Comprehensive review of the genotype-phenotype spectrum and treatment for biallelic insulin receptoropathy based on 259 affected people

Background Biallelic deleterious variants in the insulin receptor gene ( INSR ) cause severe insulin resistance (IR), historically partitioned into Donohue syndrome, Rabson–Mendenhall syndrome and type A IR. Case literature is scattered, and clinical management rests on uncontrolled case series. Genotype–phenotype relationships, natural history, prognosis and optimal management are only partially established. Methods We reviewed reports of biallelic INSR variants published over 40 years and supplemented these with cases from international referral centres, abstracting clinical, biochemical and genetic data from 259 affected individuals (226 unrelated probands; 210 distinct INSR variants). Variant burden was compared with the general population using gnomAD, and a Random Survival Forest-based prognostic model was developed, with cross-validation of feature importance. Results Complete loss-of-function (KO) variants affected 33.3% of patients vs 12.6% in gnomAD (p = 2.5×10⁻¹²), with no domain clustering. All biallelic KO patients presented in the first month of life. Disease was multi-system, encompassing profound IR with fasting hypoglycaemia, hyperglycaemia without dyslipidaemia, pre-/postnatal growth failure with soft-tissue overgrowth, cholestatic liver disease, nephrocalcinosis, cardiac hypertrophy, ovarian enlargement, coagulopathy and colonic overgrowth. Seventy-two individuals (28%) died at a median 0.36 years, most commonly from sepsis; all KO patients were deceased or censored by 36 months. The prognostic model discriminated survival beyond 2 years, stratifying the cohort into tertiles with 2-year survivals of 89.8%, 83.3% and 39.5%. Biallelic KO variants and low birthweight standard deviation score were the strongest negative predictors, with male sex also adverse. Evidence guiding therapy is weak: insulin and metformin remain pragmatic mainstays, SGLT2 inhibitors show encouraging benefit, and rhIGF-1, used in a third of cases, yielded mixed outcomes. Conclusions Biallelic insulin receptoropathy forms a clinical continuum. Routinely available presentation features permit data-driven prognostication, facilitating individualised risk assessment. Several stereotyped tissue-specific features remain poorly understood mechanistically. Strengthening of the therapeutic evidence base is urgently required.