Risk Factors for Steatorrhea in Chronic Pancreatitis: A Cohort of 2,153 Patients

This study aimed to investigate the occurrence of and determine the risk factors for steatorrhea in chronic pancreatitis (CP). It was based on analysis of both retrospectively and prospectively acquired database for CP patients admitted to our center from January 2000 to December 2013. Demographic data, course of disease, medical history, and follow-up evaluations of patients were documented in detail. Cumulative rate of steatorrhea was calculated by using the Kaplan–Meier method. For risk factor analysis, multivariate analysis by Cox proportional hazards regression model was performed. A total of 2,153 CP patients were included with a mean follow-up duration of 9.3 years. Approximately 14% (291/2,153) of CP patients presented with steatorrhea at diagnosis of CP. Cumulative rates of steatorrhea at 1, 5, 10, and 20 years after diagnosis of CP were 4.27% (95% CI: 3.42%–5.34%), 12.53% (95% CI: 10.74%–14.59%), 20.44% (95% CI: 17.37%–23.98%) and 30.82% (95% CI: 20.20%–45.21%), respectively. Male gender (HR = 1.771, p = 0.004), diabetes (HR = 1.923, p < 0.001), alcohol abuse (HR = 1.503, p = 0.025) and pancreaticoduodenectomy (HR = 2.901, p < 0.001) were independent risk factors for steatorrhea while CP in adolescents (HR = 0.433, p = 0.009) was a protective factor. In conclusion, male gender, adult, diabetes, alcohol abuse and pancreaticoduodenectomy lead to increased risk of steatorrhea in CP patients.


Clinical characteristics of steatorrhea and non-steatorrhea patients. Steatorrhea patients differed
from non-steatorrhea patients in the following aspects: more male patients, fewer adolescent patients, higher prevalence of diabetes mellitus (DM), more ACP, more patients with initial manifestations of DM or steatorrhea, and different types of abdominal pain. No significant differences between these two groups were detected in terms of age at the onset or diagnosis of CP, common bile duct (CBD) stricture, pancreatic pseudocyst, stone, development of cancer, interventional treatment strategy, and death (Table 1).
After the first successful MPD drainage. For 1,544 patients who showed no signs of steatorrhea when successful drainage of MPD was performed by interventions, the cumulative rate of steatorrhea after the first successful intervention that achieved MPD drainage was described. Steatorrhea developed in 54, 109 and 123 patients at 1, 5, and 10 years after intervention treatment, with the cumulative rates being 3.91% (95% CI: 3.00%-5.07%), 10.86% (95% CI: 8.93%-13.17%) and 15.09% (95% CI: 12.34%-18.39%), respectively (Fig. 2e). Among the 1,544 patients, 82.19% (1269/1544) achieved successful drainage of MPD via endotherapy and the remaining 17.81% (275/1544) via surgery. Log-rank test showed that endotherapy group had lower cumulative rate of steatorrhea than the surgery group (mean interval of steatorrhea: 12.93 years vs. 12.17 years, p < 0.001, Fig. 2f). Predictors for steatorrhea. All 20 variables in Table 2 were considered potential predictors of newly occurring steatorrhea, and were analyzed in the univariate analysis. Eight variables showed a p value less than 0.10. These eight variables were selected as candidates for multivariate analysis by the Cox proportional hazards regression model. The results showed that five factors were independent predictors of newly onset steatorrhea. The risk factors were male gender (hazard ratio (HR) = 1.771, p = 0.004), DM (HR = 1.923, p < 0.001), alcohol abuse (HR = 1.503, p = 0.025), and pancreaticoduodenectomy (HR = 2.901, p < 0.001) and the protective factor was CP in adolescents (HR = 0.433, p = 0.009). RAP, recurrent acute pancreatitis; RAP/P, recurrent acute pancreatitis or adominal pain without significant increasing in serum amylase; CPP, chronic pancreatic pain; CBD, common bile duct; PPC, pancreatic pseudocyst. * ACP is defined as 80 g/d alcohol consumption that lasted for no less than two years for men and 60 g/d for women. A total of 727 patients (33.8%) had a history of alcohol consumption. # Referring to pain type in the course from onset to last personal contact or death. + Pancreatic calcification was also regarded as stone(s) that located in branch pancreatic duct or ductulus 46 . Discussion PEI, which has a negative effect on nutrition absorption, is rarely confirmed at the early stage for CP patients. Severe PEI, namely, pancreatic exocrine failure, which manifests as clinical steatorrhea, is commonly diagnosed in CP patients. A substantial proportion of CP patients had PEI without overt steatorrhea, and detection of risk factors for PEI would be clinically important. But direct function tests available for PEI are invasive and with risk of complications, which makes them difficult to be used in clinical practice as well as in clinical research. For the current study, we focused on severe PEI, and presence of steatorrhea was set as the sign of severe PEI. Steatorrhea is related to increased risks of cardiovascular events, osteoporosis, fracture et al. and is a direct and decisive evidence of PEI for CP in clinical practice. We detected the prevalence of steatorrhea in CP and its predictors with a relatively large sample size (Table 3), which may help improve the outcome of CP patients with PEI. For 2,153 CP patients, the overall incidence rate of steatorrhea was 22.9% over a median follow-up period of 7.8 years after the onset of CP. The cumulative rate of steatorrhea after the diagnosis of CP was 12.5% at 5 years, whereas Sandhu et al. reported a risk rate of 28% over a median follow-up period of 3.7 years (including 159 patients) 16 . We reported a lower rate of steatorrhea with longer disease course, which might be due to fewer ACP patients in our study. Pancreatic exocrine function is dependent on adequate acinar mass and function, and alcohol is toxic to the pancreas and is likely to severely damage the pancreatic parenchyma.
We identified several predictors for steatorrhea. Possible explanations for decreased risk of steatorrhea for CP in adolescents were: (1) compared to adult CP patients, adolescent CP patients have better preservation of pancreatic function and also better repair capacity after injury; (2) most of adolescent CP cases were followed up in less than 20 years, and we expect similar cumulative risk of steatorrhea with a long-term follow-up. A previous study also showed that exocrine and endocrine insufficiency developed more slowly in early-onset CP than in late-onset CP 6 . Therefore, adolescent CP patients had a reduced risk of steatorrhea compared to adult CP patients during the equivalent period of CP course. Pancreaticoduodenectomy was identified as an independent risk factor for steatorrhea, while drainage of pancreatic duct (by ESWL/ERCP or LPJ) or partial resection of pancreas (head resection or tail resection) seemed not to increase the risk of steatorrhea. This indicates that steatorrhea as a sign of pancreatic exocrine function failure is due to diffuse and severe loss of pancreatic parenchyma. CP being a progressive disease with loss of pancreatic acinar tissue, minimally invasive method rather than resectional procedures may be a better choice given the advantage of preservation of pancreatic parenchyma 19 .
In the current study, ACP was considered when alcohol intake exceeded 80 g/d for male and 60 g/d for female for at least 2 years for CP patients in the absence of other causes, respectively. ACP patients showed a higher risk of steatorrhea, but the risk might be underestimated. There might be a proportion of patients considered as ICP who might probably have ACP. The Oriental population might be more sensitive to alcohol-related injuries than the Caucasian population because they tend to have a less-active aldehyde dehydrogenase, which is a key detoxifying enzyme for alcohol 20,21 .
Patients with DM showed higher risk for steatorrhea. Considerable evidence suggests important functional interactions between the exocrine and endocrine pancreas. The interactions occur at several regulatory levels, and the true dimension is still unknown 22 .
Pancreatic stone was analyzed as one of the potential predictors of steatorrhea, which was indicated in several studies 7,23,24 . However, we were unable to find a relationship between pancreatic stone and steatorrhea risk in the context of ductal drainage, which is in agreement with the findings of Lankisch et al. 18 . Stone in CP is commonly considered to be associated with an abnormal constitution of proteins and supersaturation of calcium carbonate in pancreatic juice. However, the detailed mechanism is still unknown, and the interaction between stone and PEI is poorly understood.
The Changhai CP Database established in 2005 has been collecting clinical data of CP patients since January 2000, and data have been collected retrospectively before January 2005 and prospectively ever since. Statistical analysis showed that there was no significant difference between the clinical characteristics of patients who were first admitted before January 2005 and those admitted after that date. Therefore, the recall bias minimally influenced the results of the study.
Our study has some limitations. First, clinical steatorrhea was used as a sign of severe PEI; dietary habits and celiac disease-related steatorrhea were not considered. Second, the retrospectively acquired data collected between 2000 and 2004 might introduce recall bias. Prospective data collection since January 2005 minimized the chance of incomplete or inaccurate data collection. Third, risk factor analysis did not include all potential factors related to the development of steatorrhea. Fourth, 603 CP patients were followed up for less than 2 years after the diagnosis of CP; among these patients, several pancreatic cancer patients may have been misdiagnosed as CP 25 . However, these limitations minimally influence the results considering the relatively large sample size of the study.
In conclusion, CP patients showed increased risk of steatorrhea for those of male gender, adults, DM, alcohol abuse and pancreaticoduodenectomy. The evaluation of risk factors in CP patients before the occurrence of steatorrhea might help determine the replacement therapy of pancreatic enzyme earlier, which ensures that severe complications related to PEI can be avoided. Prospectively conducted studies are expected to confirm the benefit of early treatment of PEI on CP patients.

Methods
This study was based on analysis of both retrospectively and prospectively acquired database.
Patients and database. Since the 1990s, an electronic medical record system (GOODWILL Inc., Beijing, China) has been used in Changhai Hospital (Shanghai, China), which has facilitated several studies on CP [26][27][28][29] . In order to track changes consistently throughout the course of CP and to facilitate the evaluation and study of  The database system was also set a reminder for investigators to call patients for clinical checkups. Aside from visits due to complaints of discomfort related to CP, all patients were periodically (at least once a year) recalled for clinical checkup and investigations. Ultrasound, MRI, or CT was selected as an evaluation modality during each follow-up visit. An evaluation of each revisit, or an evaluation via telephone inquiries for patients who did not have follow-up visits to Changhai Hospital, was added to the CP database.
In December 2013, we contacted all the patients in our database for a final evaluation, except those who were lost to follow-up or died. The duration of follow-up is defined as the duration from the onset of CP to the date of the last personal contact, death, or December 2013, whichever came first.
The study was approved by the Ethics Committee of Changhai Hospital, The Second Military Medical University, Shanghai, China according to the Helsinki Declaration. Written informed consent was obtained from all participating patients. All of the diagnostic and therapeutic modalities were carried out in accordance to the approved guidelines.
Definitions. CP was diagnosed according to the Asia-Pacific consensus of CP 30 . ACP was diagnosed when alcohol intake exceeded 80 g/d for male and 60 g/d for female for at least two years in the absence of other causes, respectively 27,31 . HP was diagnosed when the CP patient had no less than two first-degree relatives with CP or recurrent acute pancreatitis (AP), or no less than three second-degree relatives with CP or recurrent AP 32 . We defined abnormal anatomy of pancreatic duct system (including pancreas divisum and anomalous pancreaticobiliary junction) as an etiology of CP in our study, although it still remains a controversy 33 . A patient was defined as post-traumatic CP due to a definite history of abdominal trauma with imaging evidence of pancreatic injury and subsequent ductal dilation. Hyperlipidemia was considered as an etiology when blood triglyceride was higher than 1,000 mg/dL at the diagnosis of CP 34 . CP patients were considered idiopathic when none of the above etiologies were found.
Patients who were diagnosed with pancreatic cancer less than two years after the diagnosis of CP were excluded from this study 26,35 . Autoimmune pancreatitis (AIP) and groove pancreatitis (GP) were also excluded from this study: AIP is different from typical CP in terms of pathophysiology, clinical manifestations, and prognosis, while GP could hardly be differentiated from pancreatic head carcinoma until pancreaticoduodenectomy and confirmed histological findings 36 .
Diagnosis of steatorrhea was established when either of the following conditions was met: (1) chronic diarrhea with foul-smelling, oily bowel movements 37 ; (2) a positive result in quantification of fecal fat determination (fecal fat quantification was performed over a period of three days; steatorrhea was defined as a fecal fat excretion of more than 14 g/day). Treatment strategy. As a tertiary medical center, Changhai Hospital admitted patients with previous pancreas-related surgical, endoscopic, or other invasive procedures from primary medical centers. In our center, minimally invasive interventions were taken as principle methods prior to surgery: extracorporeal shock wave lithotripsy (ESWL)/endoscopic retrograde cholangiopancreatography (ERCP) for stone removal and main pancreatic duct (MPD) drainage, insertion of stents to treat dominant MPD stricture and biliary duct stricture, and endoscopic drainage for uncomplicated pseudocyst with endoscopic reach 28,35,[38][39][40][41][42][43][44][45] . For CP patients who did not experience pain, interventions were performed only when CP was complicated by CBD stricture, pancreatic portal hypertension, et al.; DM or steatorrhea was not an indication for invasive treatment of CP 13 .
Data management and statistical analysis (Fig. 1). Continuous and categorical variables were presented as mean ± SD and counts (percentages), respectively. Student's t-test or Mann-Whitney U test and χ -square test or Fisher's exact test were used as indicated. Cumulative rates of steatorrhea were calculated by using the Kaplan-Meier method after the onset of CP (all CP patients were included), after the diagnosis of CP (patients with steatorrhea at the diagnosis of CP were excluded), and after the first successful drainage of MPD (only patients who without steatorrhea at the time of the first successful drainage of MPD were included). Log-rank test was used to further analyze the difference of cumulative rates of steatorrhea between two groups (e.g., male vs. female, endotherapy vs. surgical treatment). For risk factor analysis, multivariate analysis by Cox proportional hazards regression model was performed to identify the independent predictors based on the results of univariable analyses (factors with a significance level of p < 0.10 were included in the multivariate analysis). Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. Statistical analyses were conducted at a significance level of 0.05 for all analyses. Data were analyzed by using SPSS 18.0 (SPSS, Chicago, IL, USA).