Center volume is associated with outcome following pancreas transplantation within the Eurotransplant region

W.H. Kopp, M. van Meel, H. Putter, U. Samuel, H. Arbogast, W. Schareck, J. Ringers, A.E. Braat

Chair(s): dr. Robert A. Pol, vaat-en transplantatiechirurg, UMC Groningen & dr. Wojaiech G. Polak, chirurg, Erasmus MC

Thursday 10 march 2016

9:00 - 9:10h at Zaal 5 & 6

Categories: Parallelsessie (klinisch)

Parallel session: Parallelsessie VIII - Klinisch chirurgisch en acuut

Outcome following all kinds of surgery depends on several factors, amongst these the annual volume-outcome relationship. This might also be the case in a highly complex field of pancreas transplantation. No study has investigated this relationship in a European setting.

All consecutive pancreas transplantations from January 2008 until December 2013 were included. Donor, recipient and transplant related factors were analyzed for their association with patient and graft survival. Centers were classified in equally sized groups as being low volume (<5 transplantations on average each year in the 5 preceding years), medium volume (5-13/year) or high volume (≥ 13/year).

In the study period, 1276 pancreas transplantations were included. Mean duration of follow up was 3.2 (SD 1.8) years. Low volume centers performed 396 (32.6%) transplantations, medium volume 425 (35%) transplantations and high volume 393 (32.4%) transplantations. Unadjusted 1 year patient survival was associated with center volume and was best in high volume centers, compared to medium and low volume: 96.5%, 94% and 92.3%, respectively (p=0.002). Pancreas donor risk index (PDRI) was highest in high volume centers: 1.38 vs. 1.21 in medium and 1.25 in low volume centers, (p<0.001). Unadjusted 1 year death censored pancreas graft survival did not differ significantly between volume categories: 86%, 83.2% and 81.6%, respectively (p=0.110). However, after multivariate Cox-regression analysis, higher PDRI (HR 1.61, p=0.002), retransplantation (HR 1.90, p<0.001) and higher recipient BMI (HR 1.04, p=0.044) were independent risk factors for pancreas graft failure. High center volume was protective for graft failure (HR 0.65, p=0.014) compared to low center volume. Medium volume was protective compared to low volume, albeit not statistically significant (HR 0.87, p=0.367).

Patient and graft survival following pancreas transplantation is better in higher volume centers. High volume centers have good results, even though they transplant organs with the highest risk.