In the previous post, I had left you hanging on the follow-up publication by Shapiro, et al. that was published last year (2018). I wanted to review both papers in light of the recently established precedent obtained by the FTC in a Federal court Order issued last October, in which the FTC laid out the ground rules for making health-related claims: only results from a Level 1 clinical study performed by recognized experts in the pathology will keep you from getting a call from the FTC. Sounds like kind of a pain, huh? The Shapiro et al. papers are from a Level 2 study of treating knee OA that the authors indicated was the first to be “FDA-monitored”, so presumably the Agency was involved in some fashion.
The short answer that is readily apparent in the 2018 Shapiro, et al. publication is that the passage of time didn’t improve the curious equivalence in reduction or durability of pain relief when each enrolled study subject had either saline (claimed to be a placebo) or the subject’s BMC injected into their OA knees. I explored possible hypotheses offered by the authors in their earlier publication on 6-month milestone data that showed no statistically significant difference in pain mitigation 6-months post-treatment between the saline and BMC treatments. I also offered a few of my own hypotheses to account for the outcome.
As indicated in the Shapiro, et al. follow-up publication of 1-yr post-treatment clinical outcomes, there remains no statistically significant difference between the knees receiving saline or BMC in terms of pain mitigation. The authors point out that both injections resulted in substantial pain mitigation measurable at the 1-year milestone compared to subject-reported outcomes prior to treatment. So, pain mitigation is durable, but odd given that saline was supposed to be a placebo, and placebos aren’t supposed to provide any medical benefit, just a psychological one.
One of the critical features of a Level 1 study is the use of a placebo. However, in the 2018 publication, the authors moved away from considering saline as a placebo, and instead refer to it as a “a non-harmful injected substance”. They cite literature that supports a possible analgesic role for saline, but conclude that the durable benefit associated with the saline-injected knees out to the 1-year milestone goes well beyond the time frame for therapeutic benefits reported in previous publications for saline injections. This led the authors to fall back on a systemic mechanism in which the pain mitigation associated with the BMC-injected knee would migrate over to the saline-injected knee. After all, the study subjects were their own controls.
In support of a potential systemic effect, the authors cite other literature to support a mechanism involving “stochastic homing” of progenitor cells moving through via the circulation to reach the other knee. While I have no doubt that progenitor cells can move through the circulation and home to sites of inflammation elsewhere in a study subject, the authors use the term “stochastic”, meaning random distribution, to describe the physical movement of the progenitor cells, which doesn’t convey to me a high probability of success. I, myself, am a fan of the homing of cells and their exosomes via the lymphatic systemic, which offers a less dilutive passage to get to another pathologic site. Imagine the concentration gradients needed if progenitor cells went on a trip through the blood stream and had to home in on the correct inflammatory site, as opposed to moseying along in a much less fluid-filled lymphatic conduit. There also is the issue of having progenitor cells moving through the capillary beds of the lungs, which are notorious for filtering out cells. Moving through the lymphatic system bypasses the pulmonary transit.
The 2018 Shapiro, et al. paper reported on MRI imaging performed at the 6-month milestone on all treated knees. Their MRI analysis failed to demonstrate any increase in anything remotely useful related to articular cartilage, like increased thickness, etc. They also examined radiographic images taken at the 1-year milestone and weren’t able to change the Kellgren-Lawrence classification for any of the knees. From which the authors conclude that BMC treatment in the context of their study could not be claimed to be regenerative. Furthermore, they can’t recommend the use of BMC to treat knee OA as superior to other injectable treatments.
Before moving on to how I think pushing BMC treatments into a Level 1 clinical study format that meets the FTC and FDA requirements to make health-related claims is setting the bar beyond reach, I will offer a couple of observations to provide context for the conclusions reached by Shapiro, et al.
Shapiro, et al. (2018) indicated that the time frame for the MRI was just 6-months post-treatment, and 1-year for the radiological evaluation. As it happens, Dr. Philippe Hernigou (Chief of Orthopedic Surgery, Henri Mondor Hospital, East University, Créteil, France) examined changes in articular cartilage in younger patients (18-41 yo) suffering from knee osteoarthritis associated with bilateral, avascular necrotic lesions in their knees resulting from corticosteroid therapy (2018).
The authors conducted a prospective, randomized study (but not a Level 1 clinical study) on the benefits of treating knee OA with BMC in one knee and a total knee replacement in the other. I previously have reviewed the details of the clinical study and results, but will focus on the MRI assessment of the patients’ knees. The following is a summary of salient details from the paper:
- Patients had evidence of Grade 4 avascular necrosis in both tibial plateaus and femoral condyles
- All patients had stage IV OA, with collapse, and evidence of joint space narrowing, osteophytes and other degradative changes
- Mean follow-up was 12 years
- BMC was injected into both the femoral condyle and the tibial plateau of one knee
- BMC-treated knees showed an average increase in cartilage volume (MRI assessment) of 4.2% ± 2.5% at follow-up
- The avascular necrotic lesions decreased over the same time period by 40%
- There was a statistical correlation between decrease in pain, size of the bone marrow lesions and cartilage volume
What I find most intriguing about the Hernigou paper is the relatively small change in cartilage volume observed over a mean of 12 years in the BMC-treated knees. The authors reported that knee pain decreased over time, but since the bone marrow lesions shrank in volume, while the volume of the articular cartilage increased, it isn’t possible to point to the small volume change in the articular layer as a sole source of pain relief. Also, it might be possible to infer that decreasing bone marrow lesion size reflected an improved bone marrow depot health, which in turn might have promoted a more robust articular cartilage, as was reported by Hernigou, et al., if you looked carefully and after an average of 12 years.
In view of the Hernigou, et al. findings, it isn’t surprising that Shapiro, et al. could find no evidence of changes in articular cartilage volume after just 6-months, since after an average of 12 years the patients’ knees treated intraosseously with BMC showed an increased cartilage volume of 4.2%. Another really important distinction between the two studies is that Shapiro, et al. injected the BMC into the intraarticular compartment of the knee, while Dr. Hernigou injected the subchondral compartments of the tibial plateau and femoral condyle in the BMC-treated knee. Finally, Dr. Hernigou reported that patients receiving BMC with less than 1200 CFU-F/mL of injectate didn’t show a reduction in pain, whereas all of the patients receiving more than 1200 CFU-F/mL reported mitigation of pain. This emphasizes the importance of progenitor cell concentration in the injectate in achieving beneficial outcomes. Shapiro, et al. took a 5-mL volume of BMC and diluted it to 15-mL just prior to injection, which certainly reduced the progenitor cell concentration in the injectate.
So, the negative findings of Shapiro, et al. might be explained in part due to poor protocol choices rather than an inherent failure of BMC to provide a therapeutic benefit in treating knee OA compared to the saline placebo injected knee.
Now, the question is, what drove Shapiro, et al. to design the study protocol in the manner they did? And were some of the protocol elements adopted to appease the FDA’s addiction to a Level 1 clinical study? I will consider these questions and point out some of the issues with doing Level 1 studies with BMC in the next post.