Excerpt – In this issue of JAMA Cardiology, Cohen and colleagues have performed a formal cost-effectiveness analysis of SGLT2 [sodium-glucose cotransporter-2] inhibitors for patients with heart failure and an ejection fraction more than 40%. They developed a computer-simulation model to project the long-term clinical benefits and costs for patients with HFpEF [heart failure with preserved ejection fraction] with and without SGLT2 inhibition. Their model was based on pooled estimates of baseline risk and effectiveness of SGLT2 inhibitors derived from the EMPEROR-PRESERVED and DELIVER trials. Because these trials followed up patients for a median of only 2.3 years, the authors extrapolated long-term event rates under the assumption that the therapeutic benefits on mortality, quality of life, and heart failure hospitalization would persist over a patient’s lifetime. In their primary analysis, they found that treatment with SGLT2 inhibitors led to an additional 0.19 quality-adjusted life-years (QALYs) at a cost of $26 312 with an incremental cost-effectiveness ratio of $141 200 per QALY gained. Based on current US benchmarks of less than $50 000 per QALY for high value, $50 000 to $150 000 per QALY for intermediate value, and more than $150 000 per QALY for low value, this narrowly met the standard for intermediate value. However, the study authors noted substantial uncertainty in their assessment of cost-effectiveness; in 41% of simulations, SGLT2 inhibitors had a cost per QALY exceeding $150 000 per QALY. Thus, despite their unequivocal clinical benefit, at current prices, treatment of patients with HFpEF, similar to those enrolled in the pivotal randomized trials, appears to provide only intermediate to low value.
The study by Cohen and colleagues provides much more insight into the appropriate role of SGLT2 inhibitors in patients with HFpEF, however. Through a series of sensitivity and scenario analyses, the authors provide extensive insight into those factors that influence the cost-effectiveness of SGLT2 inhibitors for patients with HFpEF. First, the economic value of SGLT2 inhibitors is highly dependent on their effect on cardiovascular mortality, which was not significantly reduced in either of the individual trials or a recent meta-analysis. Nonetheless, for their primary analysis, the authors assumed a long-term relative risk of 0.88 for cardiovascular death based on the point estimates from the pooled trial data. Under the more conservative assumption of no effect on mortality, the incremental cost-effectiveness ratio for SGLT2 inhibition increased to $373 400 per QALY gained, indicative of low economic value. The contrast between these 2 results emphasizes the need for additional data on the effect of SGLT2 inhibitors on cardiovascular mortality among patients with HFpEF.
Second, the value of SGLT2 inhibitors for HFpEF is highly dependent on their price, which not only varies substantially across payers but is also often hidden as a trade secret. The authors followed best practices by using the Federal Supply Schedule price (the price available to all federal agencies), but payers may obtain SGLT2 inhibitors at a lower cost. If payers received a 68% price reduction—a level that might be achieved by certain insurers—SGLT2 inhibitors would provide high economic value. This uncertainty is further complicated by the likely availability of generics as soon as 2025, which can be expected to improve the future cost-effectiveness of SGLT2 inhibitor therapy.
The contrast between the clinical impression that SGLT2 inhibitors are a revolutionary treatment for HFpEF and the economic assessment that they provide only intermediate to low economic value is jarring, especially in light of previous studies demonstrating that these drugs provide better economic value in patients with HFrEF [heart failure with reduced ejection fraction]. Since drug costs are similar for these 2 conditions, the difference in value is driven by the smaller absolute benefit among patients with HFpEF compared with those with HFrEF. Again, this finding may seem counterintuitive given the consistent relative risk reduction for cardiovascular mortality across the spectrum of both HFrEF and HFpEF. However, the explanation for these counterintuitive findings becomes clear when one examines the baseline risk of these 2 populations. Among patients with HFrEF, the absolute risk of cardiovascular death is substantially higher than among patients with HFpEF. Moreover, among patients with HFpEF, a higher proportion of overall mortality is noncardiovascular. The competing risk of noncardiovascular mortality reduces the absolute benefit of a therapy that targets cardiovascular outcomes. The multimorbidity complexity of patients with HFpEF underscores the challenge of improving clinical outcomes among this population.
The final key finding of this study is the substantial variation in cost-effectiveness across patient subsets. Like the overall HFpEF population, patients in the DELIVER and EMPEROR-PRESERVED trials were diverse and included patients with true HFpEF and those with heart failure with midrange ejection fraction. Not surprisingly, both the absolute clinical benefit and economic value of SGLT2 inhibition varied across this diverse population. For example, even under the most favorable assumptions, SGLT2 inhibition was projected to provide low value for patients older than 78 years (who represent a substantial subgroup of the HFpEF population) or when the left ventricular ejection fraction was 60% or more. On the other hand, as suggested by a previous study, SGLT2 inhibition may provide greater value for patients with more symptomatic HFpEF (where the quality of life gains are greater.) Among patients with both HFpEF and chronic kidney disease, SGLT2 inhibitors may also have greater long-term value by slowing the progression of chronic kidney disease.
When considering these findings, it is also important to recognize that cost-effectiveness analyses provide valuable insight into expected long-term benefits and costs for the health system but are not intended for individual patient decision-making. Not only do they generally fail to capture the patient-level heterogeneity discussed above, but patients weigh their expected benefit against affordability, which is based on out-of-pocket costs rather than overall costs. Unfortunately, in the US, many drugs that provide high value based on cost-effectiveness from a societal perspective remain unaffordable to many patients, exacerbating disparities in adoption of life-saving therapies.
Understanding the economic value of novel therapies is important for health systems to prioritize implementation in the context of limited budgets. The higher the value, the more health systems should invest in promoting adoption. Recently, the American College of Cardiology and the American Heart Association have gone so far as to restrict performance measures to therapies considered high or intermediate economic value. Nonetheless, cost-effectiveness should remain only one consideration when thinking about the overall importance of novel therapies. For SGLT2 inhibitors in HFpEF, being the sole effective drug for a condition that causes substantial morbidity warrants praise, even if it comes at a substantial cost. Even so, until substantial reductions in the cost of these agents are realized, for the HFpEF population, we will continue “living on the edge.”
Full editorial, AT Sandhu and DJ Cohen, JAMA Cardiology, 2023.3.4