Open Access Open Badges Research

The role of amputation as an outcome measure in cellular therapy for critical limb ischemia: implications for clinical trial design

Eric Benoit1, Thomas F O'Donnell1, Mark D Iafrati1, Enrico Asher2, Dennis F Bandyk3, John W Hallett4, Alan B Lumsden5, Gregory J Pearl6, Sean P Roddy7, Krishnaswami Vijayaraghavan8 and Amit N Patel9*

Author Affiliations

1 The CardioVascular Center, Tufts Medical Center, Boston, MA 02111, USA

2 Division of Vascular Surgery, Maimonides Medical Center, Brooklyn, NY 11219, USA

3 Division of Vascular & Endovascular Surgery, University of Southern Florida, Tampa, FL 33606, USA

4 Heart & Vascular Center, Roper St. Francis Medical Center, Charleston, SC 29401, USA

5 Methodist DeBakey Heart & Vascular Center, The Methodist Hospital, Houston, TX 77030, USA

6 Department of Vascular Surgery, Baylor University Medical Center, Dallas, TX 75204, USA

7 The Vascular Group, Albany, NY 12208, USA

8 Department of Vascular Surgery, Sri Ramachandra Medical College & Research Institute, Chennai, India

9 Division of Surgery, University of Utah, Salt Lake City, UT 84132, USA

For all author emails, please log on.

Journal of Translational Medicine 2011, 9:165  doi:10.1186/1479-5876-9-165

Published: 27 September 2011



Autologous bone marrow-derived stem cells have been ascribed an important therapeutic role in No-Option Critical limb Ischemia (NO-CLI). One primary endpoint for evaluating NO-CLI therapy is major amputation (AMP), which is usually combined with mortality for AMP-free survival (AFS). Only a trial which is double blinded can eliminate physician and patient bias as to the timing and reason for AMP. We examined factors influencing AMP in a prospective double-blinded pilot RCT (2:1 therapy to control) of 48 patients treated with site of service obtained bone marrow cells (BMAC) as well as a systematic review of the literature.


Cells were injected intramuscularly in the CLI limbs as either BMAC or placebo (peripheral blood). Six month AMP rates were compared between the two arms. Both patient and treating team were blinded of the assignment in follow-up examinations. A search of the literature identified 9 NO-CLI trials, the control arms of which were used to determine 6 month AMP rates and the influence of tissue loss.


Fifteen amputations occurred during the 6 month period, 86.7% of these during the first 4 months. One amputation occurred in a Rutherford 4 patient. The difference in amputation rate between patients with rest pain (5.6%) and those with tissue loss (46.7%), irrespective of treatment group, was significant (p = 0.0029). In patients with tissue loss, treatment with BMAC demonstrated a lower amputation rate than placebo (39.1% vs. 71.4%, p = 0.1337). The Kaplan-Meier time to amputation was longer in the BMAC group than in the placebo group (p = 0.067). Projecting these results to a pivotal trial, a bootstrap simulation model showed significant difference in AFS between BMAC and placebo with a power of 95% for a sample size of 210 patients. Meta-analysis of the literature confirmed a difference in amputation rate between patients with tissue loss and rest pain.


BMAC shows promise in improving AMP-free survival if the trends in this pilot study are validated in a larger pivotal trial. The difference in amp rate between Rutherford 4 & 5 patients suggests that these patients should be stratified in future RCTs.