European Journal of Cancer (2015) 51, 1389– 1404
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.ejcancer.com
Review
microRNAs with prognostic significance in pancreatic ductal adenocarcinoma: A meta-analysis Adam E. Frampton a,b,⇑,1, Jonathan Krell b,1, Nigel B. Jamieson c, Tamara M.H. Gall a, Elisa Giovannetti d, Niccola Funel e, Mireia Mato Prado b, Daniel Krell f, Nagy A. Habib a, Leandro Castellano b, Long R. Jiao a, Justin Stebbing b,⇑ a
HPB Surgical Unit, Division of Surgery, Dept. of Surgery & Cancer, Imperial College, Hammersmith Hospital Campus, Du Cane Road, London W12 0HS, UK b Division of Oncology, Dept. of Surgery & Cancer, Imperial Centre for Translational and Experimental Medicine (ICTEM), Imperial College, Hammersmith Hospital campus, Du Cane Road, London W12 0NN, UK c Academic Unit of Surgery, Faculty of Medicine, Glasgow Royal Infirmary, Alexandra Parade, University of Glasgow, G31 2ER, UK d Dept. of Medical Oncology, VU University Medical Center, Amsterdam, The Netherlands e Dept. of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Italy f Dept. of Academic Oncology, Royal Free Hospital, Pond Street, London NW3 2QG, UK Received 19 January 2015; received in revised form 6 April 2015; accepted 10 April 2015 Available online 19 May 2015
KEYWORDS microRNAs Pancreatic cancer Prognosis Meta-analysis miR-21 Biomarker
Abstract Background: Reports have described the prognostic relevance of microRNAs (miRNAs) in patients treated for pancreatic ductal adenocarcinoma (PDAC). However, many of these include small numbers of patients. To increase statistical power and improve translation, we performed a systematic review and meta-analysis to determine a pooled conclusion. We examined the impact of miRNAs on overall survival (OS) and disease-free survival (DFS) in PDAC. Methods: Eligible studies were identified and quality assessed using multiple search strategies (last search December 2014). Data were collected from studies correlating clinical outcomes with dysregulated tumoural or blood miRNAs. Studies were pooled, and combined hazard ratios (HRs) with 95% confidence intervals (CIs) were used to estimate strength of the associations. Results: Twenty studies involving 1525 patients treated for PDAC were included. After correcting for publication bias, OS was significantly shortened in patients with high tumoural miR-21 (adjusted HR = 2.48; 1.96–3.14). This result persisted when only studies adjusting for adjuvant chemotherapy were combined (adjusted HR = 2.72; 1.91–3.89). High miR-21 also predicted reduced DFS (adjusted HR = 3.08; 1.78–5.33). Similarly, we found significant
⇑ Corresponding authors at: Dept. of Surgery & Cancer, Imperial College, London, UK. 1
E-mail addresses: [email protected] (A.E. Frampton), [email protected] (J. Stebbing). The following authors contributed equally to this work: AEF and JK.
http://dx.doi.org/10.1016/j.ejca.2015.04.006 0959-8049/Ó 2015 Elsevier Ltd. All rights reserved.
1390
A.E. Frampton et al. / European Journal of Cancer 51 (2015) 1389–1404
adjusted HRs for poor OS for high miR-155, high miR-203, and low miR-34a; and unadjusted HRs for high miR-222 and high miR-10b. The small number of studies, limited number of miRNAs and paucity of multivariate analyses are the limitations of our study. Conclusions: This is the first rigorous pooled analysis assessing miRNAs as prognostic biomarkers in PDAC. Tumoural miR-21 overexpression emerged as an important predictor of poor prognosis after PDAC resection independent of other clinicopathologic factors, including adjuvant chemotherapy use. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Pancreatic ductal adenocarcinoma (PDAC) remains a leading cause of cancer-related deaths [1]. The majority of patients present with late-stage disease and even if surgical resection can be performed, most patients already have clinically undetectable micro-metastases and therefore recur quickly resulting in an extremely poor prognosis [2,3]. Currently, factors for predicting survival in PDAC are poorly defined. The traditional clinicopathologic factors, including tumour characteristics, remain the only significant and clinically used features able to predict survival after resection. However, there is often discordance between histological characteristics, tumour stage and survival [4,5]. Serum carbohydrate antigen 19-9 (CA 19-9) remains the only biomarker used in clinical practice to monitor disease progression during PDAC treatment [2], with elevated levels post-operatively correlated with resection margin positivity (R1) and development of distant metastases [6]. Advances in the molecular biology of PDAC have given hope for better disease management [7], and the identification of advantageous and/or susceptible tumour characteristics is crucial for personalising therapy and achieving longer survival. Accumulating evidence suggests that PDAC is characterised by marked genetic heterogeneity, with recent large genomic sequencing efforts demonstrating that, with the exception of a few driver mutations that currently cannot be effectively targeted (i.e. KRAS, TP53, SMAD4, CDKN2A), most mutations occur at a prevalence of
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.ejcancer.com
Review
microRNAs with prognostic significance in pancreatic ductal adenocarcinoma: A meta-analysis Adam E. Frampton a,b,⇑,1, Jonathan Krell b,1, Nigel B. Jamieson c, Tamara M.H. Gall a, Elisa Giovannetti d, Niccola Funel e, Mireia Mato Prado b, Daniel Krell f, Nagy A. Habib a, Leandro Castellano b, Long R. Jiao a, Justin Stebbing b,⇑ a
HPB Surgical Unit, Division of Surgery, Dept. of Surgery & Cancer, Imperial College, Hammersmith Hospital Campus, Du Cane Road, London W12 0HS, UK b Division of Oncology, Dept. of Surgery & Cancer, Imperial Centre for Translational and Experimental Medicine (ICTEM), Imperial College, Hammersmith Hospital campus, Du Cane Road, London W12 0NN, UK c Academic Unit of Surgery, Faculty of Medicine, Glasgow Royal Infirmary, Alexandra Parade, University of Glasgow, G31 2ER, UK d Dept. of Medical Oncology, VU University Medical Center, Amsterdam, The Netherlands e Dept. of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Italy f Dept. of Academic Oncology, Royal Free Hospital, Pond Street, London NW3 2QG, UK Received 19 January 2015; received in revised form 6 April 2015; accepted 10 April 2015 Available online 19 May 2015
KEYWORDS microRNAs Pancreatic cancer Prognosis Meta-analysis miR-21 Biomarker
Abstract Background: Reports have described the prognostic relevance of microRNAs (miRNAs) in patients treated for pancreatic ductal adenocarcinoma (PDAC). However, many of these include small numbers of patients. To increase statistical power and improve translation, we performed a systematic review and meta-analysis to determine a pooled conclusion. We examined the impact of miRNAs on overall survival (OS) and disease-free survival (DFS) in PDAC. Methods: Eligible studies were identified and quality assessed using multiple search strategies (last search December 2014). Data were collected from studies correlating clinical outcomes with dysregulated tumoural or blood miRNAs. Studies were pooled, and combined hazard ratios (HRs) with 95% confidence intervals (CIs) were used to estimate strength of the associations. Results: Twenty studies involving 1525 patients treated for PDAC were included. After correcting for publication bias, OS was significantly shortened in patients with high tumoural miR-21 (adjusted HR = 2.48; 1.96–3.14). This result persisted when only studies adjusting for adjuvant chemotherapy were combined (adjusted HR = 2.72; 1.91–3.89). High miR-21 also predicted reduced DFS (adjusted HR = 3.08; 1.78–5.33). Similarly, we found significant
⇑ Corresponding authors at: Dept. of Surgery & Cancer, Imperial College, London, UK. 1
E-mail addresses: [email protected] (A.E. Frampton), [email protected] (J. Stebbing). The following authors contributed equally to this work: AEF and JK.
http://dx.doi.org/10.1016/j.ejca.2015.04.006 0959-8049/Ó 2015 Elsevier Ltd. All rights reserved.
1390
A.E. Frampton et al. / European Journal of Cancer 51 (2015) 1389–1404
adjusted HRs for poor OS for high miR-155, high miR-203, and low miR-34a; and unadjusted HRs for high miR-222 and high miR-10b. The small number of studies, limited number of miRNAs and paucity of multivariate analyses are the limitations of our study. Conclusions: This is the first rigorous pooled analysis assessing miRNAs as prognostic biomarkers in PDAC. Tumoural miR-21 overexpression emerged as an important predictor of poor prognosis after PDAC resection independent of other clinicopathologic factors, including adjuvant chemotherapy use. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Pancreatic ductal adenocarcinoma (PDAC) remains a leading cause of cancer-related deaths [1]. The majority of patients present with late-stage disease and even if surgical resection can be performed, most patients already have clinically undetectable micro-metastases and therefore recur quickly resulting in an extremely poor prognosis [2,3]. Currently, factors for predicting survival in PDAC are poorly defined. The traditional clinicopathologic factors, including tumour characteristics, remain the only significant and clinically used features able to predict survival after resection. However, there is often discordance between histological characteristics, tumour stage and survival [4,5]. Serum carbohydrate antigen 19-9 (CA 19-9) remains the only biomarker used in clinical practice to monitor disease progression during PDAC treatment [2], with elevated levels post-operatively correlated with resection margin positivity (R1) and development of distant metastases [6]. Advances in the molecular biology of PDAC have given hope for better disease management [7], and the identification of advantageous and/or susceptible tumour characteristics is crucial for personalising therapy and achieving longer survival. Accumulating evidence suggests that PDAC is characterised by marked genetic heterogeneity, with recent large genomic sequencing efforts demonstrating that, with the exception of a few driver mutations that currently cannot be effectively targeted (i.e. KRAS, TP53, SMAD4, CDKN2A), most mutations occur at a prevalence of
