Prognostic Significance of Clinicopathological Factors Influencing Overall Survival and Event-Free Survival of Patients with Cervical Cancer: A Systematic Review and Meta-Analysis.

BACKGROUND Cervical cancer (CC) is the most frequent type of cancer among women and its poor prognosis is a main concern, while the prognostic factors for CC have still remained controversial. We conducted this systematic review and meta-analysis to identify the prognostic significance of clinicopathological factors, influencing overall survival (OS), and event-free survival (EFS) of CC patients. MATERIAL AND METHODS The electronic databases of PubMed, EmBase, and the Cochrane library were systematically searched for identification of eligible studies published until June 2021. The pooled hazard ratio (HR) with 95% confidence interval (CI) were calculated using the random-effects model. Sensitivity and subgroup analyses and assessment of publication bias were also conducted. RESULTS We selected 140 studies that involved 47 965 patients for the meta-analysis. The results revealed that age, cell type, depth of tumor invasion, the International Federation of Gynecology and Obstetrics stage, hemoglobin level, histological grade, leukocytosis, lymph node involvement, lymph-vascular space invasion, neutrophil-to-lymphocyte ratio, parametrial invasion, platelet-to-lymphocyte ratio, resection margin, squamous cell carcinoma antigen level, thrombocytosis, tumor grade, tumor size, and tumor volume were clinicopathological factors influencing OS and EFS of CC patients (P<0.05). CONCLUSIONS This study comprehensively identified the prognostic significance of clinicopathological factors, influencing OS, and EFS of CC patients. However, further large-scale prospective studies should be conducted to verify our findings and develop more accurate prognostic models for CC.

Background

Cervical cancer (CC) is a frequent gynecologic malignancy and is the primary cause of cancer-related deaths in women worldwide [1,2]. A total of 604 127 new cases and 341 831 CC-related deaths were reported in 2020, accounting for 7.7% of all cancer-related deaths in women [1]. The HPV infection rate is rising, particularly in developing countries, where the incidence and prevalence of CC are still high, which can be attributed to the lack of a universal and integrated vaccination program for CC [3,4]. The prognosis of CC could be improved by a variety of treatment strategies on the basis of the disease stage, metastasis, or recurrence [2,5]. The International Federation of Gynecology and Obstetrics (FIGO) staging system has been widely used for predicting the prognosis of CC patients, while the prognosis of CC patients with the same FIGO stage varies [6]. Several prognostic models have already been introduced to predict the prognosis of CC on the basis of tumor and demographic characteristics [7-10], but the practicality of these models is limited by uneven quality and various characteristics of clinical setting, outcomes, and predictors. Therefore, additional prognostic factors should be explored to improve the prognosis of CC patients.

We therefore attempted to construct a prognostic model using the previously defined factors to predict the prognosis of CC patients. Numerous systematic reviews and meta-analyses have been performed to identify the prognostic significance of other variables in estimating the rates of overall survival (OS) and event-free survival (EFS) [11-15]. However, the other clinicopathological characteristics influencing the prognosis of CC patients were not assessed. There is an urgent need to summarize the prognostic variables to establish more comprehensive prognostic models. We therefore conducted the present systematic review and meta-analysis to identify the prognostic factors for CC and we also investigated the prognostic significance of these factors for CC.

Material and Methods

Search Strategy and Selection Criteria

The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Statement was utilized, as described previously [16]. Studies on the prognostic significance of clinicopathological factors, influencing OS, and EFS of CC patients were selected, and the language was restricted to English. No restriction was placed on publication status, including published, in press, or in progress. The electronic databases of PubMed, EmBase, and the Cochrane library were systematically searched for retrieving potential studies published until June 2021 using the following text word or Medical Subject Heading terms: (“cervical cancer” OR “cervical carcinoma” OR “cervical intraepithelial neoplasia” OR “uterine cervix cancer”) AND (“prognosis” OR “prognostic” OR “survival” OR “recurrence”). We also manually searched the reference lists of relevant reviews and original articles to identify eligible studies.

The literature search and study selection were independently performed by 2 reviewers, and the inconsistencies between reviewers were resolved by group discussion until a consensus could be reached. The following inclusion criteria were considered: (1) Study design: prospective or retrospective studies; (2) Patients: all patients who were diagnosed with CC; (3) Exposure: the clinicopathological factors reported ≥3 studies, including patients’ age, cell type, depth of tumor invasion, FIGO stage, hemoglobin level, histological grade, leukocytosis, lymph node involvement (LNI), lymph-vascular space invasion (LVSI), neutrophil-to-lymphocyte ratio (NLR), parametrial invasion, platelet-to-lymphocyte ratio (PLR), resection margin, squamous cell carcinoma antigen (SCCA), thrombocytosis, tumor grade, tumor size, and tumor volume; and (4) Clinical outcomes: OS or EFS. Reviews and abstracts were excluded because they contain no original data or have an unclear definition of prognostic factors.

Data Collection and Quality Assessment

Two reviewers independently abstracted the following items: characteristics of studies (the first author’s full name, year of publication, the first author’s country of residence, and study design), sample size, mean or median age, FIGO stage, follow-up duration, clinical outcomes, and prognostic factors. Then, these 2 reviewers assessed the quality of each study using the Newcastle-Ottawa Scale (NOS) score, which ranges from 0–9 stars for assessment of quality of each study [17]. Studies were classified into low quality (0–6 stars), medium quality (7–8 stars), and high quality (9 stars). Any disagreement between reviewers for data collection and quality assessment was resolved via reading the full-text of the included studies by the third reviewer.

Statistical Analysis

The prognostic factors, influencing OS and EFS of CC patients were presented as hazard ratio (HR) and 95% confidence interval (CI) for each individual study, and the pooled HRs and 95% CIs were calculated using the random-effects model, as described elsewhere [18,19]. Heterogeneity among the included studies was assessed using the Cochran’s Q-statistic and the I-statistic, and a significant heterogeneity was defined as I ≥50.0% or P<0.10 [20,21]. To determine sources of heterogeneity, we performed a leave-one-out sensitivity analysis via exclusion of individual studies one at a time, and the pooled estimates were recalculated for the remaining studies [22]. Subgroup analyses were undertaken on the basis of the first author’s country of residence, FIGO stage, cutoff value, and study quality, and the subgroups were calculated using the chi-square test to explore the differences in the estimates between subgroups [23]. The Eastern countries contained Asia, while Western countries including Europe, America, and Oceania. Assessment of publication bias was carried out by using Egger’s and Begg’s tests, which compared the summary estimate of each study to its precision for outcomes that were reported in more than 5 studies [24,25]. The trim and fill method was applied to adjust pooled results if significant publication bias was observed [26]. Two-sided P<0.05 was regarded as statistically significant. The STATA 10.0 software was used to conduct the statistical analyses (Stata Corporation, College Station, TX, USA).

Results

Literature Search

The search strategy resulted in retrieving 18 912 articles, and 9141 articles were retained after exclusion of 9771 studies owning to duplicate publication. Then, 8762 studies were excluded because of irrelevant titles, the review of the reference lists of potentially relevant studies indicated 21 studies, and a total of 380 studies were retrieved for further full-text evaluations. Next, 240 studies were removed because they investigated other interventions (n=169), had inadequate outcomes (n=46), and were review articles (n=25). The remaining 140 studies were selected for the final meta-analysis (Figure 1), and characteristics of the eligible studies are presented in Table 1 [27-166].

Figure 1

The PRISMA flowchart for the literature search and the study selection.

Table 1

The baseline characteristics of included studies.

Study	Country	Study design	Sample size	Age (years)	FIGO stage	Follow-up (years)	Reported outcomes	Prognostic factors	NOS score
Sevin 1995 [27]	USA	Retro	301	43.5	I–II	5.0	DFS	DI, TS, LVSI, LNI, TV, FIGO, RM, CT, TG, age	6
Werner-Wasik 1995 [28]	USA	Retro	125	55.0	I–II	5.0	DFS	LNI, LVSI, PI, He, TS, FIGO, CT, TG	5
Tsai 1999 [29]	China	Retro	222	50.0	I–II	5.0	DFS	FIGO, TS, age, CT, SCC, He, LNI, PI, LVSI, RM	6
Lai 1999 [30]	China	Retro	891	NA	I–II	5.0	DFS	TG, FIGO, TS,DI	7
Nakanishi 2000 [31]	Japan	Retro	509	49.3	I	9.3	OS, DFS	CT, LNI, and TS	6
Hernandez 2000 [32]	USA	Retro	291	49.7	II–IV	5.0	PFS	Th, LNI, TS, age, and FIGO	7
Alfsen 2001 [33]	Norway	Retro	505	53.0	I–IV	5.0	OS	CT, LVSI, LNI, and age	7
Flores-Luna 2001 [34]	Mexico	Retro	378	52.2	I–IV	12.5	OS	FIGO, TG, TS, and age	5
Trattner 2001 [35]	Austria	Retro	113	46.1	I–II	4.7	OS	TV, LNI, LVSI, FIGO, PI, RM, CT, TG, and age	5
Yanoh 2001 [36]	Japan	Retro	751	45.0	I	> 5.0	DFS	LNI, PI, TS, DI, and LVSI	6
Takeda 2002 [37]	Japan	Retro	187	48.2	I–II	6.9	OS	FIGO, CT, LVSI, TS, DI, PI, and LNI	6
Gasinska 2002 [38]	Poland	Retro	152	55.0	I–III	2.2	OS	Age, TG, and He	6
Martin-Loeches 2002 [39]	Spain	Retro	114	49.1	I–II	10.0	OS	TS, TV, DI	5
Brun 2003 [40]	France	Retro	308	53.0	I–IV	7.8	OS	Age, TG, and PI	6
Morice 2003 [41]	France	Retro	193	37.0	I–II	5.0	OS	FIGO, TS, LVSI, and LNI	6
Kodaira 2003 [42]	Japan	Retro	164	68.0	II–III	1.9	DFS	TV, LNI, and FIGO	6
Grisaru 2003 [43]	Canada	Pro	871	42.1	I	4.1	DFS	LNI, TG, LVSI, RM, and CT	7
Huang 2003 [44]	China	Pro	157	44.0	I–II	5.0	OS, DFS	TS, age, CT	6
Shinohara 2004 [45]	Japan	Retro	130	49.0	I–II	14.4	DFS	LVSI, LNI, and DI	6
Ho 2004 [46]	China	Retro	197	47.4	I–II	5.8	OS, DFS	Age, FIGO, CT, TG, TS, DI, LVSI, LNI, PI	5
Ayhan 2004 [47]	Turkey	Retro	393	48.5	I	2.6	OS, DFS	TS, LVSI, PI,	6
Choi 2006 [48]	Korea	Retro	85	50.0	I–IV	3.0	OS, DFS	Age, CT, FIGO, TS, LNI, SCC, and He	5
Chittithaworn 2007 [49]	Thailand	Retro	205	44.2	I	4.7	DFS	DI, LVSI, RM, and LNI	5
Grigiene 2007 [50]	Lithuania	Retro	162	52.0	II–III	2.7	OS, DFS	FIGO, He	7
Horn 2007 [51]	Germany	Retro	245	43.0	II	4.5	OS	TS, LNI, FIGO	6
Atahan 2007 [52]	Turkey	Retro	183	54.0	I–III	3.8	OS, DFS	Age, PI, FIGO, TS, CT, LNI	7
Garcia-Arias 2007 [53]	Mexico	Retro	294	49.4	I–IV	2.3	OS	Le, He, age, CT, and FIGO	7
Choi 2008 [54]	Korea	Retro	143	58.0	I–IV	2.2	PFS	FIGO, TS	6
Behtash 2009 [55]	Iran	Retro	203	49.8	I–II	3.5	OS, DFS	Age, CT, FIGO, TS, LNI, PI, LVSI, DI	6
Jacobson 2009 [56]	USA	Retro	436	52.3	I–IV	8.0	OS	FIGO, CT	7
Zusterzeel 2009 [57]	Netherlands	Retro	167	42.0	I–IV	2.8	OS, DFS	FIGO, CT, TG, LVSI, DI, TS	7
Polterauer 2010 [58]	Austria	Retro	88	49.9	I–IV	3.1	OS, DFS	FIGO, TG, CT	7
Munagala 2010 [59]	India	Retro	89	46.0	I–III	5.0–7.0	OS, DFS/PFS	Age, FIGO, LNI, PI, CT, TG, and TS	6
Huang 2010 [60]	China	Retro	960	45.0	I–II	5.0	OS	FIGO,SCC, DI, PI	6
Touboul 2010 [61]	France	Retro	150	47.0	I–IV	3.6	OS	FIGO, CT, RM, LNI	7
Horn 2010 [62]	Germany	Retro	194	44.0	I–II	5.1	OS	LNI, TG, FIGO	6
Kodama 2010 [63]	Japan	Retro	97	46.0	I–II	8.4	OS, DFS	Age, FIGO, DI, TS, PI, LVSI, LNI	5
Lee 2010 [64]	Korea	Retro	134	58.0	II–IV	3.2	OS, PFS	FIGO	5
Tseng 2010 [65]	China	Pro	251	48.6	II–IV	6.3	OS	SCC, TS, PI, LNI	6
Nugent 2010 [66]	USA	Retro	111	51.0	I–IV	1.4	OS, PFS	FIGO	6
Srisomboon 2011 [67]	Thailand	Retro	680	44.5	I	4.0	DFS	LNI, LVSI, CT, DI, PI, TG, RM	6
Seamon 2011 [68]	USA	Retro	381	47.0	I–IV	3.3	OS, DFS	FIGO, CT, TG	7
Polterauer 2011 [69]	Austria	Retro	178	49.2	I–IV	3.8	OS, DFS	FIGO, LNI, TG, age, CT	7
Mabuchi 2011 [70]	Japan	Retro/Pro	536	57.5	I–IV	6.4	OS, PFS	Age, FIGO, CT, TS	6
Min 2011 [71]	China	Retro	88	NA	I–II	5.0	OS	Age, TS, CT, TG, FIGO, LNI	5
Biewenga 2011 [72]	Netherlands	Retro	710	41.0	I–II	5.2	DFS	CT, TG, DI, PI, LNI, LVSI, RM	7
Polterauer 2012 [73]	Austria	Retro	528	47.9	I–IV	3.8	OS	Age, FIGO, TS, CT, LNI, PI	7
Kim 2012 [74]	Korea	Retro	174	NA	I–IV	2.5	OS, PFS	FIGO, LNI, TS	6
Lee 2012 [75]	Korea	Retro	1,061	50.0	I–IV	4.4	OS, PFS	NLR, FIGO, CT	7
Okazawa 2012 [76]	Japan	Retro	311	51.0	I–II	5.2	PFS	Age, CT, LNI, PI, RM, DI, LVSI, TS, He	7
Wang 2012 [77]	China	Retro	179	47.0	I–IV	4.3	OS, DFS	FIGO, LNI, RM	6
Yan 2012 [78]	China	Retro	148	42.0	I	2.3	OS	Age, CT, TG, TS, DI, LVSI, LNI	5
Cibula 2012 [79]	Czech Republic	Retro	645	46.0	I–II	3.3	OS, DFS	Age, FIGO, PI, LNI	6
Singh 2012 [80]	Australia	Retro	196	NA	I–II	6.1	OS, DFS	Age, LVSI, LNI, PI, TS, DI	7
Wang 2013 [81]	China	Retro	424	NA	I–II	7.0	DFS	Age, CT, TG, FIGO, LNI	5
Tsubamoto 2013 [82]	Japan	Retro	73	47.0	I–II	5.9	OS, DFS	Age, FIGO, CT, TS, LNI	6
Song 2013 [83]	Korea	Retro	268	57.0	I–IV	5.0	OS, DFS	FIGO, age, LNI, CT, He	6
Cho 2013 [84]	Korea	Retro	185	50.0	I–II	5.9	DFS	Age, FIGO, LNI, RM, PI, TS, DI, LVSI	6
Zhang 2014 [85]	China	Retro	460	44.0	I–II	5.8	OS, PFS	FIGO, LNI, NLR	7
Horn 2014 [86]	Germany	Retro	366	40.0	I	7.8	OS, DFS	TS, LNI, TG	7
Noh 2014 [87]	Korea	Retro	1,323	50.0	I–II	6.3	OS, DFS	CT, age, FIGO, TS, LNI, PI, LVSI, DI, RM	7
Yu 2014 [88]	China	Retro	153	NA	II	5.0	DFS	TS, LVSI, LNI	6
Liu 2014 [89]	China	Retro	184	46.0	I–II	5.8	OS, DFS	Age, TS, CT, TG, FIGO, DI, LVSI, LNI	6
Kawano 2015 [90]	Japan	Retro	286	63.6	I–IV	6.7	OS	Age, FIGO, PNI, CT, TS, He, Th	7
Ruengkhachorn 2015 [91]	Thailand	Retro	331	48.6	I–II	7.0	DFS	He, CT, FIGO, PNI, PI, RM, DI, LVSI	6
Bradbury 2015 [92]	UK	Retro	92	39.5	I	4.8	OS, PFS	Age, TS, CT, TG, LVSI, LNI, RM	7
Yuan 2015 [93]	China	Retro	38	40.4	I–II	5.0	OS, DFS	PI	6
Mizunuma 2015 [94]	Japan	Retro	56	65.1	I–IV	6.7	OS, PFS	FIGO, TS, He, NLR	6
Endo 2015 [95]	Japan	Retro	84	62.0	I–IV	6.7	OS	Age, CT, He, TS, LNI	6
Zhao 2015 [96]	China	Retro	220	NA	I–II	5.0	OS, DFS	Age, FIGO, TG, CT, DI, TS, LNI	7
Takatori 2015 [97]	Japan	Retro	33	42.0	I–II	2.8	OS	Age, FIGO, TS, SCC	5
Huang 2016 [98]	China	Retro	643	NA	I–II	3.1	OS, DFS	Age, CT, TG, TS, FIGO, DI, LVSI, LNI, PI, RM	7
Li 2016 [99]	China	Retro	347	51.6	I–II	3.1	OS, DFS	Age, CT, FIGO, TG, DI, LVSI, RM, LNI, PI, SCCA	7
Cho 2016 [100]	Korea	Retro	2,456	56.0	I–IV	5.4	OS, DFS	Age, FIGO, CT, TS, LNI, He, Le, NLR	7
Matsumiya 2016 [101]	Japan	Retro	54	55.0	I–IV	1.0	OS	CT	6
Usami 2016 [102]	Japan	Retro	111	51.0	I–IV	1.4	OS	Age, CT	6
Chen 2016 [103]	China	Retro	407	44.0	I–II	5.0	OS, DFS	Age, CT, TG, DI, LVSI, LNI, FIGO, PI, PLR, NLR	6
Oishi 2016 [104]	Japan	Retro	85	55.0	IV	0.8	OS	Age, CT, TS, TG, He, SCC	5
Onal 2016 [105]	Turkey	Retro	235	57.0	I–IV	5.8	OS, PFS	Age, FIGO, TS, LNI, NLR	7
Wu 2016 [106]	USA	Retro	71	49.0	I–IV	2.1	OS, PFS	FIGO, CT, TG	6
Xia 2016 [107]	China	Retro	274	43.0	I–II	2.4	OS, DFS	Age, FIGO, CT, TS, TG, DI, LVSI, RM, PI, LNI	6
Lee 2017 [108]	Korea	Retro	231	58.0	I–IV	2.3	OS, PFS	Age, LNI, FIGO, SCC, TV	7
Barquet-Muñoz 2017 [109]	Mexico	Retro	202	49.5	I–II	5.0	OS, DFS	Age, CT, TS, DI, LVSI, RM, PI, LNI	6
Jung 2017 [110]	Korea	Retro	1,113	48.7	I–II	7.6	OS, DFS	CT, FIGO, TS, DI, LNI, LVSI, PI, RM	7
Chung 2017 [111]	Korea	Retro	103	48.0	I–II	2.4	PFS	FIGO, TS, LNI, PI, DI, LVSI	5
Zheng 2017 [112]	China	Retro	795	49.5	I–II	5.2	OS	FIGO, He, TG, LVSI, LNI, TS, PI, RM	6
Obrzut 2017 [113]	Poland	Pro	102	48.0	I–II	10.0	OS, DFS	FIGO, CT, TG, LNI, LVSI, RM	6
Cho 2017 [114]	Korea	Retro	105	NA	II	4.8	PFS	Age, CT, TS, LNI, NLR	5
Chandeying 2017 [115]	Thailand	Retro	626	45.0	I	7.7	OS, DFS	CT, age, TS, FIGO, RM, PI, LNI, LVSI, DI	7
Yokoi 2017 [116]	Japan	Retro	249	61.5	II–IV	5.0	PFS	Age, FIGO, LNI, CT, He	7
Lim 2017 [117]	Korea	Retro	180	NA	I–II	5.0	OS, DFS	PI, LNI	5
Xu 2018 [118]	China	Retro	40	45.5	I–IV	5.0	OS	Age, FIGO, LNI, LVSI, DI, TS	6
Wen 2018 [119]	China	Retro	99	NA	II–IV	4.0	DFS	Age, TS, CT, FIGO, SCC, PI	6
Joo 2018 [120]	Korea	Retro	397	45.0	I–II	4.0	OS, DFS	CT, FIGO, LNI, PI, LVSI, DI, TS	6
Dai 2018 [121]	China	Retro	302	45.1	I–II	5.0	OS	FIGO, TS, TG, DI, LVSI, PI, LNI	6
Zhu 2018 [122]	China	Retro	365	45.0	I–II	3.7	OS, PFS	Age, DI, LNI, LVSI, PI	5
Zhou 2018 [123]	China	Retro	312	46.0	I–II	4.7	OS, DFS	Age, FIGO, TS, TG, DI, LVSI, LNI	5
Liu 2018 [124]	China	Retro	98	52.0	I–III	3.1	OS, PFS	TS, LNI	5
Xie 2018 [125]	China	Retro	810	46.3	I–II	5.0	OS	FIGO, LNI	5
Taarnhøj 2018 [126]	Denmark	Retro	1,523	NA	I	5.0	DFS	FIGO, CT, age, DI, LVSI	6
Zhang 2018 [127]	China	Retro	235	46.0	I–II	6.4	OS, PFS	Age, FIGO, TS, CT, LVSI, LNI, DI, NLR	7
Je 2018 [128]	Korea	Retro	1,069	49.0	I–II	5.0	DFS	CT, PI, LVSI, DI, TS, LNI	7
Ishikawa 2018 [129]	Japan	Retro	93	NA	I–II	10.0	OS, DFS	CT, TS, DI, LVSI, PI, LNI, RM	6
Kwon 2018 [130]	Korea	Retro	259	47.0	I–II	5.8	DFS	CT, LVSI	6
Zhu 2019 [131]	China	Retro	110	51.5	I–II	4.0	OS, PFS	Age, TS, LNI, FIGO, TG, Ly	6
Yan 2019 [132]	China	Retro	347	NA	I–II	3.3	OS, PFS	Age, FIGO, LNI, TG, LVSI, DI	6
Wang 2019 [133]	China	Retro	559	51.0	I–IV	3.3	DFS	Age, SCC, FIGO, TS, LNI	7
Farzaneh 2019 [134]	Iran	Retro	307	40.4	I–III	5.0	DFS	RM, NLR	5
Sawada 2019 [135]	Japan	Retro	107	46.0	I–II	4.8	OS	FIGO, CT, TS, LNI, PI	6
Khalkhali 2019 [136]	Iran	Retro	109	50.1	I–IV	3.2	OS	Age, FIGO	5
Yildirim 2019 [137]	Turkey	Retro	104	56.0	I–IV	4.4	OS, DFS	TS, FIGO, LNI	6
Gai 2019 [138]	China	Retro	79	51.0	I–IV	5.0	OS	FIGO, LNI, LVSI	6
Chen 2019 [139]	China	Retro	88	48.0	I–II	2.2	DFS	Age, CT, FIGO, TG, LVSI	5
Guani 2019 [140]	France	Pro	139	NA	I	3.0	DFS	LNI, CT, TS, FIGO, LVSI, age	5
Huang 2019 [141]	China	Retro	458	45.0	I–II	3.9	OS	Age, TG, TS, LNI, LVSI, FIGO, NLR	7
Queiroz 2019 [142]	Brazil	Retro	127	50.8	II–IV	4.1	OS, DFS	Age, CT, TS, LNI	5
Gillani 2019 [143]	Malaysia	Pro	3,797	57.3	I–II	6.1	OS	Age, FIGO, TS, LNI, CT	6
de Foucher 2019 [144]	France	Retro	501	54.0	I–II	3.0	OS, DFS	FIGO, LNI	6
Yoshino 2019 [145]	Japan	Retro	128	65.0	I–IV	2.5	OS	FIGO, CT	6
Zhang 2019 [146]	China	Retro	89	40.5	I–IV	4.8	OS	FIGO, TS, LNI, LVSI, DI	6
Seebacher 2019 [147]	Austria	Retro	116	52.1	I–IV	1.7	OS	Age, FIGO, CT, SCC	5
Holub 2019 [148]	Spain	Retro	151	52.8	I–IV	3.7	OS	TS, FIGO, age, NLR	6
Theplib 2020 [149]	Thailand	Retro	196	41.0	I	5.0	OS, DFS	LVSI, PI, LNI, DI	6
Maulard 2020 [150]	France	Pro	238	45.9	I–IV	4.4	OS	FIGO, CT, LNI	7
An 2020 [151]	China	Retro	278	45.5	I–II	5.0	OS, DFS	Age, CT, FIGO, TG, TS, LVSI, LNI, DI, RM, He	6
Casarin 2020 [152]	Italy	Retro	428	45.0	I	4.7	DFS	TS, LVSI, TG, LNI	7
Wang 2020 [153]	China	Retro	120	59.0	I–III	3.2	OS	LNI, age, FIGO, TG, TS	6
Zyla 2020 [154]	Canada	Retro	285	41.0	I	4.0	OS, DFS	TG, CT, LVSI	6
He 2020 [155]	China	Retro	1,414	NA	I–II	3.6	OS, DFS	Age, FIGO, TS, CT, TG, DI, LVSI, PI, RM, LNI	7
Zeng 2020 [156]	China	Retro	251	46.0	I–III	3.9	OS, DFS	FIGO, LNI	6
Liu 2020 [157]	China	Retro	73	NA	I–II	5.7	OS	Age, CT, FIGO, TG, TS, SCC	5
Kim 2020 [158]	Korea	Retro	47	45.0	I–II	2.4	OS, DFS	FIGO, SCC, DI, RM, PI, LNI, LVSI	5
Anfinan 2020 [159]	Saudi Arabia	Retro	190	54.2	I–IV	3.1	OS	FIGO, TG, PI	6
Lee 2020 [160]	Korea	Retro	125	53.7	II–III	4.2	OS, DFS	Age, CT, TS, FIGO, LNI, SCC, NLR	5
Zong 2020 [161]	China	Retro	384	46.3	I–II	3.6	OS, DFS	Age, FIGO, TG, TS, PI, LVSI, DI, RM	6
Aslan 2020 [162]	Turkey	Retro	185	50.0	III	3.8	OS, DFS	Age, CT, DI, PI, TS, LVSI, RM, FIGO	7
Gülseren 2020 [163]	Turkey	Retro	194	NA	I–II	5.0	DFS	FIGO, TS, PI, LVSI	6
Kim 2021 [164]	Korea	Retro	55	52.6	I–II	4.5	DFS	Age, FIGO, PI, RM	7
Okadome 2021 [165]	Japan	Retro	82	NA	II	5.8	DFS	CT, LNI, TS	6
Buda 2021 [166]	Italy	Retro	573	45.5	I–II	3.8	DFS	Age, CT, FIGO, LVSI	6

CT – cell type; DI – depth of invasion; He – hemoglobin; Retro – retrospective; Pro – prospective; PI – parametrial invasion; Le – leukocytosis; LVSI – lymph vascular space invasion; LNI – lymph node involvement; Ly – lymphocyte; RM – resection margin; SCC – squamous cell carcinoma antigen; TG – tumor grade; Th – thrombocytosis; TS – tumor size; TV – tumor volume; NA – not available; NLR – neutrophil/lymphocyte ratio.

Characteristics of the Eligible Studies

Of 140 included studies, 7 were designed as prospective cohorts, 132 as retrospective cohorts, and the remaining 1 study had both prospective and retrospective design. The sample size of the included studies ranged from 38 to 3797, and a total of 47 965 patients were involved. Forty-seven studies were conducted in Western countries and the remaining 93 studies were performed in Eastern countries. In addition, 106 and 99 studies reported the prognostic significance of clinicopathological characteristics, influencing OS and EFS of CC patients, respectively. Moreover, 41 studies were of medium quality (7 stars), and a total of 99 studies were of low quality (6 stars (69 studies) versus 5 stars (30 studies).

Overall Survival

The summary results for the prognostic factors on OS in CC patients are shown in Figure 2. The pooled results found older patients (HR: 1.10; 95% CI: 1.00–1.20; P=0.040), cell types other than squamous type (HR: 1.64; 95% CI: 1.47–1.83; P<0.001), deep depth of tumor invasion (HR: 1.92; 95% CI: 1.53–2.40; P<0.001), high FIGO stage (HR: 2.00; 95% CI: 1.76–2.28; P<0.001), low hemoglobin level (HR: 1.84; 95% CI: 1.36–2.50; P<0.001), high histological grade (HR: 1.52; 95% CI: 1.27–1.83; P<0.001), leukocytosis (HR: 2.21; 95% CI: 1.55–3.15; P<0.001), LNI (HR: 2.59; 95% CI: 2.30–2.92; P<0.001), LVSI (HR: 2.09; 95% CI: 1.75–2.49; P<0.001), high NLR (HR: 1.69; 95% CI: 1.36–2.11; P<0.001), parametrial invasion (HR: 2.18; 95% CI: 1.84–2.59; P<0.001), high PLR (HR: 1.98; 95% CI: 1.45–2.71; P<0.001), positive resection margin (HR: 1.97; 95% CI: 1.45–2.69; P<0.001), high SCCA level (HR: 1.65; 95% CI: 1.28–2.15; P<0.001), thrombocytosis (HR: 1.69; 95% CI: 1.32–2.17; P<0.001), large tumor volume (HR: 2.87; 95% CI: 2.03–4.04; P<0.001), high tumor grade (HR: 1.74; 95% CI: 1.24–2.43; P=0.001), and large tumor size (HR: 1.81; 95% CI: 1.59–2.07; P<0.001) were associated with shorter OS. There was significant heterogeneity for age, cell type, depth of tumor invasion, FIGO stage, hemoglobin, histological grade, leukocytosis, LNI, LVSI, NLR, parametrial invasion, resection margin, tumor grade, and tumor size. The pooled conclusions were stability for OS related to cell type, depth of tumor invasion, FIGO stage, hemoglobin level, histological grade, leukocytosis, LNI, LVSI, NLR, parametrial invasion, PLR, resection margin, SCCA level, tumor grade, and tumor size (data not shown).

Figure 2

The results of the meta-analysis of the prognostic factors influencing OS.

Subgroup analysis indicated the statistically significant prognostic significance of age in OS of patients with FIGO stages I–II CC or studies with low quality; cell type did not affect OS of patients with FIGO stages III–IV CC; depth of tumor invasion did not influence OS of patients with FIGO stages III–IV or I–IV CC; high FIGO stage did not influence OS of patients with FIGO stages III–IV CC; hemoglobin level did not influence OS of patients with FIGO stages I–II or III–IV CC; LVSI was not associated with OS in patients with FIGO stages III–IV CC; parametrial invasion did not affect OS of patients with FIGO stages III–IV CC; high PLR was not associated with OS of patients with FIGO stages I–IV CC, studies conducted in the Western countries or studies with high quality; positive resection margin did not influence OS of patients with FIGO stages III–IV CC; high SCCA level was not associated with OS of patients with FIGO stages III–IV CC, according to the results of pooled analyses conducted in the Western countries, and cutoff value ≥10; high tumor grade was not associated with OS of patients with FIGO stages I–IV CC, according to the pooled analyses conducted in the Western countries, or studies with high quality; and tumor size did not influence OS of patients with FIGO stages III–IV CC (Table 2).

Table 2

Subgroup analysis for overall survival and event-free survival based on countries, FIGO stage, and cutoff value.

Prognostic factors	Outcome	Variables	Subgroups	HR and 95% CI	P value	I2 (%)	Q statistic	P value between subgroups
Age	OS	Countries	Eastern	1.11 (1.00–1.23)	0.052	61.8	<0.001	0.703
			Western	1.08 (0.86–1.36)	0.489	72.0	<0.001
		FIGO stage	I–II	1.23 (1.10–1.38)	<0.001	56.2	<0.001	0.070
			III–IV	1.13 (0.76–1.69)	0.539	0.0	0.719
			Both	0.94 (0.79–1.13)	0.524	72.7	<0.001
		Cutoff value	≥50.0	1.09 (0.97–1.23)	0.162	68.2	<0.001	0.592
			<50.0	1.13 (0.96–1.33)	0.140	58.4	<0.001
		Study quality	High	1.03 (0.86–1.24)	0.723	71.9	<0.001	0.206
			Low	1.15 (1.03–1.28)	0.016	59.8	<0.001
	EFS	Countries	Eastern	1.19 (1.02–1.38)	0.024	67.4	<0.001	0.082
			Western	1.40 (0.99–1.98)	0.061	67.5	0.002
		FIGO stage	I–II	1.31 (1.13–1.52)	<0.001	56.3	<0.001	<0.001
			III–IV	0.91 (0.59–1.40)	0.666	–	–
			Both	1.03 (0.76–1.39)	0.864	77.7	<0.001
		Cutoff value	≥50.0	1.23 (1.04–1.46)	0.016	70.1	<0.001	0.022
			<50.0	1.20 (0.96–1.51)	0.116	59.0	0.001
		Study quality	High	0.90 (0.76–1.08)	0.251	65.2	<0.001	<0.001
			Low	1.49 (1.29–1.73)	<0.001	37.6	0.019
Cell type	OS	Countries	Eastern	1.74 (1.52–1.98)	<0.001	39.9	0.007	0.047
			Western	1.44 (1.20–1.73)	<0.001	18.6	0.231
		FIGO stage	I–II	1.65 (1.43–1.91)	<0.001	24.2	0.120	0.963
			III–IV	1.58 (0.89–2.78)	0.115	0.0	0.521
			Both	1.63 (1.36–1.95)	<0.001	51.3	0.002
		Study quality	High	1.79 (1.53–2.09)	<0.001	42.6	0.015	0.049
			Low	1.50 (1.29–1.74)	<0.001	26.5	0.090
	EFS	Countries	Eastern	1.68 (1.43–1.97)	<0.001	62.9	<0.001	0.008
			Western	1.50 (1.18–1.91)	0.001	58.8	0.001
		FIGO stage	I–II	1.56 (1.31–1.86)	<0.001	65.2	<0.001	0.490
			III–IV	2.33 (1.38–3.94)	0.002	–	–
			Both	1.71 (1.37–2.13)	<0.001	59.8	0.001
		Study quality	High	1.88 (1.57–2.24)	<0.001	67.4	<0.001	0.004
			Low	1.43 (1.17–1.74)	<0.001	56.7	<0.001
Depth of invasion	OS	Countries	Eastern	2.09 (1.66–2.63)	<0.001	59.1	<0.001	0.024
			Western	1.11 (0.52–2.38)	0.790	75.3	0.003
		FIGO stage	I–II	2.09 (1.65–2.63)	<0.001	62.1	<0.001	0.015
			III–IV	0.89 (0.42–1.89)	0.761	–	–
			Both	1.01 (0.43–2.37)	0.979	58.9	0.088
		Cutoff value	≥1/2	2.02 (1.59–2.57)	<0.001	37.2	0.053	0.782
			<1/2	1.73 (1.15–2.61)	0.009	77.1	<0.001
		Study quality	High	1.75 (1.20–2.55)	0.004	67.5	0.001	0.680
			Low	2.02 (1.51–2.40)	<0.001	62.6	<0.001
	EFS	Countries	Eastern	1.83 (1.60–2.09)	<0.001	28.1	0.070	0.010
			Western	1.29 (0.75–2.22)	0.359	80.7	<0.001
		FIGO stage	I–II	1.77 (1.52–2.06)	<0.001	51.6	<0.001	0.054
			III–IV	0.93 (0.51–1.71)	0.815	–	–
			Both	0.86 (0.32–2.31)	0.765	–	–
		Cutoff value	≥1/2	1.67 (1.39–2.00)	<0.001	43.8	0.019	0.549
			<1/2	1.77 (1.37–2.29)	<0.001	60.5	<0.001
		Study quality	High	1.64 (1.23–2.18)	0.001	69.1	<0.001	0.596
			Low	1.77 (1.49–2.09)	<0.001	34.6	0.047
FIGO stage	OS	Countries	Eastern	1.86 (1.62–2.14)	<0.001	84.1	<0.001	<0.001
			Western	2.36 (1.73–3.21)	<0.001	85.9	<0.001
		FIGO stage	I–II	1.60 (1.41–1.82)	<0.001	73.4	<0.001	<0.001
			III–IV	1.47 (0.85–2.54)	0.168	–	–
			Both	2.51 (2.04–3.09)	<0.001	81.7	<0.001
		Cutoff value	IA or IB	1.92 (1.65–2.23)	<0.001	87.6	<0.001	<0.001
			II–III	2.24 (1.78–2.81)	<0.001	64.9	<0.001
		Study quality	High	2.40 (1.87–3.07)	<0.001	86.9	<0.001	<0.001
			Low	1.80 (1.57–2.06)	<0.001	78.9	<0.001
	EFS	Countries	Eastern	1.83 (1.60–2.08)	<0.001	69.1	<0.001	0.355
			Western	1.97 (1.61–2.41)	<0.001	62.4	<0.001
		FIGO stage	I–II	1.70 (1.50–1.93)	<0.001	52.6	<0.001	0.001
			III–IV	1.01 (0.55–1.83)	0.984	–	–
			Both	2.11 (1.75–2.54)	<0.001	75.5	<0.001
		Cutoff value	IA or IB	1.80 (1.59–2.04)	<0.001	68.1	<0.001	0.021
			II–III	2.04 (1.65–2.52)	<0.001	62.5	<0.001
		Study quality	High	1.70 (1.45–2.00)	<0.001	73.9	<0.001	0.023
			Low	1.99 (1.72–2.31)	<0.001	61.3	<0.001
Hemoglobin	OS	Countries	Eastern	1.56 (1.15–2.10)	0.004	58.1	0.019	0.001
			Western	3.05 (2.01–4.64)	<0.001	0.0	0.608
		FIGO stage	I–II	1.39 (0.99–1.95)	0.061	0.0	0.898	0.720
			III–IV	1.81 (0.90–3.64)	0.097	-	-
			Both	2.07 (1.34–3.19)	0.001	75.7	<0.001
		Cutoff value	10	1.94 (1.13–3.36)	0.017	80.2	<0.001	0.156
			>10	1.77 (1.39–2.27)	<0.001	0.0	0.688
		Study quality	High	2.01 (1.00–4.04)	0.050	88.0	<0.001	0.337
			Low	1.70 (1.33–2.17)	<0.001	0.0	0.740
	EFS	Countries	Eastern	1.20 (1.07–1.34)	0.002	4.3	0.401	0.004
			Western	2.25 (1.48–3.41)	<0.001	0.0	0.580
		FIGO stage	I–II	1.58 (1.19–2.09)	0.001	0.0	0.778	0.071
			III–IV	–	–	–	–
			Both	1.24 (1.03–1.50)	0.022	53.6	0.044
		Cutoff value	10	1.50 (1.11–2.04)	0.009	58.9	0.023	0.248
			>10	1.19 (1.04–1.35)	0.010	0.0	0.733
		Study quality	High	1.30 (0.86–1.96)	0.216	67.5	0.026	0.718
			Low	1.29 (1.12–1.50)	0.001	18.5	0.284
Histological grade	OS	Countries	Eastern	1.56 (1.24–1.96)	<0.001	56.3	0.004	0.460
			Western	1.48 (1.08–2.02)	0.014	55.0	0.011
		FIGO stage	I–II	1.44 (1.19–1.74)	<0.001	41.6	0.030	0.414
			III–IV	–	–	–	–
			Both	1.75 (1.13–2.72)	0.012	72.6	0.001
		Cutoff value	1	1.52 (1.20–1.92)	0.001	61.8	<0.001	0.424
			2	1.56 (1.17–2.07)	0.002	32.6	0.157
		Study quality	High	1.43 (1.16–1.76)	0.001	29.1	0.160	0.839
			Low	1.62 (1.20–2.19)	0.001	66.2	<0.001
	EFS	Countries	Eastern	1.47 (1.09–1.97)	0.011	73.4	<0.001	0.377
			Western	1.38 (1.07–1.78)	0.013	44.0	0.051
		FIGO stage	I–II	1.49 (1.17–1.89)	0.001	66.7	<0.001	0.340
			III–IV	–	–	–	–
			Both	1.24 (0.99–1.57)	0.066	0.0	0.517
		Cutoff value	1	1.47 (1.13–1.90)	0.004	72.5	<0.001	0.746
			2	1.41 (1.15–1.73)	0.001	0.0	0.447
		Study quality	High	1.43 (1.12–1.84)	0.005	64.5	0.001	0.308
			Low	1.45 (1.05–2.01)	0.025	58.7	0.013
Leukocytosis	OS	Countries	Eastern	2.20 (1.48–3.26)	<0.001	75.2	0.001	0.726
			Western	2.46 (1.15–5.26)	0.020	-	-
		FIGO stage	I–II	1.55 (1.16–2.05)	0.003	0.0	0.623	0.013
			III–IV	3.04 (1.52–6.07)	0.002	-	-
			Both	2.66 (1.53–4.64)	0.001	73.7	0.010
		Cutoff value	≥10000	2.05 (1.25–3.35)	0.004	50.6	0.132	0.242
			<10000	2.35 (1.39–4.00)	0.002	79.8	0.002
		Study quality	High	1.74 (1.18–2.56)	0.005	9.6	0.293	0.148
			Low	2.41 (1.51–3.85)	<0.001	76.6	0.002
	EFS	Countries	Eastern	2.08 (1.25–3.45)	0.005	69.6	0.011	–
			Western	–	–	–	–
		FIGO stage	I–II	1.66 (0.52–5.26)	0.389	–	–	0.642
			III–IV	–	–	–	–
			Both	2.14 (1.20–3.81)	0.010	76.8	0.005
		Cutoff value	≥10000	1.63 (0.66–4.05)	0.290	0.0	0.964	0.526
			<10000	2.22 (1.16–4.24)	0.016	84.3	0.002
		Study quality	High	2.10 (1.62–2.74)	<0.001	–	–	0.685
			Low	2.00 (0.86–4.65)	0.109	76.9	0.005
LNI	OS	Countries	Eastern	2.49 (2.17–2.85)	<0.001	71.4	<0.001	0.007
			Western	2.90 (2.29–3.67)	<0.001	60.5	<0.001
		FIGO stage	I–II	2.97 (2.57–3.43)	<0.001	65.8	<0.001	0.001
			III–IV	–	–	–	–
			Both	2.04 (1.66–2.51)	<0.001	72.3	<0.001
		Study quality	High	2.52 (2.08–3.04)	<0.001	68.5	<0.001	0.639
			Low	2.64 (2.26–3.09)	<0.001	70.7	<0.001
	EFS	Countries	Eastern	2.37 (2.03–2.77)	<0.001	81.0	<0.001	0.001
			Western	2.18 (1.75–2.72)	<0.001	61.6	<0.001
		FIGO stage	I–II	2.54 (2.14–3.01)	<0.001	81.8	<0.001	0.998
			III–IV	–	–	–	–
			Both	1.89 (1.57–2.26)	<0.001	61.5	<0.001
		Study quality	High	2.16 (1.73–2.70)	<0.001	87.0	<0.001	<0.001
			Low	2.40 (2.10–2.75)	<0.001	51.3	<0.001
LVSI	OS	Countries	Eastern	1.99 (1.64–2.43)	<0.001	63.2	<0.001	0.036
			Western	2.49 (1.72–3.60)	<0.001	36.3	0.100
		FIGO stage	I–II	2.08 (1.70–2.55)	<0.001	64.6	<0.001	0.539
			III–IV	2.10 (0.32–13.68)	0.438	–	–
			Both	2.20 (1.66–2.90)	<0.001	0.0	0.976
		Study quality	High	1.78 (1.41–2.24)	<0.001	47.4	0.029	0.046
			Low	2.30 (1.80–2.94)	<0.001	62.0	<0.001
	EFS	Countries	Eastern	1.87 (1.62–2.16)	<0.001	48.0	0.001	<0.001
			Western	1.80 (1.33–2.46)	<0.001	80.5	<0.001
		FIGO stage	I–II	1.92 (1.68–2.18)	<0.001	51.1	<0.001	<0.001
			III–IV	0.94 (0.36–2.44)	0.899	–	–
			Both	1.02 (0.95–1.09)	0.572	–	–
		Study quality	High	1.77 (1.34–2.32)	<0.001	85.6	<0.001	<0.001
			Low	1.91 (1.63–2.23)	<0.001	43.0	0.004
NLR	OS	Countries	Eastern	1.48 (1.23–1.79)	<0.001	52.9	0.038	0.001
			Western	2.50 (1.39–4.50)	0.002	50.5	0.155
		FIGO stage	I–II	1.78 (1.37–2.31)	<0.001	0.0	0.476	0.004
			III–IV	-	–	–	–
			Both	1.62 (1.22–2.14)	0.001	72.7	0.003
		Cutoff value	≥3.0	2.40 (1.75–3.28)	<0.001	0.0	0.494	<0.001
			<3.0	1.35 (1.15–1.59)	<0.001	41.1	0.131
		Study quality	High	1.58 (1.23–2.03)	<0.001	75.7	0.001	0.005
			Low	2.04 (1.43–2.92)	<0.001	0.0	0.926
	EFS	Countries	Eastern	1.56 (1.23–1.98)	<0.001	76.6	<0.001	<0.001
			Western	3.58 (2.11–6.08)	<0.001	–	–
		FIGO stage	I–II	1.99 (1.51–2.63)	<0.001	0.0	0.816	<0.001
			III–IV	-	–	–	–
			Both	1.61 (1.17–2.21)	0.003	86.1	<0.001
		Cutoff value	≥3.0	2.12 (1.28–3.52)	0.004	58.4	0.065	<0.001
			< 3.0	1.51 (1.16–1.98)	0.002	81.7	<0.001
		Study quality	High	1.65 (1.16–2.36)	0.006	85.8	<0.001	<0.001
			Low	1.85 (1.24–2.78)	0.003	60.6	0.038
Parametrial invasion	OS	Countries	Eastern	2.16 (1.81–2.58)	<0.001	31.7	0.060	0.828
			Western	2.26 (1.44–3.55)	<0.001	67.5	0.001
		FIGO stage	I–II	2.15 (1.81–2.55)	<0.001	31.4	0.053	0.024
			III–IV	1.11 (0.53–2.32)	0.782	–	–
			Both	2.26 (1.31–3.89)	0.003	68.9	0.007
		Study quality	High	1.90 (1.36–2.66)	<0.001	66.1	0.001	0.050
			Low	2.36 (1.96–2.64)	<0.001	22.9	0.146
	EFS	Countries	Eastern	1.89 (1.63–2.21)	<0.001	37.6	0.019	0.948
			Western	2.03 (1.66–2.21)	<0.001	58.0	0.015
		FIGO stage	I–II	1.96 (1.68–2.28)	<0.001	42.7	0.005	0.153
			III–IV	3.70 (1.14–11.96)	0.029	–	–
			Both	1.48 (1.01–2.15)	0.044	24.9	0.262
		Study quality	High	1.54 (1.32–1.80)	<0.001	12.1	0.321	<0.001
			Low	2.23 (1.86–2.69)	<0.001	32.7	0.056
PLR	OS	Countries	Eastern	2.20 (1.62–3.00)	<0.001	0.0	0.531	0.101
			Western	1.54 (0.73–3.25)	0.260	69.8	0.069
		FIGO stage	I–II	2.10 (1.51–2.91)	<0.001	0.0	0.486	0.342
			III–IV	–	–	–	–
			Both	1.86 (0.97–3.59)	0.062	65.6	0.055
		Cutoff value	≥150	2.59 (1.68–3.99)	<0.001	0.0	0.862	0.081
			<150	1.72 (1.12–2.65)	0.014	48.4	0.121
		Study quality	High	1.55 (0.98–2.43)	0.059	45.1	0.162	0.033
			Low	2.54 (1.76–3.66)	<0.001	0.0	0.805
	EFS	Countries	Eastern	2.47 (1.80–3.38)	<0.001	0.0	0.914	0.004
			Western	1.01 (0.60–1.70)	0.973	–	–
		FIGO stage	I–II	2.44 (1.71–3.48)	<0.001	0.0	0.779	0.058
			III–IV	–	–	–	–
			Both	1.58 (0.63–3.95)	0.333	78.8	0.030
		Cutoff value	≥150	2.59 (1.58–4.23)	<0.001	0.0	0.992	0.174
			<150	1.82 (0.96–3.46)	0.069	71.3	0.030
		Study quality	High	1.56 (0.62–3.93)	0.343	76.7	0.038	0.045
			Low	2.44 (1.72–3.46)	<0.001	0.0	0.779
Resection margin	OS	Countries	Eastern	1.88 (1.29–2.75)	0.001	65.7	0.002	0.268
			Western	2.22 (1.25–3.95)	0.006	44.1	0.111
		FIGO stage	I–II	1.89 (1.36–2.62)	<0.001	57.3	0.004	0.050
			III–IV	1.55 (0.86–2.81)	0.148	–	–
			Both	5.49 (2.09–14.41)	0.001	–	–
		Study quality	High	2.13 (1.24–3.66)	0.006	74.6	<0.001	0.569
			Low	1.75 (1.27–2.40)	0.001	18.3	0.285
	EFS	Countries	Eastern	2.16 (1.56–2.99)	<0.001	52.2	0.006	0.129
			Western	1.69 (1.20–2.37)	0.003	39.3	0.106
		FIGO stage	I–II	1.86 (1.43–2.43)	<0.001	45.2	0.012	0.005
			III–IV	1.71 (1.20–2.43)	0.003	0.0	0.925
			Both	5.62 (2.78–11.37)	<0.001	0.0	0.795
		Study quality	High	1.80 (1.33–2.45)	<0.001	53.3	0.015	0.218
			Low	2.26 (1.53–3.33)	<0.001	45.7	0.032
SCC	OS	Countries	Eastern	1.72 (1.26–2.35)	0.001	42.0	0.078	0.884
			Western	1.50 (0.92–2.45)	0.105	–	–
		FIGO stage	I–II	1.81 (1.22–2.68)	0.003	0.0	0.737	0.259
			III–IV	1.00 (0.55–1.82)	0.992	–	–
			Both	1.97 (1.25–3.10)	0.003	63.1	0.028
		Cutoff value	≥10	1.39 (0.76–2.53)	0.288	20.4	0.285	0.654
			<10	1.77 (1.29–2.42)	<0.001	45.4	0.076
		Study quality	High	2.61 (1.42–4.83)	0.002	37.0	0.204	0.019
			Low	1.36 (1.15–1.60)	<0.001	0.0	0.440
	EFS	Countries	Eastern	1.80 (1.33–2.45)	<0.001	43.7	0.087	–
			Western	–	–	–	–
		FIGO stage	I–II	1.17 (0.61–2.22)	0.637	34.4	0.218	0.079
			III–IV	–	–	–	–
			Both	2.08 (1.51–2.87)	<0.001	36.6	0.177
		Cutoff value	≥10	1.63 (0.56–4.76)	0.370	77.4	0.035	0.954
			<10	1.83 (1.33–2.53)	<0.001	37.6	0.156
		Study quality	High	1.70 (1.14–2.56)	0.010	52.4	0.122	0.469
			Low	1.86 (1.12–3.11)	0.017	48.2	0.102
Tumor grade	OS	Countries	Eastern	2.00 (1.37–2.93)	<0.001	61.5	0.016	0.007
			Western	1.07 (0.78–1.45)	0.678	0.0	0.899
		FIGO stage	I–II	2.00 (1.37–2.93)	<0.001	61.5	0.016	0.007
			III–IV	–	–	–	–
			Both	1.07 (0.78–1.45)	0.678	0.0	0.899
		Study quality	High	1.67 (0.90–3.10)	0.101	–	–	0.721
			Low	1.76 (1.21–2.57)	0.003	69.4	0.002
	EFS	Countries	Eastern	1.39 (1.14–1.71)	0.001	39.2	0.130	0.480
			Western	1.16 (0.60–2.26)	0.661	11.6	0.288
		FIGO stage	I–II	1.41 (1.17–1.70)	<0.001	30.3	0.186	0.226
			III–IV	–	–	–	–
			Both	0.89 (0.41–1.94)	0.769	–	–
		Study quality	High	1.35 (0.96–1.89)	0.084	54.1	0.113	0.754
			Low	1.38 (1.05–1.82)	0.021	29.1	0.217
Tumor size	OS	Countries	Eastern	1.76 (1.52–2.05)	<0.001	71.7	<0.001	0.004
			Western	1.95 (1.51–2.53)	<0.001	59.9	0.001
		FIGO stage	I–II	1.66 (1.41–1.97)	<0.001	70.6	<0.001	<0.001
			III–IV	1.09 (0.55–2.15)	0.811	45.3	0.176
			Both	2.17 (1.78–2.65)	<0.001	59.1	<0.001
		Cutoff value	≥4.0 cm	1.72 (1.48–2.00)	<0.001	69.6	<0.001	<0.001
			<4.0 cm	2.09 (1.61–2.70)	<0.001	64.8	<0.001
		Study quality	High	1.87 (1.52–2.31)	<0.001	66.1	<0.001	0.010
			Low	1.78 (1.51–2.11)	<0.001	70.7	<0.001
	EFS	Countries	Eastern	1.70 (1.46–1.98)	<0.001	77.8	<0.001	<0.001
			Western	1.67 (1.25–2.22)	0.001	74.4	<0.001
		FIGO stage	I–II	1.67 (1.45–1.93)	<0.001	66.9	<0.001	<0.001
			III–IV	1.59 (0.89–2.83)	0.115	–	–
			Both	1.75 (1.34–2.28)	<0.001	86.1	<0.001
		Cutoff value	≥4.0 cm	1.66 (1.39–1.98)	<0.001	78.5	<0.001	0.062
			<4.0 cm	1.76 (1.43–2.17)	<0.001	77.1	<0.001
		Study quality	High	1.48 (1.28–1.72)	<0.001	68.3	<0.001	0.053
			Low	1.90 (1.54–2.35)	<0.001	81.3	<0.001

There was significant publication bias for the prognostic significance of FIGO stage (P (Egger’s test) <0.001; P (Begg’s test)=0.749), hemoglobin level (P (Egger’s test)=0.013; P (Begg’s test)=0.119), histological grade (P (Egger’s test)=0.044; P (Begg’s test)=0.024), LVSI (P (Egger’s test)=0.026; P (Begg’s test)=0.056), NLR (P (Egger’s test)=0.001; P (Begg’s test)=0.074), PLR (P (Egger’s test)=0.020; P (Begg’s test)=0.007), tumor grade (P (Egger’s test)=0.031; P (Begg’s test)=0.048), and tumor size (P (Egger’s test)=0.006; P (Begg’s test)=0.950) in OS (Table 3). The pooled conclusion for OS were not changed after adjustment for publication bias by using the trim and fill method.

Table 3

Publication bias for clinicopathological factors.

Factors	OS		EFS
	Egger	Begg	Egger	Begg
Age	0.261	0.298	<0.001	0.010
Cell type	0.052	0.114	0.083	0.057
Depth of invasion	0.641	0.700	0.624	0.408
FIGO stage	<0.001	0.749	0.016	0.061
Hemoglobin	0.013	0.119	0.026	0.024
Histological grade	0.044	0.024	0.186	0.063
Leukocytosis	0.624	0.368	0.831	0.806
LNI	0.127	0.603	<0.001	0.460
LVSI	0.026	0.056	<0.001	0.273
NLR	0.001	0.074	0.006	0.210
Parametrial invasion	0.640	0.948	0.566	0.972
PLR	0.020	0.007	0.388	0.221
Resection margin	0.101	0.260	0.087	0.378
SCC	0.139	0.533	0.430	0.536
Tumor grade	0.031	0.048	0.568	1.000
Tumor size	0.006	0.950	<0.001	0.082

Event-Free Survival

The summary results for the prognostic factors on EFS in CC patients are shown in Figure 3. The pooled analyses indicated that older patients (HR: 1.22; 95% CI: 1.06–1.40; P=0.004), cell types other than squamous type (HR: 1.62; 95% CI: 1.42–1.86; P<0.001), deep depth of tumor invasion (HR: 1.72; 95% CI: 1.48–2.00; P<0.001), high FIGO stage (HR: 1.87; 95% CI: 1.67–2.08; P<0.001), low hemoglobin level (HR: 1.31; 95% CI: 1.12–1.53; P=0.001), high histological grade (HR: 1.43; 95% CI: 1.18–1.74; P<0.001), leukocytosis (HR: 2.08; 95% CI: 1.25–3.45; P=0.005), LNI (HR: 2.32; 95% CI: 2.03–2.64; P<0.001), LVSI (HR: 1.87; 95% CI: 1.60–2.18; P<0.001), high NLR (HR: 1.73; 95% CI: 1.33–2.25; P<0.001), parametrial invasion (HR: 1.91; 95% CI: 1.66–2.21; P<0.001), high PLR (HR: 2.05; 95% CI: 1.35–3.10; P=0.001), positive resection margin (HR: 1.99; 95% CI: 1.56–2.52; P<0.001), high SCCA level (HR: 1.80; 95% CI: 1.33–2.45; P<0.001), thrombocytosis (HR: 1.47; 95% CI: 1.08–1.98; P=0.013), large tumor volume (HR: 1.86; 95% CI: 1.40–2.47; P<0.001), high tumor grade (HR: 1.37; 95% CI: 1.14–1.66; P=0.001), and large tumor size (HR: 1.68; 95% CI: 1.48–1.90; P<0.001) were associated with shorter EFS. There was significant heterogeneity for age, cell type, depth of tumor invasion, FIGO stage, hemoglobin, histological grade, leukocytosis, LNI, LVSI, NLR, parametrial invasion, PLR, resection margin, SCCA level, and tumor size. The pooled conclusions were stability for EFS related to age, cell type, depth of tumor invasion, FIGO stage, hemoglobin level, histological grade, LNI, LVSI, NLR, parametrial invasion, PLR, resection margin, SCCA level, tumor grade, and tumor size (Data not shown).

Figure 3

The results of the meta-analysis of the prognostic factors influencing EFS.

Subgroup analysis indicated the statistically significant prognostic significance of age in EFS was observed for studies performed in Eastern countries, patients with FIGO stages I–II CC, the cutoff value of age was ≥50.0, and studies with low quality; depth of tumor invasion did not influence EFS of patients with FIGO stages III–IV or I–IV CC; high FIGO stage did not influence EFS of patients with FIGO stages III–IV CC; EFS were not affected by hemoglobin when pooled studies with high quality; histological grade did not influence EFS of patients with FIGO stages I–IV CC; leukocytosis did not impact EFS of patients with FIGO stages I–II CC, and cutoff value ≥10 000, or studies with low quality; LVSI was not associated with EFS in patients with FIGO stages III–IV or I–IV CC; PLR did not influence EFS of patients with FIGO stages I–IV CC, studies conducted in the Western countries, cutoff value <150, or studies with high quality; high SCCA level did not affect EFS of patients with FIGO stages I–II CC, or cutoff value ≥10; high tumor grade was not associated with EFS of patients with FIGO stages I–IV CC, according to the pooled analyses conducted in the Western countries, or studies with high quality; and tumor size did not influence EFS of patients with FIGO stages III–IV CC (Table 2).

There was significant publication bias for the prognostic significance of age (P (Egger’s test) <0.001; P (Begg’s test)=0.010), FIGO stage (P (Egger’s test)=0.016; P (Begg’s test)=0.061), hemoglobin level (P (Egger’s test)=0.026; P (Begg’s test)=0.024), LNI (P (Egger’s test) <0.001; P (Begg’s test)=0.460), LVSI (P (Egger’s test) <0.001; P (Begg’s test)=0.273), NLR (P (Egger’s test)=0.006; P (Begg’s test)=0.210), and tumor size (P (Egger’s test) <0.001; P (Begg’s test)=0.082) in EFS (Table 3). The pooled conclusions for EFS were not altered after adjusting for potential publication bias.

Discussion

The results of this study showed that the potential risk factors for OS and EFS were age, cell type, depth of tumor invasion, FIGO stage, hemoglobin level, histological grade, leukocytosis, LNI, LVSI, NLR, parametrial invasion, PLR, resection margin, SCCA level, thrombocytosis, tumor grade, tumor size, and tumor volume. Moreover, we noted that the first author’s country of residence could affect the prognostic significance of cell type, depth of tumor invasion, FIGO stage, hemoglobin level, LNI, LVSI, NLR, tumor stage, and tumor size in OS, and the prognostic significance of cell type, depth of tumor invasion, hemoglobin level, LNI, LVSI, NLR, PLR, and tumor size in EFS was influenced by the first author’s country of residence. Furthermore, FIGO stage could affect the prognostic significance of depth of tumor invasion, leukocytosis, LNI, NLR, parametrial invasion, resection margin, tumor grade, and tumor size in OS, and the prognostic significance of age, LVSI, NLR, resection margin, and tumor size in EFS could be influenced by FIGO stage. We also found that cutoff value affected the prognostic significance of FIGO stage, NLR, and tumor size in OS, and the prognostic significance of age, FIGO stage, and NLR in EFS could be affected by cutoff value. Finally, the study quality could affect the prognostic significance of cell type, FIGO stage, LVSI, NLR, parametrial invasion, PLR, SCCA, and tumor size in OS, while the prognostic significance of age, cell type, FIGO stage, LNI, LVSI, NLR, parametrial invasion, and PLR in EFS could affect by study quality.

A previous meta-analysis of 22 studies revealed that the prognosis of CC was influenced by advanced FIGO stage, large tumor size, LNI, LVSI, parametrial invasion, depth of tumor invasion, and radiation therapy [118]. Zhang et al retrieved 20 cohort studies and found that FIGO stage, tumor size, parametrial invasion, resection margin, LNI, depth of tumor invasion, neoadjuvant chemotherapy, and adjuvant chemotherapy could affect OS of patients with CC [167]. However, other meta-analyses investigated the prognostic factors for OS, whereas those factors for EFS were not assessed. Moreover, the prognostic significance of clinicopathological factors, influencing OS and EFS of CC patients, which could be influenced by the first author’s country of origin, FIGO stage, and cutoff value, were not evaluated. We therefore conducted the present systematic review and meta-analysis to identify the prognostic significance of clinicopathological factors influencing OS and EFS of patients with CC.

Compared with previous studies, this study revealed that FIGO stage, tumor size, parametrial invasion, resection margin, LNI, LVSI, and depth of tumor invasion could affect the prognosis of CC patients, which may be related to the fact that these factors could directly reflect distant metastasis and are associated with a poor prognosis of CC patients [168-170]. Furthermore, we studied additional prognostic factors, such as age, cell type, hemoglobin level, histological grade, leukocytosis, NLR, PLR, SCCA level, thrombocytosis, tumor grade, and tumor volume. The above-mentioned results could be explained as follows: (1) The incidence of CC varies among different age-based groups, and the FIGO stage of CC also significantly differs among various age-based groups [2]; (2) Compared with squamous cell carcinoma, patients with adenocarcinoma may tend to have other extracervical spread, associating with a poor prognosis of CC patients [171]; (3) The hemoglobin level is significantly correlated to the tumor size and infiltrative phenotypes of tumors [172,173]; Moreover, the hemoglobin level may act as a surrogate marker of tumor hypoxia, which is significantly associated with resistance to radiotherapy [174]; (4) Histological grade, tumor grade, and tumor volume are significantly correlated to tumor extension and invasion, which may influence the prognosis of CC patients; (5) Leukocytosis in CC patients is associated with a poor prognosis, which may be related to a poor response to radiation therapy [100]; (6) Increased NLR is markedly associated with a large tumor size, advanced clinical stage, and positive LNI, resulting in shorter OS and EFS [15]; (7) Elevated PLR can induce inflammatory cytokines and chemokines, promoting the progression of cancer cells [175]; (8) Increased SCCA concentration can reflect the degree of cell proliferation for patients with CC [176]; and (9) Cancer treatment can induce thrombocytosis, cytokines or growth factors, receptors, and downstream effectors, playing an important role in the prognosis of CC [177].

The current meta-analysis indicated the prognostic significance of cell type, depth of tumor invasion, FIGO stage, hemoglobin level, LNI, LVSI, NLR, PLR, tumor stage, and tumor size, which significantly differed in patients studied in the Eastern and Western countries. The results were based on the diagnosis of CC patients at various FIGO stages in different countries. Moreover, the vaccination rate in the Eastern and Western countries is different, influencing the incidence and prognosis of CC. Moreover, the effects of age, depth of tumor invasion, leukocytosis, LNI, LVSI, NLR, parametrial invasion, resection margin, tumor grade, and tumor size on the prognosis of CC patients could be influenced by FIGO stage. Additionally, the effects of age, FIGO stage, NLR, and tumor size on the prognosis of CC patients could be affected by the cutoff value.

The strengths of our study include: (1) our study contained 18 clinicopathological factors, which provide relatively comprehensive prognostic factors for CC; (2) the analysis was based on a large number of included studies, and the pooled conclusions are potentially more robust than are those of any individual study; and (3) subgroup analyses were performed for prognostic factors reported by more than 5 studies, which could assess the prognostic role of clinicopathological factors on OS and EFS according to studies’ characteristics. Several shortcoming of this study should be pointed out: (1) the majority of the included studies had a retrospective design, and selection or confounder biases were therefore inevitable; (2) the noticeable changes of the cutoff values partly expanded the range of the results of subgroup analyses; (3) the heterogeneity among the included studies was not fully explained by the results of the sensitivity and subgroup analyses; (4) the treatment strategies for CC significantly differed among the included studies, which could influence the prognosis of CC patients; (5) several other outcomes should be addressed in further large-scale prospective studies, including response to chemotherapy, remission rates, hospitalization rates, and complication rates; (6) the transparency of our study was restricted because it was not registered in PROSPERO; and (7) inherent limitations of meta-analysis of previously published articles are noteworthy.

Conclusions

This study comprehensively identified the prognostic significance of clinicopathological factors and influencing OS and EFS of patients with CC, including age, cell type, depth of tumor invasion, FIGO stage, hemoglobin level, histological grade, leukocytosis, LNI, LVSI, NLR, parametrial invasion, PLR, resection margin, SCCA level, thrombocytosis, tumor grade, tumor size, and tumor volume. However, further large-scale prospective studies should be conducted to verify our findings and develop more accurate prognostic models for CC.