What the durability gap costs a stop-loss carrier.
A gene therapy costs $2–4 million. The OBA window closes. After that, nobody confirms whether it's still working. This tool models what that gap costs — and what independent adjudication changes in the renewal corridor.
OBA Enforcement — No published dataRebate Magnitude — Range only (17–100%)Renewal Spike — UnsourcedDurability — Trial data, limited follow-upCGT PEPM — Published pricing
Model Inputs
Set the assumptions. See what moves.
Each input carries a confidence tier. The one that matters most — OBA enforcement — has never been published. That's not a limitation of the model. It's the thesis.
This model assumes one CGT claim has occurred. It models severity and exposure, not incidence probability.
Therapy Cost Published
$2.2M
Range: $700K (Luxturna) – $4.25M (Lenmeldy)
OBA Enforcement Rate No data
50%
No enforcement results have ever been published. This is an assumption with zero empirical floor.
ICER/NEWDIGS, April 2024
10-Year Efficacy Decline Modeled
10%
Modeled aggregate, not observed population rate. Driven mostly by Roctavian (FVIII declining). Most therapies show durability.
CGT Stop-Loss PEPM Published
$4.05
$4.05 = PAI/SBR (published). $4–6 inferred band. Other carriers quote, not publish.
PAI Gene Therapy Stop-Loss Solution, 2026
Rebate Magnitude (when enforced) Range only
50%
Published structures range from 17% (Zolgensma Medicaid cap) to 100% (Roctavian pro-rated WAC). No aggregate figure exists.
Novartis, BioMarin, bluebird bio warranty disclosures
Plan Size (employees)
5,000
Annual Claim Probability Estimate
0.5%
Probability that at least one plan member requires an approved CGT in a given year. Varies by population, plan size, and eligible therapies. No standard actuarial table exists.
Base Renewal Increase Published
11%
Segal: 9.4% (2024), 11.5% comparable. Reinsurance hardening ~15% for 2026.
Segal/IFEBP 2025
Why OBA enforcement is the centerpiece
No outcomes-based agreement in cell and gene therapy has ever disclosed its enforcement results. Not the rebate collected. Not the adjudication process. Not the role of any third party. The full 0–100% range is the honest input. The swing across that range is the argument for an independent adjudicator.
PharmacoEconomics 2025 scoping review; ASTCT 2026 scoping review; ICER/NEWDIGS April 2024 white paper
Model Output
The economics of the gap.
$2.2M
Gross Therapy Exposure
$0
Expected OBA Rebate
$2.2M
Net Carrier Exposure
$0
Expected Annual Loss
P(claim) × net exposure
$0
Annual CGT Premium
PEPM × plan × 12
Renewal corridor: two paths.
After a gene therapy claim, the carrier faces two paths. The carved-out path was designed for this. The unprotected path was not.
Path A — Carved Out
$0
CGT carve-out product excludes the claim from renewal calculations. Post-claim spike near zero by design.
Annual CGT premium: $0
Path B — Unprotected ISL
$0
Traditional stop-loss. Outcomes: large increase, laser, or exclusion. No published post-CGT renewal data exists — "doubling" traces to a single competitor illustration (Quantile Health), not actuarial data.
Base renewal: 0%
Post-claim spike: unsourced — carrier-specific, not modeled
OBA enforcement sensitivity.
What the net carrier exposure looks like across the full enforcement range. The swing is the thesis.
At 50% rebate magnitude
The durability cost sink
Durability failure in AAV-based gene therapies is a dead loss, not a retreatment cost. Anti-AAV antibodies form after the first dose and preclude redosing. The patient loses benefit, cannot be re-dosed, and returns to baseline therapy (e.g., factor replacement for hemophilia, chronic transfusion for SCD). At 10% modeled efficacy decline over 10 years, $0 in therapy value is unrecoverable — and downstream treatment costs resume with no offset.
What Cadence does here
Independent durability governance.
01
A durability finding the carrier can price against. It comes from outside the payer's incentive structure. It holds up under underwriting scrutiny.
02
OBA milestone adjudication no manufacturer can call conflicted. The adjudicator has no stake in which direction the outcome lands.
03
A record that outlives the agreement. Independent measurement at year three, year five, after the OBA window has closed and no one else is looking.
A plan cannot independently adjudicate its own outcomes agreement. That is the structural reason for an arm's-length party.