Tumor-infiltrating lymphocytes in patients with HER2-positive breast cancer treated with neoadjuvant chemotherapy plus trastuzumab, lapatinib or their combination: A meta-analysis of randomized controlled trials

Accepted Manuscript Systematic or Meta-analysis Studies Tumor-infiltrating lymphocytes in patients with HER2-positive breast cancer treated with neoadjuvant chemotherapy plus trastuzumab, lapatinib or their combination: A meta-analysis of randomized controlled trials C. Solinas, M. Ceppi, M. Lambertini, M. Scartozzi, L. Buisseret, S. Garaud, D. Fumagalli, E. de Azambuja, R. Salgado, C. Sotiriou, K. Willard-Gallo, M. Ignatiadis PII: DOI: Reference:

S0305-7372(17)30063-4 http://dx.doi.org/10.1016/j.ctrv.2017.04.005 YCTRV 1629

To appear in:

Cancer Treatment Reviews Cancer Treatment Reviews

Received Date: Revised Date: Accepted Date:

12 March 2017 21 April 2017 23 April 2017

Please cite this article as: Solinas, C., Ceppi, M., Lambertini, M., Scartozzi, M., Buisseret, L., Garaud, S., Fumagalli, D., de Azambuja, E., Salgado, R., Sotiriou, C., Willard-Gallo, K., Ignatiadis, M., Tumor-infiltrating lymphocytes in patients with HER2-positive breast cancer treated with neoadjuvant chemotherapy plus trastuzumab, lapatinib or their combination: A meta-analysis of randomized controlled trials, Cancer Treatment Reviews Cancer Treatment Reviews (2017), doi: http://dx.doi.org/10.1016/j.ctrv.2017.04.005

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

SYSTEMATIC REVIEW AND META-ANALYSIS

Tumor-infiltrating lymphocytes in patients with HER2-positive breast cancer treated with neoadjuvant chemotherapy plus trastuzumab, lapatinib or their combination: A meta-analysis of randomized controlled trials

C. Solinas1, M. Ceppi2, M. Lambertini3, M. Scartozzi4, L. Buisseret1,3,5, S. Garaud1, D. Fumagalli6, E. de Azambuja5, R. Salgado3,7, C. Sotiriou3,5, K. Willard-Gallo1* and M. Ignatiadis5*

Authors affiliations

1

Molecular Immunology Unit, Institut Jules Bordet, Université Libre de Bruxelles (U.L.B.), Brussels, Belgium

2

Unit of Clinical Epidemiology, IRCCS AOU San Martino-IST, Genova, Italy

3

Breast Cancer Translational Research Laboratory, Institut Jules Bordet, Université Libre de Bruxelles (U.L.B.), Brussels, Belgium

4

Medical Oncology, University of Cagliari, Cagliari, Italy

5

Department of Medical Oncology, Institut Jules Bordet, Université Libre de Bruxelles (U.L.B.), Brussels, Belgium

6

Breast International Group (BIG), Brussels, Belgium 1

7

Department of Pathology, GZA Ziekenhuizen, Sint-Augustinus campus, Wilrijk, Belgium

*contributed equally (co-last authors)

Authors emails:

Cinzia Solinas: [email protected] Marcello Ceppi: [email protected] Matteo Lambertini: [email protected] Mario Scartozzi: [email protected] Laurence Buisseret: [email protected] Soizic Garaud: [email protected] Debora Fumagalli: [email protected] Evandro de Azambuja: [email protected] Roberto Salgado: [email protected] Christos Sotiriou: [email protected] Karen Willard-Gallo: [email protected] Michail Ignatiadis: [email protected]

2

Corresponding author: Dr. Matteo Lambertini Breast Cancer Translational Research Laboratory Institut Jules Bordet, Université Libre de Bruxelles (U.L.B.) Boulevard de Waterloo, 127 (3rd floor) (1000) Brussels, Belgium +32.2.541.3768 [email protected]

3

Abstract

Background A relationship between high baseline tumor-infiltrating lymphocytes (TIL) and better outcomes has been described in early-stage HER2-positive breast cancer. Nevertheless, the magnitude of this association and whether this effect could differ based on the type of anti-HER2 agent administered remain controversial. This meta-analysis investigated the association between baseline TIL and pathologic complete response (pCR) rates in HER2-positive breast cancer patients treated with neoadjuvant chemotherapy plus trastuzumab and lapatinib either alone or in combination.

Methods A literature search covering PubMed, Embase and the Cochrane library up to October 31, 2016 identified randomized, controlled trials investigating neoadjuvant chemotherapy plus trastuzumab and lapatinib either alone or in combination where published data for pCR based on pre-treatment TIL scores was available. Two subgroups were considered: high baseline TIL vs. non-high TIL, according to each study definition. Summary risk estimates (odds ratio) and 95% confidence intervals (CI) were calculated for pCR using pretreatment TIL levels for each trial. Pooled analyses were conducted using random and fixed effects models. Interaction P-values were computed using a Monte Carlo permutation test. 4

Results A total of 5 studies (N=1,256 patients) were included. Overall, high TIL subgroup was associated with a significantly increased pCR rate (OR 2.46; 95% CI 1.36-4.43; P=0.003). No interaction was observed between TIL subgroup (high vs. non-high TIL) and response to anti-HER2 agent(s) (trastuzumab vs. lapatinib vs. their combination; P=0.747) and chemotherapy (anthracycline and taxanes vs. taxanes only; P=0.201). A stronger association between high TIL subgroup and pCR rate was observed when examining only the 4 studies using neoadjuvant anthracycline- and taxane- based chemotherapy and the 60% cut-off for high TIL (N=869, NeoALTTO excluded) with an OR of 2.88 (95% CI 2.03-4.08; P<0.001).

Conclusions In HER2-positive breast cancer, high baseline TIL are associated with increased pCR probability irrespective of anti-HER2 agent(s) and neoadjuvant chemotherapy regimen used.

Keywords

5

Tumor-infiltrating lymphocytes; neoadjuvant treatment; HER2-postive breast cancer; pathologic complete response; trastuzumab; lapatinib.

6

Key message

This meta-analysis estimated the association between tumor-infiltrating lymphocytes (TIL) and pathologic complete response (pCR) rates after neoadjuvant chemotherapy in combination with anti-HER2 therapy (trastuzumab, lapatinib or their combination). In HER2positive breast cancer, high baseline TIL are associated with increased pCR probability irrespective of anti-HER2 agent(s) and neoadjuvant chemotherapy regimen used.

7

MANUSCRIPT

Introduction

Treatment of breast cancer patients by neoadjuvant systemic therapy has traditionally been used to treat locally advanced inoperable and inflammatory cancers or preceding breast conserving surgery [1]. A recent pooled analysis from the Collaborative Trials in Neoadjuvant Breast Cancer (CTNeoBC) international working group demonstrated that pathologic complete responses (pCR) are correlated with better long term outcomes (event free survival and overall survival [OS]). This was particularly true for the more aggressive breast cancer subtypes (i.e. triple negative and HER2-positive), which currently derive the most benefit from neoadjuvant treatment [1,2]. These findings led the U.S. Food and Drug Administration (FDA) to propose that pCR is potentially a valid regulatory endpoint for drug approval [3], increasing current use of neoadjuvant therapy. Use of this approach has raised not only for the increase in the rate of breast conservative surgery, but also because it provides an opportunity to examine drug responses in vivo over a shorter period [4], which also lends itself to the analysis of new predictive biomarkers. HER2-positive breast cancer subtype has demonstrated sensitivity to current cancer therapies (chemotherapy and anti-HER2 targeted agents), making it an ideal candidate for neoadjuvant treatment. Currently, the only validated, predictive biomarker for antiHER2 agents is HER2 over-expression or amplification [5]. 8

Tumor-infiltrating lymphocytes (TIL) have emerged as potentially important prognostic and/or predictive biomarkers for breast cancer. Recent studies identified significant immune infiltrates in 10-21% of HER2-positive tumors [7–13]. Denkert et al. were the first to show a relationship between the presence of high infiltration in baseline pre-treatment biopsies and increased pCR rates in breast cancer patients given anthracycline- and taxane-based neoadjuvant chemotherapy in two randomized, controlled trials [13]. This study initiated TIL assessment on hematoxylin eosin (H&E) stained tissue sections [13], which has now been generalized into guidelines for scoring TIL in breast cancer [14]. Of note, it has been suggested that HER2-positive breast cancer patients with increased baseline TIL may benefit more from the anti-HER2 monoclonal antibody trastuzumab [9,15]. This benefit might be linked to the immune mediated effects of trastuzumab (including Antibody-Dependent Cell-mediated Cytotoxicity (ADCC)) [16] that were first observed in pre-clinical studies [17,18]. Interestingly, an increase in these immune effects might be achieved with dual anti-HER2 blockade with lapatinib, for the accumulation of HER2 receptors at the cell surface [19,20]. In a recent meta-analysis evaluating TIL in breast cancer, HER2-positive patients who received anti-HER2 treatment(s) were mixed with those who did not; therefore, the potential association between baseline TIL and pCR rates for patients receiving neoadjuvant chemotherapy combined with anti-HER2 agent(s) could not be evaluated [21]. The present meta-analysis of published randomized controlled trials addressed the association between TIL scored on pre-treatment core biopsies and pCR rates in HER2positive breast cancer patients receiving neoadjuvant chemotherapy plus trastuzumab and lapatinib.

9

Methods

A systematic review of the literature and meta-analysis was performed to identify randomized controlled trials that investigated the role of TIL in HER2-positive breast cancer patients treated with neoadjuvant chemotherapy plus trastuzumab and lapatinib either alone or in combination. A prerequisite was that TIL in pre-treatment biopsies had been scored on H&E slides using the guidelines published by the International Breast Cancer Immuno-Oncological Biomarkers Working Group (previously named as the TIL working group) [14]. The present study was conducted using a pre-planned protocol and the PRISMA guidelines [22].

Search strategy, eligibility criteria and data collection

A literature search through October 31, 2016 was performed using Medline, Embase and the Cochrane library. The search strategy was built by inputting keywords related to “breast cancer”, “HER2-positive”, “immunotherapy” and “TIL”. Boolean operators were used to combine specific keywords for each database and free text terms. The titles and abstracts of the identified studies were independently evaluated by two reviewers (C.S. and M.L.) with a consensus for discordant cases reached through discussion. Crossreferencing from relevant studies and review articles on the topic was conducted to confirm that all possible pertinent studies had been 10

retrieved. Eligible studies had to meet the following criteria: 1) English-language published study; 2) randomized controlled trial; 3) inclusion of patients with non-metastatic HER2-positive breast cancer treated with trastuzumab and lapatinib either alone or in combination prior to surgery; 4) TIL evaluation on H&E slides according to guidelines [14]; and 5) the odds ratio (OR) for pCR for each TIL subgroup (high TIL vs. non-high TIL) had to be reported or could be calculated from the published data. Non-randomized or retrospective studies and publications evaluating immune gene signatures alone were excluded from our analysis. For each eligible study, two investigators (C.S. and M.L.) independently collected the following information: first author, year of publication, study design, years of recruitment, neoadjuvant chemotherapy regimens administered, pCR and high TIL definitions by trials, sample size and source of study subjects, number of HER2-positive patients included, pCR rates in HER2-positive patients from the original trials, hormone receptor (HR) status in patients from the TIL study, and number of pCR in high TIL and non-high TIL subgroups according to each study definition.

Study Objectives

Our principal objective was to evaluate the association between TIL and pCR rates in HER2-positive breast cancer patients treated with neoadjuvant trastuzumab and lapatinib (either alone or in combination) together with chemotherapy. Baseline TIL levels 11

were considered as a categorical variable with tumors segregated into high TIL vs. non-high TIL using the threshold defined for each trial. Our secondary objective was to explore the association between TIL and pCR rates according to type of administered antiHER2 therapy and chemotherapy. Three groups based on the anti-HER2 agent used were considered: trastuzumab alone, lapatinib alone and trastuzumab plus lapatinib. Two chemotherapy groups were identified: anthracycline- and taxane- based chemotherapy vs. taxanes alone.

Statistical analysis

The association between TIL and pCR rates was evaluated by determining the probability of achieving pCR between high TIL vs. non-high TIL subgroups in all eligible studies, irrespective of treatment regimen. The association between TIL and pCR rates according to type of administered anti-HER2 therapy and chemotherapy was evaluated using the interaction test, undertaken to investigate whether the effect of one exploratory variable depended on the level of one or more variables [23]. The probability of achieving a pCR between the TIL subgroups was measured in patients treated with chemotherapy plus trastuzumab and lapatinib either alone or in combination and in patients treated with anthracycline- and taxane-based chemotherapy or taxanes alone.

12

Patients achieving a pCR in each TIL subgroup of the overall trial population were identified and then categorized by treatment regimen. OR and 95% confidence intervals (CI) with two-sided P-values were determined. The OR for pCR was calculated as the odds of pCR patients in the high TIL subgroup divided by pCR patients in the non-high TIL subgroup. An OR 1 reflected a higher pCR probability in the high TIL subgroup. A fixed effect model using the inverse variance method was fitted to the data by making the assumption that a common effect size was shared by all the studies analyzed, with differences between studies occurring only by chance. The occurrence of inconsistency between studies was investigated through the I2 statistics, used to quantify the percentage of total variation attributable to heterogeneity across studies [24]. When significant heterogeneity was detected, pooled OR was computed using the Der Simonian and Laird random effect model. This model provides pooled OR estimates that take into account the potential correlation and the heterogeneity between studies [25]. Sensitivity analyses were conducted by recalculating the pooled OR estimates by excluding one study at a time (leave-one-out procedure) [26]. Interaction P-values were estimated using a Monte Carlo permutation test [23]. Funnel plots analysis was run to detect bias in the meta-analysis, by estimating effects from individual studies with respect to their standard errors (as a measure of each study’s size or precision) [27]. A P-value <0.05 was considered statistically significant. Statistical analysis, funnel plot and forest plots were performed using STATA statistical software (StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP).

13

Results

Our literature search identified 531 entries, which included 70 of potential interest; however, 65 out of 70 entries were excluded because they did not meet the inclusion criteria (Fig. 1). The five eligible studies included in this meta-analysis [6–9] enrolled a total of 1,256 HER2-positive breast cancer patients who received chemotherapy plus trastuzumab and lapatinib either alone or in combination in the neoadjuvant setting (Table 1). The neoadjuvant chemotherapy backbone was anthracyclines and taxanes in four trials (N=869; CherLOB, GeparQuattro, GeparQuinto and GeparSixto) while taxanes alone were administered in NeoALTTO (N=387) (in this trial, anthracyclines were given after surgery). The cut-offs used for high TIL were 60% in the first four trials and 30% in NeoALTTO.

1. Association between TIL and pCR rates in the global study population.

Overall for the five trials (N=1,256), the high TIL subgroup was associated with a significant increase in pCR rate (OR 2.46; 95% CI 1.36-4.43; P=0.003) (Fig. 2). However, significant heterogeneity was observed (I 2=74.8%; P=0.003), showing a substantial inconsistency across studies. Sensitivity analysis revealed this effect was mainly driven by including data from NeoALTTO, which

14

uniquely had taxane only neoadjuvant chemotherapy and a lower cut-off for high TIL. Funnel plot analysis showed a low probability of publication bias (P=0.186) (Fig. 3). No interaction was observed between TIL subgroups (high TIL vs. non-high TIL) and different anti-HER2 agent(s) administered (P=0.747) nor between TIL subgroups and type of chemotherapy (P=0.201).

2. Association between TIL and pCR rates based on the neoadjuvant anti-HER2 therapy administered.

Trastuzumab

Four out of five selected studies evaluated the association between TIL and pCR in a trial’s arm receiving trastuzumab only with chemotherapy [7–9]. The patients in CherLOB and NeoALTTO received neoadjuvant weekly trastuzumab (26 and 18 weeks, respectively), while in GeparQuattro, GeparQuinto and GeparSixto, neoadjuvant trastuzumab was administered every three weeks (8 cycles in GeparQuattro’s anthracycline/taxane and anthracycline/taxane/concomitant capecitabine arms; 12 cycles in GeparQuattro’s anthracycline/taxane/sequential capecitabine arm; 8 cycles in GeparQuinto and 6 cycles in GeparSixto). The chemotherapy backbone was anthracyclines and taxanes in CherLOB, GeparQuattro and GeparQuinto (N=372) and taxanes alone in NeoALTTO (N=131) (Table 1). 15

In the whole group, OR for pCR in the high TIL subgroup was 1.90 (95% CI 1.21-3.00; P=0.005) with no significant heterogeneity observed (I2=0.0%; P=0.814) (Figure 4A). The sensitivity analyses estimated OR equaled to 1.90, 2.06, 1.58 and 2.03 after the sequential exclusion of CherLOB, NeoALTTO, GeparQuattro or GeparQuinto, respectively. All results were statistically significant except those obtained after removing GeparQuattro, where the same trend was observed but did not reach any statistical significance (OR 1.58 95% CI 0.86-2.89; P=0.139).

Lapatinib

Three studies evaluated the association between TIL and pCR in patients treated with chemotherapy plus lapatinib as a single anti-HER2 agent [7–9]. Neoadjuvant lapatinib was administered daily at similar doses in all the trials (1500 mg in CherLOB and NeoALTTO, reduced to 750 mg during concurrent paclitaxel in a subgroup of patients after a protocol amendment; 1250 mg in GeparQuinto). The neoadjuvant chemotherapy backbone was represented by anthracyclines and taxanes in CherLOB and GeparQuinto (N=192) and taxanes alone in NeoALTTO (N=130) (Table 1). In the whole group, OR for pCR was 1.79 (95% CI 0.99-3.23; P=0.052). No significant heterogeneity was observed (I2=57%; P=0.098) (Figure 4A). At the sensitivity analyses, the estimated OR equaled to 1.50 and 1.67 after removing CherLOB and GeparQuinto, respectively. With the exclusion of NeoALTTO, a significantly increased pCR rate was observed (OR 2.44; 95% CI 1.1716

5.11, P=0.018).

Trastuzumab plus lapatinib dual blockade

Three studies evaluated the association between TIL and pCR in patients treated with chemotherapy and dual blockade by trastuzumab plus lapatinib [6–8]. In CherLOB and NeoALTTO, neoadjuvant trastuzumab was given weekly (26 and 18 weeks, respectively) with concomitant lapatinib at a dose of 1000 mg/daily (reduced to 750 mg during concurrent paclitaxel in a subgroup of patients after a protocol amendment); in GeparSixto, neoadjuvant trastuzumab was administered every three weeks (6 cycles) and lapatinib at 750 mg/daily (18 weeks). The neoadjuvant chemotherapy backbone was represented by anthracyclines and taxanes in CherLOB and GeparSixto (N=305) and taxanes alone in NeoALTTO (N=126) (Table 1). In the overall group, OR for pCR in the high TIL subgroup was higher than in the single agent arms, but these results were no t significant (OR 3.06; 95% CI 0.84-11.20; P=0.147) (Figure 4A). Significant heterogeneity was observed (I2=79.7%; P=0.007). The sensitivity analyses found the estimated OR was 2.31 and 2.49 after CherLOB and GeparSixto removal, respectively, and did not reach any statistical significance. In contrast, the estimated OR significantly increased after removing NeoALTTO (OR 5.16 95% CI 2.79-9.54, P<0.001).

17

3. Association of TIL with pCR rates based on the neoadjuvant chemotherapy regimen.

Anthracyclines plus taxanes

Four studies evaluated the association between TIL and pCR in patients treated with anthracycline plus taxane neoadjuvant chemotherapy together with anti-HER2 agent(s) [6,8,9]. The neoadjuvant anti-HER2 treatment backbone was represented by trastuzumab in one arm of CherLOB, GeparQuattro and GeparQuinto (N=372), lapatinib in one arm of CherLOB and GeparQuinto (N=192) and the combination in one arm of CherLOB and the entire HER2-positive cohort in GeparSixto (N=305). Globally, OR for pCR in the high TIL subgroup was 2.84 (95% CI 2.01-4.01; P<0.001) (Figure 4B). No significant heterogeneity was detected (I2=59.6%; P=0.060). Sensitivity analyses found pooled OR ranged from 2.27 to 3.79 after GeparSixto and GeparQuinto were removed, respectively (Figure 4B).

Taxanes

18

The NeoALTTO trial alone evaluated the association between TIL and pCR in patients treated with taxane-based neoadjuvant chemotherapy together with anti-HER2 agent(s). The neoadjuvant anti-HER2 treatment backbone was trastuzumab for 131, lapatinib for 130 and dual blockade for 126 patients analyzed for TIL. The OR for pCR in the high TIL subgroup was 1.15 (95% CI 0.70-1.90; P=0.586) (Figure 4B).

4. Association between TIL and pCR rates in the global study population excluding NeoALTTO

Considering that NeoALTTO emerged as the main driver of heterogeneity, it was excluded in the context of an exploratory analysis in which the association between TIL and pCR rates was evaluated in the various anti-HER2 arms of the four trials using neoadjuvant anthracycline plus taxane chemotherapy and a >60% cut-off for high TIL (N=869) [6,8,9]. Overall, high TIL subgroup was associated with a significant increase in pCR rate (OR 2.88 95% CI 2.03-4.08; P<0.001) (Figure 5) and no significant heterogeneity was detected (I2=33%; P=0.176). A significant association was observed between high TIL and pCR probability in the trastuzumab (OR 2.06 95% CI 1.23-3.46; P=0.006; I2=0%, P=0.766), lapatinib (OR 2.44 95% CI 1.17-5.11; P=0.018; I2=64.5%, P=0.093) and combination arms (OR 5.16 95% CI 2.79-9.54; P<0.001; I2=0%, P=0.552) (Figure 5). This result was consistent for all patients, irrespective of the anti-HER2 treatment they received (interaction P=0.077). 19

Discussion

The principal goal of our meta-analysis was to investigate the role of baseline TIL in HER2-positive breast cancer patients treated with neoadjuvant chemotherapy together with the anti-HER2 agents trastuzumab and lapatinib used either alone or in combination. A strong association between high TIL and increased pCR rates was observed (OR 2.46; 95% CI 1.36-4.43; P=0.003). Heterogeneity between studies was significant (I 2=74.8%; P=0.003). This was principally driven by the inclusion of NeoALTTO where taxanes alone (without anthracyclines) were administered as the neoadjuvant chemotherapy backbone and because the available cutoff for high TIL was set at 30% (and not 60%). No association between TIL and pCR rates according to type of administered antiHER2 therapy and chemotherapy was detected. A recent meta-analysis of published randomized controlled trials estimated that high TIL counts correlated to a higher chance of pCR in triple negative and HER2-positive breast cancer subtypes. Specifically, a pCR rate of 56.1% was achieved in >60% TIL HER2positive tumors compared to 22.8% in the <60% TIL subgroup [21]. Nevertheless, in the HER2-positive subtype group, the aforementioned meta-analysis included also a study (GeparTrio) where no anti-HER2 agent was administered and did not include NeoALTTO [7] and the HER2-positive cohorts of GeparQuattro and GeparQuinto studies [9]. Compared to this meta-analysis, our work evaluated a more homogeneous population (all patients were treated with chemotherapy plus trastuzumab and lapatinib given 20

either alone or in combination) and larger number of patients (1,256 patients vs. 625) [21]. Our findings confirmed that the presence of high TIL in pre-treatment biopsies of primary tumors from HER2-positive breast cancer is strongly associated with increased pCR probability in patients treated with neoadjuvant chemotherapy plus trastuzumab, lapatinib either alone or in combination. We further investigated whether TIL assessment could help in guiding the choice among different systemic treatment strategies (chemotherapy and anti-HER2 agents). However, no association between TIL and pCR rates according to type of anti-HER2 therapy and chemotherapy was observed. Nevertheless, the choice among trastuzumab as single anti-HER2 blockade or its combination with lapatinb in the early setting is of more limited value nowadays. Despite the fact that the combination of trastuzumab and lapatinib significantly increased pCR rates in the neoadjuvant setting [28,29], the ALTTO trial did not show any significant improvement in disease-free survival and OS with the adjuvant anti-HER2 dual blockade trastuzumab plus lapatinib [30]. Thus, chemotherapy plus adjuvant trastuzumab as single anti-HER2 blockade remains the current standard of care in early-stage HER2-positive breast cancer. The investigation of the potential predictive value of TIL would be more relevant when evaluating the combination of trastuzumab and pertuzumab that is currently standard of care in the metastatic setting [31]. Furthermore, pertuzumab has been approved by the U.S. FDA for the neoadjuvant treatment of HER2-positive early-stage breast cancer, based on the significantly increased pCR rates observed as compared to trastuzumab alone [32]. Results of the ongoing APHINITY trial (NCT01358877) investigating the value of adding pertuzumab to trastuzumab in the adjuvant setting are still awaited. The identification of a biomarker to better select patients who would benefit the most from dual (trastuzumab plus pertuzumab) vs. mono (trastuzumab alone) anti-HER2 blockade should be 21

considered a research priority also from a public health perspective. Recent data from the CLEOPATRA trial investigating the combination of docetaxel and trastuzumab with or without pertuzumab as first-line therapy for patients with HER2-positive metastatic breast cancer revealed that 10% stromal increment of TIL was associated with better OS, but the treatment effect of pertuzumab did not differ by TIL levels [33]. In the NeoSphere trial [32], a four arms study where neoadjuvant docetaxel was given in association with trastuzumab, pertuzumab or their combination, or both anti-HER2 monoclonal antibodies were given without any chemotherapy, TIL as continuous variable did not correlate with pCR in the breast [10]. In the TRYPHAENA study [34] investigating neoadjuvant trastuzumab and pertuzumab as dual anti-HER2 blockade in association with different chemotherapy regimens (anthracycline- and taxane-based or carboplatin- and taxane-based chemotherapy), increased baseline TIL were associated with higher pCR probability after adjusting for clinico-pathological characteristics [11]. Hence, to date, data on the role of TIL as biomarkers of response to dual anti-HER2 blockade with trastuzumab and pertuzumab remain limited and not conclusive. Further research efforts are needed on this regard. Besides their clinical utility as predictors of benefit from neoadjuvant chemotherapy, TIL have the potential to become in the near future a biomarker for patients’ stratification and selection for immunotherapy. TIL might be useful to identify ideal candidates to receive immunotherapy as a boost for their pre-existing immune response, for example in patients with a residual disease after neoadjuvant treatment. Patients with low baseline TIL might benefit from treatments aiming at inducing or boosting their anti-tumor immune response [35]. 22

Some limitations should be taken into account in the interpretation of our findings. Since no individual patients’ data could be obtained, it was not possible to investigate the impact of other important factors (e.g. hormone receptor status) on the association between TIL and pCR. For the same reason, we used TIL as a categorical variable without analyzing them as a continuous variable as suggested by international guidelines [14]. Association between TIL and the residual cancer burden (RCB)-index representing a continuous parameter of response after neoadjuvant chemotherapy [36] instead of using pCR which is a binary variable (pCR vs. nonpCR) might provide in the future more valuable information on the clinical utility of TIL in the neoadjuvant setting of HER2-positive breast cancer. Despite these limitations, our meta-analysis has some important strengths: the considerable number of patients included (N=1,256) from randomized, controlled trials; the standard definition of pCR used across the studies (i.e. response in breast and axilla); the chemotherapy backbone and definition of high TIL subgroup which were homogeneous in most of the studies (4/5) [37]. Furthermore, TIL assessment in all the studies has been done following current recommendations [14], and even though not all the slides were read in the same centers and by the same pathologists, some of these pathologists successfully participated to the ring studies, which demonstrated the feasibility of a reproducible evaluation of TIL in breast cancer [38].

Conclusion

23

In patients with HER2-positive breast cancer, our meta-analysis confirmed the association between high TIL at baseline and increased pCR rates after neoadjuvant treatments with chemotherapy and anti-HER2 agent(s). This association was not different according to anti-HER2 therapy (trastuzumab, lapatinib either alone or in combination) or chemotherapy regimen (anthracyclines and taxanes vs. taxanes alone). Further investigations are needed to better evaluate the potential role of TIL in pre-treatment biopsies and residual disease to tailor treatment decisions (including both dual anti-HER2 blockade and immunotherapeutic approaches) in earlystage HER2-positive breast cancer.

Acknowledgements

Dr. Cinzia Solinas is a fellow of the Belgian Fund for Scientific Research (FNRS)-Operation Télévie. Dr. Matteo Lambertini acknowledges the support from the European Society for Medical Oncology (ESMO) for a Translational Research Fellowship at the Institut Jules Bordet in Brussels (Belgium).

Funding

This work was partially supported by a grant from the Belgian FNRS-Opération Télévie, and Les Amis de l’Institut Bordet. The study 24

funders had no role in the design of the study, the collection, analysis, or interpretation of the data, the writing of the manuscript, nor the decision to submit the manuscript for publication.

Disclosure

Prof. Mario Scartozzi received honoraria and travel grants from Roche, Celgene, Merck, Amgen and MSD outside the submitted work. Dr. de Azambuja received honoraria from Roche and travel grants from Roche and GlaxoSmithKline outside the submitted work. His institution has received research grants form Roche. Dr. Roberto Salgado receives travel support from Roche. Prof. Michail Ignatiadis was consultant for Roche. All the other authors declare no conflict of interest.

Conflict of interests Prof. Mario Scartozzi received honoraria and travel grants from Roche, Celgene, Merck, Amgen and MSD outside the submitted work. Dr. de Azambuja received honoraria from Roche and travel grants from Roche and GlaxoSmithKline outside the submitted work. His institution has received research grants form Roche. Dr. Roberto Salgado receives travel support from Roche. Prof. Michail Ignati adis was consultant for Roche. All the other authors declare no conflict of interest.

25

References

[1]

Kaufmann M, Minckwitz G von, Mamounas EP, Cameron D, Carey LA, Cristofanilli M, et al. Recommendations from an

International Consensus Conference on the Current Status and Future of Neoadjuvant Systemic Therapy in Primary Breast Cancer. Ann Surg Oncol 2012;19:1508–16. [2]

Cortazar P, Zhang L, Untch M, Mehta K, Costantino JP, Wolmark N, et al. Pathological complete response and long-term

clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet 2014;384:164–72. [3]

Prowell TM, Pazdur R. Pathological Complete Response and Accelerated Drug Approval in Early Breast Cancer. N Engl J Med

2012;366:2438–41. [4]

Graham PJ, Brar MS, Foster T, McCall M, Bouchard-Fortier A, Temple W, et al. Neoadjuvant Chemotherapy for Breast Cancer,

Is Practice Changing? A Population-Based Review of Current Surgical Trends. Ann Surg Oncol 2015;22:3376–82. [5]

Wolff AC, Hammond MEH, Hicks DG, Dowsett M, McShane LM, Allison KH, et al. Recommendations for Human Epidermal

Growth Factor Receptor 2 Testing in Breast Cancer: American Society of Clinical Oncology/College of American Pathologists Clinical Practice Guideline Update. J Clin Oncol 2013;31:3997–4013.

26

[6]

Denkert C, Minckwitz G von, 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. [7]

Salgado R, Denkert C, Campbell C, Savas P, Nuciforo P, Aura C, et al. Tumor-Infiltrating Lymphocytes and Associations With

Pathological Complete Response and Event-Free Survival in HER2-Positive Early-Stage Breast Cancer Treated With Lapatinib and Trastuzumab: A Secondary Analysis of the NeoALTTO Trial. JAMA Oncol 2015;1:448. [8]

Dieci MV, Prat A, Tagliafico E, Paré L, Ficarra G, Bisagni G, et al. Integrated evaluation of PAM50 subtypes and immune

modulation of pCR in HER2-positive breast cancer patients treated with chemotherapy and HER2-targeted agents in the CherLOB trial. Ann Oncol 2016;27:1867–73. [9]

Heppner BI, Untch M, Denkert C, Pfitzner BM, Lederer B, Schmitt WD, et al. Tumor-infiltrating lymphocytes: a predictive and

prognostic biomarker in neoadjuvant treated HER2-positive breast cancer. Clin Cancer Res 2016;22:5747-54. [10]

Bianchini G, Pusztai L, Pienkowski T, Im Y-H, Bianchi GV, Tseng L-M, et al. Immune modulation of pathologic complete

response after neoadjuvant HER2-directed therapies in the NeoSphere trial. Ann Oncol 2015;26:2429–36. [11]

Ignatiadis M, Van den Eynden GG, Salgado R, Fornili M, Desmedt C, Holgado E, et al. Tumor infiltrating lymphocytes before

and after dual HER2 blockade in HER2-amplified early breast cancer: A TRYPHAENA substudy. ASCO Meet Abstr 2016;34:11507.

27

[12]

Perez EA, Ballman KV, Tenner KS, et al. ASsociation of stromal tumor-infiltrating lymphocytes with recurrence-free survival in

the n9831 adjuvant trial in patients with early-stage her2-positive breast cancer. JAMA Oncol 2016;2:56–64. [13]

Denkert C, Loibl S, Noske A, Roller M, Müller 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. [14]

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. [15]

Loi S, Michiels S, Salgado R, Sirtaine N, Jose V, Fumagalli D, et al. Tumor infiltrating lymphocytes are prognostic in triple

negative breast cancer and predictive for trastuzumab benefit in early breast cancer: results from the FinHER trial. Ann Oncol 2014;25:1544–50. [16]

Bianchini G, Gianni L. The immune system and response to HER2-targeted treatment in breast cancer. Lancet Oncol

2014;15:e58–68. [17]

Park S, Jiang Z, Mortenson ED, Deng L, Radkevich-Brown O, Yang X, et al. The therapeutic effect of anti-HER2/neu antibody

depends on both innate and adaptive immunity. Cancer Cell 2010;18:160–70. [18]

Stagg J, Loi S, Divisekera U, Ngiow SF, Duret H, Yagita H, et al. Anti-ErbB-2 mAb therapy requires type I and II interferons and

synergizes with anti-PD-1 or anti-CD137 mAb therapy. Proc Natl Acad Sci U S A 2011;108:7142–7.

28

[19]

Scaltriti M, Verma C, Guzman M, Jimenez J, Parra JL, Pedersen K, et al. Lapatinib, a HER2 tyrosine kinase inhibitor, induces

stabilization and accumulation of HER2 and potentiates trastuzumab-dependent cell cytotoxicity. Oncogene 2008;28:803–14. [20]

Maruyama T, Mimura K, Izawa S, Inoue A, Shiba S, Watanabe M, et al. Lapatinib Enhances Herceptin-mediated Antibody-

dependent Cellular Cytotoxicity by Up-regulation of Cell Surface HER2 Expression. Anticancer Res 2011;31:2999–3005. [21]

Carbognin L, Pilotto S, Nortilli R, Brunelli M, Nottegar A, Sperduti I, et al. Predictive and Prognostic Role of Tumor-Infiltrating

Lymphocytes for Early Breast Cancer According to Disease Subtypes: Sensitivity Analysis of Randomized Trials in Adjuvant and Neoadjuvant Setting. The Oncologist 2016;21:283–91. [22]

Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and

meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015;4:1. [23]

Higgins JPT, Thompson SG. Controlling the risk of spurious findings from meta-regression. Stat Med 2004;23:1663–82.

[24]

Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58.

[25]

DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88.

[26]

Greenland S, O’ Rourke K. Meta-Analysis. In Modern Epidemiology, 3rd edition (edited by Rothman KJ, Greenland S, Lash T):

Lippincott Williams & Wilkins, 2008. [27]

Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34.

29

[28]

Baselga J, Bradbury I, Eidtmann H, Di Cosimo S, de Azambuja E, Aura C, et al. Lapatinib with trastuzumab for HER2-positive

early breast cancer (NeoALTTO): a randomised, open-label, multicentre, phase 3 trial. Lancet 18;379:633–40. [29]

Guarneri V, Frassoldati A, Bottini A, Cagossi K, Bisagni G, Sarti S, et al. Preoperative Chemotherapy Plus Trastuzumab,

Lapatinib, or Both in Human Epidermal Growth Factor Receptor 2–Positive Operable Breast Cancer: Results of the Randomized Phase II CHER-LOB Study. J Clin Oncol 2012;30:1989–95. [30]

Piccart-Gebhart M, Holmes E, Baselga J, Azambuja E de, Dueck AC, Viale G, et al. Adjuvant Lapatinib and Trastuzumab for

Early Human Epidermal Growth Factor Receptor 2–Positive Breast Cancer: Results From the Randomized Phase III Adjuvant Lapatinib and/or Trastuzumab Treatment Optimization Trial. J Clin Oncol 2016;34:1034–42. [31]

Swain SM, Baselga J, Kim S-B, Ro J, Semiglazov V, Campone M, et al. Pertuzumab, trastuzumab, and docetaxel in HER2-

positive metastatic breast cancer. N Engl J Med 2015;372:724–34. [32]

Gianni L, Pienkowski T, Im Y-H, Roman L, Tseng L-M, Liu M-C, et al. Efficacy and safety of neoadjuvant pertuzumab and

trastuzumab in women with locally advanced, inflammatory, or early HER2-positive breast cancer (NeoSphere): a randomised multicentre, open-label, phase 2 trial. Lancet Oncol 2012;13:25–32. [33]

Luen SJ, Salgado R, Fox S, Savas P, Eng-Wong J, Clark E, et al. Tumour-infiltrating lymphocytes in advanced HER2-positive

breast cancer treated with pertuzumab or placebo in addition to trastuzumab and docetaxel: a retrospective analysis of the CLEOPATRA study. Lancet Oncol 2017;18:52–62. 30

[34]

Schneeweiss A, Chia S, Hickish T, Harvey V, Eniu A, Hegg R, et al. Pertuzumab plus trastuzumab in combination with standard

neoadjuvant anthracycline-containing and anthracycline-free chemotherapy regimens in patients with HER2-positive early breast cancer: a randomized phase II cardiac safety study (TRYPHAENA). Ann Oncol 2013;24:2278–84. [35]

Savas P, Salgado R, Denkert C, Sotiriou C, Darcy PK, Smyth MJ, et al. Clinical relevance of host immunity in breast cancer:

from TILs to the clinic. Nat Rev Clin Oncol 2016;13:228–41. [36]

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. [37]

Simon RM, Paik S, Hayes DF. Use of Archived Specimens in Evaluation of Prognostic and Predictive Biomarkers. J Natl

Cancer Inst 2009;101:1446–52. [38]

Denkert C, Wienert S, Poterie A, Loibl S, Budczies J, Badve S, et al. Standardized evaluation of tumor-infiltrating lymphocytes

in breast cancer: results of the ring studies of the international immuno-oncology biomarker working group. Mod Pathol 2016;29:1155– 64.

31

Figure 1. PRISMA flow chart summarizing the process for the identification of eligible studies.

Abbreviation: TIL: tumor-infiltrating lymphocytes.

Figure 2. Association between TIL subgroups (High TIL vs. Non-high TIL) and pCR in the whole study population (any chemotherapy plus any anti-HER2 agent(s)).

Abbreviations: OR: odds ratio; CI: confidence interval; TIL: tumor-infiltrating lymphocytes; *: number (N) of pathologic complete responses (pCR)/total number of patients in the High TIL subgroup; +: N of pCR/total N of patients in the Non-high TIL subgroup; RE: random effect.

Figure 3. Funnel plot with pseudo 95% confidence limits of the High TIL effect, estimated from individual studies (horizontal axis) against the study size (vertical axis).

32

Figure 4. Association between TIL subgroups (High TIL vs. Non-high TIL) and pCR based on: A) type of anti-HER2 therapy administered; B) type of chemotherapy administered.

Abbreviations: OR: odds ratio; CI: confidence interval; TIL: tumor-infiltrating lymphocytes; *: N of pathologic complete responses (pCR)/total number of patients in the High TIL subgroup; +: N of pCR/total N of patients in the Non-high TIL subgroup; FE: fixed effect; RE: random effect.

Figure 5. Association between TIL subgroups (High TIL vs. Non-high TIL) and pCR rates in the global study population excluding NeoALTTO.

Abbreviations: OR: odds ratio; CI: confidence interval; TIL: tumor-infiltrating lymphocytes; *: N of pathologic complete responses (pCR)/total N of patients in the High TIL subgroup; +: N of pCR/total N of patients in the Non-high TIL subgroup; FE: fixed effect.

33

34

35

36

37

38

Table 1. Summary of the main features of the included studies. Study

Trial

Years of

Neoadjuvant

Neoadjuvant

Total

HER2-

pCR rate (%) in

Patients with

High TIL

pCR definition

Phase

enrolment

CT

anti-HER2

patients

positive

patients with

HER2-positive

definition

in TIL study

agents:

in the

patients in

HER2-positive

disease included

N in TIL study

original

the

disease from

in TIL study:

(N in original

trial:

original

the original

N (%)

trial)

N

trial:

study

HR

status in T study

N (%) T

L

T+

HR+

H

%

%

L

N (%)

N(

65

40

(61.9)

(38

%

CherLOB

IIb

08/2006-

12

weekly

NCT00429299

RCT

11/2010

PACLITAXEL

mg/kg [load: 4

stromal

(80

mg/mq)

mg/kg]):

intra-tumoral

[Dieci M.V. et

4

FEC

32 (36)

TIL

al, Ann Onc

(every

3

2015]

weeks,

5-

[8]

FLUOROURA

mg):

CIL:

34 (39)

600

-T

(weekly,

2

121

121 (100)

25

26

47

105 (87)

>60%

either

ypT0is/ypN0

or

-L (daily, 1500

mg/mq; EPIRUBICIN:

-T

75

mg/kg [load: 4

mg/mq;

(weekly,

2

CYCLOPHOS

mg/kg])+L

PHAMIDE:

(daily, 1000 mg):

39

600 mg/mq)

39 (46)

GeparQuattro

III

08/2005-

4 EC (every 3

-T (every

NCT00288002

RCT

11/2006

weeks,

weeks, 6 mg/kg

EPIRUBICIN:

[load: 8 mg/kg]):

90

178 (445)

[Heppner et

al,

Cancer

B.I. Cl Res

mg/mq;

3

1495

200 (13)

32

NA

NA

178 (89)

>60% stromal

ypT0is/ypN0

TIL

CYCLOPHOS PHAMIDE:

2016]

600 mg/mq)

[9]

4 DOCETAXEL (every

3

weeks,

100

mg/mq

if

alone;

75

mg/mq in the arms

with

CAPECITABIN E)

+/-

4

CAPECITABIN E

(1800

mg/mq days 114,

every

weeks)

3

§

§

[ comcomitant or sequential]

40

NR*

NR

GeparQuinto

III

11/2007-

4 EC (every 3

-T

NCT00567554

RCT

07/2011

weeks,

weeks, 6 mg/kg

[Heppner et

B.I.

al,

3

EPIRUBICIN:

[load: 8 mg/kg]):

90

162 (307)

mg/mq;

CYCLOPHOS

-L (daily, 1000-

PHAMIDE:

1250 mg) :

2016]

600

158 (308)

[9]

4

Cancer

Cl

(every

Res

mg/mq)

2572

324 (13)

30

23

NA

320 (99)

>60% stromal

ypT0is/ypN0

NR*

NR

ypT0/ypN0

160

106

(60)

(40

TIL

DOCETAXEL (every

3

weeks,

100

mg/mq)

GeparSixto

II

08/2011-

18 cycles of

-T

(every

3

NCT01426880

RCT

12/2012

concomitant

weeks, 6 mg/kg

weekly [Denkert C. et

588

273

NA

36.

266 (97)

>60%

either

8

stromal

[load: 8 mg/kg])

(CA

intra-tumoral

PACLITAXEL

+L (daily, 750

RB

TIL

al, JCO 2015]

(80

mg):

O)

[6]

NON-

266 (273)

32.

mg/mq),

(46)

NA

PEGYLATED

8

LIPOSOMAL

(no

DOXORUBICI

CA

N (20 mg/mq)

RB

+/-

O)

or

CARBOPLATI N (AUC: 2-1.5)

41

NeoALTTO

III

01/2008-

12

weekly

NCT00553358

RCT

05/2010

PACLITAXEL

mg/kg [load: 4

(80 mg/mq)

mg/kg]) :

[Salgado R. et al,

JAMA

Oncology

-T

(weekly,

2

455

455 (100)

30

25

51

387 (85)

>30% stromal

ypT0is/ypN0

TIL

130 (149) [FEC

given

after surgery]

-L (daily, 1500 mg):

2015]

131 (154)

[7]

-T

(weekly,

2

mg/kg [4 mg/kg: load]) +L (daily, 1000 mg): 126 (152)

Abbreviations: CT: chemotherapy, pCR: pathologic complete response; TIL: tumor-infiltrating lymphocytes; HR: hormone receptor; T: trastuzumab; L: lapatinib; RCT: randomized, controlled trial; NA: not applicable; NR: not reported; *in the joint analysis HR+=263 (53%) and HR-=235 (47%); AUC: area under the curve.

42

195

192

(50)

(50

Highlights  We performed a systematic review and meta-analysis of randomized controlled trials to investigate the association between tumor-infiltrating lymphocytes (TIL) and benefit from neoadjuvant chemotherapy in combination with anti-HER2 blocking agents administered as single (trastuzumab and lapatinib) or dual blockade (trastuzumab plus lapatinib). 

High baseline TIL subgroup was associated with a significantly increased rate of pathologic complete response (pCR).



No interaction was found between TIL as categorical variable and the different antiHER2 agent(s) (trastuzumab versus lapatinib versus their combination) or the type of neoadjuvant chemotherapy regimen (anthracycline and taxanes versus taxanes only).



We observed heterogeneity between studies; this was mainly driven by the results of the NeoALTTO trial, differing from the others for the type of neoadjuvant chemotherapy administered (taxanes only before surgery) and for the definition of high TIL subgroup (>30%).

 In HER2-positive breast cancer, high baseline TIL are associated with increased pCR probability irrespective of anti-HER2 agent(s) and neoadjuvant chemotherapy regimen used.

43