Infuse (rhBMP-2) and the C-word

The “take home” message from my post about the YODA studies was that Infuse was as effective as autologous bone graft at establishing a fusion.  There were concerns raised though about the number of people who were reported as having being diagnosed with a cancer, and on the surface of it there was good reason to be suspicious.  BMP’s stimulate cells to grow, and that’s what cancers do – they grow uncontrollably.  Suspicions though are one thing – research was required.

Take any group of people and you can expect a certain proportion to be diagnosed with cancer over a period of time.  Therefore, to see if a drug “causes” (increases the risk of) cancer, you need to compare a treated group to an untreated group and look at how many get diagnosed with cancer over several years.

First out of the blocks with a study was, once again, Eugene J. Carragee, who you may remember was the instigator of the first paper that focussed on the possible risks of using Infuse, that triggered off the YODA studies.  His paper, published here, looked at the data from a study that was used to trial a combination of a new non-compressible matrix with a high dose of Infuse.  In the trial, they compared two groups of people having spinal fusions.  One group received the high-dose infuse, and the other used their own autologous bone graft.  As part of the study they were tracked for several years, so Carragee, et al, looked at the number of people who were diagnosed with a cancer in both groups.  What did they find?

“The incidence rate of cancers was 6.8-fold greater in the rhBMP-2/ CRM group compared with the control group (p = 0.0026).”

WHAT?  The people receiving Infuse had an almost 7 times risk of getting cancer than those that didn’t!  “Infuse causes cancer!”

Well, settle down for a minute.  There’s a couple of things to look at here.  Firstly, the total number of patients in the study was 463, which whittled down to 292 by the five year mark.  The number of patients with a cancer across that whole group was 17, which means that the risk of cancer for the whole group was only 5.8% over five years.  There is also some dispute about how to even measure cancer incidence; if one person gets two cancers, is that one cancer “event”, or two?

This leads to the second point.  Looking at table 1, one patient was diagnosed with a basal cell carcinoma, squamous cell carcinoma of the skin, and chronic myelogenous leukaemia.  That one patient accounted for 7 of the cancer events in the Infuse group.  Another 3 events was one patient who had multiple SCC’s.  As an Australian it’s easy to see that if someone presents with a BCC or an SCC, they are then more closely followed to recurrences of those tumours.  BCCs and SCCs are also very common, so it could be that, given the small number of people in the study, the Infuse group were “unlucky” to have two patients with multiple tumours.  Or it could be the other way around…

Lastly, the dose of Infuse was extremely high.  Usually when we use Infuse, the dose is either 8.4mg or 12mg.  The dose in the study was 40mg – three to five times the usual dose.  Clearly, even if it does have some mild carcinogenic effect at 40mg, that may not apply at the lower doses used in Australia.

So that brings us to the other study, “Risk of Cancer After Lumbar Fusion Surgery With Recombinant Human Bone Morphogenic Protein-2 (rh-BMP-2).” written by a gastroenterologist and oncologist – in other words, people least likely to have a conflict of interest.  This paper looked at all lumbar fusions performed in the USA over a five year period where the Medicare record were complete and who didn’t already have a cancer diagnosis: that left them with only 146,278 patients to analyse!  They then looked at any patients who later developed one of the 26 most common cancers, and compared them to the use of a rhBMP, either -2 or -7.  Generally it is thought that the bulk of the use was for rhBMP-2.  So what did they find?

Overall, there was no association of rhBMP admin- istration with cancer incidence (hazard ratio: 0.98, 95% confidence interval [CI]: 0.95–1.02). Similarly, when individual cancer sites were considered, there were no significant differences between the 2 groups.

In simple terms, they found there was no association between use of a rhBMP and cancer.

So we have two studies to look at.  One looked at high dose rhBMP and only 292 patients.  The other was retrospective, but it looked at 146,278 patients and included all types of lumbar fusion surgery, all doses, all techniques.  Which do I put more credence to?  The latter study.

It is highly unlikely that rhBMP-2 causes cancer.