ALCOHOLIC VS. NON-ALCOHOLIC CHRONIC PANCREATITIS: SURGEONS’ PERSPECTIVE FROM A TERTIARY CENTRE IN INDIA

ABSTRACT 
Background: Although alcohol is the most common cause for chronic pancreatitis worldwide, idiopathic type is prevalent in India. Natural history and disease progression are different between these two groups. There is paucity of data comparing surgical outcome and quality of life in these patients. 
Aim: To evaluate clinical features, surgical outcome and quality of life between these two groups of patients. 
Method: All patients with chronic pancreatitis who underwent surgery were prospectively reviewed. 
Results: From 98 patients, 42 were alcoholic. Number of male and the mean age at the time of operation was significantly more in alcoholic patients. Smoking, preoperative hospital admission rate and the prevalence of local complications like inflammatory pancreatic head mass, biliary stricture and left sided portal hypertension were distinctly more common in alcoholic group. Frey procedure was required more commonly in alcoholic group. Mean postoperative hospital stay and overall postoperative complication rate were comparable between the two groups. Over a median follow up of 18 months there was significant improvement in quality of life and pain score in both the groups. Improvement of physical functioning score at follow-up was significantly more in alcoholic group but the requirement for analgesic medications were significantly more in alcoholic group. However, appetite loss was more perceived by non-alcoholic group. 
Conclusion: Alcoholic chronic pancreatitis presents with more local complications associated with chronic pancreatitis. Frey procedure is a safe and well accepted surgery in this group. Though they required more analgesic requirement in short term follow up, other aspects of quality of life are similar to non-alcoholic group.

surgeon. Only pancreatic ductal drainage (longitudinal pancreaticojejunostomy -LPJ) was performed in presence of dilated duct caused by stricture or stone, without an inflammatory head mass. A jejunal Roux loop was anastomosed side to side to whole length of pancreas from tail to head in single layer of continuous suture (3-0 or 4-0 polypropylene). We kept the length of the Roux limb around 60 cm as it could be used later for biliary bypass if the patient develops biliary stricture in future. Drainage procedure combined with a local resection as described by Frey and Smith 11 was performed in presence of an inflammatory large head mass (>4 cm or more). However, simultaneous bilioenteric bypass was performed with the same loop in presence of obstructive jaundice or common bile duct stent. Izbicki procedure was performed in small duct disease (MPD diameter ≤3 mm) as described by Izbicki and colleagues 18 . Reconstruction was similar to LPJ. Pancreaticoduodenectomy was performed only in case of severe duodenal obstruction or malignancy could not be ruled out preoperatively. Distal pancreatectomy with or without splenectomy was performed in case of extensive disease at distal body or tail region. However, patients who were found to have malignancy after histopathological examination were excluded from the study.

Data collection and follow up
Clinicopathological and operative data were collected from our prospectively maintained database. Postoperative complications were assessed according to Clavien-Dindo classification 8 and postoperative mortality was defined as death within 90 days after the operation. Postoperative pancreatic fistulas, delayed gastric emptying and postoperative haemorrhage were diagnosed and classified based on International Study Group of Pancreatic Surgery criteria 6,26,27 . The patients were followed in the outpatient department. Follow up protocol in our department is every three months for the first two years, then six monthly for next three years, then annually for rest of the life. In the follow up, following parameters were recorded: body weight, pain, analgesic requirement, need for enzyme supplementation, need for hospitalization and blood sugar (both fasting and postprandial). Diabetes mellitus or endocrine insufficiency was defined if fasting blood sugar was more than 126 mg/dl and serum glycosylated hemoglobin was more than 6.5%. Pancreatic exocrine insufficiency was considered in presence of steatorrhoea and or the need to take pancreatic enzyme supplementation. Fecal elastase estimation was not available in our institution to assess pancreatic exocrine function.

Quality of life assessment and pain score
The quality of life was assessed by the Bengali and Hindi version of European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ C-30) 10 . It consists of five functional scales, three symptom scales, a global health status/QoL scale, and six single items. Different set of questions are included in the multi-item scale. All scales and single-item measures range in score from 0 to 100. A high score in any functional scale denotes a high/healthy level of functioning, but a high score for a symptom scale/ item represents a high level of symptomatology/problems. Functional scales include physical functioning, role functioning, emotional functioning, cognitive functioning, and social functioning. Symptoms scales/items include fatigue, nausea and vomiting, pain, dyspnoea (DY), insomnia, appetite loss, constipation, diarrhoea and financial difficulties. Experience of pain from this debilitating disease was evaluated by the validated Izbicki pain score 9 having the following components: frequency of pain attacks, a visual analog scale of pain, analgesic medication used, and inability to work. All participants were requested to fill up the EORTC QLQ C30 questionnaire and INTRODUCTION C hronic pancreatitis (CP) is a progressive inflammatory disease leading to irreversible damage of the pancreas. The most common etiology in West is chronic alcohol intake whereas idiopathic pancreatitis (or tropical pancreatitis) is reported to be most common etiology in India and China, accounting for approximately 70% of all patients 13 . Recent few papers have suggested alcoholism as a dominant cause of it in India 15,22 . Though genetic study and clinical features have suggested that tropical pancreatitis is different from idiopathic chronic pancreatitis 12 , one thing is common in both idiopathic and tropical pancreatitis: absence of alcohol intake. Natural history of alcoholic pancreatitis is different from non-alcoholic 2 . One recent study has shown the different genetic pattern for alcoholic chronic pancreatitis than other chronic pancreatitis 28 . In this part of India where genetic analysis is not prevalent widely, etiology of CP can be broadly classified as alcoholic and nonalcoholic.
Pain is the most distressing symptom and most of the patients seek medical advice for disabling pain. Since chronic pancreatitis is a benign non-curable disease with a protracted natural course, pain relief as the successful outcome of the disease reflects only one aspect of the multidimensional outcome. As result, health-related quality of life as subjectively perceived by the patient, is becoming a major issue in the evaluation of any therapeutic intervention, mainly in patients with chronic diseases. In the current step-up approach, surgery is regarded as a last resort when other modalities fail. Several studies 3,4,17,18,24,25 have shown that surgery for CP leads to pain relief and substantial improvement of health related quality of life in long-term follow up. Most these studies evaluate the nature of surgery and their long-term effect on the course of the disease. However, the quality of life in any alcoholics is quite different from healthy individuals or even from non-alcoholic patients. Several psychological and social factors interplay to determine the quality of life in alcohol dependents 23 . As earlier stated the natural history of alcoholic chronic pancreatitis is different from non-alcoholic 2 , treatment modalities and outcome may differ in long-term follow up.
The aim of this study was to evaluate clinical features, surgical outcome and quality of life after surgery in alcoholic and nonalcoholic chronic pancreatitis.

METHODS
This prospective observational analysis was performed in the Department of Surgical Gastroenterology, School of Digestive and Liver Diseases, Seth Sukhlal Karnani Memorial Hospital and the Institute of Post-Graduate Medical Education&Research, Kolkata, India. Between October 2016 and October 2019 all patients who underwent surgery for chronic pancreatitis were included. Patients diagnosed with malignant disease, h/o acute illness within four weeks and inability to understand the questionnaire were excluded from the study. All patients had a detailed history and clinical examination. The diagnosis of chronic pancreatitis was based on typical history of abdominal pain and pancreatic parenchymal/ductal calcification and/or dilatation of the main pancreatic duct on imaging. Alcoholic CP was diagnosed in presence of daily alcohol intake of >80 g/day alcohol for at least five years 1 .Other investigations included complete hemogram, liver function test, and renal function test. Upper gastrointestinal endoscopy was suggested in suspicion of portal hypertension.

Operative procedures
Decision and type of surgery were finalised after multidisciplinary meeting between gastroenterologist, radiologist and primary OriginAl Article 2/7 group (p=0.00). Mean age at the time of surgery was 34.15 years and it was 39.62 and 30.05 years in alcoholic and non-alcoholic group respectively(p=0.00). There was more smoker in alcoholic group (18 vs. 6, p=0.00). Alcoholic CP also experienced frequent preoperative hospital admission due to abdominal pain (18.14 vs. 9.20, p=0.05). Pancreatic head mass and biliary stricture were distinctly predominant in alcoholic group (18 vs. 4, p=0.00 and 10 vs. 4, p=0.03, respectively). Others demographic and preoperative clinical data in alcoholic and non-alcoholic group are presented in Table 1.

Operative data
Operative procedures and data are summarized in Table 2.
The most frequent surgical procedure was Frey procedure (46.9%), closely followed by lateral pancreaticojejunostomy (43.9%). However, Frey procedure was more common in alcoholic group (28 vs. 18) whereas LPJ was distinctly more common in non-alcoholic (35 vs. 8). Most common additional procedure was cholecystectomy. Mean operative time was 248.54 min (+/-43.08) and mean blood loss was 161.73 ml (+/-121.79). There was no difference in operative time and intraoperative blood loss in alcoholic and non-alcoholic Izbicki pain score in the hospital before surgery and during each follow up visit.

Statistical analysis
All the categorical variables were expressed as percentage and the continuous variables were expressed as mean+/-SD as well as median and range value. Comparisons of various domains of QOL and pain scores before and after surgery were performed using 'paired t test'. Comparison of various domains of QOL and pain scores between alcoholic and non-alcoholic were performed by using 'individual sample t test'. p≤0.05 was regarded as significant. All calculations were performed with IBM SPSS software, version 26.0.

RESULTS
One hundred and six patients with CP underwent surgery during the study period. Eight were excluded as five were below 15 years, two had malignancy in final biopsy and one acute illness within four weeks. Thus, 98 patients were included in the present study. Of these, alcohol was aetiology in 42 cases and overall 63 were male (64.3%). In the alcoholic group male:female was 41:1 and it was 22:34 in non-alcoholic groups (p=0.16 and 0.94 respectively). Left sided portal hypertension was found in 21 patients (21.6%) and it was significantly higher in alcoholic group (14 vs. 7, p=0.02).

IVb 0
No multiorgan dysfunction in the study group

V 0
No postoperative death in the study group Most common postoperative complication was pancreatic leak (n=18). Fifteen patients had biochemical leak that was managed conservatively. Three had grade B postoperative pancreatic fistula. One had persistent drainage beyond three weeks and required repositioning of the drain. Another two developed abdominal collection for which ultrasound guided percutaneous drain was placed. Postoperative pancreatic haemorrhage occurred in eight patients. Among these, Frey procedure was performed in seven and Izbicki procedure in one. Three patients were managed conservatively and five needed re-exploration. The source of bleeding was cored out head cavity in all five. One patient of postoperative haemorrhage developed nonarteritic ischaemic optic neuritis due to severe hypotension. Hospital stay and specific complications were compared, and no difference was found between alcoholic and nonalcoholic groups (Table 4).

Follow-up data
Median follow up was 18 months (6-30). During this period, new onset diabetes mellitus was documented in three patients (two of them underwent Frey procedure, other LPJ). At the end of follow up, 33 patients (33.67%) were diabetic. Of them 23 were on insulin and 10 on oral hypoglycaemic agents. No patient developed new onset exocrine insufficiency and there was no improvement of preoperatively diagnosed exocrine insufficiency. Benign biliary stricture developed in two patients who underwent Frey procedure previously. Choledochojejunostomy was performed in both of them using the same loop that was used for pancreaticojejunostomy. One patient developed incisional hernia and prolene mesh repair was performed. Body weight was not found to be significantly improved after surgery in both the groups (Table 5).

Quality of life and pain score assessment
Overall quality of life and Izbicki pain score were significantly improved after surgery (Table 6) though 10 patients (10.2%) had incomplete pain relief at their last follow up visit. Among them six were in alcoholic group and four in non-alcoholic. In 95 patients (96.9%) there was improvement in all parameters of quality of life whereas in three improvement in global health score only. Other parameters of quality of life were not significantly improved in them. When we compared preoperative QOL and pain score between alcoholic and nonalcoholic group, we found financial difficulty to be significantly higher in alcoholic group (p=0.020). During follow up period physical functioning was significantly improved in alcoholic group (p=0.05) but appetite loss was more perceived in nonalcoholic group (p=0.03). Requirement of analgesic medications was more in alcoholic group (p=0.05, Table 7).

DISCUSSION
The term 'tropical pancreatitis' was coined by Zuidema et al 30 to describe idiopathic chronic pancreatitis in developing countries. Recent review study by Garg PK 12 had shown that tropical pancreatitis and idiopathic pancreatitis have distinct clinical course, different genetic pattern and different prognosis. However, idiopathic chronic pancreatitis (or tropical pancreatitis) is considered to be most common variety of chronic pancreatitis in India 13 . In our study we found alcoholic pancreatitis in 42.8% cases and non-alcoholic in 57.2%. Though few cases of pancreas divisum were identified, in absence of genetic analysis we were unable to differentiate between idiopathic chronic pancreatitis and tropical chronic pancreatitis. As a result, non-alcoholic etiology remains the most common etiology in our study group. A research from South India 22 indicates alcohol intake as a dominant emerging cause of CP and Halder SK et al 15 had also mentioned alcohol as most common etiology in their study population.
Alcohol intake is not common for woman in Indian culture and as a result we found distinctly male predominant in alcoholic chronic pancreatitis. We also found significantly higher mean age in alcoholic group than non-alcoholic. This reflects several years of chronic alcoholism leads to CP and also idiopathic chronic pancreatitis and tropical chronic pancreatitis both occur in 2 nd and 3 rd decade 12 . Smoking is itself a risk factor for CP, and also accelerates the progression of CP in alcoholics 19,29 . A nationwide study from India by Balakrishnan et al 5 had identified overall 28.3% smoker in CP patients and 59% of male alcoholics were smokers. In our data smoking was present in 42.8 % (18/42) in alcoholic and only 10.7% (6/56) in non-alcoholic group. This may be explained by male predominance in alcoholic CP group and, also, heavy drinkers are often heavy smokers 5,7 .
Incidence of diabetes (overall 30.6%) was similar to other reported Indian series 5,20 but the incidence of exocrine insufficiency (overall 6.1%) was lower compared to western series 14,21 . The possible explanation may be less dietary fat intake by Indian population. During follow up period only three patients developed new onset diabetes and none new onset exocrine insufficiency. The low incidence of new diabetes and new exocrine insufficiency in our study may be due to very short duration of follow-up. Exocrine and endocrine function may deteriorate over time and needs long-term follow-up. However, we were unable to document improvement of body weight. This result can be attributed to very short-term follow up period.
Head of pancreas mass was preoperatively identified in 18 cases (42.8%) in alcoholic group. Bordacahar et al 7 had found significantly more local inflammation in posterior plane of the pancreas (53% in alcoholic CP and 25% in controls, p=0.05). Biliary stricture was present in 14.3% in our study. Similar result was obtained by in a recent study by Hao et al 16 where they found common bile duct stricture in 15.8% cases. They had included factors like male, age at onset, smoking, body mass index and morphology of main pancreatic duct to develop a nomogram to predict common bile duct stricture in chronic pancreatitis patients. The clinical profile of alcoholic chronic pancreatitis is also being reflected by nature of surgery performed. Frey procedure was more frequently performed in alcoholic group whereas LPJ was most commonly in non-alcoholic. Interestingly we performed total 46 Frey procedures whereas preoperatively we identified pancreatic head mass only in 22 cases. It may be explained by poor or inadequate preoperative computed tomography scan reporting and/or the mass was less than <4 cm (not being enrolled as pancreatic head mass). This difference was reflected in amount of tissue removed in Frey procedure. Though it is not significantly different in two groups but in alcoholic group it was on higher side (10.12 vs. 8.11 g). In study by Gestic et al 14 median tissue removed in Frey was 14.8 g (7-78). In their work etiology of CP was alcohol abuse in 95.9% cases. Several papers 3,4,25 had compared Frey with Whipple or Beger procedures and was unable to find any significant difference in terms of pain relief, mortality and long-term morbidity. However, Frey procedure is technically easier and our preferred method. Again, in presence of extrahepatic/left sided portal hypertension the dissection of tunnel of Love beneath the neck of the pancreas, as required in Beger or Whipple procedures is likely to be tedious and risky 21 . In our study we found left sided portal hypertension in 21.4% cases and it was significantly higher in alcoholic group (14 vs. 7, p=0.02). Similar result was also obtained by Bordacahar et al 7 .
They found significantly more segmental portal hypertension in alcoholic CP in comparison to non-alcoholic CP (69% vs. 31%, p=0.0069).
There was no perioperative mortality in our study and overall morbidity was present in 39.79% patients which is comparable with other studies 14,20,21 . Grade A pancreatic fistula or biochemical leak was most common complication in our study. Grade B occurred only in three. Hard texture of pancreatic parenchyma and decreased capacity of secretory function may explain this very low number of clinically relevant postoperative pancreatic fistula.
In our study the EORTC QLQ C-30 questionnaire and Izbicki pain score have been used to assess the quality of life and pain control. Both were well accepted by our patients and required around 22 min (15-40) to complete. There was significant improvement in all domains of QOL after surgery (p<0.005) which was similar to that reported by various authors 4,17,18,24,25 . In our study, we found that financial difficulty was significantly higher in alcoholic patients before surgery which might be attributed to alcohol consumption itself. Again, it may reflect the different socioeconomic status in alcoholic patient 9 .
In our study we found 90% to have complete pain relief. Similar result have been reflected by various studies with very long-term follow up 14,20,21 . Factors like preoperative opiate use, continuous pattern of pain, long duration of pain(>6y), postoperative complications and age ≥38 years were suggested for incomplete pain relief after surgery 20 . Bordacahar et al 7 had also suggested early surgery and preoperative complete smoking as favourable factor for improved quality of life after surgery in alcoholic CP.
During follow up physical functioning score was significantly improved after surgery in alcoholic group. Physical functioning score involved questions associated with physical work (e.g. any trouble during strenuous exercise or long walk, need to stay in bed during day, etc.). These questions may be not very appropriate for younger patients who are not very accustomed to this type of physical activity. This might be a reason for finding a better physical functioning score in alcoholic group where most of the patients are male and relatively higher age group. Interestingly analgesic requirement was higher in alcoholic group during follow up period. Again, it may reflect the narcotic abuse in alcoholic group.
However, there are several limitations on this research. First, this study was from a large, tertiary care referral centre. Therefore, it may reflect more severe disease than patients with chronic pancreatitis who are seen in primary care or community gastroenterology practice setting. Second, our median follow up period was only 18 months. Pain control and quality of life both may change in long-term follow up. Third, we have not analyzed the factors for incomplete pain relief in the follow up period.

CONCLUSION
Alcoholic chronic pancreatitis presents with more local complications than non-alcoholic chronic pancreatitis.
Smoking is prevalent in this group and they experienced more hospital admission before surgery. Pancreatic head mass, benign biliary stricture and segmental or left sided portal hypertension are significantly more prominent in alcoholic group in compared to idiopathic or tropical pancreatitis. As a result, resectional surgery is more performed in alcoholic patients. Frey procedure is a safe and well accepted surgery with a very good short-term outcome. Though there is significant difference in some aspects of quality of life (physical functioning and appetite loss) and pain score (analgesics requirement), further long-term data needs to be evaluated for proper conclusion in these parameters.