Abstract
Background: Systemic inflammation before surgery, as evidenced by the Glasgow prognostic score (GPS), predicts postoperative complications and cancer-specific survival in various types of cancer. The aim of this study was to evaluate the significance GPS in hepatic resection for hepatocellular carcinoma (HCC). Patients and Methods: Sixty-six patients who underwent elective hepatic resections for HCC were include in the study. Patients were classified into three groups: GPS 0 [C-reactive protein (CRP)≤1.0 mg/dl and serum albumin ≥3.5 g/dl, n=54], GPS 1 [CRP >1.0 mg/dl or serum albumin <3.5 g/dl, n=11], and GPS 2 [CRP>1.0 mg/dl and serum albumin <3.5 g/dl, n=1]. We retrospectively examined the association between GPS (0 or 1) and perioperative clinical variables and outcome. Results: In univariate analysis, GPS 0 patients had significantly better preoperative the retention rate of indocyanine green at 15 minutes (ICGR15) (p=0.0418), Child-Pugh classification (p=0.0075) and model for end-stage liver disease score (p=0.0007) than did GPS 1 patients. In multivariate analysis, blood loss and GPS 1 were independent risk factors for pulmonary complications (p=0.0118 for blood loss, p=0.0143 for GPS 1), red blood cell concentration transfusion (p=0.0036 for blood loss, p=0.0117 for GPS 1) and flesh frozen plasma transfusion (p=0.0020 for blood loss, p=0.0044 for GPS 1). Albumin product transfusion, duration of operation (p=0.0478), blood loss (p=0.0420) and GPS 1 (p=0.0111) were independent risk factors. Disease-free and overall survival of GPS 0 and GPS 1 patients were comparable. Conclusion: GPS reflects preoperative patient status, and is associated with blood transfusion and pulmonary complications in elective hepatic resection for HCC.
Hepatocellular carcinoma (HCC) is one of the most common types of malignant cancer in the world (1), for which hepatic resection or transplantation remain the most effective treatments (2). In spite of improvements in surgical techniques, instruments, and perioperative management, the morbidity rate from hepatic resection has been reported as being as high as 40% (3). Although tumor stage has been the main basis for predicting long-term survival in patients undergoing surgery, it is now recognized that postoperative complications contribute to poor cancer-specific survival for various types of cancer (4, 5), including HCC (6). It has been postulated that the presence of systemic inflammatory response, as evidenced by an elevated C-reactive protein (CRP) concentration, may be related to poor cancer-specific survival, independently of tumor stage (5, 7). Moreover, the occurrence of postoperative complications such as pulmonary complications, surgical site infection (SSI) and bile leakage leads to prolonged hospital stay. Therefore, it is important to clarify the patient group which has an increased risk for postoperative complications and HCC recurrence.
Several recent studies have indicated that the measurement of the systemic inflammatory response by the combination of serum CRP and albumin concentrations, i.e. the Glasgow prognostic score (GPS), have been shown to predict cancer-specific survival (8, 9), and postoperative infectious complications (10). However, to the best of our knowledge, the relationship between the GPS and postoperative complications, disease-free survival, as well as overall survival, in patients with HCC undergoing hepatic resection has not been reported.
Univariate analysis of perioperative findings according to Glasgow prognostic score.
We hypothesized that the GPS might reflect preoperative general status and predict perioperative conditions such as blood product transfusion requirement, postoperative complications and prognosis in patients who underwent elective hepatic resection for HCC.
Patients and Methods
Between January 2001 and December 2006, 84 patients underwent hepatic resection for HCC in the Department of Surgery, Jikei University Hospital, Tokyo, Japan. Of these, 18 patients were excluded, consisting of 4 patients with concomitant microwave coagulation or radiofrequency ablation therapy, 5 patients with additional procedures for other malignancies, 8 patients due to incomplete data sets, and 1 patient lost to follow-up, leaving the remaining 66 patients for this study.
Generally, the extent of hepatic resection was determined based on the retention rate of indocyanine green at 15 minutes (ICGR15) before surgery and hepatic reserve, as described by Miyagawa et al. (11). Hemogram and chemistry profile were routinely measured for each patient preoperatively. The absolute white cell count (WBC), lymphocyte and each subset count were routinely determined in peripheral venous blood samples. The serum biochemistry data and hemogram included hemoglobin (Hb), serum total bilirubin (T-Bil), serum creatinine (Cr), serum albumin, prothrombin time-international normalized ratio (PT-INR), and CRP. The use of blood products and the amount were determined by the preference of attending surgeons based on Guidelines for Administration of Blood Products by the Japanese Ministry of Health and Welfare established in 1999 (12), as well as intraoperative blood loss, postoperative data of Hb, platelets, serum albumin, and the prothrombin time. The tumor factor (T factor) staging was based on the General Rules for the Clinical and Pathological Study of Primary Liver Cancer by the Liver Cancer Study Group of Japan (13). The type of resection was classified into two groups: anatomical resection (extended lobectomy, lobectomy, segmentectomy, or subsegmentectomy) and non-anatomical limited partial resection.
For the assessment of systemic inflammatory response using the GPS, patients were classified into three groups: patients with normal albumin (≥3.5 g/dl) and normal CRP (≤1.0 mg/dl) as GPS 0 (n=54), those with low albumin (<3.5 g/dl) or elevated CRP (>1.0 mg/dl) as GPS 1 (n=11), and both low albumin (<3.5 g/dl) and elevated CRP (>1.0 mg/dl) as GPS 2 (n=1). We then excluded GPS 2 and compared GPS0 with GPS1, using the following 25 factors: age; gender; pre-operative ICGR15; T factor based on tumor pathology; type of resection; duration of operation; intraoperative blood loss; hepatitis virus status; Child's classification; model for end-stage liver disease score (MELD score); preoperative T-Bil, PT-INR, Cr, Hb, WBC and lymphocyte count; surgical site infection (SSI); pulmonary complications; bile leakage; and the perioperative transfusion of red blood cell concentrate (RC), fresh frozen plasma (FFP), platelet concentrate (PC), or albumin product, and disease-free and overall survival by univariate and multivariate analyses.
Univariate analysis of perioperative factors according to preoperative CRP and albumin in GPS 1 patients.
The recurrence of HCC was defined as newly detected hypervascular hepatic or extrahepatic tumors by ultrasonography, computed tomography (CT), magnetic resonance image and angiography, with or without an increase in serum α-fetoprotein (AFP) or protein induced by vitamin K absence or antagonist-II (PIVKA-II). For recurrent HCC in the liver, repeated hepatic resection, local ablation therapy, or transarterial chemoembolization was given based on hepatic functional reserve judged mainly by ICGR15. Extrahepatic recurrence was mainly treated conservatively.
Pulmonary complications were defined as postoperative pneumonia; postoperative respiratory failure with pyrexia, dyspenia, and a pulmonary infiltrate on chest X-ray; or pleural effusion that required thoracentesis. SSI was defined as surgical wound infection with purulent discharge and bacterial isolation, or abdominal abscess with pyrexia. Bile leakage was defined as bile discharge from the abdominal drainage tube, or fluid collection by CT which was demonstrated as bile by aspiration or drainage.
This study was approved by the Ethics Committee of Jikei University School of Medicine.
Statistical analysis. The data were expressed as the mean±standard deviation (SD). Analysis of disease-free and overall survival was performed using the log-rank test. Univariate analysis was performed using the non-paired Student's t-test, Mann-Whitney's U t-test and Chi-square test. Multivariate analysis was performed using logisticregression analysis. All p-values were considered statistically significant when the associated probability was less than 0.05.
Results
Association between GPS and perioperative findings. Perioperative findings in GPS 0 and GPS 1 patients are outlined in Table I. Univariate analysis demonstrated that GPS 1 patients had significantly worse preoperative ICGR15 (p=0.0418), Child's classification (p=0.0075) and MELD score (p=0.0007), PT-INR (p=0.0299), Cr (p=0.0352) and Hb (p=0.0433) than those in GPS 0 patients. The incidence of postoperative pulmonary complications in GPS 1 was significantly higher than those in GPS 0 (p=0.0205). The incidences of SSI (p=0.2305) and bile leakage (p=0.1520) in both groups were comparable. The incidence of perioperative RC transfusion, FFP transfusion and albumin product transfusion in GPS 1 were significantly higher than those in GPS 0 (p=0.0115, p=0.0059, and p=0.0024, respectively). Perioperative factors of patients with elevated CRP and low albumin concentration in GPS1 are outlined in Table II. The type of resection for HCC was statistical difference. The incidence of SSI patients with elevated CRP tended to be higher than that in patients with low albumin, which however failed to achieve significance (p=0.1247). The incidence of RC transfusion in patients with elevated CRP was statistically higher than that in patients with low albumin (p=0.0221). The incidence of albumin product transfusion in patients with elevated CRP tended to be lower than that in patients with low albumin, which however failed to achieve statistical significance. Other findings were comparable between the two groups.
Kaplan-Meier curves of disease-free (A) and overall survival (B) in GPS 0 and GPS 1 patients. Both disease-free and overall survival of the two groups were comparable.
Disease-free and overall survival analysis. Figure 1 shows Kaplan-Meier curves of disease-free and overall survival in GPS 0 and GPS 1 patients. Disease-free survival rates at 1, 3, and 5 years were 79.5%, 52.4%, and 40.0% in GPS 0, 90.9%, 51.9%, and 51.9% in GPS 1, respectively (Figure 1A). Overall survival rates at 1, 3, and 5 years were 94.4%, 83.2%, and 71.0% in GPS 0, and 90.9%, 70.1%, and 70.1% in GPS 1, respectively (Figure 1B). Both disease-free and overall survival in GPS 0 and GPS 1 were comparable (p=0.5079 for disease-free survival, p=0.8272 for overall survival).
Assessment of GPS in relation to pulmonary complications and blood product transfusion. The association between incidence of pulmonary complications and perioperative factors including the GPS are outlined in Table III. In univariate analysis, female gender (p=0.0304), massive blood loss (p=0.0091) and GPS 1 (p=0.0114) were significant risk factors for postoperative pulmonary complications. In multivariate analysis, blood loss during operation (p=0.0118) and GPS 1 (p=0.0143) were found as independent risk factors.
The association between incidence of RC transfusion and perioperative factors including the GPS are outlined in Table IV. In univariate analysis, anatomical resection (p=0.0013), operation time (p=0.0008), intraoperative blood loss (p<0.0001), absence of hepatitis virus infection (p=0.0406), GPS 1 (p=0.0115) and preoperative Hb (p=0.0039) were significant risk factors for perioperative RC transfusion. In multivariate analysis, intraoperative blood loss (p=0.0036) and preoperative Hb (p=0.0278) were found as independent risk factors. The incidence of RC transfusion in GPS 1 patients tended to be higher than that in GPS1 patients, which however was not significant by multivariate analysis (p=0.0508).
The association between incidence of FFP transfusion and perioperative findings including the GPS are outlined in Table V. In univariate analysis, advanced T factor (p=0.0227), anatomical resection (p=0.0026), operation time (p=0.0173), intraoperative blood loss (p<0.0001) and GPS 1 (p=0.0059) were significant risk factors for perioperative FFP transfusion. In multivariate analysis, intraoperative blood loss (p=0.0020) and GPS 1 (p=0.0044) were found to be independent risk factors.
The association between the incidence of albumin product transfusion and perioperative findings including the GPS are outlined in Table VI. In univariate analysis, age (p=0.0461), operation time (p<0.0001), hepatitis virus status (p=0.0398), intra-operative blood loss (p<0.0001) and GPS 1 (p=0.0024) were significant risk factors for perioperative albumin product transfusion. In multivariate analysis, operation time (p=0.0478), intraoperative blood loss (p=0.0420) and GPS 1 (p=0.0111) were found to be independent risk factors.
Discussion
The GPS was first reported as a predictor of prognosis in inoperable non-small cell lung cancer in 2003 (14). The GPS had been shown to predict prognosis in patients with various inoperable tumors of the lung (15), breast (16), esophagus or stomach (17), pancreas (18), kidney (19) and colorectum (20). Therefore, the GPS may be a useful predictor of prognosis for patients with unresctable and advanced malignancies. The reasons for the association between the GPS, pre-treatment elevated CRP or low albumin concentrations, and prognosis, as well as postoperative complications, in patients with various malignancies remain unclear (10). It has been reported that pre-operative elevation of CRP was related to reduced lymphocyte numbers and suppressed lymphocyte-mediated immunity (21), and low tumor T-lymphocyte infiltration has been reported to induce tumor progression in patients undergoing curative resection for colorectal cancer (22). The systemic inflammatory response and hypoalbuminemia reflected the loss of lean tissue and protein, which suppress immune function (23, 24). Therefore, elevated CRP and low albumin concentration may reflect immunosuppression, and predict prognosis and postoperative complications in patients with various malignancies. Since the majority of patients with HCC have liver dysfunction due to hepatitis viral infection and liver cirrhosis, the GPS may be influenced by liver function. In this study, the GPS reflected preoperative, general status such as Child's classification and MELD score, and predicted the incidence of postoperative pulmonary complications and requirement for perioperative FFP and albumin products independently.
Univariate and multivariate analyses of pulmonary complications after hepatic resection.
Univariate and multivariate analyses of perioperative red blood cell concentrate given after hepatic resection.
Univariate and multivariate analyses of requirement for perioperative FFP transfusion after hepatic resection.
Univariate and multivariate analyses of perioperative findings in albumin product transfusion.
Hypoalbuminemia has been reported as one risk factor for perioperative blood transfusion, and affected disease-free as well as overall survival in patients who underwent elective hepatic resection for HCC (25). In this study, the GPS predicted the incidence of perioperative FFP and albumin product transfusion independently, but reasons for association inflammatory response and perioperative blood transfusion is still unclear. Several investigators have reported the negative impact of blood transfusion on outcome and prognosis after hepatic resection for HCC (26, 27). Further validation and assessment of the mechanisms by which the GPS affects the incidence of perioperative blood product transfusion may help to improve outcome and prognosis in patients undergoing hepatic resection for HCC.
The evidence of a significant association between the GPS and prognosis in primary operable cancer patients has been reported in colorectal (9), gastro-esophageal (28), urinary bladder (29), pancreatic (30), renal (31), and non-small cell lung (32) cancer. Several investigators have reported the impact of preoperative systemic inflammatory responses on the prognosis of patients with HCC (33-35). However, the relationship between the GPS, postoperative complications and disease-free, as well as overall, survival in patients with HCC undergoing hepatic resection has not been reported. In this study, however, GPS was not found to be associated with disease-free or overall survival.
Patient risk stratification using the GPS is easy because it is composed only of serum CRP and albumin concentration, which are routine examinations for perioperative patient management. The GPS is useful for identification of patients at risk.
Conclusion
The GPS reflects the preoperative patient status including ICGR15, Child's classification and MELD score, the need for intraoperative use of FFP and albumin, and postoperative pulmonary complications in patients undergoing hepatic resection for HCC.
- Received September 6, 2010.
- Revision received November 1, 2010.
- Accepted November 2, 2010.
- Copyright© 2010 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- Intrahepatic HCV RNA Level and Genotype 1 Independently Associate with Hepatic Reticulon 3 Expression
- A Novel Postoperative Inflammatory Score Predicts Postoperative Pancreatic Fistula After Pancreatic Resection
- Glasgow Prognostic Score Predicts Therapeutic Outcome after Pancreaticoduodenectomy for Carcinoma of the Ampulla of Vater
- Negative Impact of Surgical Site Infection on Long-term Outcomes after Hepatic Resection for Colorectal Liver Metastases
- Postoperative Peak Serum C-Reactive Protein Predicts Outcome of Hepatic Resection for Hepatocellular Carcinoma