Cystic lesions of the pancreas: an appraisal of an aggressive resectional policy adopted at a single institution during 15 years.

BACKGROUND: Although an aggressive resectional approach toward pancreatic cysts has been advocated in the past, many clinicians now deem this therapeutic strategy impractical given the rapidly increasing incidence of incidentally detected pancreatic cystic lesions. The aim of this study was to review the aggressive resectional policy toward pancreatic cysts adopted at our institution during the past 15 years.
METHODS: One hundred nine consecutive patients who underwent surgical resection of a cystic lesion of the pancreas during a 15-year period were retrospectively reviewed. To determine subsets of patients at lower risk of having a malignant cyst, the clinicopathologic features (in particular, the malignant potential) of these patients were compared as a function of 3 variables, ie, presence of symptoms, patient age, and cyst size, using univariate analyses. Results were expressed as median and range and P < .05 was considered statistically significant.
RESULTS: Forty-three (39%) of 109 patients were asymptomatic. Incidental cysts were smaller (28 [10 to 240] vs 59 [10 to 200] mm, P < .001) and were found in older patients (55.0 [18 to 77] vs 45.5 [14 to 82] years, P = .003). Overall, 14% of asymptomatic cysts, versus 35% of symptomatic cysts, were malignant (P = .016). Incidental cysts were also less likely to be premalignant or malignant compared with symptomatic cysts (47% vs 70%, P = .015). Twenty (18%) patients were elderly (73.0 [70 to 82] years old). Elderly patients had a more equal sex distribution (45% vs 76% female, P = .005) and had smaller cysts (26 [10 to 200] vs 55 [10 to 240] mm, P = .003) that involved the head of the pancreas more frequently (8 [40%] vs 17 [19%], P = .045) compared with their younger counterparts. The cohort of elderly patients also had a higher median American Society of Anesthesiologists score (2 [1 to 3] vs 1 [1 to 3], P < .001), and a higher proportion had undergone a "more" major procedure (Whipple's or total pancreatectomy) (55% vs 18%, P < .001). Not unexpectedly, surgical morbidity in the elderly was significantly higher (10 [50%] vs 24 [27%], P = .045). The operative mortality in both groups was not significantly different (1 [5%] vs 1 [1%], P = .324). The proportion of premalignant or malignant lesions in elderly patients was also similar to that in younger patients (11 [55%] vs 55 [62%], P = .574). The size of a cyst in asymptomatic patients had no correlation with its potential for malignancy.
CONCLUSIONS: Reliance on preoperative characteristics alone such as the presence of symptoms, cyst size, and patient age are not sufficiently reliable in determining the malignant potential and thus management approach toward pancreatic cysts.