Percutaneous Coronary Intervention for Heart Failure: Worth the Cost?

In the 1980s, Dr. Rahimtoola proposed that chronic myocardial ischemia can lead to hibernation, an adaptive state of left ventricular (LV) systolic dysfunction that can be completely reversed with revascularization.^1,2 These observations were supported by recovery of LV function after coronary artery bypass graft (CABG) surgery. Over the past four decades, revascularization, with the goal of recruiting hibernating myocardium has been a key component of management of chronic ischemic heart failure, even though evidence from randomized controlled trials wasn’t available until recently.

The Surgical Treatment for Ischemic Heart Failure (STICH) trial was the first to investigate the role of revascularization with CABG among 1212 patients with chronic ischemic heart failure and severely depressed LV function. Although there was no significant difference in mortality at a median follow up of 56 months, the survival curves continued to separate during extended study (STICH-ES), such that at 10 years CABG was superior to optimal medical therapy (OMT) alone.^3,4 Importantly, the benefit of CABG was most pronounced in patients who were less than 60 years old and those with greater severity of LV dysfunction.^5,6 However, there was an early hazard of mortality in patients who underwent CABG due to the inherent risk of surgery in patients with severe LV dysfunction. Due to the upfront risk of mortality with CABG, it is logical to consider that percutaneous coronary intervention (PCI) could potentially amplify the benefit of revascularization noted in STICH.

To address this question, the Revascularization for Ischemic Ventricular Dysfunction (REVIVED-BCIS2) trial randomized 700 participants with chronic ischemic heart failure, LV ejection fraction < 35%, and severe coronary artery disease with presence of viable myocardium to PCI with OMT vs. OMT alone. ⁷ Among participants in the PCI group, drug-eluting stents were used in 97% and the procedure was successful in 98% with a low overall use of circulatory support. Contrary to the findings of STICH, PCI was not superior to OMT in reducing the incidence of mortality and hospitalization for heart failure. There was also no difference in improvement in ejection fraction between the two groups. An important learning from both trials was that medical therapy was quite effective in improving heart failure outcomes including LV function.

In this issue of the Journal, Chivardi and colleagues⁸ extend the findings from the REVIVED-BCIS2 trial with their report on a cost-utility evaluation of PCI + OMT compared to OMT. The analysis was performed from the United Kingdom National Health Services perspective which uses an implicit cost-effectiveness threshold of £20,000-£30,000 per quality adjusted life year (QALY). Costs were estimated using in-trial data on healthcare resource use, while QALYs were estimated using the EQ-5D-5L score assessed at each follow up period and converted to utilities. Cost and QALY data were adjusted for differences in baseline characteristics using generalized linear models, and the adjusted cost and QALYs were used to calculate incremental cost effectiveness ratio (ICER). Over a median follow up of 3.4 years, the adjusted cost of PCI + OMT was £22,352 compared with £15,569 for OMT (a difference of £6783, with PCI+OMT being more expensive than OMT). However, there was negligible difference in adjusted QALY between the two groups over the same period (a difference of 0.015 QALY, with PCI+OMT being less effective than OMT). Thus, PCI resulted in higher cost and lower QALY (i.e., PCI was “dominated” by OMT). In sensitivity analyses, there was zero probability that PCI would be cost-effective under any scenario. Given that the survival curves were nearly super-imposed, a life-time horizon was not considered based on a low likelihood of future benefit beyond the in-trial period. These negative findings are in sharp contrast to economic analyses from the STICH-ES trial which found that CABG compared to OMT had an ICER Of ~$64,000 per QALY over a life-time horizon, which would be considered intermediate value in the U.S.^9,10

Although the negative finding of the current economic analysis is not altogether surprising, it is important nevertheless because there was a signal for improvement in health-related quality of life at 6 and 12 months, an important consideration for some patients. However, the difference in quality of life was not sustained beyond 24 months.⁷ While findings of cost-effectiveness analyses can be difficult to translate across health systems due to substantial differences in costs and outcomes, the trivial benefit of PCI in the short-term and better health outcomes with OMT in the long-term would mean that PCI is unlikely to be cost-effective in other health systems, particularly in the US where healthcare costs are substantially higher than in the UK. Important strengths of the study include the use of data collected during a clinical trial for estimating costs, and medium to long-term follow up. There are also important limitations especially missing data on EQ-5D-5L scores in a substantial proportion of patients and a lack of clear justification for conducting adjustment to estimate cost and QALY.

So how must one reconcile the divergent findings of STICH-ES and REVIVED-BCIS2, regarding the benefit of revascularization for heart failure? Even though STICH-ES showed a benefit of CABG compared with OMT alone while REVIVED-BCIS2 did not show a benefit of PCI compared with OMT alone, neither study showed that revascularization is more effective in recovering hibernating myocardium compared with medical therapy.^11,12 It is possible that because CABG bypasses proximal segments of major coronary arteries, it avoids cardiovascular events not only in severely stenosed lesions but also other non-obstructive lesions in the proximal vessels. At the same time, it is important to emphasize some key differences between patients enrolled in STICH and REVIVED-BCIS2 studies that make comparisons across the two studies difficult. First, participants enrolled in REVIVED-BCIS2 were older compared to STICH (median age: 70 years vs. 60 years) - an important distinction because the benefit of revascularization with CABG in STICH and STICH-ES was primarily seen among those younger than 60 years of age. Whether revascularization with PCI will benefit younger patients who may also be candidates for CABG was not studied in REVIVED-BCIS2. Second, even though both studies included patients with severe LV dysfunction (mean LV ejection fraction was 27%), nearly two-third of patients enrolled in REVIVED-BCIS2 did not report any angina, and more than 50% had 1- or 2- vessel CAD. In contrast, 63% of patients in STICH had angina.

Thus, it is possible that REVIVED-BCIS2 enrolled many patients in whom the severity of LV dysfunction was out of proportion to CAD severity. Revascularization is unlikely to provide any benefit in such patients. Third, in contrast to median follow up of 10 years in STICH-ES, median follow up in REVIVED-BCIS2 was only 3.4 years. However, in the absence of any signal of benefit for PCI in REVIVED-BCIS2, and the competing risk of death due to the enrolment of a much older cohort, it remains to be seen whether a benefit of PCI would emerge over longer follow up. Lastly, the intensity of GDMT used was higher in REVIVED-BCIS2 compared with STICH. Given these differences, an ongoing randomized trial comparing CABG vs. PCI for patients with ischemic cardiomyopathy will yield important insights regarding the relative benefit of PCI over CABG in this population.¹³

Despite the above, use of PCI for revascularization in chronic ischemic heart failure is not uncommon.¹⁴ In fact, the 2017 Appropriate Use Criteria rates PCI in patients with severe LV dysfunction as appropriate or maybe appropriate depending on the presence of symptoms, diabetes and number of vessels involved.¹⁵ Although the current study does not shed light on important subsets of patients (e.g., younger patients, those with high angina burden), the REVIVED-BCIS2 investigators need to be lauded for their efforts in defining the role of PCI in chronic ischemic heart failure and quantifying the cost and benefit. Collectively, these studies highlight that medical therapy should remain the cornerstone for management of heart failure. In appropriate candidates especially younger patients with few co-morbidities, revascularization with CABG has been shown to improve survival. In non-surgical candidates, PCI may be considered for angina relief, but routine use of PCI to improve survival in patients with chronic ischemic heart failure does not appear to be worth the additional cost.