IcoSema Clinical Protocol

Purpose. Standalone clinician-facing operational protocol for IcoSema — the once-weekly fixed-ratio combination of insulin icodec (basal insulin analog) and semaglutide (GLP-1 receptor agonist) marketed in the European Union under the Novo Nordisk Kyinsu brand. The protocol is structured per the Module 5 Protocol Template (twelve sections + three appendices) and operationalizes the v1.0-final IcoSema canonical at /obsidian-peptides/Peptides/IcoSema.md (2026-05-12) plus the v1.0-final Semaglutide canonical at /obsidian-peptides/Peptides/Semaglutide.md (2026-05-08, GLP-1 RA component anchor). Every effect-size claim, dose threshold, hypoglycemia-management algorithm, and counseling beat traces to the canonical pair. This protocol does not re-derive the evidence base; it operationalizes it for clinicians preparing for IcoSema’s anticipated US availability and for clinicians operating in EU jurisdictions where Kyinsu is currently marketed.

FDA-approved-for-marketing-claims — load-bearing precision (Editorial Framework §1.2.1; Pattern AA.marketing-claims). The phrase “FDA-approved” throughout this protocol is shorthand for “FDA-approved for marketing claims for [indication X].” FDA approval is indication-specific, not drug-specific. The IcoSema combination product is not FDA-approved for marketing claims as of 2026-05-13; the individual components are independently FDA-approved-for-marketing-claims (insulin icodec as Awiqli for T2D since 2024-04-23; semaglutide as Ozempic/Wegovy/Rybelsus across multiple indications). The protocol uses “FDA-approved-for-marketing-claims for [X]; off-label for [Y]; pre-FDA-approval-for-marketing-claims” precisely.

Regulatory state — load-bearing for §1 framing. IcoSema (Kyinsu) is EMA-approved-for-marketing-claims (EC marketing authorization decision 24 November 2025) for type 2 diabetes in adults whose condition remains inadequately controlled on either basal insulin OR a GLP-1 receptor agonist, used alongside diet, exercise, and oral antidiabetic medications. The fixed-ratio combination product is not FDA-approved as of 2026-05-13; FDA submission status is not publicly disclosed. This protocol prepares US clinicians for IcoSema’s anticipated availability based on COMBINE Phase 3 readout data; current US prescribing requires off-label co-administration of separate semaglutide + insulin icodec (where both are individually available) OR enrollment in active COMBINE program follow-on trials. EU clinicians prescribe Kyinsu within the labeled EMA indication.

Combo-product structural framing — load-bearing for §2, §6, §7, §9, §10. IcoSema is a fixed-ratio co-formulation, not a separate-injection regimen. The two components share a single subcutaneous injection at a manufacturer-specified ratio; clinicians cannot titrate insulin icodec and semaglutide independently. This is structurally different from co-administration of Awiqli (weekly insulin icodec) + Ozempic (weekly semaglutide) as two separate weekly injections, and from CagriSema (cagrilintide + semaglutide) which pairs an amylin analog with a GLP-1 RA. The insulin component brings IcoSema into a different clinical positioning category than pure incretin-class compounds — basal insulin replacement layered onto GLP-1 RA-mediated postprandial control, with hypoglycemia risk that pure GLP-1 RAs do not carry.

Research-education frame. Synergy provides the research; professionals decide. This protocol presents what the published COMBINE Phase 3 program shows and how the trial-program evidence translates into operational clinical decisions. Clinical positioning for individual patients is the responsibility of the clinician using their professional judgment in the context of the patient and their specific clinical situation. The Synergy disclaimer at the canonical (§Executive summary, IcoSema canonical) is incorporated by reference.

Pattern Z calibration baseline. §10 of this protocol carries verbatim-anchor compliance to the five anchors established in /Internal/Module-5-Pipeline/pattern-z-calibration-anchor.md, with IcoSema-specific calibration for the combo-product context: Anchor 5 (off-label / extrapolation transparency) is dominant pre-FDA-approval for US clinicians; Anchor 2 (compounded vs FDA-approved) is N/A pre-approval in the US (compounded combo-product market does not exist for an unapproved combo); Anchor 4 (comparator framing) is load-bearing for the IcoSema vs separate-injection regimen vs daily-combo (Xultophy / Soliqua) vs basal-bolus comparison.

Indication scope and patient phenotypes
Selection criteria (inclusion / exclusion / contraindications)
Pre-treatment workup
Initiation protocol
Maintenance protocol
Side-effect management
Plateau and non-response algorithm
Discontinuation and tapering
Combination rules
Patient counseling beats (Pattern Z verbatim-anchor-compliant; Anchor 5 dominant)
Source citations
Clinical decision tree

Appendices

A. Verification gate — four-step cycle
B. Pattern discipline self-audit (this protocol)
C. Byte-count audit and commit log

Cross-references

Canonical (primary): /obsidian-peptides/Peptides/IcoSema.md (v1.0-final, 2026-05-12)
Component canonical (GLP-1 RA): /obsidian-peptides/Peptides/Semaglutide.md (v1.0-final, 2026-05-08)
Template: /obsidian-peptides/Methodology/Protocol Template.md (v1.0)
Pattern Z anchor: /Internal/Module-5-Pipeline/pattern-z-calibration-anchor.md
Editorial framework: /obsidian-peptides/Methodology/Synergy Editorial Framework.md (v1.2)
Voice profile: /obsidian-peptides/Methodology/Voice Profile - Dr. Jeff Gross MD.md
System observations: /obsidian-peptides/Methodology/AC2-26 - System Observations.md
Bibliography (full): /Process/IcoSema/IcoSema - Bibliography.md

1. Indication scope and patient phenotypes

1.1 Purpose

This protocol prepares clinicians for IcoSema’s anticipated US availability based on the COMBINE Phase 3 readout data, and operationalizes EU prescribing for clinicians currently using Kyinsu under the EMA marketing authorization. As of 2026-05-13, IcoSema is EMA-approved-for-marketing-claims (Kyinsu, EC decision 24 November 2025) for type 2 diabetes in adults whose condition remains inadequately controlled on either basal insulin OR a GLP-1 receptor agonist, used alongside diet, exercise, and oral antidiabetic medications. The combination is pre-FDA-approval-for-marketing-claims in the US; FDA submission status is not publicly disclosed. US prescribing of the fixed-ratio combination product is not commercially available; clinicians seeking the combined mechanism in the US operate through off-label co-administration of separate semaglutide + insulin icodec (where both are individually available as Ozempic and Awiqli respectively) OR enrollment in active COMBINE program follow-on trials.

Pattern R.1 enforcement at this section: §1 opens with what IcoSema does (combined-mechanism therapy for T2D inadequately controlled on basal-insulin or GLP-1 RA monotherapy) and the EU regulatory state where it is currently marketed, before §2 turns to who the combination is for and §10 turns to off-label transparency for US clinicians. Pattern R.2 enforcement: the indication scope is T2D-only at the EMA label; chronic weight management is explicitly outside the registered indication scope. Section 1 holds this scope; downstream sections (§4 dosing, §5 maintenance, §7 non-response, §10 counseling) do not drift into CWM framing.

Pattern Z.research-precision applied throughout §1: “pre-FDA-approval-for-marketing-claims” and “COMBINE Phase 3 program with three published primary RCTs” replace bias-vocabulary like “investigational” or “experimental” or “early-stage.” The compound has a verified Phase 3 evidence base; it sits at the EMA-approved / pre-FDA-approval interface, not at an “experimental” status.

1.2 Indication categories — applied to IcoSema

IcoSema’s EMA-approved indication scope is structurally narrower than semaglutide’s six-indication scope (semaglutide covers T2D glycemic control, chronic weight management, CV risk reduction, MASH, CKD-in-T2D, plus the investigational HFpEF extension). IcoSema covers only T2D glycemic control, with the trial-program-supported sub-stratifications below. The mechanism-component complementarity rationale (basal insulin replacement + GLP-1 RA-mediated postprandial control) is positioned specifically for T2D where both axes of glucose dysregulation are clinically active.

T2D glycemic control — basal-insulin-experienced sub-population (COMBINE 1 anchor). Adults with T2D inadequately controlled on daily basal insulin therapy. Trial: COMBINE 1 (NCT05352815, Mathieu et al Lancet Diabetes Endocrinol 2025 July, PMID 40482671; n=1,291; 52-week Phase 3 RCT IcoSema vs once-weekly insulin icodec alone). Primary endpoint HbA1c reduction: IcoSema −1.55 percentage points vs icodec −0.89; between-group treatment difference −0.66% (95% CI −0.76 to −0.57; p<0.0001 for superiority). Hypoglycemia (clinically significant + severe): IcoSema 0.14 vs icodec 0.63 episodes per person-year; rate ratio 0.22 (95% CI 0.14–0.36; p<0.0001) — IcoSema substantially lower vs basal insulin alone. The mechanism explanation: the GLP-1R glucose-dependent insulin-secretion pathway partially offsets the insulin-component hypoglycemia risk, plus IcoSema patients required less total insulin given the GLP-1R contribution to glycemic control. Phenotype primary target: T2D adults on daily basal insulin (glargine, degludec, detemir, NPH, or daily basal-component of a daily combo) with HbA1c above individualized target despite background OAD therapy.
T2D glycemic control — GLP-1 RA-experienced sub-population (COMBINE 2 anchor). Adults with T2D inadequately controlled on once-weekly semaglutide 1.0 mg or comparable GLP-1 RA monotherapy. Trial: COMBINE 2 (NCT05259033, Lingvay et al Diabetologia 2025 April, PMID 39820580; n=683; 52-week Phase 3 RCT IcoSema vs once-weekly semaglutide 1.0 mg). Primary endpoint HbA1c reduction: IcoSema −1.35 percentage points vs semaglutide −0.90; between-group treatment difference −0.45% (estimated mean difference −4.85 mmol/mol, 95% CI −6.13 to −3.57; p<0.0001). Body weight change (secondary): IcoSema +0.84 kg vs semaglutide −3.70 kg; treatment difference +4.54 kg favoring semaglutide on the weight axis (95% CI 3.84 to 5.23; p<0.0001). Hypoglycemia (clinically significant + severe): IcoSema 0.042 vs semaglutide 0.036 episodes per person-year; rate ratio 1.20 (95% CI 0.53–2.69; p=0.66) — non-significant numerical increase with IcoSema vs semaglutide-monotherapy. GI adverse events: IcoSema 31.4% vs semaglutide 34.4% — comparable. Phenotype primary target: T2D adults on weekly GLP-1 RA monotherapy with HbA1c above individualized target, where adding basal insulin is the next clinically appropriate step.
T2D glycemic control — basal-bolus-eligible sub-population (COMBINE 3 anchor). Adults with T2D for whom basal-bolus insulin therapy is the clinical alternative (typically advanced T2D with progressed beta-cell failure where intensification beyond basal insulin alone is indicated). Trial: COMBINE 3 (NCT05013229, Billings et al Lancet Diabetes Endocrinol 2025 July, PMID 40482670; n=679; 52-week Phase 3 RCT IcoSema vs basal-bolus glargine U100 + 2–4 daily aspart). Primary endpoint HbA1c reduction: IcoSema −1.47 percentage points vs basal-bolus −1.40; treatment difference −0.06% (95% CI −0.22 to 0.09; p<0.0001 for non-inferiority). Body weight: treatment difference −6.72 kg favoring IcoSema (95% CI −7.58 to −5.86; p<0.0001). Hypoglycemia: rate ratio 0.12 (95% CI 0.08–0.17; p<0.0001) — IcoSema substantially lower vs basal-bolus. Injection burden: 52/year vs 730+/year for basal-bolus. Phenotype primary target: T2D adults on basal insulin who would otherwise transition to basal-bolus due to inadequate glycemic control, where the IcoSema combination provides equivalent glycemic control with substantial body-weight, hypoglycemia, and injection-burden benefits.
T2D glycemic control — insulin-naïve sub-population (COMBINE 4 — primary publication pending). Adults with T2D inadequately controlled on OAD therapy alone, where injectable therapy initiation is the next clinical step. Trial: COMBINE 4 (NCT06269107, OverallStatus = Completed per ClinicalTrials.gov v2 API; primary publication status pending in PubMed corpus as of 2026-05-13). The COMBINE 4 readout will populate this sub-population’s effect-size anchor; Section 11 Bibliography will be updated upon publication. Pattern AA.marketing-claims and Pattern V precision: pre-publication clinical use should not generalize COMBINE 1/2/3 effect-sizes verbatim to the insulin-naïve population, where insulin-tolerance baseline is absent and hypoglycemia surveillance discipline is more load-bearing.

Investigational extensions explicitly OUT OF SCOPE for the EMA Kyinsu indication:

Chronic weight management without T2D. IcoSema’s body-weight effect direction is modest weight gain vs semaglutide-monotherapy (COMBINE 2 +0.84 kg net) given the insulin-component anabolic offset; off-label use for CWM is not evidence-supported and is mechanism-counter-indicated. Patients seeking maximal weight loss alongside glycemic control are referred to semaglutide-monotherapy (Wegovy 2.4 mg) or higher-dose semaglutide (Wegovy HD 7.2 mg per STEP UP) or to tirzepatide (Zepbound 15 mg per SURMOUNT-1 / SURMOUNT-5). Pattern V direction-of-effect: combo-product weight effect is positive (gain) vs negative (loss) for the semaglutide-alone comparison; the canonical framing in §10 counseling beats does not steer patients toward IcoSema for weight-management objectives.
Cardiovascular risk reduction. No dedicated IcoSema CVOT; CV evidence inherits from semaglutide-component SELECT (Lincoff 2023 PMID 37952131) for the GLP-1 RA component only and from ONWARDS Phase 3 program safety surveillance for the insulin icodec component only. Combo-product CV outcomes are research-state-incomplete. Pattern AA precision: do not state “FDA-approved for CV risk reduction” or “EMA-approved for CV risk reduction” — neither is true for the IcoSema combination product as of 2026-05-13.
Kidney outcomes in T2D + CKD. No dedicated IcoSema kidney outcomes trial; FLOW (Perkovic 2024 PMID 38785209) provides semaglutide-component kidney outcomes inheritance only.
MASH F2–F3. No IcoSema MASH program; ESSENCE (Sanyal/Newsome 2025 PMID 40305708) provides semaglutide-component MASH outcomes inheritance only. The fixed-ratio combination is not labeled for MASH.

1.3 Phenotype-targeting taxonomy applied to IcoSema

Within the four T2D sub-population strata above, phenotype targeting refines selection. The Module 5 phenotype taxonomy (§1.3 of the Protocol Template) applied to IcoSema:

Metabolic phenotype. IcoSema’s combination addresses both fasting hyperglycemia (insulin component direct hepatic gluconeogenesis suppression) and postprandial hyperglycemia (GLP-1 RA component slowed gastric emptying + glucose-dependent insulin secretion + glucagon suppression). Phenotype primary target: T2D with both fasting and postprandial hyperglycemia patterns. Insulin-resistant obesity with progressed beta-cell failure is the canonical target; insulin-sensitive thinner T2D phenotype (potentially LADA — see §2.3) is a clinician-judgment phenotype where C-peptide assessment informs whether the basal-insulin component is mechanistically appropriate.
Adiposity distribution. Less load-bearing for IcoSema than for pure GLP-1 RA chronic-weight-management indications. Visceral-vs-subcutaneous distribution does not modify IcoSema selection within the T2D indication; both phenotypes benefit from glycemic control. Patients with substantial visceral adiposity may benefit from the semaglutide-component’s weight-neutral-to-modest-loss effect in combination, though IcoSema is not the maximal-weight-loss option (semaglutide-monotherapy or tirzepatide is — §1.2 above).
Appetite phenotype. GLP-1 RA-mediated appetite suppression is present in IcoSema (semaglutide-component effect); insulin-component partially offsets through anabolic action. Patients with hyperphagia-dominant T2D may experience partial appetite reduction on IcoSema, though the net effect is smaller than semaglutide-monotherapy.
Energy-expenditure phenotype. Less differentiated for IcoSema than for monotherapy. No specific energy-expenditure phenotype is a primary IcoSema target.
Comorbidity load. Polycondition T2D (T2D + obesity + ASCVD or T2D + CKD) is the highest-yield IcoSema phenotype, particularly where basal insulin is already part of the regimen or anticipated as the next intensification step. The polycondition phenotype with concurrent SGLT2 inhibitor + GLP-1 RA + basal insulin regimen consolidation into IcoSema + SGLT2 inhibitor is the canonical combination context (§9.2). Monocondition T2D-glycemic-control-only patients without comorbidity load are less differentiated for IcoSema vs simpler regimens (basal insulin alone titrated up; GLP-1 RA monotherapy titrated up).
Pharmacologic history.
- Basal-insulin-experienced (COMBINE 1 phenotype): the highest-yield IcoSema phenotype; established insulin-tolerance baseline; switch protocol per the COMBINE 1 trial titration scheme.
- GLP-1 RA-experienced without insulin (COMBINE 2 phenotype): adding basal insulin component to established GLP-1 RA monotherapy where HbA1c is above target; hypoglycemia surveillance discipline at switch is load-bearing.
- Basal-bolus-experienced (COMBINE 3 phenotype): regimen-simplification with substantial body-weight, hypoglycemia, and injection-burden benefits (§7.6).
- Insulin-naïve (COMBINE 4 phenotype — pending publication): structured initiation with conservative starting dose and hypoglycemia education.
- Daily-combo-experienced (Xultophy or Soliqua transition; off-trial scope): transition to once-weekly IcoSema offers daily-to-weekly injection-burden reduction; not directly characterized in COMBINE program but operationally analogous to COMBINE 1 + COMBINE 2 mixed phenotype.
Life-stage modifier.
- Reproductive-age female. IcoSema is not used during pregnancy (semaglutide-component contraindication; insulin-component active-pharmacology continuation considerations); pre-conception discontinuation per §8.4 below. Both components have ~7-8 day half-lives; ~5 half-lives = ~35-40 days pharmacokinetic clearance.
- Perimenopausal / menopausal. No specific IcoSema modifications.
- Older adult (≥65). Hypoglycemia surveillance is particularly load-bearing given autonomic dysfunction and reduced cognitive/functional reserve for hypoglycemia recognition. Conservative titration and lower HbA1c-target individualization apply (§5.4).
- Adolescent. IcoSema is not registered for pediatric use per EMA Kyinsu indication; pediatric extension is not part of the COMBINE Phase 3 program scope. Off-label pediatric use is not evidence-supported.

1.4 Cross-reference to case construction

Every IcoSema worked clinical case (presented in §4–§9 examples and in standalone case files at /Conditions/T2D Polycondition/) is constructed against the phenotype taxonomy in §1.3, using the case-construction conventions documented in M5.10 Case Study — Patient Questions. Cases anchor to one of the four COMBINE sub-population strata (basal-insulin-experienced, GLP-1 RA-experienced, basal-bolus-transition, insulin-naïve) plus comorbidity load.

1.5 Worked clinical scope — IcoSema indication detail

Indication — T2D glycemic control with COMBINE program sub-population stratification. The EMA Kyinsu indication is operationally a single regulatory indication (“T2D inadequately controlled on basal insulin OR GLP-1 RA, used alongside diet, exercise, and OADs”) that spans the four COMBINE sub-population strata. Effect-size precision varies by sub-population and is anchored to the specific COMBINE trial:

Basal-insulin-experienced phenotype on IcoSema (COMBINE 1 anchor). HbA1c reduction approximately 1.55 percentage points (per-arm absolute change from baseline) vs approximately 0.89 percentage points with icodec-alone; between-group treatment difference −0.66% (95% CI −0.76 to −0.57; p<0.0001) at 52 weeks. Pattern V.metric-axis disclosure: both per-arm absolute changes and the between-group treatment difference are reported here; downstream sections cite the between-group treatment difference unless per-arm change is specifically load-bearing. Hypoglycemia rate ratio 0.22 (clinically significant + severe combined) favoring IcoSema. Body-weight effect direction per COMBINE 1 primary publication: less weight gain with IcoSema than with icodec-alone (the GLP-1R component partial offset).
GLP-1 RA-experienced phenotype on IcoSema (COMBINE 2 anchor). HbA1c reduction approximately 1.35 percentage points vs approximately 0.90 percentage points with semaglutide-alone; between-group treatment difference −0.45% (estimated mean difference −4.85 mmol/mol, 95% CI −6.13 to −3.57; p<0.0001) at 52 weeks. Body-weight effect direction reversed vs COMBINE 1: IcoSema +0.84 kg vs semaglutide −3.70 kg; treatment difference +4.54 kg favoring semaglutide-monotherapy on the weight axis (95% CI 3.84 to 5.23; p<0.0001). Hypoglycemia rate ratio 1.20 (non-significant; p=0.66) — insulin component adds modest absolute hypoglycemia risk over GLP-1 RA-monotherapy baseline. GI tolerability: comparable (31.4% IcoSema vs 34.4% semaglutide-alone).
Basal-bolus-eligible phenotype on IcoSema (COMBINE 3 anchor). HbA1c reduction approximately 1.47 percentage points vs approximately 1.40 percentage points with basal-bolus; treatment difference −0.06% (95% CI −0.22 to 0.09; p<0.0001 non-inferiority) at 52 weeks. Body-weight effect: treatment difference −6.72 kg favoring IcoSema (95% CI −7.58 to −5.86; p<0.0001). Hypoglycemia rate ratio 0.12 favoring IcoSema. Injection burden 52/year vs 730+/year. The regimen-simplification clinical positioning is load-bearing for this sub-population.
Insulin-naïve phenotype (COMBINE 4 — pending publication). Effect-size anchors to be populated upon primary publication; structured titration with conservative starting dose and hypoglycemia education is the operational discipline.

Pattern AB.4 standing scan applied to §1.5. Every NCT and PMID above was verified by content (sponsor = Novo Nordisk; intervention = IcoSema fixed-ratio combination; indication = T2D; phase = Phase 3; status per ClinicalTrials.gov v2 API as of canonical Phase 2 verification 2026-05-12). The COMBINE 1 PMID 40482671 (Mathieu et al Lancet Diabetes Endocrinol 2025 July) is the verified-correct identifier; PMID 40347948 is not COMBINE 1 (that PMID indexes an unrelated developmental biology paper) and must not appear in the IcoSema Bibliography. This precision was resolved in the canonical Phase 4 AB-hygiene audit (B-CRIT-2 resolution, Bibliography AB-hygiene addendum 2026-05-12).

Pattern AA.marketing-claims precision in §1.5. The IcoSema combination product is EMA-approved-for-marketing-claims (Kyinsu, EC decision 24 November 2025) for the indication described; the same product is pre-FDA-approval-for-marketing-claims as of 2026-05-13. The semaglutide GLP-1 RA component is FDA-approved-for-marketing-claims for its label indications (Ozempic 2017 T2D; Wegovy 2021 CWM; Rybelsus 2019 oral T2D; SELECT 2024 CV expansion; ESSENCE 2025 MASH; Wegovy HD 7.2 mg 2026 high-dose CWM). The insulin icodec component is FDA-approved-for-marketing-claims as Awiqli for T2D basal insulin replacement (2024-04-23). The fixed-ratio combination product is structurally and regulatorily distinct from co-administration of the separately-approved components; the precision matters for §10 counseling beats and §11 citation framing.

2. Selection criteria (inclusion / exclusion / contraindications)

2.1 Purpose

Define who IcoSema is for, who it is not for, and who it must not be given to. §2 operationalizes the §1 indication scope into actionable clinical screening criteria. The section is structured per the template into inclusion (§2.2), relative exclusion (§2.3), and hard contraindications (§2.4) sub-blocks. Pattern R.1 enforcement: open with inclusion criteria — the COMBINE-enrolled and EMA-label-permitted T2D phenotype — before exclusions and contraindications.

IcoSema’s selection criteria layer two component-class contraindication sets (GLP-1 RA class contraindications inherited from the semaglutide canonical Section 2.4 + insulin-class contraindications inherited from insulin icodec’s Awiqli labeling) plus combination-specific considerations (hypoglycemia risk profile for concurrent insulin-secretagogue regimens; fixed-ratio titration constraints). Pattern AA enforcement: each contraindication is anchored to its source (FDA boxed warning class-wide for GLP-1 RAs via the semaglutide-component; EMA Kyinsu labeled contraindication; insulin-class precaution per Awiqli label).

2.2 Inclusion criteria — the COMBINE-enrolled and EMA-label-permitted phenotype

Inclusion criteria are stated as the population the COMBINE Phase 3 program enrolled and the EMA Kyinsu label permits, with primary-source anchoring to the relevant COMBINE trial protocols:

General inclusion (all four sub-populations):

Age ≥18 years per EMA Kyinsu label. No upper age bound is label-specified; COMBINE trial enrollment typically extended to age 75 or 80 depending on the specific trial. Patients ≥65 receive Pattern Z anchor 1 counseling on the load-bearing hypoglycemia-surveillance discipline (§5.4 + §6.2.5).
Adult T2D diagnosis (ADA-criteria-equivalent; HbA1c ≥6.5% on confirmation, or fasting plasma glucose ≥126 mg/dL on confirmation, or 2-hour OGTT ≥200 mg/dL, or random plasma glucose ≥200 mg/dL with classic hyperglycemic symptoms).
Inadequate glycemic control on baseline therapy per the sub-population definition. EMA Kyinsu specifies “either basal insulin OR a GLP-1 receptor agonist, used alongside diet, exercise, and oral antidiabetic medications” as the inadequate-control baseline. COMBINE trial enrollment HbA1c ranges were 7.0–10.5% (with some inter-trial variation).
Background OAD therapy at maximally tolerated dose where clinically indicated (metformin in the absence of contraindication; SGLT2 inhibitor where CV or renal benefit is sought; sulfonylurea typically discontinued or substantially dose-reduced at IcoSema initiation per §6.11.1 — see §2.3 below).
Adequate renal function within COMBINE program enrollment range. Severe renal impairment (eGFR <30 or specific trial enrollment cutoffs) is a relative exclusion (§2.3); patients with eGFR 30–60 are clinically appropriate for IcoSema with standard monitoring.
Adequate hepatic function within COMBINE program enrollment range. Severe hepatic impairment (Child-Pugh C) is a relative exclusion (§2.3).

Sub-population-specific inclusion:

Basal-insulin-experienced sub-population (COMBINE 1): adult T2D inadequately controlled on daily basal insulin therapy (glargine U100 or U300, degludec, detemir, NPH) at appropriately titrated dose. Effect-size anchor: HbA1c between-group treatment difference −0.66% favoring IcoSema (PMID 40482671). Phenotype is the highest-yield IcoSema target.
GLP-1 RA-experienced sub-population (COMBINE 2): adult T2D inadequately controlled on once-weekly semaglutide 1.0 mg or comparable GLP-1 RA at appropriately titrated dose. Effect-size anchor: HbA1c between-group treatment difference −0.45% favoring IcoSema (PMID 39820580); body-weight tradeoff +4.54 kg vs semaglutide-monotherapy (counseling beat per §10.5 below).
Basal-bolus-eligible sub-population (COMBINE 3): adult T2D on basal insulin who would otherwise transition to basal-bolus regimen (basal + 2–4 daily prandial insulin injections) due to inadequate glycemic control. Effect-size anchor: HbA1c non-inferior to basal-bolus (treatment difference −0.06%, p<0.0001 non-inferiority); substantial body-weight benefit (−6.72 kg vs basal-bolus); substantial hypoglycemia reduction (rate ratio 0.12); injection-burden reduction (52 vs 730+/year) (PMID 40482670).
Insulin-naïve sub-population (COMBINE 4 — pending publication): adult T2D inadequately controlled on OAD therapy; injectable therapy initiation is the next clinical step. Pre-publication clinical use should approach this sub-population with structured titration discipline; effect-size anchor to be populated upon publication.

2.3 Relative exclusion criteria — clinician-judgment phenotype

Relative exclusion criteria identify phenotypes where IcoSema is not contraindicated but where benefit is uncertain, risk is elevated, or trial data are sparse. The protocol structure for each: state the criterion; state the underlying concern; state the magnitude of trial evidence; state the recommended clinician-judgment posture.

Severe gastroparesis or gastroparesis-predisposing comorbidity. Inherits from semaglutide-component class concern. GLP-1 RA mechanism includes delayed gastric emptying; in established gastroparesis, anatomical and symptomatic worsening risk is elevated. COMBINE program excluded severe gastroparesis. Recommended posture: do not initiate; ARFID and BED with active purging routed to behavioral-health co-management before any GLP-1 RA-containing therapy initiation.
Active or recent (within 12 months) acute pancreatitis. Inherits from semaglutide-component class concern. Distinct from severe prior pancreatitis history (a §2.4 contraindication for severe history). Recommended posture: deferred until at least 12 months stable post-event with clinician judgment; pre-treatment lipase and amylase baseline per §3.7.
Severe gastrointestinal disease (active IBD flare, severe GERD with esophagitis). Relative — GLP-1 RA GI AE profile may exacerbate. Inherits from semaglutide-component class concern.
Diabetic retinopathy with rapid HbA1c improvement risk. Inherits from semaglutide-component class concern. SUSTAIN-6 documented retinopathy-complication signal in patients with rapid glycemic improvement; the protocol acknowledges and routes to ophthalmology pre-screening for advanced background or proliferative diabetic retinopathy. The combined-mechanism HbA1c effect at IcoSema target dose may produce rapid glycemic improvement in patients with baseline HbA1c ≥9.0; counseling and pre-screening apply.
Severe renal impairment (eGFR <15 — outside COMBINE enrollment range). Insulin-component clearance is partly renal; severe renal impairment may require dose adjustment that the fixed-ratio combination cannot accommodate independently. Recommended posture: clinician judgment with nephrology co-management; separate-injection regimen (Awiqli + Ozempic at independently-titrated doses) may be operationally preferable where independent insulin dose adjustment is needed.
Severe hepatic impairment (Child-Pugh C). Inherits from semaglutide-component class concern. COMBINE program excluded severe hepatic impairment.
Active eating disorder (anorexia nervosa, bulimia nervosa, BED with active purging). Appetite-suppression mechanism is contraindicated in disordered eating; inherits from semaglutide-component class concern.
Hypoglycemia-unawareness or recurrent severe hypoglycemia history. Insulin-component contribution to hypoglycemia risk is partially offset by the GLP-1R component, but hypoglycemia-unawareness phenotype is a relative exclusion for any insulin-containing regimen. Recommended posture: clinician judgment; CGM-mediated hypoglycemia surveillance is load-bearing; counseling and education on hypoglycemia recognition and treatment are required before initiation; lower HbA1c-target individualization applies.
Concurrent sulfonylurea or meglitinide at full dose. Relative — additive hypoglycemia risk. Standard practice at IcoSema initiation: discontinue or substantially dose-reduce the sulfonylurea (typically 50% reduction or discontinuation) per §6.11.1. Not a contraindication; a dose-adjustment-of-concurrent-agent trigger.
Concurrent insulin (basal or bolus) outside the IcoSema regimen. Relative — additive insulin dosing. The IcoSema insulin-icodec component already provides basal insulin replacement; concurrent administration of separate basal insulin would produce additive insulin dosing requiring careful titration protocol coordination, and is not the IcoSema-as-monotherapy phenotype. Recommended posture: at IcoSema initiation, discontinue separate basal insulin; bolus insulin (insulin aspart, lispro, glulisine) may be retained at clinician judgment in patients transitioning from basal-bolus where persistent postprandial hyperglycemia warrants meal-time bolus (off-label IcoSema + bolus pattern; §6.11.4).
LADA (latent autoimmune diabetes of adults). GLP-1 RAs are not first-line in confirmed autoimmune diabetes; misclassification of LADA as T2D is a Pattern V direction-of-effect risk. Recommended posture: C-peptide and GAD-65 / IA-2 antibody assessment in atypical T2D presentations (adult-onset diabetes with normal BMI, rapid progression to insulin requirement, atypical clinical course) before IcoSema initiation; positive autoimmunity reroutes to standard autoimmune-diabetes management.
Active malignancy on therapy (other than MTC/MEN-2 contraindication — see §2.4). COMBINE program typically excluded active cancer; clinician judgment with oncology co-management.

2.4 Hard contraindications — boxed warnings and labeled contraindications

Hard contraindications are absolute — IcoSema must not be initiated, and if discovered during therapy, the molecule is discontinued. Each contraindication is documented with its source classification.

Personal or family history of medullary thyroid carcinoma (MTC). Inherits from semaglutide-component FDA boxed warning, class-wide for GLP-1 RAs based on rodent C-cell tumorigenicity signal. The EMA Kyinsu label carries the equivalent contraindication. Absolute.
Multiple Endocrine Neoplasia syndrome type 2 (MEN-2). Same source — boxed-warning / labeled contraindication. Absolute.
Severe prior pancreatitis history (severe acute or chronic). Labeled contraindication or strong-precaution depending on jurisdiction; the EMA Kyinsu label carries the precaution; clinician judgment for severe prior history defaults to not-initiate.
Known serious hypersensitivity to insulin icodec, semaglutide, or excipient. Labeled contraindication.
Pregnancy. The semaglutide-component is labeled contraindicated in pregnancy; the insulin icodec component continues active pharmacology during pregnancy and is not first-line basal insulin for pregnancy (NPH or pregnancy-appropriate analog basal insulin is first-line). The IcoSema combination is labeled contraindicated in pregnancy per the EMA Kyinsu label; pregnancy is a §8 discontinuation trigger, not a §2 inclusion-screen-only criterion — a patient on IcoSema who becomes pregnant transitions out of protocol immediately.
Type 1 diabetes mellitus. IcoSema is approved for T2D specifically; T1D requires pregnancy-and-pediatric-appropriate insulin regimens (rapid-acting prandial + long-acting basal or CSII pump) and is not within the EMA Kyinsu indication. Off-label T1D use is not evidence-supported.
Diabetic ketoacidosis (DKA) — active. Insulin-component basal pharmacology is not appropriate for acute DKA management; rapid-acting IV insulin per DKA standard-of-care is required. IcoSema is not initiated during active DKA; resumption or initiation occurs after DKA resolution.

2.5 Worked clinical screening — IcoSema selection criteria detail

Inclusion screen (per sub-population).

Basal-insulin-experienced (COMBINE 1 phenotype). Adult T2D, on daily basal insulin (glargine, degludec, detemir, NPH) at appropriately titrated dose for ≥3 months, HbA1c above individualized target (typically ≥7.5% with target <7.0%, or higher individualized target where appropriate). Background OAD therapy continued at maximally tolerated dose where clinically indicated.
GLP-1 RA-experienced (COMBINE 2 phenotype). Adult T2D, on weekly GLP-1 RA (semaglutide 1.0 mg, dulaglutide 1.5 mg, semaglutide 2.0 mg) at appropriately titrated dose for ≥3 months, HbA1c above individualized target. Counseling beat at screening: body-weight tradeoff (+4.54 kg vs continued semaglutide-monotherapy at 52 weeks per COMBINE 2) is presented factually per §10 anchor 4 multi-dimensional comparator framing.
Basal-bolus-eligible (COMBINE 3 phenotype). Adult T2D, on basal insulin with HbA1c above target despite appropriate titration, where intensification to basal-bolus is the otherwise-indicated next step. The IcoSema-vs-basal-bolus counseling beat presents the regimen-simplification trade-off (52 vs 730+ injections/year; less weight gain; less hypoglycemia; equivalent HbA1c) — Pattern Z anchor 4.
Insulin-naïve (COMBINE 4 phenotype). Adult T2D on maximally tolerated OAD therapy with HbA1c above target where injectable therapy initiation is the next clinical step. Pre-publication clinical use: structured titration with conservative starting dose; effect-size expectations should not generalize verbatim from COMBINE 1/2/3 anchors.

Relative exclusions encountered at screening.

Severe gastroparesis: defer; behavioral-health and gastroenterology co-management as indicated. Severe persistent GERD with esophagitis: relative.
Recent acute pancreatitis (within 12 months): defer until at least 12 months stable post-event.
Diabetic retinopathy: ophthalmology pre-screening for proliferative DR or advanced background DR (especially if HbA1c ≥9.0).
Severe renal impairment (eGFR <15): clinician judgment with nephrology co-management; separate-injection regimen may be operationally preferable where independent insulin titration is needed.
Severe hepatic impairment (Child-Pugh C): clinician judgment.
Hypoglycemia-unawareness: CGM-mediated surveillance; counseling and education; lower HbA1c-target individualization.
Concurrent sulfonylurea at full dose: reduce or discontinue at IcoSema initiation per §6.11.1.
Concurrent separate basal insulin: discontinue separate basal insulin at IcoSema initiation.
LADA suspected: C-peptide and GAD-65 / IA-2 antibody assessment.
Active malignancy: clinician judgment with oncology co-management; MTC/MEN-2 history is hard contraindication.

Hard contraindications encountered at screening.

Personal or family history of MTC: do not initiate. FDA boxed warning (class-wide GLP-1 RAs via semaglutide-component); EMA Kyinsu equivalent contraindication.
MEN-2: do not initiate. Same source.
Severe prior pancreatitis: do not initiate (precaution / labeled cautionary use).
Known serious hypersensitivity to insulin icodec, semaglutide, or excipient: do not initiate.
Pregnancy: do not initiate; if discovered during therapy, immediate discontinuation and transition to pregnancy-appropriate insulin regimen.
T1D: do not initiate; not in indication scope.
Active DKA: do not initiate; resume or initiate after DKA resolution.

Pattern AA.marketing-claims precision in this section. “FDA boxed warning for MTC and MEN-2 (class-wide for GLP-1 RAs, inherited by IcoSema via the semaglutide-component)” is the precise regulatory framing — the boxed warning is not specific to IcoSema as a combination product but inherits from the semaglutide-component label. “Labeled contraindication” for hypersensitivity and pregnancy per the EMA Kyinsu label. “Precaution / cautionary use” for prior pancreatitis. The IcoSema combination product itself is pre-FDA-approval-for-marketing-claims in the US as of 2026-05-13; US clinicians applying these contraindications operate via the component-level FDA labels (Ozempic / Awiqli) when prescribing the separate-injection regimen.

Pattern V cross-check at this section. The hypoglycemia rate ratios across COMBINE trials (0.22 favoring IcoSema vs icodec-alone in COMBINE 1; 1.20 non-significant numerical increase vs semaglutide-monotherapy in COMBINE 2; 0.12 favoring IcoSema vs basal-bolus in COMBINE 3) are direction-of-effect-distinct depending on comparator. Pattern V discipline: do not generalize “IcoSema lowers hypoglycemia” across all comparators — the direction of effect is comparator-specific. The §6.7 hypoglycemia AE class characterization is precise: the insulin component brings hypoglycemia risk that pure GLP-1 RAs do not carry; the combined-mechanism profile is favorable vs insulin-alone comparators and modestly elevated vs GLP-1 RA-alone comparator.

3. Pre-treatment workup

3.1 Purpose

Define the pre-treatment laboratory, imaging, and clinical-assessment workup that must be completed and reviewed before IcoSema initiation. §3 is the operational handoff between §2 (selection criteria) and §4 (initiation): a patient who passes §2 screening enters §3 workup; only on workup completion does §4 dose initiation begin.

The IcoSema workup leverages the Module 5 seven-panel framework with combo-product-specific emphasis: the diabetes-specific panel (§3.3) is uniformly load-bearing across all four COMBINE sub-populations (T2D is the indication scope); the organ-baseline panel (§3.7) includes both the GLP-1 RA class items (thyroid, pancreas, ophthalmology) inherited from the semaglutide-component and the insulin-class items (no specific class-baseline beyond standard endocrine workup); the body-composition baseline (§3.8) is less load-bearing for IcoSema than for CWM indications, since IcoSema’s body-weight effect is modest. Pattern W cross-section consistency: every lab listed in §3.2–§3.8 reconciles with §5 maintenance monitoring intervals and §6 AE management triggers.

3.2 Standard metabolic panel

Applies to every IcoSema initiation. Establishes baseline metabolic state, identifies undiagnosed comorbidity, and provides reference for monitoring.

Comprehensive metabolic panel (CMP). Sodium, potassium, chloride, CO₂, BUN, creatinine, glucose, calcium, total protein, albumin, ALT, AST, alkaline phosphatase, total bilirubin. Establishes hepatic and renal baseline for class-wide precautions and for any future indication-expansion eligibility (kidney outcomes via FLOW-pattern inheritance, hepatic outcomes via MASH-pattern inheritance — both currently outside IcoSema’s EMA-approved scope but relevant for clinician awareness).
Fasting lipid panel. Total cholesterol, LDL-C, HDL-C, triglycerides, non-HDL-C, ApoB if available. Baseline for CV-risk co-management given T2D-typical ASCVD risk profile.
HbA1c. The primary glycemic-control benchmark for IcoSema initiation; informs starting dose and titration target. COMBINE program inclusion was typically HbA1c 7.0–10.5%.
Fasting plasma glucose. Informs starting dose for the insulin-component initiation. Baseline FPG ≥250 mg/dL or severely uncontrolled hyperglycemia warrants clinician judgment on starting dose conservativeness.
Body weight, height, BMI, waist circumference. Anthropometric baseline. Less load-bearing for IcoSema than for CWM-indication molecules but standard practice.
Blood pressure (seated, two readings, standardized). Baseline for CV-risk co-management.

3.3 Diabetes-specific panel (T2D — uniformly applies to IcoSema)

Applies to every IcoSema initiation given the T2D-only indication scope.

HbA1c (above; standard panel). Confirms T2D diagnosis and severity; informs individualized target setting per ADA Standards of Care (typically <7.0% for most non-frail adults; <6.5% if achievable without significant hypoglycemia; <8.0% for some older or frail-comorbidity populations).
Fasting C-peptide. Establishes endogenous insulin reserve. Particularly load-bearing for IcoSema given the insulin-component contribution: very-low C-peptide reflects advanced beta-cell failure and predicts higher absolute basal-insulin-dose requirements; LADA suspected (C-peptide low in atypical T2D presentation) reroutes per §2.3 to autoimmunity assessment.
GAD-65 antibodies and IA-2 antibodies (if LADA suspected — adult-onset diabetes with normal BMI, rapid progression to insulin requirement, atypical clinical course). Confirmed LADA reroutes to autoimmune-diabetes management; IcoSema is not first-line in confirmed autoimmune diabetes.
Diabetes complication screen (if not within the last 12 months): dilated retinal exam (also a class-wide pre-treatment requirement — §3.7), urine albumin-to-creatinine ratio (UACR), monofilament / vibratory testing for diabetic neuropathy.
CGM data review if available. Particularly load-bearing for IcoSema given the insulin-component contribution to hypoglycemia risk. Pre-treatment CGM data establishes time-in-range baseline, hypoglycemia frequency baseline, and pattern-specific (nocturnal, post-prandial, exercise-related) hypoglycemia tendencies. Patients with documented hypoglycemia-unawareness or recurrent severe hypoglycemia on baseline therapy enter §2.3 relative-exclusion clinician-judgment posture.

3.4 MASH-specific panel — not routinely indicated for IcoSema

Applies when MASH risk profile is independently present (T2D + obesity + elevated AST/ALT on standard panel) — IcoSema is not labeled for MASH; the semaglutide-component MASH evidence (ESSENCE, Sanyal/Newsome 2025 PMID 40305708) does not extend to the IcoSema combination product. Clinicians with patients carrying MASH risk profiles may choose to assess FIB-4 and refer to hepatology per standard MASH-risk practice, but this is not an IcoSema-specific workup.

3.5 Kidney-specific panel — selective for IcoSema

Applies when baseline eGFR is 30–60, where insulin-component clearance considerations and the broader T2D + CKD co-management context apply.

Serum creatinine, eGFR. Standard renal-function baseline (also in §3.2 CMP).
Cystatin C-based eGFR if creatinine-eGFR is discordant with clinical picture (e.g., low muscle mass elderly patient with apparently normal creatinine but real GFR reduction). Insulin-component clearance assessment in older or low-muscle-mass patients informs titration conservativeness.
Urine albumin-to-creatinine ratio (UACR). Quantifies albuminuria. Patients with T2D + CKD eGFR 25–75 and UACR 100–5000 are FLOW-anchored semaglutide-component candidates; the combination-product kidney-outcomes evidence is research-state-incomplete but the semaglutide-component contribution to kidney benefit is mechanistically inherited.
Urinalysis with microscopy. Rules out alternative kidney disease.
Renin-angiotensin system (RAS) blockade documentation. Patients with T2D + CKD on maximally tolerated RAS blockade have the FLOW-protocol-equivalent background therapy; IcoSema initiation does not change RAS-blockade dosing.

3.6 CV-risk-specific panel — selective for IcoSema

Applies when baseline ASCVD risk profile is high. IcoSema is not labeled for CV risk reduction; the semaglutide-component CV evidence (SELECT, Lincoff 2023 PMID 37952131) does not extend to the IcoSema combination product. CV-risk workup is per standard T2D co-management practice, not IcoSema-specific.

ECG (12-lead). Baseline rhythm and conduction status for general T2D + CV-risk workflow.
High-sensitivity troponin if symptomatic baseline. Rules out unstable ASCVD.
NT-proBNP or BNP if HFpEF-suspect. Per standard T2D + CV-risk workflow.
Echocardiogram if HFpEF-suspect.

3.7 Organ-baseline panels (thyroid, pancreas, ophthalmology — class-wide for GLP-1 RAs inherited by IcoSema)

Class-wide pre-treatment workup for any GLP-1 RA-containing protocol, inherited by IcoSema via the semaglutide-component.

Thyroid baseline. TSH at minimum; neck examination for thyroid nodules. If personal or family history of MTC or MEN-2 is identified at this step, this is a §2.4 hard contraindication and the protocol does not initiate. Routine calcitonin screening is not generally recommended.
Pancreas baseline. Serum lipase, serum amylase (lipase is more pancreas-specific). Triglycerides per §3.2. Pancreatitis-history documentation per §2.4.
Ophthalmology — dilated retinal examination. Particularly for T2D patients per the §3.3 diabetes-complication-screen overlap and per the SUSTAIN-6 retinopathy-complication signal in rapid-HbA1c-improvement sub-population. Pre-treatment dilated exam for any T2D patient with HbA1c ≥9.0 or with known background DR is the protocol-recommended posture. The combined-mechanism HbA1c effect on IcoSema may produce rapid glycemic improvement; pre-screening is load-bearing for patients with advanced background or proliferative DR.
NAION pre-screen (semaglutide-component-inherited signal). Non-arteritic anterior ischemic optic neuropathy signal has been described in semaglutide pharmacovigilance (Hathaway 2024 JAMA Ophthalmol PMID 38958939; Grauslund Danish cohort PMID 39696569; Lakhani 2025 180-country observational pharmacovigilance PMID 40383360 — observational pharmacovigilance, not meta-analysis per Pattern AB.4 precision; EMA PRAC classified NAION as “very rare” side effect, June 2025). The IcoSema combination inherits the semaglutide-component signal at fixed-ratio dose. Clinician-judgment includes pre-treatment screen for known optic-disc cupping (small or “disc-at-risk”), prior NAION, or unexplained visual symptoms. Pattern AA precision: signal under evaluation, not a labeled warning at semaglutide-component level as of 2026-05-13.

3.8 Body-composition baseline — less load-bearing for IcoSema

Applies where lean-mass concern or sarcopenia risk is clinically active (older adults; low baseline BMI; athletic populations) but not uniformly required for IcoSema given the modest body-weight effect direction.

Bioelectrical impedance analysis (BIA) or DEXA if clinically indicated.
Hand-grip strength or sit-to-stand timed test for ≥65 phenotype where sarcopenia risk during T2D management is a concern.

3.9 Worked example — IcoSema pre-treatment panel per sub-population

Basal-insulin-experienced T2D adult initiating IcoSema (COMBINE 1 phenotype). Standard metabolic panel (§3.2), diabetes-specific panel (§3.3) including HbA1c, fasting C-peptide, dilated retinal exam (within 12 months — earlier if HbA1c ≥9.0), UACR documentation, organ-baseline (§3.7) including TSH and lipase. CGM data review if available. Pre-treatment CGM is high-value for the basal-insulin-experienced phenotype given established insulin tolerance baseline and pattern-characterization.

GLP-1 RA-experienced T2D adult initiating IcoSema (COMBINE 2 phenotype). Same standard + diabetes-specific + organ-baseline panels. CGM data review particularly load-bearing for the GLP-1 RA-experienced patient who is now adding insulin component — pre-treatment glycemic pattern characterization informs starting-dose conservativeness and hypoglycemia-surveillance setup. Counseling beat at workup completion: body-weight tradeoff (+4.54 kg vs continued semaglutide-monotherapy at 52 weeks per COMBINE 2) is reviewed before titration begins.

Basal-bolus-transition T2D adult initiating IcoSema (COMBINE 3 phenotype). Same panels. CGM data review is uniformly load-bearing for basal-bolus patients given their typically more complex glycemic pattern and hypoglycemia risk profile. Pre-treatment review of basal vs prandial insulin dose distribution informs the IcoSema starting-dose calculation.

Insulin-naïve T2D adult initiating IcoSema (COMBINE 4 phenotype — pending publication). Same panels with particular emphasis on C-peptide (residual beta-cell function), LADA assessment if atypical presentation, and structured hypoglycemia education before initiation given the absence of insulin-tolerance baseline.

Pattern W cross-check applied to §3.9. Every lab in the worked-example panels above is reconciled with §5 monitoring intervals and §6 AE triggers. Lipase is monitored on symptom-prompted basis (abdominal pain), not on scheduled-interval basis. UACR and eGFR are monitored annually or per CKD severity. HbA1c is monitored quarterly during titration and semi-annually at maintenance (§5.3).

4. Initiation protocol

4.1 Purpose

Define the starting dose, titration schedule, and tolerability-management cadence for IcoSema initiation. §4 is the time-sequenced action plan from Day 0 (first dose) through target-dose attainment. Unlike semaglutide-monotherapy where the dominant initiation discipline is GI tolerability priming, IcoSema initiation has two parallel disciplines: GI tolerability priming (semaglutide-component) AND hypoglycemia-surveillance establishment (insulin-component). Both are load-bearing; the hypoglycemia-surveillance discipline is the more clinically consequential and the more sub-population-specific.

§4 is the section where Pattern Z calibration is most operationally relevant for IcoSema’s combo-product framing. The initiation period is when patient adherence is most fragile, hypoglycemia recognition is most fragile (especially for insulin-naïve patients), and clinician counseling has the greatest influence on continuation. Pattern Z anchor 1 (lead with what hypoglycemia management practice IS) and anchor 4 (multi-dimensional comparator framing for the IcoSema-vs-alternatives decision) appear in §10 counseling beats specifically because §4 initiation is the conversational anchor.

4.2 Starting dose

The starting dose is the lowest dose that has tolerability-acceptable trial-program data and serves as the titration starting point. For IcoSema, the EMA Kyinsu label specifies starting doses per sub-population:

Basal-insulin-experienced sub-population (COMBINE 1 protocol): initiation dose calibrated to the patient’s prior daily basal insulin dose at switch; the COMBINE 1 trial titration scheme used a switch-from-current-basal-insulin starting calculation with subsequent up-titration to target HbA1c.
GLP-1 RA-experienced sub-population (COMBINE 2 protocol): initiation dose at conservative starting level for the insulin-component, with the semaglutide-component dose contribution at fixed-ratio level matching established GLP-1 RA exposure.
Basal-bolus-transition sub-population (COMBINE 3 protocol): initiation dose calibrated to the patient’s prior total daily basal-bolus insulin dose with appropriate conversion factor accounting for the GLP-1R component contribution to postprandial control.
Insulin-naïve sub-population (COMBINE 4 — pending publication): conservative starting dose with structured titration; specific dosing per the COMBINE 4 trial protocol (publication will populate this section).

Refer to the current EMA Kyinsu Summary of Product Characteristics (SmPC) for precise label specifications; this protocol presents the trial-protocol-anchored operational framework, not the precise label-mandated dosing units.

4.3 Titration schedule

The IcoSema titration schedule is week-by-week dose escalation from starting dose to target dose. The once-weekly dosing schedule produces 3–4 week steady-state per dose-titration step; titration cadence reflects this PK reality. The COMBINE protocols specify titration intervals of 2–4 weeks per dose step depending on sub-population.

The titration schedule is not absolute — patient-specific tolerability (GI AE; hypoglycemia frequency; HbA1c trajectory) may extend any titration interval. The protocol documents both the standard schedule (per COMBINE protocol pace) and the slow-titration option (each interval doubled) for patients with GI tolerability challenges or hypoglycemia frequency at any step.

The fixed-ratio combination means the two components titrate proportionally; clinicians cannot adjust insulin-icodec dose independently of semaglutide-component dose. Patients whose insulin-component requirements outpace semaglutide-component tolerability (rare; GI AE typically attenuates faster than insulin requirements emerge) or whose semaglutide-component tolerability is at acceptable level but insulin-component contribution is insufficient (insulin requirements continue rising beyond fixed-ratio IcoSema target dose) may need transition to separate-injection regimen (Awiqli + Ozempic independently titrated) — §7.3 develops the comparator framing.

4.4 GI tolerability management at each titration step

The semaglutide-component of IcoSema produces the characteristic GLP-1 RA GI AE profile: nausea, vomiting, diarrhea, constipation, eructation, abdominal pain. The AE profile peaks at each dose escalation and typically attenuates within 2–4 weeks at a stable dose. COMBINE 2 reported GI adverse events at IcoSema 31.4% vs semaglutide-alone 34.4% — comparable rates suggesting the combo formulation does not amplify the semaglutide-component GI profile.

GI tolerability management at each step:

Nausea — first-line non-pharmacologic. Reduce meal size, slow eating pace, avoid greasy / high-fat meals, hydrate consistently.
Nausea — first-line pharmacologic. Ondansetron 4 mg PRN; consider scheduled dosing for moderate-severity nausea.
Vomiting — assessment. Distinguish protocol-related vomiting (expected, attenuates) from acute-pancreatitis-suspect vomiting (severe abdominal pain, persistent, with lipase elevation) — §6.5 algorithm.
Diarrhea / constipation. Bowel-pattern-specific management.
Volumetric titration adjustment. Persistent moderate-to-severe GI AE at any step triggers slow-titration — hold at current dose for an additional 2–4 weeks before next escalation, or step down to prior dose if tolerability does not stabilize.

4.5 Hypoglycemia-surveillance establishment at initiation — IcoSema-specific discipline

The insulin-component contribution to hypoglycemia risk requires structured hypoglycemia-surveillance establishment at IcoSema initiation. This is the IcoSema-specific discipline that pure GLP-1 RA monotherapy initiation does not require:

Hypoglycemia-recognition education. Patient learns autonomic signs (sweating, tremor, palpitations, hunger) and neuroglycopenic signs (confusion, blurred vision, slurred speech, dizziness, behavior change). Patient identifies personal early-warning signs through experience; in insulin-naïve patients, the first hypoglycemia event in the experiential learning curve is anticipated and counseled-for at initiation.
Glucose self-monitoring setup. BGM (blood glucose monitoring) per individual patient regimen or CGM (continuous glucose monitoring) per ADA Standards of Care for insulin-treated T2D. CGM is increasingly standard and is particularly load-bearing for IcoSema given the once-weekly PK profile and the combined-mechanism hypoglycemia-risk characterization. CGM cost: Medicare DME coverage for many patients; commercial-insurance coverage variable; cash-pay ~50−100/sensor for 14-day wear time (cost data current as of 2026-04 per the IcoSema canonical §8.5).
Glucose-restoration practice. Standard 15-15 rule (15 g rapid-acting carbohydrate; recheck glucose in 15 minutes; repeat if still <70 mg/dL). Patients carry glucose tablets or rapid-acting carbohydrate sources during routine activities.
Concurrent-medication review for additive-hypoglycemia agents. Sulfonylureas (glipizide, glyburide, glimepiride) and meglitinides (repaglinide, nateglinide) are reviewed at IcoSema initiation; standard practice is to discontinue or substantially reduce the sulfonylurea dose (typically 50% reduction or discontinuation) at IcoSema initiation per §6.11.1. Separate basal insulin is discontinued (the IcoSema insulin-icodec component replaces it).
Severe-hypoglycemia plan. Patients and family/caregivers know the response to severe hypoglycemia (Level 3 per ADA categorization — requires assistance for recovery): glucagon administration if injectable or intranasal glucagon is available; emergency medical contact; what to communicate to clinician post-event.
When-to-call-the-clinician threshold. Severe hypoglycemia requiring assistance; recurrent unexplained hypoglycemia; hypoglycemia patterns that interfere with daily activities; hypoglycemia in the context of concurrent illness or medication changes.

4.6 Early monitoring cadence

The early-monitoring cadence for IcoSema is the contact frequency during the initiation period. Typical cadence:

Week 2 (post-first-dose telehealth tolerability and hypoglycemia check).
Week 4–6 (after first titration step; in-person or telehealth; weight, BP, brief AE-and-adherence interview, hypoglycemia-pattern review via patient-reported or CGM data).
Week 8–12 (mid-titration; HbA1c reassessment for trajectory check; hypoglycemia-pattern review).
Week 16–20 (target-dose attainment confirmation; HbA1c reassessment; transition to §5 maintenance cadence).

Contact modality is practice-specific. Escalation triggers (any contact identifying severe GI AE, suspected pancreatitis, suspected gallbladder event, suspected NAION, severe or recurrent hypoglycemia → in-person evaluation within 48 hours).

4.7 Worked example — IcoSema initiation (basal-insulin-experienced phenotype, COMBINE 1 anchor)

A 58-year-old male with T2D for 12 years, BMI 33, currently on insulin glargine U100 32 units daily + metformin 2000 mg daily + empagliflozin 10 mg daily, HbA1c 8.4% with target 7.0%, no MTC/MEN-2 history, no pancreatitis history, no gastroparesis, eGFR 72, UACR 50, baseline retinal exam 6 months ago showed mild non-proliferative DR.

Selection / contraindication screen. Inclusion criteria met for COMBINE 1 phenotype (basal-insulin-experienced). No hard contraindications. No relative exclusions beyond the mild DR (counseling on rapid-HbA1c-improvement context per §3.7 and §6.8). Sulfonylurea not in current regimen — no dose adjustment needed. Separate basal insulin (glargine) will be discontinued at IcoSema initiation.

Pre-treatment workup. Standard metabolic panel + diabetes-specific panel + organ-baseline (§3.7). CGM started 2 weeks before IcoSema initiation to establish pre-treatment glycemic pattern baseline and hypoglycemia frequency on current regimen.

Initiation. Glargine discontinued at IcoSema Week 0. IcoSema starting dose per COMBINE 1 trial titration scheme (calibrated to prior basal-insulin dose with appropriate conversion factor; refer to current EMA Kyinsu SmPC for precise label specification). Counseling at initiation visit per §10.2 (initiation conversation): what IcoSema is, expected HbA1c trajectory per COMBINE 1 anchor (HbA1c reduction approximately 1.55 percentage points absolute / between-group treatment difference −0.66% vs continued glargine-alone at 52 weeks), expected GI AE profile per §6.4, hypoglycemia-surveillance establishment per §4.5, escalation pathway.

Early monitoring. Week 2 telehealth — patient reports mild nausea Day 2–4 post-injection, attenuated by Day 5; no hypoglycemia events; CGM data review shows reasonable time-in-range. Week 6 — first titration step per protocol; nausea recurrence on dose increase anticipated; ondansetron 4 mg PRN counseled. Week 12 — HbA1c interim 7.6%; on-trajectory; titration continues per protocol. Week 20 — target-dose attainment; HbA1c 7.1%; CGM shows time-in-range 72%, Level 2 hypoglycemia frequency 0.2 events per 14-day period; transition to §5 maintenance cadence.

Pattern V applied to §4.7. Effect-size anchor: COMBINE 1 between-group treatment difference −0.66% HbA1c favoring IcoSema vs icodec-alone, with the patient’s actual trajectory (8.4 → 7.1% = 1.3 percentage-point absolute reduction over 20 weeks) on-trajectory for the COMBINE 1 52-week absolute change of 1.55 percentage points in the IcoSema arm. Per-arm vs between-group disclosure (Pattern V.metric-axis): the patient’s 1.3-point reduction at Week 20 is a per-arm trajectory; the trial’s −0.66% between-group treatment difference is the population-level superiority claim vs continued icodec-alone — both are valid; the counseling beat (§10) presents them with explicit metric-axis labels.

Pattern AA applied to §4.7. IcoSema is EMA-approved-for-marketing-claims (Kyinsu) in the EU; the worked example assumes EU clinical setting. In the US, the equivalent worked example would proceed via separate-injection regimen of Awiqli (insulin icodec, FDA-approved-for-marketing-claims 2024-04-23) + Ozempic (semaglutide for T2D, FDA-approved-for-marketing-claims 2017-12) with independent titration of each component per their respective FDA labels; this is not off-label use (both standalone components are FDA-approved for T2D), but it is operationally a separate-injection regimen rather than the fixed-ratio IcoSema combination product. §10.6 develops the US-clinician off-label / pre-approval counseling beat (Anchor 5 dominant).

5. Maintenance protocol

5.1 Purpose

Define the post-titration, target-dose-attained operating state of the IcoSema protocol: target dose, monitoring intervals, dose-adjustment triggers, and the transition between maintenance and discontinuation (§8) or non-response algorithm (§7). §5 is the longest operational phase: for a T2D patient who tolerates target dose and continues therapy, IcoSema maintenance is open-ended for the duration of clinical benefit (HbA1c at individualized target with acceptable hypoglycemia and AE burden).

5.2 Target dose

The target dose is the dose at which the IcoSema primary effect (HbA1c reduction to individualized target) is achieved for the patient. Unlike semaglutide-monotherapy where the CWM target dose (2.4 mg weekly) and the T2D target dose (1.0 or 2.0 mg weekly) are label-fixed, IcoSema’s fixed-ratio combination is titrated to glycemic target rather than to a label-fixed numeric target dose — the dose individualizes to the patient’s insulin and GLP-1 RA requirements within the label-permitted dose range.

The COMBINE 1/2/3 trials titrated IcoSema dose to HbA1c target over 52 weeks; population-mean target doses are reported in the primary publications. The clinician-translation: maintain the dose at which the patient achieves and sustains HbA1c at individualized target with acceptable hypoglycemia and AE burden. Refer to the current EMA Kyinsu SmPC for label-permitted maximum dose.

5.3 Monitoring intervals

Monitoring intervals for the IcoSema maintenance phase reflect the once-weekly PK profile and the combined-mechanism monitoring framework:

HbA1c every 3 months during titration; every 3–6 months at maintenance per ADA Standards of Care. Quarterly cadence reflects the once-weekly PK schedule (~4 weeks to steady-state per dose adjustment plus assessment window).
Weight and BP every visit. Weight monitoring informs body-weight effect direction tracking (the IcoSema population-mean +0.84 kg net at 52 weeks per COMBINE 2 vs semaglutide-monotherapy comparator; individual variation around population mean).
eGFR / serum creatinine annually; quarterly during titration if baseline eGFR <60.
UACR annually for any patient with T2D + CKD risk profile.
Lipid panel annually.
CGM data review at every visit for patients on CGM; otherwise BGM-derived glycemic-pattern review.
Hypoglycemia-frequency interview at every visit (severity per ADA categorization, patterns, response). Persistent Level 2/3 hypoglycemia or recurrent events warrant dose-adjustment review (§5.4) or transition to alternative regimen (§7).
Organ-baseline reassessment annually — TSH; ophthalmology dilated exam annually for T2D + DR baseline patients (more frequent if rapid HbA1c improvement context); lipase on symptom-prompted basis only.

5.4 Dose-adjustment triggers

Dose adjustment in the IcoSema maintenance phase is driven by three trigger categories, with combo-product-specific considerations:

Target-not-met (HbA1c above individualized target on current dose with adequate observation window). At maximum tolerated IcoSema dose with persistent above-target HbA1c, the §7 non-response algorithm applies — typical options include transition to separate-injection regimen with independent insulin titration up beyond the IcoSema fixed-ratio maximum; transition to basal-bolus regimen (counseling beat per §10 on the reverse of the COMBINE 3 comparison); transition to tirzepatide where T2D + obesity context makes the SURPASS-2 head-to-head higher-effect comparator relevant.
Target-overshoot (HbA1c below individualized target with hypoglycemia frequency or severity concern). The combined-mechanism hypoglycemia-risk profile is generally favorable vs basal-insulin-alone but the insulin-component does carry hypoglycemia risk. Persistent Level 2/3 hypoglycemia at HbA1c near individualized target triggers dose-down or transition to alternative regimen. The fixed-ratio combination cannot down-titrate insulin component independently of semaglutide-component; clinicians transitioning a patient with hypoglycemia at target HbA1c may need to step down to separate-injection regimen with independent insulin dose reduction.
AE-emergent (new or worsening AE that responds to dose reduction). Persistent moderate-severity GI AE at IcoSema target dose that does not respond to symptomatic management is the most common dose-down trigger — step down to prior dose with reassessment at Month 3 (whether the lower dose maintains adequate glycemic control for the patient).

Dose-adjustment options: hold dose (maintain current); titrate up (if not at maximum permitted label dose and HbA1c above target with no hypoglycemia concern); titrate down (to prior label-permitted lower dose for tolerability); transition to non-response algorithm (§7); transition to separate-injection regimen for independent titration flexibility.

5.5 Worked example — IcoSema maintenance (continuation of §4.7 case)

The 58-year-old male from §4.7 continues at IcoSema target dose. Monitoring visits at Month 4, 7, 10, 13 in the first year on target dose; semi-annually thereafter.

Month 4 — HbA1c 6.9%, on individualized target. Weight stable from baseline. CGM time-in-range 76%; Level 2 hypoglycemia 0.1 events per 14-day period. Continue current dose.
Month 7 — HbA1c 6.8%; stable. Weight stable. Continue current dose.
Month 10 — HbA1c 6.7%; stable. Annual labs (eGFR 70, UACR 45, lipid panel stable). Annual dilated retinal exam shows stable mild non-proliferative DR. Continue current dose.
Month 13 — HbA1c 6.8%; stable. Annual TSH normal. Continue current dose, transition to semi-annual monitoring.

Dose-adjustment scenario (hypothetical). If at Month 7 the patient reports recurrent Level 2 hypoglycemia (e.g., 1–2 events per week, nocturnal pattern on CGM), dose-down to prior label-permitted lower dose with reassessment at Month 10 (whether the lower dose maintains adequate glycemic control). If hypoglycemia persists despite dose-down, consider transition to separate-injection regimen with independent insulin reduction; if HbA1c rises above target on the lower dose, consider alternative regimen per §7.

Pattern W cross-check at §5.5. The monitoring intervals above reconcile with the §3 pre-treatment panel (every monitoring lab is established as a baseline lab) and with the §6 AE-management algorithms (every AE-trigger lab is in the monitoring schedule).

6. Side-effect management

6.1 Purpose

Define the anticipatory framing and clinician response algorithms for the adverse-event categories that apply to IcoSema. §6 is the AE-by-AE-class operational reference. The IcoSema-specific structural feature of this section: hypoglycemia is positioned first (§6.7 in template-mapping terms, but operationally the dominant AE class for IcoSema) — reflecting the actual clinical-management priority for a combo-product containing basal insulin. Pure GLP-1 RA monotherapies do not carry hypoglycemia risk as a load-bearing AE class; for IcoSema the insulin-component contribution makes hypoglycemia the dominant clinical-management consideration even though GI AE is more common in absolute incidence.

Each AE-class sub-block opens with anticipatory framing per Pattern R: what the AE is, why it occurs, how common it is in the COMBINE program — before management. Pattern Z anchor 1 calibration applies to hypoglycemia framing: lead with what hypoglycemia management practice IS (a routine clinical skill for insulin-treated T2D patients), not with the prohibition framing.

6.2 Hypoglycemia AE class — IcoSema-dominant clinical-management priority

Anticipatory framing. Hypoglycemia is the dominant clinical-management safety concern for IcoSema given the insulin-component contribution. Unlike pure GLP-1 RAs (where the glucose-dependent insulin-secretion mechanism is protective against hypoglycemia in the absence of concurrent insulin or sulfonylurea), IcoSema’s insulin icodec component drives insulin signaling regardless of plasma glucose level — the canonical hypoglycemia mechanism. The GLP-1R component partially offsets through glucose-dependent insulin secretion (which acts only when plasma glucose is elevated, not when low) but does not eliminate insulin-driven hypoglycemia risk.

The COMBINE program reported hypoglycemia per ADA Standards of Care categorization:

Level 1: glucose <70 mg/dL (3.9 mmol/L) but ≥54 mg/dL (3.0 mmol/L); alert value.
Level 2: glucose <54 mg/dL (3.0 mmol/L); clinically significant hypoglycemia warranting intervention.
Level 3: severe hypoglycemia of any glucose level requiring assistance from another person for recovery.

Per-trial hypoglycemia rates (clinically significant + severe combined, episodes per person-year):

Trial	Comparator	IcoSema rate	Comparator rate	Rate ratio (95% CI)	p-value
COMBINE 1	Insulin icodec alone	0.14	0.63	0.22 (0.14–0.36)	<0.0001
COMBINE 2	Semaglutide 1.0 mg	0.042	0.036	1.20 (0.53–2.69)	0.66 (NS)
COMBINE 3	Basal-bolus (glargine + aspart)	0.21	2.23	0.12 (0.08–0.17)	<0.0001

Cross-trial pattern. IcoSema is substantially favored vs insulin comparators (rate ratios 0.22 and 0.12) and modestly elevated but non-significant vs semaglutide-alone (rate ratio 1.20). Mechanism explanation: the GLP-1R glucose-dependent insulin-secretion contribution allows lower total basal insulin dose, reducing absolute hypoglycemia events vs basal-insulin-alone or basal-bolus regimens; adding the basal insulin component to a GLP-1 RA monotherapy baseline introduces modest absolute hypoglycemia risk.

Nocturnal hypoglycemia. Nocturnal hypoglycemia (between bedtime and waking) is a recognized risk pattern for basal insulin therapy. The once-weekly dosing schedule produces relatively flat plasma profiles across the dosing interval rather than acute nocturnal peaks (per the Bolli 2025 Diabetes Obes Metab review PMID 40678871); this profile differs from short-acting and intermediate-acting insulin regimens that produce more pronounced nocturnal exposure patterns.

Identification. Patient-reported hypoglycemia events; CGM data review (Level 2 events <54 mg/dL detection; time-below-range metrics); HbA1c trajectory below individualized target with persistent hypoglycemia frequency.

First-line management. Glucose-restoration practice (15-15 rule: 15 g rapid-acting carbohydrate; recheck glucose in 15 minutes; repeat if still <70 mg/dL). Identify trigger (missed meal, increased physical activity, concurrent illness, alcohol intake, dose-adjustment timing). Reinforce hypoglycemia recognition and treatment counseling.

Concurrent-medication review. Persistent hypoglycemia patterns trigger review of concurrent medications:

Sulfonylurea or meglitinide concurrent: reduce or discontinue.
Beta-blocker concurrent: hypoglycemia-unawareness contribution; clinician judgment.
Recent dose change in any glucose-affecting medication.

Dose-adjustment context. Persistent Level 2/3 hypoglycemia warrants dose-adjustment review. IcoSema’s once-weekly PK means dose adjustments take 3-4 weeks to reach new steady-state; patients and clinicians plan adjustment timing accordingly. Fixed-ratio combination cannot down-titrate insulin component independently of semaglutide-component; clinicians may transition to separate-injection regimen for independent insulin reduction (§7.3 develops).

Escalation triggers. Recurrent severe hypoglycemia (Level 3, requiring assistance) → urgent clinician evaluation; consider transition to alternative regimen with lower hypoglycemia risk profile.

Discontinuation triggers. Hypoglycemia is not typically a discontinuation indication for IcoSema (dose-adjustment or regimen-transition is the standard response); however, recurrent severe hypoglycemia with hypoglycemia-unawareness phenotype despite dose reduction and concurrent-medication optimization may warrant transition to non-insulin-containing regimen.

Patient-counseling beat (Pattern Z anchor 1 calibration applied — lead with what hypoglycemia management practice IS): “Hypoglycemia management is standard practice for patients on insulin-containing regimens like IcoSema. You’ll learn to recognize early signs through experience — most people notice their personal pattern of warning symptoms after a few episodes. We’ll set up glucose monitoring that works for your daily routine — that might be a continuous monitor or fingerstick checks depending on your preference and insurance coverage. The 15-15 rule for treating low glucose with quick carbs is straightforward to learn. The once-weekly schedule means dose adjustments happen gradually; we’ll plan changes together when patterns suggest a change is needed. If you experience severe low glucose requiring help from someone else, or recurrent unexplained low glucose, that’s a reason to call. Most patients find the management practice straightforward once they’re set up; the combination of basal insulin and GLP-1 RA in IcoSema actually produces lower hypoglycemia rates than basal insulin alone, because the GLP-1 RA component handles the postprandial glucose control without adding to hypoglycemia risk.”

6.3 GI AE class — semaglutide-component dominant

Anticipatory framing. GLP-1 RA-mediated delayed gastric emptying, central appetite-pathway modulation, and direct GI-motility effects produce the characteristic AE profile: nausea, vomiting, diarrhea, constipation, eructation, abdominal pain. The AE profile is typically peak-at-dose-escalation and attenuates within 2–4 weeks at stable dose. COMBINE 2 reported GI adverse events at IcoSema 31.4% vs semaglutide-alone 34.4% — comparable rates suggesting the combo formulation does not amplify the semaglutide-component GI profile.

Identification. Patient-reported during early-monitoring contacts (§4) and maintenance visits (§5). Standardized severity grading via CTCAE: Grade 1 (mild), Grade 2 (moderate), Grade 3 (severe, hospitalization indicated), Grade 4 (life-threatening). Most protocol-relevant GI AEs are Grade 1–2.

First-line management. Non-pharmacologic: meal-size reduction, slow eating pace, low-fat / non-greasy meal composition, hydration emphasis. Pharmacologic: ondansetron 4 mg PRN for nausea; loperamide PRN for diarrhea; osmotic laxative for constipation; reassurance for eructation.

Escalation triggers. Grade 3 severity at any step. Persistent Grade 2 severity beyond 4 weeks at stable dose. Vomiting with severe abdominal pain (assess for pancreatitis — §6.5).

Discontinuation triggers. Grade 3 GI AE that does not resolve with dose-down. Patient preference.

6.4 Gallbladder AE class

Anticipatory framing. GLP-1 RA association with cholelithiasis and cholecystitis has been demonstrated across the trial program; mechanism is attributed to weight-loss-rate-related bile-supersaturation and direct effects on gallbladder motility. IcoSema’s modest weight-loss profile (COMBINE 2 +0.84 kg net at 52 weeks vs semaglutide-alone) reduces the absolute bile-supersaturation contribution; gallbladder risk is correspondingly modest relative to higher-weight-loss-magnitude semaglutide-monotherapy programs (Wegovy 2.4 mg, Wegovy HD 7.2 mg).

Identification. Symptom-prompted — right upper quadrant pain, post-prandial colicky pain, fever-and-pain combination. Ultrasound is the first-line imaging study.

First-line management. Symptomatic gallstones with confirmed cholelithiasis: surgical consultation. Temporary hold during peri-operative window; resume at maintenance dose post-cholecystectomy if no other contraindication has emerged.

Escalation triggers. Acute cholecystitis with systemic signs. Choledocholithiasis suspected.

Discontinuation triggers. Recurrent symptomatic gallstones post-cholecystectomy with bile-acid-related pattern: clinician judgment.

6.5 Pancreatitis AE class

Anticipatory framing. Class-level evidence (Wen et al Endocrinol Diabetes Metab 2025 October PMID 40988099; 62 GLP-1 RA RCTs; n=66,232) reported pooled RR 1.44 (95% CI 1.09–1.89, p=0.009) for acute pancreatitis with GLP-1 RAs vs comparator; authors frame as “slightly increased risk, likely minimal.” Absolute event rates remain low (<1–2% per year). IcoSema inherits the semaglutide-component pancreatitis-signal characterization; the combination product does not amplify the signal.

Identification. Symptom-prompted — persistent severe upper abdominal pain radiating to back, often with vomiting. Differential: GI AE class (§6.3) typically does not produce severe persistent localized pain.

First-line management. Acute persistent severe abdominal pain with vomiting → immediate evaluation, lipase + amylase + CMP + CBC + lipid panel (rule out hypertriglyceridemic pancreatitis), abdominal imaging. Discontinue IcoSema pending evaluation.

Escalation triggers. Confirmed acute pancreatitis → hospitalization, supportive management per pancreatitis standard-of-care.

Discontinuation triggers. Confirmed acute pancreatitis attributable to the molecule (excluding alternative etiology) → permanent discontinuation; transition to non-GLP-1 alternative for the T2D indication (basal insulin alone with potential SGLT2 inhibitor; DPP-4 inhibitor cautioned given DPP-4 class also carries pancreatitis labeled cautionary use).

6.6 NAION AE class — semaglutide-component-inherited signal under evaluation

Anticipatory framing. Multiple converging lines of evidence support a semaglutide-NAION association: Hathaway 2024 single-center cohort (PMID 38958939; HR 4.28–7.64); Grauslund Danish nationwide cohort n=424,152 (PMID 39696569; HR 2.19 at 5 years); Lakhani 2025 180-country observational pharmacovigilance study (PMID 40383360; FAERS ROR 11.12 — observational pharmacovigilance, not meta-analysis per Pattern AB.4 precision). EMA PRAC classified NAION as “very rare” side effect (June 2025; ~1 case per 10,000 person-years). Tirzepatide signal absent at same threshold (class-differentiated to semaglutide).

IcoSema inherits the semaglutide-component signal at fixed-ratio dose. Pattern AA precision: signal under evaluation, not a labeled warning at semaglutide-component level as of 2026-05-13.

Identification. Acute monocular vision loss, typically painless, often altitudinal visual field defect. Optic disc edema on exam in the acute phase.

First-line management. Urgent ophthalmology evaluation. Discontinue IcoSema pending evaluation. Cross-reference ESR/CRP to rule out arteritic etiology.

Escalation triggers. Confirmed NAION → permanent discontinuation; ophthalmology co-management; consideration of contralateral-eye risk.

Discontinuation triggers. Confirmed NAION → permanent IcoSema discontinuation regardless of indication continuation; transition to non-semaglutide-containing T2D regimen.

6.7 Injection-site AE class

Anticipatory framing. Injection-site reactions occur in a small percentage of patients on weekly subcutaneous combination injections; usually mild and self-resolving. Both insulin and GLP-1 RA components can produce injection-site reactions; the combination injection does not have a distinct injection-site AE profile beyond the component-level expectations.

First-line management. Site-rotation reinforcement (abdomen, thigh, upper arm — alternate weekly). Topical hydrocortisone for pruritus.

Discontinuation triggers. Severe / systemic hypersensitivity reaction is a labeled contraindication (§2.4) and triggers permanent discontinuation.

6.8 Diabetic retinopathy considerations — semaglutide-component-inherited

The rapid-HbA1c-improvement context that drove the SUSTAIN-6 retinopathy-complication signal in patients with established diabetic retinopathy applies to IcoSema. The combined-mechanism HbA1c effect (COMBINE 1 −0.66% treatment difference; COMBINE 3 −0.06% non-inferiority; etc.) may produce rapid glycemic improvement in patients with baseline HbA1c ≥9.0. Pre-treatment dilated retinal exam per §3.7 is load-bearing for these patients; ophthalmology co-management for advanced background or proliferative DR is the protocol-recommended posture.

6.9 Cardiovascular safety

No CV signal of concern across the semaglutide-component (SELECT PMID 37952131; SUSTAIN-6 PMID 27633186) or insulin icodec component (ONWARDS Phase 3 program safety surveillance) evidence bases. The combo-product specific CV outcomes evidence is research-state-incomplete (no dedicated IcoSema CVOT; see §1 OUT-OF-SCOPE framing).

6.10 Drug-drug interaction management (operational framework)

The IcoSema DDI framework is the union of the semaglutide-component DDI considerations (gastric emptying delay; potential effects on oral drug absorption) and the insulin-component DDI considerations (additive hypoglycemia with sulfonylurea / meglitinide; glucocorticoid glucose-elevation; concurrent insulin avoidance).

Sulfonylureas and meglitinides (additive hypoglycemia). Discontinue or substantially reduce sulfonylurea dose (typically 50% reduction or discontinuation) at IcoSema initiation per §6.11.1 of the canonical.

Glucocorticoids (glucose-elevation). IcoSema dose may require up-titration during glucocorticoid therapy and down-titration after glucocorticoid discontinuation. Hypoglycemia surveillance is load-bearing during glucocorticoid taper.

Concurrent separate basal insulin. Discontinue separate basal insulin at IcoSema initiation; the IcoSema insulin-icodec component replaces it.

Concurrent bolus insulin (insulin aspart, lispro, glulisine). Research-state-evolving — patients on IcoSema with persistent postprandial hyperglycemia despite combo-product titration may transition to IcoSema + meal-time bolus insulin per individual clinical decision (the EMA Kyinsu indication does not specifically address this off-label combination; clinical practice for analogous situations with daily combo products may apply).

Concurrent other GLP-1 RAs. Not combined with IcoSema (semaglutide-component already provides GLP-1R agonism; concurrent administration of separate GLP-1 RA produces additive GI AE without efficacy benefit).

Concurrent DPP-4 inhibitors. Not combined with IcoSema (mechanism-overlap; discontinue DPP-4 inhibitor at IcoSema initiation).

Drugs with narrow therapeutic window affected by gastric emptying delay (warfarin, levothyroxine). May require monitoring after IcoSema initiation; inherit framework from semaglutide-component DDI considerations.

6.11 Pregnancy and lactation

Inherits §8.4 pre-conception arithmetic. Both components have ~7-8 day half-lives; ~5 half-lives = ~35-40 days pharmacokinetic clearance. EMA Kyinsu label recommends discontinuation prior to planned pregnancy with appropriate washout. Pattern Z anchor 3 calibration (pregnancy-planning conversation) applies per §10.4 below.

Human pregnancy-exposure data — Parker SE et al Diabetes Obes Metab 2025 (PMID 40329607): pooled review of unplanned pregnancies from FDA- and EMA-submitted GLP-1 RA regulatory clinical trials. Incidence of congenital abnormalities appears relatively low in this pooled dataset; sample size is limited and the exposures are from unplanned-pregnancy contexts. This is the load-bearing human evidence anchor; it leads the pregnancy-planning conversation per Pattern Z anchor 3.

Discovery of pregnancy on protocol. Immediate discontinuation; transition to pregnancy-appropriate insulin regimen (typically NPH or basal-analog with bolus insulin per ADA Standards of Care for pregnancy in T2D). Obstetrics co-management.

6.12 Worked example — IcoSema AE management at maintenance

A T2D adult on IcoSema at target dose, Month 6, presents with persistent moderate nausea (Grade 2, recurring 2–3 days after each weekly injection) and 2.0 kg total weight gain from baseline (within COMBINE 2 expected trajectory for this sub-population).

Protocol response. Anticipatory framing: within trial-program-typical GI AE profile at maintenance dose; recurring 2–3-day post-injection pattern consistent with the pharmacokinetic profile of weekly semaglutide-component. First-line management: meal-size and meal-composition reinforcement; consider scheduled (not just PRN) ondansetron for the 2–3-day-post-injection window. Reassessment at Month 7 visit: if nausea remains Grade 2 and patient is on glycemic target, continue current dose; if Grade 2 nausea is unacceptable to patient (Pattern Z calibration: patient-preference-anchored, not clinician-override), dose-down to prior label-permitted lower dose with Month 9 HbA1c-trajectory reassessment.

A T2D adult on IcoSema, Month 3, presents with recurrent Level 2 hypoglycemia (3 events per week over past 2 weeks, nocturnal pattern on CGM), HbA1c trajectory 7.2 → 6.4% (overshoot below individualized target 7.0%).

Protocol response. Hypoglycemia overshoot identified. Concurrent-medication review (no sulfonylurea or separate insulin; no recent dose change in glucose-affecting medication). Dose-down to prior label-permitted lower IcoSema dose. Reassessment at 4 weeks (steady-state on new dose) — if hypoglycemia resolves and HbA1c stable at 6.5–7.0%, continue at lower dose; if HbA1c rises above 7.5%, consider transition to separate-injection regimen for independent insulin dose reduction with maintained GLP-1 RA dose (counseling beat: independent titration flexibility is the relevant trade-off).

A T2D adult on IcoSema, Month 4, develops acute painless monocular vision loss in left eye. Ophthalmology urgent evaluation confirms left optic disc edema with altitudinal visual field defect; ESR/CRP normal. Diagnosis: NAION, left eye.

Protocol response. Discontinue IcoSema. Confirmed NAION → permanent IcoSema discontinuation (semaglutide-component signal inheritance). Ophthalmology co-management. Transition to non-semaglutide-containing T2D regimen — basal insulin alone (insulin icodec via Awiqli standalone) ± non-GLP-1 OAD therapy; tirzepatide is an alternative if T2D + obesity context and patient is willing to attempt an alternative GLP-1 / GIP class (NAION signal has been described specifically for semaglutide; cross-class direction-of-effect is not established — Pattern V flag).

Pattern AA precision in §6.12. “Pancreatitis-attributable to IcoSema” (in a hypothetical pancreatitis case) is precise (alternative etiologies ruled out, temporal association); the labeled framing inherits the semaglutide-component precaution / labeled cautionary use, not labeled contraindication. “NAION signal” is precise — post-marketing pharmacovigilance signal at semaglutide-component level, EMA PRAC “very rare” classification, not labeled warning as of 2026-05-13.

7. Plateau and non-response algorithm

7.1 Purpose

Define the structured clinical-decision approach when IcoSema’s primary effect has not met its target (non-response) or has stalled below a clinically meaningful endpoint (plateau). For IcoSema’s T2D-only indication scope, the primary effect is HbA1c reduction to individualized target with acceptable hypoglycemia and AE burden. “Non-response” framing for IcoSema is therefore HbA1c-trajectory-anchored, not weight-trajectory-anchored.

7.2 Pseudo-plateau vs true plateau vs non-response — diagnostic distinctions

Pseudo-plateau. Apparent stall in HbA1c that is in fact within normal month-to-month variation or reflects on-trajectory progression at sub-target dose. Recognized by trajectory-context — comparing the patient’s curve to the COMBINE program-typical curve for the sub-population.

True plateau. Legitimate response stall — patient was responding, then HbA1c trajectory flattens or rises. Recognized by trajectory inflection plus adequate observation window (typically 2–3 months at stable dose to confirm).

Non-response. Insufficient initial effect from the start. Recognized at Month 3–6 on target dose with HbA1c substantially below the COMBINE program-typical trajectory for the patient’s sub-population.

7.3 Decision tree for plateau / non-response

Confirm adherence. Missed doses, injection-technique issues are the most common pseudo-non-response causes. Confirm via patient interview and prescription-refill audit.
Confirm trajectory-context. Plot the patient’s HbA1c curve against the COMBINE program-typical curve for their sub-population. If on-trajectory, pseudo-plateau — continue current dose, reassess at next interval.
Confirm dose attainment. Is the patient on label-permitted target dose? If not, continue titration to target; reassess at Month 3 on target.
Reassess phenotype. Does the patient’s clinical picture support a phenotype the COMBINE program enrolled? Pattern V direction-of-effect — for under-represented phenotypes (insulin-naïve sub-population pending COMBINE 4; LADA misclassification possibilities; very advanced beta-cell failure), expected effect-size may be smaller.
If true plateau / non-response confirmed at adequate observation window in T2D context:
- Transition to separate-injection regimen with independent insulin up-titration. IcoSema’s fixed-ratio combination cannot titrate insulin component independently of semaglutide-component; transitioning to Awiqli (insulin icodec standalone) at independently titrated dose + Ozempic (semaglutide at maintained or up-titrated dose to 2.0 mg per SUSTAIN-FORTE) gives independent insulin escalation beyond the IcoSema fixed-ratio constraint.
- Transition to basal-bolus regimen. The reverse of the COMBINE 3 comparison — if IcoSema does not achieve target HbA1c at maximum dose and the patient is willing to accept the basal-bolus injection burden, basal-bolus is the next intensification step. Counseling beat per §10 on the trade-offs (730+ injections/year vs 52; expected weight gain vs IcoSema’s modest profile; higher hypoglycemia rate vs IcoSema).
- Transition to tirzepatide where T2D + obesity context applies. SURPASS-2 head-to-head supports tirzepatide as the higher-effect alternative to semaglutide-1.0-mg-monotherapy for T2D glycemic control. For IcoSema non-responders in the GLP-1 RA-experienced sub-population (COMBINE 2 phenotype), transitioning to tirzepatide-monotherapy (Mounjaro) or to tirzepatide + basal insulin (separate-injection regimen) is the within-class higher-effect alternative.
- Add SGLT2 inhibitor. If not already on SGLT2 inhibitor and CV/renal-risk profile supports, adding empagliflozin / dapagliflozin / ertugliflozin is class-additive HbA1c reduction with CV and kidney benefits. Combination is well-supported and not contraindicated.
- Intensification of behavioral / nutritional / activity program. Behavioral intensification is effect-additive to pharmacotherapy in T2D management per general diabetes literature.

7.4 Worked example — IcoSema non-responder algorithm

A T2D adult on IcoSema at label-permitted target dose, Month 6, has HbA1c 8.2% (baseline 9.0%, target <7.0%) — below the COMBINE 1-typical trajectory of 1.5 percentage-point absolute reduction at 52 weeks. Adherence confirmed; dose at target; phenotype matches COMBINE 1 enrollment (basal-insulin-experienced T2D).

Algorithm walkthrough.

Step 1 — adherence. 100% adherence; prescription-refill audit confirms.

Step 2 — trajectory-context. Patient’s curve at Month 6 (8.2%, 0.8 percentage-point reduction) is below the COMBINE 1 25th-percentile trajectory at Month 6. Not on-trajectory.

Step 3 — dose attainment. On target dose; no further titration available within IcoSema label-permitted dosing.

Step 4 — phenotype reassessment. Phenotype matches COMBINE 1 enrollment. C-peptide reviewed: very low, consistent with advanced beta-cell failure. Phenotype-specific consideration: advanced beta-cell failure phenotype may have higher absolute insulin-component requirements than the IcoSema fixed-ratio can deliver.

Step 5 — non-response confirmed; T2D decision branches.

5a. Transition to separate-injection regimen. Discontinue IcoSema. Initiate Awiqli (insulin icodec standalone) at calibrated higher dose to address the insulin-component requirement; continue Ozempic 2.0 mg (semaglutide-monotherapy) at maintained or up-titrated weekly dose. Two injections per week, independent titration flexibility, mechanism-equivalent component coverage. Reassess HbA1c at 3 months post-transition.

5b. Transition to basal-bolus regimen. Discontinue IcoSema. Initiate basal insulin (glargine, degludec) titrated to fasting glucose target + prandial insulin titrated to postprandial glucose target. Counseling beat: injection-burden increase from 52 to 730+/year; expected weight gain; higher hypoglycemia rate vs IcoSema; sometimes preferred for very advanced beta-cell failure phenotypes.

5c. Add SGLT2 inhibitor if not on. If not currently on SGLT2 inhibitor and renal function supports, adding empagliflozin 10 mg daily produces additive HbA1c reduction (typically 0.5–1.0 percentage points), with CV and kidney benefits. May obviate the need for regimen-change if HbA1c reaches target with the added agent.

Pattern Z calibration anchor 4 applied at §7.4. The decision among 5a / 5b / 5c is patient-anchored, not clinician-mandated. Counseling beats present trial-program-anchored effect-size estimates for each option, present the trade-offs (injection burden, hypoglycemia rate, weight effect, cost, adherence-complexity), and the clinician-patient decision is patient-preference-driven within the medically reasonable options.

8. Discontinuation and tapering

8.1 Purpose

Define when to stop IcoSema, how to taper if tapering is indicated, and how to frame post-discontinuation expectations. For IcoSema’s T2D-only indication scope, the post-discontinuation framing differs from semaglutide-monotherapy’s CWM context: the dominant post-discontinuation expectation is insulin requirement re-emergence + GLP-1 RA effect attenuation, not the weight-regain trajectory that dominates the post-Wegovy framing. §8 addresses both axes.

8.2 When to discontinue — discontinuation triggers

Discontinuation is indicated when one of the following emerges:

Confirmed contraindication discovery (new MTC diagnosis, new MEN-2 family-history identification, new pregnancy). §2.4 hard contraindications are immediate-discontinuation triggers.
Severe AE attributable to the molecule (confirmed acute pancreatitis, confirmed NAION, severe hypersensitivity reaction). §6 AE-class-specific discontinuation triggers.
Glycemic control resolution (rare in T2D; some patients achieve sustained euglycemia with weight loss and lifestyle modification post-IcoSema; not a typical IcoSema outcome but documented in the broader T2D-remission literature).
Patient preference. Patient-anchored discontinuation decision is legitimate; the protocol’s role is to inform post-discontinuation expectations and the re-initiation pathway, not to override.
Cost / access barriers. A documented practice-level reality. Pattern Z anchor 2 framing (compounded vs FDA-approved) is N/A for the IcoSema combination product as of 2026-05-13 in the US (no compounded combo-product market exists for an unapproved combo); Pattern Z anchor 5 (off-label / extrapolation transparency for the US pre-approval state) applies for US clinicians prescribing the separate-injection regimen as an alternative.
Pre-conception planning for reproductive-age patients: discontinuation with approximately 35–40-day pharmacokinetic washout per §8.4 below.
Non-response confirmed at adequate observation window with regimen-change decision per §7.

8.3 How to taper — IcoSema-specific tapering considerations

For IcoSema in T2D context: pharmacokinetic tapering (step-down dosing) is not pharmacologically required for the GLP-1 RA component (the half-life-driven pharmacokinetic washout occurs at fixed kinetics regardless of taper schedule); however, the insulin-component requires transition planning because abrupt insulin-component discontinuation can produce hyperglycemia rebound if alternative basal insulin or alternative glucose-control regimen is not in place.

Transition vs taper. For IcoSema, the relevant pattern is typically “transition to alternative regimen” rather than “taper to nothing”:

Discontinuing IcoSema in the context of regimen change (e.g., transition to separate-injection regimen, or transition to basal-bolus, or transition to non-insulin regimen if HbA1c control allows): the alternative regimen is initiated at calibrated dose at the IcoSema discontinuation point; no IcoSema taper required.
Discontinuing IcoSema in the context of patient-preference cessation without alternative pharmacotherapy: clinician judgment on whether basal insulin alone (Awiqli or daily basal analog) is needed as a bridge; HbA1c trajectory monitoring at Months 3, 6, 12 post-discontinuation; resumption pathway if hyperglycemia returns.
Discontinuing IcoSema for pregnancy planning: scheduled discontinuation with 35–40-day washout for the IcoSema components plus transition to pregnancy-appropriate insulin regimen (NPH or basal-analog with bolus insulin per ADA Standards of Care).

Pattern AA precision: “Transition to alternative regimen” is the operational pattern within label / clinician-judgment for IcoSema discontinuation in T2D context. “Taper to nothing” is rarely appropriate for a T2D pharmacotherapy because the underlying T2D pathophysiology persists post-discontinuation.

8.4 Pre-conception washout arithmetic

Semaglutide-component elimination half-life approximately 1 week (165–184 hours). Insulin icodec elimination half-life approximately 1 week (~196 hours). Approximately 5 half-lives are required for >95% pharmacokinetic clearance — approximately 35–40 days for both components.

The EMA Kyinsu label recommends discontinuation prior to planned pregnancy with appropriate washout (refer to current label for precise interval specification). The semaglutide-component label (Wegovy / Ozempic) recommends ≥8 weeks pre-conception washout for the semaglutide-monotherapy formulations; the conservative interval accounts for pharmacokinetic and pharmacodynamic clearance plus a margin. For IcoSema, the same conservative 8-week pre-conception window is the operational counseling beat with appropriate transition to pregnancy-appropriate insulin regimen during the washout window.

8.5 Post-discontinuation framing

The trial-program data on post-IcoSema-discontinuation HbA1c trajectory anchor to the COMBINE program follow-up periods (where available). The mechanism prediction: insulin-component discontinuation produces insulin-requirement return; semaglutide-component discontinuation produces GLP-1R-effect attenuation (gastric emptying returns to baseline; postprandial glucose excursions return; appetite suppression attenuates).

Post-discontinuation patient counseling addresses:

Glycemic trajectory. HbA1c is expected to drift upward over 1–3 months post-discontinuation as both components clear. Alternative pharmacotherapy is the standard response — abrupt discontinuation without replacement is appropriate only in the context of severe AE attributable to the molecule (e.g., confirmed pancreatitis, confirmed NAION) where the alternative regimen is initiated in parallel.
Weight trajectory. For IcoSema, the modest weight effect (combination produces net +0.84 kg vs semaglutide-monotherapy comparator) means post-discontinuation weight regain is not the dominant axis (unlike semaglutide-monotherapy CWM context). Some patients experience modest weight loss post-IcoSema-discontinuation if their alternative regimen has different body-weight characteristics (e.g., transition to tirzepatide-monotherapy may produce body-weight reduction; transition to basal-bolus may produce modest weight gain).
Hypoglycemia trajectory. Post-IcoSema-discontinuation hypoglycemia risk depends on the replacement regimen. Basal-insulin-alone replacement: hypoglycemia rate higher than IcoSema (COMBINE 1 anchor: insulin icodec-alone rate 0.63 vs IcoSema 0.14 per person-year). Basal-bolus replacement: substantially higher hypoglycemia rate (COMBINE 3 anchor: basal-bolus rate 2.23 vs IcoSema 0.21). GLP-1 RA-alone replacement: hypoglycemia rate lower than IcoSema (COMBINE 2 anchor: semaglutide-alone rate 0.036 vs IcoSema 0.042).

8.6 Re-initiation pathway

A patient who discontinued IcoSema and is considering re-initiation: typically re-titration from a conservative starting dose is the protocol-recommended pattern — the tolerability re-priming (GI AE; hypoglycemia surveillance) is equivalent to initial titration; abrupt re-initiation at the prior maintenance dose is not recommended due to GI AE recurrence risk and hypoglycemia risk if insulin requirements have changed during the off-period.

§4 titration applies. Indication confirmation per §1, selection-criteria re-screening per §2, pre-treatment workup refresh per §3 if the discontinuation period exceeded approximately 12 months or if any new comorbidity has emerged.

8.7 Worked example — IcoSema discontinuation scenarios

Scenario A — pre-conception planning. A 34-year-old female on IcoSema, Month 18 on target dose, HbA1c 6.7% with target <7.0%, planning conception in approximately 4 months. Counseling beat: discuss the 8-week pre-conception window per EMA Kyinsu label-equivalent recommendation; plan transition to pregnancy-appropriate insulin regimen approximately 8 weeks before planned conception. Pattern Z calibration anchor 3 applies — pregnancy-planning counseling leads with Parker 2025 human pregnancy-exposure data (PMID 40329607; incidence of congenital abnormalities appears relatively low in the pooled dataset of GLP-1 RA unplanned-pregnancy exposures, with sample-size limitation acknowledgment), then pharmacokinetic facts (1-week half-life of each component, 35–40-day pharmacokinetic clearance, 8-week label recommendation), then the transition-to-pregnancy-appropriate-insulin plan. The patient’s reproductive-planning decision is patient-anchored; the protocol presents facts.

Scenario B — patient-preference discontinuation in remission context. A 52-year-old male on IcoSema, Month 24, has lost 8 kg through lifestyle modification (despite IcoSema’s modest weight-effect direction, individual patient outcomes vary), HbA1c 6.4% with target <7.0%, and wishes to attempt discontinuation. Counseling beat: discuss the expected HbA1c trajectory post-discontinuation (likely drift upward over 1–3 months as both components clear); transition plan to alternative pharmacotherapy (likely metformin + SGLT2 inhibitor as the maintenance regimen if HbA1c remains controlled); re-initiation pathway available if HbA1c rises above target. Pattern Z calibration: patient-preference is legitimate; protocol’s role is to inform.

Scenario C — confirmed acute pancreatitis. Permanent discontinuation per §6.5. No taper — abrupt discontinuation appropriate given AE-attributable etiology. Transition to non-GLP-1 alternative for T2D: basal insulin alone (Awiqli or daily basal analog) ± SGLT2 inhibitor; DPP-4 inhibitor cautioned given class-overlap with pancreatitis labeled cautionary use.

Scenario D — new pregnancy on protocol. A 31-year-old female on IcoSema discovers pregnancy at approximately 6 weeks gestation. The combination is contraindicated in pregnancy. Immediate discontinuation. Obstetrics co-management; transition to pregnancy-appropriate insulin regimen (typically NPH or basal-analog with bolus insulin per ADA Standards of Care for pregnancy in T2D). The pharmacokinetic clearance window (~35–40 days, with first-trimester exposure already occurred) is documented for the obstetrics record. Parker 2025 reassuring early signal (PMID 40329607) is part of the obstetrics-context counseling. Post-pregnancy and post-lactation, re-initiation decision per §8.6 if indication and selection criteria are met.

Pattern Z calibration anchor 3 precision in §8.7. Pregnancy-planning counseling leads with Parker 2025 human data, then pharmacokinetic arithmetic, then the transition-to-pregnancy-appropriate-insulin plan, then the post-pregnancy re-initiation pathway. The protocol does not steer the patient toward continued pharmacotherapy by emphasizing post-discontinuation glycemic-drift risk, nor toward discontinuation by emphasizing pregnancy-exposure risk. Both facts are presented; the patient decides within the medically reasonable framework.

9. Combination rules

9.1 Purpose

Define what stacks with IcoSema, what is contraindicated in combination, and the rationale for each combination category. §9 bridges to Module 5.12 Combination Protocols and to the lean-mass-stack protocols where applicable. For IcoSema’s T2D-only indication scope, the combination framework focuses on T2D co-management agents (SGLT2 inhibitors, metformin, OAD adjuncts) and on the IcoSema-vs-separate-injection-regimen comparator framework (§7.3, §10.5).

Pattern W cross-section consistency: every combination in this section reconciles with §2 (no patient enters combination with contraindicated agent), §6 (combinations do not mask AE-class concerns), and §10 (counseling beats are Pattern Z-anchor-compliant).

9.2 Within-Module-5 combinations

The principal within-Module-5 combinations involve combining metabolic-axis mechanism classes in T2D context:

IcoSema + metformin (continuation of background OAD therapy). Mechanism rationale: metformin’s mechanism (AMPK-mediated hepatic glucose-production suppression + insulin sensitization) is mechanism-complementary to IcoSema’s combined basal insulin replacement + GLP-1R agonism. The COMBINE program enrolled patients on background metformin per standard T2D management. Combination is class-additive on HbA1c and is the canonical IcoSema regimen anchor for T2D patients in standard practice.
IcoSema + SGLT2 inhibitor (T2D + CV/CKD context). Mechanism rationale: SGLT2 inhibitor adds glucose-osmotic-diuretic and ketogenesis-modulating effects; the combination is broadly supported by individual-agent CVOT and KOOT data for the SGLT2-class and by FLOW/SELECT inheritance for the semaglutide-component. The combination is widely co-prescribed in T2D + CKD or T2D + ASCVD context. Mechanism-additive on HbA1c; no significant hypoglycemia interaction. This is the canonical polycondition IcoSema combination for the T2D + comorbidity-load patient.
IcoSema + insulin (basal or bolus) — operational distinctions.
- Concurrent separate basal insulin: contraindicated (the IcoSema insulin-icodec component already provides basal insulin replacement; concurrent administration would produce additive insulin dosing without titration flexibility).
- Concurrent bolus insulin (insulin aspart, lispro, glulisine): research-state-evolving / off-label. Patients on IcoSema with persistent postprandial hyperglycemia despite combo-product titration may transition to IcoSema + meal-time bolus insulin per individual clinical decision; the EMA Kyinsu indication does not specifically address this off-label combination. Counseling beat (Pattern Z anchor 5): present the trial-population scope (COMBINE program did not test IcoSema + bolus insulin as a registration regimen) and the clinical-practice extrapolation context.
IcoSema + sulfonylurea (relative — additive hypoglycemia). §6.10. Discontinue or substantially reduce sulfonylurea at IcoSema initiation; combination at full sulfonylurea dose is not recommended.
IcoSema + DPP-4 inhibitor — contraindicated combination. §6.10. Mechanism-overlap (DPP-4 inhibitor preserves endogenous GLP-1; redundant with exogenous GLP-1 RA via the semaglutide-component) makes the combination not clinically additive. Discontinue DPP-4 inhibitor at IcoSema initiation.
IcoSema + thiazolidinedione (pioglitazone). Mechanism-complementary; pioglitazone adds PPAR-γ-mediated insulin sensitization. Clinical practice typically does not combine in current T2D management given pioglitazone’s secondary effects (heart failure precaution, weight gain, bladder cancer risk-stratification), but combination is not contraindicated per mechanism overlap.
IcoSema + GLP-1 RA (additional / separate) — contraindicated combination. §6.10. The semaglutide-component already provides GLP-1R agonism; concurrent administration of separate GLP-1 RA produces additive AE without efficacy benefit.

9.3 Cross-Module combinations — lean-mass and body-composition stacks (limited applicability)

Cross-Module combinations involve adding a peptide outside the metabolic-axis family. For IcoSema in T2D context, the lean-mass-preservation framing that drives semaglutide-CWM stacks (M5.6 CJC-1295 / Ipamorelin context) is less load-bearing — IcoSema’s modest weight-effect direction means lean-mass loss during weight loss is rarely the dominant concern. Cross-Module combinations are generally not standard practice for IcoSema; clinician judgment in specific patient contexts may identify limited indications.

IcoSema + tesamorelin (off-label for visceral adiposity in non-HIV T2D + obesity context). Mechanism rationale: tesamorelin (GHRH analog) reduces visceral adipose tissue. Off-label use in T2D + visceral-adiposity-dominant phenotype: clinician judgment with informed-consent and primary-source-supported clinical reasoning. Pattern AA precision: tesamorelin’s FDA-approved-for-marketing-claims indication is HIV-associated lipodystrophy; non-HIV use is off-label.
IcoSema + CJC-1295 / Ipamorelin: rarely indicated. Lean-mass concern in IcoSema-treated patients is uncommon given the modest weight-effect direction. Older adults or specific clinical contexts (e.g., concurrent steroid therapy with iatrogenic muscle loss) may have selective lean-mass-stack indications; clinician judgment.

9.4 Contraindicated combinations

IcoSema + concurrent separate basal insulin (§6.10): redundant insulin replacement.
IcoSema + concurrent separate GLP-1 RA (§6.10): redundant GLP-1R agonism.
IcoSema + DPP-4 inhibitor (§6.10): mechanism-overlap.
IcoSema + concurrent oral semaglutide (Rybelsus) (§6.10): redundant semaglutide exposure.
IcoSema + sulfonylurea at full dose (relative): excess hypoglycemia risk; sulfonylurea should be reduced or discontinued at IcoSema initiation.

9.5 IcoSema vs separate-injection regimen — comparator framework

The IcoSema-vs-separate-injection-regimen comparison is the most operationally relevant clinical decision in the IcoSema patient pathway (more so than the within-Module-5 combinations above). Separate-injection regimens combine once-weekly Awiqli (insulin icodec, FDA-approved-for-marketing-claims since 2024-04-23 for T2D) with once-weekly Ozempic (semaglutide for T2D, FDA-approved-for-marketing-claims since 2017-12) as two separate injections. The decision is multi-dimensional (Pattern Z anchor 4 multi-dimensional fact presentation):

Injection burden. IcoSema = 1 weekly injection (52/year); separate-injection regimen = 2 weekly injections (104/year). For patients sensitive to injection burden, the burden-reduction benefit is real.
Component titration flexibility. IcoSema fixed-ratio eliminates independent titration; separate-injection retains independent component dosing. For patients whose insulin and GLP-1 RA requirements may evolve differently over time, the separate-injection flexibility is meaningful.
Cost and insurance coverage. Differs by jurisdiction and insurance plan. In the US (pre-FDA-approval state for IcoSema as of 2026-05-13), the separate-injection regimen is the de-facto available pathway. In the EU (post-EMA-approval), both regimens are available; cost differential depends on the specific national-formulary pricing.
Body-weight effect. IcoSema produces modest weight gain vs semaglutide-monotherapy (+4.54 kg vs semaglutide-monotherapy in COMBINE 2). Separate-injection regimen of Awiqli + Ozempic produces effects that depend on the independent doses of each component; clinician can titrate semaglutide to higher dose (2.0 mg, the SUSTAIN-FORTE-anchored T2D dose) if weight effect is a priority.
Regulatory state for US clinicians. IcoSema is not commercially available in the US as of 2026-05-13. Separate-injection Awiqli + Ozempic is FDA-approved-for-marketing-claims for T2D and is the available regimen. Pattern Z anchor 5 dominant: US clinicians counseling patients seeking the combined-mechanism approach present the separate-injection regimen as the available evidence-supported option, with explicit acknowledgment that the IcoSema fixed-ratio combination is pre-FDA-approval and not commercially available.

9.6 Worked example — IcoSema combination scenarios

Scenario A — IcoSema + metformin + SGLT2 inhibitor (canonical polycondition T2D regimen). A 61-year-old male with T2D + CKD eGFR 42, on metformin 2000 mg + empagliflozin 10 mg daily, switches from glargine to IcoSema (basal-insulin-experienced phenotype, COMBINE 1 anchor). Combination is well-supported (mechanism-complementary; no contraindicated interactions; FLOW/SELECT inheritance for renal/CV benefit at the semaglutide-component level). Monitoring per §5.3 (HbA1c, eGFR, UACR, weight, BP). Continue metformin and empagliflozin at current doses.

Scenario B — IcoSema + metformin (sulfonylurea discontinued at switch). A 56-year-old female with T2D, on metformin 2000 mg + glimepiride 4 mg + glargine 24 units daily, HbA1c 8.0%, switches to IcoSema (basal-insulin-experienced phenotype). At IcoSema initiation: glimepiride discontinued (additive hypoglycemia risk per §6.10); glargine discontinued (separate basal insulin redundant with IcoSema insulin-icodec component). Metformin continued. Monitoring per §5.3 with particular hypoglycemia-surveillance attention in the first 4–8 weeks post-glimepiride-discontinuation.

Scenario C — IcoSema + bolus insulin (off-label, research-state-evolving). A 64-year-old male with advanced T2D, on IcoSema at target dose for 6 months, has HbA1c 7.4% with target <7.0% but persistent postprandial glucose excursions on CGM. Counseling beat (Pattern Z anchor 5 off-label / extrapolation framing): “The EMA Kyinsu indication does not specifically address adding meal-time bolus insulin to IcoSema; the clinical pattern is research-state-evolving with analogous-situation extrapolation from daily combo products. Off-label use of meal-time insulin alongside IcoSema is clinician-judgment within informed-consent. Alternative options include transition to separate-injection regimen with independent insulin titration up, or transition to basal-bolus. Let’s discuss which option matters most for your situation.”

Scenario D — IcoSema vs separate-injection regimen (US clinician pre-FDA-approval context). A US T2D adult, basal-insulin-experienced (insulin glargine 28 units daily + metformin + empagliflozin), HbA1c 8.2%, interested in transitioning to a combined-mechanism regimen. Counseling beat (Pattern Z anchor 5 off-label / extrapolation framing dominant in US pre-FDA-approval state): “The IcoSema fixed-ratio combination is approved in the EU but not yet FDA-approved as of [2026-05-13]; it is not commercially available in the US. The clinical equivalent available in the US is the separate-injection regimen of Awiqli (insulin icodec, FDA-approved 2024 for T2D) plus Ozempic (semaglutide, FDA-approved 2017 for T2D). This requires two weekly injections instead of one, but allows independent titration of each component and is FDA-approved for marketing claims for T2D. The mechanism components are identical to IcoSema; the operational distinctions are injection burden (104 vs 52 injections/year) and titration flexibility (independent vs fixed-ratio). Let’s discuss which factors matter most for your situation, and the COMBINE program evidence base for the combination mechanism direction.”

Pattern W cross-check at §9.6. Each combination above reconciles with §2 (no contraindicated agent enters the combination), §6 (AE-management algorithms apply to each component), and §10 (counseling beats are Pattern Z-anchor-compliant). The IcoSema + DPP-4 inhibitor combination (contraindicated per §9.4) is not used in any of the scenarios.

10. Patient counseling beats (Pattern Z verbatim-anchor-compliant; Anchor 5 dominant)

10.1 Purpose

Define the IcoSema patient-counseling content — the conversations the clinician has with the patient at each protocol phase. §10 is the operational anchor for Pattern Z (cumulative-tone calibration) and for Pattern R (lead-framing calibration); the counseling beats are the most reader-facing content the protocol produces and the most vulnerable to drift from “presenting facts” to “steering decisions.”

IcoSema-specific calibration adjustments to the five Pattern Z anchors:

Anchor 5 (off-label / extrapolation transparency) is DOMINANT for IcoSema in the US pre-FDA-approval state. US clinicians counseling patients about IcoSema operate in pre-FDA-approval context; the canonical counseling pattern is “the combination is approved in the EU but not FDA-approved; the available US pathway is the separate-injection regimen of Awiqli + Ozempic.” This is Anchor 5 framing.
Anchor 2 (compounded vs FDA-approved) is N/A for the IcoSema combination product pre-approval. No compounded combo-product market exists for an unapproved combo as of 2026-05-13. (Compounded semaglutide and compounded insulin icodec exist as standalone-component compounded markets in some jurisdictions; that is component-level, not combo-product-level.) The Anchor 2 framing structure (lead with what the option IS; framing operational differences as factual scope) is preserved in the Anchor 5 application — the available US pathway IS the separate-injection regimen of FDA-approved components.
Anchor 4 (multi-dimensional comparator framing) is load-bearing for the IcoSema-vs-separate-injection-regimen decision (§9.5 above) and for the IcoSema-vs-basal-bolus and IcoSema-vs-semaglutide-monotherapy decisions (§7 non-response algorithm).
Anchor 3 (pregnancy-planning research-state-leading) applies normally with the Parker 2025 (PMID 40329607) human-data lead; the IcoSema-specific adjustment is the two-component half-life arithmetic (~35–40 days for both components together) and the transition-to-pregnancy-appropriate-insulin plan during the washout window.
Anchor 1 (lead with what the option IS) applies normally with IcoSema-specific calibration for hypoglycemia management — lead with what hypoglycemia management practice IS (a routine clinical skill for insulin-treated T2D patients), not with the prohibition framing.

10.2 Initiation conversation (§4 anchor)

The initiation conversation occurs at the pre-treatment workup completion / §4 initiation visit. Required counseling beats:

What IcoSema is (Anchor 1): name, mechanism (combined basal insulin + GLP-1 RA in single weekly injection), what indication it’s being used for (T2D glycemic control where current therapy is inadequate).
What it does for the patient’s indication (Anchor 1, Anchor 4): expected effect-size with COMBINE program anchor precision per the sub-population (HbA1c reduction approximately X percentage points per the COMBINE 1/2/3 effect-size anchor relevant to the patient’s phenotype); body-weight effect direction per the comparator-class context; hypoglycemia risk profile vs the comparator regimen.
The titration schedule (§4): standard pace and slow-titration option; the patient is informed that the titration timeline is adjustable to their tolerability and hypoglycemia-pattern.
The AE profile expected at each titration step (§6.3 GI class + §6.2 hypoglycemia class anticipatory framing): nausea is anticipated, typically attenuates, management strategies. Hypoglycemia-surveillance establishment per §4.5 — the IcoSema-specific dual-discipline at initiation.
The pre-conception planning beat for reproductive-age patients (Anchor 3): Parker 2025 human pregnancy-exposure data lead; pharmacokinetic washout arithmetic; transition-to-pregnancy-appropriate-insulin plan; re-initiation pathway.
Cost / access realities (Anchor 5 dominant for US pre-FDA-approval state): present the IcoSema-vs-separate-injection-regimen pathway distinction; present insurance-coverage realities if known.
What the patient signals back if any concern emerges (escalation pathway): severe persistent abdominal pain, vomiting, acute vision change, severe or recurrent hypoglycemia, signs of severe AE.

10.3 Compounded-vs-FDA-approved counseling — N/A for IcoSema combination product pre-approval

Pattern Z anchor 2 framing is N/A for IcoSema combination product as of 2026-05-13. No compounded combo-product market exists for an unapproved combo. The compounded semaglutide market and the compounded basal-insulin markets are component-level, not combo-product-level. Clinicians and patients in the US do not have a compounded-IcoSema option to compare against; the relevant comparator framing for the US patient pathway is the separate-injection regimen of FDA-approved components (Awiqli + Ozempic), addressed per Anchor 5 (§10.6 below) and Anchor 4 (§10.5 below).

For EU clinicians where Kyinsu is commercially available: compounded combo-product market for IcoSema-equivalent fixed-ratio combinations is also research-state-incomplete. The Kyinsu commercial product is the standard pathway; compounded alternatives are not characterized in the literature.

10.4 Pregnancy-planning conversation (Anchor 3)

For reproductive-age patients on IcoSema. The conversation leads with human pregnancy-exposure research-state data, then pharmacokinetic facts, then label recommendation, then the transition-to-pregnancy-appropriate-insulin plan. Required counseling beats:

Human pregnancy-exposure data — Parker 2025 (PMID 40329607). Pooled review of unplanned pregnancies from FDA- and EMA-submitted GLP-1 RA regulatory clinical trials (semaglutide, liraglutide, dulaglutide, class-related compounds). Incidence of congenital abnormalities appears relatively low in this pooled dataset; sample size limited; exposures from unplanned-pregnancy contexts. Prospective planned-pregnancy exposure data is research-state-incomplete.
Pharmacokinetic facts (post-research-state-lead). Semaglutide-component half-life approximately 1 week. Insulin icodec component half-life approximately 1 week. Approximately 5 half-lives (~35–40 days) for >95% pharmacokinetic clearance of both components.
Label recommendation. EMA Kyinsu label recommends discontinuation prior to planned pregnancy with appropriate washout (refer to current label for precise interval). The conservative 8-week pre-conception window inherits from the semaglutide-component label.
Transition to pregnancy-appropriate insulin regimen. During the washout window (approximately 8 weeks pre-conception) and through pregnancy and lactation, the standard regimen is NPH or pregnancy-appropriate basal-analog with bolus insulin per ADA Standards of Care for pregnancy in T2D.
Animal data context. Reproductive toxicity studies in animals supported the labeled contraindication. Parker 2025 human pooled data is the load-bearing human evidence.
The patient’s reproductive-planning decision is patient-anchored. Some patients will plan conception within the next year; some not for years. The counseling beat presents the facts; the timing decision is patient-anchored.
Re-initiation pathway. Post-pregnancy and post-lactation, re-initiation per §8.6 if indication and selection criteria are met.

10.5 Comparator conversation (Anchor 4) — IcoSema vs separate-injection regimen vs semaglutide-monotherapy vs basal-bolus

Canonical Anchor 4 requires multi-dimensional fact presentation across 8+ dimensions. For IcoSema’s principal comparators (separate-injection Awiqli + Ozempic; semaglutide-monotherapy; basal-bolus regimens), the dimensions are:

Glycemic-control effect (Pattern V trial-anchored). IcoSema vs icodec-alone: COMBINE 1 between-group treatment difference −0.66% HbA1c favoring IcoSema (PMID 40482671). IcoSema vs semaglutide-1.0-mg-alone: COMBINE 2 between-group treatment difference −0.45% HbA1c favoring IcoSema (PMID 39820580). IcoSema vs basal-bolus: COMBINE 3 between-group treatment difference −0.06% non-inferiority (PMID 40482670). The separate-injection regimen (Awiqli + Ozempic at independently-titrated doses) has no direct head-to-head IcoSema comparison; mechanism-equivalent components support directional similarity.
Body-weight effect (Pattern V trial-anchored). IcoSema vs semaglutide-1.0-mg-alone: COMBINE 2 +4.54 kg treatment difference favoring semaglutide-monotherapy on weight axis (PMID 39820580). IcoSema vs basal-bolus: COMBINE 3 −6.72 kg favoring IcoSema (PMID 40482670). Patient-counseling beat: if continued weight loss is a priority, semaglutide-monotherapy (or higher-dose Wegovy / Wegovy HD 7.2 mg) is better aligned than the combination.
Hypoglycemia rate (Pattern V trial-anchored). IcoSema vs icodec-alone: COMBINE 1 rate ratio 0.22 favoring IcoSema. IcoSema vs semaglutide-1.0-mg-alone: COMBINE 2 rate ratio 1.20 non-significant. IcoSema vs basal-bolus: COMBINE 3 rate ratio 0.12 favoring IcoSema.
Injection burden. IcoSema = 1/week (52/year). Separate-injection Awiqli + Ozempic = 2/week (104/year). Daily combos (Xultophy / Soliqua) = 7/week (365/year). Basal-bolus = 4+/day (1,460+/year for 4× daily).
Component titration flexibility. IcoSema fixed-ratio constrains independent titration; separate-injection regimen retains independent component dosing; daily combos retain daily titration but constrain to component-pair ratio; basal-bolus has maximum titration flexibility.
Indication scope. IcoSema (EMA Kyinsu): T2D inadequately controlled on basal insulin OR GLP-1 RA. Semaglutide-monotherapy: T2D, CWM, CV risk reduction (SELECT BMI ≥27 + CVD), MASH F2/F3, CKD-in-T2D (multiple FDA-approved indications). Basal-bolus: T2D advanced (no specific indication-label expansion). Separate-injection Awiqli + Ozempic: T2D per individual component FDA labels.
Regulatory state for US clinicians. IcoSema is pre-FDA-approval-for-marketing-claims as of 2026-05-13; not commercially available in the US. Separate-injection Awiqli + Ozempic: both FDA-approved-for-marketing-claims for T2D (Awiqli 2024-04-23; Ozempic 2017-12). Semaglutide-monotherapy: FDA-approved-for-marketing-claims across multiple indications. Basal-bolus: FDA-approved-for-marketing-claims components are routinely available.
Cost and access. Patient-specific; insurance-coverage realities vary by jurisdiction and indication.
Comparator framing precision (Pattern V). Effect-size differentials are anchored to specific COMBINE trials and specific enrollments. The differential observed in trial populations may or may not generalize to the patient’s specific phenotype; the patient and clinician make the comparator decision with the trial-anchored estimates as one input, not the only input.

Patient-counseling beat (verbatim Pattern Z anchor 4 calibration, IcoSema-specific): “There are several evidence-based options for T2D requiring both basal insulin and GLP-1 RA effect. IcoSema offers a single weekly injection with the convenience of co-formulation; the COMBINE program showed IcoSema’s safety and efficacy in T2D inadequately controlled on either component alone. The tradeoffs are that we can’t titrate the components separately, and the body-weight effect is more modest than semaglutide-monotherapy. Separate-injection regimens of Awiqli plus Ozempic offer independent titration flexibility — two injections per week instead of one, but the components dose independently and both are FDA-approved for T2D. Daily combinations like Xultophy require daily injections with daily-frequency dose-adjustment flexibility. Basal-bolus involves four or more daily injections with substantial hypoglycemia risk. Let’s discuss which factors matter most for your situation — convenience, flexibility, weight effects, hypoglycemia tolerance, cost, and clinical priorities.”

10.6 Off-label / extrapolation conversation (Anchor 5) — DOMINANT for IcoSema US pre-FDA-approval state

For US clinicians and patients considering IcoSema or its mechanism-equivalent approach in the pre-FDA-approval state, Anchor 5 framing is dominant. Required counseling beats:

Explicit pre-FDA-approval framing. “The IcoSema fixed-ratio combination is approved in the EU under the Kyinsu brand as of November 2025 but is not FDA-approved for marketing claims in the US as of 2026-05-13. The FDA submission status is not publicly disclosed.”
Primary-source rationale for the combination mechanism approach. COMBINE 1/2/3 Phase 3 trial data anchor the combined-mechanism evidence base (Mathieu PMID 40482671; Lingvay PMID 39820580; Billings PMID 40482670). The mechanism-equivalent US-available pathway is the separate-injection regimen of Awiqli + Ozempic, both FDA-approved for T2D.
Off-label combination considerations within the US pathway. Awiqli and Ozempic as separate-injection co-administered weekly are not off-label individually (both are FDA-approved for T2D at component level); the co-administration regimen is operationally a combined-mechanism approach but uses FDA-approved components within their labeled indications. This is distinct from the IcoSema combination product, which is a separate regulatory entity in the EU (Kyinsu) and pre-FDA-approval in the US.
Informed-consent acknowledgement. Patient understands the US regulatory state (separate-injection regimen of FDA-approved components vs the IcoSema fixed-ratio combination product is pre-FDA-approval) and consents to clinician judgment within informed-consent standards.
Re-evaluation if FDA approval emerges. If FDA approval for the IcoSema combination product is granted subsequently, the use transitions from pre-FDA-approval to label-supported; the counseling beat re-anchors at that transition.
Anchor 5 mirror — off-label use of IcoSema for non-T2D weight management (patient-question context). Patients sometimes ask whether IcoSema could be used for weight management without T2D. EU clinicians: the EMA Kyinsu indication is T2D specifically. The IcoSema combination produces less weight loss than semaglutide-monotherapy at population level (COMBINE 2 +0.84 kg vs −3.70 kg) given the insulin-component anabolic offset; mechanism prediction suggests less weight loss than semaglutide-monotherapy. Off-label use of IcoSema for non-T2D weight management is not evidence-supported and is mechanism-counter-indicated. The patient-counseling beat: “For weight management without T2D, the evidence base for IcoSema is research-state-incomplete and the mechanism prediction actually suggests less weight loss than semaglutide-monotherapy would produce. The EU-approved indication is T2D glycemic control, not weight management. Semaglutide as Wegovy 2.4 mg or higher-dose options (Wegovy HD 7.2 mg per STEP UP) or tirzepatide (Zepbound) are the evidence-based pharmacotherapy for chronic weight management — let’s discuss those options if weight management is your goal.”

Patient-counseling beat (verbatim Pattern Z anchor 5 calibration, IcoSema US pre-FDA-approval state): “The IcoSema combination — basal insulin plus GLP-1 RA in a single weekly injection — is approved in Europe but not yet FDA-approved here in the US as of 2026-05-13. The FDA submission status isn’t publicly disclosed. What’s available in the US today is the separate-injection version of the same mechanism approach: Awiqli (weekly insulin icodec) plus Ozempic (weekly semaglutide), both FDA-approved for T2D and prescribed together at independently-titrated doses. The COMBINE Phase 3 program — Mathieu, Lingvay, Billings 2025 — showed the fixed-ratio combination’s safety and efficacy in T2D inadequately controlled on either component alone, against four different comparators. The mechanism components are the same in the separate-injection approach. The trade-off is two injections per week instead of one, with the upside of independent titration flexibility. We can pursue the separate-injection approach now; if IcoSema becomes FDA-approved later and convenience matters to you, we can revisit. Let’s discuss what matters most for your situation.”

10.7 Discontinuation conversation (§8 anchor)

For patient-preference, indication-resolution, or AE-driven discontinuation. Required counseling beats:

Reason for discontinuation framing. Patient-preference (legitimate; protocol’s role is to inform, not override); indication-resolution (rare in T2D; framing is non-pathologizing); AE-attribution (clinical decision; protocol-mandated where applicable).
Transition plan rather than taper to nothing. For IcoSema in T2D context, the typical pattern is transition to alternative regimen (separate-injection regimen; basal-bolus; non-insulin regimen if HbA1c control allows) per §8.3.
Post-discontinuation expectations (per §8.5): glycemic-trajectory drift upward over 1–3 months as both components clear; body-weight axis non-dominant for IcoSema (unlike Wegovy CWM context); hypoglycemia trajectory dependent on replacement regimen.
Re-initiation pathway (per §8.6): titration restarts at conservative starting dose, not at prior maintenance dose.

10.8 Pattern Z self-audit on §10 counseling beats

The five Anchors above are the self-audit checklist for §10 counseling-beat language. An IcoSema protocol’s §10 is Pattern-Z-violation-positive if any of the following appear:

IcoSema framed as “experimental” / “highly speculative” / “fringe” (Pattern Z.research-precision violation — IcoSema has a verified Phase 3 evidence base with three published primary RCTs and is EMA-approved-for-marketing-claims).
Pre-FDA-approval framing operating as dismissal rather than as factual regulatory-state precision (Anchor 5 violation; Pattern AA.marketing-claims cross-fail).
IcoSema recommended off-label for non-T2D weight management (mechanism-counter-indicated; Pattern V direction-of-effect cross-fail; the mechanism prediction is less weight loss than semaglutide-monotherapy).
Hypoglycemia framed with prohibition language vs Pattern Z anchor 1 calibration (routine clinical skill for insulin-treated patients) — bias-vocabulary like “scary” / “daunting” / “major barrier” applied to insulin-injection or hypoglycemia (Pattern Z.injection-framing / Pattern Z.research-precision violation).
Pregnancy-planning leading with regulatory contraindication framing vs Parker 2025 human-data lead (Anchor 3 violation).
Comparator framing with single-dimension comparison (e.g., weight magnitude alone) rather than multi-dimensional Anchor 4 calibration.

The §10 production agent runs the five-Anchor self-audit before handoff to the verification cycle (Appendix A).

11. Source citations

11.1 Purpose

Define the bibliography format, the evidence-hierarchy tiers, and the PMID-anchor / NCT-anchor / DOI-anchor identifier-integrity discipline. §11 is the protocol’s evidentiary spine — every effect-size claim, every dose threshold, every contraindication, every counseling-beat fact-anchor traces to a §11 citation entry. Pattern AB.1 / AB.4 standing scans operate at this section.

11.2 Bibliography format

Each citation entry contains: first author last name et al; title; journal year volume pages; PMID; DOI if available; NCT for trial reports; effect-size summary; citation context (which protocol sections cite this source).

11.3 Evidence hierarchy tiers

Tier 1 — pivotal Phase 3 RCT supporting the EMA Kyinsu approval. COMBINE 1, COMBINE 2, COMBINE 3 (peer-reviewed primary publications); COMBINE 4 (pending primary publication).
Tier 1.5 — Phase 3 head-to-head RCT within the comparator-class context. Comparator-class context inherits from semaglutide-component head-to-head trials (SURPASS-2; SURMOUNT-5) per the IcoSema canonical’s comparator framing.
Tier 2 — supporting Phase 3 evidence at component level (ONWARDS program for insulin icodec; SUSTAIN/STEP/SELECT/FLOW/ESSENCE for semaglutide); peer-reviewed reviews and meta-analyses with appropriate dataset depth.
Tier 3 — post-marketing pharmacovigilance, observational cohort, retrospective analysis, regulatory documents.

11.4 Identifier-integrity discipline — Pattern AB.1 / AB.2 / AB.4

Every PMID, NCT, and DOI verified by content (not by existence). The IcoSema canonical’s Phase 4 AB-hygiene audit identified and resolved B-CRIT-2 (the COMBINE 1 PMID correction: PMID 40482671 is COMBINE 1, Mathieu et al Lancet Diabetes Endocrinol 2025 July; PMID 40347948 is not COMBINE 1 — that PMID indexes an unrelated developmental biology paper and must not appear in the IcoSema Bibliography). This protocol inherits the verified-identifier table from the IcoSema canonical Bibliography (/Process/IcoSema/IcoSema - Bibliography.md).

11.5 IcoSema protocol bibliography — load-bearing sources

Tier 1 — COMBINE Phase 3 program (IcoSema combination product anchors).

Mathieu C, et al. Once-weekly insulin icodec with semaglutide (IcoSema) compared with once-weekly insulin icodec alone in adults with type 2 diabetes (COMBINE 1): a randomised, open-label, parallel-group, phase 3 trial. Lancet Diabetes Endocrinol 2025 July. PMID 40482671. NCT05352815. COMBINE 1; basal-insulin-experienced T2D; n=1,291; 52 weeks; HbA1c between-group treatment difference −0.66% favoring IcoSema (95% CI −0.76 to −0.57; p<0.0001); hypoglycemia rate ratio 0.22 favoring IcoSema. Cited in §1.2, §1.5, §2.5, §4.7, §6.2, §7.4, §10.5, §11.3.
Lingvay I, et al. Efficacy and safety of once-weekly IcoSema versus once-weekly semaglutide in adults with type 2 diabetes inadequately controlled on a GLP-1 receptor agonist (COMBINE 2). Diabetologia 2025 April. PMID 39820580. NCT05259033. COMBINE 2; GLP-1 RA-experienced T2D; n=683; 52 weeks; HbA1c between-group treatment difference −0.45% favoring IcoSema (95% CI −6.13 to −3.57 mmol/mol; p<0.0001); body weight +4.54 kg treatment difference favoring semaglutide-monotherapy on weight axis; hypoglycemia rate ratio 1.20 (NS, p=0.66). Cited in §1.2, §1.5, §4.7, §6.2, §7.4, §10.5, §11.3.
Billings LK, et al. Once-weekly IcoSema versus basal-bolus insulin therapy in adults with type 2 diabetes (COMBINE 3): a randomised, open-label, phase 3 trial. Lancet Diabetes Endocrinol 2025 July. PMID 40482670. NCT05013229. COMBINE 3; basal-bolus-eligible T2D; n=679; 52 weeks; HbA1c treatment difference −0.06% non-inferiority (95% CI −0.22 to 0.09; p<0.0001 non-inferiority); body weight −6.72 kg favoring IcoSema; hypoglycemia rate ratio 0.12 favoring IcoSema; injection burden 52 vs 730+/year. Cited in §1.2, §1.5, §6.2, §7.4, §10.5, §11.3.
COMBINE 4 (NCT06269107) — IcoSema vs once-daily insulin glargine U100 in T2D inadequately controlled on OADs; 40-week study; OverallStatus = Completed per ClinicalTrials.gov v2 API. Primary publication status pending in PubMed corpus as of 2026-05-13. Cited in §1.2, §1.5 (with explicit pending-publication framing).

Tier 2 — supporting IcoSema-specific peer-reviewed analyses.

Kamrul-Hasan ABM, et al. IcoSema systematic review and meta-analysis. AACE Endocrinology and Diabetes 2025 August. PMID 41048697. Meta-analysis of 3 RCTs; n=2,653; pooled HbA1c reduction 0.39 percentage points favoring IcoSema vs control; reduced hypoglycemia risk vs insulin comparators; comparable weight and SAE rates. Pattern AB.4 precision: early-cycle SR + meta-analysis with limited dataset depth (3 RCTs); not characterized as a definitive class-level analysis.
Wang H, et al. Pharmacokinetic study of IcoSema combination in Chinese type 2 diabetes patients. Diabetes Therapy 2025 October. PMID 41051695. Crossover PK study in 20 Chinese T2D patients; insulin icodec PK unaffected by combination; semaglutide peak concentration higher and earlier with IcoSema combination vs separate-component administration; sample size n=20 limits population-level generalizability.
Bolli GB, et al. Fixed-ratio basal insulin + GLP-1 RA combination products: class review. Diabetes Obes Metab 2025. PMID 40678871. Class-level review covering Xultophy, Soliqua, IcoSema; comparator-class analytical work for §7 and §9.5 comparator framing.
Umpierrez GE, et al. Commentary on COMBINE 1 and COMBINE 3. Lancet Diabetes Endocrinol 2025 July. PMID 40482672. Editorial commentary published alongside COMBINE 1 + COMBINE 3 — editorial commentary publication type per Pattern AB.4 precision, not primary trial data.

Tier 2 — insulin icodec component (Awiqli) ONWARDS Phase 3 program — supporting basal-insulin component anchor.

Rosenstock J, et al. Once-Weekly Insulin Icodec (Awiqli) vs Once-Daily Insulin Glargine in Insulin-Naive Type 2 Diabetes (ONWARDS 1). N Engl J Med 2023. PMID 37356066. NCT04460885. Insulin-naïve T2D; standalone insulin icodec safety and efficacy anchor.
Philis-Tsimikas A, et al. ONWARDS 2: Once-Weekly Insulin Icodec vs Once-Daily Insulin Degludec in Basal-Insulin-Treated T2D. Lancet Diabetes Endocrinol 2023. PMID 37148899.
Lingvay I, et al. ONWARDS 3: Once-Weekly Insulin Icodec in Insulin-Naive T2D. JAMA 2023. PMID 37354562.
Mathieu C, et al. ONWARDS 4: Once-Weekly Insulin Icodec in Basal-Bolus T2D. Lancet 2023. PMID 37156252.
Bajaj HS, et al. ONWARDS 5: Once-Weekly Insulin Icodec with Dosing-Guide App vs Daily Basal Analogs. Annals Intern Med 2023. PMID 37748181.
Russell-Jones D, et al. ONWARDS 6: Once-Weekly Insulin Icodec Basal-Bolus vs Insulin Degludec in T1D. Lancet 2023. PMID 37863084.
Philis-Tsimikas A, et al. ONWARDS program design rationale. 2022. PMID 36106652. Program-level design and rationale.

Tier 2 — semaglutide component cross-canonical inheritance (semaglutide protocol Bibliography is the primary source).

Wilding JPH, et al. STEP-1; semaglutide CWM. N Engl J Med 2021. PMID 33567185. NCT03548935.
Marso SP, et al. SUSTAIN-6; semaglutide CV outcomes in T2D. N Engl J Med 2016. PMID 27633186. NCT01720446.
Lincoff AM, et al. SELECT; semaglutide CV outcomes in non-diabetic obesity. N Engl J Med 2023. PMID 37952131. NCT03574597.
Sanyal AJ, Newsome PN, et al. ESSENCE Phase 3 trial of semaglutide in MASH. N Engl J Med 2025. PMID 40305708. NCT04822181.
Perkovic V, et al. FLOW; semaglutide CKD-in-T2D outcomes. N Engl J Med 2024. PMID 38785209. NCT03819153.
Frías JP, et al. SURPASS-2; tirzepatide vs semaglutide T2D head-to-head. N Engl J Med 2021. PMID 34170647. NCT03987919. Within-class higher-effect comparator referenced in §7 non-response algorithm.

Tier 2 — class-level pancreatitis / safety inheritance.

Wen B, et al. Acute pancreatitis with GLP-1 receptor agonists: meta-analysis of 62 RCTs (n=66,232). Endocrinol Diabetes Metab 2025 October. PMID 40988099. Pooled RR 1.44 (95% CI 1.09–1.89, p=0.009) for acute pancreatitis; authors frame as “slightly increased risk, likely minimal.” Cited in §6.5.

Tier 3 — post-marketing pharmacovigilance and signals under evaluation (semaglutide-component-inherited NAION).

Hathaway JT, et al. NAION risk in patients prescribed semaglutide. JAMA Ophthalmol 2024. PMID 38958939. Single-center cohort; HR 4.28–7.64. Signal under evaluation per Pattern AA precision.
Grauslund J, et al. NAION nationwide Danish cohort. 2024. PMID 39696569. n=424,152; HR 2.19 at 5 years. Cited in §6.6 inheritance.
Lakhani A, et al. NAION observational pharmacovigilance 180-country study. 2025. PMID 40383360. FAERS ROR 11.12; observational pharmacovigilance, not meta-analysis per Pattern AB.4 precision. Tirzepatide signal absent at same threshold (class-differentiated to semaglutide).

Tier 3 — human pregnancy-exposure pooled data (Pattern Z anchor 3 calibration lead).

Parker SE, et al. Pooled review of unplanned pregnancies from FDA- and EMA-submitted GLP-1 RA regulatory clinical trials. Diabetes Obes Metab 2025. PMID 40329607. Incidence of congenital abnormalities appears relatively low in this pooled dataset; sample size limited; exposures from unplanned-pregnancy contexts. Load-bearing for §6.11 and §10.4 Anchor 3 lead.

Tier 3 — class cancer-context inheritance.

Ko EH, et al. GLP-1 RA cancer outcomes systematic review. Annals Intern Med 2026 February. PMID 41359966. n=94,245 across 48 RCTs; 11 cancer types + 2 conditions; moderate-certainty “little or no effect” on thyroid, pancreatic, breast, kidney cancers; low-certainty “may have little or no effect” on additional tumor types. Cross-canonical inheritance from semaglutide v1.0-final.

Regulatory reference (Tier 2 within Section 10 regulatory context per Editorial Framework §1.7).

European Medicines Agency. European Public Assessment Report (EPAR) for Kyinsu (IcoSema). CHMP positive opinion 18 September 2025; EC marketing authorization decision 24 November 2025. Source of the EMA-approved-for-marketing-claims indication scope per Pattern AA.marketing-claims.

11.6 Pattern AB.4 standing scan applied to §11.5

Every PMID and NCT in this Bibliography subset has been content-verified by abstract retrieval (PMID) or ClinicalTrials.gov entry retrieval (NCT) per the IcoSema canonical Phase 2 verification (2026-05-12) and Phase 4 AB-hygiene audit (B-CRIT-2 resolution: COMBINE 1 PMID 40482671 verified-correct; PMID 40347948 explicitly excluded as developmental biology paper, not COMBINE 1). The cascade-failure pattern (PMID misattribution; NCT mis-matching) is the failure mode this verification prevents.

12. Clinical decision tree

12.1 Purpose

Define a phenotype-guided decision tree that integrates §§1–11 into a single navigable clinical-workflow reference. §12 is the operational summary clinicians reach for at point-of-care. The decision tree summarizes the authoritative content in §§1–11; if the decision tree and a §§1–11 sub-block disagree, the §§1–11 content is canonical.

12.2 High-level decision flowchart (ASCII)

                    PATIENT PRESENTS (adult T2D context)
                                |
                                v
                  [§1] Indication scope check
                  - T2D inadequately controlled on basal insulin
                    OR GLP-1 RA monotherapy? (EMA Kyinsu indication)
                  - US clinician pre-FDA-approval: separate-injection
                    regimen (Awiqli + Ozempic) is the FDA-approved
                    pathway; IcoSema fixed-ratio combo product is
                    pre-FDA-approval
                                |
                                v
                  [§2] Selection criteria
                  - Inclusion (COMBINE-enrolled phenotype)?   --> NO  --> Out of protocol
                  - Relative exclusion?       --> Clinician judgment
                  - Hard contraindication?    --> YES --> Out of protocol
                                |
                                v (Inclusion met, no contraindication)
                  [§3] Pre-treatment workup
                  - Standard metabolic + diabetes-specific panels
                  - Organ-baseline (thyroid, pancreas, ophthalmology)
                  - CGM pre-treatment baseline (hypoglycemia surveillance)
                                |
                                v
                  [§4] Initiation
                  - Starting dose per sub-population (COMBINE 1/2/3/4)
                  - GI tolerability priming + hypoglycemia
                    surveillance establishment (dual discipline)
                  - Early monitoring cadence Week 2 / 4-6 / 8-12 / 16-20
                                |
                                v (Target dose attained)
                  [§5] Maintenance
                  - HbA1c quarterly during titration; semi-annually at maintenance
                  - Weight, BP every visit; CGM data review; hypoglycemia interview
                                |
                                v
                  [§6/7/8] Branch points
                  - AE emerges? --> §6 AE-class algorithm
                    (hypoglycemia DOMINANT clinical-management
                    AE class; §6.2)
                  - Plateau / non-response? --> §7 algorithm
                    (transition to separate-injection / basal-bolus /
                    tirzepatide / add SGLT2 inhibitor)
                  - Discontinuation trigger? --> §8 transition (not taper-to-nothing)
                                |
                                v
                  [§9] Combination decisions
                  - IcoSema + metformin + SGLT2 inhibitor (canonical
                    polycondition regimen)
                  - Contraindicated: separate basal insulin, separate
                    GLP-1 RA, DPP-4 inhibitor
                  - IcoSema vs separate-injection regimen comparator
                                |
                                v
                  [§10] Counseling beats
                  - Anchor 5 DOMINANT for US pre-FDA-approval state
                  - Anchor 4 multi-dimensional comparator framing
                  - Anchor 3 pregnancy-planning Parker 2025 lead
                  - Anchor 1 hypoglycemia as routine clinical skill
                                |
                                v
                  [§11] All claims source-anchored
                  - Tier 1 / 1.5 / 2 / 3 evidence hierarchy
                  - COMBINE 1 PMID 40482671 verified
                  - Pattern AB.4 identifier-integrity standing scan
                                |
                                v
                  [Appendix A] Verification gate
                  - 4-step cycle before clinician delivery

12.3 Phenotype-guided decision matrix

Phenotype dimension	Pattern	IcoSema fit	Within-class alternatives	Cross-class additions
Basal-insulin-experienced T2D, HbA1c above target	COMBINE 1-anchored	Strong fit; switch protocol per trial titration	Awiqli alone titrated up (less HbA1c efficacy, higher hypoglycemia); separate-injection Awiqli + Ozempic	Continue metformin + SGLT2 if applicable
GLP-1 RA-experienced T2D, HbA1c above target	COMBINE 2-anchored	Strong fit on glycemic axis; weight tradeoff (+4.54 kg vs semaglutide-monotherapy)	Tirzepatide-monotherapy if higher-effect or weight-priority; semaglutide up-titration to 2.0 mg (SUSTAIN-FORTE)	SGLT2 if not already on
Basal-bolus-eligible T2D, would otherwise transition to basal-bolus	COMBINE 3-anchored	Strong fit; substantial regimen-simplification + body-weight + hypoglycemia benefits	Basal-bolus (the comparator); separate-injection regimen	SGLT2 if not already on
Insulin-naïve T2D, OAD-inadequate, injectable initiation indicated	COMBINE 4-anchored (pending publication)	Plausible fit; pre-publication clinical use requires structured titration discipline	Awiqli alone; semaglutide-monotherapy; basal insulin alone (glargine, degludec); tirzepatide-monotherapy	Continue OAD background
T2D + obesity, weight loss is a priority	Not aligned	Modest weight gain vs semaglutide-monotherapy; not the canonical fit	Semaglutide-monotherapy (Wegovy 2.4 mg or Wegovy HD 7.2 mg per STEP UP); tirzepatide (Zepbound)	Behavioral / activity intensification
T2D + CKD eGFR 25–75	Polycondition canonical	Strong fit; semaglutide-component FLOW inheritance	Semaglutide-monotherapy (FLOW dose) + separate basal insulin if needed	SGLT2 inhibitor (DAPA-CKD / EMPA-KIDNEY) additive
T2D + ASCVD	Polycondition; CV inheritance from semaglutide-component	Reasonable fit; semaglutide-component SELECT inheritance for CV benefit	Semaglutide-monotherapy SELECT-anchored; dulaglutide REWIND; liraglutide LEADER	SGLT2 for additive CV benefit
T2D adolescent	Not registered	Not applicable; IcoSema is adult-only	Semaglutide-monotherapy Ozempic 2 mg pediatric (≥10 years); liraglutide Victoza pediatric (≥10 years)	Behavioral / family-anchored intervention
LADA suspected	Pattern V risk	Not applicable; not first-line in autoimmune diabetes	Standard autoimmune-diabetes management (basal-bolus, CSII)	C-peptide and antibody assessment per §2.3
Severe gastroparesis	Relative exclusion	Caution / clinician judgment	Non-GLP-1 alternative (basal insulin alone; SGLT2; basal-bolus)	Behavioral / surgical bariatric pathway
MTC / MEN-2 personal/family hx	Hard contraindication	OUT OF PROTOCOL — boxed warning class-wide (inherited via semaglutide-component)	Non-GLP-1 alternative	Per indication; non-GLP-1 T2D management
Pre-conception planning	Anchor 3 framing	IcoSema with 8-week pre-conception washout + transition to pregnancy-appropriate insulin	Same arithmetic for all GLP-1 RA-containing regimens	Pre-pregnancy counseling pathway
Reproductive-age + active pregnancy	Pregnancy = contraindication	Immediate discontinuation; transition to NPH or pregnancy-appropriate insulin	Same — all GLP-1 RA-containing regimens pregnancy-contraindicated	Pre-pregnancy / post-pregnancy reinitiation per §8.6
Older adult (≥65)	Hypoglycemia surveillance load-bearing	Conservative titration; lower HbA1c-target individualization	Same considerations for all insulin-containing regimens	CGM if available
Hypoglycemia-unawareness	Relative exclusion	CGM-mediated surveillance; counseling; clinician judgment	Non-insulin alternative regimens if hypoglycemia-unawareness severe	CGM is uniformly load-bearing
US pre-FDA-approval context	Anchor 5 dominant	Not commercially available US; separate-injection Awiqli + Ozempic is available alternative	Separate-injection regimen (FDA-approved components)	Anchor 5 counseling beat per §10.6
Prior IcoSema non-response at target dose	§7 algorithm	Out of IcoSema; transition	Separate-injection regimen with independent insulin up-titration; basal-bolus; tirzepatide	Add SGLT2 inhibitor if not on

12.4 Worked example — IcoSema phenotype-guided decision tree application

A 67-year-old male presents with T2D (12 years), BMI 31, HbA1c 8.4% with target 7.0%, current regimen: metformin 2000 mg + empagliflozin 10 mg + insulin glargine U100 38 units daily. Documented ASCVD (prior CABG 5 years ago, on aspirin and atorvastatin). UACR 220 mg/g, eGFR 58 (CKD stage 3a). No MTC / MEN-2 family history. No pancreatitis history. Mild non-proliferative DR on annual dilated exam 4 months ago. No hypoglycemia events on current regimen in the last 12 months. EU patient (Kyinsu commercially available).

Decision-tree walkthrough.

Step 1 — Indication scope. T2D inadequately controlled on basal insulin (glargine) — EMA Kyinsu indication is satisfied. Polycondition phenotype: T2D + ASCVD + CKD + obesity.

Step 2 — Selection criteria. Inclusion criteria met (basal-insulin-experienced, COMBINE 1 phenotype). No hard contraindications. Relative exclusion: mild DR is documented (counseling on rapid-HbA1c-improvement context per §6.8); eGFR 58 is within COMBINE 1 enrollment range; older adult phenotype warrants hypoglycemia-surveillance attention per §5.4.

Step 3 — Pre-treatment workup. Standard metabolic + diabetes-specific panel + organ-baseline (§3.7). Update dilated retinal exam given baseline mild DR + anticipated rapid HbA1c improvement. CGM started 2 weeks before IcoSema initiation. UACR and eGFR documented; RAS blockade status confirmed (patient on lisinopril at maximally tolerated dose).

Step 4 — Initiation. Glargine discontinued at IcoSema Week 0. Empagliflozin continued. Metformin continued. IcoSema starting dose per COMBINE 1 trial titration scheme. Counseling at initiation visit per §10.2 (initiation conversation): IcoSema mechanism, expected HbA1c trajectory per COMBINE 1 anchor (HbA1c reduction approximately 1.55 percentage points absolute / between-group treatment difference −0.66% vs continued glargine-alone at 52 weeks), expected GI AE profile per §6.3, hypoglycemia-surveillance establishment per §4.5 (load-bearing given older adult phenotype), escalation pathway. Pattern Z anchor 1 calibration applied to hypoglycemia management framing.

Step 5 — Maintenance. Target dose attained Week 20. HbA1c at Months 4, 7, 10, 13 in Y1. UACR and eGFR annually. Lipid panel annually. Weight + BP every visit. CGM data review every visit.

Step 6 — Branch points. Anticipated GI AE profile manage as standard. Non-response algorithm (§7) not anticipated given polycondition trial-program support; if HbA1c not at target at Month 6, consider dose adjustment or transition to separate-injection regimen for independent insulin up-titration.

Step 7 — Combination decisions. Continue metformin + empagliflozin (canonical polycondition T2D + ASCVD + CKD regimen). Continue RAS blockade (lisinopril). Continue aspirin and atorvastatin (CV co-management). Lean-mass stack not indicated.

Step 8 — Counseling beats. Initiation conversation per §10.2. Anchor 4 multi-dimensional comparator framing presented at initiation (the patient was offered IcoSema vs continued glargine + future addition of separate Ozempic; chose IcoSema for the single-injection convenience). Anchor 3 not applicable (no pregnancy planning context). Anchor 5 not dominant for this EU patient (Kyinsu is commercially available in this jurisdiction). Anchor 1 hypoglycemia framing applied at initiation given older adult phenotype.

Step 9 — Source citations. All claims above traced to §11.5 Bibliography: COMBINE 1 (PMID 40482671) — primary effect-size anchor for the basal-insulin-experienced sub-population; SELECT (PMID 37952131) — semaglutide-component CV inheritance; FLOW (PMID 38785209) — semaglutide-component kidney inheritance; ONWARDS 1 (PMID 37356066) — insulin icodec component standalone basal-insulin anchor.

Step 10 — Verification gate. Per Appendix A, this protocol-application case passes through the four-step cycle if formalized as a case write-up.

Pattern Z calibration self-audit at §12.4. The decision-tree walkthrough presents the patient’s options factually with trial-anchored effect-size estimates; the polycondition phenotype is anchored to the COMBINE 1 sub-population plus semaglutide-component inheritance from FLOW and SELECT without inflating combo-product-specific CV / renal outcomes (which are research-state-incomplete per §1 OUT-OF-SCOPE framing); the combination is presented as standard polycondition regimen (Pattern Z compliant — what the combination IS for this phenotype); off-label / extrapolation transparency is not dominant for this EU patient (Kyinsu commercially available; combination is on-label for this T2D + comorbidity-load patient).

Appendix A. Verification gate — four-step cycle

The IcoSema Protocol passes through the four-step verification cycle per the template Appendix A discipline.

A.1 Step 1 — Production agent drafts

The production agent applies the twelve-section structure with Pattern R / R.1 / R.2 framing discipline at section-architecture-design step, Pattern V direction-of-effect anchoring (combo-product hypoglycemia rate ratios are comparator-specific: 0.22 favoring IcoSema vs icodec-alone; 1.20 non-significant vs semaglutide-monotherapy; 0.12 favoring IcoSema vs basal-bolus), Pattern AA.marketing-claims regulatory-state precision (EMA-approved-for-marketing-claims as Kyinsu since 24 November 2025; pre-FDA-approval-for-marketing-claims as of 2026-05-13), Pattern AB.1 identifier integrity at draft step (COMBINE 1 PMID 40482671 verified; PMID 40347948 explicitly excluded). §10 self-audit against the five Pattern Z calibration anchors complete with IcoSema-specific calibration (Anchor 5 dominant for US pre-FDA-approval state; Anchor 2 N/A pre-approval).

A.2 Step 2 — Primary-Source-Verification-Agent iteration 1

The PSV agent retrieves PMIDs in §11 Bibliography and verifies content matches (Mathieu et al Lancet Diabetes Endocrinol 2025 July at PMID 40482671 = COMBINE 1; Lingvay et al Diabetologia 2025 April at PMID 39820580 = COMBINE 2; Billings et al Lancet Diabetes Endocrinol 2025 July at PMID 40482670 = COMBINE 3). Retrieves NCTs at ClinicalTrials.gov v2 API and verifies sponsor / intervention / indication / phase / status (NCT05352815 = COMBINE 1; NCT05259033 = COMBINE 2; NCT05013229 = COMBINE 3; NCT06269107 = COMBINE 4 OverallStatus = Completed, primary publication pending). Runs Pattern AB.4 standing scan for cascade locations — the B-CRIT-2 PMID 40347948 false-attribution-as-COMBINE-1 was resolved in the canonical Phase 4 AB-hygiene audit and is not present in this protocol.

A.3 Step 3 — Independent-Adversarial-Reviewer-Agent iteration 1

The IAR agent six-axis scrutiny:

Framing. §1 opens with what IcoSema does for whom (combined-mechanism T2D therapy where current therapy is inadequate; EMA-approved indication), not with what it doesn’t do (chronic weight management is explicitly OUT OF SCOPE; CV / renal outcomes are research-state-incomplete for the combo product). §6.2 hypoglycemia opens with anticipatory framing (what hypoglycemia management practice IS, per Pattern Z anchor 1 calibration). §10 carries Pattern Z verbatim-anchor compliance with IcoSema-specific calibration.
Completeness. All twelve sections substantively populated. Worked examples present at §1.5, §2.5, §3.9, §4.7, §5.5, §6.12, §7.4, §8.7, §9.6, §12.4. All required cross-references present.
Consistency (Pattern W). Structural-count claims consistent end-to-end: four COMBINE sub-populations (§1.2), four sub-population dosing schemes (§4.2), four sub-population inclusion criteria (§2.2), four sub-population worked examples (§3.9 + §4.7 + §5.5). §3 panel inventory reconciles with §5 monitoring intervals and §6 AE-trigger labs.
Direction-of-effect (Pattern V). Hypoglycemia rate ratios are comparator-specific (0.22 vs icodec-alone; 1.20 NS vs semaglutide-monotherapy; 0.12 vs basal-bolus) and are not generalized across comparators — Pattern V compliant. Body-weight effect direction is comparator-specific (+4.54 kg vs semaglutide-monotherapy on weight axis; −6.72 kg vs basal-bolus) and is not generalized. Pattern V.metric-axis disclosure: per-arm absolute changes and between-group treatment differences are labeled explicitly. Pattern V.estimand-axis: the COMBINE 1 dual-estimand framing for HbA1c is per-trial-protocol; the canonical IcoSema-protocol presentation uses the between-group treatment difference as the primary anchor with per-arm changes disclosed.
Regulatory precision (Pattern AA.marketing-claims). EMA-approved-for-marketing-claims (Kyinsu, EC decision 24 November 2025) for the IcoSema combination product. Pre-FDA-approval-for-marketing-claims as of 2026-05-13. Component-level: insulin icodec FDA-approved as Awiqli 2024-04-23 for T2D; semaglutide FDA-approved across multiple indications. The distinction between combo-product regulatory state and component regulatory state is preserved throughout.
Identifier integrity (Pattern AB.1 / AB.2 / AB.4 standing scans). COMBINE 1 PMID 40482671 verified-correct; PMID 40347948 explicitly excluded (B-CRIT-2 resolution from canonical Phase 4 AB-hygiene audit). NCT05352815 = COMBINE 1. All component-level PMIDs (ONWARDS, SUSTAIN, STEP, SELECT, FLOW, ESSENCE) inherit verified state from the semaglutide canonical Bibliography.

A.4 Step 4 — Dr. Gross verification gate

Dr. Gross applies the final clinical-judgment review. Items for his attention include:

Hypoglycemia AE class positioning at §6.2 — the IcoSema-specific structural decision to position hypoglycemia as the dominant clinical-management AE class (immediately after acute safety overview) reflecting actual clinical priority for an insulin-containing combo product. Does the §6.2 anticipatory framing match how an experienced clinician would actually counsel an insulin-naïve or basal-insulin-experienced patient on the hypoglycemia-surveillance discipline at IcoSema initiation?
Anchor 5 dominance in §10 for US pre-FDA-approval state — does the §10.6 off-label / extrapolation framing accurately reflect what US clinicians need to communicate to patients pre-FDA-approval? Specifically: does the framing “the separate-injection regimen of Awiqli + Ozempic is the available US pathway” capture the clinical reality, or does it understate / overstate the access reality?
Pattern Z anchor 2 N/A determination — confirm that no compounded IcoSema-equivalent combo-product market exists pre-FDA-approval, and that the protocol’s Anchor 2 N/A treatment is correct as of 2026-05-13.
§7.3 non-response branches — the transition-to-separate-injection-regimen branch (5a) is a load-bearing IcoSema-specific clinical decision that the COMBINE program did not directly test but that the fixed-ratio constraint logically requires. Does the §7 framing accurately reflect the practice pattern?
§8.3 transition-vs-taper framing — the IcoSema-specific discontinuation discipline (transition to alternative regimen vs taper to nothing) reflects the T2D-pathophysiology-persists framing. Does this match how clinicians actually approach IcoSema discontinuation?
§9.2 IcoSema + bolus insulin off-label combination — research-state-evolving classification with §10.6 Anchor 5 framing applied. Does this match the practice-pattern reality?
§12.3 phenotype matrix completeness — does the matrix cover the clinically relevant phenotypes a clinician would actually navigate at point-of-care, or are there phenotypes missing?

Appendix B. Pattern discipline self-audit (this protocol)

B.1 Pattern R / R.1 / R.2

§1 opens with what IcoSema does for whom (EMA-approved indication for T2D inadequately controlled on basal insulin OR GLP-1 RA monotherapy; combined-mechanism therapy). §2 opens with inclusion criteria (the COMBINE-enrolled phenotype) before relative exclusions and hard contraindications. §6.2 hypoglycemia opens with anticipatory framing — what hypoglycemia management practice IS — before management algorithms. §7 opens with diagnostic distinctions (pseudo-plateau vs true plateau vs non-response). §8 opens with discontinuation triggers framed as legitimate clinical pathway, with the transition-to-alternative-regimen framing as the typical T2D-context pattern. Section ordering is design-time-locked; framing discipline applied at design step.

B.2 Pattern V — direction-of-effect verification

Every effect-size claim has primary-source anchor and population qualification. Hypoglycemia rate ratios are comparator-specific (0.22 favoring IcoSema vs icodec-alone in COMBINE 1; 1.20 non-significant vs semaglutide-monotherapy in COMBINE 2; 0.12 favoring IcoSema vs basal-bolus in COMBINE 3) and are not generalized across comparators. Body-weight effect direction is comparator-specific (+4.54 kg vs semaglutide-monotherapy on weight axis; −6.72 kg vs basal-bolus) and is not generalized.

Pattern V.metric-axis disclosure. Per-arm absolute changes (COMBINE 1: IcoSema −1.55 percentage points vs icodec −0.89 percentage points HbA1c) and between-group treatment differences (COMBINE 1: −0.66%, 95% CI −0.76 to −0.57; p<0.0001 for superiority) are labeled explicitly. The protocol cites the between-group treatment difference as the primary anchor with per-arm changes disclosed where load-bearing.

Pattern V.estimand-axis disclosure. The COMBINE 1 dual-estimand framing for HbA1c + weight endpoints is per-trial-protocol; this protocol presents the trial-anchored between-group treatment difference as the primary anchor and acknowledges the estimand-axis at §1.5 with explicit per-arm-vs-between-group labeling.

B.3 Pattern W — cross-section enumeration consistency

Structural-count claims are reconciled end-to-end: four COMBINE sub-populations (basal-insulin-experienced, GLP-1 RA-experienced, basal-bolus-eligible, insulin-naïve) appear consistently in §1.2, §2.2, §3.9, §4.2, §7.4, §11.5, §12.3. Three peer-reviewed COMBINE primary publications (COMBINE 1 PMID 40482671; COMBINE 2 PMID 39820580; COMBINE 3 PMID 40482670) plus one pending (COMBINE 4 NCT06269107) appear consistently. §3 panel inventory reconciles with §5 monitoring intervals and §6 AE-trigger labs.

B.4 Pattern Z — 5-anchor calibration in §10

§10 self-audit against the five Pattern Z anchors with IcoSema-specific calibration:

Anchor 1 (lead with what the option IS). §6.2 hypoglycemia framing leads with hypoglycemia management as routine clinical skill for insulin-treated patients. §10.2 initiation conversation leads with what IcoSema is mechanistically.
Anchor 2 (compounded vs FDA-approved). N/A for IcoSema combination product pre-approval; explicitly noted at §10.3.
Anchor 3 (pregnancy-planning). §10.4 leads with Parker 2025 human pregnancy-exposure data (PMID 40329607), then pharmacokinetic facts, then label recommendation, then transition-to-pregnancy-appropriate-insulin plan.
Anchor 4 (multi-dimensional comparator framing). §10.5 presents IcoSema vs separate-injection regimen vs semaglutide-monotherapy vs basal-bolus across 9+ dimensions (glycemic control, body-weight, hypoglycemia, injection burden, titration flexibility, indication scope, regulatory state, cost, comparator framing precision).
Anchor 5 (off-label / extrapolation transparency) — DOMINANT for IcoSema US pre-FDA-approval state. §10.6 presents the pre-FDA-approval regulatory state with the separate-injection regimen of Awiqli + Ozempic as the available US pathway; counseling beat explicitly addresses the EU-vs-US regulatory state distinction.

B.5 Pattern AA — regulatory-claim precision

Every regulatory claim is precise. IcoSema combination product: EMA-approved-for-marketing-claims (Kyinsu, EC decision 24 November 2025) for T2D inadequately controlled on basal insulin OR GLP-1 RA monotherapy; pre-FDA-approval-for-marketing-claims as of 2026-05-13. Component-level: insulin icodec FDA-approved-for-marketing-claims as Awiqli 2024-04-23 for T2D; semaglutide FDA-approved-for-marketing-claims across multiple indications. The combo-product-vs-component regulatory-state distinction is preserved throughout.

B.6 Pattern AB.1 / AB.2 / AB.4

PMIDs and NCTs verified at draft step (Pattern AB.1) and inherit canonical-verified state (Pattern AB.2). Pattern AB.4 cascade scan completed: COMBINE 1 PMID 40482671 verified-correct; PMID 40347948 explicitly excluded (developmental biology paper, not COMBINE 1; B-CRIT-2 resolved in canonical Phase 4 AB-hygiene audit). NCT05352815 / NCT05259033 / NCT05013229 / NCT06269107 verified content-correct.

B.7 Pattern N.1 — peptide vs small-molecule classification

IcoSema is a peptide combination product (insulin icodec is an engineered peptide with C20 fatty-diacid acylation; semaglutide is an engineered peptide with C18 fatty-diacid acylation). Pattern N.1 compliant — peptide path is correct location.

B.8 IcoSema-specific pattern-discipline notes

Pattern Z.research-precision applied to “pre-FDA-approval-for-marketing-claims” language replaces bias-vocabulary like “investigational” or “experimental” or “early-stage” (these would be Pattern Z.research-precision violations given IcoSema has a verified Phase 3 evidence base with three published primary RCTs and EMA approval).
Pattern Z.injection-framing applied at §4.5 and §10.2 — hypoglycemia management and weekly injection administration are framed as routine clinical skills, not as scary / daunting / major barriers.
Pattern AA.marketing-claims applied uniformly — every “FDA-approved” or “EMA-approved” statement is anchored to the specific indication and date; pre-FDA-approval state is precise per the canonical (FDA submission status not publicly disclosed).

Appendix C. Byte-count audit and commit log

C.1 Protocol length

Final byte-count audit. Word count approximately 22,000 words across the twelve sections and three appendices; within the 20,000–25,000 word target range per the task brief. Section length distribution proportional to evidence-density: §1, §6, §10, §11 are the highest-evidence-density sections (anchor sections — indication scope, AE management, counseling beats, citations) and are correspondingly longest. §12 (decision tree) is the shortest content section because it is operational summary rather than primary content.

C.2 Worked-example coverage

All twelve sections carry a worked IcoSema example or worked-walkthrough. §1.5 indication detail (four sub-populations with effect-size anchors); §2.5 selection criteria detail (four sub-populations); §3.9 pre-treatment panels (four sub-populations); §4.7 initiation worked example (COMBINE 1 phenotype); §5.5 maintenance worked example; §6.12 AE management worked examples (GI persistent, hypoglycemia overshoot, NAION); §7.4 non-response algorithm worked example; §8.7 discontinuation scenarios (pre-conception, patient-preference, pancreatitis, new pregnancy); §9.6 combination scenarios (canonical polycondition, sulfonylurea-discontinuation, off-label bolus addition, US pre-FDA-approval); §10 counseling beats with verbatim Pattern Z anchor calibration; §11 Bibliography subset; §12.4 phenotype-guided decision tree application.

C.3 Cross-reference inventory

The protocol references: /obsidian-peptides/Peptides/IcoSema.md (canonical primary), /obsidian-peptides/Peptides/Semaglutide.md (component canonical, GLP-1 RA component anchor), /obsidian-peptides/Methodology/Protocol Template.md, /Internal/Module-5-Pipeline/pattern-z-calibration-anchor.md, /obsidian-peptides/Methodology/Synergy Editorial Framework.md (v1.2), /obsidian-peptides/Methodology/Voice Profile - Dr. Jeff Gross MD.md, /obsidian-peptides/Methodology/AC2-26 - System Observations.md, /Process/IcoSema/IcoSema - Bibliography.md.

C.4 Six-commit production audit

The protocol was produced in six strict incremental commits per the production-discipline standard:

Commit 1 — Frontmatter, header front-matter, intro framing, TOC, methodology cross-references, §1 Indication scope, §2 Selection criteria. Commit hash 904234f.
Commit 2 — §3 Pre-treatment workup, §4 Initiation, §5 Maintenance. Commit hash ff24d8c.
Commit 3 — §6 Side-effect management (hypoglycemia-dominant), §7 Non-response algorithm. Commit hash 7bfe2a3.
Commit 4 — §8 Discontinuation, §9 Combination rules. Commit hash 160c28a.
Commit 5 — §10 Counseling beats, §11 Source citations. Commit hash bf868ce.
Commit 6 — §12 Clinical decision tree, Appendix A Verification gate, Appendix B Pattern discipline self-audit, Appendix C Byte-count audit. Commit hash (this commit).

Each commit followed the CLAUDE.md commit-message template with Protocol: IcoSema §X-Y (Commit N/6) subject prefix and the standard Co-Authored-By trailer.

C.5 Document version and next-iteration trigger

Protocol version v1.0. The next-iteration triggers are: (a) COMBINE 4 primary publication (will populate §1.2 / §1.5 insulin-naïve sub-population effect-size anchor and §11.5 Bibliography); (b) FDA approval of the IcoSema combination product (will shift §10.6 Anchor 5 dominance from US-pre-FDA-approval to US-post-FDA-approval framing); (c) post-launch RWE accumulation (will populate §8.5 post-discontinuation framing and §9 IcoSema-vs-separate-injection-regimen comparator with real-world data); (d) any subsequent Module 5 protocol production cycle that surfaces new sub-patterns warranting v1.1 patch. Iteration history will be appended at this section as it accrues.

Self-audit completion: 2026-05-13. Awaiting Dr. Gross verification gate (Appendix A.4).

IcoSema Protocol