Platelet-Rich
Plasma for Knee Osteoarthritis
A new study
shows promise, as long as you don't look too closely.
By Alex
Hutchinson
Published: February
22, 2013
Remember a few years ago
when platelet-rich plasma (PRP) injections were the hottest thing in
sports medicine? The evidence that PRP could help tendon injuries was
preliminary at best, but Tiger Woods and a ton of other top athletes
were doing it, so it was only a matter of time before better studies
emerged. Except that they didn't -- study after study produced results that were ambiguous at best.
The hype has died down a bit, but my sense is that PRP (which involves
removing some of your blood, spinning it to concentrate the purported
healing power of the platelets, then reinjecting it at the injury site)
remains quietly popular.
The most recent PRP study I saw tested its potential to treat knee osteoarthritis, and the press release was very encouraging:
A study by researchers from Hospital for Special Surgery has shown that platelet-rich plasma (PRP) holds great promise for treating patients with knee osteoarthritis. The treatment improved pain and function, and in up to 73% of patients, appeared to delay the progression of osteoarthritis, which is a progressive disease... "This is a very positive study," said Brian Halpern, M.D., chief of the Primary Care Sports Medicine Service at Hospital for Special Surgery, New York City, and lead author of the study.
What I found most interesting was that the study used objective
outcomes to measure success, rather than just asking the patients
whether they felt better. Again from the press release:
"The problem with a lot of the PRP studies is that most people have just used subjective outcome instruments, such as pain and function scores," said Hollis Potter, M.D., chief of the Division of Magnetic Resonance Imaging at Hospital for Special Surgery, another author of the study. "But even when patients are blinded, they know there has been some treatment, so there is often some bias interjected into those types of studies. When you add MRI assessment, it shows you the status of the disease at that time, regardless of whether the patient is symptomatic or asymptomatic..."
This sounded very promising, so I dug up the full paper from the Clinical Journal of Sports Medicine
to check it out. I'll leave aside the subjective measures of pain and
stiffness and so on, because this was an unblinded case series with no
control group -- based on the results produced even by placebo
injections in other PRP studies, I don't think there's much useful
information we can gather from those measures.
But what about the objective MRI data? In this case, the images were
analyzed and scored by radiologists who were blinded as to whether the
images were "before" or "after," which is great. While it's a little
ambiguous, it appears 12 patients (a total of 15 knees) were assessed
with MRI. The headline number that osteoarthritis progression was
delayed in 73% of patients seems to come from the fact that 11 of the 15
knees showed "no significant worsening" 12 months after the PRP
treatment. Now, the crucial question is: under normal circumstances
(i.e. if they'd had a control group), what sort of changes would they
have expected to see? Here's what they say:
This is in contrast to some longitudinal studies that suggest an annual decrease of up to 4% to 6% of cartilage volume in knee osteoarthritis compartments.
If you follow the references they cite for this statement, you find this paper (full text freely available here).
Sure enough, they report that after 12 months, they observed a decrease
in cartilage volume of "3.7 +/- 3.0% for global cartilage and -5.5 +/-
4.3% for the medial compartment [of the knee joint]." You'll notice that
the standard deviation of these measurements is almost as big the
measurement itself, which means you need a reasonably large sample size
to observe the effect (the original study had 102 subjects). So what
were the actual numbers (and standard deviations) in the PRP study? As
far as I can tell, they don't report them anywhere! They just say that
the measurements "did not reach statistical significance." Does this
mean they observed a volume change of -0.0 +/ 0.5%? Or -3.0 +/- 6.0%? I
have no idea.
Why does this matter? There's a big difference between failing to
observe a statistically significant change and successfully showing that
a parameter doesn't change (i.e. that the progression of the
disease is delayed). Doing the latter was always going to be extremely
difficult with just 12 patients -- but not publishing the results makes
it impossible to judge whether they did. (On the other hand, they did
include a large table with detailed values for seven different
subjective measures, with a total of 70 different numbers). As the press
release stated, "The problem with a lot of the PRP studies is that most
people have just used subjective outcome instruments." This one didn't,
but if it doesn't reveal the results of the objective measures, then
what's the difference?
Source:
Δεν υπάρχουν σχόλια :
Δημοσίευση σχολίου