Gene therapy, a novel medical treatment altering a patient's genome, holds immense potential. However, the high cost raises affordability concerns. While four gene therapies were FDA-approved by December 2022 (voretigene neparvovec, onasemnogene abeparvovec-xioi, brexucabtagene autoleucel, and etranacogene dezaparvovec), their high prices (e.g., $425,000 per eye for voretigene neparvovec, $2.1 million per patient for onasemnogene abeparvovec-xioi) have sparked debate. Drug companies justify these prices by citing significant clinical benefits, high R&D costs (billions of dollars and decades of time, with late-stage trials being particularly expensive), and the inherent risks of development. The success rate of drug development programs is low (13.8% reach FDA approval). Moreover, insurance coverage varies significantly, with many plans not covering approved therapies or imposing restrictive policies. This leaves uninsured or underinsured patients unable to afford treatment. Aggregate spending on gene therapies is unknown due to a lack of publicly available data on patients treated annually. Previous research provides limited estimates (e.g., Quinn et al., 2019) without disaggregation by gene therapy or disease. This study aims to estimate the expected annual fiscal impact of gene therapy in the U.S. by creating and implementing a novel financial model that simulates future approvals, patient populations, and prices.

The introduction adequately references existing literature. It highlights the high cost of gene therapies, the variability in insurance coverage, and the lack of comprehensive data on treatment numbers and spending. The authors cite previous work by Quinn et al. (2019) estimating future patient numbers but acknowledge its limitations. They emphasize the absence of systematic data on U.S. spending on gene therapies, setting the stage for their own comprehensive analysis.

The study utilizes a simulation model (flowcharted in Figure 1) to estimate the future financial impact of gene therapy. The methodology involves several key steps: 1. **Identifying Gene Therapies:** The authors identified 109 unique gene therapy clinical trials (Phase 2/3 or 3) from Citeline Trial Trove and ClinicalTrials.gov databases (before December 31, 2019). Compassionate use trials were excluded. Trials without U.S. sites were included due to the possibility of FDA approval based on foreign trial data. Diseases were categorized into oncology, rare disease, and general disease. 2. **Probability of Success:** A simulation estimated the likelihood of success for each trial, informed by prior studies on drug development success rates by therapeutic area from the MIT Laboratory of Financial Engineering's Project ALPHA. The model accounted for correlation between development programs (assumed at 90%, with sensitivity analysis showing insensitivity to this parameter). 3. **Time to Approval:** Time to approval was simulated using triangular distributions for the time between trial completion and BLA submission (median 182.5 days) and between BLA submission and FDA approval (median 182.5 days). Missing trial end dates were imputed using a fitted gamma distribution. 4. **Number of Patients:** The simulation estimated the number of treated patients over time based on disease prevalence and incidence data from various sources (e.g., SEER). Equations 1 and 2 were used to derive prevalence from incidence or vice versa. An exponential decay function (Equation 3) with a half-life parameter (λ) modeled the treatment of existing patients. A ramp function (Equation 4) modeled patient penetration over time, varying by disease category (rare diseases having faster adoption). 5. **Expected Market Pricing:** The model estimated prices based on quality-adjusted life years (QALYs) gained, using a calibrated price per QALY. Calibration was done using existing gene therapy prices (onasemnogene abeparvovec-xioi, voretigene neparvovec) and CAR-T therapies (tisagenlecleucel, axicabtagene ciloleucel). Two separate calibrations were performed to account for potential QALY differences. Equation 7 was used to estimate the Price of Gene Therapy. 6. **Simulation:** One million iterations were performed to calculate the mean number of gene therapy patients and total spending. The 5th and 95th percentiles were reported as upper and lower bounds. 7. **Sensitivity Analysis:** The impact of ±20% changes in key variables (maximum penetration rate, QALY gained, price per QALY, probability of success, patient numbers, time to approval) was assessed on cumulative spending and peak monthly spending. The effect of correlation between development programs and the proportion of existing patients seeking treatment in the first year were also examined.

The simulation results predict: * **Expected Approvals:** 18.3 gene therapies (90% CI: [14.0, 23.0]) between January 2020 and January 2034. * **Patient Numbers:** A peak of 7911 patients treated per month in July 2025 (CI: [3978, 12,477]), declining to 5424 by December 2034 (CI: [2778, 8350]). Cumulative patients treated: 1.09 million (CI: [0.595 M, 1.66 M]) by December 2034. Cancer patients are expected to constitute the largest group. * **Spending:** A peak monthly spending of $2.11 billion (CI: [1.01B, 3.88B]) in April 2026, declining to $1.62 billion (CI: [0.624B, 2.98B]) per month. Cumulative discounted spending: $241 billion (CI: [123B, 402B]) by December 2034. Average annual spending: $20.4 billion. * **QALYs:** An average increase of 5.12 life-years per treated patient, with an average cost of $43,110 per QALY gain. * **Spending by Payer:** Estimated annual spending of $8.1 billion by Medicare, $5.44 billion by Medicaid, and $12.2 billion by private sources. * **Sensitivity Analysis:** The model's key findings (cumulative spending, peak monthly spending) showed linear scaling with changes in maximum penetration rate, QALY gained, and price per QALY. Other variables showed sublinear or opposite directional effects. Introducing new gene therapy programs increased estimates of approvals, patients, and spending by around 15-25%.

The estimated $20.4 billion in average annual spending represents a significant financial impact. This is a lower bound, as the model makes conservative assumptions (e.g., excluding trials after December 2019, not considering cost savings from fewer treatment sessions). The price estimation based solely on QALYs might over- or underestimate spending, as it doesn't account for cost savings compared to standard care or the potential for price reductions due to competition. However, the inclusion of additional FDA-approved therapies after 2020 adds credence to the model's accuracy. The high cost per QALY ($43,110) is notable but needs to be considered within the context of overall U.S. healthcare spending and tax revenue. The authors discuss potential funding mechanisms, including adjustments to Medicare and Medicaid budgets and the possibility of a national gene therapy reinsurance company to manage risk and coordinate access. This approach would address concerns such as adverse selection and offer cost-effectiveness through Centers of Excellence.

This study provides valuable estimates of the potential financial impact of gene therapy in the U.S., offering insights for decision-makers. The model, while having limitations, offers a framework for incorporating new data and assumptions. The findings highlight the need for strategic planning and innovative financing mechanisms to ensure patient access while managing the financial burden on payers.

The study's limitations include its reliance on conservative assumptions (e.g., excluding trials after 2019, not fully accounting for cost-saving potential, lack of robust competition data). The price estimation based solely on QALYs might over or underestimate aggregate spending, and the model simplifies aspects of patient treatment and insurance coverage. The authors acknowledge these limitations and emphasize the model's adaptability to future data and refinements.