Of the total study population of 357 PCI patients with concomitant PADs, 70% had invasive treatment for PADs while 30% was treated non-invasively. These two patient groups, which turned out not to differ in demographics, and clinical and PCI procedural characteristics, showed at 3-year follow-up no statistically significant difference in the primary clinical endpoint TVF (20.5% vs. 16.0%), the secondary endpoint MACE (23.3% vs. 20.4%), and all other secondary endpoints. All-cause mortality (12.1% vs. 8.3%) and cardiac mortality (7.4% vs. 2.8%) rates were numerically higher in study patients with previous invasive PADs treatment; yet, these dissimilarities did not reach statistical significance, possibly due to sample size. Highly detailed information on previous PADs treatment was available in 131 patients who also participated in a dedicated clinical PADs registry. Baseline characteristics and long-term outcome following PCI of patients who previously had invasive treatment for PADs were found to be similar in patients with an exclusively non-invasive treatment for PADs. Furthermore, the all-cause mortality rate was higher in patients with rest pain (Fontaine stage III) than in patients with a pain-free walking distance of more than 200 m (Fontaine stage IIa).

To the best of our knowledge, the present analysis is the first to assess the impact of the invasiveness of previous PADs treatment on long-term outcome after PCI. In former studies, various approaches of detecting and defining PADs have been applied [6,7,8,9, 13, 24, 25]. As a result, the distribution of treatment types for PADs differs between studies.

For instance, peripheral artery disease can be assessed by measuring the ankle-brachial index, but usually only 10–30% of all patients with a decreased ankle-brachial index report symptoms consistent with classic claudication [26]. When using the ankle-brachial index, it is fair to assume that 7–13% of the patients undergoing (coronary angiography or) PCI have undiagnosed PADs [24, 25]. Yet, if the presence of PADs is defined based on the assessment of medical records, only patients who currently are (or previously were) symptomatic are classified as PADs patients, while asymptomatic patients with abnormal ankle-brachial index are not. In clinical practice, when patients present with anginal symptoms, information about previous PADs treatment is generally available from the medical record, while measurements of the ankle-brachial index often are not.

Furthermore, the vascular regions included in the definition of PADs differ between studies. Many studies included not only atherosclerotic disease in the lower limb but also cerebrovascular disease [8, 10], aortic pathologies [6, 7, 27], or atherosclerotic disease in all non-coronary arteries except for the aorta [9, 13]. Such between-study differences in the definition of PADs result in substantial differences between the respective study populations, which may lead to dissimilar findings when assessing the potential impact of PADs on clinical outcomes after PCI.

Previously, PADs with a decreased ankle-brachial index were associated with an increased risk of cardiovascular and all-cause mortality, ischemic or hemorrhagic stroke, repeated target lesion revascularization, major bleeding, and major adverse cardiac and cerebrovascular events at follow-up with a duration of 1 to 4 years [24, 28, 29]. Even in patients with borderline decreased ankle-brachial index, higher event rates were observed 18 months after PCI for the composite endpoint of major adverse cardiac and cerebrovascular events as well as for stroke [30].

Studies that defined PADs based on data from the medical history (rather than ankle-brachial index-measurements) showed quite similar results [9, 11, 15, 24, 28, 29]. A previous analysis of PADs patients from a series of four randomized PCI trials found that patients with a history of PADs, based on medical records, had higher 3-year risks of target vessel failure, repeated target vessel revascularization, MACE, and all-cause mortality than patients without PADs [15. The findings of that analysis are in accordance with former studies that evaluated clinical outcomes after PCI in patients with PADs based on medical records. In those studies, the presence of PADs was associated with lower rates of procedural success [1, 9] and with higher in-hospital and 2-year rates of stroke, myocardial infarction, MACE, clinically relevant bleeding, and mortality [1, 6, 9,10,11,12, 27]. In addition, PCI patients with concomitant PADs had higher all-cause mortality rates at long-term follow-up after PCI [6, 10].

Hence, regardless of the way of defining PADs, PCI patients with PADs had a worse clinical outcome than those without PADs. Although the aforementioned studies included somewhat different patient populations with varying treatments for PADs, the presence of PADs had a similar, unfavorable impact on long-term clinical outcome following PCI. In the present study in PCI patients with concomitant PADs, patients with a history of invasive treatment had on average a higher (maximum) Fontaine stage than patients with exclusively non-invasive treatment for PADs, as may be expected. Yet, between PCI patients with invasive and non-invasive previous treatment for concomitant PADs, we found no significant difference in clinical outcome after PCI.

While PADs itself may directly account for some post-PCI cardiovascular events and mortality, the presence of PADs can be seen as an indication of greater plaque burden and more progressive atherosclerotic disease in the vasculature, including the coronary arteries [1, 2, 14]. The findings of the present study suggest that the presence of PADs –regardless of the invasiveness of previous PADs treatment– should be seen as a marker of an increased event risk after a PCI. Moreover, among PCI patients with concomitant PADs, patients with a higher Fontaine stage may show a higher mortality than PADs patients with a pain-free walking distance of more than 200 m (Fontaine stage IIa). While detailed information on PADs location and previous treatment is valuable when choosing a vascular access site for PCI, the present study shows that high-risk patients for PCI can be identified by straightforwardly consulting the medical records, searching for known PADs irrespective of the invasiveness of PADs treatment. Knowledge of these findings may be particularly useful when considering PCI during Heart Team discussions, and when informing patients about their individual adverse event risk.

The results of the present post-hoc study should be considered hypothesis-generating. Nevertheless, this study is the first to assess the potential impact of previous PADs treatment on clinical outcome after PCI with contemporary drug-eluting stents. Pooled individual patient-level data from four PCI all-comer trials were studied, as adverse event rates were relatively low. The relatively low adverse event rates may reflect progress in coronary stent design and concomitant pharmacological therapy (including antithrombotic strategy), rather than missing adverse events in these randomized clinical trials with high follow-up, external monitoring, and independent event adjudication. Of all PADs patients who underwent PCI at the Medisch Spectrum Twente (Enschede, the Netherlands), medical records were reviewed for details of their PADs treatment. Of patients included in the other hospitals detailed information about PADs was not available. In addition, information about PADs type, symptoms, and treatment was unavailable in some trial participants. In addition, the patient group with non-invasive treatment for PADs may be somewhat heterogeneous, as not only patients with mild PADs may have been included in this group but also some patients who were too frail to undergo invasive treatment. The collection of detailed data on anatomical PADs severity could have been of interest, but in many study patients such details were not available.

Furthermore, as a result of the PADs definition used, undiagnosed PADs have been missed. Assessing the ankle-brachial index in all PCI patients could have provided further insights. Yet, one should not expect from a randomized PCI trial that routine measurements of the ankle-brachial index are performed, especially not in the large-sized group of patients who underwent PCI for acute myocardial infarction. Moreover, such approach would not have reflected current routine clinical practice that all-comer PCI trials typically strive to emulate.