Over the past few posts, I have been reviewing the situation clinicians and others find themselves in if they want to make health-related claims about regenerative therapies currently in use by the community. The FTC has gotten a Federal court to embrace the requirement that health-related claims need to be based on results from Level 1 double-blinded, placebo-controlled clinical studies performed by experts in the particular pathology being treated. The focus of the FTC’s court Order was amniotic stem cell products for which they stated there was no clinical utility due to the lack of any human clinical studies. However, the court Order broadly covers any stem cell-containing preparation, including PRP and BMC. So, if a sales rep comes up to you and offers some magical stem cell-containing goop, send them packing unless they provide Level 1 human clinical data.
In the previous post, I had gone over the difficulty in trying to do a Level 1 clinical study on the potential benefit of treating knee OA with intraossesous injections of autologous BMC. I just don’t see how it is possible to do sham intraosseous injections, or who in their right mind would undertake to inject a placebo intraossesously. My suggestion was to skip the drama and contortions and just do a Level 4 study. However, if you want to do a Level 4 study on intraossesous BMC injections to treat knee OA, you might want to think about adopting a few design elements that could boost the quality of care as you practice regenerative medicine:
- Anticipate treating a fairly large number of patients (i.e., 50) for a single, qualifying pathology (i.e., intraossesous injections into the femoral condyle and the tibial plateau for treating adjacent tissue knee OA)
- Qualify all patients who are to receive a BMC treatment with specific metrics before treatment (e.g., VAS, KOOS, etc.)
- Analyze the BMA and BMC preparations prior to injection, so at least you will have hemoanalytic data on the injectate
- If you are feeling philanthropic, pay for having CFU-F done by a reputable 3rd-party laboratory on the BMC injectates
- Track the patients with your selected metrics for at least 2-years, and hope they don’t move away
- Don’t pick and choose; treat all patients presenting with the pathology for whom a BMC therapy is a valid treatment option who agree to pay for the treatment (BMC treatments are cash pay) and record their results, in order to maintain a consecutive series for evaluation
- Record all adverse events as they arise, and report any serious adverse events to the FDA
As far as I can tell, none of the foregoing is a violation of any rule or requirement set by the FDA. Clearly, as a physician practicing medicine with autologous BMC, the FDA would applaud your characterizing the pre-treatment status of patients and tracking their outcomes over time, including adverse events. Having compositional data on the injected BMC, like a hemoanalysis, also should warm the Agency’s heart, since that type of information means you can refine your practice of regenerative medicine should trends pop out. As an added benefit, at least you will have quantitative data on each BMC placing you in a position to make more meaningful statements on the quality of the injectate besides indicating that it was some shade of “red”. I am not sure what the FDA thinks of CFU-F data, but that isn’t your problem; if you have it, correlate it.
Once enough of your patients have reached a two-year milestone, you could collate all of the data you have and look for correlations among patient demographics, reported outcomes, perhaps a two-year radiological assessment of the Kellgren-Lawrence score, hemoanalysis, and, my favorite, CFU-F data. Run the numbers and if something looks interesting, and with two-year data being in short supply in this area of the literature, you could consider publishing your findings. The process to publish the data requires that you go to an Institutional Review Board (IRB), which is qualified to handle stem cell-based clinical studies, to request an exemption from informed consent.
The reason you want an exemption is that you already have, as a part of your practice of medicine, all of the clinical data you will include in the manuscript, so you will have no further contact with the patient for the period of the two-year milestone study. Since they won’t be involved in your retrospective review, there is no need for them to consent to have you study them and report their clinical outcomes. None of this is devious, nor deceptive, since you are keeping good records as a standard of care, and you don’t know that in two-years’ time anything will be worthy of publication.
What will be some of the arguments used against you and your study? The classic one will be clinician bias, due to the absence of a placebo control and no blinding. Your counter argument is that you don’t see a way to do a Level 1 study with an intraossesous delivery of BMC to the tibial plateau and femoral condyle. Or you could revert to the use of historical controls who received another therapy for knee OA, matched to within an eyelash of the patients in your intraossesous-treatment cohort. Including historical controls will lift your study above a true Level 4, but requires a very large data base of patients treated previously with another therapy for whom you have all of the metrics out to two years.
Also, it will be argued that the patients being treated are self-selecting, since they will be paying for the therapy, so not all patients presenting with the condition will be treated. I am not quite sure how the fact that patients willing to pay selects out responders from non-responders to the BMC treatment. Obviously, both richer and poorer people suffer from OA, so this argument seems to imply that those who pay will heal differently or better than non-payers. There might be an argument to be made in a broader sense, if you take into consideration the socio-economic status of the patients being treated, but that is a question well beyond the basic one of assessing if BMC injected intraosseously produces a favorable outcome in treating knee OA.
Another argument is that the patients know they are being treated and psychologically will be pre-disposed to report a beneficial outcome. That is why you need to run the study for 2-years, and to use objective clinical data like a Kellgren-Lawrance score pre-treatment and 2-year post-treatment (you might try to get a blinded evaluation of images). You still will encounter naysayers who will argue that the effect is just a placebo even out to 2-years, at which point, ask the person to name a “real” injection-based treatment that performs as well as the results you might obtain. Viscosupplementation in an OA knee? For two years? No way, or at least not for a reasonable number of patients.
Finally, if the naysayers persist, take comfort in the statement made ever so long ago by Jonathan Swift (author of Gulliver’s Travels, if you were about to search for his name):
There are none so blind as those who refuse to see.
Plenty of blindness on autologous BMC therapy going around, including at the FDA and FTC, but they don’t practice medicine and you do. Just be careful about the health-related claims you make.