BASIL-3: More Questions Than Answers About Drug-Coated Devices in CLTI

Compared with plain old balloon angioplasty (POBA), drug-eluting stent and drug-coated balloon (DCB) technologies do not improve amputation-free survival in patients with chronic limb-threatening ischemia (CLTI), the BASIL-3 trial concludes.

The study also included an analysis that suggested DCBs are unlikely to be cost-effective under the current willingness-to-pay thresholds in the United Kingdom, though BASIL-3 statisticians said there was some indication that DES potentially may be cost-effective.

Following the presentation of the topline results yesterday at the 2024 Charing Cross International Symposium, the study investigators all struggled with questions of how to interpret the results and deflected criticism about various issues ranging from the primary endpoint they chose to the possibility that the trial was underpowered.

“I suppose we'd have liked to have had a more clear-cut result like we had with BASIL-2, but the nature of these trials is you don't always get that, and we try to report that data to you completely, honestly, with no spin—just telling you what we found,” said lead investigator Andrew W. Bradbury, MBChB (University of Birmingham, England).

In response to a question from moderator Andrew Holden, MBChB (Auckland Hospital, New Zealand), regarding what the audience should take away from the results, Bradbury said, “I don't really think it's for us to tell you how to use these data.”

BASIL-3 is part of a series of trials that have attempted to fill evidence gaps regarding best treatment strategies for CLTI following a conclusion from the National Institute for Health and Care Excellence (NICE) in 2012 that there was insufficient evidence to support the use of DES and DCB in the treatment of PAD. The study’s CLTI patients were randomized to one of three treatment arms: POBA with or without BMS (n = 160), DCB with or without BMS (n = 161), and DES alone (n = 159), with the primary endpoint being amputation-free survival, the same endpoint used in BASIL-2 and BASIL-1.

I don't really think it's for us to tell you how to use these data. Andrew W. Bradbury

Eric Secemsky, MD (Beth Israel Deaconess Medical Center, Boston, MA), who was not involved in BASIL-3, told TCTMD that while trials like this are necessary for making determinations from a cost-effectiveness standpoint, and notably it forged ahead despite facing two major disruptions including from COVID-19, he wasn’t surprised to see numerous seasoned investigators in the audience challenging the study team following the presentation—or surprised that they often seemed perplexed by the answers.

“Although the investigators thought it was well powered, it was looking at an endpoint that isn't particularly sensitive for the impact of drug-coated device therapy,” Secemsky said. “That outcome combines amputation, which is related, and death, which can be unrelated to treatment. I think that it would be nice to have seen a larger trial driven by direct endpoints related to what drug-coated therapy offers [because] we know that it is important for delaying repeat intervention.”

Similarly, William A. Gray, MD (Main Line Health/Lankenau Heart Institute, Wynnewood, PA), who spoke from the audience after the presentation, said he, too, was concerned about the endpoint selection. Although he wouldn‘t expect to see an impact on mortality from drug-coated devices, Gray noted, repeat intervention and other considerations like major amputation “might have been a more salient endpoint given the dominance of death in this population.”

Bradbury responded that, based on his group’s experience, “I don't think I agree with your premise that what procedure you do for these folks has no bearing upon mortality.”

That may be true when surgery, which could impact mortality, is part of the comparison, Gray said. But “when I'm talking about across-the-board endovascular choice between antiproliferative [devices] and nonproliferative [POBA], I have no expectation of [a] mortality effect,” he specified.

BASIL-3 Clinical and Cost-effectiveness Results

Patients in the three treatment arms were well matched at baseline, with a median age of 72 years (35% female). More than half of the patients in all three arms had diabetes, approximately one-third had chronic kidney disease, and around 20% had tissue loss in their legs. More than 90% of patients in all three groups underwent revascularization within 2 weeks of randomization.

Over 7-year follow-up, amputation-free survival was 34% in the POBA arm, 40% in the DCB arm, and 42% in the DES arm. Comparing POBA to DCB yielded a HR of 0.84 (97.5% CI 0.61-1.16), while comparison of DES and POBA yielded a HR of 0.83 (97.5% CI 0.60-1.15).

The lower range of the confidence intervals narrowly missed or bordered the 40% reduction in amputation-free survival that the investigators had set as the target difference and for which the trial was powered. This led senior statistician Catherine Moakes, MSc (University of Birmingham), to note in her presentation of the results that “we cannot conclude that DCB or DES did not offer potentially smaller clinical benefits that BASIL-3 was not powered to detect.”

In a Kaplan-Meier analysis, there was some crossing of the curves, suggesting that the treatment effect may not be constant over time. An additional nonproportional analysis that allowed the hazard ratio to change over time showed that the difference between DCB versus POBA remained relatively constant; however, the hazard ratio for DES versus POBA, while similar in the early part of the study, evolved to favor DES.

Approximately 25% of patients in the DCB and DES groups were considered nonadherent due primarily to not receiving the allocated treatment. A separate analysis of the primary endpoint to account for this showed few differences from the intention-to-treat results.

The median 7-year survival rates were 40% in the POBA group, 44% in the DCB group, and 50% in the DES group. That yielded hazard ratios of 0.86 for DCB versus POBA (97.5% CI 0.61-1.20) and 0.79 for DES versus POBA (97.5% CI 0.56-1.11). Moakes again stressed the wide confidence intervals, with additional analyses showing varying treatment effects over time suggesting better outcomes with both DCB and DES versus POBA at all time points, but with a high degree of uncertainty at early and late time points due to few events or few patients at risk.

Major amputation occurred in 14% of the POBA group, 11% of the DCB group, and 16% of the DES group, with the cumulative incidence being lower overall in the DCB group from 1 year through the entirety of follow-up.

Rates of major adverse limb events were 37%, 35%, and 36%, respectively for POBA, DCB, and DES. For DCB versus POBA, the relative risk was 0.92 (97.5% CI 0.66-1.28) and for DES versus POBA the RR was 0.98 (97.5% CI 0.71-1.34).

Rates of MACE occurred in 31%, 39%, and 38% of the POBA, DCB, and DES groups, respectively. This yielded relative risks for DCB versus POBA of 1.28 (97.5% CI 0.90.1.80) and for DES versus POBA of 1.29 (97.5% CI 0.91-1.83).

For the economic analysis presented by Jesse Kigozi, PhD (University of Birmingham), the BASIL-3 senior health economist, there were minimal differences in costs and outcomes among the three groups. However, compared with POBA, treatment with DCB was less costly by about £250, but with wide confidence intervals. Accounting for quality-adjusted life-years (QALY) showed DCB to be less effective than POBA by -0.007. Neither difference was statistically significant. At the NICE willingness-to-pay threshold of £20,000, the probability of a DCB strategy being more cost effective than POBA was 52%.

DES, on the other hand, were less costly by about £700 compared with POBA and resulted in additional QALY of 0.048. The probability of a DES strategy compared with POBA being cost-effective was 76% at the current threshold and 64% if the threshold was £15,000.

Many Questions Remain

In addition to the delay in procedures and follow-up caused by the COVID-19 pandemic, BASIL-3 was one of the trials that was stopped during what has now become known as the “Katsanos pause,” a reference to the publication of a controversial 2018 meta-analysis that led to a period where paclitaxel-coated DES and DCB for peripheral interventions were taken out of use around the world because of safety concerns. Ultimately, the mortality signal within the meta-analysis was refuted by numerous data sets, including a patient-level, pooled meta-analysis with 5-year follow-up that showed no evidence of an increased risk of death.

Martin Björck, MD, PhD (Uppsala University, Sweden), who was in the audience during the presentation, spoke up to say he was surprised to see such small numbers of patients randomized into BASIL-3 given that the SWEDEPAD trial was able to randomize over 2,000 patients from a small country. Björck was the lead investigator of the SWEDEPAD series of trials, which were halted at the same time as BASIL-3.

“We have struggled to get colleagues in the UK to recruit patients,” Bradbury responded, added a plea for British surgeons to enroll their patients in clinical trials.

“In the end, we have to be realistic about what we can achieve given the size of the country, given the willingness to randomize, given the relatively low use of these drug devices in the UK following NICE, and also within the funding envelope,” he added.

Björck also questioned whether some of the deaths in all three arms of the trial could have been COVID-related, which could potentially dilute the power of the study, noting that different countries registered these deaths very differently during the pandemic.

“You're quite right, there may be patients there that we've missed,” Bradbury said. “But I don't see any obvious reason why that number of deaths should be different in the three arms.”

Björck further suggested that the investigators consider combining the DCB and DES groups and comparing them with POBA, something that Bradbury and colleagues say is planned.

Thomas Zeller, MD, PhD (Universitäts-Herzzentrum Freiburg-Bad Krozingen, Germany), who also spoke from the audience, issued another challenge, saying: “I’m not sure if you have been asking the right question.” CLTI, he noted, is far more complex than what is happening in the superficial femoropopliteal artery (SFA).

“So, the question remains what is revascularization of the femoral segment adding to limb outcome if you have impaired outflow? Meaning, is the isolated improvement of inflow really affecting limb outcome in CLTI?” he said. “If not, then it would not be surprising that whatever you do in the SFA does not really change the outcome.”

Holden, meanwhile, found fault with the cost-effectiveness argument favoring DES. “In simple terms, the up-front cost is more expensive,” he said. “That can only be offset by a lower intervention rate or hospitalization, and yet we were told that the intervention rates were no different. This is one thing that, I must say, doesn't follow all the other evidence.”

According to Kigozi, “the confidence intervals are so wide the cost could have gone whichever direction and in some of the subgroup analyses we've looked at, it does slightly cross over on the other side, by maybe £100 pounds or £50 pounds.”

Lastly, several people, including Secemsky, took issue with whether the Katsanos meta-analysis had a direct impact on the trial and its participants.

“You have to imagine that patients who came back into that trial [after it was paused] were likely somewhat different,” Secemsky said. “It’s not at all unreasonable that people willing to be randomized into a trial like this, being told what they were told about the Katsanos meta-analysis, probably were not the same as those in the pre-Katsanos era.”  

The BASIL-3 results should be confirmed in larger populations “before any real conclusions can be made,” he added, “because drug therapy has little risk to the patient and only can benefit, even if that means reducing the time to needing another intervention.”