By Daniel Lubelski, BA, and Thomas E. Mroz, MD
Cleveland Clinic is a non-profit academic medical center. Advertising on our site helps support our mission. We do not endorse non-Cleveland Clinic products or services Policy
In the five decades since their osteoinductive potential was first characterized, a multitude of bone morphogenetic proteins (BMPs) have been discovered and studied intensively. At the forefront of this group of soluble, low-molecular-weight glycoproteins is recombinant human BMP-2 (rhBMP-2), thanks to its unique, robust capacity to induce bone formation. BMP-2 participates in several signaling pathways and has roles in many important functions, including bone and cartilage growth and embryologic development.
rhBMP-2 has gained much attention in recent years due to controversy regarding its use in spine surgery. While rhBMP-2 was extolled after initial randomized controlled trials demonstrated high fusion rates without use of autogenous bone graft (a perceived merit of the product), more recent reports of associated complications have led some to halt or limit use of rhBMP-2 in fusion procedures.
These controversies have challenged the spine community to more extensively investigate the relative risks and benefits of rhBMP-2 in spine surgery.
A promising alternative to autogenous bone graft
The success of spine fusion surgery is often contingent on achieving solid bony fusion at the surgically treated levels. Autogenous bone graft, typically obtained from the patient’s iliac crest, is considered the gold standard for graft material to promote fusion. It is osteoconductive, osteoinductive and osteogenic. Unfortunately, using iliac crest bone graft can be associated with morbidity, including donor-site pain.
rhBMP-2, first approved by the FDA in 2002 for use with the LT-CAGE™ Lumbar Tapered Fusion Device in single-level anterior lumbar interbody fusion (ALIF) procedures, was thought to obviate the need for iliac crest bone graft. Since its approval, rhBMP-2 has been used off label for fusion surgery in other regions of the spine. In 2011-2012, between 148,000 and 165,000 patients received InFUSE™ Bone Graft, the commercialized recombinant form of BMP-2, as part of spine surgery.
The rapid adoption of rhBMP-2 by spine surgeons was undoubtedly spurred by positive early trials reporting that rhBMP-2 improved fusion rates, reduced operative time and blood loss, and had few to no complications.
Complications cloud the promise
Within a few years, however, reports of numerous complications were published.
In the first systematic review of complications associated with rhBMP-2, we identified 31 articles between 1990 and June 2009 that reported a mean incidence of 44 percent graft resorption, 25 percent graft subsidence and 27 percent interbody cage migration for posterior lumbar spine interbody fusion. Other notable adverse events included heterotopic bone formation, seroma formation and postoperative radiculitis. For anterior cervical fusion, the studies reported a mean incidence of 5.8 percent postoperative soft tissue problems, such as dysphagia and difficulty breathing.
Reports related to the adverse events in the anterior cervical spine led the FDA to issue a notification in July 2008 warning surgeons of potential minor and major complications of rhBMP-2.
Insights from the Cleveland Clinic experience
We proceeded to retrospectively study the Cleveland Clinic experience to identify complications associated with rhBMP-2 in lumbar fusion. We reviewed all patients between January 2002 and September 2010 who received rhBMP-2 with posterior/transforaminal lumbar interbody fusion (PLIF/TLIF), posterolateral fusion (PLF) or ALIF. Overall, 547 patients were reviewed, with a mean follow-up of 17 months. No differences in complication rates were found among the various surgical approaches, and, notably, complication rates were similar to those of historical controls who did not receive rhBMP-2.
Since rhBMP-2 is now frequently reserved for “harder to fuse” patients, such as smokers or those with prior spine surgery, it is noteworthy that in our cohort these two factors were significantly associated with an increased incidence of complications such as radiculitis, reoperation and pseudoarthrosis.
In our aforementioned study, urologic complications were rarely reported. However, urologic complications, particularly retrograde ejaculation (RE) associated with ALIF, received substantial attention following a report from Stanford University that RE occurred more frequently in patients receiving rhBMP-2 compared with control patients.
Digging deeper into urologic complications
To determine the Cleveland Clinic experience, we retrospectively investigated all male patients who had undergone ALIF with (n = 59) and without (n = 51) rhBMP-2 between 2002 and 2010. We found that the incidence of overall urologic complications did not differ between the two groups, occurring in 22 percent of those receiving rhBMP-2 vs. 20 percent of controls. There was likewise no significant difference in rates of RE, which was reported by 8 percent of patients in both groups.
Notably, only one patient had RE recorded in the electronic medical record, and only after telephone interviews with specific questions about urologic complications did we identify the extent of this complication (though it was not associated with rhBMP-2 use).
Turning research attention back to benefits too
The limitations of retrospective reviews are well recognized, however, and prospective evaluation of rhBMP-2 is needed to fully understand the risk-benefit ratio of this biologic. Understanding the complications is important for evaluating the utility of rhBMP-2 in spine surgery, but it is not enough. Defining the benefits of rhBMP-2 is needed as well.
We are currently systematically reviewing the literature to identify the reported fusion rates, a correlate to clinical success, associated with lumbar spine surgery. We have found that when the reported data are analyzed in combination, ALIF and PLF have significantly higher fusion rates in those who received rhBMP-2 as compared with controls, but there is no difference in fusion rates between groups for PLIF/TLIF. Moreover, there is limited long-term fusion rate evidence for PLIF/TLIF.
Our understanding of both the advantages and disadvantages of rhBMP-2 is far from complete, and further prospective investigation is needed before conclusions can be drawn.
Role-refining quest continues
The fact that BMP-2 forms bone is indisputable, as it is demonstrated by a multitude of preclinical and clinical studies. A mounting body of evidence suggests that its osteoinductive potential is largely predicated on the type of biological environment (i.e., type of fusion performed), the dose and the carrier of BMP-2 used.
Early enthusiasm for rhBMP-2 and its potential to obviate the need to use the patient’s own bone for fusion certainly has been tempered by its more recently defined adverse event profile. One other important consideration is the financial cost associated with use of this biologic, which is beyond the scope of this article.
In the coming years, it is likely that the attention now focused on the merits, drawbacks and costs associated with rhBMP-2 will ultimately help define its optimal indication and application for spine fusion procedures.
Mr. Lubelski is a medical student at Cleveland Clinic Lerner College of Medicine with a special interest in neurosurgery and spine surgery.
Dr. Mroz is a spine surgeon in Cleveland Clinic’s Center for Spine Health.