A Bad Knee Arthritis MSC Injection Meta-analysis Part 2
As I said yesterday, I would dig deeper into the recent meta-analysis that purported to show that “MSC Injections” didn’t work for knee arthritis. Today we’ll get into more problems with that paper.
Monkeying Around with Placebos
A placebo is a treatment we know doesn’t work that can be given to a patient in such a way that the patient believes that they received the real treatment. Yesterday we learned that a meta-analysis requires that we group the results of the same or very similar treatments. The paper in question failed because it lumped together the results of 6 VASTLY different treatments. It also fails at another simple test, it also has a placebo heterogeneity problem.
The paper we dissected yesterday and are also looking at today bears the title: “Intra-articular Mesenchymal Stromal Cell Injections are no Different than Placebo in the Treatment of Knee Osteoarthritis: A Systematic Review and Meta-analysis of Randomized Controlled Trials.” (1) Hence we assume from the start that they will be comparing studies that used MSC injections to treat knee arthritis against a placebo. Right? Wrong.
What Is a Control Group?
In the gold standard of research called a randomized controlled trial (RCT), you compare the treatment being tested to a control group. These are the patients who received the placebo or another more established therapy. The goal of the RCT is to compare the results of one to the other.
Here are the control groups that the “MSC” injections paper used in the 13 studies reviewed:
- Normal saline
- Hyaluronic acid
- Conditioned media
- Ringers lactate
- Conservative care
Only one of these is a common placebo, which is normal saline as discussed in the title. We might be able to consider “ringers lactate” (the proper name is the US is lactated ringers (LR)) and “plasma-lyte” (PL) placebos as they are also commonly used intravenous fluids (like normal saline), but searching the millions of scientific studies listed at the US Library of Medicine only yielded 10 hits under the search “lactated ringers knee joint”. None of these were on using this solution as a placebo. In fact, one study demonstrated that washing out the knee with LR showed positive short-term effects on knee arthritis (2). No studies using PL as a knee arthritis placebo were uncovered. Hence, certainly, LR and possibly PL wouldn’t necessarily qualify as placebos, as we don’t know with any certainty that they don’t help knee arthritis pain and function.
We also have more placebo issues above. Let’s start with another obvious one, hyaluronic acid (HA). These are the knee gel shots used for knee osteoarthritis as a treatment. This therapy isn’t a placebo but in fact, a known traditional HA treatment called viscosupplementation. While we can certainly argue how well this approach works as a treatment for knee arthritis, as the literature goes both ways here, what’s not debatable is that HA is not a saline placebo (3,4). The same would hold true for “conservative care” like physical therapy.
Finally, we have a study which was one of the 13 included which used “conditioned media” as a control treatment. The problem here is that they used the culture media in which the stem cells were growing as a control injection. We know from several different studies that this in and of itself could be a knee arthritis therapy (5).
Differences Are the Death Nell of a Meta-Analysis
As I have shown above, these 13 studies can’t be compared directly in a meta-analysis as they use vastly different placebo groups. The same holds true for a systematic review. In fact, some of these “placebos” are definitely standard of care therapies and some are possible treatments for knee arthritis. Hence, just like comparing different “MSC injection” therapies that had nothing to do with each other, this study falls apart again at it’s most basic construction, the fact that differences in these studies would never allow them to be lumped together or compared side by side.
Having Fun with Stats
There are lies, damned lies, and statistics.”-Mark Twain
Most people, like Mark Twain, hate statistics. Hence, I won’t bore you with the problems here that are more obtuse and that will be part of our formal rebuttal of this messy meta-analysis. However, there’s a high-level stats problem here that my wolf-dog could find. Let’s dig in superficially.
10 of the 13 studies showed that the “MSC injections” were more effective than the control group. In fact, it’s only when you combine the data from these 13 studies (which you now know that you can’t do with a straight face) that the conclusion can be made that the whole dataset showed that “MSC injections” were ineffective. Hence, no amount of lying with stats here gets you around bad study design.
The upshot? I am amazed that any journal, let alone a good one, greenlit this insanity disguised as a meta-analysis and systematic review to be published. The fact that it did either shows how little smart physicians know about the literature in regenerative medicine or that smart physicians were willing to overlook some serious systemic warts to get the anti-interventional orthobiologic message in print. Either one is not pretty.
(1) Dai W, Leng X, Wang J, Shi Z, Cheng J, Hu X, Ao Y. Intra-articular Mesenchymal Stromal Cell Injections are no Different than Placebo in the Treatment of Knee Osteoarthritis: A Systematic Review and Meta-analysis of Randomized Controlled Trials. Arthroscopy. 2020 Oct 21:S0749-8063(20)30846-X. doi: 10.1016/j.arthro.2020.10.016. Epub ahead of print. PMID: 33098949.
(2) Edelson R, Burks RT, Bloebaum RD. Short-term effects of knee washout for osteoarthritis. Am J Sports Med. 1995 May-Jun;23(3):345-9. doi: 10.1177/036354659502300317. PMID: 7661265.
(3) Rutjes AW, Jüni P, da Costa BR, et al. Viscosupplementation for Osteoarthritis of the Knee: A Systematic Review and Meta-analysis. Ann Intern Med. 2012;157:180–191. doi: 10.7326/0003-4819-157-3-201208070-00473
(4) Strand V, McIntyre LF, Beach WR, Miller LE, Block JE. Safety and efficacy of US-approved viscosupplements for knee osteoarthritis: a systematic review and meta-analysis of randomized, saline-controlled trials. J Pain Res. 2015;8:217–228. Published 2015 May 7. doi:10.2147/JPR.S83076
(5) Simental-Mendía M, Lozano-Sepúlveda SA, Pérez-Silos V, Fuentes-Mera L, Martínez-Rodríguez HG, Acosta-Olivo CA, Peña-Martínez VM, Vilchez-Cavazos F. Anti‑inflammatory and anti‑catabolic effect of non‑animal stabilized hyaluronic acid and mesenchymal stem cell‑conditioned medium in an osteoarthritis coculture model. Mol Med Rep. 2020 May;21(5):2243-2250. doi: 10.3892/mmr.2020.11004. Epub 2020 Feb 26. PMID: 32323772.