Over the past few decades, cystic neoplasms of the pancreas (CNPs) have been identified with increasing frequency due to the widespread availability of better-quality imaging studies,1 which have shown a prevalence of 2 to 15%.2–4

CNPs are commonly detected incidentally during the approach to a non-specific abdominal symptom. These lesions make up a histologically heterogeneous group, with a wide spectrum of biologic and clinical behavior. They consist of predominantly benign lesions, lesions with malignant potential, and malignant lesions,5–7 with an overall low risk of malignancy (0.5–1.5%) and annual progression (0.5%).8–10

The management of patients with CNPs may be complex and vary considerably among the different subtypes. An accurate classification of pancreatic cystic lesions is crucial and often based on clinical presentation and radiologic characteristics.

In general, premalignant lesions require resection and/or follow-up and malignant lesions require resection, whereas benign or indolent lesions may be observed.

The aim of this study was to describe the clinical and histopathologic characteristics and postoperative results of CNPs resected over the past 22 years at a tertiary care referral center for the management of said lesions.

A retrospective, observational, cross-sectional study was conducted, following the methodology described in the STROBE checklist guidelines.

All patients who underwent CNP resection at our center within the time frame of January 2000 and December 2022 were reviewed. Preoperative, perioperative, and postoperative information was analyzed, including the indication for surgery, pancreatic resection technique employed, surgery duration, intraoperative blood loss, histopathologic report, postoperative complications (reported according to the Clavien-Dindo classification, considering a classification ≥ IIIA a major complication), hospital stay, the development of a biliary or pancreatic fistula, delayed gastric emptying, the need for reintervention, tumor recurrence, and survival in the long-term follow-up. Patient information was obtained from the hospital’s electronic clinical records, as well as from outpatient consultation follow-up notes. Patients who met the inclusion criteria but did not participate were not identified, given that the information came from a retrospective hospital registry, and so all available cases were included. The data collection was completed in September 2024. The variables with missing information were reviewed and registered; data missing from the case records were considered “unavailable” and not assigned a value.

One hundred and thirty-nine patients who met the inclusion criteria were identified and included in the study. Table 1 summarizes their demographic and clinical characteristics. The median follow-up duration was 30 (12−56) months. The main symptom was abdominal pain (76.3%), with a lower percentage of patients presenting with weight loss (27.3%) (8 kg [3–15]). Diagnosis was incidental in 16 patients (11.5%) undergoing an imaging study for another cause. Tumor location was very similar, with 47 (33.8%) cases in the head of the pancreas, 47 (33.8%) in the neck/body of the pancreas, and 45 (32.4%) in the tail of the pancreas.

The approach was open surgery in 134 cases (96.4%), utilizing laparoscopy in only 5 of them (3.6%). There was also a significant increase in the number of surgeries performed by year since 2009, with a maximum peak of 14 cases of resected CNPs at our center in 2016. A later descent in frequency occurred in 2020 (Fig. 1). The most common surgical procedures performed were distal pancreatectomy with splenectomy (48.9%) and pancreatoduodenectomy or the Whipple procedure (30.9%) (Table 2). Concerning the postoperative results, the median surgical procedure duration for all cases was 210 min (165–300) and the estimated intraoperative blood loss was 450 mL (250–800), with 31 patients (22.3%) requiring intraoperative blood transfusion. Four patients (2.9%) required vascular reconstruction during surgery and a postoperative drain was placed in 126 (90.6%). Fifty-five patients (39.6%) presented with a complication. The most frequent was pancreatic fistula (34 patients, 24.5%), and most commonly grade B (19 patients, 13.7%). Median hospital stay was 11 days (7–16) and was longer in the group presenting with a postoperative complication, compared with the patients that did not (15 days [8–24] vs 9 days [6–14], p = 0.001). Nine patients (6.5%) required surgical reintervention and two (1.4%) died in the immediate postoperative period due to vascular complications (Table 3).

Figure 1.

Number of surgeries performed by year for resection of the cystic neoplasms of the pancreas at our center.

In the postoperative result analysis by subgroup, based on surgical technique, the group that underwent pancreatoduodenectomy had longer surgery duration and greater estimated intraoperative blood loss (Fig. 2). They also had greater need of intraoperative blood transfusions, higher frequency of postoperative complications and of delayed gastric emptying, and longer hospital stay. There were no statistically significant differences in the rest of the postoperative results (Table 4).

Figure 2.

Box diagrams that compare the surgical time and estimated intraoperative bleeding in patients who underwent pancreatoduodenectomy and those who underwent other surgical procedures, with a p < 0.001 and p = 0.001, respectively, on the Mann-Whitney U test.

Regarding histopathologic classification after resection, the most common histologic subtypes were solid pseudopapillary neoplasm (SPN) (49 patients, 35.3%), followed by mucinous cystadenoma (mucinous cystic neoplasm [MCN]) (45 patients, 32.4%), and serous cystadenoma (serous cystic neoplasm [SCN]) (23 patients, 16.5%) (Fig. 3). A lesion with malignant potential (SPN, MCN, intraductal papillary mucinous neoplasm [IPMN]) was found in 112 patients (80.5%), with a 92.8% R0 resection rate. Eighteen cases (12.9%) corresponded to IPMNs, two of which arose from the main branch, four from the secondary branch, four were mixed, and the originating branch could not be identified in eight cases. The phenotype was intestinal in three cases, pancreatobiliary in one case, gastric in another, and the phenotype could not be identified in the remaining cases. There was also an increase in the diagnostic frequency of this lesion starting in 2009 (Figs. 4 and 5).

Figure 3.

Frequency of the histologic subtypes found in the cystic neoplasms of the pancreas after their resection.

Figure 4.

Frequency of diagnosis of each of the histologic subtypes in nonmalignant lesions, studied by 5-year periods. There was a rising trend starting in the 2010-2014 period for all lesions, with a later descent in the 2020–2022 period, a result of the COVID-19 pandemic, with no statistically significant differences between each of the time periods, according to the chi-square test (p = 0.343).

Figure 5.

Number of the intraductal papillary mucinous neoplasm (IPMN) diagnoses by year at our center. There was an increase in frequency starting in 2009, maintaining the rising trend.

When the postoperative results based on lesion histologic subtype were compared, there were statistically significant differences between the subtypes in relation to resection surgery duration. The post-hoc analysis showed a significant difference between the SPN and MCN groups, in which the SPN group had a longer median duration (p = 0.035). There was also a greater frequency of surgical site infection in the IPMN group (IPMN 22.2% vs 3.4% for the rest of the lesions, OR 8 [95% CI 1.7–35.6]), as well as of grade B delayed gastric emptying (IPMN 11.1% vs 0% for the rest of the lesions, p = 0.017). A trend toward a greater frequency of postoperative complications was observed in the IPMN and SPN groups, along with a greater frequency of pancreatic fistula in the MCN group, without reaching statistical significance when compared with the other groups (Table 5).

The frequency of malignant lesions was low, with only four cases (2.8%); two (1.4%) were the mucinous cystadenocarcinoma subtype, one (0.7%) was solid pseudopapillary carcinoma, and one (0.7%) was intraductal papillary mucinous carcinoma. Of the malignant lesions, one had tumor extension into the lymph nodes and lymphovascular invasion, with negative margins, corresponding to one of the mucinous cystadenocarcinoma cases (5/15 positive lymph nodes), and one had perineural invasion and positive tumor margins, corresponding to the case of solid pseudopapillary carcinoma. R0 resection was achieved in the rest of the malignant lesions (3 cases, 75%), with a median of 10.5 (2–25) resected lymph nodes. Of those four patients, the female patient with mucinous cystadenocarcinoma received 12 sessions of 5-fluorouracil and leucovorin as adjuvant treatment, with the last application at seven months after the surgery; there was later evidence of peritoneal carcinomatosis and the presence of metastatic lesions in the lung three years after the surgery; tumor progression was documented and the patient was later lost to follow-up. The female patient with intraductal papillary mucinous carcinoma died 4.4 years after resection, due to acute tubular necrosis, in the context of active tuberculosis; no recurrence of complications associated with the surgery were documented during her follow-up. The other two cases were lost to follow-up after the surgery.

The median tumor size of all the CNPs was 55 mm (35.5–95.5), whereas it was 66.5 mm (43.2–92) for the four malignant lesions (p = 0.677). Finally, in the long-term follow-up of the patients with CNPs with malignant potential, two (1.7%) presented with tumor recurrence and disease progression; both were patients with SPN. Table 6 summarizes the oncologic characteristics and composition of the CNPs.

CNPs are a relevant group of tumors due to their wide variety of presentations, ranging from practically benign, to premalignant, to openly malignant tumors, with a heterogeneous prognosis.1

Throughout 22 years of experience in the management of CNPs at a tertiary care referral center, we have seen a high frequency of neoplasms with malignant potential (80.5%). SPNs (35.3%) and MCNs (32.4%) stand out at our center as the most frequent subtypes, followed by SCNs (16.5%) and IPMNs (12.9%). The overall postoperative complication rate was 39.6%, of which 24.5% were classified as major complications (Clavien-Dindo ≥ IIIA). The most frequent complication was pancreatic fistula (24.5%), primarily grade B (13.7%). Perioperative mortality was low (1.4%), similar to that at high-volume centers (0.5−0.9%),1,8,9 and lower than that reported in smaller case series (6-9%).11,12 A total of 1.7% of the patients, specifically those with SPNs, had tumor recurrence, highlighting the need for long-term follow-up.

Regarding the histologic distribution of the CNPs, there were significant regional variations. In high-volume international centers, such as Massachusetts General Hospital8 and Memorial Sloan Kettering,1 in the United States, and the Pancreatic Surgery Unit of the University of Verona,9 in Italy, IPMNs and MCNs predominated, whereas in our case series the SPNs stood out notably, partially coinciding with previous studies in Mexico,11 but contrasting with that reported in other regions13−14 (Table 7). Differing from the above, in another case series conducted in China, SCNs were the most common (35.1%), followed by SPNs (31.5%), whereas MCNs had a lower prevalence,15 similar to that found in other Latin American case series.16–18 Those differences could be attributed to demographic variations, as well as to the relatively low rate of incidental diagnoses at our center (11.5%), compared with international rates (15–25%), possibly conditioned by differences in access to diagnostic imaging studies and their quality.19

A significant increase in the frequency of diagnosed cystic tumors has been seen over time, coinciding with a higher number of surgeries performed on our population. In particular, since 2009, surgical interventions have increased progressively, reaching a peak of 14 cases of CNPs operated on in 2016, followed by a descent in 2020. Said increase was accompanied by greater incidental detection of the neoplasms, an important percentage of the diagnoses, reflecting a growing trend in the identification of CNPs through imaging studies carried out for other causes. The increase in IPMN cases was especially striking, with a considerable increase in the frequency of their diagnosis in 2009. That phenomenon is consistent with the findings by Valsangkar et al.,8 who reported a notable increase in the relative number of IPMNs over the past three decades, going from being diagnosed in only 2% of cases before 1991, to currently being diagnosed in 49% of patients with resected CNPs. The above contrasts with that described in other Mexican case series, in which no cases were reported.11,13,14 Even though some authors question whether this increase reflects a true rise in incidence or is simply a result of greater detection due to advances in imaging techniques and pathologic classifications, the results in our population suggest a similar trend, possibly attributed to better diagnostic capacity and greater clinical awareness about mucinous neoplasms.

Unlike the worldwide trend toward performing minimally invasive procedures, open surgery widely predominates at our center (96.4%), most likely influenced by the anatomic complexity and considerable tumor size seen in our cases. This is comparable with the case series by Chapa et al. in Mexico, in which only 11% of the cases were laparoscopic.11 In contrast, international studies have reported a gradual increase in the use of laparoscopy, which is associated with fewer complications and reduced hospital stay.19,20

The surgical procedures performed more frequently on our population were distal pancreatectomy with splenectomy (48.9%) and pancreatoduodenectomy or the Whipple procedure (30.9%), achieving R0 resection rates (92.8%), differing from the 82% reported in other countries,1 which is similar to rates at other international centers. However, the appearance of major postoperative complications (Clavien-Dindo ≥ IIIA in 24.5%) underlines the need to optimize preoperative strategies and intraoperative management, especially in high-complexity procedures, such as pancreatoduodenectomy, which showed longer surgery duration, greater intraoperative blood loss, and longer hospital stay, similar to that reported in other case series.21–23

Regarding postoperative complications, pancreatic fistula presented in 24.5% of the patients, a trend also reported in other studies,8,11 with a greater frequency in MCNs. Complications were also higher in the group that underwent pancreatoduodenectomy (30.2% vs 21.8%), coinciding with reports that emphasize the morbidity associated with the procedure.24

Lastly, the frequency of malignancy was low, consistent with that reported in other studies, with a percentage below 3% in resected CNPs.8

Our results underline the importance of a multidisciplinary approach and the continuous evaluation of results to improve surgical safety and the quality of care of patients with CNPs.

The present study provides a broad vision of the clinical and surgical profile of CNPs in our population. The high prevalence of SPNs and MCNs highlights the importance of considering population and geographic factors upon interpreting epidemiologic data. Even though the low frequency of malignancy observed in the CNPs is consistent with that reported in the literature, the presence of recurrence in some cases emphasizes the need for continuous and personalized follow-up.

Surgical management showed high complete resection rates (R0), with postoperative complications and pancreatic fistulas comparable to that reported at international centers. It is an area of continuous improvement, through the implementation of more advanced safe resection techniques.

These findings strengthen the relevance of a multidisciplinary focus for optimizing management and improving outcomes in this complex disease.

No specific grants were received from public sector agencies, the business sector, or non-profit organizations in relation to this study.