Advances in spinal fusion surgery have raised concerns about the risk of cancer in some procedures.

Here’s an interesting article that appeared in the Back Letter that reviews the main points from a recent presentation at the North American Spine Society.

Cancer Risks With Spinal Fusion Surgery

The controversy surrounding the use of the bone growth factor BMP-2 (bone morphogenetic protein-2) in spinal fusion surgery continues to ripple across the spine field.

BMP-2, of course, is the genetically engineered growth factor developed to speed bony healing of the spine (InFuse, Medtronic, Minneapolis). It may be employed in as many as 100,000 fusion procedures every year. At this year’s annual meeting of the North American Spine Society (NASS) in Chicago, the most eagerly anticipated presentation—and the one most widely reported in the mass media—was a new analysis by Eugene J. Carragee, MD, and colleagues on cancer risk associated with the use of BMP-2 in fusion surgery.

Any strong link between BMP-2 and cancer could knock BMP-2 completely out of routine use in spinal surgery and lead to thorny liability problems—not to mention tragic outcomes for patients seeking surgical solutions for common back problems. So surgeons and patients are understandably concerned about this issue.

No Cancer Link, According to Medtronic

Medtronic, the developer of BMP-2, had reported at FDA hearings in 2010 that there were no statistically significant associations between BMP-2 use and cancer in any of the clinical trials submitted to the FDA.“

All the results support a conclusion that there is not a direct relationship between rhBMP-2 and malignancy,” according to Medtronic. (See FDA, 2010.)

Carragee directly contradicted Medtronic’s position in his presentation at NASS. In the new analysis, he and his colleagues found that patients who received BMP-2 in the trials performed under FDA auspices (i.e. in pooled study data) were more than twice as likely to develop cancer compared with control subjects. (See Carragee et al. [a], 2011.)

Patients who received a higher dose of BMP-2 in the AMPLIFY trial (a randomized trial evaluating BMP-2 in posterolateral fusion) were more than three times as likely to develop cancer over the course of three-year follow-up as control subjects.

“Despite strict exclusion criteria in the AMPLIFY trial—which excluded patients with a history of serious cancer—a strong association between BMP-2 and cancer was seen by a variety of statistical methods,” Carragee reported. And these differences were statistically significant.

Because of those strict exclusion criteria in the AMPLIFY trial, Carragee suggested that his new analysis may actually underestimate cancer risk related to BMP-2 in a more typical spine surgery population.

Carragee and colleagues are not alone in finding an association between BMP-2 and cancer. Carragee cited an FDA analysis that came to a similar conclusion.

“The primary statistical concern is the apparent association with malignancy,” according to the FDA. “In this regard, there were higher rates of cancer events with the use of the AMPLIFY product in the pivotal study, which were not contradicted by all of the pooled Medtronic trials using BMP-2. In addition, higher rates of malignancy were observed when considering all high-dose use of BMP-2. Therefore, this issue requires careful consideration and a cautious path forward.” (See Carragee et al. [a], 2011.)

“We agree with this statement,” said Carragee.

“The Spine Literature Has Been Silent on This Issue”

Despite these reported associations, Carragee observed that there has been almost no discussion of the reported cancer risk in studies published by Medtronic-supported researchers. “The spine literature has been silent on this issue,” Carragee said ominously. “There have been reports from the AMPLIFY trial without discussion of the cancer data,” he explained.

Six researchers who took part in the AMPLIFY trial did recently report in a letter to the Spine Journal that there were more cancers in the BMP-2 group than in the control group in the AMPLIFY trial over two-year follow-up, but that the differences did not achieve statistical significance. And these researchers felt that the diversity of cancers found in the BMP-2 group “did not suggest a causal link.” (See Dimar et al., 2011.)

What Should I Tell My Patients Next Tuesday?

Attendees at NASS peppered Carragee with questions after his presentation. Los Angeles-area spine surgeon Robert Watkins, MD, asked Carragee the question that is on many surgeons’ minds: “What do I tell my patients?”

“Do you feel InFuse [BMP-2] causes cancer? I am a little unclear based on your study,” said Watkins. “Is there a ‘yes’ or ‘no’ answer?”

“I agree with the FDA’s analysis that there is a strong association with cancer,” Carragee responded. “Proof of causation of cancer, however, requires a very high evidence bar—as 40 years of litigation from the tobacco industry about smoking and cancer have demonstrated.”

Watkins then asked a more specific question. “Getting down to a clinical level, what should I do on Tuesday? What type of advice should I give my patients?”

“In an individual patient, the overall odds that the use of BMP-2 will lead to cancer is very low,” said Carragee. But the association is still there, he added. And the association appears to be stronger with higher doses. It may be stronger in older patients who have less resiliency to cancer. And it may be stronger in patients with a propensity for cancer (i.e. various risk factors for its development).

Another attending surgeon asked if the cancers were a local response to the application of BMP-2 in bone grafting. “Are these cancers all localized to the pelvic/abdominal region?” he asked. “We have always heard that BMP-2 has a transient local effect,” he commented.

“The tumors were all over the place,” Carragee responded. “In the BMP-2 group, there was an assortment of tumors—and not the mix you would expect to see in this population.” They included lung cancer, pancreatic cancer, prostate cancer, stomach cancer, ovarian cancer, and breast cancer among others.

The potential cancer mechanisms aren’t clear. “It is unlikely that these cancers stem from a direct carcinogenic effect [where BMP-2 is the root cause of the cancer],” said Carragee. “The lead time is too short.

“I think the most likely hypothesis is that there is a cancer-promotion effect related to the use of BMP-2,” he explained.

The Stanford spine researcher suggested that there is a balance between cancer formation and immune system surveillance in the human body.

“As people get into their forties, fifties, and sixties, there is a certain level of cancer mutation going on in the body, which the immune system wipes out,” said Carragee. But theoretically, if some drug or compound upsets that balance, there can be a rapid escalation in cancer growth.

A Study of Two Sets of Data

Carragee’s presentation was brief, and it is difficult to extract complex study data from a series of slides. However, here is a capsule description of the new analysis.

In one part of the study, Carragee et al. looked at pooled data from all the studies Medtronic submitted to the FDA comparing fusion with BMP-2 and fusion with autogenous bone graft.

“These data showed that there were about twice as many cancers in the BMP-2 group,” Carragee commented. “There is about a 1% chance of cancer over time. And if you split the data up into no BMP, low-dose BMP, and high-dose BMP, there appears to be a dose-response curve.”

Carragee acknowledged, however, that this type of analysis has limitations.

“What are the problems with the pooled data?,” he asked. “[The groups] are not well matched, there are mixed dosages, the approaches are different, the uptake is different, and carriers are different,” he answered.


Carragee noted that the best data come from individual randomized controlled trials (RCTs), such as the AMPLIFY trial, which was designed in a way that allowed the detection of cancer.

The AMPLIFY RCT was a non-inferiority trial that included 463 men and women with chronic back pain and single-level “degenerative disc disease.” The researchers randomly allocated the subjects to one of two treatment approaches: (1) instrumented posterolateral fusion with a concentrated dose of BMP-2 in a compression-resistant matrix (239 patients); or (2) instrumented posterolateral fusion using a bone graft harvested from the iliac crest (224 patients). (See FDA, 2010, for study details.)

Carragee pointed out that the two study groups in AMPLIFY were remarkably well-matched at baseline. As mentioned above, there were strict exclusion criteria. All patients with current cancer or a history of serious cancer were excluded. (Of patients with a history of cancer, only those with a history of minor skin cancers were allowed to participate.

The AMPLIFY trial also had reasonable follow-up—more than 70% of patients were followed for more than three years.

“The analyses we did at Stanford were to look at the number of cancer events that occurred in both groups by years two and three after surgery—and to analyze the results both by the number of enrolled and the numbers followed,” said Carragee. The researchers also performed a variety of secondary analyses.

Overall, there were 20 cancers in the BMP group at study closure compared with only five in the control group. There were multiple cancers in three patients in the BMP-2 group and none in the control group. There were three deaths due to cancer in the BMP group and one in the control group.

Multi-Fold Differences in Cancer Rates Between Study Groups

In the primary analysis involving one or more cancers per patient, analyzed by initial enrollment, after year two roughly four times as many patients in the BMP-2 group developed cancer compared with the non-BMP group (p = .04). After three years, roughly five times as many patients in the BMP-2 group developed cancer compared with the control group (p = .02).

To give an idea of the magnitude of cancer incidence here is a quote from Carragee. “Looking at the per-patient analysis, at years one, two, and three, 2%, close to 4%, and 5% [respectively] of the high-dose BMP-2 group [developed one or more cancers], while 1%, or a little more than 1%, of the non-BMP group [developed cancer].”

When the researchers looked at the occurrence of one or more cancers by actual follow-up numbers, there was a similar pattern of results. At the end of year two, there were roughly five times as many cancers in the BMP-2 group as in the control group (p = .02). After three years, there were almost four times as many cancers in the BMP-2 group compared with the control group (p = .03).

As mentioned above, Carragee closed his presentation by pointing out that there was a “strong association” between the use of BMP-2 and cancer in multiple statistical analyses involving different projections of the data.


Carragee EJ et al. (a), Cancer association with BMP-2, Presented at the annual meeting of the North American Spine Society, Chicago, 2011; as yet unpublished.

Carragee EJ et al. (b), A critical review of recombinant human bone morphogenetic protein-2 trials in spinal surgery: Emerging safety concerns and lessons learned, The Spine Journal, 2011; 11:471–91

Dimar JR et al., Reply to “A critical review of recombinant human bone morphogenetic protein-2 trials in spinal surgery: Emerging safety concerns and lessons learned,” The Spine Journal, 2011; epub ahead of print

FDA, FDA Panel Pack: Amplify rhBMP-2 Matrix, June 2010 (Read Full Report)

From The Back Letter >>>

Do you have additional questions or concerns about cancer risks with spinal fusion surgery? Leave your question or comment in the section below.

Before you go, do us a favor and click the like button to share this information with your friends on Facebook and (if you don’t mind) the +1 button to let Google know you liked what we have to say about back pain. We appreciate your willingness to help us get the word out.

And… don’t forget to reserve your seat to the next Virtual Back Pain Clinic webinar. The event is absolutely free. Register Here