It’s always interesting to see the back and forth on bone marrow versus fat as a source for stem cells. Given that the FDA has long declared the same-day fat stem cell procedure that breaks down the tissue in which stem cells are imprisoned (SVF) as an illegal and unapproved drug, you would think this would be a moot point. However, it still comes up. Now, a new paper purports to reverse the findings of many other papers that show fat is inferior to bone marrow for orthopedic uses, stirring the pot yet again. So it’s time to delve deeply into this publication to see how it came up with its unusual findings.
I’ve blogged before on the issue of what the head-to-head research shows for bone marrow and fat stem cells helping cartilage repair. The last time I did this was in the spring of 2015, and at that point, it was clear that bone marrow beat fat in 13 consecutive and different studies; whereas, I could find no studies that went the other way. So with more than a dozen papers heading in one direction, any time an article comes out that shows the opposite, one needs to look very carefully at the results.
For whatever reason, the FDA has declared that adipose stem cell procedures are illegal drugs. I don’t personally agree with this designation, but that’s what has been clear for many years. However, that hasn’t stopped many physicians from floating the regulatory risk and breaking down fat to get stem cells. So let’s explore the two types of fat procedures out there.
First, the most straightforward procedure using fat is a graft. This means that you take a fat sample via liposuction and then centrifuge and wash the fat. What do you get? Concentrated fat with the stem cells regrettably imprisoned inside collagen and unavailable to help the patient. Despite this not being a stem cell procedure, we have any number of clinics that will claim it is, and since patients don’t know the difference, in the minds of the clinic, no harm no foul. Then there’s a real fat stem cell procedure where the cells are released from their collagen prison. While in bone marrow, this isn’t needed because the stem cells are freely available in solution. When you add this step, the FDA declares the substance an illegally manufactured drug, even if that process occurs in a physician’s office or a surgery center. So if you’ve had this procedure in the U.S., you’ve had an illegal drug.
First, every study ever published by every advocate for fat stem cells always begins with the same song and dance about bone marrow aspiration being an invasive and awful procedure. This new research paper is no different. This is pretty funny to a physician who has performed lots of both bone marrow aspirates (BMAs) and liposuctions. As I’ve blogged before, bone marrow aspiration is the less invasive procedure.
So what’s real here? First, liposuction is the most brutal procedure I have ever personally performed. Why? Your job is to liquefy structural tissue (fat) while it’s still in the body and then suck it out. To do this, you have to literally beat the tissue up enough so that it’s left in pieces small enough to be sucked through a cannula. Whereas with a BMA, you place a larger needle into bone (which has the consistency of hard plastic) and draw out what looks like thick blood (bone marrow aspirate). Pretty simple.
What does the research say about the complication rate of both of these procedures? Even the least invasive form of liposuction is 1,000 times more invasive than a bone marrow aspiration. So much for the idea that BMA is more invasive than liposuction.
The paper first states that the authors took 60 ml of bone marrow and 120 ml of fat. This difference in draw volumes is confusing as the authors are comparing the number of stem cells in the tissues, so they should have taken the same amount. A table in the paper is even more confusing. Table 1 says that the authors used 4 ml of each for testing, but that they also had 68.5 g of adipose tissue and 60 ml of bone marrow aspirate? These numbers don’t add up. So I have no idea how much of each tissue they compared.Second, the paper says that the authors took 60 ml of bone marrow aspirate from a single site, which is a great way to ensure that you reduce the number of stem cells in the sample. I have produced a video on this issue, so watch that for the reasons why this type of single site draw reduces yield.
Third, the bone marrow was concentrated with a commercially available bedside centrifuge, while the fat was processed by hand to extract stem cells. Why is this an issue? As an example, the size of the stem cell fraction in a 60 ml sample is about 2 ml. The machine produced 6 ml, so about two-thirds of that sample was contamination that didn’t have stem cells. So the bone marrow sample was very dilute. The adipose stem cell sample didn’t suffer from this same issue as the lab process had many steps designed to concentrate the stem cell fraction in the sample as much as possible. This difference in processing is a critical problem as the study aims to compare stem cell numbers per unit volume.The authors found that the nucleated cell yield of BMA was much higher. This means that the total number of cells in bone marrow was more. However, it’s unknown, based on the volume issues reported above, if we should adjust for the fact that one-half as much BMA was drawn or not. In this case, the difference would be 12 times more. If we don’t adjust, it’s 6 times more. We certainly know that the bone marrow machine produced a more dilute sample than is possible, which also messes up this comparison.
It’s also concerning that an automated cell counter was used, as for SVF these have been notoriously less accurate and inflate total nucleated cell counts as they often count random lipid droplets as cells. This artificial fat cell inflation likely happened here as BMA in other studies has been shown to have 100X more total cells than fat (see video above for reference). So it’s unknown why this fat sample would have so many more nucleated cells. Which brings us back to the issue of counting lipid droplets as cells, which explains the discrepancy.
The authors claim 4X the number of adherent stem cells in SVF, but when you adjust for the fact that they began with twice as much fat, it’s really only 2X more. Should we make this adjustment? Unsure based on the confusing issue of the volumes harvested and tested. If we adjust for the fact that the bone marrow concentrate was diluted by a factor of two-thirds by contamination as a function of the poor processing job accomplished by the machine, we should also likely make this adjustment as well. Suffice it to say that this comparison is not an apples to apples one.
Flow cytometry uses the markers on cells to identify that cell type. The problem with stem cells is that there is no single marker we can use, so we have to use many. In this study, however, the authors don’t use enough markers to identify stem cells accurately. Why? The ISCT has a minimal definition for the markers you must use. For mesenchymal stem cells (MSCs) this must include: CD105, CD73, and CD90, and lack expression of CD45, CD34, CD14 or CD11b, CD79 alpha or CD19, and HLA-DR surface molecules. Only some of these markers were used here.
While the flow cytometry markers section takes shortcuts by not using the appropriate requisite number of MSC markers, there are other issues as well. More importantly, all stem cell percentages are expressed as a percentage of nucleated cells. The authors state that the BMC has 6X more nucleated cells (we think, see discussions on the issue of the volume of tissues tested), so when they report that SVF has an MSC percentage of the total of 4.28% and BMC has a percentage of 0.42. However, this bone marrow percentage number must by multiplied by 6 to compare apples and apples to fat (compare absolute stem cell numbers rather than a comparison to the number of nucleated cells). So 0.42 X 6 = 2.52%. However, if we adjust for things like the poorly accomplished bone marrow draw, then these numbers may change.
The research tried to access how well both tissues could produce cartilage. To accomplish this goal, cells are cultured with heavy-handed chemical clues for the stem cells to turn to cartilage. Regrettably, this type of cell culture and conditioning isn’t available to physicians who use same-day stem cell procedures.
What’s weird about the rest of this study is what I discussed above. As of spring of 2015, we had more than a dozen papers that showed that bone marrow stem cells produce more and better-quality cartilage and bone than fat stem cells. This difference makes common sense as bone marrow stem cells are directly involved in both types of repair in the body whereas stomach or buttocks fat does not participate in either of these things naturally. This paper reports that they are equivalent or that fat has a slight edge in a few areas. So for this section, I can’t explain how the results go counter to 13 other papers. More interesting is how this paper ignores the contrary results in its discussion section, which, by definition, would be a huge problem in getting it published.
The upshot? This most recent paper has serious issues in comparing bone marrow and fat stem cells. Some of these are obvious; some we’ll never know how the authors got contrary results. In the meantime, I’m sure we’ll see many fat stem cell advocates use this paper as proof of the superiority of these cells, but, in fact, it’s really just a poorly constructed experiment with outlier results.
About the Author
Christopher J. Centeno, M.D. is an international expert and specialist in regenerative medicine and the clinical use of mesenchymal stem cells in orthopedics. He is board certified in physical medicine as well as rehabilitation and in pain management through The American Board of Physical Medicine and Rehabilitation.…