Medicare Part D spent $17.39 billion on apixaban in 2023. In 2013, it spent $24.7 million on the same drug. That's not a typo.

Over the decade between 2013 and 2023, total Medicare Part D spending grew by $131.77 billion. Apixaban, a blood thinner used to prevent stroke in patients with atrial fibrillation, consumed 13.18% of that entire growth on its own. No other drug came close.

A Decade of Drug Spending, Rewritten by Five Molecules

The scale of apixaban's rise is difficult to contextualize. A 70,435.8% increase over ten years. The drug that led Part D spending in 2013, esomeprazole magnesium (a proton pump inhibitor), cost $2.32 billion that year. Apixaban's 2023 bill was more than seven times that figure.

But apixaban wasn't alone in reshaping the program's cost structure. The next four drugs by absolute spending growth, semaglutide, empagliflozin, dulaglutide, and rivaroxaban, added another $29.47 billion combined. All five of them ranked in the top five most costly drugs in 2023. None of them were in the top 20 in 2013, with the exception of rivaroxaban, which grew 1,241.6% from $410.8 million to $5.51 billion.

The top six drugs by absolute growth account for more than 38% of the program's entire decade-long spending increase. For every dollar Medicare Part D spent more in 2023 than in 2013, roughly 38 cents went to these six drugs. That concentration matters: it means the program's cost trajectory is being driven by a small number of high-volume, high-price therapies, not by broad-based inflation across thousands of drugs.

The Turnover No One Predicted

Fifteen of the top 20 most costly drugs in 2023 were not in the top 20 in 2013. The drugs that defined Part D spending a decade ago, esomeprazole, fluticasone/salmeterol, rosuvastatin, aripiprazole, duloxetine, have largely been displaced by cardiovascular agents, GLP-1 receptor agonists, SGLT2 inhibitors, and oncology drugs.

This is a structural shift, not a pricing anomaly. The drugs that fell off the top 20 list were mostly generics or near-generic branded products whose costs declined as competition increased. The drugs that replaced them are largely still under patent, with no biosimilar or generic competition in the period covered. Patients on these newer therapies face different cost-sharing dynamics, and the program faces a fundamentally different spending profile than it did when Part D was designed.

Lenalidomide's trajectory illustrates a different dimension of this problem. It grew 307.5% from $1.18 billion in 2013 to $4.82 billion in 2023, accounting for 2.76% of total Part D growth. Unlike the cardiovascular and metabolic drugs, lenalidomide treats multiple myeloma, a cancer with a relatively small patient population. Its cost growth reflects price increases and extended use, not a broad expansion of the treated population.

Semaglutide Is Still Accelerating

Semaglutide reached $9.67 billion in Part D spending in 2023, starting from zero in 2013. It's already the second-largest driver of Part D spending growth over the decade. GLP-1 prescribing has continued to expand well beyond the 2023 data captured here, driven by approvals for obesity treatment in addition to type 2 diabetes management.

The comparison to apixaban is instructive. Apixaban's growth was driven by a large, well-defined patient population (atrial fibrillation affects millions of older Americans) and a decade of clinical adoption. Semaglutide's potential patient population is orders of magnitude larger. If its trajectory follows even a fraction of apixaban's arc, its share of total Part D spending growth will look very different in the next analysis.

Given that semaglutide reached $9.67 billion in Part D spending in 2023 starting from zero in 2013, how much of the next decade's spending growth will a single GLP-1 drug consume?