Invasive intraductal papillary mucinous neoplasm and pancreatic ductal adenocarcinoma : a comparison in clinicopathology and long-term outcome

Background: The resections for both pre-malignant and invasive intraductal papillary mucinous neoplasm (inv-IPMN) have increased the last decades. Long term outcome, and the impact of adjuvant chemotherapy, non-regional lymph node status and recurrence pattern on overall survival (OS) is known for conventional pancreatic ductal adenocarcinoma (PDAC), but not so for inv-IPMN.

Aims: I) Investigate differences and similarities in clinicopathology and overall survival between patients resected for inv-IPMN and PDAC. II) Elucidate whether the raised numbers of pancreatic resections for inv-IPMN in combination with the improvement in OS recent years have influenced outcome. III) Assess the prognostic significance of para-aortal lymph node (PALN) involvement in patients resected for inv-IPMN and PDAC in the pancreatic head. IV) Explore the impact of adjuvant chemotherapy and spatio-temporal recurrence pattern on overall survival for inv-IPMN compared with PDAC.

Methods: All studies were retrospective observational studies of consecutive patients ≥ 18 years of age resected for inv-IPMN and PDAC. Study I-III were single-centre studies of in total 515 patients resected at Karolinska University Hospital between 2009-2018, Study IV was a national multi-centre study of patients resected in Sweden between 2010–2019. Clinicopathological variables were analysed in multivariable Cox regression models. Outcome was assessed by calculating two- or three-year OS rate and estimating OS using the Kaplan-Meier model. Survival functions were compared with log-rank test.

In study I were clinicopathological variables also analysed in multivariable logistic regression models. Study II only comprised patients with PDAC or inv-IPMN in the pancreatic head who underwent partial or total pancreatoduodenectomy including PALN resection. In study III, that only included patients residing in the Stockholm area, different initial recurrence sites and time frames as well as predictors for death including the impact of adjuvant chemotherapy were assessed with multivariable logistic and Cox regressions. In study IV, clinicopathological variables were retrieved from the Swedish national pancreatic and periampullary cancer registry. The effect on death was assessed in two multivariable Cox regression models, one for patients resected 2010-2015, one for patients resected 2016-2019.

Results: In study I, 513 patients were included, 122 inv-IPMN and 391 PDAC. The proportion resected inv-IPMN and two-year OS increased during the study period. In Kaplan-Meier survival analysis, inv-IPMN had more favourable median OS compared to PDAC. In multivariable Cox Regression analysis, tumour type was not a predictor for death.

In study II, 403 patients were included, 89 inv-IPMN and 314 PDAC. PALN were metastatic in 16% and there was no difference between the groups. N0- and N2-stage were present in 16% and 53% respectively for patients with inv-IPMN compared to 6% and 65% respectively for patients with PDAC (p=0.007). Median OS was 12.7 and 22.7 months in the presence or absence of PALN metastases respectively (p<0.001), and similar in N2-stage regardless the presence of PALN status. PALN status was not an independent prognostic factor.

In study III, 396 patients were included, 92 inv-IPMN and 304 PDAC. Both recurrence rate and death rate within three-years were lower for inv-IPMN compared to PDAC. The most common recurrence patterns were multi-site (25%), single-site liver (21%) and single-site locoregional (10%) recurrence. The most important predictors for death were multi-site, single-site peritoneal and single-site liver recurrence. These predictors were less common in inv-IPMN compared to PDAC. Adjuvant chemotherapy had similar effect in the two groups.

In study IV, 1909 patients were included, 293 inv-IPMN and 1616 PDAC. Tumour type was an independent predictor for death in the 2016-2019 cohort, but not in the 2010-2015 cohort. In Kaplan-Meier survival analysis, inv-IPMN was associated with longer median OS in stage N0-1 and in stage M0 compared to PDAC. However, in stage T2-4 and stage N2 median OS was similar, whereas median OS in stage M1 was even shorter for inv-IPMN compared to PDAC.

Conclusions: Inv-IPMN seemed to have favourable survival outcome compared to PDAC in lower stages, and similar to worse in higher. Outcome was dependent on the combination of a pronounced increase in resected inv-IPMN and a concurrent hazard reduction for death within 2 years during the study period. PALN status is not an independent risk factor for death and does not influence survival in N2-staged disease. The M1-stage for PALN positivity may therefore need reconsideration. Resected inv-IPMN exhibited a less aggressive recurrence pattern than PDAC that translated into a more favourable overall survival.

List of scientific papers

I. Outcome after resection for invasive intraductal papillary mucinous neoplasia is similar to conventional pancreatic ductal adenocarcinoma. Holmberg, M.; Ghorbani, P.; Gilg, S.; Del Chiaro, M.; Arnelo, U.; Löhr, J-M., Sparrelid, E. Pancreatology. 2021;21(7):1371-7.
https://doi.org/10.1016/j.pan.2021.07.009

II. Prognostic impact of para-aortic lymph node status in resected pancreatic ductal adenocarcinoma and invasive intraductal papillary mucinous neoplasm – Time to consider a reclassification? Linder S.; Holmberg, M.; Engstrand, J.; Ghorbani, P.; Sparrelid, E. Surg Oncol. 2022; 41:101735.
https://doi.org/10.1016/j.suronc.2022.101735

III. Impact of initial recurrence pattern on overall survival for intraductal papillary mucinous neoplasia – a comparison with conventional pancreatic ductal adenocarcinoma. Holmberg, M.; Kordes, M.; Liljefors M.; Ghorbani, P.; Löhr, J-M.; Sparrelid, E. Pancreatology. 2022;22(5):598-607.
https://doi.org/10.1016/j.pan.2022.04.007

IV. Outcome after resection for invasive intraductal papillary mucinous neoplasia compared to conventional pancreatic ductal adenocarcinoma – a nationwide study from Sweden. Holmberg, M.; Radkiewicz, C.; Strömberg C.; Öman M.; Ghorbani, P.; Löhr, J-M.; Sparrelid, E. [Submitted]