In the previous post, I had completed my review of the clinical findings published in a second paper by Shapiro, et al. confirming the utility of injecting saline into one OA-afflicted knee and BMC into the other OA-afflicted knee when treating patients with bilateral knee OA. Certainly, their approach is a novel way for treating bilateral knee OA. The authors offered up a number of hypotheses in their two publications to account for their results, and I offered up a few of my own, some of which point to structural issues with the design of the clinical study protocol.
For example, saline was used as a so-called placebo injection, since in order to dance to the FDA’s and FTC’s tune a highly valued clinical study must compare the treatment to a placebo control. The authors certainly did that, but as they acknowledged and as I have reviewed, saline probably isn’t a good placebo. So, if you are negotiating with the FDA to do an Agency-blessed clinical study try to think of a compound other than saline that could be injected that truly is neutral. Since writing the posts on the issues with considering saline to be a placebo, I have been trying to come up with an alternate compound, without luck. A key problem is that it must not disrupt the ionic strength at the injected site, and nothing beats isotonicity like good ol’ saline.
You might be wondering why Shapiro, et al. chose to do a study in which patients were their own controls. Join the club. One of the hallmarks of BMC-based therapies is that BMC relies on the activity of living cells to promote a therapeutic benefit. Thus, various cells in the BMC once implanted at the site of pathology will sense the inflammatory status and respond by taking over the feedback loops that might be promoting inflammation or by organizing the interweaving of catabolic and anabolic enzymatic activities that are required to remodel pathologic tissues. And it should surprise no one that progenitor cells migrate in our bodies to treat sites of pathology. For example, as indicated in Laing, et al. (2007), CD34+ cells were found to have been mobilized in patients who had experienced a tibial fracture, with a peak for CD34+ cells in their peripheral blood occurring around Day 3 post-fracture.
Consequently, in the Shapiro, et al. papers the authors quite rightly point to the possibility of having a benefit of BMC injection in one knee aiding in the mitigation of pain in the saline-injected knee via a systemic effect. As the authors acknowledged (2018):
Future studies must include larger populations and compare BMAC treatment in unilateral knee joint disease to placebo control in order to eliminate the possibility of a systemic effect.
Clearly, there is no reason to repeat the potentially suspect clinical protocol used in Shapiro, et al., so something else needs to be adopted.
From a technical point of view, a true Level 1 study will involve randomization into a test and control group, the study subjects will be blinded as to which treatment they receive, and the evaluators of the subjects also will be blinded as to which subject received what therapy. It also is possible in some cases to blind the physician providing the therapy. For example, a covered syringe could be used to deliver a control fluid versus a test fluid like PRP or BMC.
What I thought would be an obvious problem—having a control patient undergo a bone marrow aspiration, but not be treated with the BMC—turns out not to bother the FDA at all. During a presentation at the recently concluded Interventional Orthopedics Foundation Conference (Broomfield, CO), I learned that the FDA was just fine having all enrolled study subjects undergo a bone marrow aspiration, but those designated as controls wouldn’t receive their BMC. Instead, the control subjects will receive an injection of corticosteroids, which I am sure will have them thanking their lucky stars. That the control subjects will run a risk, however small, involving an invasive procedure like marrow aspiration, and not receive any potential benefit from the aspiration, appears to be a small price the FDA is willing to let them pay in order to fulfill the requirements of a Level 1, double-blinded, placebo-controlled clinical trial.
One other complication in using BMC to treat knee OA is that an intraarticular injection might not be the best route to mitigate associated pain and delay disease progression. I summarized results from Dr. Philippe Hernigou in the previous post in which patients suffering from both OA and avascular necrosis of the tibial plateau and the femoral condyle reported durable pain relief in the knee treated following intraosseous BMC injections into all four bone marrow compartments, including an increased volume of articular cartilage and reduction of the necrotic lesions over an average of 12-year follow-up. The trick is to inject the BMC intraosseously, rather than intraarticularly. Dr. Hernigou started treating patients with avascular necrosis of the femoral head more than 30-years ago, which requires an intraosseous delivery of BMC to the lesion in the femoral head. He has expanded his intraosseous approach to a number of other pathologies, including rotator cuff repair.
I will continue to explore in the next post the implications of trying to do Level 1 clinical studies in which autologous BMC is delivered intraosseously to the pathologic site and suggest some alternatives to doing a Level 1 clinical trial.