The importance of assessing clinical trial data in an objective, balanced way has been borne out by recent history in the vascular surgery space, according to Ziv Haskal (University of Virginia, Charlottesville, USA). This is particularly pertinent given the latest drug-coated balloon (DCB) breakthroughs, including three-year data from the IN.PACT AV Access trial (Medtronic), presented for the first time at the 2022 Charing Cross (CX) Symposium (26–28 April, London, UK). Haskal also warns against “optimistic exuberance” and emphasises the need to remain grounded in “durable, hard reality” when drawing conclusions from these studies.
Speaking to Renal Interventions at the Society of Interventional Radiology (SIR) annual scientific meeting (11–16 June, Boston, USA), Haskal noted: “Industry trials are always designed to show positivity—and there is nothing wrong with that, because of the substantial stakes and costs of testing a new therapy, and a desire to define the most rarefied cohort where success and signal can be discerned. But the problem is that we naturally extrapolate the resulting data to everybody.
“And, we know that, when you continue to replicate things, signals becomes weaker over time as populations broaden or undergo further testing, or control therapies do better than expected.” On discrepancies between perceptibly positive IN.PACT AV Access results and the more maligned Lutonix AV trial (BD), Haskal outlines the possibility that—while comparatively better outcomes in the former have been attributed to drug-device superiority—the enrolled IN.PACT AV Access population may have been “honed” more effectively to show a certain signal.
The flipside of this, Haskal continues, is that heterogenous studies that are not industry-sponsored, such as the UK-based PAVE trial, carry “natural variability”. The trial, in which BD played no role other than providing the paclitaxel-coated balloons used, ultimately failed to provide evidence of additional benefit with DCBs following plain balloon angioplasty in patients with arteriovenous fistula (AVF) restenosis.
Haskal believes the design of PAVE, including its less selective patient population and the fact high-pressure plain balloon fistuloplasties were performed in all cases, could have contributed to the DCB group not outperforming its comparator. “It does not degrade or diminish the UK trial,” he adds, “but, rather, it asks the question: Are we now starting to replicate [DCB data] in a controlled fashion, but with more representative populations?”
Discrepancies in quality between the plain balloon angioplasties that DCBs are being compared to in such trials are also an important factor to consider, according to Haskal, who says the optimal combination of proper balloon sizing, inflation time, and possibly the use of high-pressure balloons, “may, in fact, provide better results”. “Was high-quality, conventional balloon angioplasty more consistently performed in the PAVE trial?” he queries.
Lessons from history
Here, Haskal draws on a previous instance of clinical data being more indicative of study design than drug-device performance. “The original sirolimus-eluting stent, Cypher [Cordis], was supposed to be the benchmark that would put drug-eluting stents onto the market,” he says. “Why did it fail? Because results with the bare metal stents it was trialled against were better than expected. The conventional therapy simply performed better.”
Haskal points out that, while limus drugs are now ironically beginning to “circle back around” in vascular access— for example, with 12-month results from the ISABELLA trial (MedAlliance), which were also presented at CX 2022—they may have “lost ground” due to factors that were unique to that trial, and there being no “appetite” to test them again at the time. He also likens the real-world variability that appears to have impacted PAVE to a similar phenomenon seen in the ATTRACT trial, which “did not kill” deep-vein thrombosis (DVT) interventions but instead led to further, more concentrated studies.
“I think the theme here is the importance of not simply jumping on the newest upgrade or latest model,” Haskal notes. “The challenge is that, having spent a long time in the field and having been through these cycles of seeing multiple generations of devices evolve, disappear and then return, you risk being viewed as recalcitrant, or resistant to embracing new technologies. Newer generations will always jump on that—but we still have to educate on [and exercise] a natural, daily cynicism and scepticism with clinical trials.”
Haskal also concludes by highlighting that, while DCB safety is now considered a “foregone conclusion” by many, the evidence base for this may be flawed, as patient numbers in studies like IN.PACT AV Access are likely too small to detect or rule out increased cardiovascular risks, “as we do not know their potential effect size”.