Outcomes of Neoadjuvant Chemotherapy for Invasive Intraductal Papillary Mucinous Neoplasm Compared with de Novo Pancreatic Adenocarcinoma

All consecutive patients with a confirmed pathological diagnosis of PDAC or I-IPMN who underwent pancreatic resection between 1 January 2011 and 30 January 2022 in a high-volume center for pancreatic surgery were included in this study. The differentiation between I-IPMN and de novo PDAC is routinely assessed by the institution pathologists by looking for the presence of an IPMN environment around PDAC. Chart review was performed retrospectively and included demographical characteristics, surgical variables, tumor characteristics, imaging characteristics, and follow-up data. Demographical characteristics included age, sex, tumor histology, Eastern Cooperative Oncology Group (ECOG) and American Society of Anesthesiologists (ASA) score, while surgical variables included type of pancreatic resection, concomitant vascular resection, and Clavien–Dindo grade¹⁰ of postoperative complications assessed within 90 days from surgery. Tumor characteristics included tumor size, staging according to the 8th edition of the American Joint Committee on Cancer (AJCC) staging system for pancreatic adenocarcinoma, preoperative CA19.9, number of harvested and positive lymph nodes, resection margins, tumor grading, perineural invasion and lymphovascular invasion, and follow-up data included overall survival (OS) and disease-free survival (DFS). Disease recurrence was defined as ‘early recurrence’ when occurring in the first 6 months after surgery.

The primary outcome of this study was pathological response to neoadjuvant therapy (NAT), and was defined according to the guidelines of the College of American Pathologists.¹¹ Pathological response score was assigned as follows: complete response (score 0), near complete response (score 1), partial response (score 2), and minimal/no response (score 3). Subjects with scores of 0 or 1 were grouped together as ‘major response’, while scores of 2 or 3 were grouped together as ‘partial/no response’. Secondary outcomes were radiological response as defined by the Response Evaluation Criteria in Solid Tumors (RECIST) criteria,¹² CA19.9 response defined as normalization of marker level (≤37 UI/mL) after NAT regardless of initial CA19.9 level, and positron emission tomography (PET) response. Non-secretor patients were excluded from the CA19.9 response analysis. PET response was defined as ‘major response’ when tumor fluorodeoxyglucose (FDG) uptake was below hepatic FDG uptake, and similar to the background pancreatic tissue or ‘minor response’ if tumor FDG uptake was persistent or higher than the background.

Data were collected in a Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) database and then transferred to SPSS statistics. Statistical analysis was performed using SPSS statistics version 28.0 (IBM Corporation, Armonk, NY, USA). Continuous variables were tested for normality using the Shapiro–Wilk test. Normally distributed continuous variables were expressed as the mean ± standard deviation and compared using Student’s t-test, and non-normally distributed variables were expressed as the median with interquartile range and compared using the Mann–Whitney U test. Categorical variables were expressed as number of events (%) and were compared using the Chi-square test. A univariate and multivariate logistic regression analysis was performed to assess correlation to tumor recurrence, and a univariate and multivariate Cox regression analysis was performed to assess correlation to OS. Variables with a p-value < 0.05 at univariate analysis were selected for multivariate analysis. The OS and DFS were compared using Kaplan–Meier curves and log-rank test. A p-value of < 0.05 was considered as the cut-off for statistical significance. When data regarding a variable were not available, the patient was excluded from that specific variable analysis.

The present study involves a large cohort of resected pancreatic cancer patients who were treated in a high-volume tertiary center for pancreatic surgery. I-IPMNs accounted for 9% of surgical resection for PDAC, which is similar to prior reported rates.¹³ The aim of this study was to compare the OS and the response to NAT of I-IPMN and PDAC. In the observed cohorts we observed a significantly longer OS of I-IPMN, however no differences in pathological, CA19.9, PET, and radiological response to NAT between I-IPMN and PDAC was observed.

From the demographical data observed in this study, there are some differences in the average I-IPMN and de novo PDAC populations. I-IPMN patients tend to be older, they undergo a distal and total pancreatectomy slightly more often, they are less likely to need a vascular resection, they receive NAT less often, and they have fewer major postoperative complications. The increased age of I-IPMN patients might be the consequence of a slow process of malignant degeneration.¹⁴ The higher proportion of total pancreatectomies is instead probably to be attributed to the nature of these pancreatic cystic neoplasms, as they can present with an involvement of the whole gland. As I-IPMN patients were less often undergoing a pancreaticoduodenectomy and vascular resection,¹⁵ they also had fewer major postoperative complications, but the 90-day mortality after surgery did not differ. The lower use of NAT in I-IPMN patients can be partially explained by the 37% preoperative uncertainty of malignancy and the lower proportion of tumor with vascular involvement, but even by accounting for these patients it appears that in current clinical practice I-IPMNs with a known malignancy are more likely to receive upfront surgery, even though current medical oncology guidelines do not mention a different approach to I-IPMN and PDAC.^8,9 The role of NAT in I-IPMN has not been thoroughly investigated and currently available literature on the subject is scarce; however, in a large cohort of 240 resected I-IPMN patients, it was reported that NAT was used in only 2.5% of cases.¹⁶ Regarding histological differences between the two cohorts, I-IPMNs had more well-differentiated (G1) tumors, more T1 tumors, and were less likely to invade perineurally. Nonetheless, the nodal involvement was similar and the overall staging distribution of I-IPMN and PDAC did not differ significantly.

The role of adjuvant chemotherapy in I-IPMN has recently been discussed, with several retrospective studies^17-19 suggesting that it might not improve OS in the absence of nodal involvement. Unfortunately, as only a minority of patients with an I-IPMN undergo NAT, the I-IPMN cohort was small and no strong recommendations are possible on the basis of this study. Despite the small statistical power of this study, to our knowledge, this is the first study to assess the response of I-IPMN to NAT and to compare it with de novo PDAC. This makes the study a relevant first step in understanding the topic. In light of our observations, the scarce use of NAT in I-IPMN might not be justified. In the case of preoperative confirmation of malignancy with fine needle aspiration, the use of NAT in I-IPMN could provide the same advantages that it provides in PDAC.^20,21 This is particularly true when radiological suspicion of nodal involvement is present at preoperative disease staging, as adjuvant therapy has been proven beneficial in these settings.^17-19

A secondary result obtained from this study is the suggestion that I-IPMNs are not the same disease as de novo PDAC. Similar to other observations in the literature,²² I-IPMNs have been associated with a more favorable outcome than PDAC, with a higher OS and DFS. Another indicator of a less aggressive biology was the lower rate of early recurrence in the first 6 months after surgery. Moreover, at multivariate analysis, the I-IPMN subtype of pancreatic cancer was an independent predictor of both lower risk of recurrence and death.

The retrospective nature of this study and the small cohort of I-IPMNs undergoing NAT make it hard to draw strong conclusions. Selection and time bias may have influenced treatment decisions, including the use of NAT in I-IPMN. The small size of the I-IPMN cohort may have led our analysis to be underpowered to detect clinically significant differences in some comparisons with PDAC.

Pancreatic adenocarcinoma arising from an IPMN is independently associated with longer OS and DFS after surgical treatment when compared with de novo PDAC. Patients with I-IPMNs are less likely to receive NAT before surgical resection, but when undergoing NAT, they have a similar response compared with patients affected by de novo PDAC. These data show that the potential efficacy of NAT on I-IPMN is similar to that of PDAC and thus may be considered in selected high-risk patients with I-IPMN similar to PDAC indications.