Clinical trial identifiers for MSCs

A shiny app to explore the characterisation of mesenchymal stromal cells in clinical trial reports

Genever Lab

Department of Biology

University of York

York YO10 5DD


Prepared by Emma Rand in support of:

Wilson, A. J., Rand, E., Webster, A. J., & Genever, P. G. (2021). Characterisation of mesenchymal stromal cells in clinical trial reports: analysis of published descriptors. Stem cell research & therapy, 12 (1), 360.


Background: Mesenchymal stem or stromal cells are the most widely used cell therapy to date. They are heterogeneous, with variations in growth potential, differentiation capacity and protein expression profile depending on tissue source and production process. Nomenclature and defining characteristics have been debated for almost 20 years, yet the generic term “MSC” is used to cover a wide range of cellular phenotypes. Against a documented lack of definition of cellular populations used in clinical trials, our study evaluated the extent of characterization of the cellular population or study drug.

Methods: A literature search of clinical trials involving mesenchymal stem/stromal cells was refined to 84 papers upon application of pre-defined inclusion/exclusion criteria. Data were extracted covering background trial information including location, phase, indication, tissue source, and details of clinical cell population characterisation (expression of surface markers, viability, differentiation assays and potency/functionality assays). Descriptive statistics were applied, and tests of association between groups were explored using Fisher's Exact Test for Count Data with simulated p-value.

Results: Twenty-eight studies (33.3%) include no characterization data. Forty-five (53.6%) reported average values per marker for all cell lots used in the trial, and 11 (13.1%) studies included individual values per cell lot. Viability was reported in 57% of studies. Differentiation was discussed: osteogenesis (29% of papers) adipogenesis (27%) and chondrogenesis (20%); and other functional assays arose in 7 papers (8%). Extent of characterization was not related to clinical phase of development. Assessment of functionality was very limited and did not always relate to likely mechanism of action.

Conclusions: Extent of characterization was poor and variable. Our findings concur with those in other fields including bone marrow aspirate and platelet-rich plasma therapy. We discuss the potential implications of these findings for the use of mesenchymal stem or stromal cells in regenerative medicine, and the importance of characterization for transparency and comparability of literature.

Visualisation tools

Barplots for categorical variables

These barplots allow you to explore the number of trials for various categorical variables.

Where there are many categories, or category names are long, barplots with flipped axes are often clearer. You can select an additional categorical variable using the fill option. A test for the association between the x-axis variable and the fill variable is given below the plot.

Each point represents a publication - hover over the point for the publication id number. The extent of characterisation is given by the number of tests done without values being reported, with values being reported, or the sum of these (total).

Boxplots for categorical variable

A test for a difference in the extent of characterisations between different levels of the categorical variable is given below the plot.

The number of trials testing for an attribute and reporting values, testing without reporting values or not testing for each characterisation attribute.

In detail for each trial. The box indicates ISCT markers

Each point represents a publication - hover over the point for the publication id number.

Where there are no error bars - the publication reported average values for all cell lots used in the trial.

Where error bars are included - either the standard error on the average values for all cell lots used in the trial reported by the publication or the standard error calculated from individual values reported per cell lot.

ISCT markers

Other markers

The following markers were not reported in any paper: CD133, CD146, CD271, STRO-1, MSCA-1, SSEA-4

Figure 1. Literature search strategy and results. (A) The schematic shows search terms, refinements and exclusions used. Numbers refer to the total number of papers remaining at each stage. (B) Reported characteristics for MSCs in clinical research studies: data elements captured for this analysis. Basic information on the trial included clinical phase, indication, route of administration and mechanism(s) of action. Specifics of the cell source included donor details, tissue source and usage (allogeneic/autologous) and the descriptor used by the study: stem/stromal cells or other nomenclature. Aspects of characterization reported in the study were captured, focusing on assessment of viability, phenotypic profile, differentiation capacity and potency evaluations. Reference to ISCT minimal criteria for identification of MSC was also recorded.

Figure 2. Background trial information. (A) Origin of clinical research publications, ranked by number from each country represented in the analysis. (B) Clinical trials reported in literature by clinical phase, ranked by most commonly represented phase of clinical study. (C) Route of administration, ranked by most commonly used in the studies. (D) Indications addressed by the clinical studies, ranked by most commonly represented indication.

Figure 3. Background information on cells used in clinical trials. (A) Sources of tissue from which MSCs were derived. (B) Reported use of autologous and allogeneic MSCs (C) Nomenclature used to describe the cells used in the clinical trials.

Figure 4. Extent and stringency of characterization. (A) Number of articles reporting each category of characterization. (B) Stringency of characterization reported at each clinical phase of development (coloured as in A). (C) Number of phenotypic markers, and viability, evaluated in articles that reported values/averages.

Figure 5. Phenotypic characterization and viability. The minimal criteria recommended by ISCT for identification of MSC are shown between the black bars on the y-axis. (A) Analysis of individual markers reported in the clinical data set, showing whether an attribute was performed with results reported, whether it was performed but no results stated, or not mentioned in the study report. (B) Number of studies that addressed each attribute, defined by extent of reporting for each marker. Required expression or absence of a marker according to the ISCT recommendation is indicated on the y-axis.

Figure 6. Differentiation and other functionality assessments. (A) Frequency of functionality assessments. (B) Nomenclature (stem/stromal) in relation to potential mechanism of actions relevant to each study indication. (C) Evaluation of MSC differentiation capacity (multi-potentiality) in relation to the mechanism of action anticipated for each study.