3.1. Literature Search and Study Characteristics
According to the previous search strategy, 25961 citations were obtained from the online database from January 2010 to January 2020. Then, 1325 records were excluded by viewing title and abstract. Among the remaining 187 records, 175 citations were removed with various reasons. Finally, 11 full-text studies were suitable for this meta-analysis[9-19](Fig.1). Besides the characteristics, quality evaluation, and demographics of the included studies are summarized in Table 1.
3.2. Data Analysis
3.2.1. The second stage hepatectomy
Eleven literatures described the completion rate of two-step surgery in two groups, including a total of 867 patients[9-19]. There were 247 cases in ALPPS group and 620 cases in PVE group. The completion rate of two-step surgery was significantly different between the two groups (OR=11.25, 95%CI: 5.64~22.43, Z=6.87, P<0.001) (Fig. 2A).
3.2.2. The sufficient FLR growth
The rate of FLR after the first operation was described in 10 articles, including 845 patients[9-14, 16-19]. There were 239 cases in ALPPS group and 606 cases in PVE group. The difference in FLR growth rate between the two groups was statistically significant (MD=46.85, 95% CI:4~89.70, Z=2.41, P=0.03) (Fig. 2B).
3.2.3. The time to stage II operation
A total of 671 patients were included in 9 articles describing the time to stage II operation[9, 11, 12, 14-19]. There were 188 cases in ALPPS group and 483 cases in PVE group. The difference between the two groups was statistically significant (MD=-22.85, 95% CI:-33.87~-11.84, Z=4.07, P<0.001) (Fig. 2C).
3.2.4. The rate of R0 resection
The R0 removal rate was described in five literatures, involving a total of 307 patients[9, 10, 13, 14, 18]. There were 128 cases in ALPPS group and 179 cases in PVE group. The difference of R0 removal rate between the two groups was statistically significant (OR=2.29, 95%CI: 1.07~4.90, Z=2.13, P=0.03) (Fig. 2D).
3.2.5. Mortality within 90 days of perioperative period
All included literatures described the case fatality rate within 90 days of the perioperative period, including a total of 791 patients[9-19]. Among them, 240 cases in ALPPS group and 451 cases in PVE group showed no significant difference in case fatality rate within 90 days between the two groups (OR=1.31, 95%CI: 0.73~2.34, Z=0.91, P=0.36) (Fig. 2E).
3.2.6. Overall postoperative complication rate
All included literatures described the overall incidence of postoperative complications (including biliary fistula, sepsis, abdominal infection, etc.), including a total of 791 patients[9-19]. Among them, 240 cases in ALPPS group and 451 cases in PVE group showed no significant difference in the overall complication rate between the two groups (OR=1.14, 95%CI: 0.63~2.07, Z=0.44, P=0.66) (Fig. 2F).
3.2.7. The incidence of postoperative liver failure
The incidence of postoperative liver failure was described in 7 literatures, involving a total of 428 patients[11-17]. There were 119 cases in ALPPS group and 309 cases in PVE group, and there was no significant difference in the incidence of postoperative liver failure between the two groups (OR=0.83, 95%CI: 0.42~1.63, Z=0.54, P=0.59) (Fig. 2G).
3.2.8. Postoperative hospital stay
The total length of stay was described in 4 articles, including 283 patients. There were 77 cases in ALPPS group and 206 cases in PVE group. There was no significant difference in total hospitalization time between the two groups (OR=4.71, 95%CI: -1.25~10.66, Z=1.55, P=0.12) (Fig. 2H).
3.2.9. Immunohistochemical analysis of Ki-67 in tumor tissue
Only one literature described the expression of ki-67 after surgery[13]. At first hepatectomy, Ki-67 expression was evident in 28.2±42.7% of cells in the ALPPS group and 51.7±35.6% in the PVE group (P=0.09). At second hepatectomy, expression was 20.5±24.7% in the ALPPS group and 54.5±26.9% in the PVE group (P=0.01). No difference of Ki-67 expression in liver tumor cells was seen at first hepatectomy, but expression at second hepatectomy was greater in the PVE group than the ALPPS group.
3.3. Publication Bias and Sensitivity Analysis
On the basis of the overall incidence of perioperative complications in the ALPPS group and the PVE group, the Bgger’s test was performed to assess the publication bias of the literature. Funnel plots indicated a symmetric distribution of included studied. Begg tests demonstrated that there was no potential publication bias among studies (Begg test, P=0.94). Sensitivity analyses have confirmed the robustness of the results (Fig. 3).