Search results and characteristics of eligible studies
The systematic literature search returned 885 records (Fig. 1). Finally, 11 potentially eligible RCTs [33–42, 54] and 11 case-control studies [43–53] were considered, consisting of approximately 8 052 participants. The population of patients in each study varied from 50 to 1 060 cases, and the follow-up time ranged from 3.4 to 120 months. Six and 12 publications had available data for the OS and DFS analyses respectively. Nineteen studies provided evidence of the prognostic value of TILs for PCR. Cut-off chosen were 10% (n = 2); 30% (n = 2); 40% (n = 1); 50% (n = 3); 60% (n = 7); 70% (n = 1). The continuous variable most used per 10% increment. The majority of NACT regimen contains anthracycline and taxane. Trastuzumab or lapatinib were typically used in HER2-positive patients. The basic characteristics and target outcomes extracted from the included studies are listed in Table 1.
Literature quality evaluation results
We evaluated the risk of bias for all included prospective studies (n = 11). The main sources of bias were related to blinded (participants and personnel). The risk of bias assessments for each cohort and evaluations for each domain across full reported studies are shown in Supplementary Fig. 1a and Fig. 1b.
Correlation of TILs with clinicopathological parameters
The correlations between clinicopathologic characteristics and TILs were analyzed in Table 2. Total TILs levels were not associated with tumor grade, tumor size and patients age, but there was association with shorter DFS in node-positive breast cancer patients (pooled HR = 3.340, 95%CI, 2.280–4.890). Unfortunately, we had not analyzed the relationship of Ki-67, hormone receptor or HER2 status due to the limited data.
TILs and PCR
The 17 studies were eligible to be assessed for TILs and PCR. For binary variable, high level of TILs was correlated with better PCR rate, pooled OR = 3.182, 95%CI, 2.549–3.973, P = 0.000, and no significant heterogeneity (Fig. 2a). In a subgroup analysis of different types of breast cancer, for HER2- positive breast cancer (pooled OR = 2.329, 95%CI, 1.174–4.621, P = 0.016), HER2-negative breast cancer (pooled OR = 3.386, 95%CI, 2.242–5.115, P = 0.000) and TNBC (pooled OR = 7.571, 95%CI, 3.631–15.784, P = 0.000), higher total TILs level was correlated with better PCR (Fig. 2b, Fig. 2c and Fig. 2d ), but not for ER negative and ER positive breast cancer (Fig. 2e and Fig. 2f ).
In a subgroup analysis, high level of TILs subtypes, CD8+ TILs (polled OR = 3.300, 95%CI, 1.730–6.294, P = 0.000) and Foxp3+ TILs (polled OR = 2.353, 95%CI, 1.273–4.347, P = 0.006) also predicted better pathological response to NACT. For HER2-positive patients received trastuzumab or lapatinib, meta- analysis showed that high levels of TILs may not enhance the efficacy of an anti-HER2 therapy by trastuzumab (pooled OR = 0.670, 95%CI, 0.081–5.504, P = 0.709) and lapatinib (pooled OR = 1.307, 95%CI, 0.123–13.878, P = 0.824).
Cut-off value of TILs and PCR
We used predefined cut-off points to distinguish high and low TIL subgroups. In studies (types of breast cancer not clearly classified in the literature) and where the cut-off value for TILs was ≥ 10%, the high level of TILs can predict PCR rate better. This cut-off value of TILs applied to the TNBC patients, either. However, for HER2-positive breast cancer patients, subgroup analysis showed in 2 studies where the cut-off value for TILs was 10%, the pooled OR for PCR was 0.683 (95% CI, 0.029–16.014, P = 0.813). If we grouped studies with a cut-off of TILs ≥ 30%, the OR was 2.631 (95% CI, 1.739–3.982, P = 0.000) and no significant heterogeneity (P = 0.104, I2 = 45.2%).
TILs and DFS
A total of 8 studies were eligible to be assessed for TILs and DFS. When TIL level was assessed as a continuous variable (per 10% increase), patients with increased TIL level in breast cancer had significantly longer DFS than did patients with lower TIL levels (pooled HR = 0.95, 95%CI, 0.92–0.98, P = 0.0003) (Fig. 3a). The pooled data suggested both iTILs (pooled HR = 0.91, 95%CI, 0.84–0.98, P = 0.020) and sTILs (pooled HR = 0.96, 95%CI, 0.93–0.98, P = 0.003) were associated with better DFS.
In the HER2-positive patients (pooled HR = 0.96, 95%CI, 0.94–0.97, P < 0.00001) and TNBC patients (pooled HR = 0.85, 95%CI, 0.78–0.92, P < 0.0001), TILs were significantly associated with better DFS (Fig. 3b and Fig. 3c). The association between TILs and survival was similar between iTILs and sTILs. Subgroup analyses according to TILs subtypes, Both CD8+ (polled HR 1.00, 95% CI 1.00–1.00) and CXCL13 (polled HR 0.55, 95% CI 0.38–0.78) predicted better for DFS after NACT. On the contrary, the meta-analysis confirmed that for any 10% increase of FOXP3+ level there was a poor DFS (polled HR 1.11, 95% CI 0.76–1.62).
TILs and OS
We assessed TILs as a prognostic marker for OS from 6 studies. Meta-analysis results showed that breast cancer patients with high level of TILs showed a favorable OS after NACT (pooled HR = 0.90, 95%CI, 0.85–0.95, P < 0.0001) (Fig. 4a). Both iTILs (pooled HR = 0.86, 95%CI, 0.75–0.99, P = 0.030) and sTILs (pooled HR = 0.91, 95%CI, 0.86–0.96, P = 0.0006) achieved similar results. In a subgroup analysis of types of breast cancer, a 10% increase in TILs was associated with longer OS in HER2-positive breast cancer (pooled HR = 0.93, 95% CI, 0.87–0.99, P = 0.010) and TNBC (pooled HR = 0.86, 95%CI, 0.79–0.93, P = 0.0003) (Fig. 4b and Fig. 4c).
Publication bias
Funnel plot analysis which was performed to assess the publication bias of the selected studies for the pooled PCR rate (Supplementary Fig. 2). Visual inspection of analysis indicated some evidence of asymmetry, but Egger’s tests indicated that there was no significant publication bias, P > 0.5. Limited data for OS outcome indicators, so we did not do the funnel plot.
Sensitivity analysis
To assess the impact of each included study, a sensitivity analysis was performed. After excluding each study, similar results were observed.