Bone marrow cells for acute myocardial infarction – meta-analysis of controlled clinical trials

by Alexey Bersenev on September 4, 2012 · 0 comments

in Uncategorized

Cell therapy of acute myocardial infarction (AMI) is the field of regenerative medicine with a great promise and high expectations. The field is not very new. First clinical trials have been started more than a decade ago. There are more than a hundred of clinical trials ongoing around the world, which are assessing different types of cells for treatment of AMI. Are they fulfilling our expectations?

A biggest part of these trials, involve autologous bone marrow cells. Usually these cells are unseparated mononuclear cells, frequently labeled by trialists as “stem cells”. Some trials assessing enriched stem/ progenitor cell populations (CD34+ or CD133+) or mobilized into bloodstream bone marrow hematopoietic progenitors. I’d like to bring your attention to meta-analysis of randomized controlled trials, published earlier this year in Cochrane Database.

This is the most valuable and comprehensive analysis, which was ever been done in this field. The authors included, 33 controlled randomized trials (CRTs), conducted in 17 countries with total number of patients 1765. I’d like to highlight a few key reasons, which make this meta-analysis exceptional:

  1. Very strict inclusion criteria. For example, they exclude all published trials which had no clear indication to randomization and control groups, acute phase of AMI, involved allogeneic cells and so on.
  2. The analysis included CRTs, published in other than English languages. The authors actually translated them to make it clear.
  3. The authors dug very deeply into details of clinical protocols. For example, the risk of biases (sequence of randomization, type of blinding, use of co-interventions, reporting bias, sponsorships …), the composition of control media (concentrations of heparin or serum) in placebo group and so on.
  4. Three analysts extracted data and two of them assessed data independently with following meta-analysis. Data were collected from multiple sources, including many literature databases (libraries), web-based clinical trial registries, conferences proceedings and many others.
  5. To clarify the missing data, 18 authors were contacted via email – 9 responded, 6 – provided requested data. All clarified issues or those awaiting resolution were carefully listed in the report.

Well, the quality of this meta-analysis is very very impressive! This report look like a book and include 209 pages with multiple detailed tables. What are the results?

  1. Was no difference observed in mortality and morbidity (restenosis, hospital readmission, reinfarction…) between “cells” versus “placebo” groups.
  2. Bone marrow cells significantly improved functional characteristics of infarcted myocardium (such as LVEF or LVESV/ LVEDV, size of infarct and heart wall motion) short-term and long-term.
  3. There was a positive correlation observed between cell dose heart function (LVEF). There was also a correlation between timing of transplant and heart function.
  4. The cell therapy procedures were safe. There was no significant difference observed in adverse events and number of complications (with a special emphasis on arrhythmias) between “cells” and “placebo” groups. One trial reported many adverse events linked to mobilization protocol by G-CSF and not associated with cell transplantation procedure.
  5. Overall, efficacy of autologous bone marrow cells in AMI could be estimated as “moderate” and significant.

Interestingly, only 3 trial teams attempted to measure engraftment of transplanted cells. At time point about 30 min after transplantation, as much as ~ 3-10% of cell remained in myocardium. Only one team showed about 3.2% of cell retention at 24 hours time point.

What lessons can we learn from this analysis? How can we improve the future assessments? A couple of major points:
1. The main hurdle for analysis – a heterogeneity of protocols and clinical end-points. The authors called for standardization of (A) cell composition and preparation and (B) relevant end-points for assessment of clinical outcome. In the discussion, the authors mentioned that without standards, any meaningful analysis with confident conclusions is not possible. Right now, there is no consensus between different groups on cell preparation and clinical outcome measures.
2. A sample size – should be good enough to assess safety and efficacy:

In general, the sample sizes are small in all trials included, perhaps with the exception of one trial (Janssens 2006) where the number of participants in each arm of the trial is close to 100. Only nine trials used power calculations to estimate the minimum number of participants to be randomised in the trial. Larger randomised trials with appropriate power calculations will be required to define therapy-related adverse events.

As we know, angioplasty alone allows to reduce mortality in AMI patients dramatically. So, it’s very hard to show impact on mortality of any additional intervention:

However, because mortality rates after successful revascularization of the culprit arteries are very low, larger number of participants would be required to assess the full clinical effect of this treatment.

Well, nothing like “great” or “exceptional” or even “good”. I’d like to highlight one more time – the results so far are moderate. Does it look like a revolution in medicine? Not to me. I’m not sure how much improvement of mortality was expected (all patients had angioplasty before cells), but boosting in heart function was. And, in terms of functional improvement, the cells seem to work, but the question is (i) at what degree and (ii) how confident we can be about it without the standards. Pretty much all studies used LVEF as a scientific marker of heart function. But, as the authors noted in discussion, it’s a surrogate marker and how it translates in clinical effects is not very clear. One more thing – promising (as claimed by many business reports and regenerative medicine promoters) economic value. None of the trials reported duration of hospitalization, none of them assessed cost-benefit. Taking in account a moderate functional improvement and absence of changes in mortality/ morbidity, the cost calculations would be important future task.

Many trials, which have been mentioned in report still ongoing and/ or awaiting assessment. The report will be updated accordingly completion of the trials.

{ 0 comments… add one now }

Leave a Comment

Previous post:

Next post: