- Email: [email protected]
Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis
Pancreatology xxx (2018) 1e8
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis Jianbin Ye, Junjie Wen, Yunshan Ning*, Yan Li** School of Laboratory Medicine and Biotechnology, Southern Medical University, Guangzhou, 510515, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 26 December 2017 Received in revised form 26 September 2018 Accepted 28 September 2018 Available online xxx
Background: At present, pancreatic ductal adenocarcinoma (PDAC) is a fetal disease lack of effective prognostic and therapeutic methods resulting in high mortality. The Notch signaling has been demonstrated being up- or down-regulated in many cancers, but the effects in pancreatic ductal adenocarcinoma are still controversial. Moreover, the available cases in an individual study are of small samples. Therefore, it is essential to define the effect of Notch signaling in pancreatic ductal adenocarcinoma with larger samples. Methods: Conducted from 6 eligible studies and 463 pancreatic ductal adenocarcinoma patients, this was the first meta-analysis to analyze the correlation between the Notch signal pathway and pancreatic ductal adenocarcinoma. All data were sourced from The National Center for Biotechnology Information, Web of Science and Cochrane. The articles which matched the inclusion criteria were included. All included data were analyzed and performed by Review Manager 5.3. Results: The results indicated that high expression of Notch signaling proteins was associated with poor overall survival of pancreatic ductal adenocarcinoma patients (pooled hazard ratio>2.00; P < 0.001). Moreover, poor survival was related to high expression of Notch3 (pooled hazard ratio: 2.05; confidence interval: 1.49e2.82; P < 0.001) and DLL4 (pooled hazard ratio: 2.13; confidence interval: 1.37e3.32; P < 0.001). Conclusions: This meta-analysis supports that Notch signaling proteins may be available as prognostic factors for pancreatic ductal adenocarcinoma progression and patient survival. Higher expression of Notch signaling proteins indicated poor survival of pancreatic ductal adenocarcinoma patients. Targeting Notch signaling components, especially Notch3 protein, would be beneficial for therapies. © 2018 Published by Elsevier B.V. on behalf of IAP and EPC.
Keywords: Notch Overall survival Pancreatic ductal adenocarcinoma Meta-analysis
Introduction WHO reported that pancreatic cancers were the seventh most common cause of cancer deaths resulting in 330,000 deaths globally [1]. And the number further increased to 411,600 in 2015 [2]. Pancreatic ductal adenocarcinoma (PDAC), which accounts for 85%
Abbreviation: PDAC, pancreatic ductal adenocarcinoma; IHC, immunohistochemistry; RT-PCR, real-time reverse transcription polymerase chain reaction; OS, Overall survival; HR, hazard ratio; CI, confidence interval; PanIN, pancreatic intraepithelial neoplasia. * Corresponding author. School of Laboratory Medicine and Biotechnology, Southern Medical University, No. 1838 North Guangzhou Street, Guangzhou, 510515, China. ** Corresponding author. School of Laboratory Medicine and Biotechnology, Southern Medical University, No. 1838 North Guangzhou Street, Guangzhou, 510515, China. E-mail addresses: [email protected] (Y. Ning), [email protected] (Y. Li).
of pancreatic cancer, typically has an inferior prognosis and survival after diagnosis [1]. In the United States, PDAC is the most lethal of all cancer types, with approximately 48,000 new cases and 40,000 deaths annually [3]. In China, estimated new case and deaths in 2015 were 90,000 and 79,000 respectively [4]. Most PDAC is diagnosed at advanced stages. There are usually no symptoms in the early stages, and the symptoms are not specific. Even with complete resection and negative results from analyses of tumor margins, long-term survival after surgery is poor, and tumors recur in all patients virtually [5]. Therefore, finding or using active early diagnosis markers and new curative targets are crucial. Notch signaling is an essential cell-to-cell communication pathway penetrating into cell survival, apoptosis, proliferation, differentiation, and cell fate control. Notch signaling contains four receptors (Notch1, Notch2, Notch3, Notch4), five ligands (DLL1, DLL3, DLL4, Jagged1, Jagged2) and a series of downstream target proteins like Hes1 and Hey1 [6,7]. A review of 112 articles
https://doi.org/10.1016/j.pan.2018.09.014 1424-3903/© 2018 Published by Elsevier B.V. on behalf of IAP and EPC.
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
2
J. Ye et al. / Pancreatology xxx (2018) 1e8
demonstrated that Notch receptors were somehow involved in pancreatic tumor cells [8]. Furthermore, a certain percentage of patients with higher expression of Notch proteins like Notch1 and Notch3 showed poorer overall survival [9e14]. However, another review reported that the Notch signaling pathway could serve as both promotor and inhibitor in the different stage of pancreas [15]. Recent articles found that Notch1 might function as a tumor suppressor in pancreatic cancer [16,17]. Therefore, to make a convictive result about the correlation between Notch and PDAC, we did a systematic review by using a larger scale of clinical data. The current reports of Notch in PDAC which are focused on the cell lines, mouse models and xenografts are lacking further understanding of the effects of the Notch signaling on patients. Here, we presented the first systematic review and meta-analysis to evaluate the impact of Notch signaling on PDAC patients and determine whether Notch signaling could be an independent prognostic factor or therapeutic target.
provided according to criteria by Newcastle-Ottawa quality assessment scale [20]. The full score is 9. All the meta-analyses were performed using Review Manager 5.3 (Cochrane Collaboration, Oxford, UK). Statistical heterogeneity between studies was assessed using the chi-square test with significance set at p < 0.10, and heterogeneity was quantified using the I [2] statistic. The random-effects model was used if there was heterogeneity between studies; otherwise, the fixed-effects model was used. The P value of hypothesis testing <0.01 should be considered as a significant difference. The risk of bias and sensitivity of individual studies were analyzed by the Review Manager 5.3. If the studies locate inside the funnel with an even distribution around the vertical, it should be considered no significant publication bias. The studies which would change the results significantly in sensitivity analysis should be removed or reconsidered.
Results Methods Characteristics of included studies Ethics statement and guidelines The present systematic review and meta-analysis did not involve animal experiments or direct human trials, neither a particular ethics review nor an ethical approval was therefore necessary. The study was performed in accordance with the Preferred Reporting Items for Systematic Reviews and MetaAnalyses statement [18,19]. Search methods for identification of studies To evaluate the expression of Notch signaling members as the predictive markers for outcomes of PDAC, a literature search was performed in June 2018 without restriction to regions, publication types, or languages. All publication types were searched, but only the qualified articles met criteria were included and analyzed. The primary sources were the electronic databases of PubMed, Web of Science and Cochrane Library. The search item for PubMed and Web of Science was (((((pancreatic cancer) OR pancreatic carcinoma) OR pancreatic adenocarcinoma) OR neuroendocrine tumor) OR solid pseudo papillary neoplasm) AND notch. The search item for Cochrane Library was notch (All text). The search item “Notch” could include all Notch signaling proteins because it was “Notch” in all text. When multiple reports describing the same population were published, the most recent or complete report was used. Inclusion criteria All prospective (cohort study) and retrospective studies (casecontrol study) which compared high/detected Notch group with low/not-detected Notch group should be evaluated. Studies that met the following criteria were included: a) Patients were diagnosed to have pancreatic cancer; b) Notch signaling expression was measured by immunohistochemistry (IHC) or real-time reverse transcription polymerase chain reaction (RT-PCR) in primary pancreatic cancer tissue; c) Overall survival (OS) should be analyzed, and the hazard ratio (HR) and confidence interval (CI) could be extracted. Data from the included studies were extracted and summarized independently by two of the authors (NY and LY). Any disagreement was resolved by the adjudicating senior author (YJ). Quality assessment and statistical analysis The observational studies were rated for the level of evidence
The details and process of studies screening were shown in Fig. 1. All possible studies about Notch on PDAC have been analyzed and only the qualified studies that match the inclusion criteria were included. The four reviews excluded did not provide the details of clinical data. The characteristics of 6 included studies were illustrated in Tables 1e3. All six included articles were observational studies based on human PDAC samples. The methodological quality of these six studies has been qualified according to NewcastleOttawa quality assessment scale [20]. The hazard ratio (HR) and confidence interval (CI) of overall survival (OS) were provided from the articles of Zhou, J. X., et al. [13], Drouillard, A. et al. [14], Cao, F., et al. [9], Mann, C.D., et al. [10] and Chen, H.T., et al. [11]. HR and CI were not offered in the article by Doucas, H., et al. [12] directly, but HR and CI could be extracted from the survival curve by software named Engauge. Totally 463 patients of PDAC and 47 normal controls were included. The patients were all diagnosed as PDAC histologically, and the average age was around 60.
Notch1, Notch3, and Hes1 were expressed more frequently in PDAC compared with normal pancreas Notch expression detected by IHC was summarized in Table 2. As displayed in Fig. 2, Notch1 was expressed in 128 of 181 tissue samples (70.7%) while no Notch1 was expressed in 35 uninvolved pancreases (0%) [10,14]. Notch3 was expressed in 216 of 305 tissue samples (70.8%) while no Notch3 was expressed in 47 uninvolved pancreases (0%) [10,12,13]. Hes1 was expressed in 136 of 164 tissue samples (82.9%) while 8 in 35 uninvolved pancreases (22.86%). Since not all articles were involved in normal pancreas, the normal pancreas analyzed here only extracted from 2 articles [10,12]. Above all, Notch1, Notch3, and Hes1 were expressed more frequently in PDAC compared with normal pancreas.
Higher expression of Notch3 and DLL4 was associated with reduced overall survival PDAC patients with higher Notch3 [10,12,13] and DLL4 [10,14] expressions had worse survival than the patients with low expressions. As displayed in Fig. 3, survival time of patients was significantly lower with higher Notch3 (pooled HR: 2.05; CI: 1.49e2.82; P < 0.0001) and DLL4 (pooled HR: 2.13; CI: 1.37e3.32; P ¼ 0.0009) expression respectively [11,14].
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
J. Ye et al. / Pancreatology xxx (2018) 1e8
3
Fig. 1. Flow diagram of studies identified, included, and excluded.
Table 1 Characteristics of included studies. Study
Year Design Histology Operation
Quality Patient Uninvolved Age Histological grade Detection Follow score number pancreas (median) (differentiation) & analysis up (months)
Zhou, J. X. et al.
2016 R
PDAC
Surgically resected
101
Drouillard, 2016 R A. et al. [14] Cao, F. et al. 2015 R [9]
PDAC
Chen, H.T. 2012 R et al. [11] Mann, C.D. 2012 R et al. [10] Doucas, H. 2008 R et al. [12]
NA
55.3
Duodenopancreatectomy, Distal pancreatectomy 86
NA
66
PDAC
Duodenopancreatectomy, Distal pancreatectomy. 72
NA
66.5
PDAC
89
NA
58
PDAC
Total pancreatectomy, Pancreaticoduodenectomy, or pyloruspreserving pancreaticoduodenectomy Surgically resected.
42
35
64
PDAC
Unresected Surgically resected
50 23
10
NA 66
Well (33), moderately/ poorly (68) Well/moderately (69), poorly (17) Well (7), moderately (33, poorly (32) Well (1), moderately (70), poorly (18) Well (4), moderately (19), poorly (18) NA NA
IHC, OS
12
4
IHC, OS
104
4
IHC, OS
37
4
IHC, OS
48
4
IHC, OS
122
4
IHC IHC, OS
NA 27
4
Abbreviation: PDAC: Pancreatic ductal adenocarcinoma; R: retrospective; IHC: Immunohistochemistry; OS: overall survival; NA: data not available.
Higher expression of Notch components was associated with reduced overall survival The original details of the overall survival of each study were shown in Table 3. Two studies focused on DLL4 [11,14], one study focused on Hes1 [9], three studies focused on Notch3 [10,12,13] and one study was involved in Hey1 [10]. All these four proteins were parts of Notch signaling. And the positive detection or high expression of these Notch components could imply the activation of Notch signaling to a certain extent. The overall survival of patients who expressed Notch components was analyzed and shown in Fig. 4. The analysis indicated that higher expression of Notch
components would lead to poor survival (Fig. 4a, pooled HR: 2.06, CI: 1.60e2.66, P < 0.00001; Fig. 4b, pooled HR: 2.11, CI: 1.64e2.73, P < 0.00001). Sensitivity analysis and publication bias The HRs of all studies analyzed were larger than 2.00. Sensitivity analysis showed that the pooled HR was not changed significantly with any studies removed. Fig. 5 showed four funnel plots (a, b, c, d) of studies analyzed above, which were respectively corresponding to the analysis in Fig. 3a, b, 4a, b. All studies located inside the 95% CIs, with an even
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
4
J. Ye et al. / Pancreatology xxx (2018) 1e8
Table 2 Notch expression detected by IHC from 6 studies. study
Detection
Zhou, J. X. et al. Drouillard, A. et al.
Cao, F. et al. Chen, H.T. et al. Mann, C.D. et al.
Doucas, H. et al.
Location
IHC IHC IHC IHC IHC IHC IHC IHC IHC IHC IHC IHC
Gene
Cytoplasm, cell membrane NA NA NA NA tumor stromal vessels Nuclear Nuclear Nuclear Nuclear Nuclear Nuclear
Notch3 DLL4 Notch1 Notch3 Hes1 DLL4 Notch1 Notch3 Notch4 Hes1 Hey1 Notch3
Normal
patients
Low/ND
High/D
Low/ND
High/D
NA NA NA NA NA NA 35 35 35 27 35 12
NA NA NA NA NA NA 0 0 0 8 0 0
46 3 7 21 19 50 46 27 52 9 43 13
55 86 82 68 53 32 46 65 40 83 49 10
Abbreviation: IHC: Immunohistochemistry; Low: low expression; ND: not detected; High: high expression; D: detected.
Table 3 The details of overall survival and Notch expression from 6 studies. study
Patients for OS analysis
Gene detected
Detection IHC score
Patient number
Multivariate analysis, hazard ratio (95% CI)
P Univariate analysis, hazard value ratio (95% CI)
P value
Low/ High/ ND D Zhou, J. X. 101 et al. Drouillard, A. 83 et al. Cao, F. et al. 72 Chen, H.T. et al. Mann, C.D. et al. Doucas, H. et al.
Notch3
IHC
DLL4
IHC
Hes1
IHC
82
DLL4
IHC
42 42 23
Notch3 Hey1 Notch3
IHC IHC IHC
Low (-, þ), high (þþ, þþþ) Low (<50%), high (>50%) Low (<25%), high (>25%) Low (<50%), high (>50%) – (<10%), þ (>10%) – (<10%), þ (>10%) – (<5%), þ (>5%)
46
55
1.960 (1.181e3.252)
0.009 2.927 (1.876e4.569)
<0.001
33
50
2.05 (1.1e3.6)
0.02
0.005
19
53
2.012 (1.549e10.214)
0.001 2.154 (1.987e11.212)
<0.001
50
32
2.24 (1.14e4.38)
0.019 NA
NA
22 32 13
20 11 10
2.671 (1.030e6.928) 3.591 (1.441e8.946) 1.95 (1.21e3.16)
0.053 2.541 (1.213e5.324) 0.006 2.998 (1.326e6.778) 0.036 NA
0.013 0.008 NA
2.12 (1.34e3.63)
Abbreviation: IHC: Immunohistochemistry; OS: overall survival; Low: low expression; ND: not detected; High: high expression; D: detected.
distribution around the vertical, indicating no obvious publication bias. Discussion Pancreatic ductal adenocarcinoma (PDAC) is a gastrointestinal tumor with a high degree of malignancy, which is difficult to be diagnosed and treated. Many dysregulated signal pathways have been characterized in PDAC, such as Notch signaling [21]. As our analysis presented, Notch signaling was over-expressed in some patients and associated with a lower survival rate. Not all patients over-express the same Notch proteins, but up-regulation of any Notch components, especially Notch3 and DLL4, should be an alarm. Notch signal pathway not only works as an accelerator to PDAC as we introduced, but also induces chemoresistance to antineoplasm drugs like gemcitabine [22,23]. Therefore, targeting Notch signaling is a feasible way to develop a novel therapeutic treatment for PADC. Generally, activation of Notch signaling was associated with poor outcome. Deficiency of Notch2 but not Notch1 stopped pancreatic intraepithelial neoplasia (PanIN) progression, prolonged survival, and led to a phenotypical switch toward anaplastic pancreatic cancer with epithelial-mesenchymal transition [24]. Aberrantly high expressed Jagged2 could enhance PDAC both migration and invasion with affecting cell proliferation, similar to Notch1 [25]. Increased expression of Notch1 and Hes1 was discovered in gemcitabine-treated cell lines and led to treatment failure [26]. Many proteins in Notch signaling were alternative
therapeutic targets. A Notch2/Notch3 antagonist (tarextumab) inhibited tumor growth and decreased tumor-initiating cell frequency in pancreatic tumor xenografts [27]. Inhibition of MAML (an essential co-activator of the canonical Notch signaling-mediated transcription) could delay PanIN formation [28]. According to our analysis results, Notch3 was detected only in PDAC patients by IHC, and higher Notch3 expression showed worse survival [10,12,13]. In the study of Eto, K., et al. [23], it demonstrated an extremely similar result that low Notch3 mRNA expression was significantly associated with longer overall survival in a multivariate analysis (P < 0.0094). Also, our analysis implied that higher Notch3 and DLL4 expression indicated poor survival [11,14]. Thus, Notch3 and DLL4 seem to be the most promising therapeutic target of Notch signal pathway in PDAC. Many researchers focused on one of Notch signaling restrict enzyme, g-secretase. Yabuuchi, S., et al. [29], showed that PF03084014, a selective g-secretase inhibitor, had greater antitumor activity in combination with gemcitabine in PDAC xenografts. PF-03084014 inhibited the cleavage of the nuclear Notch1 intracellular domain and targeted Hes-1 and Hey-1. PF-03084014 also targeted cancer stem cells within pancreatic tumors. Similar consequences were demonstrated in studies using other g-secretase inhibitors on PDAC xenografts [30e32]. But only focusing on gsecretase seemed to be not good enough. According to a phase II study of a g-secretase inhibitor RO4929097 in patients with previously treated metastatic PDAC, RO4929097 was well-tolerated [33]. And the use of g-secretase inhibitor led to intestinal goblet cell hyperplasia, thymus atrophy, a decrease in lymphocytes, and
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
J. Ye et al. / Pancreatology xxx (2018) 1e8
5
Fig. 2. Forest plot of the comparison of Notch1, Notch3 and Hes1 expression between PDAC patients and normal pancreas. Events: Notch proteins were detected or under high expression in patients. MeH: Mantel-Haenszel method; CI: confidence interval.
Fig. 3. (a) Forest plot of overall survival between PDAC patients with high/detected Notch3 and low/not detected. (b) Forest plot of overall survival between PDAC patients with high/detected DLL4 and low/not detected. CI: confidence interval.
alterations in hair color [34]. Because g-secretase is only one of the many enzymes in Notch signaling and it also participates in many other pathways, the therapeutic target could be more specific and effective like Notch3 and DLL4 in our analytic results. In fact, cancers near pancreas shared a similar association with Notch signaling, like gastric cancer, colon cancer, colorectal cancer,
and cholangiocarcinoma. According to a meta-analysis of Notch and gastric cancers [35], higher expression of Notch3 and Jagged1 was found in diffuse-type gastric cancer. Jagged1 was also significantly over-expressed in poor differentiation type. And overexpression DLL4 was associated with advanced T stage, N stage and TNM stage in gastric cancer patients. In colorectal cancer, high
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
6
J. Ye et al. / Pancreatology xxx (2018) 1e8
Fig. 4. Forest plot of overall survival between PDAC patients with high/detected Notch components and low/not detected. (a) Notch components including Notch3, DLL4, and Hes1; (b) Notch components including Notch3, DLL4, Hes1 and Hey1.
Fig. 5. Funnel plot illustrating meta-analysis of overall survival. SE: standard error; OR: odds ratio. The circles represent the articles. (a) Corresponding to the three studies in Fig. 3a. (b). Corresponding to the two studies in Fig. 3b. (c) Corresponding to the six studies in Fig. 4a. (d). Corresponding to the six studies in Fig. 4b.
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
J. Ye et al. / Pancreatology xxx (2018) 1e8
expression of Hes1 was found significantly correlated with distal metastasis in 320 patient samples [36]. Up-regulation of DLL4 was closely related to metastasis and drug resistance in colorectal and colon cancer, leading to poor prognosis [37,38]. Similarly, Notch1 and Hes1 were found increased in colon cancer cells which were resistant to regorafenib [39]. Especially, nuclear Notch3 expression was associated with tumor recurrence in patients with stage II and III colorectal cancers [40]. Through gene profile and cell surface proteome analysis, Notch signaling was found involved in cholangiocarcinoma [41,42]. More studies illustrated the same result that excessive activation of Notch signaling promotes the development of digestive cancers like Notch signaling in PDAC. However, more and more research found Notch signaling pathway had different effects in different situations of PDAC. A study pointed out that Notch signaling might inhibit PanIN development at an early stage but promote PanIN progression at later stages once lesions are established [15]. Recent studies have also demonstrated tumor suppressive effects of Notch signaling pathway. A study found that missing Notch expression in the malignant tumor and loss of Notch1 in a model of K-ras-induced PDAC led to tumor progression, suggesting that the Notch1 can serve as a tumor suppressor in PDAC [16,17]. To sum up, the effect of Notch signaling pathway in PDAC is complex. It might be related to the crosstalk. PDAC progression is associated with various signaling pathway like KRAS signaling, WNT signaling, TGF-beta signaling. In the field of molecular biology, the exact role of the Notch pathway in pancreatic cancer is still a controversy. Therefore, considering the clinical situation, we need to analyze clinical data to conclude the association between Notch and PADC patient. Adopting only six studies and 463 patients and 47 health controls in this meta-analysis may be a limitation that should be mentioned, but all six studies elucidated similar results. Moreover, there was no publication bias as our analysis shown. Therefore, the results of this meta-analysis should be convincing. But if possible, the data covered should be enlarged. Our study is the first meta-analysis to review all articles about the effects of Notch on human PDAC to provide the scientific foundations for policy, medicine, social science, business, ecology, and other fields. The results demonstrated the idea that higher expression of Notch proteins indicates poor survival of patients with PDAC, which will help understand the function of Notch in PDAC and provide a possible prognostic index of PDAC and even a therapeutic target based on a more significant sample number. Conflicts of interest
[3] [4] [5]
[6] [7] [8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
All the authors have no competing financial interests. [24]
Acknowledgments This research was supported by the Natural Science Foundation of China (No.81470831), NSFC-Guangdong Province Joint Key Project (U1401223), Science and Technology Planning Project of Guangdong Province, China (No. 2016B090919019), Science and Technology Planning Project of Guangzhou City, China (201604010057), Innovative experiment program of college students of Guangdong Province, China (No. 201612121034, 201812121036, 201812121110, 201812121231, 201812121234). References [1] Stewart B, Wild CP. World cancer report 2014. World Health Organization; 2014. [2] Global, regional, and national life expectancy, all-cause mortality, and cause-
[25]
[26] [27]
[28]
[29]
[30]
7
specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the global burden of disease study 2015. Lancet (London, England) 2016;388:1459e544. Ryan DP, Hong TS, Bardeesy N. Pancreatic adenocarcinoma. N Engl J Med 2014;371:2140e1. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, et al. Cancer statistics in China, 2015. CA A Cancer J Clin 2016;66:115e32. Yeo CJ, Cameron JL, Lillemoe KD, Sitzmann JV, Hruban RH, Goodman SN, et al. Pancreaticoduodenectomy for cancer of the head of the pancreas. 201 patients. Ann Surg 1995;221:721e31. discussion 731-723. Kopan R, Ilagan MX. The canonical notch signaling pathway: unfolding the activation mechanism. Cell 2009;137:216e33. Bray SJ. Notch signalling: a simple pathway becomes complex. Nat Rev Mol Cell Biol 2006;7:678e89. Damaskos C, Karatzas T, Kostakis ID, Nikolidakis L, Kostakis A, Kouraklis G. Nuclear receptors in pancreatic tumor cells. Anticancer Res 2014;34: 6897e911. Cao F, Li J, Sun H, Liu S, Cui Y, Li F. Hes 1 is essential for chemoresistance induced by stellate cells and is associated with poor prognosis in pancreatic cancer. Oncol Rep 2015;33:1883e9. Mann CD, Bastianpillai C, Neal CP, Masood MM, Jones DJ, Teichert F, et al. Notch3 and hey-1 as prognostic biomarkers in pancreatic adenocarcinoma. PloS One 2012;7, e51119. Chen HT, Cai QC, Zheng JM, Man XH, Jiang H, Song B, et al. High expression of delta-like ligand 4 predicts poor prognosis after curative resection for pancreatic cancer. Ann Surg Oncol 2012;19(Suppl 3):S464e74. Doucas H, Mann CD, Sutton CD, Garcea G, Neal CP, Berry DP, et al. Expression of nuclear notch3 in pancreatic adenocarcinomas is associated with adverse clinical features, and correlates with the expression of stat3 and phosphorylated akt. J Surg Oncol 2008;97:63e8. Zhou JX, Zhou L, Li QJ, Feng W, Wang PM, Li EF, et al. Association between high levels of notch3 expression and high invasion and poor overall survival rates in pancreatic ductal adenocarcinoma. Oncol Rep 2016;36:2893e901. Drouillard A, Puleo F, Bachet JB, Ouazzani S, Calomme A, Demetter P, et al. Dll4 expression is a prognostic marker and may predict gemcitabine benefit in resected pancreatic cancer. Br J Canc 2016;115:1245e52. Avila JL, Kissil JL. Notch signaling in pancreatic cancer: oncogene or tumor suppressor? Trends Mol Med 2013;19:320e7. Krausch M, Kroepil F, Lehwald N, Lachenmayer A, Schott M, Anlauf M, et al. Notch 1 tumor expression is lacking in highly proliferative pancreatic neuroendocrine tumors. Endocrine 2013;44:182e6. Hanlon L, Avila JL, Demarest RM, Troutman S, Allen M, Ratti F, et al. Notch1 functions as a tumor suppressor in a model of k-ras-induced pancreatic ductal adenocarcinoma. Cancer Res 2010;70:4280e6. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the prisma statement. Int J Surg 2010;8: 336e41. Knobloch K, Yoon U, Vogt PM. Preferred reporting items for systematic reviews and meta-analyses (prisma) statement and publication bias. J Craniomaxillo-Fac Surg: Official Pub European Assoc Cranio-Maxillo-Fac Surg 2011;39:91e2. Stang A. Critical evaluation of the newcastle-ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603e5. Knudsen ES, O'Reilly EM, Brody JR, Witkiewicz AK. Genetic diversity of pancreatic ductal adenocarcinoma and opportunities for precision medicine. Gastroenterology 2015;150(1):48e63. Du X, Zhao YP, Zhang TP, Zhou L, Chen G, Wang TX, et al. Alteration of the intrinsic apoptosis pathway is involved in notch-induced chemoresistance to gemcitabine in pancreatic cancer. Arch Med Res 2014;45:15e20. Eto K, Kawakami H, Kuwatani M, Kudo T, Abe Y, Kawahata S, et al. Human equilibrative nucleoside transporter 1 and notch3 can predict gemcitabine effects in patients with unresectable pancreatic cancer. Br J Canc 2013;108: 1488e94. Mazur PK, Einw€ achter H, Lee M, Sipos B, Nakhai H, Rad R, et al. Notch2 is required for progression of pancreatic intraepithelial neoplasia and development of pancreatic ductal adenocarcinoma. Proc Natl Acad Sci Unit States Am 2010;107:13438e43. Hu Y, Su H, Li X, Guo G, Cheng L, Qin R, et al. The notch ligand jagged2 promotes pancreatic cancer metastasis independent of notch signaling activation. Mol Canc Therapeut 2015;14:289e97. Lee JY, Song SY, Park JY. Notch pathway activation is associated with pancreatic cancer treatment failure. Pancreatology 2014;14:48e53. Yen WC, Fischer MM, Axelrod F, Bond C, Cain J, Cancilla B, et al. Targeting notch signaling with a notch2/notch3 antagonist (tarextumab) inhibits tumor growth and decreases tumor-initiating cell frequency. Clin Canc Res: Official J Am Assoc Canc Res 2015;21:2084e95. Thomas MM, Zhang Y, Mathew E, Kane KT, Maillard I, Pasca di Magliano M. Epithelial notch signaling is a limiting step for pancreatic carcinogenesis. BMC Canc 2014;14:862. Yabuuchi S, Pai SG, Campbell NR, de Wilde RF, De Oliveira E, Korangath P, et al. Notch signaling pathway targeted therapy suppresses tumor progression and metastatic spread in pancreatic cancer. Cancer Lett 2013;335:41e51. Palagani V, Bozko P, El Khatib M, Belahmer H, Giese N, Sipos B, et al. Combined inhibition of notch and jak/stat is superior to monotherapies and impairs
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
8
J. Ye et al. / Pancreatology xxx (2018) 1e8
pancreatic cancer progression. Carcinogenesis 2014;35:859e66. [31] Palagani V, El Khatib M, Kossatz U, Bozko P, Muller MR, Manns MP, et al. Epithelial mesenchymal transition and pancreatic tumor initiating cd44þ/ epcamþ cells are inhibited by gamma-secretase inhibitor ix. PloS One 2012;7, e46514. [32] Mizuma M, Rasheed ZA, Yabuuchi S, Omura N, Campbell NR, de Wilde RF, et al. The gamma secretase inhibitor mrk-003 attenuates pancreatic cancer growth in preclinical models. Mol Canc Therapeut 2012;11:1999e2009. [33] De Jesus-Acosta A, Laheru D, Maitra A, Arcaroli J, Rudek MA, Dasari A, et al. A phase ii study of the gamma secretase inhibitor ro4929097 in patients with previously treated metastatic pancreatic adenocarcinoma. Invest N Drugs 2014;32:739e45. [34] Panza F, Frisardi V, Imbimbo BP, Capurso C, Logroscino G, Sancarlo D, et al. Review: gamma-secretase inhibitors for the treatment of alzheimer's disease: the current state. CNS Neurosci Ther 2010;16:272e84. [35] Du X, Cheng Z, Wang YH, Guo ZH, Zhang SQ, Hu JK, et al. Role of notch signaling pathway in gastric cancer: a meta-analysis of the literature. World J Gastroenterol 2014;20:9191e9. [36] Yuan R, Ke J, Sun L, He Z, Zou Y, He X, et al. Hes1 promotes metastasis and predicts poor survival in patients with colorectal cancer. Clin Exp Metastasis
2015;32:169e79. [37] Negri FV, Crafa P, Pedrazzi G, Bozzetti C, Lagrasta C, Gardini G, et al. Strong notch activation hinders bevacizumab efficacy in advanced colorectal cancer. Future Oncol (London, England) 2015;11:3167e74. [38] Zheng Y, Hao Z, Ding Y, Wang Q, Li S, Xiao G, et al. Expression of delta-like 4 (drosophila) and vascular endothelial growth factor a in colon cancer and association with tumour angiogenesis. J Int Med Res 2015;43:535e43. [39] Mirone G, Perna S, Shukla A, Marfe G. Involvement of notch-1 in resistance to regorafenib in colon cancer cells. J Cell Physiol 2015;231(5):1097e105. [40] Ozawa T, Kazama S, Akiyoshi T, Murono K, Yoneyama S, Tanaka T, et al. Nuclear notch3 expression is associated with tumor recurrence in patients with stage ii and iii colorectal cancer. Ann Surg Oncol 2014;21:2650e8. [41] Xue TC, Zhang BH, Ye SL, Ren ZG. Differentially expressed gene profiles of intrahepatic cholangiocarcinoma, hepatocellular carcinoma, and combined hepatocellular-cholangiocarcinoma by integrated microarray analysis. Tumour Biol: J Int Soc Oncodevelopmental Biol Med 2015;36:5891e9. [42] Stephenson B, Shimwell N, Humphreys E, Ward D, Adams D, Martin A, et al. Quantitative assessment of the cell surface proteome to identify novel therapeutic targets in cholangiocarcinoma. Lancet (London, England) 2015;385(Suppl 1):S94.
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis Jianbin Ye, Junjie Wen, Yunshan Ning*, Yan Li** School of Laboratory Medicine and Biotechnology, Southern Medical University, Guangzhou, 510515, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 26 December 2017 Received in revised form 26 September 2018 Accepted 28 September 2018 Available online xxx
Background: At present, pancreatic ductal adenocarcinoma (PDAC) is a fetal disease lack of effective prognostic and therapeutic methods resulting in high mortality. The Notch signaling has been demonstrated being up- or down-regulated in many cancers, but the effects in pancreatic ductal adenocarcinoma are still controversial. Moreover, the available cases in an individual study are of small samples. Therefore, it is essential to define the effect of Notch signaling in pancreatic ductal adenocarcinoma with larger samples. Methods: Conducted from 6 eligible studies and 463 pancreatic ductal adenocarcinoma patients, this was the first meta-analysis to analyze the correlation between the Notch signal pathway and pancreatic ductal adenocarcinoma. All data were sourced from The National Center for Biotechnology Information, Web of Science and Cochrane. The articles which matched the inclusion criteria were included. All included data were analyzed and performed by Review Manager 5.3. Results: The results indicated that high expression of Notch signaling proteins was associated with poor overall survival of pancreatic ductal adenocarcinoma patients (pooled hazard ratio>2.00; P < 0.001). Moreover, poor survival was related to high expression of Notch3 (pooled hazard ratio: 2.05; confidence interval: 1.49e2.82; P < 0.001) and DLL4 (pooled hazard ratio: 2.13; confidence interval: 1.37e3.32; P < 0.001). Conclusions: This meta-analysis supports that Notch signaling proteins may be available as prognostic factors for pancreatic ductal adenocarcinoma progression and patient survival. Higher expression of Notch signaling proteins indicated poor survival of pancreatic ductal adenocarcinoma patients. Targeting Notch signaling components, especially Notch3 protein, would be beneficial for therapies. © 2018 Published by Elsevier B.V. on behalf of IAP and EPC.
Keywords: Notch Overall survival Pancreatic ductal adenocarcinoma Meta-analysis
Introduction WHO reported that pancreatic cancers were the seventh most common cause of cancer deaths resulting in 330,000 deaths globally [1]. And the number further increased to 411,600 in 2015 [2]. Pancreatic ductal adenocarcinoma (PDAC), which accounts for 85%
Abbreviation: PDAC, pancreatic ductal adenocarcinoma; IHC, immunohistochemistry; RT-PCR, real-time reverse transcription polymerase chain reaction; OS, Overall survival; HR, hazard ratio; CI, confidence interval; PanIN, pancreatic intraepithelial neoplasia. * Corresponding author. School of Laboratory Medicine and Biotechnology, Southern Medical University, No. 1838 North Guangzhou Street, Guangzhou, 510515, China. ** Corresponding author. School of Laboratory Medicine and Biotechnology, Southern Medical University, No. 1838 North Guangzhou Street, Guangzhou, 510515, China. E-mail addresses: [email protected] (Y. Ning), [email protected] (Y. Li).
of pancreatic cancer, typically has an inferior prognosis and survival after diagnosis [1]. In the United States, PDAC is the most lethal of all cancer types, with approximately 48,000 new cases and 40,000 deaths annually [3]. In China, estimated new case and deaths in 2015 were 90,000 and 79,000 respectively [4]. Most PDAC is diagnosed at advanced stages. There are usually no symptoms in the early stages, and the symptoms are not specific. Even with complete resection and negative results from analyses of tumor margins, long-term survival after surgery is poor, and tumors recur in all patients virtually [5]. Therefore, finding or using active early diagnosis markers and new curative targets are crucial. Notch signaling is an essential cell-to-cell communication pathway penetrating into cell survival, apoptosis, proliferation, differentiation, and cell fate control. Notch signaling contains four receptors (Notch1, Notch2, Notch3, Notch4), five ligands (DLL1, DLL3, DLL4, Jagged1, Jagged2) and a series of downstream target proteins like Hes1 and Hey1 [6,7]. A review of 112 articles
https://doi.org/10.1016/j.pan.2018.09.014 1424-3903/© 2018 Published by Elsevier B.V. on behalf of IAP and EPC.
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
2
J. Ye et al. / Pancreatology xxx (2018) 1e8
demonstrated that Notch receptors were somehow involved in pancreatic tumor cells [8]. Furthermore, a certain percentage of patients with higher expression of Notch proteins like Notch1 and Notch3 showed poorer overall survival [9e14]. However, another review reported that the Notch signaling pathway could serve as both promotor and inhibitor in the different stage of pancreas [15]. Recent articles found that Notch1 might function as a tumor suppressor in pancreatic cancer [16,17]. Therefore, to make a convictive result about the correlation between Notch and PDAC, we did a systematic review by using a larger scale of clinical data. The current reports of Notch in PDAC which are focused on the cell lines, mouse models and xenografts are lacking further understanding of the effects of the Notch signaling on patients. Here, we presented the first systematic review and meta-analysis to evaluate the impact of Notch signaling on PDAC patients and determine whether Notch signaling could be an independent prognostic factor or therapeutic target.
provided according to criteria by Newcastle-Ottawa quality assessment scale [20]. The full score is 9. All the meta-analyses were performed using Review Manager 5.3 (Cochrane Collaboration, Oxford, UK). Statistical heterogeneity between studies was assessed using the chi-square test with significance set at p < 0.10, and heterogeneity was quantified using the I [2] statistic. The random-effects model was used if there was heterogeneity between studies; otherwise, the fixed-effects model was used. The P value of hypothesis testing <0.01 should be considered as a significant difference. The risk of bias and sensitivity of individual studies were analyzed by the Review Manager 5.3. If the studies locate inside the funnel with an even distribution around the vertical, it should be considered no significant publication bias. The studies which would change the results significantly in sensitivity analysis should be removed or reconsidered.
Results Methods Characteristics of included studies Ethics statement and guidelines The present systematic review and meta-analysis did not involve animal experiments or direct human trials, neither a particular ethics review nor an ethical approval was therefore necessary. The study was performed in accordance with the Preferred Reporting Items for Systematic Reviews and MetaAnalyses statement [18,19]. Search methods for identification of studies To evaluate the expression of Notch signaling members as the predictive markers for outcomes of PDAC, a literature search was performed in June 2018 without restriction to regions, publication types, or languages. All publication types were searched, but only the qualified articles met criteria were included and analyzed. The primary sources were the electronic databases of PubMed, Web of Science and Cochrane Library. The search item for PubMed and Web of Science was (((((pancreatic cancer) OR pancreatic carcinoma) OR pancreatic adenocarcinoma) OR neuroendocrine tumor) OR solid pseudo papillary neoplasm) AND notch. The search item for Cochrane Library was notch (All text). The search item “Notch” could include all Notch signaling proteins because it was “Notch” in all text. When multiple reports describing the same population were published, the most recent or complete report was used. Inclusion criteria All prospective (cohort study) and retrospective studies (casecontrol study) which compared high/detected Notch group with low/not-detected Notch group should be evaluated. Studies that met the following criteria were included: a) Patients were diagnosed to have pancreatic cancer; b) Notch signaling expression was measured by immunohistochemistry (IHC) or real-time reverse transcription polymerase chain reaction (RT-PCR) in primary pancreatic cancer tissue; c) Overall survival (OS) should be analyzed, and the hazard ratio (HR) and confidence interval (CI) could be extracted. Data from the included studies were extracted and summarized independently by two of the authors (NY and LY). Any disagreement was resolved by the adjudicating senior author (YJ). Quality assessment and statistical analysis The observational studies were rated for the level of evidence
The details and process of studies screening were shown in Fig. 1. All possible studies about Notch on PDAC have been analyzed and only the qualified studies that match the inclusion criteria were included. The four reviews excluded did not provide the details of clinical data. The characteristics of 6 included studies were illustrated in Tables 1e3. All six included articles were observational studies based on human PDAC samples. The methodological quality of these six studies has been qualified according to NewcastleOttawa quality assessment scale [20]. The hazard ratio (HR) and confidence interval (CI) of overall survival (OS) were provided from the articles of Zhou, J. X., et al. [13], Drouillard, A. et al. [14], Cao, F., et al. [9], Mann, C.D., et al. [10] and Chen, H.T., et al. [11]. HR and CI were not offered in the article by Doucas, H., et al. [12] directly, but HR and CI could be extracted from the survival curve by software named Engauge. Totally 463 patients of PDAC and 47 normal controls were included. The patients were all diagnosed as PDAC histologically, and the average age was around 60.
Notch1, Notch3, and Hes1 were expressed more frequently in PDAC compared with normal pancreas Notch expression detected by IHC was summarized in Table 2. As displayed in Fig. 2, Notch1 was expressed in 128 of 181 tissue samples (70.7%) while no Notch1 was expressed in 35 uninvolved pancreases (0%) [10,14]. Notch3 was expressed in 216 of 305 tissue samples (70.8%) while no Notch3 was expressed in 47 uninvolved pancreases (0%) [10,12,13]. Hes1 was expressed in 136 of 164 tissue samples (82.9%) while 8 in 35 uninvolved pancreases (22.86%). Since not all articles were involved in normal pancreas, the normal pancreas analyzed here only extracted from 2 articles [10,12]. Above all, Notch1, Notch3, and Hes1 were expressed more frequently in PDAC compared with normal pancreas.
Higher expression of Notch3 and DLL4 was associated with reduced overall survival PDAC patients with higher Notch3 [10,12,13] and DLL4 [10,14] expressions had worse survival than the patients with low expressions. As displayed in Fig. 3, survival time of patients was significantly lower with higher Notch3 (pooled HR: 2.05; CI: 1.49e2.82; P < 0.0001) and DLL4 (pooled HR: 2.13; CI: 1.37e3.32; P ¼ 0.0009) expression respectively [11,14].
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
J. Ye et al. / Pancreatology xxx (2018) 1e8
3
Fig. 1. Flow diagram of studies identified, included, and excluded.
Table 1 Characteristics of included studies. Study
Year Design Histology Operation
Quality Patient Uninvolved Age Histological grade Detection Follow score number pancreas (median) (differentiation) & analysis up (months)
Zhou, J. X. et al.
2016 R
PDAC
Surgically resected
101
Drouillard, 2016 R A. et al. [14] Cao, F. et al. 2015 R [9]
PDAC
Chen, H.T. 2012 R et al. [11] Mann, C.D. 2012 R et al. [10] Doucas, H. 2008 R et al. [12]
NA
55.3
Duodenopancreatectomy, Distal pancreatectomy 86
NA
66
PDAC
Duodenopancreatectomy, Distal pancreatectomy. 72
NA
66.5
PDAC
89
NA
58
PDAC
Total pancreatectomy, Pancreaticoduodenectomy, or pyloruspreserving pancreaticoduodenectomy Surgically resected.
42
35
64
PDAC
Unresected Surgically resected
50 23
10
NA 66
Well (33), moderately/ poorly (68) Well/moderately (69), poorly (17) Well (7), moderately (33, poorly (32) Well (1), moderately (70), poorly (18) Well (4), moderately (19), poorly (18) NA NA
IHC, OS
12
4
IHC, OS
104
4
IHC, OS
37
4
IHC, OS
48
4
IHC, OS
122
4
IHC IHC, OS
NA 27
4
Abbreviation: PDAC: Pancreatic ductal adenocarcinoma; R: retrospective; IHC: Immunohistochemistry; OS: overall survival; NA: data not available.
Higher expression of Notch components was associated with reduced overall survival The original details of the overall survival of each study were shown in Table 3. Two studies focused on DLL4 [11,14], one study focused on Hes1 [9], three studies focused on Notch3 [10,12,13] and one study was involved in Hey1 [10]. All these four proteins were parts of Notch signaling. And the positive detection or high expression of these Notch components could imply the activation of Notch signaling to a certain extent. The overall survival of patients who expressed Notch components was analyzed and shown in Fig. 4. The analysis indicated that higher expression of Notch
components would lead to poor survival (Fig. 4a, pooled HR: 2.06, CI: 1.60e2.66, P < 0.00001; Fig. 4b, pooled HR: 2.11, CI: 1.64e2.73, P < 0.00001). Sensitivity analysis and publication bias The HRs of all studies analyzed were larger than 2.00. Sensitivity analysis showed that the pooled HR was not changed significantly with any studies removed. Fig. 5 showed four funnel plots (a, b, c, d) of studies analyzed above, which were respectively corresponding to the analysis in Fig. 3a, b, 4a, b. All studies located inside the 95% CIs, with an even
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
4
J. Ye et al. / Pancreatology xxx (2018) 1e8
Table 2 Notch expression detected by IHC from 6 studies. study
Detection
Zhou, J. X. et al. Drouillard, A. et al.
Cao, F. et al. Chen, H.T. et al. Mann, C.D. et al.
Doucas, H. et al.
Location
IHC IHC IHC IHC IHC IHC IHC IHC IHC IHC IHC IHC
Gene
Cytoplasm, cell membrane NA NA NA NA tumor stromal vessels Nuclear Nuclear Nuclear Nuclear Nuclear Nuclear
Notch3 DLL4 Notch1 Notch3 Hes1 DLL4 Notch1 Notch3 Notch4 Hes1 Hey1 Notch3
Normal
patients
Low/ND
High/D
Low/ND
High/D
NA NA NA NA NA NA 35 35 35 27 35 12
NA NA NA NA NA NA 0 0 0 8 0 0
46 3 7 21 19 50 46 27 52 9 43 13
55 86 82 68 53 32 46 65 40 83 49 10
Abbreviation: IHC: Immunohistochemistry; Low: low expression; ND: not detected; High: high expression; D: detected.
Table 3 The details of overall survival and Notch expression from 6 studies. study
Patients for OS analysis
Gene detected
Detection IHC score
Patient number
Multivariate analysis, hazard ratio (95% CI)
P Univariate analysis, hazard value ratio (95% CI)
P value
Low/ High/ ND D Zhou, J. X. 101 et al. Drouillard, A. 83 et al. Cao, F. et al. 72 Chen, H.T. et al. Mann, C.D. et al. Doucas, H. et al.
Notch3
IHC
DLL4
IHC
Hes1
IHC
82
DLL4
IHC
42 42 23
Notch3 Hey1 Notch3
IHC IHC IHC
Low (-, þ), high (þþ, þþþ) Low (<50%), high (>50%) Low (<25%), high (>25%) Low (<50%), high (>50%) – (<10%), þ (>10%) – (<10%), þ (>10%) – (<5%), þ (>5%)
46
55
1.960 (1.181e3.252)
0.009 2.927 (1.876e4.569)
<0.001
33
50
2.05 (1.1e3.6)
0.02
0.005
19
53
2.012 (1.549e10.214)
0.001 2.154 (1.987e11.212)
<0.001
50
32
2.24 (1.14e4.38)
0.019 NA
NA
22 32 13
20 11 10
2.671 (1.030e6.928) 3.591 (1.441e8.946) 1.95 (1.21e3.16)
0.053 2.541 (1.213e5.324) 0.006 2.998 (1.326e6.778) 0.036 NA
0.013 0.008 NA
2.12 (1.34e3.63)
Abbreviation: IHC: Immunohistochemistry; OS: overall survival; Low: low expression; ND: not detected; High: high expression; D: detected.
distribution around the vertical, indicating no obvious publication bias. Discussion Pancreatic ductal adenocarcinoma (PDAC) is a gastrointestinal tumor with a high degree of malignancy, which is difficult to be diagnosed and treated. Many dysregulated signal pathways have been characterized in PDAC, such as Notch signaling [21]. As our analysis presented, Notch signaling was over-expressed in some patients and associated with a lower survival rate. Not all patients over-express the same Notch proteins, but up-regulation of any Notch components, especially Notch3 and DLL4, should be an alarm. Notch signal pathway not only works as an accelerator to PDAC as we introduced, but also induces chemoresistance to antineoplasm drugs like gemcitabine [22,23]. Therefore, targeting Notch signaling is a feasible way to develop a novel therapeutic treatment for PADC. Generally, activation of Notch signaling was associated with poor outcome. Deficiency of Notch2 but not Notch1 stopped pancreatic intraepithelial neoplasia (PanIN) progression, prolonged survival, and led to a phenotypical switch toward anaplastic pancreatic cancer with epithelial-mesenchymal transition [24]. Aberrantly high expressed Jagged2 could enhance PDAC both migration and invasion with affecting cell proliferation, similar to Notch1 [25]. Increased expression of Notch1 and Hes1 was discovered in gemcitabine-treated cell lines and led to treatment failure [26]. Many proteins in Notch signaling were alternative
therapeutic targets. A Notch2/Notch3 antagonist (tarextumab) inhibited tumor growth and decreased tumor-initiating cell frequency in pancreatic tumor xenografts [27]. Inhibition of MAML (an essential co-activator of the canonical Notch signaling-mediated transcription) could delay PanIN formation [28]. According to our analysis results, Notch3 was detected only in PDAC patients by IHC, and higher Notch3 expression showed worse survival [10,12,13]. In the study of Eto, K., et al. [23], it demonstrated an extremely similar result that low Notch3 mRNA expression was significantly associated with longer overall survival in a multivariate analysis (P < 0.0094). Also, our analysis implied that higher Notch3 and DLL4 expression indicated poor survival [11,14]. Thus, Notch3 and DLL4 seem to be the most promising therapeutic target of Notch signal pathway in PDAC. Many researchers focused on one of Notch signaling restrict enzyme, g-secretase. Yabuuchi, S., et al. [29], showed that PF03084014, a selective g-secretase inhibitor, had greater antitumor activity in combination with gemcitabine in PDAC xenografts. PF-03084014 inhibited the cleavage of the nuclear Notch1 intracellular domain and targeted Hes-1 and Hey-1. PF-03084014 also targeted cancer stem cells within pancreatic tumors. Similar consequences were demonstrated in studies using other g-secretase inhibitors on PDAC xenografts [30e32]. But only focusing on gsecretase seemed to be not good enough. According to a phase II study of a g-secretase inhibitor RO4929097 in patients with previously treated metastatic PDAC, RO4929097 was well-tolerated [33]. And the use of g-secretase inhibitor led to intestinal goblet cell hyperplasia, thymus atrophy, a decrease in lymphocytes, and
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
J. Ye et al. / Pancreatology xxx (2018) 1e8
5
Fig. 2. Forest plot of the comparison of Notch1, Notch3 and Hes1 expression between PDAC patients and normal pancreas. Events: Notch proteins were detected or under high expression in patients. MeH: Mantel-Haenszel method; CI: confidence interval.
Fig. 3. (a) Forest plot of overall survival between PDAC patients with high/detected Notch3 and low/not detected. (b) Forest plot of overall survival between PDAC patients with high/detected DLL4 and low/not detected. CI: confidence interval.
alterations in hair color [34]. Because g-secretase is only one of the many enzymes in Notch signaling and it also participates in many other pathways, the therapeutic target could be more specific and effective like Notch3 and DLL4 in our analytic results. In fact, cancers near pancreas shared a similar association with Notch signaling, like gastric cancer, colon cancer, colorectal cancer,
and cholangiocarcinoma. According to a meta-analysis of Notch and gastric cancers [35], higher expression of Notch3 and Jagged1 was found in diffuse-type gastric cancer. Jagged1 was also significantly over-expressed in poor differentiation type. And overexpression DLL4 was associated with advanced T stage, N stage and TNM stage in gastric cancer patients. In colorectal cancer, high
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
6
J. Ye et al. / Pancreatology xxx (2018) 1e8
Fig. 4. Forest plot of overall survival between PDAC patients with high/detected Notch components and low/not detected. (a) Notch components including Notch3, DLL4, and Hes1; (b) Notch components including Notch3, DLL4, Hes1 and Hey1.
Fig. 5. Funnel plot illustrating meta-analysis of overall survival. SE: standard error; OR: odds ratio. The circles represent the articles. (a) Corresponding to the three studies in Fig. 3a. (b). Corresponding to the two studies in Fig. 3b. (c) Corresponding to the six studies in Fig. 4a. (d). Corresponding to the six studies in Fig. 4b.
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
J. Ye et al. / Pancreatology xxx (2018) 1e8
expression of Hes1 was found significantly correlated with distal metastasis in 320 patient samples [36]. Up-regulation of DLL4 was closely related to metastasis and drug resistance in colorectal and colon cancer, leading to poor prognosis [37,38]. Similarly, Notch1 and Hes1 were found increased in colon cancer cells which were resistant to regorafenib [39]. Especially, nuclear Notch3 expression was associated with tumor recurrence in patients with stage II and III colorectal cancers [40]. Through gene profile and cell surface proteome analysis, Notch signaling was found involved in cholangiocarcinoma [41,42]. More studies illustrated the same result that excessive activation of Notch signaling promotes the development of digestive cancers like Notch signaling in PDAC. However, more and more research found Notch signaling pathway had different effects in different situations of PDAC. A study pointed out that Notch signaling might inhibit PanIN development at an early stage but promote PanIN progression at later stages once lesions are established [15]. Recent studies have also demonstrated tumor suppressive effects of Notch signaling pathway. A study found that missing Notch expression in the malignant tumor and loss of Notch1 in a model of K-ras-induced PDAC led to tumor progression, suggesting that the Notch1 can serve as a tumor suppressor in PDAC [16,17]. To sum up, the effect of Notch signaling pathway in PDAC is complex. It might be related to the crosstalk. PDAC progression is associated with various signaling pathway like KRAS signaling, WNT signaling, TGF-beta signaling. In the field of molecular biology, the exact role of the Notch pathway in pancreatic cancer is still a controversy. Therefore, considering the clinical situation, we need to analyze clinical data to conclude the association between Notch and PADC patient. Adopting only six studies and 463 patients and 47 health controls in this meta-analysis may be a limitation that should be mentioned, but all six studies elucidated similar results. Moreover, there was no publication bias as our analysis shown. Therefore, the results of this meta-analysis should be convincing. But if possible, the data covered should be enlarged. Our study is the first meta-analysis to review all articles about the effects of Notch on human PDAC to provide the scientific foundations for policy, medicine, social science, business, ecology, and other fields. The results demonstrated the idea that higher expression of Notch proteins indicates poor survival of patients with PDAC, which will help understand the function of Notch in PDAC and provide a possible prognostic index of PDAC and even a therapeutic target based on a more significant sample number. Conflicts of interest
[3] [4] [5]
[6] [7] [8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
All the authors have no competing financial interests. [24]
Acknowledgments This research was supported by the Natural Science Foundation of China (No.81470831), NSFC-Guangdong Province Joint Key Project (U1401223), Science and Technology Planning Project of Guangdong Province, China (No. 2016B090919019), Science and Technology Planning Project of Guangzhou City, China (201604010057), Innovative experiment program of college students of Guangdong Province, China (No. 201612121034, 201812121036, 201812121110, 201812121231, 201812121234). References [1] Stewart B, Wild CP. World cancer report 2014. World Health Organization; 2014. [2] Global, regional, and national life expectancy, all-cause mortality, and cause-
[25]
[26] [27]
[28]
[29]
[30]
7
specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the global burden of disease study 2015. Lancet (London, England) 2016;388:1459e544. Ryan DP, Hong TS, Bardeesy N. Pancreatic adenocarcinoma. N Engl J Med 2014;371:2140e1. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, et al. Cancer statistics in China, 2015. CA A Cancer J Clin 2016;66:115e32. Yeo CJ, Cameron JL, Lillemoe KD, Sitzmann JV, Hruban RH, Goodman SN, et al. Pancreaticoduodenectomy for cancer of the head of the pancreas. 201 patients. Ann Surg 1995;221:721e31. discussion 731-723. Kopan R, Ilagan MX. The canonical notch signaling pathway: unfolding the activation mechanism. Cell 2009;137:216e33. Bray SJ. Notch signalling: a simple pathway becomes complex. Nat Rev Mol Cell Biol 2006;7:678e89. Damaskos C, Karatzas T, Kostakis ID, Nikolidakis L, Kostakis A, Kouraklis G. Nuclear receptors in pancreatic tumor cells. Anticancer Res 2014;34: 6897e911. Cao F, Li J, Sun H, Liu S, Cui Y, Li F. Hes 1 is essential for chemoresistance induced by stellate cells and is associated with poor prognosis in pancreatic cancer. Oncol Rep 2015;33:1883e9. Mann CD, Bastianpillai C, Neal CP, Masood MM, Jones DJ, Teichert F, et al. Notch3 and hey-1 as prognostic biomarkers in pancreatic adenocarcinoma. PloS One 2012;7, e51119. Chen HT, Cai QC, Zheng JM, Man XH, Jiang H, Song B, et al. High expression of delta-like ligand 4 predicts poor prognosis after curative resection for pancreatic cancer. Ann Surg Oncol 2012;19(Suppl 3):S464e74. Doucas H, Mann CD, Sutton CD, Garcea G, Neal CP, Berry DP, et al. Expression of nuclear notch3 in pancreatic adenocarcinomas is associated with adverse clinical features, and correlates with the expression of stat3 and phosphorylated akt. J Surg Oncol 2008;97:63e8. Zhou JX, Zhou L, Li QJ, Feng W, Wang PM, Li EF, et al. Association between high levels of notch3 expression and high invasion and poor overall survival rates in pancreatic ductal adenocarcinoma. Oncol Rep 2016;36:2893e901. Drouillard A, Puleo F, Bachet JB, Ouazzani S, Calomme A, Demetter P, et al. Dll4 expression is a prognostic marker and may predict gemcitabine benefit in resected pancreatic cancer. Br J Canc 2016;115:1245e52. Avila JL, Kissil JL. Notch signaling in pancreatic cancer: oncogene or tumor suppressor? Trends Mol Med 2013;19:320e7. Krausch M, Kroepil F, Lehwald N, Lachenmayer A, Schott M, Anlauf M, et al. Notch 1 tumor expression is lacking in highly proliferative pancreatic neuroendocrine tumors. Endocrine 2013;44:182e6. Hanlon L, Avila JL, Demarest RM, Troutman S, Allen M, Ratti F, et al. Notch1 functions as a tumor suppressor in a model of k-ras-induced pancreatic ductal adenocarcinoma. Cancer Res 2010;70:4280e6. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the prisma statement. Int J Surg 2010;8: 336e41. Knobloch K, Yoon U, Vogt PM. Preferred reporting items for systematic reviews and meta-analyses (prisma) statement and publication bias. J Craniomaxillo-Fac Surg: Official Pub European Assoc Cranio-Maxillo-Fac Surg 2011;39:91e2. Stang A. Critical evaluation of the newcastle-ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603e5. Knudsen ES, O'Reilly EM, Brody JR, Witkiewicz AK. Genetic diversity of pancreatic ductal adenocarcinoma and opportunities for precision medicine. Gastroenterology 2015;150(1):48e63. Du X, Zhao YP, Zhang TP, Zhou L, Chen G, Wang TX, et al. Alteration of the intrinsic apoptosis pathway is involved in notch-induced chemoresistance to gemcitabine in pancreatic cancer. Arch Med Res 2014;45:15e20. Eto K, Kawakami H, Kuwatani M, Kudo T, Abe Y, Kawahata S, et al. Human equilibrative nucleoside transporter 1 and notch3 can predict gemcitabine effects in patients with unresectable pancreatic cancer. Br J Canc 2013;108: 1488e94. Mazur PK, Einw€ achter H, Lee M, Sipos B, Nakhai H, Rad R, et al. Notch2 is required for progression of pancreatic intraepithelial neoplasia and development of pancreatic ductal adenocarcinoma. Proc Natl Acad Sci Unit States Am 2010;107:13438e43. Hu Y, Su H, Li X, Guo G, Cheng L, Qin R, et al. The notch ligand jagged2 promotes pancreatic cancer metastasis independent of notch signaling activation. Mol Canc Therapeut 2015;14:289e97. Lee JY, Song SY, Park JY. Notch pathway activation is associated with pancreatic cancer treatment failure. Pancreatology 2014;14:48e53. Yen WC, Fischer MM, Axelrod F, Bond C, Cain J, Cancilla B, et al. Targeting notch signaling with a notch2/notch3 antagonist (tarextumab) inhibits tumor growth and decreases tumor-initiating cell frequency. Clin Canc Res: Official J Am Assoc Canc Res 2015;21:2084e95. Thomas MM, Zhang Y, Mathew E, Kane KT, Maillard I, Pasca di Magliano M. Epithelial notch signaling is a limiting step for pancreatic carcinogenesis. BMC Canc 2014;14:862. Yabuuchi S, Pai SG, Campbell NR, de Wilde RF, De Oliveira E, Korangath P, et al. Notch signaling pathway targeted therapy suppresses tumor progression and metastatic spread in pancreatic cancer. Cancer Lett 2013;335:41e51. Palagani V, Bozko P, El Khatib M, Belahmer H, Giese N, Sipos B, et al. Combined inhibition of notch and jak/stat is superior to monotherapies and impairs
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014
8
J. Ye et al. / Pancreatology xxx (2018) 1e8
pancreatic cancer progression. Carcinogenesis 2014;35:859e66. [31] Palagani V, El Khatib M, Kossatz U, Bozko P, Muller MR, Manns MP, et al. Epithelial mesenchymal transition and pancreatic tumor initiating cd44þ/ epcamþ cells are inhibited by gamma-secretase inhibitor ix. PloS One 2012;7, e46514. [32] Mizuma M, Rasheed ZA, Yabuuchi S, Omura N, Campbell NR, de Wilde RF, et al. The gamma secretase inhibitor mrk-003 attenuates pancreatic cancer growth in preclinical models. Mol Canc Therapeut 2012;11:1999e2009. [33] De Jesus-Acosta A, Laheru D, Maitra A, Arcaroli J, Rudek MA, Dasari A, et al. A phase ii study of the gamma secretase inhibitor ro4929097 in patients with previously treated metastatic pancreatic adenocarcinoma. Invest N Drugs 2014;32:739e45. [34] Panza F, Frisardi V, Imbimbo BP, Capurso C, Logroscino G, Sancarlo D, et al. Review: gamma-secretase inhibitors for the treatment of alzheimer's disease: the current state. CNS Neurosci Ther 2010;16:272e84. [35] Du X, Cheng Z, Wang YH, Guo ZH, Zhang SQ, Hu JK, et al. Role of notch signaling pathway in gastric cancer: a meta-analysis of the literature. World J Gastroenterol 2014;20:9191e9. [36] Yuan R, Ke J, Sun L, He Z, Zou Y, He X, et al. Hes1 promotes metastasis and predicts poor survival in patients with colorectal cancer. Clin Exp Metastasis
2015;32:169e79. [37] Negri FV, Crafa P, Pedrazzi G, Bozzetti C, Lagrasta C, Gardini G, et al. Strong notch activation hinders bevacizumab efficacy in advanced colorectal cancer. Future Oncol (London, England) 2015;11:3167e74. [38] Zheng Y, Hao Z, Ding Y, Wang Q, Li S, Xiao G, et al. Expression of delta-like 4 (drosophila) and vascular endothelial growth factor a in colon cancer and association with tumour angiogenesis. J Int Med Res 2015;43:535e43. [39] Mirone G, Perna S, Shukla A, Marfe G. Involvement of notch-1 in resistance to regorafenib in colon cancer cells. J Cell Physiol 2015;231(5):1097e105. [40] Ozawa T, Kazama S, Akiyoshi T, Murono K, Yoneyama S, Tanaka T, et al. Nuclear notch3 expression is associated with tumor recurrence in patients with stage ii and iii colorectal cancer. Ann Surg Oncol 2014;21:2650e8. [41] Xue TC, Zhang BH, Ye SL, Ren ZG. Differentially expressed gene profiles of intrahepatic cholangiocarcinoma, hepatocellular carcinoma, and combined hepatocellular-cholangiocarcinoma by integrated microarray analysis. Tumour Biol: J Int Soc Oncodevelopmental Biol Med 2015;36:5891e9. [42] Stephenson B, Shimwell N, Humphreys E, Ward D, Adams D, Martin A, et al. Quantitative assessment of the cell surface proteome to identify novel therapeutic targets in cholangiocarcinoma. Lancet (London, England) 2015;385(Suppl 1):S94.
Please cite this article in press as: Ye J, et al., Higher notch expression implies poor survival in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis, Pancreatology (2018), https://doi.org/10.1016/j.pan.2018.09.014