J Gastrointest Surg (2014) 18:1987–1993 DOI 10.1007/s11605-014-2657-3
ORIGINAL ARTICLE
Strong YB-1 Expression Predicts Liver Recurrence Following Resection for Colorectal Metastases Francesco Ardito & Vincenzo Arena & Maria Vellone & Gennaro Grande & Ilaria Pennacchia & Francesco Majellaro & Ivo Giovannini & Fabio M. Vecchio & Gennaro Nuzzo & Felice Giuliante
Received: 16 May 2014 / Accepted: 2 September 2014 / Published online: 18 September 2014 # 2014 The Society for Surgery of the Alimentary Tract
Abstract Introduction The Y-box binding protein-1 (YB-1) is a multifunctional oncoprotein involved in the proliferation and aggressiveness of cancer cells. The aim of this study was to determine whether strong YB-1 expression in neoplastic cells of colorectal liver metastases (CRLM) may have an impact on liver disease-free survival following liver resection. Materials and Methods Immunohistochemistry was performed to evaluate YB-1 in 66 patients who underwent liver resection for CRLM. YB-1 expression was classified as weak (low-staining intensity) and strong (high-staining intensity). Results YB-1 expression was observed in the cytoplasm of all CRLM. YB-1 expression was weak in 17 patients (25.8 %) and strong in 49 patients (74.2 %). Liver recurrence rate was significantly higher in the strong than in the weak expression group: 55.1 vs. 23.5 % (p=0.023). Multivariable logistic regression analysis showed that YB-1 strong expression was the only independent risk factor for liver recurrence. The 5-year specific liver disease-free survival rate was 76.0 % in the weak expression group and 41.5 % in the strong expression group (p=0.034). These results were not influenced by clinical prognostic factors of tumor recurrence. Conclusions This is the first study showing that the degree of YB-1 expression in tissue specimens of CRLM predicts liver recurrence following liver resection. Keywords Colorectal liver metastases . Liver recurrence . YB-1 . Liver resection . Molecular markers
Introduction Resection of colorectal liver metastases (CRLM) is widely accepted as the only potentially curative treatment that can offer a chance of long-term survival. However, despite recent improvements in the multidisciplinary treatment of CRLM, Preliminary data have been presented at the 10th Congress of the European-African Hepato-Pancreato-Biliary Association (May 29–31, 2013; Belgrade, Serbia). F. Ardito (*) : M. Vellone : G. Grande : I. Giovannini : G. Nuzzo : F. Giuliante Hepatobiliary Surgery Unit, Department of Surgery, School of Medicine, Catholic University of the Sacred Heart, L.go A. Gemelli, 8, I-00168 Rome, Italy e-mail: [email protected] V. Arena : I. Pennacchia : F. Majellaro : F. M. Vecchio Department of Pathology, School of Medicine, Catholic University of the Sacred Heart, Rome, Italy
the 5-year overall survival following liver resection still differs – in the reported surgical series, ranging from 351 3 to more than – 50 %.4 6 Moreover, tumor recurrence following liver resection still remains an oncologic problem which is reported in about half of resected patients, usually occurring within the first 3 years after resection.7 The possibility to accurately predict the outcome and the risk of tumor recurrence following liver resection for CRLM is fundamental in order to stratify and select patients to perioperative chemotherapy and to liver resection. Historically, prognosis after liver resection has been assessed according to preoperative clinical factors and in particular by prognostic – scoring systems.8 10 However, several studies demonstrated a poor correlation between these scoring systems and survival. Their clinical utility to the general population still remains , controversial, showing a limited external validation.11 12 Variations observed in the overall and disease-free survival of patients with similar prognostic scores suggest that other factors may play a role in determining outcome, such as biologic and molecular markers which may better correlate with patient-specific tumor biology.13 Indeed, the most important limit of traditional prognostic factors is that they evaluate
1988
the disease just at diagnosis, without providing any information about tumor biology. Recently, there has been an increasing interest in the use of biologic and molecular markers which are expected to better stratify prognosis after liver , resection for CRLM.13 14 The Y-box binding protein-1 (YB-1) is a multifunctional protein which is involved in the oncogenesis, proliferation, aggressiveness, and drug resistance of cancer cells.15 Elevated YB-1 protein levels are correlated with cancer progression and poor prognosis because it can function as an oncoprotein which, when present at elevated levels, leads to increased tumor cell proliferation. Correlation between YB-1 expression and prognosis has been evaluated in various tumors (breast cancer, 16 ovarian cancer,17 non-small cell lung cancer,18 gastric cancer, 19 and prostate cancer.20) To our knowledge, no studies have evaluated the prognostic impact of YB-1 expression in tissue specimens of CRLM. The aim of this study was to evaluate the impact of strong YB-1 expression on liver disease-free survival following liver resection in patients with CRLM.
Materials and Methods Between January 2004 and December 2010, 318 liver resections were performed with curative intent for CRLM at our unit. These patients were identified from a prospective database established in our unit from January 1987 for all consecutive admissions related to possible liver resection. A simple random sample was performed on these 318 liver resections, in order to collect a minimum sample of four patients for each year. Only patients with liver-only colorectal metastases were included. Patients with extrahepatic disease and patients who underwent liver resection combined with radiofrequency ablation were excluded from this analysis. Sixty-six patients were identified by the simple random sample, and they are the object of this study. Before surgery, all patients underwent physical examination and had their history taken. Routine blood tests and measurement of serum tumor marker (CEA) were routinely performed. Preoperative imaging assessment included abdominal ultrasound and abdominal computed tomography (CT) scan or magnetic resonance, and CT scan of the chest. Liver Resections Liver resections were defined according to the International Hepato-Pancreato-Biliary Association terminology. 21 Resection of three or greater than three liver segments was classified as major hepatectomy. The surgical technique for , liver resection used in our unit was previously described.22 23 Intraoperative ultrasonography was usually performed to
J Gastrointest Surg (2014) 18:1987–1993
assess the number of CRLM and their relationship with major vascular structures and to guide parenchymal transection. Immunohistochemistry Cellular YB-1 expression was determined by immunohistochemistry in the surgical specimens of CRLM of the 66 patients. All samples were routinely fixed for one night in 4 % buffered formalin and embedded in paraffin. Sections were cut 4 μm thick and were tested to the biotin-avidin immunoperoxidase reaction using the labeled streptavidin biotin (LSAB) method for YB-1 [YB1 Monocl Rabbit (EP2708Y) UCS; Morlupo, Italy]. In detail, sections for immunohistochemistry were prepared with 3-aminopropyltriethoxysilane (Sigma Chemicals; Milan, Italy) and dried for one night at 37 °C in order to ensure proper adhesion. Then, they were deparaffinized, rehydrated, and immersed in 0.3 % H2O2 with methanol for 10 min in order to inhibit endogenous peroxidase and non-specific binding. Sections were also heated in 0.5 % Triton-x (PBS) by microwaving for 10 min and then cooled at room temperature for 20 min for antigen retrieval. Endogenous biotin had been saturated by a biotin blocking kit (Vector Laboratories, Burlingame, CA). Sections were incubated at room temperature for 30 min with YB-1 antibody (dilution 1:100). The binding was visualized with the use of the biotinylated antibody (secondary antibody). The streptavidin-biotin peroxidase complex was developed with diaminobenzidine. Finally, the slides have been counterstained with hematoxylin. Negative controls were obtained by omitting the primary antibody. Slides were analyzed by two observers who were blinded to the clinical data. Each specimen was scored according to the presence and the intensity of staining within the tumor. According to the intensity of staining within the tumor, each specimen was given a score of 1 (weak) or 2–3 (strong) in case of weak or moderate-to-strong positivity, respectively.
Statistical Analysis The following data were collected for each patient: demographics; site of primary tumor; primary tumor nodal involvement; size, number, and distribution of CRLM; type of CRLM (synchronous or metachronous); use of preoperative chemotherapy; grade of YB-1 expression. The clinical risk score (MSKCC risk score) developed by Fong et al.10 was used to compare patients. Operative details included type of liver resection, rate of intraoperative blood transfusions, and radicality of liver resection. When the surgical-free margin was 0 mm, or there was exposed tumor along the transection plane, liver resection was classified as R1-resection.
J Gastrointest Surg (2014) 18:1987–1993
Early results included postoperative morbidity and 60-day mortality. Late results included liver recurrence rate and the 5-year liver disease-free survival rate which was assessed at first metastatic liver relapse following liver resection. Differences between subgroups were tested using the Pearson chi-square test. A multivariable logistic regression model was then built to estimate the odds of liver recurrence following the surgical procedure. Variables significant at the 0.2 level in the univariate analysis were used in model building. The final model included only variables significant at the 0.05 level at the multivariable analysis. The results of the logistic regression model are expressed as odds ratios with 95 % confidence intervals. The Kaplan-Meier method was used to analyze actuarial liver disease-free survival. Statistical analysis was performed using the SPSS software for Windows version 13.0 (SPSS, Chicago, IL, USA).
Results Sixty-six patients were included in the study. There were 44 men (66.7 %) and 22 women (33.3 %). The mean age of the patients was 66±10 years (range 41–80). Site of the primary tumor was colon in 40 patients (60.6 %) and rectum in 26 (39.4 %). Lymph nodes of the primary tumor were involved in 41 patients (62.1 %). The mean preoperative CEA level was 170 ng/mL (range 1.0–3423.5). Thirty-eight patients (57.6 %) had synchronous hepatic metastases. Thirty-four hepatectomies (51.5 %) were performed for multiple metastases. Mean size of CRLM was 4.2 cm (range 1.0–14.0). Distribution of CRLM was bilobar in 18 patients (27.3 %). The MSKCC risk score was
ORIGINAL ARTICLE
Strong YB-1 Expression Predicts Liver Recurrence Following Resection for Colorectal Metastases Francesco Ardito & Vincenzo Arena & Maria Vellone & Gennaro Grande & Ilaria Pennacchia & Francesco Majellaro & Ivo Giovannini & Fabio M. Vecchio & Gennaro Nuzzo & Felice Giuliante
Received: 16 May 2014 / Accepted: 2 September 2014 / Published online: 18 September 2014 # 2014 The Society for Surgery of the Alimentary Tract
Abstract Introduction The Y-box binding protein-1 (YB-1) is a multifunctional oncoprotein involved in the proliferation and aggressiveness of cancer cells. The aim of this study was to determine whether strong YB-1 expression in neoplastic cells of colorectal liver metastases (CRLM) may have an impact on liver disease-free survival following liver resection. Materials and Methods Immunohistochemistry was performed to evaluate YB-1 in 66 patients who underwent liver resection for CRLM. YB-1 expression was classified as weak (low-staining intensity) and strong (high-staining intensity). Results YB-1 expression was observed in the cytoplasm of all CRLM. YB-1 expression was weak in 17 patients (25.8 %) and strong in 49 patients (74.2 %). Liver recurrence rate was significantly higher in the strong than in the weak expression group: 55.1 vs. 23.5 % (p=0.023). Multivariable logistic regression analysis showed that YB-1 strong expression was the only independent risk factor for liver recurrence. The 5-year specific liver disease-free survival rate was 76.0 % in the weak expression group and 41.5 % in the strong expression group (p=0.034). These results were not influenced by clinical prognostic factors of tumor recurrence. Conclusions This is the first study showing that the degree of YB-1 expression in tissue specimens of CRLM predicts liver recurrence following liver resection. Keywords Colorectal liver metastases . Liver recurrence . YB-1 . Liver resection . Molecular markers
Introduction Resection of colorectal liver metastases (CRLM) is widely accepted as the only potentially curative treatment that can offer a chance of long-term survival. However, despite recent improvements in the multidisciplinary treatment of CRLM, Preliminary data have been presented at the 10th Congress of the European-African Hepato-Pancreato-Biliary Association (May 29–31, 2013; Belgrade, Serbia). F. Ardito (*) : M. Vellone : G. Grande : I. Giovannini : G. Nuzzo : F. Giuliante Hepatobiliary Surgery Unit, Department of Surgery, School of Medicine, Catholic University of the Sacred Heart, L.go A. Gemelli, 8, I-00168 Rome, Italy e-mail: [email protected] V. Arena : I. Pennacchia : F. Majellaro : F. M. Vecchio Department of Pathology, School of Medicine, Catholic University of the Sacred Heart, Rome, Italy
the 5-year overall survival following liver resection still differs – in the reported surgical series, ranging from 351 3 to more than – 50 %.4 6 Moreover, tumor recurrence following liver resection still remains an oncologic problem which is reported in about half of resected patients, usually occurring within the first 3 years after resection.7 The possibility to accurately predict the outcome and the risk of tumor recurrence following liver resection for CRLM is fundamental in order to stratify and select patients to perioperative chemotherapy and to liver resection. Historically, prognosis after liver resection has been assessed according to preoperative clinical factors and in particular by prognostic – scoring systems.8 10 However, several studies demonstrated a poor correlation between these scoring systems and survival. Their clinical utility to the general population still remains , controversial, showing a limited external validation.11 12 Variations observed in the overall and disease-free survival of patients with similar prognostic scores suggest that other factors may play a role in determining outcome, such as biologic and molecular markers which may better correlate with patient-specific tumor biology.13 Indeed, the most important limit of traditional prognostic factors is that they evaluate
1988
the disease just at diagnosis, without providing any information about tumor biology. Recently, there has been an increasing interest in the use of biologic and molecular markers which are expected to better stratify prognosis after liver , resection for CRLM.13 14 The Y-box binding protein-1 (YB-1) is a multifunctional protein which is involved in the oncogenesis, proliferation, aggressiveness, and drug resistance of cancer cells.15 Elevated YB-1 protein levels are correlated with cancer progression and poor prognosis because it can function as an oncoprotein which, when present at elevated levels, leads to increased tumor cell proliferation. Correlation between YB-1 expression and prognosis has been evaluated in various tumors (breast cancer, 16 ovarian cancer,17 non-small cell lung cancer,18 gastric cancer, 19 and prostate cancer.20) To our knowledge, no studies have evaluated the prognostic impact of YB-1 expression in tissue specimens of CRLM. The aim of this study was to evaluate the impact of strong YB-1 expression on liver disease-free survival following liver resection in patients with CRLM.
Materials and Methods Between January 2004 and December 2010, 318 liver resections were performed with curative intent for CRLM at our unit. These patients were identified from a prospective database established in our unit from January 1987 for all consecutive admissions related to possible liver resection. A simple random sample was performed on these 318 liver resections, in order to collect a minimum sample of four patients for each year. Only patients with liver-only colorectal metastases were included. Patients with extrahepatic disease and patients who underwent liver resection combined with radiofrequency ablation were excluded from this analysis. Sixty-six patients were identified by the simple random sample, and they are the object of this study. Before surgery, all patients underwent physical examination and had their history taken. Routine blood tests and measurement of serum tumor marker (CEA) were routinely performed. Preoperative imaging assessment included abdominal ultrasound and abdominal computed tomography (CT) scan or magnetic resonance, and CT scan of the chest. Liver Resections Liver resections were defined according to the International Hepato-Pancreato-Biliary Association terminology. 21 Resection of three or greater than three liver segments was classified as major hepatectomy. The surgical technique for , liver resection used in our unit was previously described.22 23 Intraoperative ultrasonography was usually performed to
J Gastrointest Surg (2014) 18:1987–1993
assess the number of CRLM and their relationship with major vascular structures and to guide parenchymal transection. Immunohistochemistry Cellular YB-1 expression was determined by immunohistochemistry in the surgical specimens of CRLM of the 66 patients. All samples were routinely fixed for one night in 4 % buffered formalin and embedded in paraffin. Sections were cut 4 μm thick and were tested to the biotin-avidin immunoperoxidase reaction using the labeled streptavidin biotin (LSAB) method for YB-1 [YB1 Monocl Rabbit (EP2708Y) UCS; Morlupo, Italy]. In detail, sections for immunohistochemistry were prepared with 3-aminopropyltriethoxysilane (Sigma Chemicals; Milan, Italy) and dried for one night at 37 °C in order to ensure proper adhesion. Then, they were deparaffinized, rehydrated, and immersed in 0.3 % H2O2 with methanol for 10 min in order to inhibit endogenous peroxidase and non-specific binding. Sections were also heated in 0.5 % Triton-x (PBS) by microwaving for 10 min and then cooled at room temperature for 20 min for antigen retrieval. Endogenous biotin had been saturated by a biotin blocking kit (Vector Laboratories, Burlingame, CA). Sections were incubated at room temperature for 30 min with YB-1 antibody (dilution 1:100). The binding was visualized with the use of the biotinylated antibody (secondary antibody). The streptavidin-biotin peroxidase complex was developed with diaminobenzidine. Finally, the slides have been counterstained with hematoxylin. Negative controls were obtained by omitting the primary antibody. Slides were analyzed by two observers who were blinded to the clinical data. Each specimen was scored according to the presence and the intensity of staining within the tumor. According to the intensity of staining within the tumor, each specimen was given a score of 1 (weak) or 2–3 (strong) in case of weak or moderate-to-strong positivity, respectively.
Statistical Analysis The following data were collected for each patient: demographics; site of primary tumor; primary tumor nodal involvement; size, number, and distribution of CRLM; type of CRLM (synchronous or metachronous); use of preoperative chemotherapy; grade of YB-1 expression. The clinical risk score (MSKCC risk score) developed by Fong et al.10 was used to compare patients. Operative details included type of liver resection, rate of intraoperative blood transfusions, and radicality of liver resection. When the surgical-free margin was 0 mm, or there was exposed tumor along the transection plane, liver resection was classified as R1-resection.
J Gastrointest Surg (2014) 18:1987–1993
Early results included postoperative morbidity and 60-day mortality. Late results included liver recurrence rate and the 5-year liver disease-free survival rate which was assessed at first metastatic liver relapse following liver resection. Differences between subgroups were tested using the Pearson chi-square test. A multivariable logistic regression model was then built to estimate the odds of liver recurrence following the surgical procedure. Variables significant at the 0.2 level in the univariate analysis were used in model building. The final model included only variables significant at the 0.05 level at the multivariable analysis. The results of the logistic regression model are expressed as odds ratios with 95 % confidence intervals. The Kaplan-Meier method was used to analyze actuarial liver disease-free survival. Statistical analysis was performed using the SPSS software for Windows version 13.0 (SPSS, Chicago, IL, USA).
Results Sixty-six patients were included in the study. There were 44 men (66.7 %) and 22 women (33.3 %). The mean age of the patients was 66±10 years (range 41–80). Site of the primary tumor was colon in 40 patients (60.6 %) and rectum in 26 (39.4 %). Lymph nodes of the primary tumor were involved in 41 patients (62.1 %). The mean preoperative CEA level was 170 ng/mL (range 1.0–3423.5). Thirty-eight patients (57.6 %) had synchronous hepatic metastases. Thirty-four hepatectomies (51.5 %) were performed for multiple metastases. Mean size of CRLM was 4.2 cm (range 1.0–14.0). Distribution of CRLM was bilobar in 18 patients (27.3 %). The MSKCC risk score was
