Introduction
Description of the condition
Description of the interventions
Why it is important to do this overview
Methods
Protocol and registration
Criteria for considering reviews for inclusion
Inclusion and exclusion criteria
Types of participants
Types of intervention
-
(1)One combination chemotherapy versus another;
-
(2)chemotherapy + targeted therapy versus chemotherapy;
-
(3)combined chemotherapy versus single-drug chemotherapy;
-
(4) observation versus chemotherapy or targeted therapy.
Types of outcome measure
-
(1) Indicators related to prognosis: overall survival (OS), disease-free survival (DFS), and progression-free survival (PFS).
-
(2) Efficacy-related indicators: disease control rate (DCR), disease response rate (DRR), and overall response rate (ORR).
Search methods for the identification of reviews
Data collection and analysis
Data extraction and management
Assessment of methodological quality
-
High: No or one non-critical weakness: the systematic review provided an accurate and comprehensive summary of the results of the available studies that addressed the question of interest [25].
-
Moderate: More than one non-critical weakness: the systematic review had more than one weakness but no critical flaws. It may provide an accurate summary of the results of the available studies that were included in the review [25].
-
Low: One critical flaw with or without non-critical weaknesses: the review had a critical flaw and may not provide an accurate and comprehensive summary of the available studies that addressed the question of interest [25].
-
Critically low: More than one critical flaw with or without non-critical weaknesses: the review had more than one critical flaw and should not be relied on to provide an accurate and comprehensive summary of the available studies [25].
Study | Q1 | Q2* | Q3 | Q4* | Q5 | Q6 | Q7* | Q8 | Q9* | Q10 | Q11* | Q12 | Q13* | Q14 | Q15* | Q16 | AMSTAR-2 overall quality |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ALESSANDRO RIZZO, 2020 [22] | Y | N | N | Y | N | Y | N | PY | PY | N | Y | Y | Y | Y | N | Y | Critically low |
Lawrence Chen, 2016 [16] | Y | Y | Y | Y | N | Y | N | PY | PY | N | N | N | Y | Y | N | Y | Critically low |
Ting Zheng, 2020 [19] | Y | N | N | PY | N | Y | N | PY | PY | N | N | Y | Y | Y | Y | Y | Critically low |
Xin ZHUANG, 2017 [21] | Y | N | N | PY | N | Y | Y | PY | PY | N | Y | Y | Y | Y | Y | Y | Low |
Heng Liu, 2014 [18] | Y | N | Y | PY | N | Y | Y | PY | PY | N | Y | Y | Y | Y | Y | Y | Low |
Sheng Zhao, 2016 [17] | Y | N | N | Y | N | Y | N | PY | PY | N | Y | Y | Y | Y | Y | Y | Critically low |
Alessandro Rizzo, 2022 [26] | Y | N | N | Y | Y | Y | N | PY | PY | N | N | Y | Y | Y | N | Y | Critically low |
Wen-Jie Ma,2020 [27] | Y | N | Y | Y | Y | Y | N | Y | PY | N | Y | Y | Y | Y | Y | Y | Critically low |
Julien Edeline, 2022 [28] | Y | N | Y | N | N | N | N | Y | N | N | N | Y | Y | N | N | Y | Critically low |
Abdel-Rahman O, 2018 [29] | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | High |
Yan Li, 2019 [30] | Y | N | N | Y | Y | Y | N | Y | N | N | N | N | N | N | Y | Y | Critically low |
Jie Ying, 2019 [31] | Y | PY | N | Y | N | Y | N | PY | PY | N | N | Y | N | N | Y | Y | Critically low |
Wei Zheng, 2019 [32] | Y | N | N | Y | N | Y | N | PY | N | N | N | N | N | Y | Y | Y | Critically low |
Yanfeng Jiang, 2021 [33] | Y | N | Y | PY | N | Y | N | PY | PY | N | N | N | Y | Y | Y | Y | Critically low |
Quality evaluation of evidence
-
Study type: Evidence based on randomized controlled studies was initially defined as high level; evidence based on retrospective studies was initially defined as low level.
-
Risk of bias: If relevant evidence was from studies with high risk of bias, the quality level of randomized trials and observational studies may be reduced. No serious limitations – Rating down 0 level; Severe limitations – Rating down 1 level; extremely severe limitations – Rating down 2 levels [35].
-
Indirectness: There was a large difference or no direct comparison between the populations, interventions, or outcomes in relevant systematic review.– Rating down 1 level [36].
-
Inconsistency: After discussing the prior hypothesis that may explain the source of heterogeneity, the research results remained inconsistent (heterogeneity).– Rating down 1 level [37];
-
Imprecision: Outcomes conformed to the OIS standard and the corresponding 95% CIs did not contain invalid values.– Precision – Rating down 0 level; Outcomes did not conform to the OIS standard.– Imprecision – Rating down 1 level; Outcomes conformed to the OIS standard, but the corresponding 95% CI contained invalid value; the CI did not exclude significant benefits or hazards. – Imprecision – Rating down 1 level [38].
-
Publication bias: If suspected, the quality of evidence should be rated down at least 1 level [39].
-
Large effect: The relative risk of direct evidence was large (RR = –5 or RR = 0.5–0.2) without reasonable confounding. – Large effect – Rating up 1 level; The relative risk of direct evidence was very large (RR > 5 or RR < 0.2) and had no risk of bias or serious problems related to accuracy.– Very large effect – Rating up 2 levels [40];
-
Dose–response gradient: Rating up 1 level [40];
-
Plausible confounding: Reasonable residual confounding will further support the conclusion of efficacy inference. – Rating up 1 level [40];
Data synthesis
Results
Description of included reviews
First author, Year | Original article retrieval time | Journal | Total No. of included studies | Type of study | Study design | Type of Chemotherapy | Subgroup No. of included studies | Intervention (No.of cases) | Control (No.of cases) | Sample size |
---|---|---|---|---|---|---|---|---|---|---|
ALESSANDRO RIZZO,2020 [22] | 2019/11/02 | IN VIVO | 4 | Systematic Review and Meta-analysis | RCTs | G-based + anti-EGFR vs. G-based | 4 | Gem-based + anti-EGFR (228) | Gem-based CHT alone (222) | 450 patients |
Lawrence Chen, 2016 [16] | 2016/04 | MEDICINE | 15 | Meta-analysis | RCTs | GP vs. G | 2 | GP(202) | G(184) | 386 patients |
GP + anti-EGFR vs. GP | 4 | GP + anti-EGFR (316) | GP(313) | 629 patients | ||||||
Ting Zheng, 2020 [19] | 2018/12 | ONCOLOGY RESEARCH AND TREATMENT | 5 | Meta-analysis | Restrospective; RCTs | GP vs. FP | 5 | GP(311) | FP(416) | 727 patients |
Xin ZHUANG, 2017 [21] | 2016/07 | JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE ANDTECHNOLOGY-MEDICAL SCIENCES | 7 | Meta-analysis | RCTs | G-based + anti-EGFR vs. G-based | 6 | Gem-based + anti-EGFR (432) | Gem-based CHT alone (423) | 855 patients |
Heng Liu, 2014 [18] | 2013/11 | WORLD JOURNAL OF GASTROENTEROLOGY | 7 | Meta-analysis | RCTs | G-based vs. non-G-based | 4 | Gem-based(150) | non-G-based(153) | 303 patients |
G-based vs. G | 3 | Gem-based(275) | G (280) | 555 patients | ||||||
Sheng Zhao, 2016 [17] | 2016/03 | ONCOTARGETS AND THERAPY | 6 | Systematic Review and Meta-analysis | RCTs | G-based + anti-VEGFR/EGFR vs. G-based | 6 | G-based + anti-VEGFR/ EGFR(NA) | G-based(NA) | 855 patients |
Alessandro Rizzo, 2022 [26] | 2021/12/08 | EXPERT REVIEW OF GASTROENTEROLOGY & HEPATOLOGY | 2 | Meta-analysis | RCTs | Fluoropyrimidine-based doublet CHT vs. ASC or 5-FU/LV | 2 | Fluoropyrimidine-based doublet CHT(169) | ASC or 5-FU/LV(167) | 336 patients |
Wen-Jie Ma, 2020 [27] | 2019/06/12 | HBP | 5 | Meta-analysis | RCTs | Fluoropyrimidine-based vs. Observation | 4 | Fluoropyrimidine-based (381) | Observation(358) | 739 patients |
G-based vs. Observation | 3 | G-based(246) | Observation(238) | 484 patients | ||||||
Julien Edeline, 2022 [28] | NA | EUROPEAN JOURNAL OF CANCER | 2 | Meta-analysis | RCTs | G-based vs. Observation | 2 | G-based(212) | Observation(207) | 419 patients |
Abdel-Rahman O, 2018 [29] | 2017/06 | COCHRANE DATABASE OF SYSTEMATIC REVIEWS | 7 | Meta-analysis | RCTs | G + S-1 vs. S-1 | 2 | G + S-1(76) | S-1(75) | 151 patients |
Yan Li,2019 [30] | 2018/10/06 | FRONTIERS IN ONCOLOGY | 25 | Network Meta-analysis | Restrospective; RCTs | Folfox-4 vs. Observation | NA | Folfox-4(NA) | Observation(NA) | NA |
XP vs. GP | NA | XP(NA) | GP(NA) | NA | ||||||
G + S-1 vs. GC | NA | GS(NA) | GC(NA) | NA | ||||||
Jie Ying, 2019 [31] | 2018/04 | CRITICAL REVIEWS IN ONCOLOGY HEMATOLOGY | 32 | Network Meta-analysis | Restrospective; RCTs | G-based CHT vs. single CHT(mainly FU alone) | 15 | GEM-based CHT(233) | single CHT(396) | 629 patients |
Fluoropyrimidine-based CHT vs. single TAs | 15 | Fluoropyrimidine-based CHT(527) | single TAs(150) | 677 patients | ||||||
Taxanes-based CHT vs. single CHT(mainly FU alone) | 9 | Taxanes-based CHT(54) | single CHT(396) | 450 patients | ||||||
Fluoropyrimidine-based CHT vs. single CHT(mainly FU alone) | 8 | Fluoropyrimidine-based CHT (244) | single CHT(348) | 592 patients | ||||||
Wei Zheng, 2019 [32] | 2017/11 | JOURNAL OF CANCER | 16 | Network Meta-analysis | RCTs | G + S-1 vs. G | NA | G + S-1(NA) | G(NA) | NA |
G + S-1 vs. 5-FU | NA | G + S-1(NA) | 5-FU(NA) | NA | ||||||
CapC vs.5-FU | NA | CapC(NA) | 5-FU(NA) | NA | ||||||
GEMOX vs. 5-FU | NA | GEMOX (NA) | 5-FU(NA) | NA | ||||||
FP vs. 5-FU | NA | FP(NA) | 5-FU(NA) | NA | ||||||
Yanfeng Jiang, 2021 [33] | 2020/08/10 | FRONTIERS IN ONCOLOGY | 24 | Network Meta-analysis | RCTs | Observation vs. Folfox-4 | NA | Observation(NA) | Folfox-4(NA) | NA |
Observation vs. C-GEMOX | NA | Observation(NA) | C-GEMOX(NA) | NA | ||||||
Observation vs. GEMOX + erlotinib | NA | Observation(NA) | GEMOX + erlotinib(NA) | NA | ||||||
GP + cediranib vs. 5-FU | NA | GP + cediranib(NA) | 5-FU(NA) | NA | ||||||
GP vs. 5-FU | NA | GP(NA) | 5-FU(NA) | NA | ||||||
G + S-1 vs. 5-FU | NA | GS(NA) | 5-FU(NA) | NA | ||||||
C-GEMOX vs. 5-FU | NA | C-GEMOX(NA) | 5-FU(NA) | NA | ||||||
RAM + GP vs. 5-FU | NA | RAM + GP(NA) | 5-FU(NA) | NA | ||||||
MER + GP vs. 5-FU | NA | MER + GP(NA) | 5-FU(NA) | NA | ||||||
XELOX vs. G + XELOX | NA | XELOX(NA) | G + XELOX(NA) | NA | ||||||
XELOX vs. GEMOX + erlotinib | NA | XELOX(NA) | GEMOX + erlotinib(NA) | NA |
Objectives and scope of the reviews
-
One review reported the comparison between fluoropyrimidine + cisplatin and gemcitabine + cisplatin [19];
Study characteristics and populations
Methodological quality of included systematic reviews
First author, Year | Type of Chemotherapy | Clinical Outcome | MA metric | AMSTAR2 | GRADE |
---|---|---|---|---|---|
ALESSANDRO RIZZO,2020 [22] | G-based + anti-EGFR vs. G-based | OS | HR | Critically low | Low |
PFS | HR | Critically low | Low | ||
ORR | RR | Critically low | Very Low | ||
Toxicities–Neutropenia | RR | Critically low | Low | ||
Toxicities–Thrombocytopenia | RR | Critically low | Very Low | ||
Toxicities–Skin rash | RR | Critically low | Moderate | ||
Toxicities–Diarrhea | RR | Critically low | Very Low | ||
Toxicities–Fatigue | RR | Critically low | Very Low | ||
Lawrence Chen,2016 [16] | GP vs. G | Duration of OS | MD | Critically low | Low |
OS | HR | Critically low | Low | ||
Duration of PFS | MD | Critically low | Low | ||
PFS | HR | Critically low | Low | ||
ORR | OR | Critically low | Low | ||
GP + anti-EGFR vs. GP | Duration of OS | MD | Critically low | Low | |
OS | HR | Critically low | Low | ||
Duration of PFS | MD | Critically low | Low | ||
PFS | HR | Critically low | Low | ||
ORR | OR | Critically low | Low | ||
Ting Zheng,2020 [19] | GP vs. FP | ORR | RR | Critically low | Very Low |
DCR | RR | Critically low | Very Low | ||
PFS/TTP | HR | Critically low | Low | ||
OS | HR | Critically low | Low | ||
Toxicities–Neutropenia | NA | Critically low | Very Low | ||
Toxicities–Anemia | NA | Critically low | Very Low | ||
Toxicities–Trombocytopenia | NA | Critically low | Very Low | ||
Toxicities–Nausea/Vomiting | NA | Critically low | Very Low | ||
Toxicities–Anorexia | NA | Critically low | Very Low | ||
Toxicities–Nephropathy | NA | Critically low | Very Low | ||
Toxicities–Neuropathy | NA | Critically low | Very Low | ||
Xin ZHUANG,2017 [21] | G-based + anti-EGFR vs. G-based | Toxicities–Neutropenia | OR | Low | Very Low |
Toxicities–Thrombocytopenia | OR | Low | Very Low | ||
Toxicities–Anemia | OR | Low | Very Low | ||
Toxicities–Peripheral neuropathy | OR | Low | Very Low | ||
Toxicities–Increased AST/ALT | OR | Low | Very Low | ||
Heng Liu,2014 [18] | G-based vs. non-G-based | DRR | OR | Low | Low |
DCR | OR | Low | Very Low | ||
PFS | OR | Low | Low | ||
OS | OR | Low | Low | ||
Toxicities–Leukopenia | OR | Low | Moderate | ||
Toxicities–Anemia | OR | Low | High | ||
Toxicities–Neutropenia | OR | Low | Very Low | ||
Toxicities–Thrombocytopenia | OR | Low | Very Low | ||
Toxicities– Increased ALT level | OR | Low | Very Low | ||
G-based vs. G | Toxicities–Leukopenia | OR | Low | Low | |
Toxicities–Anemia | OR | Low | Very Low | ||
Toxicities–Neutropenia | OR | Low | Low | ||
Toxicities–Thrombocytopenia | OR | Low | Very Low | ||
Toxicities– Increased ALT level | OR | Low | Very Low | ||
Sheng Zhao,2016 [17] | G-based + anti-VEGFR/EGFR vs. G-based | Toxicities– Nausea | RR | Critically low | Very Low |
Toxicities– Vomiting | RR | Critically low | Very Low | ||
Toxicities– Diarrhea | RR | Critically low | Moderate | ||
Alessandro Rizzo,2022 [26] | Fluoropyrimidine-based doublet CHT vs. ASC or 5-FU/LV | OS | HR | Critically low | Low |
DCR | OR | Critically low | Moderate | ||
ORR | OR | Critically low | Moderate | ||
Wen-Jie Ma,2020 [27] | Fluoropyrimidine-based vs. Observation | OS | HR | Critically low | Moderate |
G-based vs. Observation | OS | HR | Critically low | Moderate | |
Julien Edeline,2022 [28] | G-based vs. Observation | RFS-All Patients | HR | Critically low | Low |
RFS-R1 resection Patients | HR | Critically low | Low | ||
RFS-N + tumor Patients | HR | Critically low | Low | ||
OS-All Patients | HR | Critically low | Low | ||
OS-R1 resection Patients | HR | Critically low | Low | ||
OS-N + tumor Patients | HR | Critically low | Low | ||
Abdel-Rahman O,2018 [29] | G + S-1 vs. S-1 | All-cause mortality at 1 year | RR | High | Very Low |
ORR(S-1 vs.G + S-1) | RR | High | Moderate | ||
Toxicities–Grade 1—4 Anaemia | RR | High | Very Low | ||
Toxicities–Grade 1—4 Thrombocytopenia | RR | High | Moderate | ||
Toxicities–Grade 1—4 Neutropenia | RR | High | Moderate | ||
Toxicities–Febrile Neutropenia | RR | High | Very Low | ||
Yan Li,2019 [30] | FOLFOX-4 vs. Observation | OS | HR | Critically low | — |
XP vs. GP | OS | HR | Critically low | — | |
G + S-1 vs. GC | OS | HR | Critically low | — | |
Jie Ying,2019 [31] | G-based CHT vs. single CHT(mainly FU alone) | DCR | RR | Critically low | — |
Fluoropyrimidine-based CHT vs. single TAs | DCR | RR | Critically low | — | |
Taxanes-based CHT vs. single CHT(mainly FU alone) | DCR | RR | Critically low | — | |
Fluoropyrimidine-based CHT vs. single CHT(mainly FU alone) | 1-year OS | RR | Critically low | — | |
Wei Zheng,2019 [32] | G + S-1 vs. G | ORR | OR | Critically low | — |
OS | HR | Critically low | — | ||
G + S-1 vs. 5-FU | ORR | OR | Critically low | — | |
OS | HR | Critically low | — | ||
CapC vs.5-FU | ORR | OR | Critically low | — | |
GEMOX vs. 5-FU | OS | HR | Critically low | — | |
FP vs. 5-FU | OS | HR | Critically low | — | |
Yanfeng Jiang,2021 [33] | Observation vs. Folfox-4 | PFS | HR | Critically low | — |
Observation vs. C-GEMOX | PFS | HR | Critically low | — | |
Observation vs. GEMOX + erlotinib | PFS | HR | Critically low | — | |
GP + cediranib vs. 5-FU | ORR (5-FU vs.GP + cediranib) | OR | Critically low | — | |
Toxicities–Neutropenia | OR | Critically low | — | ||
GP vs. 5-FU | Toxicities–Neutropenia | OR | Critically low | — | |
G + S-1 vs. 5-FU | Toxicities–Neutropenia | OR | Critically low | — | |
C-GEMOX vs. 5-FU | Toxicities–Neutropenia | OR | Critically low | — | |
RAM + GP vs. 5-FU | Toxicities–Neutropenia | OR | Critically low | — | |
MER + GP vs. 5-FU | Toxicities–Neutropenia | OR | Critically low | — | |
XELOX vs. G + XELOX | Toxicities–Vomiting | OR | Critically low | — | |
XELOX vs. GEMOX + erlotinib | Toxicities– Diarrhea | OR | Critically low | — |
Certainty of evidence
First author, Year | Type of Chemotherapy | Study design | Clinical Outcome | Effects model | MA metric | Estimates | 95%CI-low | 95%CI-up | p-value | I2% | P-value for TES | P-value forEgger test | MA Quality Assessment |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ALESSANDRO RIZZO,2020 [22] | G-based + anti-EGFR vs. G-based | RCTs | OS | Random | HR | 0.82 | 0.64 | 1.06 | 0.13 | 33 | 0.2472 | NA | Cochrane ROB Tool |
PFS | Fixed | HR | 0.88 | 0.73 | 1.08 | 0.22 | 0 | 1 | NA | ||||
ORR | Fixed | RR | 1.34 | 0.91 | 1.99 | 0.14 | 0 | 1 | NA | ||||
Toxicities–Neutropenia | Fixed | RR | 1.95 | 1.13 | 3.36 | 0.02 | 0 | 0.2367 | NA | ||||
Toxicities–Thrombocytopenia | Fixed | RR | 1.69 | 0.99 | 2.87 | 0.05 | 0 | 1 | NA | ||||
Toxicities–Skin rash | Fixed | RR | 18.11 | 5.13 | 63.91 | < 0.00001 | 0 | 0.8343 | NA | ||||
Toxicities–Diarrhea | Fixed | RR | 1.65 | 0.89 | 3.04 | 0.11 | 0 | 1 | NA | ||||
Toxicities–Fatigue | Fixed | RR | 2.01 | 0.91 | 4.44 | 0.09 | 0 | 1 | NA | ||||
Lawrence Chen,2016 [16] | GP vs. G | RCTs | Duration of OS | Random | WMD | -3.52 | -5.14 | -1.35 | 0.0008 | 0 | — | NA | Cochrane ROB Tool |
OS | Random | HR | 0.65 | 0.53 | 0.79 | < 0.00001 | 0 | 0.8264 | NA | ||||
Duration of PFS | Random | WMD | -2.60 | -3.81 | -1.40 | < 0.00001 | 0 | — | NA | ||||
PFS | Random | HR | 0.63 | 0.52 | 0.76 | < 0.00001 | 0 | 0.8544 | NA | ||||
ORR | Random | OR | 0.53 | 0.31 | 0.88 | 0.02 | 0 | 0.6220 | NA | ||||
GP + anti-EGFR vs. GP | Duration of OS | Random | WMD | -1.49 | -2.56 | -0.43 | 0.006 | 0 | — | NA | |||
OS | Random | HR | 0.90 | 0.70 | 1.15 | 0.39 | 0 | 1 | NA | ||||
Duration of PFS | Random | WMD | -0.07 | -1.91 | 1.77 | 0.94 | 0 | — | NA | ||||
PFS | Random | HR | 0.79 | 0.63 | 0.99 | 0.04 | 0 | 1 | NA | ||||
ORR | Random | OR | 0.56 | 0.38 | 0.83 | 0.003 | 0 | 0.9783 | NA | ||||
Ting Zheng,2020 [19] | GP vs. FP | Restrospective;RCTs | ORR | Fixed | RR | 1.13 | 0.80 | 1.58 | > 0.05 | 44.9 | 0.482 | 0.187 | Cochrane ROB Tool; NOS |
DCR | Fixed | RR | 1.02 | 0.91 | 1.13 | > 0.05 | 12 | 0.7508 | 0.209 | ||||
PFS/TTP | Fixed | HR | 0.95 | 0.86 | 1.05 | > 0.05 | 0 | 1 | NA | ||||
OS | Fixed | HR | 1.06 | 0.98 | 1.14 | > 0.05 | 37.3 | 0.2265 | NA | ||||
Toxicities–Neutropenia | NA | NA | 16.7 (9.3–25.8) vs. 19.3 (3.7–43.3) | < 0.001 | 91.4 | — | NA | ||||||
Toxicities–Anemia | NA | NA | 5.6 (1.1–13.3) vs. 13.1 (7.8–19.5) | < 0.001 | 81.5 | — | NA | ||||||
Toxicities–Trombocytopenia | NA | NA | 6 (2.7–10.5) vs. 10.3 (2.7–22.1) | < 0.001 | 79.5 | — | NA | ||||||
Toxicities–Nausea/Vomiting | NA | NA | 5.7 (4–7.7) vs. 7.8 (5.6–10.3) | < 0.001 | 0 | — | NA | ||||||
Toxicities–Anorexia | NA | NA | 2.2 (1–3.7) vs. 3.1 (0.2–9.3) | < 0.001 | 64.9 | — | NA | ||||||
Toxicities–Nephropathy | NA | NA | 1.1 (0.2–2.7) vs. 2.9 (0.7–6.6) | < 0.001 | 67.6 | — | NA | ||||||
Toxicities–Neuropathy | NA | NA | 0.9 (0.3–1.8) vs. 2.6 (1.4–4.1) | < 0.001 | 47.9 | — | NA | ||||||
Xin ZHUANG,2017 [21] | G-based + anti-EGFR vs. G-based | RCTs | Toxicities–Neutropenia | Fixed | OR | 1.37 | 0.89 | 2.12 | 0.15 | 0 | 0.3165 | NA | Jadad Scale |
Toxicities–Thrombocytopenia | Fixed | OR | 1.4 | 0.83 | 2.39 | 0.21 | 48 | 1 | NA | ||||
Toxicities–Anemia | Fixed | OR | 1.21 | 0.62 | 2.38 | 0.57 | 0 | 1 | NA | ||||
Toxicities–Peripheral neuropathy | Fixed | OR | 1.52 | 0.81 | 2.88 | 0.19 | 0 | 0.2096 | NA | ||||
Toxicities–Increased AST/ALT | Fixed | OR | 1.4 | 0.82 | 2.39 | 0.22 | 0 | 1 | NA | ||||
Heng Liu,2014 [18] | G-based vs. non-G-based | RCTs | DRR | Fixed | OR | 1.39 | 0.81 | 2.40 | 0.24 | 48 | 0.8733 | 1.00 | Jadad Scale |
DCR | Random | OR | 1.48 | 0.43 | 5.07 | 0.53 | 81 | 0.8312 | 0.23 | ||||
PFS(months) | Random | OR | 1.78 | -0.39 | 3.96 | 0.11 | 99 | — | 0.55 | ||||
OS(months) | Random | OR | 1.51 | -1.37 | 4.38 | 0.3 | 99 | — | 1.00 | ||||
Toxicities–Leukopenia | Random | OR | 7.17 | 1.43 | 36.08 | 0.02 | 64 | 0.8917 | NA | ||||
Toxicities–Anemia | Fixed | OR | 7.04 | 2.59 | 19.12 | 0.00001 | 0 | 0.1356 | NA | ||||
Toxicities–Neutropenia | Random | OR | 4.63 | 0.95 | 22.50 | 0.06 | 82 | 0.9924 | NA | ||||
Toxicities–Thrombocytopenia | Random | OR | 2.79 | 0.66 | 11.81 | 0.16 | 70 | 0.3348 | NA | ||||
Toxicities– Increased ALT level | Fixed | OR | 1.11 | 0.56 | 2.23 | 0.76 | 46 | 1 | NA | ||||
G-based vs. G | Toxicities–Leukopenia | Random | OR | 1.82 | 1.13 | 2.94 | 0.01 | 0 | 1 | NA | |||
Toxicities–Anemia | Fixed | OR | 1.96 | 1.07 | 3.62 | 0.03 | 54 | 0.3617 | NA | ||||
Toxicities–Neutropenia | Random | OR | 1.78 | 1.19 | 2.66 | 0.005 | 0 | 0.6637 | NA | ||||
Toxicities–Thrombocytopenia | Random | OR | 1.13 | 0.6 | 2.14 | 0.71 | 0 | 1 | NA | ||||
Toxicities– Increased ALT level | Fixed | OR | 0.76 | 0.47 | 1.25 | 0.29 | 66 | 1 | NA | ||||
Sheng Zhao,2016 [17] | G-based + anti-VEGFR/EGFR vs. G-based | RCTs | Toxicities– Nausea | NA | RR | 1.01 | 0.41 | 2.47 | 0.98 | No significant | — | NA | Jadad Scale |
Toxicities– Vomiting | NA | RR | 0.71 | 0.31 | 1.60 | 0.41 | No significant | — | NA | ||||
Toxicities– Diarrhea | NA | RR | 2.48 | 1.2 | 5.10 | 0.014 | No significant | — | NA | ||||
Alessandro Rizzo,2022 [26] | Fluoropyrimidine-based doublet CHT vs. ASC or 5-FU/LV | RCTs | OS | Fixed | HR | 0.63 | 0.49 | 0.80 | < 0.0001 | 0 | 0.5867 | NA | Cochrane ROB Tool |
DCR | Fixed | OR | 5.18 | 3.3 | 10.23 | NA | 84 | 0.9772 | NA | ||||
ORR | Random | OR | 3.24 | 1.18 | 8.92 | NA | 0 | 1 | NA | ||||
Wen-Jie Ma,2020 [27] | Fluoropyrimidine-based vs. Observation | RCTs | OS | Random | HR | 0.83 | 0.7 | 0.99 | 0.04 | 13 | 0.5350 | 0.62 | Jadad Scale |
G-based vs. Observation | OS | Random | HR | 0.91 | 0.74 | 1.12 | 0.37 | 2 | 1 | 0.62 | |||
Julien Edeline,2022 [28] | G-based vs. Observation | RCTs | RFS-All Patients | NA | HR | 0.91 | 0.71 | 1.16 | 0.46 | NA | — | NA | NA |
RFS-R1 resection Patients | NA | HR | 1.10 | 0.58 | 2.07 | 0.77 | NA | — | NA | ||||
RFS-N + tumor Patients | NA | HR | 0.86 | 0.60 | 1.23 | 0.40 | NA | — | NA | ||||
OS-All Patients | NA | HR | 1.03 | 0.78 | 1.35 | 0.85 | NA | — | NA | ||||
OS-R1 resection Patients | NA | HR | 1.25 | 0.63 | 2.49 | 0.52 | NA | — | NA | ||||
OS-N + tumor Patients | NA | HR | 0.99 | 0.67 | 1.46 | 0.94 | NA | — | NA | ||||
Abdel-Rahman O,2018 [29] | G + S-1 vs. S-1 | RCTs | All-cause mortality at 1 year | Random | RR | 0.61 | 0.33 | 1.13 | 0.12 | 76 | 0.2928 | NA | Cochrane ROB Tool |
ORR(S-1 vs.G + S-1) | Random | RR | 2.46 | 1.27 | 4.57 | 0.0073 | 0 | 0.5223 | NA | ||||
Toxicities–Grade 1—4 Anaemia | Random | RR | 1.26 | 1.00 | 1.59 | 0.052 | 0 | 1 | NA | ||||
Toxicities–Grade 1—4 Thrombocytopenia | Random | RR | 2.45 | 1.39 | 4.32 | 0.0019 | 0 | 0.6803 | NA | ||||
Toxicities–Grade 1—4 Neutropenia | Random | RR | 3.30 | 1.04 | 10.50 | 0.043 | 66 | 0.4714 | NA | ||||
Toxicities–Febrile Neutropenia | Random | RR | 2.97 | 0.32 | 27.87 | 0.34 | 0 | 1 | NA | ||||
Yan Li,2019 [30] | Observation vs. FOLFOX-4 | Restrospective;RCTs | OS | NA | HR | 3.40 | 1.70 | 6.70 | NA | NA | — | NA | NA |
XP vs. GP | OS | NA | HR | 0.74 | 0.51 | 1.10 | NA | NA | — | NA | |||
G + S-1 vs. GC | OS | NA | HR | 1.10 | 0.71 | 1.50 | NA | NA | — | NA | |||
Jie Ying,2019 [31] | G-based CHT vs. single CHT(mainly FU alone) | Restrospective;RCTs | DCR | NA | RR | 1.36 | 1.04 | 1.80 | 0.012 | NA | — | Funnel plot only | NOS |
Fluoropyrimidine-based CHT vs. single TAs | DCR | NA | RR | 0.78 | 0.61 | 1.00 | 0.03 | NA | — | Funnel plot only | |||
Taxanes-based CHT vs. single CHT(mainly FU alone) | DCR | NA | RR | 1.54 | 1.02 | 2.32 | 0.02 | NA | — | Funnel plot only | |||
Fluoropyrimidine-based CHT vs. single CHT(mainly FU alone) | 1-year OS | NA | RR | 0.51 | 0.29 | 0.87 | 0.006 | NA | — | Funnel plot only | |||
Wei Zheng,2019 [32] | G + S-1 vs. G | RCTs | ORR | NA | OR | 4.72 | 1.31 | 17.02 | NA | NA | — | Funnel plot only | NA |
OS | NA | HR | 0.43 | 0.20 | 0.93 | NA | NA | — | Funnel plot only | ||||
G + S-1 vs. 5-FU | ORR | NA | OR | 9.08 | 1.56 | 89.20 | NA | NA | — | Funnel plot only | |||
OS | NA | HR | 0.51 | 0.28 | 0.96 | NA | NA | — | Funnel plot only | ||||
CapC vs.5-FU | ORR | NA | OR | 5.46 | 1.07 | 56.63 | NA | NA | — | Funnel plot only | |||
GEMOX vs. 5-FU | OS | NA | HR | 0.57 | 0.32 | 0.96 | NA | NA | — | Funnel plot only | |||
FP vs. 5-FU | OS | NA | HR | 1.88 | 1.07 | 3.16 | NA | NA | — | Funnel plot only | |||
Yanfeng Jiang,2021 [33] | Observation vs. Folfox-4 | Restrospective;RCTs | PFS | NA | HR | 2.88 | 1.05 | 7.93 | NA | NA | — | Funnel plot only | Cochrane ROB Tool |
Observation vs. C-GEMOX | PFS | NA | HR | 2.82 | 1.20 | 6.62 | NA | NA | — | Funnel plot only | |||
Observation vs. GEMOX + erlotinib | PFS | NA | HR | 3.21 | 1.38 | 7.56 | NA | NA | — | Funnel plot only | |||
GP + cediranib vs. 5-FU | ORR (5-FU vs.GP + cediranib) | NA | OR | 0.13 | 0.02 | 0.87 | NA | NA | — | Funnel plot only | |||
Toxicities–Neutropenia | NA | OR | 0.04 | 0 | 0.65 | NA | NA | — | Funnel plot only | ||||
GP vs. 5-FU | Toxicities–Neutropenia | NA | OR | 0.06 | 0.01 | 0.50 | NA | NA | — | Funnel plot only | |||
G + S-1 vs. 5-FU | Toxicities–Neutropenia | NA | OR | 0.05 | 0 | 0.55 | NA | NA | — | Funnel plot only | |||
C-GEMOX vs. 5-FU | Toxicities–Neutropenia | NA | OR | 0.08 | 0.01 | 0.60 | NA | NA | — | Funnel plot only | |||
RAM + GP vs. 5-FU | Toxicities–Neutropenia | NA | OR | 0.03 | 0 | 0.38 | NA | NA | — | Funnel plot only | |||
MER + GP vs. 5-FU | Toxicities–Neutropenia | NA | OR | 0.03 | 0 | 0.41 | NA | NA | — | Funnel plot only | |||
XELOX vs. G + XELOX | Toxicities–Vomiting | NA | OR | 0.07 | 0 | 0.98 | NA | NA | — | Funnel plot only | |||
XELOX vs. GEMOX + erlotinib | Toxicities– Diarrhea | NA | OR | 0.09 | 0.01 | 0.63 | NA | NA | — | Funnel plot only |
Effect of interventions
Gemcitabine-based Chemotherapy + Targeted Therapy versus Gemcitabine-based Chemotherapy
Gemcitabine-based Chemotherapy versus Gemcitabine monotherapy
Fluoropyrimidine + Cisplatin versus Gemcitabine + Cisplatin
Gemcitabine-based Chemotherapy versus Non-Gemcitabine-based Chemotherapy
Fluoropyrimidine-based Chemotherapy
Gemcitabine-based Chemotherapy versus Observation
Comparison between chemotherapy and targeted therapy regimens in the network meta-analysis
Discussion
Summary of main results
Overall completeness and applicability of evidence
-
High: We are certain that the real effect value is close to the estimated effect value, and further research is unlikely to change our confidence in the estimated effect value [42].
-
Moderate: We have moderate confidence in the estimated value of the effect, and the real value may be close to the estimated value, but the two values may be quite different [42].
-
Low: Our confidence in the estimated value of the effect is limited, and the actual value may be quite different from the estimated value [42].
-
Very Low: We have little confidence in the estimated value of the effect. The real value is likely to be quite different from the estimated value, and any estimated value of the effect is very uncertain [42].
-
(1) clear evidence of benefit;
-
(2) clear evidence of harm;
-
(3) clear evidence of no effect or equivalence; and.
-
(4) unknown benefit or harm or no effect or equivalence.