Effects of Stents or Prostanoids on Critical Limb Ischemia

Reference: Hsu CCT, Kwan GNC, Singh D, Rophael JA, Anthony C, van Driel ML. Angioplasty versus stenting for infrapopliteal arterial lesions in chronic limb-threatening ischaemia. Cochrane Database Syst Rev. 2018;12:CD009195. doi:10.1002/14651858.CD009195.pub2

Rationale: The goals of CLI treatment are to alleviate ischemic rest pain, heal ischemic lesions, improve quality of life, prevent limb loss, and prolong life. Percutaneous transluminal angioplasty has temporarily improved below-knee circulation. It remains unclear as to whether adding an endovascular metal alloy stent would prolong or improve CLI outcomes after PTA.

Objective: The authors conducted a Cochrane review of the literature comparing CLI outcomes after infrapopliteal PTA with versus without stenting in patients with CLI. 

Methods: A Cochrane vascular specialist searched Cochrane Vascular Specialized Register, AMED, CENTRAL, MEDLINE, EMBASE, and CINAHL databases, plus World Health Organization International Clinical Trials Registry Platform and ClinicalTrials.gov trials registries to June 25, 2018, for randomized controlled trials (RCTs) and quasi-RCTs comparing clinical outcomes of patients with CLI who received PTA of their infrapopliteal arterial lesions using a balloon to expand the affected arterial lesion as compared with any form of stenting to expand a similar lesion. Studies of bare metal, drug-eluting, and bioabsorbable stents were included. Patients meeting the definition of CLI had ischemic rest pain, tissue loss, or gangrene. This included those with Fontaine stage III or Rutherford category 4 or 5 CLI with a stenotic (> 50% luminal loss) or occluded infrapopliteal artery, including tibiofibular trunk, anterior tibial artery, posterior tibial artery, or fibular artery. Two reviewers independently selected RCTs, rated their quality using GRADE criteria, and summarized extracted data with arbitration by a third reviewer as needed. Outcomes summarized included procedural technical success or complications, target artery patency, major amputation, and mortality. Odds ratios were calculated for the likelihood of each outcome reported for patients receiving PTA alone as compared with PTA plus a stent, with statistical significance set at P = .05.

Results: Among 7 studies qualifying for meta-analysis, the only statistically significant result for intent-to-treat comparisons was 3 times more likely technical success of the procedure when the stent was added to PTA (5 studies on 276 patients; P = .03). Ankle-to-brachial systolic pressure ratios (ABI) improved comparably from baseline to 6-month or 12-month values for both PTA and stent groups with no significant difference between groups. Only 1 study analyzed complete ulcer healing 12 months after PTA (n = 33) or stent placement (n = 21), reporting a higher percentage of ischemic ulcers healed in the PTA group (P = .006) in the 44 patients completing the study to 12 months. There were no consistent effects on primary patency, complications, amputations, or mortality across the RCTs. Quality of life was not measured in any study qualifying for inclusion. All studies reported all outcomes measured, but all had a high risk of bias due to the lack of blinding of participants and staff. Outcomes assessment was not blinded in 5 studies. Cardiovascular risk factors were more likely in those receiving stents in 1 study. In another study, patients receiving stents also received perioperative clopidogrel, while patients with PTAs did not. There was a higher incidence of PTAs per patient, as 1 balloon could be applied to more than 1 lesion site on a patient. These differences may have biased study results, reducing comparability of groups. Most RCTs reported data on patency for only 6 months, with only 2 RCTs extending this to 12 months. Inconsistent use of anticoagulant and/or antiplatelet medications within and across RCTs may have further confounded these results.

Authors’ Conclusions: There is insufficient high-quality evidence to suggest adding stent insertion to a standard PTA improves clinical outcomes as compared with a PTA without stenting. Stent insertion could be reserved for use when arterial dissection is encountered.  

Reference: Vietto V, Franco JVA, Saenz V, Cytryn D, Chas J, Ciapponi A. Prostanoids for critical limb ischaemia. Cochrane Database Syst Rev. 2018;1:CD006544. doi:10.1002/14651858.CD006544.pub3

Rationale: Unchecked CLI progresses to loss of the affected limb or patient mortality. Prostanoid drugs have reported activities on endothelial cells, vascular smooth muscle cells, and platelets that may improve the course of PAD or its end-stage progression to CLI. 

Objective: Conduct a literature search of RCTs supporting safety and efficacy of prostanoids for managing patients with CLI who were considered unsuitable for rescue or reconstructive interventions.

Methods: A Cochrane Vascular Information Specialist searched the Cochrane Central Register of Controlled Trials and the Cochrane Vascular Specialized Register, a weekly updated compendium of MEDLINE, CINAHL, EMBASE, and OVID databases as well as the Allied and Complementary Medicine Database through January 17 for RCTS and derivative references of at least 80% of participants with CLI of atherosclerotic origin who were treated with a prostanoid, including prostaglandin (PG) E₁_, I₂, iloprost, beraprost, cisaprost, ciptostene, clinprost, ecraprost, or taprostene, compared with a placebo or other pharmacologic agent. Studies of surgical or non-pharmacological medical comparators were excluded. Primary outcome measures were cardiovascular-related mortality, major or minor amputations, quality of life, or adverse events related to treatment. Secondary outcomes were ABI, rest pain, or analgesic drug use measured using a validated pain scale or questionnaire, ulcer healing quantified as decreased ulcer area, or presence/absence of granulation tissue, major amputations above or below the knee, minor amputations of partial feet or fingers, or all-cause mortality. Authors independently verified RCT inclusion from the full text of each relevant study, resolving disagreements by consensus. Meta-analysis was performed with P < .05, indicating statistical significance for overall safety and efficacy effects of prostanoids versus placebo and for prostanoid subsets even if the RCTs were not homogeneous. Results and risk of bias for individual studies were summarized and a quantitative synthesis of reported adverse events was prepared for each type of prostanoid. 

Results: There were 33 studies with a total of 4477 patients who qualified for meta-analysis. Only 8 RCTs were at a low risk of attrition bias, with less than a 10% rate for participant withdrawal before study end and intent-to-treat analysis. No consistent benefit was reported for prostanoids as compared with a placebo on cardiovascular or all-cause mortality. Those treated with iloprost experienced fewer total amputations (6 RCTs on 1205 patients; P = .02) or major amputations (5 RCTs on 1125 patients; P = .0079). These effects lacked statistical significance in the combined analysis of all prostanoids versus placebos. More adverse events were experienced by those receiving PGE₁ as compared with placebo (4 RCTs on 451 patients; P < .00001) or iloprost as compared with PGE₁ (4 RCTs on 511 patients; P < .00001) or as compared with placebo (3 RCTs on 378 patients; P < .00001). More patients receiving iloprost as compared with placebo experienced CLI-related rest pain relief (6 RCTs on 733 patients; P = .002), a result confirmed in 1 RCT reporting reduced analgesic use for patients receiving iloprost as compared with placebo (n = 100; P = .037). Ulcer healing was improved by prostanoids as compared with placebos (P = .019), a result driven mainly by 4 RCTs of iloprost (n = 632; P = .012) and 1 RCT of ciprostene (n = 211; P = .00075). Improved ulcer healing and rest pain with iloprost were no longer significant when only the 2 high-quality studies were analyzed.

Authors’ Conclusions: Low-quality evidence supported a lack of prostanoid efficacy as compared with placebos on the incidence of cardiovascular mortality. High-quality evidence suggested a similar absence of prostanoid effects on total amputations. Moderate-quality evidence supported rest pain and ulcer healing benefits of prostanoids compared with placebos, offset by increased adverse events associated with prostanoid use.

Both of these evidence summaries^6,7 described interventions with insufficient evidence that improves amputation and mortality outcomes for patients with CLI, though some non-blinded studies improved healing, rest pain, and major amputation rates. It is an important wake-up call that neither search returned a single RCT reporting patient quality of life, which is similar to that of terminal cancer patients.⁸ Are researchers measuring CLI outcomes that matter to patients?⁹ Another important insight these studies reveal is that a 1-treatment approach may not work for these patients with multiple challenges. It may be more fruitful to address all factors causing each patient’s CLI. For any given patient these may include nicotine use, diabetes mellitus, hyperlipidemia, or lack of calf muscle exercise. If uncontrolled, such factors become powerful sources of unidentified variability in any clinical trial, obscuring clinical efficacy of any one intervention. For example, phase III neuroischemic diabetic foot ulcer RCTs that do not control hyperglycemia, nicotine use, or consistency of study-foot offloading waste clinical resources, money, and patients’ time. Now that clinicians are beginning to understand that a univariate approach is inadequate for resolving CLI,⁵ Hsu et al⁶ took an important step toward identifying stenting as 1 intervention that lacks sufficient evidence of improved patient-centered outcomes when combined with PTA. Well-designed, unbiased RCTs are needed to determine whether the nonblinded visual assessments reporting improved technical success associated with stent placement support its addition to a standardized CLI protocol. Hsu et al⁶ also recommended harmonization of terms, definitions, and outcome measurements for CLI patients as well as rigorous control of adjuvant procedures, such as anticoagulant or antiplatelet use in RCTs. Sound research suggests a quality CLI study protocol should monitor, control, and analyze PAD risk factors, including hyperlipidemia, nicotine use, diabetes mellitus, cardiovascular comorbidities, and genetic factors. Standardized adjunctive calf muscle exercises can improve outcomes and merit inclusion in preoperative and postoperative CLI protocols.^5,10 Monitoring and resolving patient CLI risk factors are important aspects of vascular management that may improve consistency of results of CLI research and clinical practice.