Impact of TROP2 expression on prognosis in solid tumors: A Systematic Review and Meta-analysis

Over-expression of TROP2 (the trophoblast cell surface antigen 2) was reported to predict poor prognosis in various solid tumors in number of studies. However, the results remained not comprehensive. Therefore, we here carried out this meta-analysis of relevant studies published on this topic to quantitatively evaluate the clinicopathological significance of TROP2 in solid tumors. Relevant articles were identified through searching the PubMed, Web of Science and Embase database. The primary outcomes were overall survival (OS) and disease-free survival (DFS). In this meta-analysis, 16 studies involving 2,569 participants were included, and we drew the conclusion that TROP2 overexpression was significantly associated with poor OS (pooled HR = 1.896, 95% CI = 1.599–2.247, P < 0.001) and short DFS (pooled HR = 2.336, 95% CI = 1.596–3.419, P < 0.001). Furthermore, the subgroup analysis revealed that the associations between TROP2 overexpression and the outcome endpoints (OS or DFS) were significant in in patients with female genital system neoplasms, as well in gastrointestine neoplasms. In addition, subgroup analysis found no difference HR across populations of different descent.Taken together, TROP2 overexpression was associated with poor survival in human solid tumors. TROP2 may be a valuable prognosis predictive biomarker and a potential therapeutic target in human solid tumors.

Inclusion criteria. To be eligible for inclusion in this meta-analysis and data extraction, studies had to: (a) evaluation of TROP2 expression for predicting prognosis in cancer, (b) provide hazard ratios (HRs) with 95% confidence intervals (CIs) or enable calculation of these statistics from the data presented (c) classify TROP2 expression as "high" and "low" or "positive" and "negative". (d) studies published in English.
Exclusion criteria. Exclusion criteria were: (a) literatures published as letters, editorials, abstracts, reviews, case reports and expert opinions; (b) experiments performed in vitro or in vivo, but not based on patients; (c) articles without the HRs and 95% CI or K-M survival curves dealing with overall survival, disease-free survival; (d) The follow-up duration was shorter than 3 years. Data extraction. All data from the included studies were extracted independently and carefully by two reviewers using a standardized form. Disagreement was resolved through independently extracting data from the original article by the third author, and consensus was reached by discussions. The meta-analysis of TROP2 expression was based on two outcome endpoints: OS and DFS. Several different parameters, if reported, were extracted from each paper, including the first author's surname, publication year, country of origin, number of patients analyzed, types of measurement, and score for TROP2 assessment, cut-off values to determine TROP2 overexpression, OS and DFS. The main features of these eligible studies are summarized in Table 1. OS was defined as the interval between initial diagnosis and death. DFS was the interval between initial diagnosis and recurrent or between disease progress and death. The multivariate HR was extracted to assess prognostic value of TROP2 expression. For the articles in which prognosis was plotted only as the Kaplan-Meier curves, the Engauge Digitizer V4.1 was then used to extract survival data, and the estimates of the HRs and 95% CIs were calculated by Tierney's method 30 . All studies were assessed by Newcastle-Ottawa Scale (NOS). The quality scores ranged from 6 to 8, suggesting that the methodological quality was high.
Statistical analysis. The data collected from each qualified paper of outcomes was used to evaluate the associations between TROP2 expression and solid cancer prognosis through meta-analysis. Pooled HRs and 95% CIs for two outcome endpoints (OS, DFS) were calculated. Subgroup analysis was performed when there were at least three studies in each subgroup. Statistical heterogeneity was assessed using the Q test, and a P value > 0.10 suggested a lack of heterogeneity among studies. We also quantified the effect of heterogeneity using , values of about 50% may be considered "moderate" and values of > 75% may be considered "high" 31 . According to the absence or presence of heterogeneity, random effects model or fixed effects model was used to merge the HR, respectively. Without statistical heterogeneity, a fixed effects model was employed to calculate the pooled HRs, otherwise random effects model was used 32 . Funnel plots and the Egger's test were employed to estimate the possible publication bias 33 . If a publication bias did exist, its influence on the overall effect was assessed by the Duval and Tweedie's trim and fill method 34 . Sensitivity analysis was also conducted to find out if certain single article could influence the overall result. Statistical analyses were conducted using Stata 14.0 (StataCorp, College Station, TX). P values for all comparisons were two-tailed and statistical significance was defined as p < 0.05 for all tests, except those for heterogeneity.

Results
Demographic characteristics. A total of 126 articles were retrieved by a literature search of the PubMed, Embase, and Web of Science databases, using different combinations of key terms. As showed in the search flow diagram (Fig. 1), 126 records were initially retrieved using the predefined search strategy. Because of repeated data, 50 records were removed. After browsing the retrieved titles and abstracts, 56 records were excluded due to no relevant endpoint provided. The remaining 22 records were downloaded as full-text and carefully accessed one by one. Among them, 6 studies were excluded, including one study that was experimental study, four that without prognosis data, one in chinese. As a result, 16 published studies including 2,569 patients that met the inclusion norm were finally selected for the meta-analysis, which assessed the relevance between TROP2 expression and solid tumor prognosis. The median sample-size was 97, with a wide range from 26 to 620. Among all cohorts, Mongoloid (n = 11) became the major race of literatures, followed by Caucasian (n = 5 Full-text articles excluded: • without prognosis data (n=4) •experimental studie (n=1) •study in chinese (n=1)

Studies included in the meta-analysis (n=16)
Full-text articles assessed for eligibility (n=22) Articles excluded because title and/or abstra ct were not relevant to the endpoint of the study(n=54)     Publication bias and sensitivity analysis. Begg's funnel plot and Egger's test were utilized to estimate the publication bias of the included literatures. The shapes of the funnel plots for the OS and DFS showed no evidence of obvious heterogeneity (Fig. 4), and Egger's tests revealed publication bias concerning OS (P = 0.061) but not DFS (P = 0.653). Therefore, we performed trim and fill method to make pooled HR more reliable, and the P value was also < 0.01(data not shown). Sensitivity analyses were further performed to determine the robustness of the results described above. No individual study dominated this meta-analysis, and the removal of any single study had no significant effect on the overall conclusion (Fig. 5).

Discussions
High TROP2 expression had been reported to promote cancer progression and predict of poor prognosis of cancer patients 13,18,24,25,35 . Many clinical studies investigated the prognostic value of TROP2 over-expression. Most of these studies, however, include only limited numbers of patients and their conclusions remain not comprehensive. This current meta-analysis is the first complete overview of all reported clinical studies exploring the impact of TROP2 expression on prognosis of many solid tumors. We systematically evaluated survival data of 2,569 solid tumor patients included in 16 different studies. Overall, these results clearly show that high TROP2 expression was a poor prognostic factor in solid tumors, with both results of poor OS (pooled HR = 1.896, 95% CI = 1.599-2.247, P < 0.001) and poor DFS (pooled HR = 2.336, 95% CI = 1.596-3.419, P < 0.001). Similarly, subgroup analysis revealed the associations between TROP2 overexpression and poor OS and DFS were significant within Mongoloid and Caucasian. When data was stratified according to cancer types, the results showed the prognostic value of TROP2 over-expression was significant in female genital system neoplasms and in gastrointestine neoplasms.
To our knowledge, the present study is the first and most full-scale meta-analysis systemically exploring the possible prognostic role of TROP2 up-regulation in solid tumors. Our quantitative results strongly supported the current mainstream viewpoint that an undesirable impact of TROP2 redundancy was correlated with the OS and DFS.
Additionally, several important implications in this meta-analysis were displayed. First, high TROP2 expression may be a general poor prognostic marker in solid tumors. In this meta-analysis, we included thirteen different cancer types, including gastric cancer 14,15 , nasopharyngeal carcinoma 16 , gallbladder cancer 17 , cervical cancer 18 , extranodal NK/T cell lymphoma 11 , lung cancer 9,19 , laryngeal squamous cell carcinoma 20 , Hilar Cholangiocarcinoma 24 , colon cancer 21,22 , pancreatic cancer 25 , squamous cell carcinoma of the oral cavity 26 , ovarian carcinoma 28 , endometrioid endometrial carcinoma 27 . The pooled results from these cancer types demonstrated that high TROP2 expression was associated with poor OS and DFS, and this finding can be extended to all solid tumors. Second, we demonstrated that high TROP2 expression correlated with poor OS and DFS in Mongoloid and Caucasian patients, as well in female genital system neoplasms and in gastrointestine neoplasms. Finally, it underlines the potential to develop TROP2 as a valuable therapeutic target and prognostic biomarker for solid tumors. Apart from the inspiring outcomes, limitations still lay in this quantitative meta-analysis. First of all, most of the included studies were designed as retrospective studies, and such studies are more likely to be published if they have positive results than if they have negative results. Furthermore, the method for assessing TROP2 expression and definition of TROP2 positivity were inconsistent. Besides, some studies did not provide complete data 36 or published in English 23 were excluded in statistics. Therefore, our estimate of the associations between increased TROP2 and outcomes may have been overestimated.
In conclusion, high TROP2 expression in solid tumor tissues association with poor survival was clearly demonstrated in the present meta-analysis. We suggest that TROP2 may be a useful prognostic biomarker and a promising therapeutic target for solid tumors. Nevertheless, further studies related to specific tumor types and perspectives are required to corroborate the clinical utility of TROP2 expression in solid tumors.