Chinese Medical Journal 2013;126(1):147-153
A meta-analysis of the long-term effects of chronic pancreatitis surgical treatments: duodenum-preserving pancreatic head resection versus pancreatoduodenectomy

LÜ Wen-ping,  SHI Qing,  ZHANG Wen-zhi,  CAI Shou-wang,  JIANG Kai ,  DONG Jia-hong

LÜ Wen-ping (Department of Hepatobiliary Surgery, Chinese People’s Liberation Army General Hospital, Beijing 100853, China)

SHI Qing (Department of Microbiology, Peking University Health Science Center, Beijing 100191, China)

ZHANG Wen-zhi (Department of Hepatobiliary Surgery, Chinese People’s Liberation Army General Hospital, Beijing 100853, China)

CAI Shou-wang (Department of Hepatobiliary Surgery, Chinese People’s Liberation Army General Hospital, Beijing 100853, China)

JIANG Kai (Department of Hepatobiliary Surgery, Chinese People’s Liberation Army General Hospital, Beijing 100853, China)

DONG Jia-hong (Department of Hepatobiliary Surgery, Chinese People’s Liberation Army General Hospital, Beijing 100853, China)

Correspondence to:DONG Jia-hong,Department of Hepatobiliary Surgery, Chinese People’s Liberation Army General Hospital, Beijing 100853, China (Tel: . Fax:. E-mail:dongjh301@163.com)
Keywords
chronic pancreatitis; surgical treatments; long-term postoperative effect; meta-analysis
Abstract
Background  Surgery is regarded as the most effective treatment to relieve pain and reduce complications in chronic pancreatitis (CP). Two major strategies exist: duodenum-preserving pancreatic head resection (DPPHR) and pancreatoduodenectomy (PD). Many studies suggest that DPPHR offers advantages during surgery and in the short-term; however, the long-term effects have not been thoroughly investigated. We analyzed the long-term outcomes of DPPHR and PD, over follow-up times of at least 1 year, to determine the optimal surgical treatment for CP.
Methods  We systemically reviewed all CP surgical treatment reports, and only included randomized controlled trials (RCT) comparing DPPHR and PD, excluding unqualified studies using several pre-specified criteria. When multiple publications of a single trial were found, the most comprehensive current data were selected. Characteristics of the study populations and long-term postoperative outcome parameters were collected. The quality of the studies and data was analyzed using RevMan 4.2 software.
Results  Five trials were qualified for meta-analysis, with 261 participants in total (114 in the DPPHR group and 147 in the PD group). There were no significant differences in the age, gender, or indications for surgery of each group. At the mean of 5.7-year (1–14 years) follow-up examination, DPPHR and PD resulted in equally effective pain relief, exocrine and endocrine function, and similar mortality rates (P >0.05); however, DPPHR patients had improved global quality of life and weight gain, and reduced diarrhea and fatigue (P <0.05).
Conclusion  DPPHR and PD result in equal pain relief, mortality, and pancreatic function; however, DPPHR provides superior long-term outcomes.
Chronic pancreatitis (CP) is a common disorder of the gastrointestinal tract, and the causative factors include biliary lithiasis, autoimmune or genetic predisposition, and anatomical variations.1 Pain is the leading symptom, developing in 85.0% of CP patients. During the early stages of CP, pain is intermittent, but can become persistent. Furthermore, as there is a concurrent loss of pancreatic parenchymal tissue, other symptoms related to exocrine and endocrine insufficiencies develop as the disease progresses.2
 
To date, numerous interventions have been used for the treatment of CP. Medical management may temporarily relieve pain and improve digestion and nutrition; however, there is a risk of developing tolerance or dependency to medication, and the therapeutic effects do not last long. In addition, sphincterotomy, endoscopic stone extraction, extracorporeal shock wave lithotripsy, and stenting of the pancreatic duct may lead to unsatisfactory long-term results. Most patients with CP eventually require surgical intervention to treat the intractable abdominal pain. Surgery has been shown to yield significant pain relief and improved quality of life compared to endoscopic drainage procedures.3 The major purpose of surgical treatment for CP is to relieve pain, reduce complications involving the adjacent organs, preserve pancreatic function, and restore quality of life for the patient.4
 
There are two principal surgical strategies for CP: pancreaticoduodenectomy (PD) and duodenum- preserving pancreatic head resection (DPPHR). PD has been performed in a standard fashion since it was first reported in 1980.5 Many clinical reports have suggested that PD provides effective pain relief and has a low incidence of complications.6 In 1985, pylorus-preserving pancreatoduodenectomy (ppPD) was described, which leads to satisfactory postoperative weight gain, prevents symptoms associated with gastrectomy, and can be performed with an extremely low mortality rate.7 However, over the last two decades, DPPHR has become increasingly popular for the treatment of CP. DPPHR was specifically developed to treat complications induced by an inflammatory mass in the pancreatic head8 and was first reported by Beger et al.9 In DPPHR, the pylorus, duodenum, and extrahepatic bile ducts are preserved after subtotal resection of the pancreatic head.9 DPPHR was first modified by Frey and Smith,10 who described an extended drainage procedure with local pancreatic head excision. The subsequent modifications by Gloor et al11 combine the advantages of both techniques described by Beger and Frey et al,9,10 and avoid delicate division of the pancreatic neck anterior to the portal vein.
 
It is important to determine the optimal surgical strategy for patients with CP. In 2008, Diener et al12 conducted a meta-analysis comparing DPPHR and PD, which showed that DPPHR and PD were equally effective in terms of pain relief, overall morbidity, and reduction of postoperative endocrine insufficiency. However, DPPHR was superior with regard to several peri- and post-operative parameters of effectiveness and quality of life. In order to further delineate the advantages and disadvantages of the DPPHR and PD procedures, we conducted a more thorough current literature search and meta-analysis to compare the long-term outcomes of the two surgical treatments.
 
METHODS
 
Search strategy and study selection criteria
Two independent reviewers searched the following electronic databases: Medline, Biosis, Embase, Wos and the Cochrane library database. The final search was performed on 23 May, 2011. The search was not restricted by language, and the search strategy used a combination of “pancreatitis” AND “pancreatectomy” OR “Whipple OR Beger OR Frey OR Bern” as the MeSH terms and text words. Additionally, we also manually examined reference lists and contacted authors when necessary. All the included studies were required to meet the following three criteria: (1) the study populations were patients diagnosed with CP who underwent a surgical treatment; (2) the aim of the trial was to compare the effectiveness of DPPHR (described by Beger, Frey or Bern et al) with PD; (3) the trial was a randomized controlled trial; and (4) the postoperative follow-up time was  ³1 year.
 
Data collection and extraction
Two principle types of information were extracted from the included studies: (1) the demographic characteristics of the study populations (age, gender, history of symptoms, and operative indications) and (2) long-term postoperative outcome parameters, including mortality, functional scales, and symptom scales. If more than one publication was generated from a single study, the most comprehensive and up-dated data were selected for the meta-analysis.
 
Assessment of the quality of the studies
The quality of trials which were designed with control and treatment groups was assessed using Review Manager (Version 4.2.2 for Windows; The Cochrane Collaboration, Oxford, UK).
 
A number of criteria were adopted for this assessment, including randomization, heterogeneity, and selection bias of the arms in the study, and using CONSORT criteria for assessment of RCT quality.13
 
Statistical methods
Cochrane Collaboration’s tool was used to assess the risk of bias for each trial. Relative risk (RR) and 95% confidence intervals (CI) were adopted for dichotomous outcomes. Continuous outcome variables were expressed as the weighted mean difference (WMD) and 95% CI, unless differing validated scales were reported. When the interquartile range (iqr) was given instead of standard deviations (SD), we converted the data using the formula SD = 0.25 × iqr.12 The Chi-squared distribution test was performed and the I-squared statistic (I2) was calculated, in order to determine whether significant heterogeneities existed between trials. If the studies were highly heterogeneous, the overall outcomes were calculated with the random effects model; otherwise, outcomes were calculated with the fixed effect model. A P value less than 0.05 was considered statistically significant.
 
RESULTS
 
Search results
Five trials met the inclusion criteria and were included in the meta-analysis. Figure 1 shows a flow diagram of the selection and exclusion of papers for this study. The included studies were published between 1995 and 2009. The follow-up duration in all studies was equal to or more than 1 year, with a mean follow-up period of 5.7 years (range, 1–14 years). In this study, the trials by Müller et al14 and Strate et al15 were included as more current reports of the data published by Buchler et al16 and Izbicki et al,4 respectively, which were included in the previous analysis by Diener et al.12
 
 

view in a new window
Figure 1. The quality of reporting of meta-analyses (QUOROM) chart for the meta-analysis.
 
Characteristics of the study population
Of the five trials suitable for meta-analysis, four were conducted in Germany while the trial by McClaine et al17 was performed in the USA. In total, 261 participants were randomized into DPPHR (n=114) or PD groups (n=147). Both study groups were similar with regard to age (43.5 years in DPPHR vs. 45.6 years in PD), gender (66 males and 27 females in DPPHR vs 85 males and 44 females in PD), and duration of symptoms (6.2 years in DPPHR vs. 5.6 years in PD; Table 1). The indications for operation were intractable pain, bile duct stenosis, pancreatic duct stenosis, pancreatic head mass, and duodenal obstruction (detailed data were not shown), with no differences between the indications for the patients in each group.
 

view in a new window
Table 1. Characteristics of the group populations for the studies included in the DPPHR and PD meta-analysi
 
Methodological quality of the trials
All five enrolled trials were RCTs and four of them reported the details of the random allocation of patients.14,15,18,19 Only two trials provided further information about allocation concealment.15,19 None of the studies included mentioned blinded status or procedures. Among all five studies, one applied predefined follow-up periods of 1 year18 and duration of follow-up in the remaining studies varied among patients. Heterogeneous study quality was shown with sample size, standardization of study interventions, consistency of follow-up, and outcome assessment.
 
Mortality and follow-up outcomes
Twenty-three patients died during follow-up, 12/114 in the DPPHR group and 11/147 in the PD group (P=0.71, Figure 2). The main causes of death were unrelated to chronic pancreatitis, and included liver cirrhosis and myocardial infarction. In addition, some patients failed to return questionnaires and several patients were lost to follow-up in the original studies. A total of 206 patients were available for meta-analysis (102 in the DPPHR group and 104 in the PD group).
 
Exocrine and endocrine insufficiency outcomes
Data regarding post-operative exocrine insufficiency was available in all the five trials. Three trials presented data from the standard pancreatic exocrine evaluations, stool elastase level,18 and pancreatic enzyme use.14,17 No significant differences were found in exocrine insufficiency between the two groups (P=0.17; Figure 3).
 
Of the four trials reporting endocrine insufficiency data, 22 of 69 (31.9%) patients in the DPPHR group and 39 of 83 (47.0%) patients in the PD group had endocrine insufficiency; these differences were not significantly different (Figure 4; P=0.06). Müller et al14 also presented data on the postoperative onset of diabetes mellitus, which occurred in 4/11 of DPPHR and 6/9 of PD patients, P=0.327).
 
Evaluation of quality of life and outcome scales
Global quality of life was assessed using the European Organization for Research and Treatment of Cancer (EORTC) Quality-of-Life Questionnaire (QLQ). The analysis of the long-term outcomes of the DPPHR and PD procedures suggested that the patients in the DPPHR group had a significantly better global quality of life compared to the PD group (RR 11.57; 95% CI, 6.51–16.64; P <0.01; I2 =0% data heterogeneity; Table 2, Figure 5). For assessment of long-term pain relief, pain scores were used in three trials14,15,17 (Table 2), while pain relief ratios were presented in the other two (17/20 in the DPPHR group vs. 18/20 in the PD group in Farkas et al18 and 14/20 in the DPPHR group vs. 12/20 in the PD group in Klempa et al19). Overall, the pain results were not significantly different between the two groups (WMD –1.44; 95% CI, –11.56–8.68; P=0.78; I2=0 data heterogeneity).
 

view in a new window
Table 2. Meta-analysis of DPPHR and PD outcomes
 

view in a new window
Figure 2. Meta-analysis of mortality for the DPPHR and PD procedures.
Figure 3. Meta-analysis of exocrine insufficiency for the DPPHR and PD procedures.
Figure 4. Meta-analysis of endocrine insufficiency for the DPPHR and PD procedures.
 
 
Significant improvements in diarrhoea and fatigue were observed in the DPPHR group compared to the PD group (P <0.01, Figures 6 and 7, Table 2). Patients in both groups of three trials exhibited equal body weight loss (WMD –15.15; 95% CI, –41.52–11.23; P=0.26; I2=95.2% data heterogeneity; Table 2). In the other two trials, weight gain was assessed, and 31/40 (77.5%) DPPHR patients gained weight compared to 12/41 (29.3%) PD patients,18,19 showing that DPPHR significantly increased the ability to gain weight (P <0.01). There were no significant differences observed between mouth dryness and the other outcome parameters in the two groups (P >0.05, Table 2).
 

view in a new window
Figure 5. Meta-analysis of global quality of life for the DPPHR and PD procedures.
Figure 6. Meta-analysis of diarrhoea for the DPPHR and PD procedures.
Figure 7. Meta-analysis of fatigue for the DPPHR and PD procedures.
 
 
DISCUSSION
 
The optimal surgical procedure for CP treatment should relieve pain, reduce mortality and morbidity, and improve quality of life. Currently, PD remains the standard option for CP, while the DPPHR technique has been adopted at varying levels in many countries. One reason for this trend is that doctors are more familiar with the PD procedure, and importantly, many retrospective reviews have demonstrated that the PD procedure has become increasingly safe,20 while arguing that overall, there is no evidence to suggest that DPPHR is superior to PD.
 
However, many other studies have suggested that DPPHR is superior to PD in many aspects. DPPHR is less invasive than PD, and preservation of the entire duodenum is an important factor for maintenance of pancreatic function.21 Although DPPHR has a higher complication rate than ppPD in benign or premalignant lesions, it results in a better long-term outcome, as it preserves the entire duodenum and a normal biliary tree.22 DPPHR is also considered to offer the best short-term outcome2 and quality of life,15 and provide excellent functional results, with a relatively low postoperative morbidity and short hospital stay.23 A 5-year long-term outcome analysis showed that patient self-assessed quality of life and pain intensity after the Beger procedure was superior than after classic PD.24
 
In our analysis of long-term outcomes, QLQ scores showed that patients in the DPPHR group had better global quality of life than the PD group (WMD 11.57, P <0.01). The extent of pain was assessed using forms filled out by participants, and showed no significant differences in pain relief between the two groups, similar to the analysis by Diener et al (RR 1.08; P=0.46).12 However, Büchler et al16 reported that DPPHR resulted in a lower incidence of pain than PD 6 months after the operation (12/16 in DPPHR vs. 6/15 in PD, P <0.05), and concluded that DPPHR led to less pain initially, but this advantage was gradually lost with time.
 
There are some limitations in this meta-analysis, including small sample sizes, inadequate allocation concealment, different period of follow-up, lack of a specific definition of long-time follow up and loss of population during follow-up. With regard to these problems, we have searched as many as well-known databases, all trials included selected study population with defined inclusion and exclusion criteria with no significant differences and randomly divided them into PD and DPPHR groups. However, the methodological quality of the trials included was not high, and most of them provided inadequate information regarding allocation concealment.
 
In our study, 114 and 147 patients were included in DPPHR and PD groups respectively, and the number is too small to analyze publication bias, which may be due to low methodological quality. Although there is a lack of a specific definition of long-time follow up, all trials conducted follow-up equal or more than 1 year. We selected the relative longer ones for two trials which have published a short and a long-term follow-up studies one after another.14,15 Eventually, 12 and 43 patients were lost for various reasons during the follow-up in DPPHR and PD groups respectively; no significant difference was found between two groups.
 
DPPHR was initially introduced to reduce the postoperative impairment in pancreatic function; however, our analysis suggests that DPPHR may not improve preservation of pancreatic function, as the differences in the DPPHR and PD groups were not statistically significant (P >0.05). However, Diener et al12 reported a significant superiority of the DPPHR in terms of less exocrine insufficiency, while Strobel et al25 reported that the initial functional advantages of DPPHR were lost over the period of follow-up. It is likely that neither DPPHR nor PD can prevent a progressive loss of exocrine and endocrine function, which emphasizes the fact that development of pancreatic insufficiency does not depend on the surgical procedure,26 but instead may be related to the features of CP. Additionally, the lack of superiority in pain relief provided by DPPHR may be related to the fact that exocrine insufficiency is the strongest independent risk factor for pain at follow-up.27
 
Body weight is a good indication of physical status, and in our analysis, both groups exhibited an increase in body weight, which was significantly higher in the DPPHR group compared to PD. DPPHR offered additional advantages over PD, including a significantly reduced incidence of diarrhoea and fatigue. While we did not observe significant differences in the other outcome parameters between the two groups in this meta-analysis, other studies have found that patients who underwent PD were more likely to be thirsty and lose their appetite.14,17
 
In conclusion, this meta-analysis indicates that DPPHR and PD are equally effective with regard to pain relief and preservation of pancreatic function in CP; however, at long-term follow up, DPPHR results in a better quality of life and improved long term outcome improvement in several measures including weight gain, diarrhoea, and fatigue.
 
REFERENCES
 
1.  Mayerle J, Lerch MM. Is it necessary to distinguish between alcoholic and nonalcoholic chronic pancreatitis? J Gastroenterol 2007; 42: 127-130.
2.  Bachmann K, Kutup A, Mann O, Yekebas E, Izbicki JR. Surgical treatment in chronic pancreatitis timing and type of procedure. Best Pract Res Clin Gastroenterol 2010; 24: 299-310.
3.  Díte P, Ruzicka M, Zboril V, Novotný I. A prospective, randomized trial comparing endoscopic and surgical therapy for chronic pancreatitis. Endoscopy 2003; 35: 553-558.
4.  Izbicki JR, Bloechle C, Broering DC, Knoefel WT, Kuechler T, Broelsch CE. Extended drainage versus resection in surgery for chronic pancreatitis: a prospective randomized trial comparing the longitudinal pancreaticojejunostomy combined with local pancreatic head excision with the pylorus-preserving pancreatoduodenectomy. Ann Surg 1998; 228: 771-779.
5.  Beger HG, Witte C, Krautzberger W, Bittner R. Experiences with duodenum-sparing pancreas head resection in chronic pancreatitis. Chirurg 1980; 51: 303-307.
6.  Schmidt CM, Powell ES, Yiannoutsos CT, Howard TJ, Wiebke EA, Wiesenauer CA, et al. Pancreaticoduodenectomy: a 20-year experience in 516 patients. Arch Surg 2004; 139: 718-727.
7.  Braasch JW, Rossi RL. Pyloric preservation with the Whipple procedure. Surg Clin North Am 1985; 65: 263-271.
8.  Farkas G, Leindler L, Daróczi M, Farkas G Jr. Organ-preserving pancreatic head resection in chronic pancreatitis. Br J Surg 2003; 90: 29-32.
9.  Beger HG, Krautzberger W, Bittner R, Büchler M, Limmer J. Duodenum-preserving resection of the head of the pancreas in patients with severe chronic pancreatitis. Surgery 1985; 97: 467-473.
10.  Frey CF, Smith GJ. Description and rationale of a new operation for chronic pancreatitis. Pancreas 1987; 2: 701-707.
11.  Gloor B, Friess H, Uhl W, Büchler MW. A modified technique of the Beger and Frey procedure in patients with chronic pancreatitis. Dig Surg 2001; 18: 21-25.
12.  Diener MK, Rahbari NN, Fischer L, Antes G, Büchler MW, Seiler CM. Duodenum-preserving pancreatic head resection versus pancreatoduodenectomy for surgical treatment of chronic pancreatitis: a systematic review and meta-analysis. Ann Surg 2008; 247: 950-961.
13.  Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000; 283: 2008-2012.
14.  Müller MW, Friess H, Martin DJ, Hinz U, Dahmen R, Büchler MW. Long-term follow-up of a randomized clinical trial comparing Beger with pylorus-preserving Whipple procedure for chronic pancreatitis. Br J Surg 2008; 95: 350-356.
15.  Strate T, Bachmann K, Busch P, Mann O, Schneider C, Bruhn JP, et al. Resection vs drainage in treatment of chronic pancreatitis: long-term results of a randomized trial. Gastroenterology 2008; 134: 1406-1411.
16.  Büchler MW, Friess H, Müller MW, Wheatley AM, Beger HG. Randomized trial of duodenum-preserving pancreatic head resection versus pylorus-preserving Whipple in chronic pancreatitis. Am J Surg 1995; 169: 65-70.
17.  McClaine RJ, Lowy AM, Matthews JB, Schmulewitz N, Sussman JJ, Ingraham AM et al. A comparison of pancreaticoduodenectomy and duodenum-preserving head resection for the treatment of chronic pancreatitis. HPB (Oxford) 2009; 11: 677-683.
18.  Farkas G, Leindler L, Daróczi M, Farkas G Jr. Prospective randomised comparison of organ-preserving pancreatic head resection with pylorus-preserving pancreaticoduodenectomy. Langenbecks Arch Surg 2006; 391: 338-342.
19.  Klempa I, Spatny M, Menzel J, Baca I, Nustede R, Stöckmann F, et al. Pancreatic function and quality of life after resection of the head of the pancreas in chronic pancreatitis. A prospective, randomized comparative study after duodenum preserving resection of the head of the pancreas versus Whipple’s operation. Chirurg 1995; 66: 350-359.
20.  Russell RC, Theis BA. Pancreatoduodenectomy in the treatment of chronic pancreatitis. World J Surg 2003; 27: 1203-1210.
21.  Ito K. Duodenum preservation in pancreatic head resection to maintain pancreatic exocrine function (determined by pancreatic function diagnostant test and cholecystokinin secretion). J Hepatobiliary Pancreat Surg 2005; 12: 123-128.
22.  Pedrazzoli S, Canton SA, Sperti C. Duodenum-preserving versus pylorus-preserving pancreatic head resection for benign and premalignant lesions. J Hepatobiliary Pancreat Sci 2011; 18: 94-102.
23.  Varghese TK, Bell RH. Duodenum-preserving head resection for chronic pancreatitis: an institutional experience and national survey of usage. Surgery 2007; 142: 588-593.
24.  Möbius C, Max D, Uhlmann D, Gumpp K, Behrbohm J, Horvath K, et al. Five-year follow-up of a prospective non-randomised study comparing duodenum-preserving pancreatic head resection with classic Whipple procedure in the treatment of chronic pancreatitis. Langenbecks Arch Surg 2007; 392: 359-364.
25.  Strobel O, Buchler MW, Werner J. Duodenum-preserving pancreatic head resection: technique according to Beger, technique according to Frey and Berne modifications. Chirurg 2009; 80: 22-27.
26.  Malka D, Hammel P, Sauvanet A, Rufat P, O’Toole D, Bardet P, et al. Risk factors for diabetes mellitus in chronic pancreatitis. Gastroenterology 2000; 119: 1324-1332.
27.  Riediger H, Adam U, Fischer E, Keck T, Pfeffer F, Hopt UT, et al. Long-term outcome after resection for chronic pancreatitis in 224 patients. J Gastrointest Surg 2007; 11: 949-960.
  1. National Natural Science Foundation of China,No. 30772493 and No. 81170429;