Our aim was to examine the association of pretreatment tumor-infiltrating lymphocyte (TIL) count and PD-L1 levels with pathologic complete response (pCR) and assess immune marker changes following treatment in tumor specimens from the S0800 clinical trial, which randomized patients to bevacizumab + nab-paclitaxel, followed by doxorubicin/cyclophosphamide (AC) versus two control arms without bevacizumab (varying sequence of AC and nab-paclitaxel). TILs were assessed in 124 pre- and 62 posttreatment tissues (including 59 pairs). PD-L1 was assessed in 120 pre- and 43 posttreatment tissues (including 39 pairs) using the 22C3 antibody. Baseline and treatment-induced immune changes were correlated with pCR and survival using estrogen receptor (ER) and treatment-adjusted logistic and Cox regressions, respectively. At baseline, the mean TIL count was 17.4% (17% had zero TILs, 9% had ≥50% TILs). Posttreatment, mean TIL count decreased to 11% (5% had no TILs, 2% had >50% TILs). In paired samples, the mean TIL change was 15% decrease. Baseline PD-L1 was detected in 43% of cases (n = 5 in tumor cells, n = 29 stroma, n = 18 tumor + stroma). Posttreatment, PD-L1 expression was not significantly lower (33%). Higher baseline TIL count and PD-L1 positivity rate were associated with higher pCR rate even after adjustment for treatment and ER status (P = 0.018). There was no association between TIL counts, PD-L1 expression, and survival due to few events. In conclusion, TIL counts, but not PD-L1 expression, decreased significantly after treatment. Continued PD-L1 expression in some residual cancers raises the possibility that adjuvant immune checkpoint inhibitor therapy could improve survival in this patient population. Mol Cancer Ther; 17(6); 1324–31. ©2018 AACR.

Neoadjuvant (preoperative) chemotherapy is increasingly used in the treatment of early-stage breast cancer (1) because it leads to higher breast conservation rates among locally advanced cancers, to smaller surgical resection in stage II cancers (2, 3) and the extent of residual cancer provides important prognostic information (4). Pathologic complete response (pCR), defined as no invasive cancer in the breast or lymph nodes after neoadjuvant chemotherapy, is an indicator of excellent survival, whereas extensive residual disease indicates poor prognosis. Patients with residual disease may receive additional chemotherapy, which can improve survival in triple-negative breast cancers (TNBC; ref. 5) or could participate in clinical trials designed for this high-risk population (NCT02954874, NCT02445391, NCT02101385).

Tumor-infiltrating lymphocytes (TIL), or immune-related gene expression signatures, are predictive of higher pCR rates (6–8) and are also associated with better survival in TNBC, HER2+, and high-risk estrogen receptor–positive (ER+) breast cancers (9–12). Surprisingly, high expression of immune checkpoint molecules, such as PD-1 (programmed death receptor 1) and PD-L1 (programmed death ligand 1), which downregulate antitumor immune effector mechanisms, is also associated with higher pCR rate and better prognosis (13–16). This is due to the strong correlation between PD-L1 expression and TIL counts, and it is also suggested that high expression of this checkpoint molecule does not completely eliminate the benefits of antitumor immunosurveillance in lymphocyte-rich cancers.

Several ongoing neoadjuvant clinical trials test whether addition of an immune checkpoint inhibitor to standard-of-care chemotherapy could increase pCR rates and improve survival in early-stage breast cancers, particularly TNBC (17). Understanding how chemotherapy influences the tumor immune microenvironment could assist in designing future studies and develop biomarkers. Preclinical evidence supports that some of the antitumor activity of cytotoxic agents is mediated by antitumor immune response (18). Tumor cell injury from chemotherapy may trigger neoantigen formation, dendritic cell activation, antigen cross-presentation, and cytokine release that ultimately lead to induction of tumor-specific cytotoxic T cells (19). Some chemotherapy drugs can also inhibit myeloid-derived immune suppressor cells and FOXP3+ regulatory T cells (20). VEGF in the tumor microenvironment enhances expression of PD-1 and other inhibitory checkpoints involved in T-cell exhaustion, and this effect can be reverted by antiangiogenic agents such as bevacizumab (21).

S0800 (NCT00856492) was a randomized three-arm phase II trial that assessed whether inclusion of bevacizumab with neoadjuvant chemotherapy could improve pCR rates in HER2, locally advanced, or inflammatory breast cancer (IBC). The three arms of the trial were weekly nab-paclitaxel and bevacizumab followed by dose-dense doxorubicin/cyclophosphamide (ddAC; Arm A), nab-paclitaxel followed by ddAC (Arm B), and ddAC followed by nab-paclitaxel (Arm C). Patients were randomly allocated (2:1:1) to the three arms, but for the primary efficacy analysis, the two non-bevacizumab arms (B and C) were combined. The primary efficacy results were reported earlier (22) and showed that bevacizumab increased pCR rate from 21% to 36% (P = 0.019). In TNBC, the improvement in pCR rate was even higher (29% vs. 59%; P = 0.014), whereas in ER+ cancer, the improvement did not reach statistical significance (18% vs. 24%; P = 0.41). There was also a trend for improved event-free survival (EFS) with the addition of bevacizumab in the TNBC subset (P = 0.06). The main objectives of the current study were to (i) examine the association of pretreatment TIL and PD-L1 levels with pCR and (ii) assess changes in TIL counts and PD-L1 expression between pre- and posttreatment tissues from the S0800 clinical trial. We also assessed associations between these immune markers and EFS and overall survival (OS).

Patients

Baseline core needle biopsies and posttreatment surgical resection specimens were collected prospectively during the trial. Formalin-fixed paraffin-embedded (FFPE) blocks or unstained cut sections were submitted to the SWOG tissue bank. Of the 215 patients registered for the S0800 trial, 211 were available for efficacy analysis, 134 patients had pretreatment and 63 had posttreatment FFPE tissues with consent for research, including 59 paired pre- and posttreatment tissues (CONSORT diagram, Fig. 1). TIL counts could be assessed in 124 pre- and 62 posttreatment tissues, including 59 paired cases. For the remaining cases, the submitted tissue did not contain cancer or the staining procedure failed. PD-L1 IHC could be generated for 120 pretreatment, 43 posttreatment and 39 paired specimens. The missing cases had no adequate tissues for IHC. Patient characteristics of the entire cohort and the current biomarker study subpopulation are shown in Table 1. pCR was determined by the local pathologists, and pCR was defined as the absence of any residual invasive cancer, with or without ductal carcinoma in situ, in the breast and axilla (ypT0/is ypN0). All surgical pathology reports were reviewed centrally for accuracy by the study chair (Z. Nahleh) without the knowledge of treatment assignment. The current biomarker analysis was approved by the Yale Cancer Center Human Investigations Committee.

Figure 1.

CONSORT diagram of samples used in the study.

Figure 1.

CONSORT diagram of samples used in the study.

Close modal
Table 1.

Demographic and disease characteristics for the overall trial population and the immune marker subset

S0800 TotalImmune study
Eligible and maintained consent 211 134 
IBC or LABC 
 IBC 24 (11.4%) 12 (9.0%) 
 LABC 187 (88.6%) 122 (91.0%) 
HR status 
 HR+: ER+ or PR+ 144 (68.2%) 93 (69.4%) 
 HR: ER and PR 67 (31.8%) 41 (30.6%) 
Randomized treatment 
 No bevacizumab 113 (53.5%) 73 (54.5%) 
 Bevacizumab 98 (46.5%) 61 (45.5%) 
Primary outcome   
 No pCR 152 (72.0%) 97 (72.4%) 
 pCR 59 (28.0%) 37 (27.6%) 
S0800 TotalImmune study
Eligible and maintained consent 211 134 
IBC or LABC 
 IBC 24 (11.4%) 12 (9.0%) 
 LABC 187 (88.6%) 122 (91.0%) 
HR status 
 HR+: ER+ or PR+ 144 (68.2%) 93 (69.4%) 
 HR: ER and PR 67 (31.8%) 41 (30.6%) 
Randomized treatment 
 No bevacizumab 113 (53.5%) 73 (54.5%) 
 Bevacizumab 98 (46.5%) 61 (45.5%) 
Primary outcome   
 No pCR 152 (72.0%) 97 (72.4%) 
 pCR 59 (28.0%) 37 (27.6%) 

Abbreviations: HR, hormone receptor; IBC, inflammatory breast cancer; LABC, locally advanced breast cancer.

TIL assessment

TILs were assessed by two pathologists (V. Pelekanou and B. Wasserman) on hematoxylin and eosin (H&E)–stained full sections following the scoring guidelines of the International TILs Working Group (23). In cases with pCR, the tumor bed was examined and scored. Stromal TIL scores were defined as the percentage of tumor stromal area that was occupied by mononuclear inflammatory cells. Inflammatory infiltrates in the stroma of noninvasive lesions and normal breast structures were excluded from TIL counts. The two scores were averaged to obtain the mean TIL percentage.

PD-L1 IHC

PD-L1 IHC was performed on 5-μm whole tissue sections using the FDA-cleared 22C3 assay on the Dako Link 48 platform following the manufacturer's instructions as reported previously (24). For controls, we used the control slide from DAKO 22c3 pharmDx assay that includes a PD-L1–positive (NCI-H226) and a PD-L1–negative (MCF-7) cell line (Supplementary Fig. S1A) and also a tissue microarray assembled in our laboratory that contains 100 spots of randomly selected cases of placenta, tonsil, lung cancer, and cell lines that express broad ranges of PD-L1 (Supplementary Fig. S1B; ref. 24). Two breast pathologists (V. Pelekanou and Y.-C. Lo) scored independently both the tumor and stromal cell compartments as a percentage of cells with PD-L1 signal at any intensity. When greater than 10% absolute difference in percent positive score was observed, the pathologists jointly reviewed the case to arrive at consensus; otherwise, the average of the two pathologists' scores was used as the final PD-L1 percent. PD-L1 positivity threshold was set at ≥1% of either tumor or stromal cells. A similar 1% threshold, using the same 22C3 antibody, was used to select PD-L1–positive metastatic TNBC for anti–PD-1 therapy in a clinical trial that reported an overall response rate of 18.5% with single-agent pembrolizumab in metastatic TNBC (25).

Statistical analysis

All available specimens were used in this study, and sample size was determined by tissue availability. The primary outcome was pCR. Associations with pCR rate were evaluated either using contingency table analyses (using exact methods) or modeled with logistic regression. TIL counts were classified into approximate quartiles, and the four quartile categories were tested either as an ordinal variable or as a categorical variable. The logistic regression analyses were adjusted for hormone receptor status and randomized treatment. The secondary outcomes were OS defined as time from registration to death due to any cause and EFS. Events included progression prior to surgery, local or distant recurrence postsurgery, or death from any cause. Patients without an event were censored at the last known follow-up time. OS and EFS were analyzed using Cox regression adjusting for ER status and randomized treatment assignment. Hazard ratios and 95% confidence intervals (95% CI) are presented. Changes in immune marker levels between pre- and posttreatment samples were compared by pCR outcome or hormone receptor status using a Wilcoxon nonparametric test.

TIL count before and after chemotherapy and its association with outcome

At baseline (n = 124), the mean TIL count was 17.4% (median 10%); 17% of cases had zero TILs, and 9% had ≥50% TILs. Baseline mean and median (15%) TIL percentages were nominally higher in ER (n = 39, mean 20.8%, median 15%) compared with ER+ cancers (n = 85, mean 15.8%, median 7.5%), but these differences did not reach statistical significance (Wilcoxon P = 0.11). Classifying baseline TILs into approximate quartiles showed a significant association with pCR in an ordinal trend test (P = 0.008), but were not significant when treated as four distinct categories (Fisher exact test P = 0.07; Table 2). This pattern remained in a logistic regression adjusting for treatment and ER status (Ptrend = 0.019; categorical P = 0.12). Using TIL counts as a continuous variable adjusted for ER status and bevacizumab treatment, every 10% increase in TILs increased the odds of pCR with an OR of 1.21 (95% CI, 0.99–1.48; P = 0.07).

Table 2.

pCR rates in approximate quartiles of percent TIL categories

No pCRpCRTotal
Baseline TIL quartilen (%)n (%)n
1: <5% 25 (83%) 5 (17%) 30 
2: 5%–10% 27 (75%) 9 (25%) 36 
3: 11%–25% 21 (68%) 10 (32%) 31 
4: 26%–90% 14 (52%) 13 (48%) 27 
Total 87 (70%) 37 (30%) 124 
No pCRpCRTotal
Baseline TIL quartilen (%)n (%)n
1: <5% 25 (83%) 5 (17%) 30 
2: 5%–10% 27 (75%) 9 (25%) 36 
3: 11%–25% 21 (68%) 10 (32%) 31 
4: 26%–90% 14 (52%) 13 (48%) 27 
Total 87 (70%) 37 (30%) 124 

There was no significant association between EFS, OS, and baseline TIL counts either as continuous variable (P = 0.36, P = 0.10) or as quartiles (P = 0.49, P = 0.32). The median follow-up of this study was only 3 years, and only 24 EFS and 19 OS events occurred during this time, which limits the power of the survival analyses for the entire study or for ER subsets.

In the post-NAC samples (n = 62), the mean and median TIL counts were 11% and 7.5%, respectively; 5% of cases had zero TILs, and 2% had ≥50% TILs. In paired pre- and posttreatment samples (n = 59), TIL counts decreased in 78% of cases in the posttreatment samples with a mean change of 15% decrease in TILs. In the remaining 22% of cases (n = 13), TIL counts increased. Among these patients, 3 had pCR. Decrease in TILs was not associated with pCR (P = 1.00), ER status (P = 0.27), or bevacizumab treatment (P = 0.35). Figure 2A shows the distribution of pre- and posttreatment TIL counts in paired samples. Cases with residual disease (n = 44) had lesser absolute TIL decrease (Wilcoxon P = 0.041) than cases with pCR (n = 15) where the tumor bed was assessed (Fig. 2B). The posttreatment decrease in TILs was also observed after excluding cases with pCR. Representative TIL images are shown in Fig. 3A and B.

Figure 2.

TIL counts in baseline tumors and residual disease. A, Distribution of TIL percentage counts before and after neoadjuvant chemotherapy. B, Change in TIL count before and after neoadjuvant chemotherapy in paired samples grouped by pathologic response category (n = 15; no-pCR, n = 44). The mean change was 11% in cases with no-pCR and 26% in cases with pCR (Wilcoxon test P = 0.04).

Figure 2.

TIL counts in baseline tumors and residual disease. A, Distribution of TIL percentage counts before and after neoadjuvant chemotherapy. B, Change in TIL count before and after neoadjuvant chemotherapy in paired samples grouped by pathologic response category (n = 15; no-pCR, n = 44). The mean change was 11% in cases with no-pCR and 26% in cases with pCR (Wilcoxon test P = 0.04).

Close modal
Figure 3.

Representative images of TILs and PD-L1 chromogenic staining. A, Baseline H&E of a case with high TIL count (×40 magnification; scale bar, 100 μm). B, Posttreatment H&E of the same case with decrease of the TIL infiltrate. C, In this case, the PD-L1 immunostaining is mostly observed in cells morphologically compatible with macrophages or fibroblasts. D, Example of PD-L1 immunostaining that is not in lymphocytes in a tumor with high lymphocytic infiltration; staining is localized to cells that are morphologically compatible with macrophages. E, A case with baseline high PD-L1 expression in tumor cells. F, Example of a PD-L1–negative case.

Figure 3.

Representative images of TILs and PD-L1 chromogenic staining. A, Baseline H&E of a case with high TIL count (×40 magnification; scale bar, 100 μm). B, Posttreatment H&E of the same case with decrease of the TIL infiltrate. C, In this case, the PD-L1 immunostaining is mostly observed in cells morphologically compatible with macrophages or fibroblasts. D, Example of PD-L1 immunostaining that is not in lymphocytes in a tumor with high lymphocytic infiltration; staining is localized to cells that are morphologically compatible with macrophages. E, A case with baseline high PD-L1 expression in tumor cells. F, Example of a PD-L1–negative case.

Close modal

PD-L1 expression before and after chemotherapy and its association with outcome

At baseline, PD-L1 expression was detected in 52 of 120 (43%) cases mostly in the stroma (n = 5 PD-L1 staining in tumor only, n = 29 stroma only, n = 18 tumor + stroma). Stromal and tumor PD-L1 percentages were moderately but statistically significantly correlated (Pearson r = 0.56; P < 0.0001). Most of PD-L1 immunostaining in the stroma was observed not on TILs but macrophages and morphologically fibroblast-like cells (Fig. 3C and D). The correlation between baseline PD-L1 expression and TIL count was weak and nonsignificant (tumor cell PD-L1 vs. TILs r = 0.21, P = 0.27; stromal PD-L1 vs. TILs r = 0.21, P = 0.25). Cases with PD-L1 expression at baseline, either in the stroma or in tumor cells, or both, had significantly higher pCR rates 63% versus 37%, compared with cases lacking PD-L1 expression (Fisher exact test P = 0.008). Examples of PD-L1–positive and negative cases are shown in Fig. 3E and F, respectively. In ER and treatment-adjusted logistic regression, every 10% increase in baseline stromal cell PD-L1 percentage had an OR of 3.02 for pCR (95% CI, 1.55–5.89; P = 0.001). Baseline tumor cell PD-L1 expression was not associated with pCR (P = 0.10), which may be due to the limited number of such cases (n = 23). We also did not observe a significant association between baseline PD-L1 expression and EFS (P = 0.93) and OS (P = 0.48).

Posttreatment, PD-L1 expression was seen in 14 of 43 (33%) cases (n = 6 stroma only, n = 8 tumor + stroma). Posttreatment stromal (r = 0.59; P = 0.0002) and tumor cell (r = 0.42; P = 0.014) PD-L1 expression were significantly correlated with posttreatment TIL count. In the 39 paired cases, PD-L1 expression was negative in both the pre- and posttreatment samples in 20 cases, positive in 10, positive at baseline but negative in the posttreatment sample in 6, and negative at baseline but positive after chemotherapy in 3 cases. In these paired samples, posttreatment stromal PD-L1 expression decreased on average by 1%, which did not reach statistical significance (P = 0.44; Fig. 4). The decrease in stromal PD-L1 expression was slightly greater among those with pCR (mean 3.6%; min −12.5%; max 25.05%) than those with residual disease (mean 0.5%; min −31.5%; max 17.5%), but this difference was not statistically significant (Wilcoxon P = 0.77). There was a slight, 0.3% mean, nonsignificant increase in tumor cell PD-L1 expression in residual disease (n = 31). These results suggest that PD-L1 expression remained stable in the tumor microenvironment before and after chemotherapy with or without bevacizumab.

Figure 4.

PD-L1 expression at baseline and in residual disease of paired samples. PD-L1 decrease in expression from baseline to follow-up by residual disease of 39 paired samples. PD-L1 percent decrease from baseline to posttreatment is shown in the box plot classified by cases with pCR (n = 15) or not (no pCR, n = 24).

Figure 4.

PD-L1 expression at baseline and in residual disease of paired samples. PD-L1 decrease in expression from baseline to follow-up by residual disease of 39 paired samples. PD-L1 percent decrease from baseline to posttreatment is shown in the box plot classified by cases with pCR (n = 15) or not (no pCR, n = 24).

Close modal

In this study, we examined changes in TIL count and PD-L1 expression after neoadjuvant chemotherapy and assessed associations between these immune parameters at baseline and pCR rate and survival. The randomized design of the S0800 trial also allowed us to test for interaction between the immune markers and bevacizumab added to paclitaxel/ddAC chemotherapy, although the small sample size limits the power of this analysis. At baseline, 17% of cancers had zero TILs and 9% were TIL predominant. These findings confirm that most breast cancers (73%) contain small but detectable number of TILs, with a median TIL count around 10%. At baseline, PD-L1 expression was observed in 43% of cases. PD-L1 signal was mainly detected in stromal cells (90%), whereas cancer cells stained positive in only 44% of the cases. These observations are consistent with other reports showing that in breast cancer, unlike other tumor types, stromal cells, including TILs but also macrophages and morphologically fibroblast-like cells, are the primary sites of PD-L1 expression (26–28). This suggests that in breast cancer, interruption of PD-1/PD-L1 signaling between various types of immune cells, rather than (or in addition to) between tumor cells and immune cells, is an important mechanism of action of PD-1/PD-L1–targeting antibodies.

We observed that higher baseline TIL counts and PD-L1 positivity were associated with increasing probability of pCR as reported previously (6–8, 13–16). This supports the hypothesis that chemotherapy response is partly mediated by activated cytotoxic T cells (18–21), and frequent PD-L1 expression provides rationale for combining immune checkpoint inhibitors with chemotherapy to increase pCR rates (29–31). One could hypothesize that PD-L1 expression is a sign of an incomplete negative feedback to a robust antitumor immune response. Indeed, PD-L1 expression is highly correlated with the presence of immune effector cells and immune activation signals (6, 8, 10, 11, 13).

Because of the availability of posttreatment tissues, we could examine treatment-induced changes in TIL counts and PD-L1 expression. We anticipated an overall increase in these parameters as clinical (32) and preclinical studies suggested that chemotherapy can render tumor cells more immunogenic (18–21). Preclinical studies also suggested that PD-L1 expression on cancer cells is stimulated by chemotherapy and suppressed by VEGF (21, 33). However, we observed a significant decrease in TIL count, while PD-L1 expression did not change significantly from baseline to posttreatment tissues, either overall or in the bevacizumab-treated arm. Other investigators have also reported chemotherapy-induced decrease of CD3 (total lymphocytes), CD4 (T cells), and CD20 (B cells)-positive cells (34), and gene expression analysis of paired pre- and posttreatment samples demonstrated depletion of immune-related mRNAs in residual cancer (35). These observations suggest that either chemotherapy has a cytotoxic effect on TILs or as the size of the primary tumor decreases in response to therapy, the immunogenic target decreases and the corresponding antitumor immune reaction also winds down. Our finding that the greatest decrease in TILs between matched pre-/posttreatment samples coincides with pCR supports the hypothesis that after complete eradication of the cancer from the breast, the immune response also resolves.

Our small sample size and few recurrence events prevented us from assessing the prognostic impact of TILs in residual cancer. However, several studies demonstrated that higher TIL counts in the residual cancer correlate with better survival after chemotherapy (36, 37). These observations suggest that cancers that remain “immunogenic” after chemotherapy may continue to be subjected to antitumor immunosurveillance that can reduce the risk of distant recurrence. This hypothesis provides a rationale to explore adjuvant immunotherapy in breast cancers with residual disease, such as the currently accruing SWOG S01418/NRG BR006 trial (NCT02954874).

An important limitation of the S0800 trial, designed 10 years ago, is that it included both ER+ and ER patients. Because of the small sample size, no separate, adequately powered analysis could be done by ER subgroups even though today, we recognize the distinct immunologic and molecular characteristics (6, 10–12) and different chemotherapy sensitivities of ER+ and ER cancers. Sampling bias could also have influenced our pre- and posttreatment comparisons, as the pretreatment immune marker assessments were done on core needle biopsies, whereas the posttreatment samples were surgically resected tissues. However, we previously studied the impact of tumor sampling on immune markers and examined TIL subpopulation counts between biopsies from different regions of the same cancer (38). Our results showed that the average lymphocyte score across multiple fields of view from a single biopsy is reasonably representative of the whole cancer.

Our results confirm that higher pretreatment TIL count (as quartiles) and PD-L1 expression are associated with greater probability of pCR, independently of bevacizumab administration. This finding is consistent with the hypothesis that chemotherapy-induced tumor response is partially mediated by immune cells and provides rationale for exploring immune checkpoint inhibitors in the neoadjuvant treatment setting to further increase pCR rates. Several clinical trials now test this hypothesis in the clinic. We also demonstrated that TIL counts were lower in postchemotherapy tissues, while PD-L1 expression remained the same. The continued PD-L1 expression in many residual cancers raises the possibility that anticancer immunosurveillance persists and might be further augmented by adjuvant immune checkpoint inhibitor therapy.

J.R. Gralow is a consultant/advisory board member for AstraZeneca, Merck, Novartis, Pfizer, Puma, and Roche. D.L. Rimm reports receiving commercial research grants from AstraZeneca, Navigate, PerkinElmer, and Ultivue and is a consultant/advisory board member for Astra Zeneca, Cell Signaling Technology, Merck, and Ultivue. No potential conflicts of interest were disclosed by the other authors.

The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Conception and design: V. Pelekanou, W.E. Barlow, Z.A. Nahleh, D. Hayes, G.N. Hortobagyi, P. Porter, L. Pusztai

Development of methodology: V. Pelekanou, D.L. Rimm, L. Pusztai

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): V. Pelekanou, Z.A. Nahleh, B. Wasserman, Y.-C. Lo, M.-K. von Wahlde, G.N. Hortobagyi, J. Gralow

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): V. Pelekanou, W.E. Barlow, C. Hatzis, D.L. Rimm, L. Pusztai

Writing, review, and/or revision of the manuscript: V. Pelekanou, W.E. Barlow, Z.A. Nahleh, B. Wasserman, Y.-C. Lo, D. Hayes, G.N. Hortobagyi, J. Gralow, D. Tripathy, B. Szekely, C. Hatzis, D.L. Rimm, L. Pusztai

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): V. Pelekanou, W.E. Barlow, Z.A. Nahleh, B. Wasserman, Y.-C. Lo, D. Hayes, G.N. Hortobagyi, J. Gralow, D. Tripathy, B. Szekely, C. Hatzis, D.L. Rimm, L. Pusztai

Study supervision: V. Pelekanou, D. Hayes, G.N. Hortobagyi, J. Gralow, L. Pusztai

Other (Chair of the SWOG Breast Cancer Translational Medicine Committee within which this study was conducted, using specimens from a previously conducted SWOG trial, and was integrally involved in discussions of the objectives and performance of this study): D.F. Hayes

Research reported in this publication was supported by the NCI of the NIH under Award Numbers CA180888 (to C. Blanke), CA180819 (to M. LeBlanc), CA180826 (to C. Fuchs), CA180801 (to M. Zalupski), and CA180858 (to C. Eng); a grant from Gilead Sciences to (D.L. Rimm); and in part by Genentech (Roche), Abraxis BioScience (Celgene), Helomics (to SWOG), the Breast Cancer Research Foundation (to D.L. Rimm, C. Hatzis, and L. Pusztai), and the Susan Komen Foundation for The Cure (to L. Pusztai).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Mougalian
SS
,
Soulos
PR
,
Killelea
BK
,
Lannin
DR
,
Abu-Khalaf
MM
,
DiGiovanna
MP
, et al
Use of neoadjuvant chemotherapy for patients with stage I to III breast cancer in the United States
.
Cancer
2015
;
121
:
2544
52
.
2.
Killelea
BK
,
Yang
VQ
,
Mougalian
S
,
Horowitz
NR
,
Pusztai
L
,
Chagpar
AB
, et al
Neoadjuvant chemotherapy for breast cancer increases the rate of breast conservation: results from the national cancer database
.
J Am Coll Surg
2015
;
220
:
1063
9
.
3.
Boughey
JC
,
Peintinger
F
,
Meric-Bernstam
F
,
Perry
AC
,
Hunt
KK
,
Babiera
GV
, et al
Impact of preoperative versus postoperative chemotherapy on the extent and number of surgical procedures in patients treated in randomized clinical trials for breast cancer
.
Ann Surg
2006
;
244
:
464
70
.
4.
Symmans
WF
,
Peintinger
F
,
Hatzis
C
,
Rajan
R
,
Kuerer
H
,
Valero
V
, et al
Measurement of residual breast cancer burden to predict survival after neoadjuvant chemotherapy
.
J Clin Oncol
2007
;
25
:
4414
22
.
5.
Toi
ML-J
,
Lee
ES
,
Ohtani
S
,
Im
Y-H
,
Im
S-A
,
Park
B-W
, et al
A phase III trial of adjuvant capecitabine in breast cancer patients with HER2-negative pathologic residual invasive disease after neoadjuvant chemotherapy (CREATE-X, JBCRG-04)
.
Proceedings of the 38th Annual CTRC-AACR San Antonio Breast Cancer Symposium
: 
2015 Dec 8–12
;
San Antonio, TX
.
Philadelphia (PA)
:
AACR
; 
2015
.
6.
Denkert
C
,
Loibl
S
,
Noske
A
,
Roller
M
,
Muller
BM
,
Komor
M
, et al
Tumor-associated lymphocytes as an independent predictor of response to neoadjuvant chemotherapy in breast cancer
.
J Clin Oncol
2010
;
28
:
105
13
.
7.
Denkert
C
,
von Minckwitz
G
,
Brase
JC
,
Sinn
BV
,
Gade
S
,
Kronenwett
R
, et al
Tumor-infiltrating lymphocytes and response to neoadjuvant chemotherapy with or without carboplatin in human epidermal growth factor receptor 2-positive and triple-negative primary breast cancers
.
J Clin Oncol
2015
;
33
:
983
91
.
8.
Iwamoto
T
,
Bianchini
G
,
Booser
D
,
Qi
Y
,
Coutant
C
,
Shiang
CY
, et al
Gene pathways associated with prognosis and chemotherapy sensitivity in molecular subtypes of breast cancer
.
J Natl Cancer Inst
2011
;
103
:
264
72
.
9.
Adams
S
,
Gray
RJ
,
Demaria
S
,
Goldstein
L
,
Perez
EA
,
Shulman
LN
, et al
Prognostic value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1199
.
J Clin Oncol
2014
;
32
:
2959
66
.
10.
Bianchini
G
,
Qi
Y
,
Alvarez
RH
,
Iwamoto
T
,
Coutant
C
,
Ibrahim
NK
, et al
Molecular anatomy of breast cancer stroma and its prognostic value in estrogen receptor-positive and -negative cancers
.
J Clin Oncol
2010
;
28
:
4316
23
.
11.
Rody
A
,
Holtrich
U
,
Pusztai
L
,
Liedtke
C
,
Gaetje
R
,
Ruckhaeberle
E
, et al
T-cell metagene predicts a favorable prognosis in estrogen receptor-negative and HER2-positive breast cancers
.
Breast Cancer Res
2009
;
11
:
R15
.
12.
Loi
S
,
Sirtaine
N
,
Piette
F
,
Salgado
R
,
Viale
G
,
Van Eenoo
F
, et al
Prognostic and predictive value of tumor-infiltrating lymphocytes in a phase III randomized adjuvant breast cancer trial in node-positive breast cancer comparing the addition of docetaxel to doxorubicin with doxorubicin-based chemotherapy: BIG 02–98
.
J Clin Oncol
2013
;
31
:
860
7
.
13.
Bottai
G
,
Raschioni
C
,
Losurdo
A
,
Di Tommaso
L
,
Tinterri
C
,
Torrisi
R
, et al
An immune stratification reveals a subset of PD-1/LAG-3 double-positive triple-negative breast cancers
.
Breast Cancer Res
2016
;
18
:
121
.
14.
Wimberly
H
,
Brown
JR
,
Schalper
K
,
Haack
H
,
Silver
MR
,
Nixon
C
, et al
PD-L1 expression correlates with tumor-infiltrating lymphocytes and response to neoadjuvant chemotherapy in breast cancer
.
Cancer Immunol Res
2015
;
3
:
326
32
.
15.
Schalper
KA
,
Velcheti
V
,
Carvajal
D
,
Wimberly
H
,
Brown
J
,
Pusztai
L
, et al
In situ tumor PD-L1 mRNA expression is associated with increased TILs and better outcome in breast carcinomas
.
Clin Cancer Res
2014
;
20
:
2773
82
.
16.
Liu
B
,
Cui
J
,
Sun
J
,
Li
J
,
Han
X
,
Guo
J
, et al
Immunolocalization of MMP9 and MMP2 in osteolytic metastasis originating from MDA-MB-231 human breast cancer cells
.
Mol Med Rep
2016
;
14
:
1099
106
.
17.
Pusztai
L
,
Karn
T
,
Safonov
A
,
Abu-Khalaf
MM
,
Bianchini
G
. 
New strategies in breast cancer: immunotherapy
.
Clin Cancer Res
2016
;
22
:
2105
10
.
18.
Sistigu
A
,
Yamazaki
T
,
Vacchelli
E
,
Chaba
K
,
Enot
DP
,
Adam
J
, et al
Cancer cell-autonomous contribution of type I interferon signaling to the efficacy of chemotherapy
.
Nat Med
2014
;
20
:
1301
9
.
19.
Vincent
J
,
Mignot
G
,
Chalmin
F
,
Ladoire
S
,
Bruchard
M
,
Chevriaux
A
, et al
5-Fluorouracil selectively kills tumor-associated myeloid-derived suppressor cells resulting in enhanced T cell-dependent antitumor immunity
.
Cancer Res
2010
;
70
:
3052
61
.
20.
Roselli
M
,
Cereda
V
,
di Bari
MG
,
Formica
V
,
Spila
A
,
Jochems
C
, et al
Effects of conventional therapeutic interventions on the number and function of regulatory T cells
.
Oncoimmunology
2013
;
2
:
e27025
.
21.
Voron
T
,
Colussi
O
,
Marcheteau
E
,
Pernot
S
,
Nizard
M
,
Pointet
AL
, et al
VEGF-A modulates expression of inhibitory checkpoints on CD8+ T cells in tumors
.
J Exp Med
2015
;
212
:
139
48
.
22.
Nahleh
ZA
,
Barlow
WE
,
Hayes
DF
,
Schott
AF
,
Gralow
JR
,
Sikov
WM
, et al
SWOG S0800 (NCI CDR0000636131): addition of bevacizumab to neoadjuvant nab-paclitaxel with dose-dense doxorubicin and cyclophosphamide improves pathologic complete response (pCR) rates in inflammatory or locally advanced breast cancer
.
Breast Cancer Res Treat
2016
;
158
:
485
95
.
23.
Salgado
R
,
Denkert
C
,
Demaria
S
,
Sirtaine
N
,
Klauschen
F
,
Pruneri
G
, et al
The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TILs Working Group 2014
.
Ann Oncol
2015
;
26
:
259
71
.
24.
Rimm
DL
,
Han
G
,
Taube
JM
,
Yi
ES
,
Bridge
JA
,
Flieder
DB
, et al
A prospective, multi-institutional, pathologist-based assessment of 4 immunohistochemistry assays for PD-L1 expression in non-small cell lung cancer
.
JAMA Oncol
2017
;
3
:
1051
8
.
25.
Nanda
R
,
Chow
LQ
,
Dees
EC
,
Berger
R
,
Gupta
S
,
Geva
R
, et al
Pembrolizumab in patients with advanced triple-negative breast cancer: phase Ib KEYNOTE-012 Study
.
J Clin Oncol
2016
;
34
:
2460
7
.
26.
McLaughlin
J
,
Han
G
,
Schalper
KA
,
Carvajal-Hausdorf
D
,
Pelekanou
V
,
Rehman
J
, et al
Quantitative assessment of the heterogeneity of PD-L1 expression in non-small-cell lung cancer
.
JAMA Oncol
2016
;
2
:
46
54
.
27.
Kluger
HM
,
Zito
CR
,
Barr
ML
,
Baine
MK
,
Chiang
VL
,
Sznol
M
, et al
Characterization of PD-L1 expression and associated T-cell infiltrates in metastatic melanoma samples from variable anatomic sites
.
Clin Cancer Res
2015
;
21
:
3052
60
.
28.
Cimino-Mathews
A
,
Thompson
E
,
Taube
JM
,
Ye
X
,
Lu
Y
,
Meeker
A
, et al
PD-L1 (B7-H1) expression and the immune tumor microenvironment in primary and metastatic breast carcinomas
.
Hum Pathol
2016
;
47
:
52
63
.
29.
Nanda
R
,
Liu
MC
,
Yau
C
,
Asare
S
,
Hylton
N
,
Van't Veer
L
, et al
Pembrolizumab plus standard neoadjuvant therapy for high-risk breast cancer (BC): results from I-SPY 2
.
J Clin Oncol
35:15s
, 
2017
(
suppl; abstr 506
).
30.
Pusztai
L
,
Silber
A
,
Wysong Hofstatter
E
,
Chung
GG
,
Horowitz
NR
,
Lannin
DR
, et al
Safety of MEDI4736 (anti-PD-L1 antibody) administered concomitant with weekly nab-paclitaxel and dose dense doxorubicin/cyclophosphamide (ddAC) as neoadjuvant chemotherapy for stage I-III triple negative breast cancer (TNBC): a phase I/II trial
.
J Clin Oncol
35:15s
, 
2017
(
suppl; abstr 572
).
31.
Schmid
P
,
Park
YH
,
Muñoz-Couselo
E
,
Kim
S-B
,
Sohn
J
,
Im
S-A
, et al
Pembrolizumab (pembro) + chemotherapy (chemo) as neoadjuvant treatment for triple negative breast cancer (TNBC): preliminary results from KEYNOTE-173
.
J Clin Oncol
35:15s
, 
2017
(
suppl; abstr 556
).
32.
Demaria
S
,
Volm
MD
,
Shapiro
RL
,
Yee
HT
,
Oratz
R
,
Formenti
SC
, et al
Development of tumor-infiltrating lymphocytes in breast cancer after neoadjuvant paclitaxel chemotherapy
.
Clin Cancer Res
2001
;
7
:
3025
30
.
33.
Zhang
P
,
Su
DM
,
Liang
M
,
Fu
J
. 
Chemopreventive agents induce programmed death-1-ligand 1 (PD-L1) surface expression in breast cancer cells and promote PD-L1-mediated T cell apoptosis
.
Mol Immunol
2008
;
45
:
1470
6
.
34.
Garcia-Martinez
E
,
Gil
GL
,
Benito
AC
,
Gonzalez-Billalabeitia
E
,
Conesa
MA
,
Garcia Garcia
T
, et al
Tumor-infiltrating immune cell profiles and their change after neoadjuvant chemotherapy predict response and prognosis of breast cancer
.
Breast Cancer Res
2014
;
16
:
488
.
35.
Gonzalez-Angulo
AM
,
Iwamoto
T
,
Liu
S
,
Chen
H
,
Do
KA
,
Hortobagyi
GN
, et al
Gene expression, molecular class changes, and pathway analysis after neoadjuvant systemic therapy for breast cancer
.
Clin Cancer Res
2012
;
18
:
1109
19
.
36.
Dieci
MV
,
Criscitiello
C
,
Goubar
A
,
Viale
G
,
Conte
P
,
Guarneri
V
, et al
Prognostic value of tumor-infiltrating lymphocytes on residual disease after primary chemotherapy for triple-negative breast cancer: a retrospective multicenter study
.
Ann Oncol
2014
;
25
:
611
8
.
37.
Ladoire
S
,
Arnould
L
,
Apetoh
L
,
Coudert
B
,
Martin
F
,
Chauffert
B
, et al
Pathologic complete response to neoadjuvant chemotherapy of breast carcinoma is associated with the disappearance of tumor-infiltrating foxp3+ regulatory T cells
.
Clin Cancer Res
2008
;
14
:
2413
20
.
38.
Mani
NL
,
Schalper
KA
,
Hatzis
C
,
Saglam
O
,
Tavassoli
F
,
Butler
M
, et al
Quantitative assessment of the spatial heterogeneity of tumor-infiltrating lymphocytes in breast cancer
.
Breast Cancer Res
2016
;
18
:
78
.