Session Item

Early diagnosis of malignant spinal cord compression (SCC) is crucial as pre-treatment neurologic status is the major determinant of outcome. In metastatic castrate resistant prostate cancer (mCRPC) SCC is a significant cause of disease-related morbidity. In the PROMPTS trial we investigated whether screening for SCC with spinal MRI, with pre-emptive treatment if radiological SCC (rSCC) was detected, reduced the incidence of clinical SCC (cSCC) in asymptomatic mCRPC patients.

PROMPTS is a phase III parallel-group, randomised controlled superiority trial.  CRPC patients aged at least 18 years with spinal metastases without related back pain or neurological symptoms, no previous SCC, and no spinal MRI in previous 12 months were eligible. Participants were randomly allocated (1:1 ratio) to control (no MRI) or screening spinal MRI. Allocation was not masked. A validated epidural spinal cord compromise scale (ESCC) was used. Pre-emptive treatment and 6-monthly spinal MRI were offered to patients with screen-detected rSCC. The primary endpoint was incidence of cSCC at 12 months.

Between Feb 26, 2013 and April 25, 2017, we randomly assigned 420 men from 45 UK centres to control (n=210) or screening MRI (n=210).  Median age was 74 years (IQR:68-79), 53% (222/420) had normal alkaline phosphatase and median PSA was 48.0ng/ml (IQR:17-162). rSCC was detected at screening in 61/200 (30.5%) intervention group patients. Concordance of local and central assessments of rSCC was good (92.4%). At 12 months, the cumulative incidence of cSCC was 6.7% (95% CI 3.8-10.6) in the control group and 4.3% (2.1-7.7) in the intervention group (Gray's test p=0.119, HR:0.64 95%CI:0.37-1.11, Fig 1). In the intervention group cSCC developed more commonly in the rSCC +ve cases (11.5%) than the rSCC -ve cases (1.3%) and the rSCC-ve group had significantly less cSCC than the control group (Gray's test p=0.04, Fig2). Intervention for rSCC was with radiotherapy (RT) in 50/61 (82%) cases, only 4% progressed at the treated sites. RT was not given to 18 sites with  early rSCC (ESCC 1a-b,17: 1c, 1) but none progressed at 6 months. At the time of cSCC 70% of patients were ambulant and 18% of non-ambulant patients regained function post-RT. More spinal RT was given in the intervention than control group (86vs49 courses) but the use of additional systemic therapy was significantly less by 12 months (54%vs70%, Gray's test p=0.003).

We found no statistically significant differences in incidence of cSCC between the intervention and control groups. MRI screen-detected early rSCC lesions do not always progress to cSCC with contemporary systemic management of CRPC and observation may be reasonable for ESCC grade 1a/b lesions. However these patients remain at substantial risk of developing new sites of cSCC. Further efforts to better identify patients at high risk for rSCC and cSCC are warranted to refine selection of groups for screening spinal MR