An individual patient data meta-regression for continence recovery following radical prostatectomy

Mungovan S F1, Graham P L2, Sandhu J S3, Scardino P T3, Coakley F V4, Matsushita K5, Ha H K6, Tienza A7, Choi S8, Kim S C9, Jeong S J10, Patel M I11

Research Type


Abstract Category

Prostate Clinical / Surgical

Abstract 252
Male Incontinence
Scientific Podium Short Oral Session 13
Wednesday 29th August 2018
16:45 - 16:52
Hall C
Male Incontinence Imaging Anatomy Surgery
1. The Clinical Research Institute and Westmead Private Physiotherapy Services, Westmead Private Hospital, Sydney, Australia, 2. Department of Statistics, Macquarie University, Australia, 3. Urology Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, 4. School of Medicine, Oregon Health and Science University, Oregon USA, 5. Department of Urology, Juntendo University, Tokyo, Japan, 6. epartment of Urology, Pusan National University Hospital, Busan, Korea, 7. Department of Urology, Clinica Universidad de Navarra, Spain, 8. Department of Urology, Seoul Medical Center, Korea, 9. Department of Urology, Haeundae Paik Hospital, Inje University, Busan, Korea, 10. Department of Urology, Seoul National University Bundang Hospital, Korea, 11. Department of Urology, Westmead Hospital, Sydney and Discipline of Surgery, Sydney Medical School, The University of Sydney, Australia

Sean F Mungovan



Hypothesis / aims of study
Following radical prostatectomy (RP), urinary incontinence is a predictable consequence, occurring in the majority of patients. The time to achieve continence (continence recovery) after RP is variable. The mechanism for the time dependent recovery of urinary incontinence is not clearly understood. Preoperative patient risk factors that affect continence recovery following RP have been reported. The preoperative length of the membranous urethral length (MUL) which is measured via T2 weighted images and age are two prognostic patient factors that can affect the recovery time to continence following RP. MUL is important because the structure and function of the membranous urethra is inherently associated with the urinary sphincter complex. A comprehensive understanding of prognostic risk factors is potentially of value to clinicians when counselling patients in clinical practice prior to surgery. Our aim was to i) Undertake an individual patient data (IPD) meta-regression for the recovery of continence at 6 and 12 months following RP from studies reporting prognostic factors, ii) Propose a multivariate logistic model for predicting postoperative continence recovery.
Study design, materials and methods
The corresponding authors of the 12 studies that were included in our previously published systematic review that examined the effect of MUL on the continence recovery following RP were invited to provide their individual patient datasets comprising i) patient characteristic data including age, prostate volume, MUL measurements ii) continence recovery (defined as no pad usage) at 6 and 12 months following RP. A one-step individual patient data (IPD) logistic meta-regression with a random-effect to control for within-study clustering was undertaken.
Six datasets reporting continence recovery (defined as no pad usage) at 6 months (4747 patients) and 6 datasets at 12 months (5241 patients) were analysed. Patients with a longer MUL (p<0.001), younger age (p<0.001) or smaller prostate volume (p≤0.003) had significantly higher odds of continence recovery at 6 and 12 months in both the univariate and multivariate models. Although prostate volume was statistically significant, the effect was very small and is likely not to be of clinical significance.Higher Gleason biopsy scores were associated with poorer outcomes in the short term (p≤0.022) but  not in the long term (p≤0.060). Using the multivariate model, it is possible to say for example that a 50 year old patient with a 9mm MUL and prostate volume of 35cm3 has a probability of return to continence at 6 and 12 months of 76% and 87% respectively. A 70 year old patient also with a 9mm MUL and a prostate volume of 35cm3 has a 56% and 69% probability of a return to continence at 6 and 12 months respectively.
Interpretation of results
A longer preoperative MUL, younger age and a smaller prostate volume are significantly and positively associated with a more rapid return to continence in men following RP. Our logistic modelling results can be translated into practice when counselling patients prior to surgery about the probability of a return to continence at 6 and 12 months following RP. IPD meta-regression provides more appropriate summaries of the effect of predictors as opposed to standard meta-regression analyses that may be subject to ecological bias.
Concluding message
A longer preoperative MUL and younger age are patient specific factors that are associated with a more rapid return to continence following RP.
Funding Dr Jaspreet S Sandhu were supported in part through the NIH/NCI Cancer Center Support Grant P30 CA008748. Clinical Trial No Subjects None