1 Dept. Ofurology, AO Papa Giovanni XXIII, Bergamo

PREDICTIVE ACCURACY OF NEPHROMETRIC SCORES CAN BE IMPROVED BY ADDING CLINICAL PATIENT CHARACTERISTICS: A NOVEL ALGORITHM COMBINING ANATOMIC TUMOUR COMPLEXITY, BODY MASS INDEX AND CHARLSON CO-MORBIDITY INDEX to depict PERIOPERATIVE COMPLICATIONS AFTER NEPHRON SPARING SURGERY

Roscigno Marco1, Francesca Ceresoli1,Capitanio Umberto2, Rayan Matloob2, Richard Naspro1, Deiana Gianfranco1, Dehò Federico1 Ettore Di Trapani2, Carenzi Cristina2, ,Montorsi Francesco2, Bertini Roberto2, Da Pozzo Luigi Filippo1.

1 Dept. ofUrology, AO Papa Giovanni XXIII, Bergamo.

2 Dept. of Urology, Vita-Salute San Raffaele University, Milan.

Objective:Even though tumor complexity represents the most important predictor of genitourinary complications, the overall rate of surgical complications may be influenced also by other patient characteristics. We developed a user-friendly algorithm, that combined anatomical features, BMI and Charlson Co-morbidity Index (CCI), to predict perioperative complications in patients undergoing nephron-sparing surgery (NSS).

MATERIALS AND METHODS: From 2010 to 2012, we prospectively collected 320 consecutive patients treated with open NSS. Patients underwent open trans-peritoneal NSS via a median xifo-umbilical approach or an extraperitoneal NSS through a flank incision. Complications within 40 perioperative days were collected and graded according to the modified Clavien-Dindo Classification System (CCS), and stratified into minor (CCS 1-2) and major complications (CCS 3-5). Multivariable logistic regression analyses using backward selection tested the predictive value of age, gender, BMI (≤25 vs >25), CCI (0 vs 1-2 vs >2), PADUA score (6-7 vs 8-9 vs ≥10) or R.E.N.A.L. nephrometry score (4-6 vs 7-9 vs 10-12), and surgical approach on overall complication rate. Finally, the most parsimonious risk model in predicting the outcome of interest was developed. The ability of the score of this risk model to predict overall complications was tested in multivariate analysis.

RESULTS: One-hundred sixty three patients underwent open extraperitoneal NSS through a flank incision, while 157 patients underwent a transperitoneal approach. Mean patient age was 63±12 years. The mean tumor diameter was 3.3±1.5 cm. The median PADUA score was 9 (range 6-13). The median R.E.N.A.L. score was 7. Mean ischemia time was 20±9 min (median 19 min). A novel algorithm integrating anatomical features and patient characteristics was generated (Risk of Surgical Complication Score – RoSCo Score).Specifically the score included 3 independent variables: anatomical tumour features (PADUA score 6-9 = 1 point; PADUA score ≥10 = 3 points or R.E.N.A.L. score 4-9 = 1 point; R.E.N.A.L. score ≥10 = 5 point); 1,2 and 3 points for CCI 0, 1-2 and >2, respectively; 1 point for BMI ≤25 and 2 points for BMI >25. Patients were categorized according to our score algorithm as 3-8 when using PADUA score. Patients with score 3-4, 5-6 and 7-8 had 14%, 27% and 52% overall complication rate. Our score algorithm was 3-10 when using R.E.N.A.L. score. Patients with score 3-4, 5-6 and 7-10 had 18%, 33% and 51% overall complication rate. the predictive accuracy of RoSCo score for surgical complications was 68%, with a gain of 4% relatively to the use of PADUA only (64%) and of 6% relatively to the use of R.E.N.A.L. score only.

DISCUSSION: The RoSCo score not only accounted for anatomic tumor complexity but also for additional clinical characteristics that may affect non genito-urinary complications.

CONCLUSIONS:This novel tool could help clinicians better stratifypatients candidates to undergo NSS into subgroups with different risks of perioperativecomplications.