Autologous bone marrow cell injection reduces one-year amputation rate for Rutherford 4 and non-diabetic Rutherford 5 patients

Keisin Wang

Fifty-two-week outcomes of the MOBILE trial indicate a “significant reduction in major amputations” among Rutherford 4 and non-diabetic Rutherford 5 critical limb ischaemia patients following the use of intramuscular injection of autologous bone marrow cells (cBMA).

Keisin Wang (Indiana, USA) presented the 52-week data at the 2017 Vascular Annual Meeting (VAM; 30 May–3 June), telling delegates that amputation-free survival for cBMA all-comer patients was improved compared with placebo treatment, but not significantly.

Wang explained that critical limb ischaemia is directly responsible for over 50,000 amputations each year in the USA. Although surgical revascularisation remains the gold standard treatment, as many as 30% of critical limb ischaemia patients “have no revascularisation options”. MOBILE “seeks to provide a therapeutic option for this difficult cohort,” Wang said.

The MOBILE trial is the first completed phase III US study investigating cell-based therapy for cardiovascular disease. The double-blind, placebo controlled-trial involved 24 participating US centres with 155 patients randomised to either cBMA or placebo treatment at a ratio of 3:1 (n=119:36), respectively. This gave the study group an 80% power to detect a 60% reduction of risk for major amputation and death. Wang explained that this was based on a 40% failure rate for control and 20% for cBMA groups. These groups were also stratified by Rutherford classification and the presence of diabetes. Both groups received best medical therapy, made up of regimented wound care, antiplatelet therapy, statins, and tight control of morbidities.

Wang said that the patient demographics had “excellent balance regardless of age and gender.” There was no difference in the rates of diabetic enrolment between groups, and the team were “particularly proud of our recruitment of African-American patients (11.1% for the placebo group and 20.2% for the cBMA group), as this group has been historically poorly represented in critical limb ischaemia trials.”

Patients of Rutherford class 4 or 5, who could not be surgically revascularised and with ankle brachial index <0.6 or toe-brachial index <0.4 were included. Patients were excluded if they had HbA1C >10%, creatinine >2.5mg/dL or haemodialysis, or underwent amputation within 30 days of enrolment.

For the cBMA intervention, aspirated bone was placed in a centrifuge tube. Marrow was then concentrated with centrifugation for 15 minutes. The concentrated bone marrow cells were then collected and 0.75cc was injected intra-muscularly 1.5 inches deep at 35 to 40 sites in the index limb. Placebo treatment was identical but with sham bone marrow aspiration and empty needle penetration of muscle beds.

There was no significant difference between the two treatment arms for serious adverse events. In the placebo group, 69.4% (n=25) of patients had severe adverse events compared with 58.8% (n=70) in the cBMA group (p=0.33). In the placebo group, this included four (11.1%) cases of death and four (11.1%) respiratory failures, while in the cBMA group there were five (4.2%) deaths and two (17%) respiratory failures. There was also no significant difference in any adverse event rate, at 97.2% (n=35) in the placebo group and 98.3% (n=117) in the cBMA group (p=0.55). Wang noted, “We were very broad and aggressive in our definition of adverse and serious adverse events, which explains the high rate shown.”

The rate of composite all-cause mortality plus amputation—the trial’s primary regulatory endpoint—was 30.5% (n=11) in the placebo group and 20.2% (n=24) in the cBMA group (p=0.22) (HR 95% CI, 0.7 [0.3–1.59]). The incidence of major amputation—the trial’s therapeutic endpoint—was 22.2% (n=8) in the placebo group and 16% (n=19) in the cBMA group (p=0.39) (hazard ratio [HR] 95% CI, 0.64 [0.31–1.31]).

Wang and colleagues stratified the patients at the time of enrolment to gain a closer understanding of the effect of cBMA on Rutherford 4/5 patients and diabetic patients. They found that for Rutherford class 4 patients (n=84), the rate of one-year amputation was 26.3% (n=5) in the placebo group and an improved 7.7% (n=5) for cBMA (p=0.041) (HR 95% CI, 0.27 [0.08–0.95]). For Rutherford 5 patients (n=71) on the other hand, a lower one-year rate of amputation was reported in the placebo group (17.7%, n=3) than in the cBMA group (25.9%, n=14) (p=0.55) (HR 95% CI, 1.46 [0.42–5.1]).

In non-diabetic patients (n=92), one-year amputation rate was better in the cBMA group at 10% (n=7) than in the placebo group at 27.7% (n=6) (p=0.046) (HR 95% CI, 0.33 [0.11–0.98]). However, this was reversed in diabetic patients (n=63) with the placebo group performing better with a one-year amputation rate of 14.3% (n=2) versus 24.5% (n=12) in the cBMA group (p=0.48) (HR 95% CI, 1.72 [0.38–7.69]).

Excluding Rutherford 5 diabetic patients left a total of 124 patients. For these 124 patients, the rate of amputation at one year was 9.6% (n=9) in the cBMA group compared with 26.7% (n=8) in the placebo group (p=0.021) (HR 95% CI, 0.33 [0.13–0.84]). This translated to a number needed to treat to prevent one amputation value of approximately six.

In terms of secondary endpoints, Wang noted that there was a significant increase in transcutaneous oxygen pressure in the cBMA group at one-year follow-up, from 20.4mmHg to 32.4mmHg. In the placebo group, pressure increased less markedly from 23.2mmHg to 24.8mmHg. Wang suggested that the “increased transcutaneous oxygen pressure suggests that cBMA improves microvascular perfusion in the ischaemic limb.” For the other secondary endpoints, there was no difference between the two treatment groups for pain, ankle brachial index, toe-brachial index, or six-minute walk test outcomes.

“MOBILE shows that cBMA treatment continues to have an excellent safety profile compared to placebo in these patients,” Wang concluded. “The statistical power of the trial was limited by a small control population in Rutherford 5 diabetics as well as the 3:1 randomisation. Additionally the actual event rate in the control group was only 30% versus a predicted 40% in our power analysis. However, this 30% can now serve as a future benchmark for critical limb ischaemia in randomised controlled studies.”