Immunotherapy for early breast cancer: too soon, too superficial, or just right?

•The investigation of immunotherapeutic agents for breast cancer has shifted from the metastatic to early setting.•Different scenarios and designs exist clinical trials testing immunotherapy in cancer.•Combinatorial strategies may increase activity across different subtypes.•Biomarkers predicting upfront resistance, treatment response, immune-related toxicities are needed.•Collaborative research is required understanding available data rational design future trials. Immunotherapy emerged as a new modality cancer, its use approved combination with chemotherapy first-line therapy triple-negative overexpressing PD-L1. As immune checkpoint inhibitors alone have modest advanced there growing interest combinatorial modalities, particularly their rapid development disease setting. The plethora ongoing comes at time when solid still imperfect. This review offers perspective on efforts establish efficacy safety cancer. Breast (BC) heterogeneous disease, both histological molecular level often segregated into five intrinsic subtypes: luminal A, B, human epidermal growth factor receptor (HER2)-positive, (TNBC), normal-like.1Wagner J. Rapsomaniki M.A. Chevrier S. et al.A single-cell atlas tumor ecosystem cancer.Cell. 2019; 177: 1330-1345.e18Abstract Full Text PDF PubMed Scopus (185) Google Scholar, 2Luen Virassamy B. Savas P. al.The genomic landscape interaction host immunity.Breast. 2016; 29: 241-250Abstract (103) 3Hammerl D. Smid M. Timmermans A.M. al.Breast genomics immuno-oncological markers guide therapies.Semin Cancer Biol. 2018; 52: 178-188Crossref (45) Scholar Besides varying expression targetable receptors, these subtypes also distinct respect profiles, differing quantity composition tumor-infiltrating lymphocytes (TILs) microenvironment, PD-L1 expression, tumor-associated antigens, mutational burden (TMB).2Luen 4Sobral-Leite Van de Vijver K. Michaut al.Assessment subtypes, relation mutation rate, BRCA1-like status, cells survival.Oncoimmunology. 7: e1509820Crossref (42) 5Stanton S.E. Adams Disis M.L. Variation incidence magnitude systematic review.JAMA Oncol. 2: 1354-1360Crossref (224) Importantly, previous study Genome Atlas analyzing profiles BC not one was identified immunologically quiet, reinforcing possibility leveraging pre-existing immunity boost response against BC.6Thorsson V. Gibbs D.L. Brown S.D. cancer.Immunity. 48: 812-830.e14Abstract (1119) Scholar,7Gatti-Mays M.E. Balko J.M. Gameiro S.R. al.If we build it they will come: targeting cancer.npj Cancer. 5: 37Crossref (49) Although initially slower compared other tumors, number investigating been exponentially over past few years.8Esteva F.J. Hubbard-Lucey V.M. Tang al.Immunotherapy targeted combinations cancer.Lancet 20: e175-e186Abstract (132) Notably, being conducted parallel setting, an unprecedented approach, drugs traditionally tested setting before A likely explanation this approach notion that tumors become less sensitive or responsive time, reflected, example, by decrease TILs low levels signatures samples comparison primary tumors.9Aftimos P.G. Antunes De Melo e Oliveira Hilbers F. al.First report AURORA, international group (BIG) screening initiative (MBC) patients (pts).Ann 30: III48Abstract 10Hutchinson K.E. Yost Chang C.-W. al.Comprehensive profiling poor-risk paired recurrent cancers reveals phenotype shifts.Clin Res. 2020; 26: 657Crossref (21) 11Szekely Bossuyt Li X. al.Immunological differences between cancer.Ann 2232-2239Abstract (90) comprehensive overview describing current BC. Moreover, discuss immunotherapy-based We selected biomarkers, propose reflection costs benefits shifting investigations 15 September 2020, 82 were Clinicaltrials.gov EURODACT BC, either monotherapy treatments (Figure 1A). Regarding yet recruiting, 49 13 already completed recruitment, 7 reported preliminary final results. Of studies, 69 (84%) immunomodulators adaptive mostly T-cell cell via PD-1 (n = 34) 35). With populations, all represented, most TNBC B 1B). Different settings studied, majority neoadjuvant 64, 78.0%) including window-of-opportunity (WOP) 27, 33%). Studies adjuvant post-neoadjuvant indications represent 10% 12% efforts, respectively. evaluating anti-PD-1/PD-L1 agents, only eight strategy. 61) evaluate chemotherapy, immunotherapy, therapies, oncolytic viruses, local such radiotherapy cryotherapy, vaccines, novel signaling pathway inhibitors. chord diagram depicted Figure 2 shows under aiming augment 3 illustrates (ICIs) popular scenario nowadays pathological complete (pCR) endpoint. Because pCR weaker surrogate long-term benefit than HER2-positive TNBC, sometimes replaced combined percentages residual (RCB) 0 1 subtype. mature event-free survival (EFS) should be ultimate goal any therapy, but requires large sample studies powered RCB After surgery, allows more definitive endpoint, usually invasive disease-free (DFS), exposes many unnecessary escalation. very rare able identify who truly benefited strategy after trial concluded. When occurs target population narrower, prior drug exposure might made resistant approaches. Interestingly, small WOP carried full course using biomarker changes Those intended used ‘discovery tools’ without diagnostic therapeutic intent. They advantage short period (usually weeks) untreated (in case induction treatment), thus avoiding pitfall therapies affecting outcomes evaluation endpoints. allow quick assessment (on biologic endpoint) ultimately merits investigation. However, some drawbacks exposition treatment-naive unknown limited fact heterogeneity limit utility correlative research. Additionally, absence validated biomarkers limits prediction benefit. agent presence activation system through neoantigens. In addition, select following two main approaches: 4) (i) classic subtype-driven which investigator chooses subtype test selects subjects classical criteria, (ii) sophisticated immunogenicity-driven patient selection informed characteristics, beyond standard subtypes. Following based (pre-)clinical hypothesis pre-defined threshold efficacy, further steps (continuing randomized interruption). unlikely pursued pharmaceutical companies because suitable faster approval commercial interests. It role academia promote designs, could facilitate identification active well prompt elimination ineffective regimens. Data rationale results described below. acknowledge important listed discussed manuscript would deserve ad-hoc Pre-clinical suggest several can exert immunostimulatory effect, inhibiting immunosuppressive and/or activating effector increasing infiltration immunogenicity.12Heinhuis K.M. Ros W. Kok al.Enhancing antitumor combining tumors.Ann 219-235Abstract (97) 13Kersten Salvagno C. Visser Exploiting immunomodulatory properties chemotherapeutic improve success immunotherapy.Front Immunol. 2015; 6: 516Crossref 14Galluzzi L. Senovilla Zitvogel secret ally: immunostimulation anticancer drugs.Nat Rev Drug Discov. 2012; 11: 215-233Crossref (480) 15Emens L.A. Middleton G. interplay chemotherapy: harnessing potential synergies.Cancer Immunol 3: 436Crossref (339) 16Galluzzi Buqué A. Kepp O. effects conventional agents.Cancer Cell. 28: 690-714Abstract (723) For instance, anthracyclines induce immunogenic death dendritic subsequent specific responses.12Heinhuis examples, taxanes reduce cells, TILs,13Kersten Scholar,14Galluzzi reprogram macrophages,17Wanderley C.W. Colón D.F. Luiz J.P.M. al.Paclitaxel reduces reprogramming macrophages M1 profile TLR4-dependent manner.Cancer 78: 5891PubMed while eribulin transforming factor-β signaling18Yoshida T. Ozawa Y. Kimura al.Eribulin mesilate suppresses experimental metastasis reversing epithelial-mesenchymal transition (EMT) mesenchymal-epithelial (MET) states.Br J 2014; 110: 1497-1505Crossref (210) Scholar,19Ueda Saeki Takeuchi H. al.In vivo imaging eribulin-induced reoxygenation patients: bevacizumab.Br 114: 1212-1218Crossref (58) FOXP3.20Goto Kashiwagi Asano promotes responses locally cancer.Anticancer 38: 2929-2938PubMed Most chemo-immunotherapy approaches generated TNBC. Some reasons why focused characterized robust infiltrate, higher instability, non-synonymous mutations subtypes.2Luen Scholar,21Denkert von Minckwitz Darb-Esfahani al.Tumour-infiltrating prognosis cancer: pooled analysis 3771 treated therapy.Lancet 19: 40-50Abstract (548) 22Loi Drubay al.Tumor-infiltrating prognosis: individual early-stage cancers.J Clin 37: 559-569Crossref (200) 23Mittendorf E.A. Philips A.V. Meric-Bernstam al.PD-L1 cancer.Cancer 361Crossref (584) highest unmet medical need. Initially, investigated ICIs setting.24Nanda R. Chow L.Q.M. Dees E.C. al.Pembrolizumab phase Ib KEYNOTE-012 study.J 34: 2460-2467Crossref (756) 25Adams Schmid Rugo H.S. previously cohort II KEYNOTE-086 study.Ann 397-404Abstract (225) 26Dirix L.Y. Takacs I. Jerusalem al.Avelumab, anti-PD-L1 antibody, 1b JAVELIN study.Breast Res Treat. 167: 671-686Crossref (273) 27Emens Cruz Eder J.P. al.Long-term analyses atezolizumab study.JAMA 74-82Crossref (250) While rates unselected population, those whose contained high numbers expression.27Emens Subsequently, (Table 1). Chemo, chemotherapy; CPS, positive score; EC/AC, epirubicin /doxorubicin cyclophosphamide; EFS, survival; HR, hazard ratio; ITT, intention treat; OR, odds OS, overall pCR, response; PD-L1, programmed ligand; PFS, progression-free-survival; TILs, infiltrating lymphocytes; Triple negative IMpassion130 nab-paclitaxel placebo TNBC.28Schmid al.Atezolizumab cancer.N Engl Med. 379: 2108-2121Crossref (1433) co-primary endpoints progression-free (PFS) (OS) intention-to-treat (ITT) subgroup, if ITT turned out statistically superior.28Schmid PFS population: 7.2 versus 5.5 months [hazard ratio (HR) 0.80; 95% confidence interval (CI) 0.69-0.92] 7.5 5.0 (HR 0.62; CI 0.49-0.78). clinically meaningful OS suggested receiving atezolizumab: 25.0 18.0 0.71; 0.54-0.94).28Schmid Based results, health authorities Final years follow-up pointed towards benefit, restricted disease.29Emens IMpassion130: pivotal III + vs +nab-paclitaxel 31: S1142-S1215Google An initial Keynote355 trial, assessed pembrolizumab chemotherapies (paclitaxel, nab-paclitaxel, gemcitabine carboplatin), demonstrated PDL-1 population.30Cortes Cescon D.W. al.KEYNOTE-355: randomized, double-blind, inoperable cancer.J 1000Crossref contrast, IMpassion131 paclitaxel, did show regardless PD-L1.31Milles Primary IMpassion131, double-blind placebo-controlled randomised paclitaxel (PAC) +/- (atezo) unresectable advanced/metastatic (mTNBC).Ann Further better understand Possible explanations include concomitant steroids clarified. degree backbone affects effectiveness completely elucidated. single-arm, rate 25% irrespective status chemotherapy.32Tolaney S.M. Kalinsky Kaklamani V.G. Ib/II (ERI) plus (PEMBRO) (mTNBC) (ENHANCE 1).J 1015Crossref TONIC occurred doxorubicin (35%) cisplatin (23%) followed nivolumab/doxorubicin.33Voorwerk Slagter Horlings H.M. al.Immune enhance sensitivity blockade: trial.Nat 25: 920-928Crossref (218) These ‘induction treatments’ shown associated upregulation genes involved PD-1/PD-L1 cytotoxicity pathways.33Voorwerk choice could, part, explain conflicting observed I-SPY2 first promising setting.34Nanda Liu M.C. Yau high-risk (BC): I-SPY 2.J 2017; 35: 506Crossref below GeparNuevo study, received durvalumab every 4 weeks epirubicin/cyclophosphamide.35Loibl Untch Burchardi N. addition anthracycline taxane-based 1279-1288Abstract (135) included given nab-paclitaxel.35Loibl seen cohort, where 61.0% 41.4% (P 0.035).35Loibl NeoTRIPaPD-L1 failed demonstrate carboplatin/nab-paclitaxel surgery chemotherapy.36Gianni Huang C.-S. Egle al.Abstract GS3-04: pathologic triple negative, NeoTRIPaPDL1 Michelangelo study.Cancer 80 (GS3-04)Google discordance baseline within arms37Bianchini Tumor (TILs), dynamics trial.Ann contributed result. KEYNOTE-522 evaluated platinum-based newly diagnosed TNBC.38Schmid Cortes Pusztai 382: 810-821Crossref (311) Patients four cycles paclitaxel/carboplatin doxorubicin/epirubicin cyclophosphamide, placebo.38Schmid endpoint—pCR 602 patients—a significant (64.8% 51.2%; P < 0.001). independent difference larger node-positive disease: 20.6% 6.3%.38Schmid At 18 follow-up, trend superior EFS arm (91.3% 85.3%; HR: 0.63; 0.43-0.93). IMpassion031 (nab-paclitaxel, cyclophosphamide) met endpoint demonstrating improvement (58% 41%; 0.004) expression.39Mittendorf Zhang Barrios C.H. al.Neoadjuvant sequential anthracycline-based (IMpassion031): randomised, trial.Lancet. 396: 1090-1100Abstract reinforce idea microenvironment seems considerable needed order Hormone receptor-positive considered ‘cold’ tumor, lower TMB, expression.2Luen Scholar,4Sobral-Leite Scholar,5Stanton