Limiting Amputation

Limiting the extent or incidence of lower limb amputations is among the highest reported patient priorities.¹ A series of systematic reviews and meta-analyses included amputation rates in individuals with chronic diabetic foot ulcers (DFUs) subjected to different types of offloading,² adjunctive therapies,³ or debridement methods.⁴ The latter study reported a meta-analysis of 4 randomized controlled trials (RCTs) supporting the efficacy of larval debridement in reducing the relative risk (RR) of amputation to 43% of those using standard wound care or surgical debridement (P < .05). However, the RCTs were of low quality, so readers are cautioned that debridement decisions should be “based on the available expertise, patient preferences, the clinical context, and cost.”⁴ This month’s Evidence Corner summarizes 2 recent systematic reviews and meta-analyses exploring the effects of topical interventions on lower limb amputation: one exploring the effects of autologous cell therapy,⁵ the other on negative pressure wound therapy (NPWT).⁶ 

Reference: Rigato M, Monami M, Fadini GP.  Autologous cell therapy for peripheral arterial disease: systematic review and meta-analysis of randomized, non-randomized, and non-controlled studies. Circ Res. 2017;120(8):1326–1340. 

Rationale: Peripheral arterial disease (PAD) and resulting critical limb ischemia (CLI) can lead to amputations and death for the 50% of those with CLI who are ineligible for revascularization. One of the few therapeutic alternatives offered to these patients is treatment with autologous bone marrow cells. 

Objective: Conduct a critical systematic review and meta-analysis of clinical studies exploring safety and efficacy of autologous cell therapy (ACT) for patients with intractable PAD and/or CLI.

Methods: Authors searched PubMed, ISI Web of Science, Scopus, ClinicalTrials.gov, and Cochrane Central Register of Controlled Trials for clinical evidence or guidelines addressing the effects of autologous stem cells or synonyms of patients with PAD and/or CLI. English publications of uncontrolled or nonrandomized controlled studies or RCTs on 8 or more subjects with severe, intractable PAD or CLI were included in the analysis if they reported the major outcome: proportion of subjects undergoing a major (ie, above-ankle) amputation. Secondary outcomes analyzed were amputation-free survival, complete wound healing, all-cause mortality, and surrogate outcomes: perfusion recorded as ankle-to-brachial systolic blood pressure index (ABI) or transcutaneous partial pressure of oxygen (TcPO₂), subject-reported pain (normalized to a 0–4 scale across studies), and pain-free walking distance. Using intent-to-treat analysis, RR was calculated for dichotomous variables like amputation, amputation-free survival, wound healing, and mortality. No reported amputations in some trials may generate underestimates of amputation rates. Continuous variables were analyzed as mean difference between groups at the end of study. Study quality was assessed independently by 2 observers using the Cochrane risk of bias tool. Publication bias was estimated using funnel plots. RevMan software (Review Manager; The Cochrane Collaboration, London, UK) was used to analyze results, applying metaregression primary analyses for RCT outcomes and secondary analyses on all controlled trials. All trials were included in a tertiary analysis of baseline characteristics and surrogate outcomes.

Results: Most of the 19 RCTs on 837 patients with a mean of 8.1 months’ follow-up were small and underpowered to identify statistically significant results. Cell therapy injected into the calf muscle (15 RCTs), intra-arterially (3 RCTs), or both (1 RCT) varied among RCTs, using bone marrow mononuclear cells (8 RCTs), concentrate (2 RCTs) or mesenchymal stem cells (3 RCTs), or mobilized peripheral blood CD34⁺ or CD133⁺ or mononuclear cells (5 RCTs). One study each used cell therapy expanded in vivo from bone marrow mesenchymal stem cells and macrophages or from peripheral blood pro-angiogenic cells. Comparators were either vehicle or saline placebo (11 RCTs) or no treatment (8 RCTs). Pooling all 19 cell therapy RCTs revealed a 37% reduction in amputation rate (P < .0004), increased likelihood of amputation-free survival (P < .01), and 59% increased likelihood of complete wound healing during the study (P = .002). Ankle-to-brachial index and TcPO₂ increased from baseline on study (P < .0003) while patient pain decreased on study (P = .0001). These effects were significant only in studies where cell therapy was administered in the calf muscle. Effects on mortality and on pain-free walking distance were not significant. When the 3 RCTs with low-risk of bias compared cell therapy to placebo, effects on all outcomes approached but did not reach statistical significance. No dose effect was found for increasing dosage frequency or cell numbers or CD34⁺ percentage and amputation outcomes. More cell therapy than control patients experienced nonsevere adverse events, usually injection site reactions or musculoskeletal disorders (P < .05).

Authors’ Conclusions: Although there are several limitations to this meta-analysis, cell therapy proved safe and was associated with mild transient adverse events related to local administration. Based on these findings, cell therapy may offer benefit as a new “standard of care” for up to 50% of CLI patients who are not candidates for revascularization.

Reference: Liu S, He CZ, Cai YT, Xing QP, Guo YZ, Chen ZL, Su JL, Yang LP. Evaluation of negative-pressure wound therapy for patients with diabetic foot ulcers: systematic review and meta-analysis. Ther Clin Risk Manag. 2017;13:533–544.

Rationale: Diabetic foot ulcers contribute to 85% of nontraumatic lower limb amputations and 13% to 17% of the deaths of those with diabetes mellitus (DM). Negative pressure wound therapy is reportedly safe, efficacious, and cost effective in managing DFUs, but these 3 traits have never been tested together in 1 study.

Objective: Conduct a systematic review and meta-analysis of efficacy, safety, and cost effectiveness of NPWT on DFUs.

Methods: Authors searched the Cochrane Library, MEDLINE, EMBASE, Ovid, and Chinese Biological Medicine databases up to June 30, 2016, for RCTs and similar derivative studies comparing the effects of NPWT with standard dressings on DFUs and chronic foot ulcers of any pathology on individuals with DM. Using the Preferred Reporting Items for Systematic Reviews (PRISMA) checklist, 2 independent reviewers extracted data on study design, Cochrane tool ratings of study risk of bias and quality, patient and wound baseline traits, and outcomes and parameters of NPWT and control therapy. Using fixed or random models according to heterogeneity of the data, they calculated 95% confidence intervals for weighted mean differences between treatments for continuous outcomes like healing time or RR of dichotomous outcomes such as presence or absence of amputation. 

Results: Among the 11 qualifying RCTs on 1044 patients, the average duration of NPWT treatment ranged from 14 to 112 days. All studies verified the experimental and control groups’ comparability at baseline, though most showed some risk of bias. Five RCTs (610 patients) showed a combined 48% increase in likelihood of healing during the study (RR = 1.48; P < .0001) in DFUs treated with NPWT compared with standard dressings (typically moist gauze). Similarly significant effects favored NPWT for reduction in DFU area (6 RCTs on 289 patients; P = .00001) and depth (3 RCTs on 71 patients; P = .00001). Combined incidence of amputation for patients receiving NPWT was 3.5% compared with 12.1% for standard dressing controls (3 RCTs on 520 patients; P = .001). Cost of resources required to heal a DFU using NPWT were reportedly lower than those for standard dressings in 2 studies but were not calculated similarly, so those were not compared. Treatment-related adverse events were comparable for NPWT and standard dressings (P = .68). 

Authors’ Conclusions: This meta-analysis of 11 RCTs extends the effectiveness of NPWT to DFUs and postoperative surgical wounds on the diabetic foot, though  the limitations of the analyzed studies suggest the merit of adding robust RCT research to support NPWT treatment.

Although there are times when amputation is clinically necessary, salvaging as much of a patient’s limb as possible is a high priority for minimizing related clinical, economic, patient, and social burdens. Both studies reviewed here seem to offer opportunities to reduce the likelihood of amputation and preserve as much viable lower limb tissue as possible. These studies differ in the level of amputation studied, wound etiology, and analyses. The systematic review and meta-analysis of PAD/CLI cell therapy effects⁵ studied amputations at the ankle and above and explored whether the effect on amputation diminished in 3 RCTs at less risk of bias. The study of NPWT effects on DFUs⁶ was performed with the same rigor but included studies of non-DFU wounds on individuals with diabetes, so it is not clear that the results generalize to those with a DFU. It also included minor and major amputations, not just those above the ankle as was done on the PAD/CLI review.⁵ Finally, the DFU study⁶ did not explore whether the effects were diminished in less biased, placebo-controlled, blind evaluated RCTs. It is not possible to understand why the effects of cell therapy became less compelling in the placebo-controlled RCTs in the PAD/CLI study and only approached statistical significance in the 3 least-biased RCTs.⁵ Were patients inspired by the placebo effect or were the numbers of patients so low in this subset analysis that the statistical tests were underpowered to reveal statistical significance of these clinically important differences between cell therapy and placebo groups? Does the same phenomenon occur in the NPWT RCT evidence? Will a few more unbiased RCTs tip the balance toward efficacy for these potentially useful therapies? Only more rigorous research will tell. For patients and professionals who cannot wait for that research, these 2 clinical trials offer opportunities for thoughtful consideration.

This article was not subject to the WOUNDS peer-review process.