- (2009) Volume 10, Issue 1
Nehal S Shah, Ajith K Siriwardena
Hepatobiliary Surgical Unit, Manchester Royal Infirmary. Manchester, United Kingdom
Received August 13th, 2008 - Accepted October 29th, 2008
Context Evidence to guide selection of optimal surgical treatment for patients with painful chronic pancreatitis is limited. Baseline assessment data are limited and thus patients in different centres may be presenting at different stages of their illness. Objective This study undertakes a systematic overview of reports of elective surgical intervention in chronic pancreatitis with particular reference to reporting of quality of life and baseline assessment and relation between disease and type of procedure. Methods A computerised search of the PubMed, Embase and Cochrane databases was undertaken for the period January 1997 to March 2007 yielding 46 manuscripts providing data on 4,626 patients undergoing elective surgery for chronic pancreatitis. The median number of patients per study was 71 (range: 4-484). The median period for recruitment of patients was 10 years (range: 2-36 years). Results An externally validated quality of life questionnaire is reported in 8 (17.4%) of 46 manuscripts covering 441 (9.5%) of 4,626 patients. Formal comparison of pre-operative and post-operative pain scores was provided in 15 (32.6%) of manuscripts. Only seven (15.2%) reports provide a formal rationale or indication for selection of the type of elective surgical procedure for a stated disease variant and these papers cover 481 (10.4%) patients. Conclusion In conclusion, this study demonstrates that there is a lack of standardization between units of the criteria for operative intervention in painful chronic pancreatitis. At a minimum, formal quality of life testing using a validated system should be undertaken in all patients prior to elective surgery for painful chronic pancreatitis.
DPPHR: duodenum-preserving pancreatic head resection
Chronic pancreatitis is characterised by recurrent abdominal pain and progressive destruction of pancreatic parenchyma leading to exocrine and endocrine deficiency [1, 2, 3]. The spectrum of disease is variable and ranges from mild attacks to severe debilitating disease. The incidence of chronic pancreatitis is not well characterised but a French study reports figures in the order of 5-10 cases per 100,000 population [4]. The pathological changes within the pancreas vary and include dilatation of the main pancreatic duct (either uniformly or in a characteristic segmental form), parenchymal calcification and cystic change. These changes can either be localised or diffusely spread throughout the gland [1].
Surgical treatment for chronic pancreatitis has been reported to provide good relief of symptoms [5, 6]. Types of abdominal surgical intervention for chronic pancreatitis can be broadly categorised into pancreatic resectional procedures or drainage operations. Newer operations include minimally invasive division of the splanchnic nerves in the thoracic cavity [7]. However, the evidence to guide the selection of any given intervention is limited. In particular, reporting of baseline clinical assessment data are limited and thus patients in different centres may be presenting and undergoing surgery at different stages of their illness. Also, the information to guide selection between resection and drainage is limited and a number of centres appear to promote an operation or procedure developed within their particular unit. This study undertakes a systematic overview of elective surgery for chronic pancreatitis with particular reference to assessment of indication for surgery, type of surgery employed and relation between indication and type of surgery.
A computerised search of the PubMed, Embase and Cochrane databases was performed using the search engine ADITUS for the period from June 1997 to June 2007. The Medical Subject Headings (MeSH) “Chronic pancreatitis” and “Surgery” were used. The results were combined using Boolean operators to yield a total of 374 articles. Only articles providing original information were retained. Articles which were reviews, case reports, referred to patients with known pancreatic malignancy or which did not providing original information were excluded. In cases of sequential publication of data, the manuscript providing the most detail was retained. These exclusions produced a final study population of 46 manuscripts providing data on 4,626 patients undergoing elective surgery for chronic pancreatitis. The manuscripts were then reviewed and data extracted on demographic profile including aetiology of chronic pancreatitis, gender, country of publication and study period. Specific information was then sought on whether chronic pancreatitis was defined in the manuscript, whether information was provided on the indication for surgery, whether quality of life assessments were undertaken and on the type of surgery undertaken with particular reference to whether information was provided on the relationship between the indication for surgery and the type of surgery undertaken. Data on outcome of surgery were also sought. All manuscripts were reviewed independently by two authors and a final agreed dataset utilised for analysis.
Two by two tables are analysed by Fisher’s exact and the Mann-Whitney U tests using the Statsdirect software package (version 2.6.5., www.statsdirect.com).
Demographic Profile
The study population comprised 4,626 patients reported in 46 studies [8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52]. The median number of patients per study was 71 (range 4-484) (Table 1). Twenty six (56.5%) reports originated from Europe, 14 (30.4%) from the United States of America, 5 (10.9%) from Asia/Far-East and 1 (2.2%) from Mexico. The median study period for recruitment (including follow-up) of patients was 10 years (range: 2-36 years). There were 3,337 (72.1%) males. The most common aetiology was alcohol in 2,765 (61.2%) patients (note that aetiology was not available for 110 patients). Other aetiologies included idiopathic chronic pancreatitis in 893 (19.8%), tropical chronic pancreatitis in 209 (4.6%) and “others” in 649 (14.4%).
Description of Indications for Surgery
We found one only study without indications (Olah A, et al. [19]); therefore, the indication for surgery was described in 45 of 46 articles (97.8%) for a total of 4,539 patients. Of the listed indications, abdominal pain was the principal factor in 3,833 (84.4%) patients (Table 2). Formal comparison of pre-operative and post-operative pain scores was provided in 15 manuscripts (32.6%) (Table 3). Similarly, pre and postoperative analgesic usage was reported in 18 (39.1%).
Type of Surgery
Pancreatic head resectional surgery (either conventional Whipple or pylorus-preserving variant) was the most frequently undertaken procedure in this series being undertaken in 1,206 (26.1%) (Table 2). Duodenum-preserving pancreatic head resection (DPPHR) was undertaken in 745 (16.1%). Distal pancreatectomy (with or without splenic preservation) was undertaken in 563 (12.2%). Total pancreatectomy (with or without spleen and duodenum preservation) was undertaken in 122 (2.6%). Of drainage procedures, lateral pancreatico-jejunostomy was the most frequently reported being described in 989 (21.4%). In terms of the distribution of procedures across centres, the number of services reporting the use of DPPHR was13, of which 4 services reported a combined total of 16 patients undergoing DPPHR. The number of centres reporting the use of the V-shaped excision of pancreatic head was 1 (in 37 patients). Only seven (15%) reports provide a formal rationale or indication for selection of the type of elective surgical procedure for a stated disease variant and these papers cover 481 (10.4%) patients.
Operative Outcome
There were 179 (3.9%) in-hospital deaths after surgery. One thousand and seventy five (23.2%) patients had one or more surgical complications. Thirty eight (82.6%) reports gave information on a total of 1,609 (39.7%) patients with post-operative diabetes from a total population of 4,056 patients in these reports. Prior to surgery, 908 (22.0%) of 4,129 patients were reported to have diabetes mellitus. After surgery (interval not consistently reported), 1,609 (39.7%) of 4,056 were diabetic and this difference was significant (P<0.001; Fisher’s exact test).
Fifteen studies provided data on pain scores before and after surgery (Table 3). Comparison of pooled preoperative pain scores to those after surgery (accepting the variable median follow-up) shows a significant reduction in pain (P<0.002; Mann-Whitney U test).
Use of Quality of Life Assessments
An externally validated quality of life questionnaire such as either the short form 36 (SF-36) or the European Organisation for Research and Treatment of Cancer’s quality of life questionnaire with pancreasspecific module (EORTC QLQ C-30 plus PAN-26) [53] is reported in 8 (17.3%) of 46 manuscripts covering 441 (9.5%) of 4,626 patients (Table 2). Seven of these studies reported quality of life assessment comparing pre and post operative values. In addition, 5 studies have described the outcome/quality of life using non-validated questionnaire.
This article has undertaken a systematic overview of reports on elective surgery for chronic pancreatitis. With 4,626 patients undergoing elective surgery included in the study population of 46 reports, this is one of the largest patient cohorts reported. It is accepted that despite stringent computerised database searches and the precaution of cross-checking databases it is likely that not all relevant articles will have been captured and that there will be a bias toward publication of manuscripts with “positive” results. Nevertheless, the study population is thought to be representative of surgical practice in chronic pancreatitis. Other important methodological limitations are likely to include variation in definition of chronic pancreatitis, differences in case selection and probably differences in interpretation of the exact nature of a specific surgical procedure (such as the DPPHR) between reports. However, interpretation of the data presented here while keeping these limitations in mind has highlighted several interesting and consistent trends across reports
First, our study shows that almost twice as many reports originate from Europe as compared to the United States of America (Table 1). The reason for this difference is not apparent from these studies and may be multifactorial but could also reflect differences in clinical practice rather than true differences in disease incidence.
Second, it can be seen that median recruitment time to these studies is lengthy (Table 1), suggesting that even in specialist units, relatively few patients are submitted to surgery. In turn, this may mean that individual units are less able to sustain randomized comparative trials and also that collective experience of the surgical management of chronic pancreatitis may be relatively limited.
The findings that alcohol was the most frequent etiologic agent and pain the most often quoted indication for surgery are not new.
Although “pain and complications” were the most frequently cited indications for surgery (Table 2), the relative dearth of use of pre-operative pain scores or quality of life assessments could be regarded as a critical limitation of the published literature. It is difficult to compare between reports (in particular reports recommending different procedures for chronic pancreatitis) when it is not certain whether all reports are referring to the same baseline disease burden or pain levels. Similarly, it is not possible from the wide variety of operative interventions undertaken in patients with “pain and complications” to match type of surgical intervention to disease presentation. It could be further argued that the wide variation in duration of disease interval prior to surgery (Table 2) is indirect evidence that centres are reporting operative interventions on patients at different disease stages.
Although all reports refer to elective surgery in patients with chronic pancreatitis, some operative procedures are carried out for the complications of chronic pancreatitis rather than undertaking resection or drainage of the diseased gland and hence the range of reported interventions include pseudocyst drainages (Table 2).
In a chronic and essentially stable although disabling condition, it was a surprising finding to report that so few studies undertook formal quality of life assessment. One explanation could be that questionnaires validated for use in chronic pancreatitis have only recently become widely available. However, over a decade ago, Izbicki’s group were using quality of life assessment questionnaires to assess the outcome of surgery [44].
In terms of the type of surgery undertaken, although the Whipple-type pancreatic head resection was the most frequently undertaken, it appears that a contemporary trend is more towards duodenumpreserving pancreatic head resection. Of particular interest is the limited information provided in terms of matching of operative procedure to disease variant. Further, it appears that particular units undertake procedures developed within those units.
Currently available pointers for selection of treatment option include head resection by pancreaticoduodenectomy where there may be concerns about underlying malignancy and drainage in patients with duct dilatation in the absence of a pancreatic head mass. However, the treatment of patients with small duct disease remains unclear and in our study, patients with this type of disease were treated by the Izbicki operation by one group [9] but by other procedures by others [21, 43]. Similarly, criteria for case selection for duodenum preserving head resection compared to pancreaticoduodenectomy are unclear. A rational allocation of role in this management algorithm for thoracoscopic splanchnic nerve division is also required
It is encouraging to note that there is more recent published evidence providing long-term outcome data on a randomized trial comparing Whipple pancreaticoduodenectomy to duodenum-preserving pancreatic head resection [54].
In conclusion, this study demonstrates that there remains a pressing lack of standardization between units in relation to the criteria for operative intervention in painful chronic pancreatitis. In a vulnerable patient group and in the setting of an essentially stable disease, this lack of concordance is clearly highlighted in this overview and is difficult to justify in contemporary surgical practice. From our findings, it would seem logical to suggest that if clinicians are to undertake surgery for chronic pancreatitis that at a minimum, formal quality of life testing using a validated system is undertaken in all patients prior to elective surgery for painful chronic pancreatitis. In addition, longer-term follow-up is required before reporting of data.
None
Papers based on this manuscript were read at the 8th World Congress of the International Hepato-Pancreato-Biliary Association, Mumbai, India, 29th February 2008 and appeared in abstract form (HPB 2008; 10: Suppl 1) and also as a poster at the American Hepato-Pancreato-Biliary Association, Fort Lauderdale, USA, March 2008 (HPB 2008; 10: Suppl 2).
