Balloon Kyphoplasty Improves Both Roland-Morris Disability Scores and Bone Pain among Cancer Patients with Vertebral Compression Fractures: Interim Analysis of Results from a Phase IV Random Trial

Destructive vertebral lesions are a common source of morbidity among patients with cancer. Balloon kyphoplasty is a minimally invasive surgical procedure performed for patients with painful vertebral compression fractures (VCFs) with the goal of reducing pain and disability and improving quality of life. We report the results of the first randomized trial among cancer patients with VCFs to assess the efficacy and safety of this procedure. Twenty-one sites in Europe, the United States, Canada and Australia enrolled 134 patients after consent and ethical review board approval. Adult patients diagnosed with a variety of cancers and ≤ 3 painful VCFs (VAS ≥ 4) were randomly assigned to immediate kyphoplasty (N=70) or nonsurgical supportive care (N=64). Patients with primary bone tumors, osteoblastic tumors or solitary plasmacytoma at the fracture site were excluded as well as patients with spinal cord compression. The primary objective was to determine the change in the Roland-Morris Disability questionnaire, a 0- (no disability) to 24-point (maximum disability) instrument validated for assessing back-specific physical functioning, at one month. Back pain was also assessed using a validated 0- (no pain) to 10-point (worst pain imaginable) numerical rating scale. Data from a preplanned interim analysis of this ongoing study are now reported. For pain and function, patients with complete data that has been evaluated through one month are included whereas all enrolled patients were analyzed for safety. Mean patient age was 64 years (range 37 to 88), 58% were female, and tumor types included multiple myeloma (36%), cancers of the breast (20%), lung (8%), prostate (6%) and other sites (30%). At study enrollment, 23% of patients were on daily corticosteroids and 48% had received bisphosphonates within 12 months of study entry. Prior to randomization, single VCFs were identified by the local investigators in 43% of patients; an equal proportion (29%) of patients had 2 or 3 fractures. Among the kyphoplasty and nonsurgical cohorts, 59 and 56 subjects, respectively were evaluable for the efficacy analysis. At baseline, average Roland-Morris scores were similar between the groups; 17.7 and 18.3 points for kyphoplasty and non-surgical-treated patients, respectively. However, at one month, there were marked differences between the two groups with a mean improvement for patients randomized to kyphoplasty of −8.3 points (95% CI −6.2 to −10.5) whereas those receiving non-surgical care showed no significant change (−0.1 points, 95% CI 0.9 to −1.0; p<0.0001 for difference). Mean baseline pain scores were also not different between the two groups (7.2 and 7.3 points for the kyphoplasty and nonsurgical groups, respectively). At one week, kyphoplasty-treated patients showed significant improvement in their back pain (−3.6 points, 95%CI, −2.8 to −4.4) whereas those patients treated non-surgically had no change in their pain (−0.3 points, 95%CI, 0.1 to −0.8; p<0.0001 for difference). Similar results for pain were obtained at one month; kyphoplasty resulted in a −4.1 point change (95%CI, −3.2 to −4.9) and those patients treated non-surgically had no change in their pain (−0.5 points; 95%CI, 0.04 to −1.0; p<0.0001 for difference). There was no significant difference in the number of patients with serious adverse events between the kyphoplasty (16) and nonsurgical (10) groups at one month. None of the serious adverse events in the kyphoplasty group were related to the devices used, including bone cement extravasation; one serious adverse event in the form of an intra-operative non-Q-wave myocardial infarction resolved and was attributed to anesthesia. This randomized study shows, at a pre-planned interim analysis time point, that patients with cancer-related VCFs treated with immediate balloon kyphoplasty show a marked reduction in back disability and pain at one month compared to non-surgical treatment. Pain reduction was also statistically significantly improved within one week postoperatively. Importantly, minimally clinically important differences for the Roland-Morris and pain scales used in this trial have been determined from previous studies and are estimated to be approximately 2.5 and 2.0 points, respectively. Thus, these improvements in disability and pain with balloon kyphoplasty were both statistically and clinically significant and achieved without an increase in adverse events.