Carragee for the Prosecution

Eugene J. Carragee, M.D., editor in chief of The Spine Journal (TSJ), resumed his prosecutorial debating style in the journal’s latest issue as he responded to a letter to the editor citing omissions from the June 2011 issue of The Spine Journal and his own critical review of 13 early rhBMP2 studies.

The letter to the editor in question came from six frequently published (Journal of Bone and Joint Surgery and SPINE among other journals) and esteemed authors concerning one of their studies which Carragee et al. had selected for criticism in the June 2011 issue of The Spine Journal. In their letter this group of investigators found that Carragee had omitted key, material information which might well have changed Carragee’s conclusions regarding their study.

‘
Courtesy The Spine Journal and ElsevierCarragee’s response, published in the November issue of TSJ, left almost no room for further discussion and will, we expect, prove to be chilling to any other investigators who also found, as we did at OTW, a pattern of omission and error in the June 2011 issue of The Spine Journal.

The authors of the letter to the editor were Drs. Dimar, Glassman, Burkus, Pryor, Hardacker and Carreon.

Here are the citations for the two studies in question:

• Dimar JR, Glassman SD, Burkus JK, et al. Clinical and radiographic analysis of an optimized rhBMP-2 formulation as an autograft replacement in posterolateral lumbar spine arthrodesis. J Bone Joint Surg Am 2009;91:1377–86.

• Dimar JR, Glassman SD, Burkus KJ, Carreon LY. Clinical outcomes and fusion success at 2 years of single-level instrumented Posterolateral fusions with recombinant human bone morphogenetic protein-2/compression resistant matrix versus iliac crest bone graft. Spine 2006;31:2534–9; discussion 2540.

Who to Believe?

In 2006 Dimar, Glassman, Burkus, and Carreon published, in SPINE, the results from their part of a prospective, randomized FDA investigational device exempt (IDE) study of a new dose of rhBMP2 in a novel, compression resistant carrier for instrumented posterolateral spine fusions. The authors enrolled 98 patients. Forty-five were randomized to the iliac crest bone graft group. Fifty-three were randomized to the BMP/compression resistant matrix group.

Two years after the last patient was enrolled, the authors published a summary of their results. The six page report offered three conclusions and a table which listed 12 categories of adverse events for both the BMP patients and the control patients. Here are the conclusions:

1. Both patient groups, the ICBG group and the rhBMP2/CRM group, improved after surgery. Specifically, the two groups exhibited similar clinical outcomes as measured by the Short Form 36, Oswestry Low Back Pain Disability Index, leg and back pain scores at the two year mark after surgery.

2. The group receiving rhBMP2/CRM had less operative blood loss and a shorter operative time (no second surgery to harvest bone graft).

3. The rhBMP2/CRM group had a higher fusion rate based on two-year CT scans

Twelve categories of complications were listed in a table for both the rhBMP2/CRM and ICBG patients. Carragee said in his June 2011 article that the Dimar study did not mention BMP related complications. In a direct refutation of Carragee, OTW reproduced the Dimar et al. table in our critique of the June 2011 issue of TSJ.

Four years later, in July 2010, the FDA published an executive summary of the entire FDA monitored rhBMP2/CRS study—which included the Dimar et al. patients along with patients from 28 other study sites for a total of 463 patients of which 234 received the rhBMP2/CRS (now branded AMPLIFY) and 229 received ICBG.

The FDA’s summary ran 69 pages.

In that FDA report, the complications associated with the broader study were reproduced in detail. Ultimately, the range, type and severity of those complications were too much for the FDA and the AMPLIFY dosage of rhBMP2 was not approved for commercial sale.

All of this information was made available for free to the general public. The FDA’s conclusions were more extensive than those in the Dimar et al. study but, and this is key, they were consistent with and supported each of the three Dimar, et al. conclusions. Here is a brief list of those FDA conclusions which, like Dimar et al., looked at the 24-month outcomes.

• Investigational group (rhBMP2/CRS) and control groups (ICBG) were comparable in demographic and baseline characteristics (page 24).

• The average operative time was statistically less in the AMPLIFY patient group (page 25).

• The average blood loss was statistically less in the AMPLIFY patient group (page 25).

• At 24-month follow up, the rate of adverse events was NOT statistically significant between the two groups—87.4% of the AMPLIFY patients had an adverse event versus 87.9% of the control patients—[emphasis added] (page 26)

• At 24 month follow up, the AMPLIFY group had statistically lower rate of non-union adverse events—4.2% for AMPLIFY versus 10.3% for control (page 27).

But FDA Didn’t Approve AMPLIFY…

The FDA’s review, however, looked at data which was collected after the Dimar et al. study in 2006. With five-year data, the FDA decided that AMPLIFY, which has different pharmacokinetics and pharmacodynamics than InFuse, had too many safety questions to be approved for sale. Specifically, the FDA said (Under the FDA rules, ANY adverse event of any kind from any source was included):

• Of 25 categories of adverse events listed by the FDA, the AMPLIFY patients exhibited a statistically significantly greater incidence rate of adverse event than the control group in five areas (page 30). Those five areas were:

       • Arthritis/bursitis
       • Back and/or leg pain
       • Neurological adverse event
       • Trauma
       • Cancer

• There were a high number of serious back and/or leg pain adverse events in both groups but the rate in the AMPLIFY group was higher—10.0% versus 8.0% (page 31).

• The total number of serious adverse events was less for the AMPLIFY patients than for the control patients—52.7% versus 55.8% (page 31).

• The rate of cancer, however, tended to be higher among the AMPLIFY patients than for the control patients. Here we will quote directly from the FDA document: “Although the rates (of cancer) in the investigational group tended to be higher than those in the control, they are not statistically different at the 24 month analysis. However, statistical significance is borderline between the AMPLIFY and control groups when all cancer events though [sic] the 2010 Annual Review are considered” (page 33).

• The total number of AMPLIFY patient who reported having cancer at the 60-month follow up was 12 (out of a total population of 234) and the total number of control patients who reported cancer at the 60-month follow up was 5 (total control population of 229) (page 33).

• At 60-month follow up, the number of adverse events increased for both groups (page 27).

• At 60-month follow up, the AMPLIFY group had a lower rate of adverse events than the control group—92.9% for AMPLIFY versus 93.8% for control (page 27).

Bottom line: the FDA said: “Long-term data, although not intended as the primary outcome measure of success, suggests a less favorable profile. There is a concerning number of cancers in this study and all rhBMP-2 clinical spine studies. Recombinant BMP-2 has systemic effects, not unlike any other drug, and the medical community does not have enough information that relates to its long term pharmacological effects.”

So, the System Worked? 

Looking back and having read both Dimar and the FDA review, it seems to us that the six page study that was published in SPINE in 2006 should have been more comprehensive and could have included more information about complications.

But the Dimar study, in contrast to Carragee’s work, was unemotional, data driven and when it came to conclusions, landed on three simple ones and clearly kept the door open for more study. It did NOT promote the product in any way that this reader could discern.

Carragee for the Prosecution

So we return to the question of who to believe. The problem with Carragee is that he is aggressively promoting a point of view and is omitting information which would modify his conclusions. These omissions are clearly troubling to a growing number of experienced and respected North American Spine Society (NASS) members. When an author is omitting information, the question of bias and agenda become paramount.

Courtesy: North American Spine SocietyWe have disclosed some of Carragee omissions in previous OTW articles.

Carragee’s style of aggressive promotion has found its way into the form of argumentation that he employed in his response to the letter from Dimar et al. Simply put, Carragee is employing a prosecutorial style—which is a style we would not customarily expect to find in the pages of The Spine Journal.

What is a prosecutorial style of argumentation? 

There are three basic elements:

1. Impugn the credibility if not integrity of any writer who would argue against the proposition.

2. Avoid ambiguity. Ignore information which might conflict with the narrative of the case.

3. Find a simple storyline and hammer it home. For example: ‘corrupt surgeons spin study data and therefore NASS’s membership lives dangerously.’ 

In Carragee’s response to Dimar et al., these elements of the prosecutorial style were, we believe, in full view.

Examples of Carragee’s Prosecutorial Style

Here, for example, is a selection from Carragee’s response to the Dimar, Glassman, Burkus et al.’s letter to the editor of The Spine Journal; “The authors received substantial sums for consulting, royalties, and other support from Medtronic and had extensive financial ties with this manufacturer for many years before and after publication.”  Then, two paragraphs later, Carragee writes: “It is clear, however, that multimillion dollar corporate funding was received to perform both of these studies, a fact, unfortunately, omitted or denied by the original authors in both disclosures.”

Clearly Carragee is accusing these authors of taking “multi-million corporate funding” to perform the studies. First of all, Carragee’s comment is factually incorrect. OTW has been checking with these authors and so far, every one of them has told us that they received no compensation for performing the studies. None. At all. Secondly, as we documented in earlier articles, the money Carragee is referring to is for any payment made for anything at anytime paid to any single author listed in the study. Carragee made no effort to find out if any of these payments could, in fact, be reasonably connected to these studies or rhBMP2.

So, as we documented in earlier articles, a researcher (and we used Dr. Scott Boden as an example—although the same is true of other researchers mentioned by Carragee) who refused compensation for conducting the study was then credited by Carragee with receiving millions of dollars of royalties that were actually paid to an entirely different surgeon for a purpose entirely unrelated to rhBMP2 or the study in question.

But perhaps most worrisome, as we documented previously in OTW (Medtronic and Carragee on Collision Course in Court?, October 25, 2011), one lawsuit has already been filed which cites Carragee’s work as a key piece evidence for the plaintiff.

The fact that Carragee uses the prosecutorial style of argumentation, we believe, encourages the plaintiff’s bar.

We at OTW are trying to apply more light than heat to this issue, but as the evidence of omissions and mistakes associated with Carragee’s work continues to grow and as Carragee increasingly assumes the mantle of prosecutor if not also judge and jury, we are becoming ever more worried and concerned.

Space in this article precludes us from listing every material omission or mistake that sadly now characterizes Carragee’s work in TSJ regarding rhBMP2 and the investigators who have studied the compound.

But, stay tuned. We are collecting examples of Carragee’s omissions and errors and will be supplying them to both our readers and NASS soon.