When I was serving as a Senior Advisor at the Centers for Medicare & Medicaid Services, one of the most alarming meetings I had was with a group of oncologists shortly after the launch of Chimeric Antigen Receptor (CAR)-T cell therapy. The first gene therapy approved by the FDA, CAR-T therapy reprograms a patient’s own immune cells to fight disease. This personalized approach worked for certain blood cancers that did not respond to existing treatments. However, the oncologists said that accessing CAR-T therapy was nearly impossible, because hospitals lost tens to hundreds of thousands of dollars every time the therapy was provided to a patient in Medicare.
It took until the fall of 2020 – three years after FDA approval – for Medicare to consistently pay an adequate amount for CAR-T therapy. Now that same agonizing process awaits a raft of promising gene therapies in the pipeline – with lives in the balance – unless policymakers take action.
The financial losses that the oncologists described resulted from Medicare policies that do not accommodate new therapies that differ dramatically from existing ones. After an inpatient hospital stay, a Medicare patient is assigned a DRG or “Diagnosis Related Group” based on the patient’s clinical conditions and treatments received. Medicare pays hospitals a set amount for each DRG to cover the costs of an inpatient stay, including medicines and therapies.
For new therapies that lack a DRG, Medicare identifies the most similar existing DRG and pays that amount. Meanwhile, CMS gathers data on the actual cost to hospitals of the new therapy and associated services. Once the agency has sufficient data, CMS can use it to establish a new DRG.
The most similar existing intervention to CAR-T therapy was a bone marrow transplant. However, payment for that DRG was only $35,000-$40,000, far lower than CAR-T’s list price of $373,000. No one argued that CAR-T therapy was overpriced by a factor of ten. The Institute for Clinical and Economic Review (ICER) actually found that at its list price, CAR-T therapy was within commonly-cited thresholds of cost-effectiveness. Nevertheless, Medicare paid hospitals based on the bone marrow transplant DRG, as the law and regulations required.
Medicare does make certain payments based on a hospital’s charges that partially cover gaps between DRG amounts and costs. “New technology add-on payments” or NTAPs are paid for substantial clinical improvements over existing therapies. However, at the time CAR-T was approved, an NTAP could not exceed 50% of the cost of a new technology. Also available are “outlier payments,” which offset a fraction of costs above a threshold.
To complicate things further, the DRG-based payment approach only applies to inpatient cases. For hospitals that provided CAR-T therapy in an outpatient department, Medicare paid the average sales price of CAR-T therapy plus an additional six percent of this price – an amount significantly greater than Medicare’s inpatient rate. However, the first CAR-T therapies called for inpatient monitoring. Financial incentives conflicted with clinical best practice.
Hospitals were left trying to navigate Medicare’s policies and offset as many of their costs as possible – leaving many patients without the therapy they needed. By the fall of 2020, CMS had sufficient data on the costs of CAR-T therapy to establish a new DRG. However, with the FDA predicting that by 2025 ten to twenty cell and gene therapy products – which could treat any number of diseases and conditions – will be approved annually, CMS must prepare for the next innovation that is a radical departure from available therapies.
In the near-term, CMS should increase the percentage of costs that an NTAP can cover to around 80%. While I was at CMS, I was proud that we increased this percentage from 50% to 65%. As with every dial that CMS controls, this one is tricky to set. Covering too high a percentage could discourage manufacturers from offering discounts to hospitals. Insurance plans in Medicare Part D negotiate rebates that add up to around 30% of prescription drug list prices, but it is unrealistic to expect individual hospitals to negotiate that level of discount from manufacturers of new therapies. Part D plans to negotiate with manufacturers of retail prescription drugs, many of which are in highly competitive therapeutic classes.
Increasing the NTAP percentage would help, but doing this alone would not address the difference in Medicare’s payment amounts between inpatient and outpatient sites of service. And CMS can’t count on its Center for Medicare and Medicaid Innovation (CMMI) to solve this reimbursement problem. CMMI models must be cost-saving or cost-neutral relative to the status quo, but the problem with CAR-T therapy was inadequate payment.
A comprehensive solution will require legislative change. Congress should pull gene therapies entirely out of Medicare’s buy-and-bill system for hospitals. A pathway could be created for the government to purchase gene therapies directly from manufacturers for Medicare patients, as the government has done for all doses of COVID-19 monoclonal antibody therapies and vaccines. A more nimble approach is urgently needed, as science is moving faster than Medicare’s payment policies.
Jeet Guram, MD, is a visiting scholar at the USC Schaeffer Center for Health Policy & Economics. He was a Senior Advisor to the Administrator of the Centers for Medicare & Medicaid Services and to the Commissioner of the Food and Drug Administration.