Pancreas Vol. 22, No. 4, pp. 378–382 © 2001 Lippincott Williams & Wilkins, Inc., Philadelphia Outcome of Surgical Treatment for Chronic Calcifying Pancreatitis Paolo Massucco, Marco Calgaro, Franco Bertolino, Carlo Bima, *Giovanni Galatola, and Lorenzo Capussotti Departments of Surgery and *Gastroenterology, Ospedale Mauriziano “Umberto I,” Turin, Italy Objective: To evaluate the short and long-term results of surgical treatment of calcifying chronic pancreatitis in our center. Patients and methods: We studied 55 consecutive patients operated on for chronic calcifying pancreatitis during a period of 12 years. The mean follow-up period was 6.2 years. Main outcome measures were operative mortality and morbidity, degree of pain control, diabetes onset, survival, and causes of death. Results: The etiology was alcoholic in 48 patients and idiopathic in seven patients. A resection was performed in 78% of cases and a by-pass procedure was performed in 22%. Operative mortality was 3.6%; morbidity was 21.8%. A ductal adenocarcinoma was found in 3.6% of cases. The alcohol withdrawal rate was 78%. Complete pain control was achieved in 71.4% of the patients. Among diabetes, cirrhosis, type of surgery, smoking and alcohol abuse history, only alcohol with- drawal was associated with pain control (p < 0.03). A late reintervention was needed in only one patient in the by-pass group. Five and 10-year survival rates for the entire population were 80% and 61%, respectively. Among alcohol, cirrhosis, diabetes, and type of surgery, only the former was associated with survival (p < 0.003). Five-year actuarial survival was 55.6% for patients who continued drinking compared with 86.3% for ex-alcoholics. Conclusions: Surgical resection should be performed when required by the anatomical conditions because it was associated with good long-term pain control and low postoperative and late morbidity. Alcohol withdrawal has a key role for effective control of pain and prolonged survival. Key Words: Chronic pancreatitis—Surgical therapy. Chronic pancreatitis may require surgical treatment when pain becomes intractable or when local complications occur. The ideal operation should provide the best long-term results with minimal functional impairment; however, the optimal surgical technique has never been established. Two main types of operation have been developed based on different theories of pain pathogenesis. The drainage approach is based on the hypothesis of ductal and/or parenchyma hypertension; the resective approach is based on the hypothesis of local inflammation and perineuritis. The choice should be tailored to the pattern of disease in each patient, but clear guidelines are lacking. Moreover, the impact of surgical intervention on the natural history of pancreatitis is not yet well understood. We analyzed the experience of surgical treatment of chronic calcifying pancreatitis in a single institution during a period of 12 years, with particular focus on the surgical strategy adopted and the long-term outcome. PATIENTS AND METHODS Between 1985 and 1997, 55 patients (54 men and one woman) were operated on for chronic calcifying pancreatitis in our center. The mean age at surgery was 46.2 years (range, 27–81 years). The diagnostic work-up included ultrasonography and endoscopic retrograde pancreatography in most of the patients, and a contrast-enhanced CT scan in all of them. All the cases of chronic pancreatitis were confirmed by histologic examination. Follow-up data were collected by periodic visits (every 6 months in the first 2 years, then yearly). At the end of the follow-up period, all patients underwent telephone Received March 9, 2000; revised manuscript accepted September 12, 2000. Address correspondence and reprint requests to Dr. P. Massucco, I Department of Surgery, Ospedale Mauriziano “Umberto I,” Largo Turati, 62, 10128 Torino, Italy. E-mail: pmassucco@mauriziano.it 378 OUTCOME OF SURGICAL TREATMENT FOR CHRONIC CALCIFYING PANCREATITIS interviews, following a standardized questionnaire. Patients lost to follow-up were censored at time of last contact. Patients with pancreatic cancer were excluded from survival analysis. Outcome measures were operative mortality and morbidity, degree of pain control, diabetes onset, need for late reinterventions, late mortality, and causes of death. Complete pain control was defined as “no need of analgesic drugs,” partial pain control was defined as “need of minor analgesics,” and no pain control was defined as “need of narcotics or hospital admissions.” Diabetes was diagnosed following World Health Organization criteria. We defined as “late postoperative onset” all cases of diabetes diagnosed more than 6 months after surgery. Statistical analysis was performed using the K-square or Fisher exact test, as appropriate. Survival curves were calculated by the Kaplan-Meier method and compared by the log rank test. RESULTS The etiology of the disease was alcoholic in 48 patients and idiopathic in seven patients (nonfamilial, nonobstructive). Mean alcohol consumption was 210 g/day (range, 130–380). The rate of smokers was 94%, with 85% of them being heavy smokers (one or more packs/day). Cirrhosis and diabetes were found in 18% and 15.4% of the population, respectively. Indications to surgery were intractable pain in 33 patients (60%), suspicion of malignancy in nine (16%), gastrointestinal hemorrhage in five (9%), enlarging pseudocysts in two, bile duct stenosis in two, and pancreatic ascites in one. In the last three cases, we operated on patients who had undergone a previous distal pancreatectomy. The causes of reintervention were one case of pain relapse, one biliary stenosis, and one pseudoaneurysm of the splenic artery stump. Among the cases listed above, gastrointestinal hemorrhage was caused by bleeding in a pseudocyst in three cases, by Wirsung hemorrhage in one, and by a failed attempt to achieve radiologic transgastric pseudocyst drainage in one. The latter case and one of the three cases of bleeding pseudocysts required emergency surgical treatment. Morphologic findings are summarized in Table 1. Pseudocysts (31 cases) were single in 90% of cases and symptomatic in 94%. The site was the pancreatic head in 17 cases, the isthmus in one, the corpus in four, and the tail in nine. Pseudocysts of the head were associated with bile duct and/or duodenal stenosis in 53% of cases; those in the tail were associated with splenic vein thrombosis in 56% of cases. MPD dilatation (10 cases) 379 TABLE 1. Morphologic alterations Morphologic alterations Calcifications Pseudocysts Bile duct stenosis MPD dilatation (>7 mm) Head mass Splenic vein thrombosis Duodenal stenosis MPD disruption 64% 56% 35% 18% 16% 13% 9% 4% MPD ⳱ main pancreatic duct. was isolated in one case, and was associated to bile duct stenosis in two cases and to pseudocyst in seven cases. Table 2 shows the surgical techniques adopted. The 60-day operative mortality was 3.6%. The causes of death were one case of pulmonary embolism in a cirrhotic patient who underwent pseudocyst drainage, and one case of multiorgan failure in a man operated on in an emergency for massive bleeding from a pseudocyst eroding the portal vein. There was no mortality among the 27 patients who underwent duodenopancreatectomy. Morbidity was 21.8%. Reintervention was required in four cases: one duodenal necrosis after subtotal pancreatectomy, one postoperative occlusion due to adhesions, and two cases of hemoperitoneum. Other causes of morbidity included two pancreatic fistulas, two cases of delayed gastric emptying after pylorus preserving pancreatoduodenectomy, two cases of postoperative pneumonia, one cardiac arrhythmia, and one case of fever of unknown origen. Ductal adenocarcinoma was present in two cases (3.6% of population). Complete follow-up data were available for 49 patients. The mean follow-up period was 74.4 months. Alcohol withdrawal rate was 78%. Only 27% of drinkers stopped smoking. Among the alcoholics who stopped drinking, 68% continued to smoke. None of the alcoholTABLE 2. Interventions Interventions Resection duodenopancreatectomy distal pancreatectomy subtotal pancreatectomy total pancreatectomy isthmus resection Drainage pseudocysts MPD bile duct 78% 27 (10 PPPD) 11 2 2 (1 PPPD) 1 22% 6 3 3 +7 MPD drainage + 2 bile duct + 1 gastric PPPD ⳱ pylorus preserving pancreatoduodenectomy. Pancreas, Vol. 22, No. 4, 2001 380 P. MASSUCCO ET AL. ics ceased smoking without simultaneously ceasing drinking. We achieved complete pain control in 71.4% of patients, partial pain control in 21.4%, and no pain control in 7.2%. All the patients presently alive declared that their quality of life was improved by the operation, except one patient who acquired hepatitis C virus-related chronic hepatitis from a blood transfusion. In a univariate analysis, including diabetes, cirrhosis, the type of surgery, cessation of smoking and cessation of drinking, only alcohol withdrawal was associated with pain control (Table 3). The prevalence of diabetes was 59.2%; 15.4% of the patients were diabetics before surgery. The incidence of postoperative diabetes was 44.8%, with 50% of cases having late-onset diabetes. The incidence of early postoperative diabetes was 27.3% after distal pancreatectomy and 13.6% after duodenopancreatectomy (p ⳱ 0.31). A late reintervention was needed in one patient who, after undergoing biliary bypass, required a second operation for a duodenal stricture. During the follow-up period, 14 patients died. Five and 10-year actuarial survival rates for the entire population were 80% and 61%, respectively. Five-year actual survival rate was 73.2%. A univariate analysis was performed, including alcohol, cirrhosis, diabetes, and the type of surgery. The effect of smoking cessation on survival could not be evaluated because of the small number of ex-smokers. From the analysis, only alcohol was associated with survival rate. Analyzing the alcoholic pancreatitis group, patients who continued drinking had 5-year and 10-year actuarial survival rates of 55.6% and 18.5% vs 86.3% and 76.6% for patients who stopped alcohol intake (p < 0.003). Actual 5-year survival was 50% and 84.6%, respectively. The survival curves are shown in Figure 1. At follow up 3 years after completion of the study, the causes of death were end-stage cirrhosis in seven cases, cardiovascular accidents in three, duodenal ulcer perforation in one, leukemia in one, and unknown in two. In four patients who were still alive at the time of this writing, cancers possibly associated with cigarette smoking or alcohol use developed: one of the mouth, one of the lung, one of the larynx, and one of the bladder. TABLE 3. Univariate analysis of factors influencing pain control Factors Pain No pain p value diabetes vs no diabetes cirrhosis vs no cirrhosis resection vs bypass alcoholics vs ex-alcoholics smokers vs ex-smokers 9/4 3/8 11/12 5/6 10/2 20/16 4/34 28/8 4/27 23/10 0.39 0.14 0.29 0.032 0.21 Pancreas, Vol. 22, No. 4, 2001 FIG. 1. Kaplan-Meier survival curves of alcoholics versus exalcoholic patients. DISCUSSION The main indications for the surgical treatment of chronic pancreatitis are pain, when judged incapacitating by the patient; local complications; and the suspicion of malignancy. However, even if a general agreement exists about indications, the best surgical technique is not yet established. Disagreement is a result of the uncertainty on pain pathogenesis. Based on different theories on pain pathogenesis, two modalities of treatment have been proposed: drainage and resection. Drainage is based on the hypothesis that pain is caused by parenchymal and/or ductal hypertension (1) and on the theory of “compartment syndrome” (2). Drainage is generally adopted in cases of large duct pancreatitis (MPD > 5–7 mm) or when a pattern of “chain of lakes” is present at pancreatography. The advantage of such an approach consists in sparing parenchyma, resulting in less functional impairment and decreasing the incidence of postoperative diabetes. The disadvantages are a higher rate of late reintervention for obstructive complications or recurrent pain, and the difficulty in ruling out cancer. The resective approach is based on the theory of peripancreatic neuritis (3). Lower morbidity and mortality were reported for distal resections, but results on pain control are poorer (4). According to Warshaw, the role of this technique should be limited to selected cases (5). The rationale of duodenopancreatectomy is the Longmire theory that the pancreatic head is a “pacemaker” of the disease. The advantages include the surgical radicality in the case of suspected cancer, a better long-lasting pain control, and a lower rate of late reinterventions. The disadvantages include a higher mortality and morbidity and, possibly, a higher late-mortality rate (6). The duodenum-preserving technique proposed by Beger (7) should provide better glycemic control and nutritional results. OUTCOME OF SURGICAL TREATMENT FOR CHRONIC CALCIFYING PANCREATITIS More recently, mixed techniques (resection/drainage) have been proposed by Frey (8) and Izbicki (9), but their role is not yet well established. In our experience, despite our desire to preserve pancreatic function, pure drainage techniques have been rarely used because of the local condition of disease. A pure drainage has been performed in the following cases: dilated MPD not associated with pseudocysts or multiple obstructive complications (biliary and duodenal); and isolated symptomatic or enlarging pseudocyst. These features were rarely found in our patients. Moreover, bypass surgery is reserved for patients with impaired liver function. In all other cases, we prefer resection. In our opinion, when a pancreatic head enlargement is present, duodenopancreatectomy is mandatory for two reasons: to treat pain and because of the risk of missing a pancreatic cancer. Furthermore, this approach is preferable for treating pain in the absence of MPD dilatation because, in agreement with the Longmire hypothesis, we think that, in this situation, a better pain control may be obtained by removing the head. Finally, we prefer to perform resection in cases of multiple obstructive complications that would otherwise require the association of three or more bypasses, or when a pseudocyst is present in association with obstructive complications. Distal resections have been performed in selected cases: such as splenic vein thrombosis and pseudocysts in the pancreatic tail with MPD dilatation. Subtotal distal pancreatectomy, performed in the first part of our experience, has been definitively abandoned. Total pancreatoduodenectomy was adopted for patients without MPD dilatation and who were already diabetics before surgery. Following these indications, we performed a resection in 65% of cases, a pure drainage in 22%, and a resection associated to MPD drainage in 13%. By this mainly resective approach, we have obtained a good and longlasting control of pain, with low morbidity and no need for late reintervention, except for one case in the bypass group. These results are similar to those reported in other surgical series in which a high percentage of resections were performed (10,11). In contrast, in the series in which a prevalent drainage approach is used, a rate of late reintervention up to 60% has been reported (12). We observed a high prevalence of diabetes (59.2%). Other authors have reported rates of postoperative diabetes higher than 50% in recent large series of pancreatic resection for chronic pancreatitis (11,13,14). In his outcome analysis of duodenopancreatectomy, Traverso reported an incidence of postoperative diabetes of 32% (15). By examining the event curve, he concluded that postoperative diabetes is related more to the natural history of the disease than to resection itself. In our expe- 381 rience, early postoperative diabetes accounts for 22.4% of all diabetic patients, while 15.4% of the patients were diabetics at presentation, and another 22.4% developed a late-onset postoperative diabetes, more probably related to a “burning out” process. Concerning the different techniques, distal pancreatectomy seems to be associated with a higher incidence of early postoperative diabetes than duodenopancreatectomy, as reported by other authors (16), but the number of cases is too small for a correct statistical analysis. Global survival of our patients is lower than an agematched general population. In the multicentric study of the Pancreas Study Group, the 10-year survival of a mixed surgical-medical population of chronic pancreatitis was 70% (17). Significant factors on survival were alcohol use, smoking, and cirrhosis, but not previous surgery. Also, in our patients, causes of death were related mainly to the persistence of alcohol use and smoking, or the presence of cirrhosis. In contrast with previous reports (18), we did not observed a higher late mortality after pancreatoduodenectomy by comparison with distal resections or bypass procedures. In conclusion, resective therapy and, particularly, duodenopancreatectomy should be avoided whenever possible. They can be applied without concern when required by the local condition of the disease because this approach gives a good and long-lasting control of pain, with a low postoperative and late morbidity and does not affect the long-term survival. Moreover, the high postoperative rate of diabetes is partially related to the natural history of the disease. Alcohol withdrawal is the key point for achieving effective results on pain control and survival rate. REFERENCES 1. Ebbehoj N, Svendsen LB, Madsen P. Pancreatic tissue pressure chronic obstructive pancreatitis. Scand J Gastroenterol 1984;4: 1066–8. 2. Karanjia ND, Widdison AL, Leung F. Compartment syndrome in experimental obstructive chronic pancreatitis: effects of decompressing the main pancreatic duct. Br J Surg 1994;81:259–64. 3. Bockmann DE, Buechler M, Malfertheimer P, et al. Analysis of nerves in chronic pancreatitis. Gastroenterology 1988;94:1459–63. 4. Rattner DW, Fernandez-del-Castillo C, Warshaw AL. Pitfalls of distal pancreatectomy for relief of pain in chronic pancreatitis. Am J Surg 1996;171:142–6. 5. Sawyer R, Frey CF. Is there still a role for distal pancreatectomy in surgery for chronic pancreatitis? Am J Surg 1994;168:6–8. 6. Fékété F, Msika S, Gayet B, et al. Evolution of indications and results of the treatment of alcoholic chronic pancreatitis in men: study of 222 cases. Ann Chir 1991;45:209–17. 7. Beger HG, Buchler M. Duodenum-preserving resection of the head Pancreas, Vol. 22, No. 4, 2001 382 8. 9. 10. 11. 12. P. MASSUCCO ET AL. of the pancreas in chronic pancreatitis with inflammatory mass in the head. World J Surg 1990;14:83–7. Frey CF, Amikura K. Local resection of the head of the pancreas combined with longitudinal pancreaticojejunostomy in the management of patients with chronic pancreatitis. Ann Surg 1994;220: 492–507. Izbicki JR, Bloechle C, Broering DC, et al. Longitudinal V-shaped excision of the ventral pancreas for small duct disease in severe chronic pancreatitis. Ann Surg 1998;227:213–19. Evans JD, Wilson PG, Carver C, et al. Outcome of surgery for chronic pancreatitis. Br J Surg 1997;84:624–9. Buhler L, Schmidlin F, de Perrot M, et al. Long-term results after surgical management of chronic pancreatitis. Hepatogastroenterology 1999;46:1986–9. Hakaim AG, Brougham TA, Vogt DP, et al. Long-term results of the surgical management of chronic pancreatitis. Am Surg 1994; 60:306–8. Pancreas, Vol. 22, No. 4, 2001 13. Ho HS, Frey CF. Current approach to the surgical management of chronic pancreatitis. Gastroenterologist 1997;5:128–36. 14. Rumstadt B, Forssmann K, Singer MV, et al. The Whipple partial duodenopancreatectomy for the treatment of chronic pancreatitis. Hepatogastroenterology 1997;44:1554–9. 15. Traverso LW, Kozarek RA. Pancreatoduodenectomy for chronic pancreatitis: anatomic selection criteria and subsequent long-term outcome analysis. Ann Surg 1997;226:429–38. 16. Jalleh RP, Williamson RCN. Pancreatic exocrine and endocrine function after operations for chronic pancreatitis. Ann Surg 1992; 216:656–63. 17. Lowenfels AB, Maisonneuve P, Cavallini G, et al. Prognosis of chronic pancreatitis: an internetional multicenter study. Am J Gastroenterol 1994;89:1467–71. 18. Sastre B, Carabalona B, Crespy B, et al. Immediate and late results of surgical treatment for chronic pancreatitis: apropos of 127 operated cases. Ann Chir 1990;44:333–7.