Advertisement

Long Term Outcomes After Revascularisations Below the Knee with Paclitaxel Coated Devices: A Propensity Score Matched Cohort Analysis

Open AccessPublished:August 14, 2020DOI:https://doi.org/10.1016/j.ejvs.2020.06.033

      Objective

      Endovascular revascularisation has become a standard approach for below knee lesions and paclitaxel coated devices have been widely used in patients with chronic limb threatening ischaemia. A recent meta-analysis reported higher mortality in paclitaxel coated devices compared with uncoated devices in femoropopliteal lesions. This study aimed to determine long term outcomes in below the knee interventions using paclitaxel coated devices in routine vascular care using a large and contemporary cohort.

      Methods

      A large cohort was created using all inclusive health insurance claims data of patients covered by the second largest insurance fund in Germany. The cohort included patients with index revascularisation of arteries below the knee performed from 1 January 2010, to 31 December 2018. Only patients with first paclitaxel coated device exposure were included. The study cohort was stratified into balloon vs. stent treatment and patients with paclitaxel coated devices were matched with uncoated devices using propensity score. Outcomes were evaluated using the Kaplan–Meier method and Cox regression.

      Results

      There were 14 738 patients (mean age 77.6 years, 43.6% female) and 6 568 matched patients included in the study. Increasing use of paclitaxel coated devices was observed during the study period (6% in 2010 vs. 31% in 2018, p < .001), and a total of 2 611 (39.8%) deaths occurred within five years of follow up. In the propensity score matched Cox model, a paclitaxel related reduction of five year mortality (hazards ratio, HR 0.84, 95% confidence interval, CI 0.78–0.91), amputation or death (HR 0.87, 95% CI 0.81–0.94), and cardiovascular event or death (HR 0.86, 95% CI 0.80–0.92) were observed.

      Conclusion

      In this propensity score matched cohort, reduced long term all cause mortality, reduced rates of amputation or death and cardiovascular event or death were observed at five years after the use of paclitaxel coated devices when compared with uncoated devices for the treatment of chronic limb threatening ischaemia.

      Keywords

      In this retrospective cohort study of 14 738 patients and 6 568 propensity score matched patients with index revascularisation below the knee between 1 January 2010 and 31 December 2018, a reduction was observed in long term all cause mortality and the combined endpoints of amputation or death and cardiovascular event or death five years after the use of paclitaxel coated devices when compared with uncoated devices for the treatment of chronic limb threatening ischaemia. The study addressed a key question in vascular medicine and using long term real world evidence does not confirm the potential harm reported in randomised controlled trials.

      Introduction

      More than 200 million patients worldwide and 40 million in Europe are affected by peripheral arterial occlusive disease (PAOD).
      • Fowkes F.G.
      • Rudan D.
      • Rudan I.
      • Aboyans V.
      • Denenberg J.O.
      • McDermott M.M.
      • et al.
      Comparison of global estimates of prevalence and risk factors for peripheral artery disease in 2000 and 2010: a systematic review and analysis.
      Chronic limb threatening ischaemia (CLTI) marks the end stage of PAOD or diabetic foot syndrome leading to high amputation rates, morbidity, and mortality, and is commonly caused by severe multilevel atherosclerosis including the below knee (BK) segment. Endovascular revascularisation for infrapopliteal disease has evolved as a primary approach in high risk CLTI patients.
      • Conte M.S.
      • Bradbury A.W.
      • Kolh P.
      • White J.V.
      • Dick F.
      • Fitridge R.
      • et al.
      Global vascular guidelines on the management of chronic limb-threatening ischemia.
      ,
      • Feldman D.N.
      • Armstrong E.J.
      • Aronow H.D.
      • Gigliotti O.S.
      • Jaff M.R.
      • Klein A.J.
      • et al.
      SCAI consensus guidelines for device selection in femoral-popliteal arterial interventions.
      Nevertheless, outcomes of percutaneous transluminal angioplasty (PTA) with or without bare metal stenting (BMS) remain suboptimal,
      • Mustapha J.A.
      • Finton S.M.
      • Diaz-Sandoval L.J.
      • Saab F.A.
      • Miller L.E.
      Percutaneous transluminal angioplasty in patients with infrapopliteal arterial disease: systematic review and meta-analysis.
      ,
      • Aboyans V.
      • Ricco J.B.
      • Bartelink M.E.L.
      • Bjorck M.
      • Brodmann M.
      • Cohnert T.
      • et al.
      Editor's choice - 2017 ESC guidelines on the diagnosis and treatment of peripheral arterial diseases, in collaboration with the European society for vascular surgery (ESVS).
      especially in BK lesions, because of the high incidence of restenosis and occlusion caused by neointimal hyperplasia. To diminish the restenosis rate, drug eluting balloons (DEB) and stents (DES) have been invented using the antiproliferative effect of paclitaxel on vascular smooth muscle and endothelial cell proliferation.
      • Jing L.
      • Peng X.
      • Xie M.J.
      • Yu Z.Y.
      • Wang W.
      Different responses of cell cycle between rat vascular smooth muscle cells and vascular endothelial cells to paclitaxel.
      Since the first publications in the early 2000s, several randomised controlled trials (RCT) on paclitaxel coated devices (PCX) in peripheral vascular interventions (PVI) have been conducted. They involved mainly femoropopliteal
      • Tepe G.
      • Schnorr B.
      • Albrecht T.
      • Brechtel K.
      • Claussen C.D.
      • Scheller B.
      • et al.
      Angioplasty of femoral-popliteal arteries with drug-coated balloons: 5-year follow-up of the THUNDER trial.
      • Werk M.
      • Albrecht T.
      • Meyer D.R.
      • Ahmed M.N.
      • Behne A.
      • Dietz U.
      • et al.
      Paclitaxel-coated balloons reduce restenosis after femoro-popliteal angioplasty: evidence from the randomized PACIFIER trial.
      • Candy N.
      • Ng E.
      • Velu R.
      Paclitaxel-coated balloon reduces target lesion revascularization compared with standard balloon angioplasty.
      but also BK lesions,
      • Zeller T.
      • Beschorner U.
      • Pilger E.
      • Bosiers M.
      • Deloose K.
      • Peeters P.
      • et al.
      Paclitaxel-coated balloon in infrapopliteal arteries: 12-month results from the BIOLUX P-II randomized trial (BIOTRONIK'S-first in man study of the passeo-18 LUX drug releasing PTA balloon catheter vs. the uncoated passeo-18 PTA balloon catheter in subjects requiring revascularization of infrapopliteal arteries).
      • Zeller T.
      • Baumgartner I.
      • Scheinert D.
      • Brodmann M.
      • Bosiers M.
      • Micari A.
      • et al.
      Drug-eluting balloon versus standard balloon angioplasty for infrapopliteal arterial revascularization in critical limb ischemia: 12-month results from the IN.PACT DEEP randomized trial.
      • Liistro F.
      • Porto I.
      • Angioli P.
      • Grotti S.
      • Ricci L.
      • Ducci K.
      • et al.
      Drug-eluting balloon in peripheral intervention for below the knee angioplasty evaluation (DEBATE-BTK): a randomized trial in diabetic patients with critical limb ischemia.
      giving promising results regarding target lesion revascularisation and major amputation rate. Interestingly, there were more complications in the drug eluting arm of one trial leading to serious concerns over higher amputation rates.
      • Zeller T.
      • Baumgartner I.
      • Scheinert D.
      • Brodmann M.
      • Bosiers M.
      • Micari A.
      • et al.
      Drug-eluting balloon versus standard balloon angioplasty for infrapopliteal arterial revascularization in critical limb ischemia: 12-month results from the IN.PACT DEEP randomized trial.
      The remarkable rise of paclitaxel coated balloons and stents was shattered in 2018 by a meta-analysis of summary level trial data presenting increased all cause mortality two and five years following application of PCX in femoropopliteal arteries.
      • Katsanos K.
      • Spiliopoulos S.
      • Kitrou P.
      • Krokidis M.
      • Karnabatidis D.
      Risk of death following application of paclitaxel-coated balloons and stents in the femoropopliteal artery of the leg: a systematic review and meta-analysis of randomized controlled trials.
      While the FDA analyses confirmed the findings of the meta-analysis, several studies using real world data from registries or health insurance claims data either reported none or even a positive influence of PCX on long term outcomes.
      • Secemsky E.A.
      • Kundi H.
      • Weinberg I.
      • Jaff M.R.
      • Krawisz A.
      • Parikh S.A.
      • et al.
      Association of survival with femoropopliteal artery revascularization with drug-coated devices.
      • Freisinger E.
      • Koeppe J.
      • Gerss J.
      • Goerlich D.
      • Malyar N.M.
      • Marschall U.
      • et al.
      Mortality after use of paclitaxel-based devices in peripheral arteries: a real-world safety analysis.
      • Krawisz A.K.
      • Secemsky E.A.
      Paclitaxel-based devices for the treatment of PAD: balancing clinical efficacy with possible risk.
      • Behrendt C.A.
      • Sedrakyan A.
      • Peters F.
      • Kreutzburg T.
      • Schermerhorn M.
      • Bertges D.J.
      • et al.
      Long term survival after femoropopliteal artery revascularisation with paclitaxel coated devices: a propensity score matched cohort analysis.
      More recently, a second meta-analysis of summary level trial data reiterated safety concerns of PCX for BK application in patients suffering from chronic limb threatening ischaemia.
      • Katsanos K.
      • Spiliopoulos S.
      • Kitrou P.
      • Krokidis M.
      • Paraskevopoulos I.
      • Karnabatidis D.
      Risk of death and amputation with use of paclitaxel-coated balloons in the infrapopliteal arteries for treatment of critical limb ischemia: a systematic review and meta-analysis of randomized controlled trials.
      Against the backdrop of this discussion, another controversy concerned a possible association between PCX and increased amputation rates in BK CLTI patients.
      • Zeller T.
      • Baumgartner I.
      • Scheinert D.
      • Brodmann M.
      • Bosiers M.
      • Micari A.
      • et al.
      Drug-eluting balloon versus standard balloon angioplasty for infrapopliteal arterial revascularization in critical limb ischemia: 12-month results from the IN.PACT DEEP randomized trial.
      However, neither the involved RCT nor the meta-analysis was initially powered to investigate this issue in the longer term.
      The current study sought to determine the long term outcomes of BK interventions using PCX in CLTI patients with large scale population based health insurance claims data. The study hypothesis was that PCX is associated with an increase in long term overall mortality, amputation, or death, and cardiovascular events or death in real world data from Germany.

      Methods

       BARMER cohort

      The longitudinal data of Germany's second largest insurance fund, BARMER, includes the outpatient and inpatient medical care provided to approximately 9.4 million German citizens (13.2% of Germany's population) involving 6.2 million hospitalisations between 2010 and 2018. The BARMER cohort includes nationally generalisable data with comparable gender and age distribution to the entire German population and has been widely used for cardiovascular research. The database contains longitudinal information for each person including date of birth, start and end of insurance episodes, and date of death until 31 December 2018. A regular random sample validation of internal and external validity is performed by the Medical Service of the Health Funds (MDK) in Germany, and various validation studies have been published before.
      • Czwikla J.
      • Jobski K.
      • Schink T.
      The impact of the lookback period and definition of confirmatory events on the identification of incident cancer cases in administrative data.
      • Hoffmann F.
      • Andersohn F.
      • Giersiepen K.
      • Scharnetzky E.
      • Garbe E.
      Validation of secondary data. Strengths and limitations.
      • Ohlmeier C.
      • Langner I.
      • Hillebrand K.
      • Schmedt N.
      • Mikolajczyk R.
      • Riedel O.
      • et al.
      Mortality in the German Pharmacoepidemiological Research Database (GePaRD) compared to national data in Germany: results from a validation study.
      • Langner I.
      • Ohlmeier C.
      • Zeeb H.
      • Haug U.
      • Riedel O.
      Individual mortality information in the German Pharmacoepidemiological Research Database (GePaRD): a validation study using a record linkage with a large cancer registry.
      The International Classification of Diseases in its German Modification (ICD-10-GM) was used to identify diagnoses and Operations and Procedures Codes (OPS) coding to identify procedures. The German OPS code is adapted to the International Classification of Procedures in Medicine (ICPM). For identifying medical prescriptions, the German version of the international Anatomical Therapeutic Chemical (ATC) classification was used. This study is part of a larger project on outcomes of PAOD patients after revascularisation. Further details regarding this cohort can be found in a published study protocol on optimal pharmacological treatment and the ongoing claims study (clinicaltrials.gov NCT03909022).
      • Peters F.
      • Kreutzburg T.
      • Kuchenbecker J.
      • Debus E.
      • Marschall U.
      • L'Hoest H.
      • et al.
      A retrospective cohort study on the provision and outcomes of pharmacological therapy after revascularization for peripheral arterial occlusive disease: a study protocol.
      The precise analyses in this paper were not pre-specified.

       Study population

      A cohort of patients with CLTI (Fontaine stages III to IV) in the crural segment below the knee was created. All patients aged ≥ 40 years were included with an endovascular PVI comprising stent/balloon revascularisation in the crural arteries. The index hospitalisation for PVI was from 1 January 2010 to 31 December 2018, with follow up until 31 December 2018. Five year lookback was used in the BARMER dataset to create relevant comorbidities (available ICD-10-GM data going back to 2005) and to ensure first paclitaxel exposure (available procedure codes for drug coated devices going back to 2008).
      Patients were included whose primary diagnosis was CLTI (I70.22–24 until 2014 and I70.23–25 since 2015) or CLTI as secondary diagnosis in combination with a primary diagnosis of diabetic foot syndrome (E10.50–51, E10.7, E11.50–51, E11.7), other peripheral vascular diseases (I73), arterial embolism and thrombosis (I74), cellulitis of finger and toe including acute lymphangitis (L03.01–02, L03.11), or chronic ulcer of skin and gangrene (L98.4, R02) using the ICD-10-GM.
      Patients who received at least one index drug coated balloon/stent during the study period were assigned to the paclitaxel group. If the patient received a stent and balloon at the same time, this was defined as a stent procedure. Additional information and coding criteria for DES or DCB and identification of CLTI patients by Fontaine stage can be found in the Table S1. Patients with hybrid interventions (open surgical repair [OSR] and PVI) exposed to paclitaxel before 2010 or patients who received a major amputation prior to the index stay or patients with missing information on age, sex, and follow up (∼0.48%) were excluded using complete case deletion. The non-paclitaxel group included plain balloon angioplasty (PBA) and bare metal stent (BMS). Patients with PBA or BMS at index stay but exposed to paclitaxel (e.g., DCB or DES) at later revascularisation of the lower limbs were assigned to the paclitaxel group according to the date of the second procedure. If patients were exposed to consecutive treatments using paclitaxel coated devices during the study period, the initial application of paclitaxel (first exposure) was used as index treatment.
      For the baseline characteristics, the comorbidity groups were categorised by ICD-10-GM codes separated into 30 Elixhauser comorbidity groups
      • Quan H.
      • Sundararajan V.
      • Halfon P.
      • Fong A.
      • Burnand B.
      • Luthi J.C.
      • et al.
      Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data.
      ,
      • Elixhauser A.
      • Steiner C.
      • Harris D.R.
      • Coffey R.M.
      Comorbidity measures for use with administrative data.
      during five years before the first PAOD diagnosis (lookback). The linear van Walraven score (vWS) is a weighted sum score ranging from −19 to +89 based on the Elixhauser groups and has been validated, wherein high scores represent a higher risk of in hospital death.
      • van Walraven C.
      • Austin P.C.
      • Jennings A.
      • Quan H.
      • Forster A.J.
      A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data.

       Statistical analysis

      Baseline characteristics of the patients were summarised with means and standard deviations (SD) for normally distributed variables, and medians and interquartile ranges (IQR) for non-normally distributed variables. For discrete variables, percentages and relative risk differences were used including 95% Wald CI (significant if 0 outside the interval).
      The t test and Wilcoxon rank sum tests were used to test for differences between exposure groups, and the Cochrane Armitage test was used for trend test of the proportion of paclitaxel usage. The discharge year, age, sex, number of prior hospital admissions and PAOD related outpatient admissions, number of different prescriptions during the prior year, and prescription for antithrombotics, antiplatelets, oral anticoagulants, dual antithrombotic therapy (antiplatelet and anticoagulant), lipid lowering drugs, antihypertensives, antidiabetics, analgesics, hypnotics and sedatives, antidementia, antidepressives during the prior year, all the Elixhauser coding groups (except AIDS/HIV), and prior stroke or transient ischaemic attack (TIA), dyslipidaemia, coronary artery disease, smoking, and prior myocardial infarction were used as variables for propensity score (PS) matching.
      The primary outcome was all cause mortality with the end of the follow up in December 2018. Secondary outcomes were the composite endpoints amputation or death (major amputation or all cause mortality) and cardiovascular event or death (myocardial infarction, stroke or TIA, and all cause mortality). Follow up times were censored after five years to compute robust five year rates. All outcomes were estimated using Cox proportional hazards models stratified by stent and balloon treatment. All Cox models were additionally adjusted for age, sex, Fontaine stage, prior myocardial infarction, congestive heart failure, cardiac arrhythmias, valvular disease, hypertension, neurodegenerative disorders, chronic pulmonary disease, hypothyroidism, renal failure, liver disease, metastatic cancer, solid tumour without metastasis, obesity, fluid and electrolyte disorders, deficiency anaemia, psychosis, depression, lipid lowering drugs, antithrombotics, antiplatelets, oral anticoagulants, dual antithrombotic therapy (antiplatelet and anticoagulant), antihypertensives, antidiabetics, antidementia during year before admission, number of different prescriptions during year before admission, number of previous inpatient admissions total (incl. index), and number of prior PAOD outpatient admissions.
      As a sensitivity analysis, Cox models were computed without additional regression adjustment, used an intention to treat approach for assigning study groups based on the first revascularisation only, excluded patients switching from non-paclitaxel revascularisation to paclitaxel revascularisation during follow up, excluded patients experiencing any outcome events within 30 days after hospital discharge (landmark analysis), and adjusting also for post-discharge medication therapy (Tables S4 and S5). Further sensitivity models excluded patients with concomitant PCX devices in the femoropopliteal or abdominal and pelvic arteries, excluded patients treated in centres with mean annual revascularisation volume above 200 cases, adjusted the model also for log mean annual revascularisation volume, stratified the sample by low and high volume at the median of 68 cases annually and stratified the sample by low and high rate of PCX (ratio of all PCX cases and all cases) using 0.14 as the threshold (Table S6).
      The primary exposure of interest was any index (first) PCX application. PAOD severity specific PS matching (greedy 1:1 matching) was applied to adjust for observed confounding. To measure the validity of the matching algorithm, standardised differences were used (values above 0.1 or below −0.1 deemed to indicate meaningful differences). Details of the logistic regression model and a comparison of quality for PS matching are reported in Tables S2 and S3.
      All analyses were performed with software SAS version 9.04 (SAS Institute, NC, USA). Results were reported using the REporting of studies Conducted using Observational Routinely collected health Data (RECORD) statement,
      • Benchimol E.I.
      • Smeeth L.
      • Guttmann A.
      • Harron K.
      • Hemkens L.G.
      • Moher D.
      • et al.
      [The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement].
      the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) statement,
      • von Elm E.
      • Altman D.G.
      • Egger M.
      • Pocock S.J.
      • Gotzsche P.C.
      • Vandenbroucke J.P.
      • et al.
      The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.
      and following international recommendations on medical device evaluation studies.
      • Sedrakyan A.
      • Cronenwett J.L.
      • Venermo M.
      • Kraiss L.
      • Marinac-Dabic D.
      • Bjorck M.
      An international vascular registry infrastructure for medical device evaluation and surveillance.
      Cox regression models and visualisations were performed with software R version 3.3.3 (The R Foundation for Statistical Computing, Vienna, Austria). Illustrations were designed using Adobe Illustrator version 24.0.1 (Adobe Systems Software Ireland Ltd., Dublin, Republic of Ireland).

      Results

      This study included a total of 14 738 patients with CLTI hospitalised for BK revascularisation during the study period with average age 77.6 ± 9.9 years (43.6% females), and 6 568 PS matched patients undergoing PVI from 1 January 2010 to 31 December 2018 (Fig. 1). An increasing proportion of PCX was observed during the study period (6% in 2010 vs. 31% in 2018, p < .001).
      Figure 1
      Figure 1Flow chart of this propensity score matched retrospective analysis of health insurance claims to identify the study cohort of patients with peripheral arterial occlusive disease (PAOD) and chronic limb threatening ischaemia treated endovascularly with plain balloon angioplasty (PBA), bare metal stent (BMS), drug coated balloon (DCB) or drug eluting stent (DES) to determine long term outcomes of below knee interventions using paclitaxel coated devices in routine vascular care.

       Unmatched analyses for demographics and comorbidities

      The median follow up was 606 days (interquartile range from 214 to 1 295 days, 3.5 years). The longest follow up was 8 years. Among all CLTI patients, 2 192 (14.9%) suffered from rest pain and 12 546 (85.1%) had been revascularised for wound healing disorders. There was a prior diagnosis of stroke or TIA in 1934 patients (13.1%), 6 792 patients (46.1%) had a prior diagnosis of coronary artery disease, 2 283 (15.5%) patients had prior myocardial infarction, and 9 319 (63.2%) had previously been diagnosed with diabetes. The median van Walraven comorbidity score was 12 (5–21) and the median length of hospital stay was 11 days (5–20). The baseline characteristics of the entire unmatched cohort are presented in Table 1. When compared with the control group, patients exposed to PCX received a higher mean number of medical prescriptions during one year before admission (14.5 vs. 13.8, p < .001), including the prescription of any antithrombotic medication (62.5% vs. 57.1%), and lipid lowering drugs (49.7% vs. 43.7%).
      Table 1Baseline characteristics of the entire unmatched study cohort of this retrospective analysis of health insurance claims of patients with chronic limb threatening ischaemia (CLTI) treated endovascularly for arterial lesions below the knee. The patients were assigned to the paclitaxel (PCX) group if they were treated with drug eluting devices vs. control patients never exposed to drug eluting devices during the observational period
      Characteristics of the unmatched study cohortsPCX (n = 3 284)Control (n = 11 454)p value
      p values at α = .05 were considered statistically significant.
      or RD (95% CI)
      Age – y77.37 ± 9.7777.66 ± 9.94.14
      No. of prescriptions during one year prior to admission14.50 ± 6.9613.81 ± 6.73<.001
      No. of prior hospitalisations5.91 ± 4.505.31 ± 4.14<.001
      Median follow up time (IQR) – d539.50 (213.00, 1112.50)627.00 (214.25, 1346.00)<.001
      Female sex1 529 (46.6)4 901 (42.8)−3.77 (−5.7 – −1.84)
      Significant values at α = .05.
      van Walraven Score > 92 009 (61.2)7 032 (61.4)0.22 (−1.67–2.11)
      Prior stroke or TIA411 (12.5)1 523 (13.3)0.78 (−0.51–2.07)
      Smoking314 (9.6)957 (8.4)−1.21 (−2.33 – −0.08)
      Significant values at α = .05.
      Dyslipidaemia1 902 (57.9)5 967 (52.1)−5.82 (−7.74 – −3.9)
      Significant values at α = .05.
      Obesity658 (20.0)2 335 (20.4)0.35 (−1.21–1.9)
      Prescription of lipid-lowering drugs during one year prior to admission1 632 (49.7)5 003 (43.7)−6.02 (−7.95 – −4.08)
      Significant values at α = .05.
      Hypertension3 000 (91.4)10 273 (89.7)−1.66 (−2.77 – −0.55)
      Significant values at α = .05.
      Coronary artery disease1 577 (48.0)5 215 (45.5)−2.49 (−4.43 – −0.55)
      Significant values at α = .05.
      Prior myocardial infarction540 (16.4)1 743 (15.2)−1.23 (−2.65–0.2)
      Congestive heart failure1 503 (45.8)5 269 (46.0)0.23 (−1.7–2.17)
      Valvular disease678 (20.6)2 293 (20.0)−0.63 (−2.19–0.94)
      Cardiac arrhythmias1 537 (46.8)5 265 (46.0)−0.84 (−2.77–1.1)
      Antithrombotic drug prescriptions during one year prior to admission
       Antithrombotics2 052 (62.5)6 543 (57.1)−5.36 (−7.25 – −3.47)
      Significant values at α = .05.
       Antiplatelets1 464 (44.6)4 008 (35.0)−9.59 (−11.5 – −7.68)
      Significant values at α = .05.
       Oral anticoagulation1 035 (31.5)3 327 (29.0)−2.47 (−4.26 – −0.68)
      Significant values at α = .05.
       Antiplatelets and oral anticoagulation378 (11.5)812 (7.1)−4.42 (−5.61 – −3.23)
      Significant values at α = .05.
      Chronic pulmonary disease545 (16.6)2 038 (17.8)1.2 (−0.26–2.65)
      Diabetes, complicated1 718 (52.3)6 235 (54.4)2.12 (0.18–4.06)
      Prescription of antidiabetics during one year prior to admission1 632 (49.7)5 853 (51.1)1.4 (−0.54–3.34)
      Renal failure1 859 (56.6)6 204 (54.2)−2.44 (−4.37 – −0.52)
      Significant values at α = .05.
      Liver disease208 (6.3)767 (6.7)0.36 (−0.59–1.31)
      Data are provided as n (%) or mean ± standard deviation unless stated otherwise. CLTI = chronic limb threatening ischaemia; PCX = paclitaxel coated devices; RD = risk difference with 95% Wald confidence intervals (CI); IQR = interquartile range; TIA = transient ischaemic attack.
      p values at α = .05 were considered statistically significant.
      Significant values at α = .05.

       Demographics in matched cohorts

      Demographics and comorbidities of the comparison groups after PS matching are presented in Table 2. The median follow up was longer in the paclitaxel group compared with the controls (540 vs. 472 days, p < .001).
      Table 2Baseline characteristics of the matched study cohort of this retrospective analysis of health insurance claims of patients with chronic limb threatening ischaemia (CLTI) treated endovascularly for arterial lesions below the knee. The patients were assigned to the paclitaxel (PCX) group if they were treated with drug eluting devices vs. control patients never exposed to drug eluting devices during the observational period
      Characteristics of the matched study cohortsPCX (n = 3 284)Control (n = 3 284)p value
      p values at α = .05 were considered statistically significant. No data analysed for RD reached statistical significance.
      or RD (95% CI)
      Age – y77.37 ± 9.7777.30 ± 9.90.79
      No. prescriptions during one year prior to admission14.50 ± 6.9614.34 ± 6.77.37
      No. of prior hospitalisations5.91 ± 4.505.84 ± 4.73.50
      Median follow up time (IQR) – d539.50

      (213.00, 1112.50)
      471.50

      (175.00, 995.00)
      <.001
      Female sex1 529 (46.6)1 502 (45.7)−0.82 (−3.23–1.59)
      van Walraven Score > 92 009 (61.2)2 003 (61.0)−0.18 (−2.54–2.18)
      Prior stroke or TIA411 (12.5)430 (13.1)0.58 (−1.04–2.19)
      Smoking314 (9.6)318 (9.7)0.12 (−1.3–1.55)
      Dyslipidaemia1 902 (57.9)1 920 (58.5)0.55 (−1.84–2.93)
      Obesity658 (20.0)666 (20.3)0.24 (−1.7–2.18)
      Prescription of lipid lowering drugs during one year prior to admission1 632 (49.7)1 582 (48.2)−1.52 (−3.94–0.9)
      Hypertension3 000 (91.4)3 013 (91.7)0.4 (−0.95–1.74)
      Coronary artery disease1 577 (48.0)1 567 (47.7)−0.3 (−2.72–2.11)
      Prior myocardial infarction,540 (16.4)516 (15.7)−0.73 (−2.51–1.05)
      Congestive heart failure1 503 (45.8)1 523 (46.4)0.61 (−1.8–3.02)
      Valvular disease678 (20.6)656 (20.0)−0.67 (−2.62–1.28)
      Cardiac arrhythmias1 537 (46.8)1 477 (45.0)−1.83 (−4.24–0.58)
      Antithrombotic drug prescriptions during one year prior to admission
       Antithrombotics2 052 (62.5)2 020 (61.5)−0.97 (−3.32–1.37)
       Antiplatelets1 464 (44.6)1 468 (44.7)0.12 (−2.28–2.53)
       Oral anticoagulation1 035 (31.5)1 000 (30.5)−1.07 (−3.3–1.17)
       Antiplatelets and oral anticoagulation378 (11.5)366 (11.1)−0.37 (−1.9–1.17)
      Chronic pulmonary disease545 (16.6)537 (16.4)−0.24 (−2.04–1.55)
      Diabetes, complicated1 718 (52.3)1 723 (52.5)0.15 (−2.26–2.57)
      Prescription of antidiabetics during one year prior to admission1 632 (49.7)1 624 (49.5)−0.24 (−2.66–2.17)
      Renal failure1 859 (56.6)1 847 (56.2)−0.37 (−2.76–2.03)
      Liver disease208 (6.3)222 (6.8)0.43 (−0.77–1.62)
      Data are provided as n (%) or mean ± standard deviation unless stated otherwise. CLTI = chronic limb threatening ischaemia; PCX = paclitaxel coated devices; RD = risk difference with 95% Wald confidence intervals (CI); TIA = transient ischaemic attack.
      p values at α = .05 were considered statistically significant. No data analysed for RD reached statistical significance.

       Primary and secondary outcome in matched cohorts

      The short and long term outcomes are presented in Table 3. Only the length of stay was significantly different between the comparison groups. Patients exposed to PCX experienced a shorter median length of hospital stay (9 vs. 10 days, p < .001). A total of 2 611 (39.8%) deaths occurred within the five years of follow up in the matched cohort.
      Table 3Peri-operative (in hospital) outcomes and long term event rates (death, myocardial infarction, major amputation) of the matched study cohort of this retrospective analysis of health insurance claims of patients with chronic limb threatening ischaemia (CLTI) treated endovascularly for arterial lesions below the knee. The patients were assigned to the paclitaxel (PCX) group if they were treated with drug eluting devices vs. control patients never exposed to drug eluting devices during the observational period
      Peri-operative and long term outcomesPCX (n = 3 284)Control (n = 3 284)p value
      p values at α = .05 were considered statistically significant.
      or RD (95% CI)
      Median length of stay (IQR) – d9.00 (5.00, 17.00)10.00 (5.00, 19.00)<.001
      Length of stay ≥ 8 d1 919 (58.4)2 137 (65.1)6.64 (4.29–8.98)
      Significant values at α = .05.
      Acute limb ischaemia258 (7.9)228 (6.9)−0.91 (−2.18–0.35)
      Bleeding282 (8.6)260 (7.9)−0.67 (−2 – 0.66)
      Acute myocardial infarction34 (1.0)40 (1.2)0.18 (−0.33–0.69)
      Acute renal failure241 (7.3)241 (7.3)0 (−1.26–1.26)
      Acute respiratory failure87 (2.6)101 (3.1)0.43 (−0.38–1.23)
      Post-operative delirium64 (1.9)81 (2.5)0.52 (−0.19–1.23)
      Pneumonia73 (2.2)81 (2.5)0.24 (−0.49–0.98)
      Stroke or TIA13 (0.4)9 (0.3)−0.12 (−0.4–0.16)
      Sepsis or SIRS39 (1.2)49 (1.5)0.3 (−0.25–0.86)
      In hospital death71 (2.2)90 (2.7)0.58 (−0.17–1.33)
      Discharged to rehabilitation29 (0.9)40 (1.2)0.33 (−0.16–0.83)
      Discharged to nursing home103 (3.1)114 (3.5)0.33 (−0.53–1.2)
      Death within 30 d155 (4.7)181 (5.5)0.79 (−0.27–1.86)
      Death within 5 y1 256 (38.2)1 355 (41.3)3.01 (0.65–5.38)
      Significant values at α = .05.
      Major amputation within 5 y251 (7.6)214 (6.5)−1.13 (−2.37–0.11)
      Myocardial infarction within 5 y325 (9.9)332 (10.1)0.21 (−1.24–1.66)
      Stroke or TIA within 5 y303 (9.2)286 (8.7)−0.52 (−1.9–0.86)
      Data are provided as n (%) unless stated otherwise. CLTI = chronic limb threatening ischaemia; PCX = paclitaxel coated devices; RD = risk difference with 95% Wald confidence intervals (CI); TIA = transient ischaemic attack; SIRS = systemic inflammatory response syndrome.
      p values at α = .05 were considered statistically significant.
      Significant values at α = .05.

       Cox proportional hazards analyses in matched cohorts

      The results of the Cox proportional hazards models for all cause mortality, major amputation or death, and cardiovascular events or death after five years are reported in Fig. 2. In PS matched cohorts, a paclitaxel related reduction of five year all cause mortality was observed for patients treated with PCX in the combined DCB and DES analysis (HR 0.84, 95% CI 0.78–0.91) but also separately for DCB (HR 0.84, 95% CI 0.77–0.92) and DES (HR 0.76, 95% CI 0.61–0.94).
      Figure 2
      Figure 2Forest plots of all cause mortality (A), cardiovascular events or death (B), and amputation or death (C) after five years by treatment approach by drug coated balloon (DCB) or drug eluting stent (DES) in patients with chronic limb threatening ischaemia in the below knee arteries using propensity score matched cohorts with hazards ratios and 95% confidence intervals (balloon vs. stent vs. both approaches merged together).
      Treatment with DCB was associated with a lower risk of cardiovascular events or death (HR 0.86, 95% CI 0.79–0.93) also seen in the combined DCB and DES analysis (HR 0.86, 95% CI 0.80–0.92) but not for DES alone.
      Amputation or death occurred significantly less often in PCX in the combined analysis (HR 0.87, 95% CI 0.81–0.94) but also separately for DCB (HR 0.89, 95% CI 0.82–0.96) and DES (HR 0.74, 95% CI 0.60–0.92) (Fig. 2).
      The Kaplan–Meier curves and log rank test results for all cause mortality, major amputation or death, and cardiovascular events or death after five years are presented in Fig. 3. In these analyses using PS matched cohorts, treatment with paclitaxel coated devices was significantly associated with better outcomes when compared with treatment with uncoated devices.
      Figure 3
      Figure 3Cumulative Kaplan–Meier estimates of all cause mortality (A), cardiovascular events or death (B), and amputation or death (C) after five years by endovascular treatment approach with (red line) or without (blue line) paclitaxel (PCX) containing device in patients with chronic limb threatening ischaemia of the below knee arteries using propensity score matched cohorts including 95% Wald confidence interval and log rank test (p value).

       Sensitivity analysis

      The results were robust throughout the sensitivity analyses (see Tables S5 and S6). Effect size was generally smaller and less often significant in the intention to treat design perhaps as a result of the smaller sample size (n = 4 830) and less strict assignment of study groups according to PCX exposure. The same was true when stratifying the sample by hospital volume and PCX rate. Estimates for balloons and stents combined varied between HR 0.83 and HR 0.94 for all cause mortality, HR 0.85 and 0.96 for amputation or death, and between 0.86 and 0.97 for cardiovascular event or death.

      Discussion

      In this PS matched retrospective analysis of health insurance claims, an overall high long term mortality was observed after five years but no indication of increased risk of mortality, amputation or death and cardiovascular events or death after treatment with PCX when compared with uncoated devices.
      In Cox proportional hazards models, a paclitaxel related reduction of all cause mortality, amputation, or death, and cardiovascular event or death could be observed in the unstratified analyses involving both DCB and DES. In stratified analyses, only DES was not associated with a reduction in cardiovascular event or death.
      This non-specific benefit of PCX was confirmed by Kaplan–Meier analyses. These results emphasise that PCX is safe both for DES and DCB.
      After the implementation of PCX for PAOD treatment, various prospective trials led to their widespread use in the United States (US) and Europe.
      • Zeller T.
      • Noory E.
      • Beschorner U.
      • Bohme T.
      • Mpj Reijnen M.
      Outstanding effectiveness of paclitaxel-based technologies for the treatment of femoropopliteal artery occlusive disease on the potential expense of increased late all-cause mortality? No reason to panic.
      This trend was confirmed in this cohort as the proportion of PCX use significantly increased during the study period (30% PCX use in 2018). In 2018, a meta-analysis of PCX in femoropopliteal lesion reported increased all cause mortality in summary level data, which unleashed a global debate on possible harms with consecutive regulatory actions in the US and Europe. One major point of criticism concerning this meta-analysis, besides an insufficient involvement of CLTI and BK treatment, was that several RCTs included in the meta-analysis were underpowered and had no adequate long term follow up. To overcome this limitation, various real world studies have been published and the results in this field remain diametrically opposed.
      • Secemsky E.A.
      • Kundi H.
      • Weinberg I.
      • Jaff M.R.
      • Krawisz A.
      • Parikh S.A.
      • et al.
      Association of survival with femoropopliteal artery revascularization with drug-coated devices.
      ,
      • Krawisz A.
      • Secemsky E.A.
      Real-world data collection regarding paclitaxel treatment.
      • Behrendt C.-A.
      • Sedrakyan A.
      • Peters F.
      • Kreutzburg T.
      • Schermerhorn M.
      • Bertrges D.J.
      • et al.
      Long term survival after femoropopliteal artery revascularizations with paclitaxel-coated devices – a propensity score matched cohort analysis.
      • Bertges D.J.
      • Sedrakyan A.
      • Sun T.
      • Eslami M.H.
      • Schermerhorn M.
      • Goodney P.P.
      • et al.
      Mortality after paclitaxel coated balloon angioplasty and stenting of superficial femoral and popliteal artery in the vascular quality initiative.
      For PCX in BK lesions and CLTI, there currently are only a few RCTs available with low participant numbers and almost completely missing long term data, leading to a larger gap of evidence when compared with the treatment of femoropopliteal lesions. The RCTs comparing paclitaxel coated balloons with standard PTA were BIOLUX P-II,
      • Zeller T.
      • Beschorner U.
      • Pilger E.
      • Bosiers M.
      • Deloose K.
      • Peeters P.
      • et al.
      Paclitaxel-coated balloon in infrapopliteal arteries: 12-month results from the BIOLUX P-II randomized trial (BIOTRONIK'S-first in man study of the passeo-18 LUX drug releasing PTA balloon catheter vs. the uncoated passeo-18 PTA balloon catheter in subjects requiring revascularization of infrapopliteal arteries).
      IN.PACT DEEP,
      • Zeller T.
      • Baumgartner I.
      • Scheinert D.
      • Brodmann M.
      • Bosiers M.
      • Micari A.
      • et al.
      Drug-eluting balloon versus standard balloon angioplasty for infrapopliteal arterial revascularization in critical limb ischemia: 12-month results from the IN.PACT DEEP randomized trial.
      DEBATE-BTK,
      • Liistro F.
      • Porto I.
      • Angioli P.
      • Grotti S.
      • Ricci L.
      • Ducci K.
      • et al.
      Drug-eluting balloon in peripheral intervention for below the knee angioplasty evaluation (DEBATE-BTK): a randomized trial in diabetic patients with critical limb ischemia.
      Lutonix BTK,
      • Mustapha J.A.
      • Brodmann M.
      • Geraghty P.J.
      • Saab F.
      • Settlage R.A.
      • Jaff M.R.
      • et al.
      Drug-coated vs uncoated percutaneous transluminal angioplasty in infrapopliteal arteries: six-month results of the Lutonix BTK trial.
      as well as a trial by Haddad et al. including a total of 1097 patients.
      • Haddad S.E.
      • Shishani J.M.
      • Qtaish I.
      • Rawashdeh M.A.
      • Qtaishat B.S.
      One year primary patency of infrapopliteal angioplasty using drug- eluting balloons: single center experience at king Hussein medical center.
      Twelve month mortality was reported as 9.4% vs. 6.0% (DEB vs. PTA, p = .58) in BIOLUX P-II, 10.1% vs. 8.1% (DEB vs. PTA, p = .55) in IN.PACT DEEP, and 7.7% vs. 4.5% (DEB vs. PTA, p = .40) in DEBATE-BTK. Lutonix BTK reported a freedom from all cause mortality at 180 days of 96.8% vs. 96% (DEB vs. PTA, p = .70), whereas all cause mortality presented by Haddad et al. was 20.8% vs. 13.3% for DEB vs. PTA 2016. A meta-analysis of RCTs showed no significant difference regarding major adverse events (29.0% in DEB vs. 38.8% in PTA, p = .48) at 12 month follow up.
      • Wu R.
      • Tang S.
      • Wang M.
      • Li Z.
      • Yao C.
      • Wang S.
      Drug-eluting balloon versus standard percutaneous transluminal angioplasty in infrapopliteal arterial disease: a meta-analysis of randomized trials.
      Similarly, a Cochrane analysis from 2016 including femoropopliteal and BK interventions did not find any significant difference for amputation or death in a subgroup analysis of BK and CLTI patients at 12 month follow up.
      • Kayssi A.
      • Al-Atassi T.
      • Oreopoulos G.
      • Roche-Nagle G.
      • Tan K.T.
      • Rajan D.K.
      Drug-eluting balloon angioplasty versus uncoated balloon angioplasty for peripheral arterial disease of the lower limbs.
      In contrast, this PS matched cohort revealed a PCX related reduction of all cause mortality, lower risk of amputation or death, and lower risk of cardiovascular events or death during long term follow up. Besides non-random assignment to treatment, a reason for these conflicting results when compared with RCT data and meta-analyses might be the larger sample size and consequently higher statistical power to determine longer term outcomes after PCX interventions comprised in the current analysis. The observed long term survival benefit in the PCX group vs. uncoated devices could be explained by the different quality of medical therapy or other unobserved variables. Against that backdrop, the positive impact of statins is well known as one of the main pillars in optimal medical treatment with a positive impact on long term outcomes.
      • Reynolds K.
      • Mues K.E.
      • Harrison T.N.
      • Qian L.
      • Chen S.
      • Hsu J.Y.
      • et al.
      Trends in statin utilization among adults with severe peripheral artery disease including critical limb ischemia in an integrated healthcare delivery system.
      • Subherwal S.
      • Patel M.R.
      • Kober L.
      • Peterson E.D.
      • Jones W.S.
      • Gislason G.H.
      • et al.
      Missed opportunities: despite improvement in use of cardioprotective medications among patients with lower-extremity peripheral artery disease, underuse remains.
      • De Martino R.R.
      • Eldrup-Jorgensen J.
      • Nolan B.W.
      • Stone D.H.
      • Adams J.
      • Bertges D.J.
      • et al.
      Perioperative management with antiplatelet and statin medication is associated with reduced mortality following vascular surgery.
      Hence, valid societal guidelines clearly recommend the permanent prescription of statins after lower limb revascularisations.
      • Aboyans V.
      • Ricco J.B.
      • Bartelink M.E.L.
      • Bjorck M.
      • Brodmann M.
      • Cohnert T.
      • et al.
      Editor's choice - 2017 ESC guidelines on the diagnosis and treatment of peripheral arterial diseases, in collaboration with the European society for vascular surgery (ESVS).
      ,
      • Venermo M.
      • Sprynger M.
      • Desormais I.
      • Bjorck M.
      • Brodmann M.
      • Cohnert T.
      • et al.
      Editor's choice - follow-up of patients after revascularisation for peripheral arterial diseases: a consensus document from the European society of cardiology working group on aorta and peripheral vascular diseases and the European society for vascular surgery.
      In this study, higher rates of statin use and other important drugs groups were observed (see Table S4) among PCX cases both during the year before admission and during the month after. Adjustments were made only for these differences before admission and at 30 days after discharge in the landmark cohort to avoid correlation of medical therapy and outcome. Future studies may tackle the hypothesis of whether better long term management with respect to medical therapy may explain the advantage of PCX cases using a time dependent design.
      Besides, PCX might have a positive impact on survival of unknown cause, that is first seen after more than 12 months and currently, there are no further long term data available comparing uncovered devices with PCX BK.
      The overall mortality after five years reported in the current study, independent of the device, is remarkably high and most probably an expression of advanced age, severe multisite artery disease, and multiple comorbidities including coronary and cerebrovascular disease in CLTI patients. Estimated five year survival rates in the PADI trial,
      • Spreen M.I.
      • Martens J.M.
      • Knippenberg B.
      • van Dijk L.C.
      • de Vries J.P.M.
      • Vos J.A.
      • et al.
      Long term follow-up of the PADI trial: percutaneous transluminal angioplasty versus drug-eluting stents for infrapopliteal lesions in critical limb ischemia.
      which compared BK BMS vs. DES in 137 CLTI patients were comparable to the current study results (37.0% BMS vs. 37.7% in the DES group).
      Besides severe comorbidities, amputation itself is a relevant risk factor for death in PAOD patients. Klaphake et al. found overall mortality after major amputation to be 44%, 66%, and 85% after one, three, and five years.
      • Klaphake S.
      • de Leur K.
      • Mulder P.G.
      • Ho G.H.
      • de Groot H.G.
      • Veen E.J.
      • et al.
      Mortality after major amputation in elderly patients with critical limb ischemia.
      Interestingly, the current analysis found a significantly better amputation free survival after five years in the PCX group compared with uncoated devices. Correspondingly, the PADI trial also reported a higher amputation free survival in the DES group (31.8% vs. 20.4%, p = .043).
      • Spreen M.I.
      • Martens J.M.
      • Knippenberg B.
      • van Dijk L.C.
      • de Vries J.P.M.
      • Vos J.A.
      • et al.
      Long term follow-up of the PADI trial: percutaneous transluminal angioplasty versus drug-eluting stents for infrapopliteal lesions in critical limb ischemia.
      In stratified analysis, DES but not DCB was associated with a lower risk of cardiovascular events or death and amputation or death. Literature comparing paclitaxel coated balloons with paclitaxel coated stents in BK lesions is scarce. While the meta-analysis by Katsanos et al. did not detect a relationship between paclitaxel and amputation risk,
      • Katsanos K.
      • Spiliopoulos S.
      • Diamantopoulos A.
      • Karnabatidis D.
      • Sabharwal T.
      • Siablis D.
      Systematic review of infrapopliteal drug-eluting stents: a meta-analysis of randomized controlled trials.
      a more recent review reported a harm signal with respect to amputation free survival (13.7% vs. 9.4%, p = .080).
      • Katsanos K.
      • Spiliopoulos S.
      • Kitrou P.
      • Krokidis M.
      • Paraskevopoulos I.
      • Karnabatidis D.
      Risk of death and amputation with use of paclitaxel-coated balloons in the infrapopliteal arteries for treatment of critical limb ischemia: a systematic review and meta-analysis of randomized controlled trials.
      The authors presumed that major adverse events, that is limb and also systemic, might be caused by non-target paclitaxel embolisation. The results of the current study do not confirm this safety concern for more general or for more specific outcomes.
      In summary, the current study could demonstrate that PCX in BK lesions is associated with a reduction in long term all cause mortality, amputation, or death, and cardiovascular event or death in CLTI patients. Further prospective trials with appropriate statistical power are needed to illuminate the various technical and morphological aspects of BK revascularisations and their impact on long term outcomes in a more detailed way.
      This study has limitations. It is possible that the experience of the centre performing the procedures and the follow up of those patients play a major role in increasing the overall survival of the target population. It cannot be ruled out that a considerable proportion of the improved outcomes in paclitaxel exposed patients are explained by unobserved patient differentials or other aspects of healthcare. Furthermore, it was not possible to validly collect information regarding specific devices, doses of paclitaxel delivered, or paclitaxel applied during coronary or cancer treatment. There is a very small proportion of coatings other than paclitaxel, making a few of the codings used less specific. In addition, the primary purpose of the data collection should be considered when using it for secondary purposes, and all research data should undergo validation. Health insurance funds in Germany perform random cross checks with patient files on a regular basis. Prior validation studies revealed high validity of major outcomes such as mortality in health insurance claims data.
      • Czwikla J.
      • Jobski K.
      • Schink T.
      The impact of the lookback period and definition of confirmatory events on the identification of incident cancer cases in administrative data.
      • Hoffmann F.
      • Andersohn F.
      • Giersiepen K.
      • Scharnetzky E.
      • Garbe E.
      Validation of secondary data. Strengths and limitations.
      • Ohlmeier C.
      • Langner I.
      • Hillebrand K.
      • Schmedt N.
      • Mikolajczyk R.
      • Riedel O.
      • et al.
      Mortality in the German Pharmacoepidemiological Research Database (GePaRD) compared to national data in Germany: results from a validation study.
      • Langner I.
      • Ohlmeier C.
      • Zeeb H.
      • Haug U.
      • Riedel O.
      Individual mortality information in the German Pharmacoepidemiological Research Database (GePaRD): a validation study using a record linkage with a large cancer registry.
      In the current analysis, 53.8% of PCX cases were also revascularised at other sites than the infrapopliteal arteries during their index stay. Yet, the results were robust to the exclusion of these patients during sensitivity analyses. Further, patients with CLTI have competing risks for mortality that may mask the mortality effects of devices. Known confounders were controlled for in an effort to minimise this possibility. However, there are probably additional confounders not available in health insurance claims data, for example, more detailed information about disease severity or quality of medical therapy. Residual unobserved confounding might be a reason for the conflicting results between RCTs and observational studies. It must be highlighted that retrospective observational studies are merely hypothesis generating. Only a properly powered RCT would be suitable to validly confirm or refute signals for causative effects. Lastly, although the fee for service reimbursement system in Germany probably motivates interventionalists to perform inpatient procedures rather than outpatient procedures, there might be another target population not included in this study. The same limitation is valid for patients insured by other insurance providers or patients changing their health insurance company during the study period, although this is known to be uncommon in the target population of interest. However, it is believed that paclitaxel distribution is comparable.

      Conclusions

      In this PS matched cohort, a reduction was observed in long term all cause mortality, amputation, or death, and cardiovascular event or death five years after the use of PCX when compared with uncoated devices for the treatment of CLTI.

      Acknowledgements

      The authors are grateful for the kind support and submission of data from the BARMER.

      Conflict of interest

      ESD has received funding by Bayer and Terumo . The other authors declare no conflicts of interest.

      Funding

      The presented analyses were part of the IDOMENEO project (Funded by the German Joint Federal Committee ; grant-no. 01VSF16008 ). The GermanVasc research group is additionally supported by the German Stifterverband and the CORONA Foundation (grant no. S199/10061/2015 ), by the German Joint Federal Committee (RABATT study; grant-no. 01VSF18035 ), and receives research funds from Bayer Vital GmbH, from B. Braun Melsungen AG , and from the Foundation of the University Heart and Vascular Centre Hamburg (UHZ). The funders had no influence on design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. AS received funding from US Food and Drug Administration through grant U01FD005478 . The funder had no influence on design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Fowkes F.G.
        • Rudan D.
        • Rudan I.
        • Aboyans V.
        • Denenberg J.O.
        • McDermott M.M.
        • et al.
        Comparison of global estimates of prevalence and risk factors for peripheral artery disease in 2000 and 2010: a systematic review and analysis.
        Lancet. 2013; 382: 1329-1340
        • Conte M.S.
        • Bradbury A.W.
        • Kolh P.
        • White J.V.
        • Dick F.
        • Fitridge R.
        • et al.
        Global vascular guidelines on the management of chronic limb-threatening ischemia.
        Eur J Vasc Endovasc Surg. 2019; 58 (e33): S1-S109
        • Feldman D.N.
        • Armstrong E.J.
        • Aronow H.D.
        • Gigliotti O.S.
        • Jaff M.R.
        • Klein A.J.
        • et al.
        SCAI consensus guidelines for device selection in femoral-popliteal arterial interventions.
        Catheter Cardiovasc Interv. 2018; 92: 124-140
        • Mustapha J.A.
        • Finton S.M.
        • Diaz-Sandoval L.J.
        • Saab F.A.
        • Miller L.E.
        Percutaneous transluminal angioplasty in patients with infrapopliteal arterial disease: systematic review and meta-analysis.
        Circ Cardiovasc Interv. 2016; 9e003468
        • Aboyans V.
        • Ricco J.B.
        • Bartelink M.E.L.
        • Bjorck M.
        • Brodmann M.
        • Cohnert T.
        • et al.
        Editor's choice - 2017 ESC guidelines on the diagnosis and treatment of peripheral arterial diseases, in collaboration with the European society for vascular surgery (ESVS).
        Eur J Vasc Endovasc Surg. 2018; 55: 305-368
        • Jing L.
        • Peng X.
        • Xie M.J.
        • Yu Z.Y.
        • Wang W.
        Different responses of cell cycle between rat vascular smooth muscle cells and vascular endothelial cells to paclitaxel.
        J Huazhong Univ Sci Technolog Med Sci. 2014; 34: 370-375
        • Tepe G.
        • Schnorr B.
        • Albrecht T.
        • Brechtel K.
        • Claussen C.D.
        • Scheller B.
        • et al.
        Angioplasty of femoral-popliteal arteries with drug-coated balloons: 5-year follow-up of the THUNDER trial.
        JACC Cardiovasc Interv. 2015; 8: 102-108
        • Werk M.
        • Albrecht T.
        • Meyer D.R.
        • Ahmed M.N.
        • Behne A.
        • Dietz U.
        • et al.
        Paclitaxel-coated balloons reduce restenosis after femoro-popliteal angioplasty: evidence from the randomized PACIFIER trial.
        Circ Cardiovasc Interv. 2012; 5: 831-840
        • Candy N.
        • Ng E.
        • Velu R.
        Paclitaxel-coated balloon reduces target lesion revascularization compared with standard balloon angioplasty.
        J Vasc Surg. 2017; 65 (e10): 558-570
        • Zeller T.
        • Beschorner U.
        • Pilger E.
        • Bosiers M.
        • Deloose K.
        • Peeters P.
        • et al.
        Paclitaxel-coated balloon in infrapopliteal arteries: 12-month results from the BIOLUX P-II randomized trial (BIOTRONIK'S-first in man study of the passeo-18 LUX drug releasing PTA balloon catheter vs. the uncoated passeo-18 PTA balloon catheter in subjects requiring revascularization of infrapopliteal arteries).
        JACC Cardiovasc Interv. 2015; 8: 1614-1622
        • Zeller T.
        • Baumgartner I.
        • Scheinert D.
        • Brodmann M.
        • Bosiers M.
        • Micari A.
        • et al.
        Drug-eluting balloon versus standard balloon angioplasty for infrapopliteal arterial revascularization in critical limb ischemia: 12-month results from the IN.PACT DEEP randomized trial.
        J Am Coll Cardiol. 2014; 64: 1568-1576
        • Liistro F.
        • Porto I.
        • Angioli P.
        • Grotti S.
        • Ricci L.
        • Ducci K.
        • et al.
        Drug-eluting balloon in peripheral intervention for below the knee angioplasty evaluation (DEBATE-BTK): a randomized trial in diabetic patients with critical limb ischemia.
        Circulation. 2013; 128: 615-621
        • Katsanos K.
        • Spiliopoulos S.
        • Kitrou P.
        • Krokidis M.
        • Karnabatidis D.
        Risk of death following application of paclitaxel-coated balloons and stents in the femoropopliteal artery of the leg: a systematic review and meta-analysis of randomized controlled trials.
        J Am Heart Assoc. 2018; 7e011245
        • Secemsky E.A.
        • Kundi H.
        • Weinberg I.
        • Jaff M.R.
        • Krawisz A.
        • Parikh S.A.
        • et al.
        Association of survival with femoropopliteal artery revascularization with drug-coated devices.
        JAMA Cardiol. 2019; 4: 332-340
        • Freisinger E.
        • Koeppe J.
        • Gerss J.
        • Goerlich D.
        • Malyar N.M.
        • Marschall U.
        • et al.
        Mortality after use of paclitaxel-based devices in peripheral arteries: a real-world safety analysis.
        Eur Heart J. 2019; ([Epub ahead of print])https://doi.org/10.1093/eurheartj/ehz698
        • Krawisz A.K.
        • Secemsky E.A.
        Paclitaxel-based devices for the treatment of PAD: balancing clinical efficacy with possible risk.
        Curr Treat Options Cardiovasc Med. 2019; 21: 57
        • Behrendt C.A.
        • Sedrakyan A.
        • Peters F.
        • Kreutzburg T.
        • Schermerhorn M.
        • Bertges D.J.
        • et al.
        Long term survival after femoropopliteal artery revascularisation with paclitaxel coated devices: a propensity score matched cohort analysis.
        Eur J Vasc Endovasc Surg. 2020; 59: 587-596
        • Katsanos K.
        • Spiliopoulos S.
        • Kitrou P.
        • Krokidis M.
        • Paraskevopoulos I.
        • Karnabatidis D.
        Risk of death and amputation with use of paclitaxel-coated balloons in the infrapopliteal arteries for treatment of critical limb ischemia: a systematic review and meta-analysis of randomized controlled trials.
        J Vasc Interv Radiol. 2020; 31: 202-212
        • Czwikla J.
        • Jobski K.
        • Schink T.
        The impact of the lookback period and definition of confirmatory events on the identification of incident cancer cases in administrative data.
        BMC Med Res Methodol. 2017; 17: 122
        • Hoffmann F.
        • Andersohn F.
        • Giersiepen K.
        • Scharnetzky E.
        • Garbe E.
        Validation of secondary data. Strengths and limitations.
        Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz. 2008; 51: 1118-1126
        • Ohlmeier C.
        • Langner I.
        • Hillebrand K.
        • Schmedt N.
        • Mikolajczyk R.
        • Riedel O.
        • et al.
        Mortality in the German Pharmacoepidemiological Research Database (GePaRD) compared to national data in Germany: results from a validation study.
        BMC Public Health. 2015; 15: 570
        • Langner I.
        • Ohlmeier C.
        • Zeeb H.
        • Haug U.
        • Riedel O.
        Individual mortality information in the German Pharmacoepidemiological Research Database (GePaRD): a validation study using a record linkage with a large cancer registry.
        BMJ Open. 2019; 9e028223
        • Peters F.
        • Kreutzburg T.
        • Kuchenbecker J.
        • Debus E.
        • Marschall U.
        • L'Hoest H.
        • et al.
        A retrospective cohort study on the provision and outcomes of pharmacological therapy after revascularization for peripheral arterial occlusive disease: a study protocol.
        BMJ Surg Interv Health Tec. 2020; 2e000020
        • Quan H.
        • Sundararajan V.
        • Halfon P.
        • Fong A.
        • Burnand B.
        • Luthi J.C.
        • et al.
        Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data.
        Med Care. 2005; 43: 1130-1139
        • Elixhauser A.
        • Steiner C.
        • Harris D.R.
        • Coffey R.M.
        Comorbidity measures for use with administrative data.
        Med Care. 1998; 36: 8-27
        • van Walraven C.
        • Austin P.C.
        • Jennings A.
        • Quan H.
        • Forster A.J.
        A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data.
        Med Care. 2009; 47: 626-633
        • Benchimol E.I.
        • Smeeth L.
        • Guttmann A.
        • Harron K.
        • Hemkens L.G.
        • Moher D.
        • et al.
        [The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement].
        Z Evid Fortbild Qual Gesundhwes. 2016; 115–116: 33-48
        • von Elm E.
        • Altman D.G.
        • Egger M.
        • Pocock S.J.
        • Gotzsche P.C.
        • Vandenbroucke J.P.
        • et al.
        The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.
        J Clin Epidemiol. 2008; 61: 344-349
        • Sedrakyan A.
        • Cronenwett J.L.
        • Venermo M.
        • Kraiss L.
        • Marinac-Dabic D.
        • Bjorck M.
        An international vascular registry infrastructure for medical device evaluation and surveillance.
        Eur J Vasc Endovasc Surg. 2017; 53: 600-602
        • Zeller T.
        • Noory E.
        • Beschorner U.
        • Bohme T.
        • Mpj Reijnen M.
        Outstanding effectiveness of paclitaxel-based technologies for the treatment of femoropopliteal artery occlusive disease on the potential expense of increased late all-cause mortality? No reason to panic.
        Vasa. 2019; 48: 109-111
        • Krawisz A.
        • Secemsky E.A.
        Real-world data collection regarding paclitaxel treatment.
        Endovasc Today. 2019; 18: 72-76
        • Behrendt C.-A.
        • Sedrakyan A.
        • Peters F.
        • Kreutzburg T.
        • Schermerhorn M.
        • Bertrges D.J.
        • et al.
        Long term survival after femoropopliteal artery revascularizations with paclitaxel-coated devices – a propensity score matched cohort analysis.
        Eur J Vasc Endovasc Surg. 2020; 59: 587-596
        • Bertges D.J.
        • Sedrakyan A.
        • Sun T.
        • Eslami M.H.
        • Schermerhorn M.
        • Goodney P.P.
        • et al.
        Mortality after paclitaxel coated balloon angioplasty and stenting of superficial femoral and popliteal artery in the vascular quality initiative.
        Circ Cardiovasc Interv. 2020; 13e008528
        • Mustapha J.A.
        • Brodmann M.
        • Geraghty P.J.
        • Saab F.
        • Settlage R.A.
        • Jaff M.R.
        • et al.
        Drug-coated vs uncoated percutaneous transluminal angioplasty in infrapopliteal arteries: six-month results of the Lutonix BTK trial.
        J Invasive Cardiol. 2019; 31: 205-211
        • Haddad S.E.
        • Shishani J.M.
        • Qtaish I.
        • Rawashdeh M.A.
        • Qtaishat B.S.
        One year primary patency of infrapopliteal angioplasty using drug- eluting balloons: single center experience at king Hussein medical center.
        J Clin Imaging Sci. 2017; 7: 31
        • Wu R.
        • Tang S.
        • Wang M.
        • Li Z.
        • Yao C.
        • Wang S.
        Drug-eluting balloon versus standard percutaneous transluminal angioplasty in infrapopliteal arterial disease: a meta-analysis of randomized trials.
        Int J Surg. 2016; 35: 88-94
        • Kayssi A.
        • Al-Atassi T.
        • Oreopoulos G.
        • Roche-Nagle G.
        • Tan K.T.
        • Rajan D.K.
        Drug-eluting balloon angioplasty versus uncoated balloon angioplasty for peripheral arterial disease of the lower limbs.
        Cochrane Database Syst Rev. 2016; CD011319
        • Reynolds K.
        • Mues K.E.
        • Harrison T.N.
        • Qian L.
        • Chen S.
        • Hsu J.Y.
        • et al.
        Trends in statin utilization among adults with severe peripheral artery disease including critical limb ischemia in an integrated healthcare delivery system.
        Vasc Med. 2020; 25: 3-12
        • Subherwal S.
        • Patel M.R.
        • Kober L.
        • Peterson E.D.
        • Jones W.S.
        • Gislason G.H.
        • et al.
        Missed opportunities: despite improvement in use of cardioprotective medications among patients with lower-extremity peripheral artery disease, underuse remains.
        Circulation. 2012; 126: 1345-1354
        • De Martino R.R.
        • Eldrup-Jorgensen J.
        • Nolan B.W.
        • Stone D.H.
        • Adams J.
        • Bertges D.J.
        • et al.
        Perioperative management with antiplatelet and statin medication is associated with reduced mortality following vascular surgery.
        J Vasc Surg. 2014; 59 (e1): 1615-1621
        • Venermo M.
        • Sprynger M.
        • Desormais I.
        • Bjorck M.
        • Brodmann M.
        • Cohnert T.
        • et al.
        Editor's choice - follow-up of patients after revascularisation for peripheral arterial diseases: a consensus document from the European society of cardiology working group on aorta and peripheral vascular diseases and the European society for vascular surgery.
        Eur J Vasc Endovasc Surg. 2019; 58: 641-653
        • Spreen M.I.
        • Martens J.M.
        • Knippenberg B.
        • van Dijk L.C.
        • de Vries J.P.M.
        • Vos J.A.
        • et al.
        Long term follow-up of the PADI trial: percutaneous transluminal angioplasty versus drug-eluting stents for infrapopliteal lesions in critical limb ischemia.
        J Am Heart Assoc. 2017; 6e004877
        • Klaphake S.
        • de Leur K.
        • Mulder P.G.
        • Ho G.H.
        • de Groot H.G.
        • Veen E.J.
        • et al.
        Mortality after major amputation in elderly patients with critical limb ischemia.
        Clin Interv Aging. 2017; 12: 1985-1992
        • Katsanos K.
        • Spiliopoulos S.
        • Diamantopoulos A.
        • Karnabatidis D.
        • Sabharwal T.
        • Siablis D.
        Systematic review of infrapopliteal drug-eluting stents: a meta-analysis of randomized controlled trials.
        Cardiovasc Intervent Radiol. 2013; 36: 645-658