Abstract
Background: An inflammation-based prognostic score, the modified Glasgow prognostic score (mGPS), has been reported to be useful for predicting postoperative survival in patients with various types of cancer. However, no studies have investigated whether the mGPS can predict biliary stent (BS) patency in patients undergoing BS placement for unresectable malignant biliary obstruction (UMBO). Aim: To evaluate the usefulness of the mGPS for predicting BS patency in patients undergoing intraoperative placement of uncovered expandable metallic stents (EMSs) for UMBO. Patients and Methods: The mGPS was calculated as follows: patients with both an elevated C-reactive protein (CRP) level (>1.0 mg/dl) and hypoalbuminemia (<3.5 g/dl) were allocated a score of 2. Patients with only an elevated CRP level were allocated a score of 1, and patients without an elevated CRP level were allocated a score of 0. EMS patency was compared by Kaplan–Meier analysis and log-rank test between the two groups (mGPS 0 vs. mGPS 1 or 2). The significant risk factors for EMS occlusion were investigated by Cox proportional hazards model analysis. Results: Kaplan–Meier analysis revealed that patients with mGPS 1 (n=7) and 2 (n=19) had a lower EMS patency rate (p=0.014) than patients with mGPS 0 (n=37). Although univariate analyses revealed that a high serum total bilirubin level, stent-in-stent placement, and mGPS 1 or 2 were significant risk factors predictive of EMS occlusion, multivariate analysis demonstrated that no independent risk factors were significant. Conclusion: mGPS is a significant predictor of EMS patency in patients undergoing intraoperative placement of BS.
- Biliary stent
- Glasgow prognostic score
- mGPS
- stent patency
- unresectable malignant biliary obstruction
Biliary decompression has an important impact on the prognosis and quality of life of patients with unrescetable malignant biliary obstruction (UMBO). Biliary decompression can be achieved by several types of procedure, such as surgical biliary bypass, endoscopic transpapillary biliary stenting, percutaneous transhepatic biliary stent insertion, and external biliary drainage (1-3). Although surgical biliary bypass including choledochojejunostomy or cholecystojejunostomy has been performed in patients with distal UMBO diagnosed during surgical exploration, biliary decompression using surgical biliary bypass is very difficult in patients with proximal UMBO due to gallbladder carcinoma, hilar cholangiocarcinoma and cholangiocellular carcinoma (4). In patients with proximal UMBO, biliary stenting (BS) or external bilary drainage is performed after surgery. Thus, BS has become a standard strategy for biliary decompression of proximal UMBO or distal UMBO diagnosed by recently improved imaging devices such as computed tomography (CT), magnetic resonance imaging (MRI) and positron-emission tomography (PET) (5).
Two types of materials are used for BS: metal and plastic. Metallic BSs can help to improve prognosis because not only do they maintain their patency for longer than plastic stents, but they also allow antitumor therapy in the absence of biliary occlusion (6). However, metallic are more expensive than plastic ones, and are difficult to remove once they become occluded (7, 8). On the other hand, plastic BSs can easily be removed and exchanged after occlusion (7, 8). Although covered metallic BSs have been recommended for use in patients with distal UMBO because they remain patent for longer than uncovered stents, placement of a covered metallic BS for proximal UMBO is contraindicated because of the risk of occlusion at the branched intrahepatic bile duct. Therefore, uncovered metallic BSs are selected for biliary decompression in patients with proximal UMBO. However, there is still no broad consensus regarding the indications for use of these stents. Although there are various strategies for palliation of UMBO, it would be valuable to obtain indicators of prognosis and BS patency that would not only permit accurate stratification of patients but also improve clinical decision-making, thus leading to the development of rational treatment strategies for patients with UMBO.
It has been very difficult to predict outcome and BS patency in patients with UMBO using conventional clinical parameters such as tumor markers. However, a number of studies have investigated the Glasgow prognostic score (GPS), an inflammation-based prognostic scoring system that takes into account only the serum levels of C-reactive protein (CRP) and albumin for prognostication of patients with various types of advanced cancer (9-16). Moreover, a modified version of the GPS (mGPS) has recently been used for prognostication of mortality in patients with colorectal cancer (17-20). Recently, we reported that the mGPS was clearly able to predict outcome after intraoperative placement of a BS for UMBO (21). Therefore, we hypothesized that the mGPS might also be able to predict BS patency in this context. Therefore we conducted the present study to evaluate the usefulness of the mGPS for prediction of BS patency in such patients.
Patients and Methods
Between April 2001 and December 2012, 66 patients underwent intraoperative placement of BS for UMBO diagnosed by surgical exploration at Dokkyo Medical University Hospital. None of the patients underwent surgical biliary bypass including choledochojejeunostomy or cholecystojejunostomy. For the purposes of the present study, 63 patients who underwent intraoperative placement of an uncovered expandable metallic BS were enrolled. The remaining three patients underwent intraoperative placement of a plastic BS or a covered expandable metallic BS, and were therefore excluded from the study because of the different nature of the stents employed. Intraoperative placement of BS was performed in patients with UMBO. All such inoperable lesions were initially diagnosed during surgical exploration, and the methods and indications for intraoperative placement of BS were outlined in our previous article (4).
Routine laboratory measurements including hemoglobin (Hb), the serum levels of total bilirubin (T-Bil), alkaline phosphatase (ALP), CRP, and albumin, and tumor markers such as carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) were carried-out on the initial day to exclude any inflammatory effect of preoperative biliary drainage or cholangiography. No patients received preoperative chemotherapy or irradiation.
The GPS was estimated as described previously (9, 10). Patients with both an elevated CRP level (>1.0 mg/dl) and hypoalbuminemia (<3.5 g/dl) were allocated a score of 2. Patients with only one of these biochemical abnormalities were allocated a score of 1, and patients with neither were allocated a score of 0. Recently, the GPS has been modified based on evidence that hypoalbuminemia in patients without an elevated CRP concentration has no significant association with cancer-specific survival. Therefore, patients with an elevated CRP were assigned a mGPS of 1 or 2 depending on the absence or presence of hypoalbuminemia (17-20), and mGPS was defined as follows: mGPS 0, CRP <1.0 mg/dl; mGPS 1, CRP >1.0 mg/dl; mGPS 2, CRP >1.0 mg/dl and albumin <3.5 g/dl.
BS occlusion was diagnosed on the basis of ultrasonographic findings such as intra-hepatic bile duct dilation or clinical signs of prolonged cholangitis with an increased serum bilirubin level. Liver failure with jaundice or uncontrollable ascites was excluded from BS occlusion. BS patency was defined as the period from BS insertion to occlusion or death in patients without BS occlusion. We retrospectively reviewed a database for the 63 patients to evaluate the usefulness of the mGPS for prediction of BS occlusion, and the patients were divided into two groups (mGPS 0 vs. mGPS 1 and mGPS 2) for comparison of BS patency.
Data are expressed as the mean with the 95% confidence interval or median with range. Differences between the two groups divided according to the mGPS (mGPS 0 and mGPS 1, 2) were analyzed using the Mann–Whitney U-test or chi-squared test. The Cox proportional hazards model was used to identify valuable predictors of BS occlusion. Kaplan–Meier analysis and log-rank test were used to compare BS patency in patients divided according to the mGPS. Statistical analyses were performed using the SPSS statistics software package version 16.0 (SPSS Inc., Chicago, IL, USA) at a significance level of p<0.05.
Results
The study included 35 men and 28 women with a median age of 68 years (range=48-85 years). The mGPS was 0 in 37 patients, 1 in seven patients, and 2 in 19 patients. One patient had a pancreatic endocrine tumor, 29 had pancreatic cancer, 17 had cholangiocarcinoma, 8 had gallbladder cancer, 1 had hepatocellular carcinoma, 6 had cholangiocellular carcinoma, and one had duodenal cancer. Preoperative biliary drainage was performed in 57 patients. The distal end of the expandable metallic BS was passed through the papilla of Vater in 31 cases because biliary stricture was located nearby. Expandable metallic BS-in-expandable metallic BS placement was performed in 11 cases because of the presence of a long biliary stricture or tumor involvement of both the right and left hepatic bile ducts. The mean operation time for the palliative procedures overall was 250 min, and the mean estimated intraoperative blood loss was 171 ml.
The BS patency outcomes for the 63 patients who underwent intraoperative placement of BS are shown in Table I. There were no significant differences in the period of BS patency when the patients were separated on the basis of individual clinical characteristics, which included age (<70/ >70 years), gender (male/female), American Society of Anesthesiologists (ASA) grade (1, 2/3), presence of pancreatic cancer (yes/no), use of preoperative biliary drainage (yes/no), site of biliary stricture (upper/lower), Hb, T-Bil, ALP, albumin, CEA, CA19-9, reason for unresectability (local invasion/distant metastasis) and stent location on the distal side (bile duct/duodenum). However, siginificant differences were observed for CRP, mGPS (0/1/2), stent-in-stent placement (yes/no) and use of postoperative chemotherapy (yes/no) (22). Gemcitabine was administered postoperatively in most patients (31/37 cases, 84%).
Table II shows the relationships between clinical background characteristics and mGPS in patients undergoing intraoperative placement of BS for UMBO. Although age, gender, ASA grade, pancreatic cancer, preoperative biliary drainage, site of biliary stricture, reason for unresectability, stent location on the distal side and use of chemotherapy did not differ significantly between the two mGPS groups, stent-in-stent placement was performed more frequently in patients with mGPS 1 or 2 than in those with mGPS 0.
Table III shows the relationships between clinicolaboratory characteristics and mGPS. The median serum level of Hb in patients with mGPS 1 or 2 was significantly lower than in those with mGPS 0. The median serum level of T-Bil in patients with mGPS 1 or 2 was significantly higher than in those with mGPS 0. The median period of BS patency in patients with mGPS 0 was significantly longer than in those with mGPS 1 or 2. However, there were no significant differences according to mGPS for clinicolaboratory characteristics such as age, ALP, CEA and CA19-9.
The mean period of BS patency was 242 days. The overall 6-month, 1-year, and 2-year BS patency rates were 66.3%, 44.7% and 23.8%, respectively. Kaplan–Meier analysis and log-rank test demonstrated significant differences in BS patency rate between patients with a mGPS of 0 and those with a mGPS of 1 and 2 (p=0.014) (Figure 1). The 6-month and 1-year BS patency rates in patients with mGPS of 0 were 78.3% and 57.0%, and those in patients with a mGPS of 1 and 2 were 53.3% and 23.7%, respectively. During the period of observation, 27 patients (42.9%) were diagnosed as having BS occlusion after intraoperative BS placement. Among these patients, 14 (37.8%) had a mGPS of 0 and 13 (50.0%) had a mGPS of 1 or 2.
Univariate analysis revealed that an elevated serum level of T-Bil, a decreased serum level of albumin, stent-in-stent placement and a higher mGPS were associated with BS occlusion (Table IV). Multivariate analysis using the above five selected factors demonstrated that none were significantly associated with BS occlusion (Table IV).
Discussion
The GPS has become established as a predictor of survival in patients with various types of cancers since it was first proposed by Forrest et al. on the basis of a cohort study of patients with inoperable non-small cell lung cancer in 2004 (9, 10). Although many studies have investigated the GPS for prognostication in patients with various types of cancer, none have addressed the relationship between the GPS and BS patency in patients with UMBO (23). To our knowledge, this is the first retrospective study to have done so. We found that the mGPS was useful for predicting postoperative BS patency in patients undergoing intraoperative BS placement for UMBO on the basis of Kaplan–Meier analysis and univariate analysis using the Cox proportional hazards model.
In most retrospective studies, the clinical backgrounds of the enrolled patients differed. In this study, however, the patients' backgrounds were almost uniform, there being no significant differences between the mGPS groups and the clinical background characteristics (Table II), despite the retrospective design. Although multivariate analysis revealed that the mGPS was not a significant independent risk factor for BS occlusion, the mGPS appeared to be a useful indicator for stratification of patients according to BS patency.
There are several strategies for palliative biliary decompression and relief of jaundice in patients with UMBO, including surgical biliary bypass, endoscopic retrograde biliary drainage using a plastic or metallic BS, and external biliary drainage including percutaneous transhepatic biliary drainage and percutaneous transhepatic gallbladder drainage. Among them, endoscopic retrograde biliary drainage and surgical biliary bypass have become standard procedures (24, 25). Therefore, it is important to predict patient outcome and BS patency when performing rational palliative therapy for biliary decompression. Although the initial cost of metallic BS placement is higher than that for plastic BS, the period of patency is longer (25). Therefore, use of a metallic BS reduces the need for additional procedures such as percutaneous transhepatic or endoscopic retrograde biliary drainage for treatment of stent occlusion after initial stent placement (25). Recent cost-effectiveness analyses have revealed that patients receiving metallic BS with a survival period of more than four to six months have greater economic benefit than patients receiving plastic BS (7, 8, 26). Therefore, in terms of economy, plastic BS placement may be more advantageous for UMBO in patients who are considered unlikely to survive for four months.
Accurate prediction of the period of BS patency using conventional preoperative clinical parameters is very difficult in patients with UMBO. In the present study, however, the mGPS was able to clearly divide the patients into two groups, and a higher mGPS was associated with a shorter period of expandable metallic BS patency in patients who received intraoperative BS placement. Therefore, the mGPS may be a feasible classification for deciding on the type of palliative treatment for biliary decompression in patients with UMBO. For example, metallic BS placement would seem to be suitable for biliary decompression in patients with mGPS 0, because such patients had a better outcome than patients with mGPS 1 or 2 in terms of BS patency rate. On the other hand, placement of a plastic BS would seem to be suitable for biliary decompression in patients with mGPS 2, because in this study the median period of patency of expandable metallic BSs in patients with mGPS 2 was less than four months (Table III). Although it is difficult to decide the indications for biliary decompression on the basis of mGPS alone, it may be helpful for stratification and decision making in patients with UMBO even if they have distant metastasis.
Both CRP and albumin are laboratory parameters that are much cheaper and faster to determine than conventional tumor markers such as serum CEA and CA19-9. Therefore, the mGPS can be used for routine evaluation on initial admission because of its low cost and great convenience. On initial admission, the Hb level was significantly lower and T-Bil was significantly increased in patients with mGPS 1, 2 (Table III). It is speculated that these phenomena could represent the presence of cancer-induced cachexia in such patients. Patients with mGPS 1 or 2 may have more advanced disease or severe UMBO in comparison with patients with mGPS 0, because the former had a higher T-Bil level.
In conclusion, this retrospective study has demonstrated that the mGPS might be applicable for routine preoperative clinical assessment and treatment planning in patients with UMBO. Because a preoperatively high mGPS is associated with short BS patency, the mGPS would be a feasible and helpful tool for decision making about treatment for biliary decompression in patients with UMBO.
Footnotes
-
Conflicts of interest
None.
-
Grant Support
We received no grant support for this study.
- Received March 1, 2014.
- Revision received May 7, 2014.
- Accepted May 8, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved