The aim of this study was to determine the incidence and assess specific risk factors in the postoperative development of retrograde ejaculation (RE) in men treated for degenerative lumbar disc disease at the L4–5 or L5–S1 level with stand-alone anterior interbody implants with or without recombinant human bone morphogenetic protein–2 (rhBMP-2).
Patients enrolled in 5 prospective, randomized, multicenter FDA-approved investigational device exemption studies were observed for a minimum of 2 years to assess the rate of RE. Five hundred eight men with symptomatic single-level lumbar degenerative disc disease with up to Grade 1 spondylolisthesis underwent anterior lumbar interbody surgery with stand-alone anterior implants at either L4–5 or L5–S1. All patient self-reported and physician-documented adverse events were recorded over the entire course of follow-up. In the investigational groups, 207 patients were treated with an open surgical procedure using dual paired constructs and rhBMP-2 on an absorbable collagen sponge. The control groups (n = 301) were treated with lumbar fusion cage implants and iliac crest autograft or a metal-on-metal disc arthroplasty device. Multivariate analyses of RE were performed to assess the influence of treatment (rhBMP-2), surgical approach, and treated level. Data were analyzed for each trial individually and for the data pooled from the 5 trials.
Retrograde ejaculation occurred at the highest rates in the earliest clinical trial. Of the 146 men, 6 (4.1%) developed RE postoperatively. In subsequent studies, the rates of RE ranged from 0% to 2.1%. Combining the data from the 5 trials, RE was reported in 7 (3.4%) of the 207 patients who received the rhBMP-2 treatment compared with 5 (1.7%) of the 301 patients who received the autograft or lumbar disc treatment (p = 0.242, Fisher exact test). Cases of RE were reported in 7 (1.6%) of 445 patients who underwent a retroperitoneal spinal exposure; 5 RE cases were reported in 58 patients (8.6%) who underwent a transperitoneal approach. The difference in surgical approaches was significant (p = 0.007, Fisher exact test). There was no difference in the rate of RE based on the lumbar level exposed (p = 0.739). Multivariate analyses were consistent with the conclusions from Fisher exact tests. In the initial rhBMP-2 trial, after adjusting for effects of surgical approach and treated level, the difference in RE between the treatment groups (rhBMP-2 vs autograft or disc arthroplasty) was not significant (p = 0.177); however, the difference in RE between the retroperitoneal and transperitoneal approaches was significant (p = 0.029).
In these 5 prospective randomized trials involving anterior lumbar interbody surgery, the use of rhBMP-2 was associated with a higher incidence of RE (3.4% vs 1.7%) but did not reach statistical significance. Based on surgical approach, the difference in rates of RE was statistically significant. This study reports on the outcomes of 5 prospective randomized FDA-approved investigational device exemption trials. Registration for studies became law in 2007. Four of these trials were completed before the law went into effect. The registration number for the lumbar disc arthroplasty trial is NCT00635843.