New data mean indications for rhBMP-2 need re-defining


By Jan Sys

The true benefit of bone graft substitutes, such as rhBMP-2, is avoiding donor site morbidity. Studies show that the incidence of donor site pain following iliac crest bone graft prelevation ranges from 2.8 to 60%1, and this wide variation is because some studies reported that patients experienced pain two years postoperatively while others reported no pain at the donor site compared with the contralateral side after follow-up of more than one year.

Data from the early studies indicate that surgical time, blood loss, hospital stay, reoperation rate, and median time to return to work were lower and fusion rates were significantly higher with rhBMP-2. This led to the use of the product steadily increasing over the past 10 years to the extent that it has almost replaced iliac crest bone graft as the gold standard in fusion surgery at the lumbar spine. However while rhBMP-2 is widely used for all patient groups in the USA, its use in the Europe has been more cautious.

Since 2007, complications of rhBMP-2 in transforaminal lumbar interbody fusion or posterior lumbar interbody fusion surgery have been reported. For example, the product has been associated with radiculitis, endplate resorption, osteolysis, ectopic bone formation, subsidence, cage migration, non-union, inflammatory cyst or seroma formation, and carcinogenicity. Furthermore, retrograde ejaculation and urinary retention were increasingly reported in anterior lumbar interbody fusion surgery2-3.

In a prospective randomised controlled trial4, in patients with posterior lumbar interbody fusion surgery, both rhBMP-2 and iliac crest bone graft were shown to achieve uncomplicated fusion. However, trabecular bone formation and bone density were slower in the group who received rhBMP-2. Also, there was a high incidence of inflammatory cyst formation (10.5%), endplate resorption (100%), osteolysis (36.8%) and ectopic bone formation (36.8%) in the rhBMP-2 group even though only two thirds of the usual dosage was used. These adverse events remained asymptomatic and no increase of ectopic bone could be noted after two years.

This paper and other recent papers have suggested that adverse events with rhBMP-2 were under- or even not reported at all in the first publications—consequently exaggerating the potential benefits of bone substitutes². This was stated in a retrospective review of FDA documents3.

Consequently, indications for the use of rhBMP-2 in lumbar spinal fusion surgery should be re-defined. In primary surgery in the young, professionally active, male population, rhBMP-2 should be used with caution and iliac crest bone graft should be regarded as the gold standard. For long fusions, other alternatives (allograft or demineralised bone matrix etc.) should be considered because carcinogenicity of rhBMP-2 is mainly suspected when used in high dosage. There might remain a place for rhBMP-2 in posterior lumbar spine surgery without exposure of the neural elements, in revision surgery, in scoliosis surgery in the elderly population, and in anterior lumbar fusion in females. Patients should be informed thoroughly and their consent for the use of rhBMP-2 is essential.

Also, dosage determination should be reconsidered, as adverse events occurred when only two thirds (8mg) of the usual dosage of 12mg was used. Further research is needed to improve binding of the product to the collagen carrier and avoid displacement outside the disc space. Creating a barrier between neural elements and the disc space with a hydrogel sealant system may reduce the incidence of ectopic bone formation. Hydrogel sealant can reduce ectopic bone formation but it can lead to other adverse events. A case of a cauda equina syndrome has been reported, which is a major complication with sometimes irreversible functional loss.


Jan Sys is at Department of Orthopaedics and Traumatology, Sint Basius Hospital, Belgium


1. Howard et al. Spine J 2011; 11:534–37

2. Carragee et al. Spine J 2011; 11:463–68

3. Carragee et al. Spine J 2011; 11:471–91

4. Michielsen et al (inc Sys J) J Bone Joint Surg Am 2013; 95: 881–87