In the previous post, I had started my review of a paper published by Shapiro, et al., earlier this year (View Paper), in which they showed that treating a patient’s knee osteoarthritis with saline gave the same degree of pain relief as the use of bone marrow concentrate (BMC) in the other arthritic knee. Crazy, right? I suggest you read the previous post and this one before you load up on the stock of companies that manufacture sterile saline for injection, because I believe their study is deeply flawed.
Please see the previous post for a description of the clinical study and my initial misgivings. There are a number of additional issues that should be considered:
- The authors claim to have used the bone marrow aspiration method advocated by Dr. Philippe Hernigou (Chief of Orthaepedic Surgery, Henri Mondor Hospital, University of Paris, East, France), but they used a 60cc syringe, while Dr. Hernigou uses a 10cc syringe during aspiration. No big deal, you are thinking? Dr. Hernigou published a study (View Study Here) in which aspirating with a 10cc syringe resulted in a 300% increase in MSC content compared to a 50cc syringe. So, volume of syringe has been shown to play a role in the quality of the bone marrow aspirate.
- While I don’t practice medicine, I have spoken with a fair number of physicians who inject stuff into patients’ knees and from what they have told me, the upper limit on injection volume is 10 mL per knee. So, the injection of 15 mL into each joint seems excessive. In fact, with that large of a volume of saline, I suspect the joint experienced a mini-lavage, which would have the effect of washing out some of the “bad” biomolecules that contribute to pain.
- The authors state that they used a CD marker profile established by the International Society of Cell Therapy (ISCT) to detect MSCs in their fresh preparations, except the ISCT CD marker panel was developed and is used for MSCs that have been cultured for a while. The utility of this panel to count MSCs present in the patients’ fresh preparations is dubious in my opinion, and could have contributed to an overestimation of the numbers of MSCs injected. I will note that the authors didn’t comment on performing a CFU-F assay, which has been accepted and widely reported in the literature as an estimate of MSC content in bone marrow samples.
- They claimed to have assessed the HSC (hematopoietic stem cell) content by “…coexpression (sic) of CD45 and CD34 surface markers.” But it is unclear what negative CD markers were used to further discriminate the HSCs, since usually HSCs also are reported as CD31-.
- The percentage recovery of various cell types was not reported in the paper. The authors indicated they had characterized both the BMA (pre-processing sample) and BMC (post-processing sample), so it should have been an easy calculation to report out the percentage recovery for MSCs, HSCs and other cellular components.
- I wasn’t able to calculate the percentage recovery of the MSCs present in the patients’ preparations, but I was able to calculate the recovery rate of mononuclear cells (MNCs): 26%. Now, that isn’t a very high recovery rate for the class of cells that would include MSCs. As it happens, the recovery rate of MSCs in the Arteriocyte’s system, which was used in this study to generate the BMC, has been reported to be 39% (Hegde, et al. 2014; View Study Here), while higher than my calculated MNC recovery rate, it still is pretty dismal.
What does all of this mean? I think the authors don’t have a good handle on the number of MSCs (or HSCs for that matter) that were injected, so it is quite possible that the BMC-treated knees didn’t have a sufficient number of MSCs to show a more significant decrease in pain scores, or at least this could be one contributing factor. Dr. Hernigou has shown in numerous publications that a therapeutic benefit is correlated to the total number and concentration of the MSCs (as estimated by CFU-F assay) injected, so having a lower total MSC number or lower concentration of MSCs injected might have reduced the potential therapeutic benefit. In fact, Shapiro, et al. intentionally diluted the BMC from 5 mL to 15 mL, a step that guaranteed a reduced concentration of progenitor cells—exactly the opposite of what a clinician would want to do when using a patient’s BMC for therapy.
I also will point out that despite the randomization of knees with type of injection fluid, a review of the pain scores and ranges reveals that upper range values for the various pain metrics reported at baseline (see Table 3 on Page 5) show a consistent pattern of having a lower value for the saline-injected knees compared to the BMC-injected knees. So, while the authors reported that there was no statistical difference in the median pain scores reported by the study subjects between the BMC- and the saline-injected knees at baseline, there does seem to be a difference in the range of the pain scores. In fact, the lower limit of the ICOAP total pain score was “0” for the saline group, but “18” for the BMC group.
What also is curious is the fact that several of the pain metrics have “0” as the minimum pain score reported by patients at baseline, which means that one or more patients had one or more knees that had no pain. So, a fair question to ask is why was that patient included in a clinical study when the main clinical outcome is to assess the pain level? How much lower than “0” can you go? Patient-related issues always plague a clinical study, but differences in the pain profiles of the two treatment groups, and the inclusion of patients with prior knee surgery certainly doesn’t seem to me to support the sweeping negative conclusion reached in Shapiro, et al. that BMC “…cannot be recommended for the regular treatment of osteoarthritis of the knees.”
While I applaud Dr. Shapiro and co-workers for attempting to be quantitative by analyzing the composition of each patient’s BMC, I think the design of the study, along with the other issues I have mentioned above, seriously limits the validity of their conclusion. Their publication shows the difficulty of doing blinded studies involving bone marrow if you don’t do sham bone marrow aspirations or design a single knee study in which patients receive BMC only. I also question their use of sterile saline as a “placebo”.
One of the themes of the Shapiro et al. clinical study is that autologous cell therapy needs to be treated like a standard small molecule drug trial, which is what makes their study so great in their minds. However, one thing I have learned over my 14 years of involvement in translational cell therapy is that cells aren’t drugs, and autologous cell-based therapy isn’t as easily studied as small molecule drug therapies are.