Starting from 2014, I was trying to capture results of clinical studies in cell therapy. Today, I’d like to share some results of this attempt. I decided to narrow down my analysis to regenerative medicine, since most of cell-based therapies with published results belong to this category.
Inclusion criteria and definitions
- Clinical study defined as any study, which involves administration of alive cells as therapy in human and includes more than 3 subjects. It may include: registered clinical trials and “other human studies”.
- Analysis of published clinical trials could be interim (trial is not completed), final (trial is completed) or retrospective (post hoc analysis, long-term outcome).
- “Other clinical studies” may include (i) cohort studies, (ii) case series, (iii) clinical cases (3 subjects or more) and (iiii) undefined reports. “Other clinical studies” are not registered in clinical trial databases or could not be ID’d. Reports of compassionate use or hospital exemption could belong to both categories – registered as trial or have no ID.
- All published in medical literature reports of studies results, indexed in PubMed database.
- Regenerative medicine is defined (for the purpose of this analysis) as administration of alive cells to replaces or regenerates human cells, tissues or organs, to restore or establish normal function.
- Results of any clinical study, appeared in PubMed and/or journal web-site in the period between Jan. 1, 2014 to December 31, 2014.
- Mesenchymal stromal cells as immunomodulators were included in the analysis, because of multiple mechanisms of action, including tissue repair (example: GVHD)
- Conference reports, companies press releases and mass media mentions.
- Clinical cases – reports, which describe less than 3 cases.
- Studies, which only describe design of the trial, but not results.
- Other than “regenerative medicine” studies, which include cellular immunotherapies with immune cells in oncology, hematopoietic cell transplantation in hematology.
I set very loose filter for search of clinical studies results by using PubMed database with query “stem cell”. I receive notifications about new publications from this query via RSS feed. “Stem cell” query captures about 50-100 publications daily, <1% of them are clinical studies. I did random check to assess RSS feed accuracy and confirmed that it captures everything. I find that such filter allows to capture the most (if not all) clinical studies, related to cell-based therapies (by any type of cells) in regenerative medicine. I also compared “stem cell” filter with “cell therapy” and some others. A stronger filter in PubMed – “stem cell” + “clinical trial” (check box), misses >90% of clinical studies.
Total number of studies and data capture
- I was able to identify 116 clinical studies, which involve administration of cells as regenerative medicine.
- Reports of cell-based regenerative medicine clinical studies results appear with a rate of 1 every 3 days.
- I captured link to original publication, trial ID (if any), indication, country, number of subjects, type of cells, cell manipulations and some other additional information.
- You can look at sample of raw data here.
- 59 of 116 studies (~51%) did not have indication to any clinical trial ID.
- 59 registered clinical trials IDs were identified in 57 published studies (2 studies described 2 registered trials).
- About 80% of all trials were registered in NCT database (ClinicalTrials.gov).
A great variety of countries published results of clinical studies in 2014. The biggest number of reports came from China. US was the second biggest contributor. I summarized major contributing countries (more than 4 reports) in this figure:
As one may predict, Mesenchymal Stromal Cells (MSC) was the major cell type (35%) in cell-based regenerative medicine clinical studies. Mobilized hematopoietic stem/progenitor cells (HPC-A) and bone marrow mononuclear cells (BM MNC) were another popular type of cells for tissue regeneration. Interestingly, 2 reports were about results of studies, involved embryonic stem cell-based (ESC) products. Few studies used 2 different types of cells cell simultaneously or concurrently.
I loosely judged all results of clinical studies as (1) positive, (2) mixed/ inconclusive and (3) failed.
Positive are the results, reported no safety and feasibility concerns and/or provided at least some evidence of efficacy.
Mixed/ inconclusive results included:
- end points are defined, but were not analyzed/ described in report (example)
- efficacy end points were not defined or different end points were reported (example)
- some end points were met, but some missed (example)
- only some subsets of patients responded – highly depends on approach to analysis (example)
- number of patients is too low to conclude efficacy (example)
- efficacy declined or deterioration observed in different time points (example)
Failed studies were considered, based on termination of studies due to safety issues, lack of feasibility and/ or lack of efficacy. Failed efficacy usually reported by authors as lack of difference between control group (example: placebo) and experimental (“cells”) group. Few trials were not described as failed by authors, but judged as such by me, based on lack of significance between groups and missed end points.
The first figure shows total number of positive, mixed/ inconclusive and failed studies from analysis of all reported studies (116):
Next, I looked at only registered clinical trials and broke it down by phases – 1, 1/2, 2, 2/3 and 3:
Only one trial is failed in phase 1 and almost 90% of them reported as “positive”. The only trial, which failed in phase 1 was designed with efficacy end points. There were no trials, which failed safety. A number of mixed/inconclusive and failed trials were increased from phase 1 to 2/3. Because efficacy of therapy is usually assessed in phase 2 trials, studies labeled as “phase 1/2” are not necessarily included efficacy end points. Despite lack of any controls many studies concluded by authors as “positive” or “promising”. Placebo control was very rare. Most controls include: (i) historic, (ii) baseline and (iii) standard therapy.
Limitations of analysis
- The data are not validated by independent researcher.
- Definitions, inclusion and exclusion criteria, data interpretation were set by me and could be subjective.
- Thorough analysis was limited by the lack of access to some publications (I’d estimate ~10%).
- Only PubMed database was used for data capture.
- Results of many studies were very hard to interpret, because end points were not defined or/and were not reported, absence of control groups, low number of subjects. Mixed/ inconclusive results category is especially difficult to analyze. It could be interpreted differently by different people.
It was a snapshot of some data that I was able to capture in 2014. Please feel free to give me feedback and discuss these data in comments. I’m open to suggestions and collaboration.
How to cite:
Bersenev Alexey. Results of regenerative medicine clinical studies from 2014. CellTrials blog. March 1, 2015. Available: http://celltrials.info/2015/03/01/results-regenmed-studies-2014