Triple Negative Breast cancer in 2020 An Update.pptx

update in the recent advances in TNBC treatment covering chemotherapy options available. personalised medicine in TNBC 2020. genetic variation

1. Personalized Treatment In Metastatic Triple-negative Breast Cancer A Perspective In 2020

2. TNBC An Overview ~ 15% of all breast cancers do not express ER, PR, and HER-2 TNBC is generally recognized as chemo-responsive Nonetheless, vast majority TNBCs have poorer survival compared to other subtypes Clinical course of early-stage TNBC entails a very high 5-year recurrence rate (around 30%-50%), Recurrence rate dramatically peaks within the first 2-3 years from diagnosis, reflecting hostile biology of these tumors mTNBC : patients progress rapidly on chemotherapy, with short median OS between 12 - 18 months

3. Clinical, Epidemiological and Therapeutic Heterogeneity of TNBC ESMO Open 2017;2:e000208. doi:10.1136/ esmoopen-2017-000208

4. Site of 1 st TNBC Recurrence in NCCN Lin et al, ASCO 2009

5. Molecular Heterogenity of Breast Cancer Perou et al using gene expression profiling classified breast cancer into 4 prognostically distinct intrinsic molecular subtypes: luminal A luminal B HER enriched basal-like (BL) Majority of the TNBC phenotype belong to the BL molecular subtype TNBC and BL tumors have a lot of biological similarities like Genomic instability Prevalence of high grade & markedly proliferative tumors Expression of epidermal growth factor receptor (EGFR) High frequency of germline BRCA1 (gBRCA) mutations] Extraordinary prevalence of p53 mutations (>80%)

6. Molecular Landscape Of TNBC : Refined By Seminal Study From Vanderbilt Group Analyzed publically available data on gene expression profiles from587 TNBC cases Study identified 6reproducible gene expression subtypes of TNBC with unique gene expression pattern: 2 BL subtypes (BL-1, BL-2) Immunomodulatory subtype (IM) Mesenchymal subtype (M) Mesenchymal stem-like subtype (MSL) Luminal androgen receptor subtype (LAR)

7. Molecular Landscape of TNBC –Lehman Study Composed from publically available data on gene expression profiles from 587 TNBC cases BL1: Basal like 1 BL2: Basal like 2 IM: Immunomodulatory M: Mesenchymal MSL: Mesenchymal stem cell-type LAR: Androgen luminal receptor MET: Mesenchymal Epithelial Transition. Medicina Universitaria. 2016;18(71):105-114 Representative TNBC cell lines of above subtypes to show that analysis of distinct gene expression profiles can inform therapy selection Example: BL-TNBC is characterized by DNA-repair deficiency, and relevant cell models responded to cisplatin treatment

8. Further Refinement of Vanderbilt classification IM and MSL subtypes were essentially related to tumor sample impurities, attributed to the presence of tumor-infiltrating lymphocytes (TILs) and tumor-associated stromal cells, respectively Revised TNBC molecular subtypes comprised only 4 tumor-specific subtypes (BL-1, BL-2, M, and LAR), With remarkable divergence in histopathological features, molecular aberrations, and pattern of relapse Additionally, the 4 TNBC subtypes showed a significant difference in the rate of pCR when treated by standard neo-adjuvant chemotherapy, with 46% of BL1-patients achieving pCR, compared to 12% PCR rate BL2 29% PCR rate M 15% PCR rate LAR respectively ( P = .04).

9. Refinement of 6 subtypes by Lehman Study Group Previously described IM and MSL subtypes were essentially related to tumor sample impurities, attributed to presence of tumor-infiltrating lymphocytes (TILs) and tumor-associated stromal cells, respectively 2 Revised TNBC consists of only 4 molecular subtypes BL-1, BL-2, M, and LAR) that show significant difference in histopathology, molecular aberrations, relapse rates 2 Neoadjuvant PCR rates according to subtypes with standard Chemo 46% BL1 , 12% BL2 29% M ,15% LAR ( P = .04) 3 References 1. (Diagram )Oncotarget. 2017; 8:73329-73344 2. Breast J .2020;26:69–80. 3.PLoS ONE . 2016;11(6):e0157368.    

10. Baylor Group Profiling Baylor group performed RNA and DNA profiling analyses on 198 TNBC tumors and identified 4 stable TNBC Subtype  BL Lymphocyte depleted  BL Lymphocyte predominant,  Mesenchymal  LAR which were also characterized by different molecular profiles and displayed distinct prognosis

11. Genomic Heterogeneity of TNBC And Identification of Novel Targets Breast J.2020;26:69–80

12. Genetic events in TNBC Most commonly observed genetic events in TNBC are TP53 mutation reported in 80% all tumors Myc amplification reported in 40% of all tumors Some less frequent molecular events are as follows AR Signaling DNA damage response defects PIK3CA/AKT mutations, EGFR, and FGFR2 amplification. Unique immunogenicity encountered in some BL subtypes has Provided a good rational to investigate the role of immune checkpoint inhibitors (ICIs) in these tumors Unfortunately, these 2 aberrations are not possible to target in the clinic at the present time Potentially targetable and have been recently explored in many clinical trials Recently approved in combination with Chemotherapy Breast J.2020;26:69–80

13. Metastatic TNBC Treatment Options

14. Bevacizumab in MBC Conjoint analysis of the 3 pivotal trials E2100, AVADO and RIBBON showed clinically meaningful benefits with addition on bevacizumab to first-line chemotherapy , compared to single-agent chemotherapy in mTNBC patients (n = 621) Significant improvement in mPFS from 5.4 to 8.1 months ( P < .0001) ORR increased from 23% to 42%, in bevacizumab-treated patients Median OS was similar in both groups (18.9 and 17.5 months ( HR = 0.96; 95% CI 0.79-1.16) Use of bevacizumab combined with chemotherapy—particularly paclitaxel is endorsed by most of international guidelines, because of limited treatment options, for mTNBC patients Pre-clinical experiments have shown frequent amplification of(VEGF) in TNBC cell lines

15. Platinum Based Treatment

16. TNBC : Platinum salts seemingly exert a greater therapeutic effect among compared to other standard chemotherapeutics TNT Clinical Trial

17. Objective Response – According to BRCA 1/ 2 Status

18. Targeting DNA Damage & Response Defects

19. Targeting DNA Damage Response Defects Intrinsic genomic instability in BL-TNBCs has been strongly linked to DNA repair defects BL intrinsic subtype is enriched for tumors with BRCA1 mutations Majority of CA -breast patients with g BRCA mutations (around 75%) have a TN phenotype BRCA1 and BRCA2 are among the DNA repair genes, which are involved in repair of double-stranded DNA breaks (DSB) 2 BRCA genes are key components of DNA homologous recombination repair (HRR) pathway, which is an error-free process essential for maintaining genomic stability Breast J.2020;26:69–80

20. MOA of PARP Inhibitors Endogenous single-strand breaks (SSB) occur frequently in proliferating cells , and SSB are repaired mostly by PARP-dependent base excision repair (BER) pathway Efficient SSB repair is essential for the survival of cells. PARP inhibitors inhibit PARP and thus the repair of SSB by BER Unrepaired SSB can be converted to double-strand breaks (DSB) that are toxic to cells homologous recombination (HR) is the major pathway to repair such lesion during cell replication. HR-proficient cells can repair DSB originated from SSB to ensure genome stability and cell survival while HR-deficient cells that cannot repair those DSB undergo apoptosis and eventually cell death

21. Comparison of 3 phase III studies of PARPi in treatment of mBC patients with gBRCA mutations OlympiAD, EMBRACA and BROCADE-3] Breast J.2020;26:69–80

22. Comparison between the 3 phase III studies using PAPRi in the treatment of mBC patients with gBRCA mutations [OlympiAD, EMBRACA and BROCADE-3] Breast J.2020;26:69–80

23. Concern : carboplatin as a control arm would have been ideal control to test efficacy of olaparib. However, this trial used a control of physician’s choice which was not a true choice because the use of carboplatin was specifically disallowed OS was similar between the groups: a median of 19.3 versus 19.6 months. 95% confidence interval for the hazard ratio of the OS spans from 0.63 to 1.29, meaning by the data are compatible with not only a 37% reduction in mortality but also a 29% increase in mortality Robson M, Im S-A, and Senkus E, et al (2017) Olaparib for metastatic breast cancer in patients with a germline BRCA mutation N Engl J Med 377(6) 523–533. Gyawali B. The OlympiAD trial: who won the gold?.  Ecancermedicalscience . 2017;11:ed75. Published 2017 Dec 6. doi:10.3332/ecancer.2017.ed75

24. EMBRACA is largest randomized trial evaluating a PARP inhibitor in patients with advanced breast cancer and a germline BRCA1/2 mutation Talazoparib resulted in prolonged progression-free survival vs physician’s choice of therapy by blinded central review HR: 0.54 (95% CI, 0.41, 0.71); P < .0001

25. PARPi vs platinum's in mTNBC patients Data from ovarian cancer studies have clearly demonstrated that sensitivity to PARP- i is dependent on sensitivity to prior platinum therapy For example, In a study on 50 patients treated by olaparib for their relapsing OC, Fong et al have reported an ORR of 15% in patients with platinum-refractory disease (progression while receiving platinum or within 0-3 months after the last dose of platinum) ORR of 60% in patients with platinum sensitive disease (progression after >6 months) ABRAZO study could further confirm “prior platinum sensitivity” as a strong predictor to ensuing efficacy of PARPi treatment. ORR to talazoparib was 48% in sensitive disease (progression after >6 months following the last dose of platinum) vs 10% in platinum resistant disease. It appears that “ no disease progression for >6 months after platinum's treatment” is an important clinical biomarker to predict sensitivity to subsequent PARPi, which strongly suggests the presence of some cross-resistance between platinums and PARPi. Ovarian Cancer Breast Cancer Breast J.2020;26:69–80

26. Consciously OlympiAD and EMBRACA studies have included minority of platinum pretreated patients (21% and 16%, respectively) In both studies clinical evidence of platinum sensitivity was required for patients’ inclusion Nevertheless, efficacy data of PARP- i in platinum pretreated patients were less pronounced in both studies, although it was still non significantly superior to TP

27. In the clinic decision making in BRCA-mutant mTNBC First-line PARP- i and carboplatin both seem to be effective and rational treatment options However, it is unclear which agent should be used first While there is no head-to-head data comparing platinums vs PARPi, yet safety profile and QOL can reasonably favor 2 approved PARPi as initial therapy Although above sequence is not evidence-based, still it is an extrapolation from some pre-clinical and clinical data demonstrating platinum sensitivity in PARP inhibitor– resistant gBRCA-mutant tumors When PARPi are not accessible/affordable  then first-line carboplatin would stand as a rational treatment option as indicated by the TNT study. More recently, combination of carboplatin and paclitaxel was shown to provide remarkable efficacy results in gBRCA-mutant mBC, with a mPFS of 12 months and a m OS of >24 months , and hence it can be also considered as another evidence-based treatment option in these patients Breast J.2020;26:69–80

28. Curr Oncol. 2018 Jun;25(S1):S142-S150 Screening Guidelines for TNBC

29. Immunomodulation in TNBC

30. Targeting Immunomodulatory Features In TNBC Conventionally, breast cancers are not highly immunogenic tumors, in view of their relatively low mutation burden. In TNBC many studies confirmed that these tumors Have higher mutation burden Higher TILs compared to other breast cancer subtypes Moreover, RNA sequencing data from TCGA showed a higher expression of PD-L1 mRNA in TNBC non-TNBC ( P < .001) Which was shown to be associated with a greater cytotoxic T cell infiltrate, compared with PDL1-negative tumors ( P < .0001)

31. MOA of Immunotherapy Immune evasion happens by 1.Activation of immune inhibitory molecule (CTLA-4) by intratumoral dendritic cells 2.Tumor cells may directly inhibit cytotoxic t lymphocytes through overexpression of the (PD-L1) molecule PDL1binds to PD1receptor on cytotoxic T-lymphocytes leading to their apoptosis Ipilimumab and tremelimumab bind to and inactivate CTLA-4 preventing its activation PD1-PD-L1 binding may be preventing by blocking either PD1 (nivolumab, pembrolizumab) or PD-L1 (durvalumab, avelumab, Atezolizumab) cytotoxic T-lymphocyte associated protein 4

32. Rationale to Investigate ICIs in TNBC Peculiar immuno-genomic features of TNBC gave a strong rational to investigate the role ICIs in these patients. Early Clinical studies have clearly shown: Efficacy of either pembrolizumab (a PD-1 inhibitor), or Atezolizumab (a PD-L1 inhibitor), as single agents, was quite modest, in unselected previously treated mTNBC (ORR 4.7% and 6%, respectively) However, occasional responders to either agent had a 1-year survival rate of 100% Those with SD still had a good 1-year survival rate of ~65% %. In contrast to 2 nd line data, performance of frontline single-agent pembrolizumab and Atezolizumab was quite acceptable in early trials (ORR 21.4% and 24%, respectively).

33. Response to Single Agent Anti PDL1/PD -1 Single agent response rate of 20-25% in 1L Very low single agent responses in 2 nd line & beyond

34. Keynote 119 : Study Design

35. Pembrolizumab vs Chemo in 2L /3L TNBC : OS by PDL1 CPS Score

36. KEYNOTE-119 trial results summary Study did not meet any of its primary or secondary efficacy endpoints . In ITT analysis the mOS was 9.9 Months (Pembro) and 10.8(TPC) months , respectively (HR = 0.97; 95% CI 0.82-1.15) However in preplanned analysis according PD-L1 expression (defined as % of PD-L1 + CPS by the 22C3 assay) : performance of pembrolizumab improved in study population with CPS ≥ 1 & further pronounced in with CPS ≥ 10 Exploratory analysis of patients with CPS ≥ 20 (17% of study population) Showed median OS 14.9 Months in favor pembrolizumab vs chemotherapy 12.5 months respectively, (HR 0.58, 95% CI 0.38-0.88), Also meaningful increase in the ORR 26.3% and 11.5%, respectively was seen

37. Potential Positioning of Pembrolizumab Results of KEYNOTE-119 trial though not practice changing ,could provide an important note concerning the potential benefit of pembrolizumab in minority of mTNBC patients, whose tumors are PD-L1 strongly positive In view of acceptable safety, treating clinician may opt to use this drug as an “off-label therapy” beyond the frontline setting, in mTNBC patients with CPS ≥ 20 by the 22C3 assay.

38. Impassion 130 Study : Updated Analysis

39. Progression Free Survival Overall Survival

40. 2 nd interim results show no significant OS benefit with atezolizumab versuS placebo either in the intention-to-treat analysis or in PD-L1-positive patients Median OS in ITT population was 21.0 months in atezolizumab group and 18.7 months in the placebo group Nonsignificant HR of 0.86 ( P = .0777). PD-L1 positive patients Median OS was 25.0 months for atezolizumab and 18.0 months for placebo, with a nonsignificant hazard ratio of 0.71

41. Overall Survival by PDL1 Positivity

42. Summary -Results * Not statistically significant Statistical significance was not demonstrated in ITT But a 7.0 -month improvement in median OS observed in PD-L1+ pts with atezo + nP (25.0 mo) vs placebo + nP18.0 mo HR, 0.71 [95% CI: 0.54, 0.93]

43. Biomarker Analysis PD-L1 IC was determined to be highly predictive of Atz-nabPx efficacy Majority of PD-L1 TC+ tumors were also PD-L1 IC+. Intratumoral CD8, but not sTILs, were well correlated with PD-L1 IC CD8 was predictive of Atz-nabPx efficacy for PFS/OS sTILs were only predictive of PFS benefit Local vs central TNBC assessment was consistent in most patients. In addition, local and central lab-defined TNBC patients were determined to derive similar benefit from Atz-nabPx PD-L1 on tumor cells (TC) and tumor-infiltrating immune cells (IC); Stromal tumor -infiltrating lymphocytes ( sTILs ) IMpassion130: Efficacy in immune biomarker subgroups from phase III study of atezolizumab + nab-paclitaxel in patients with treatment-naïve, locally advanced or metastatic TNBC SABCS: 41ST ANNUAL SAN ANTONIO BREAST CANCER SYMPOSIUM — EMENS LA, LOI S, RUGO HS, SCHNEEWEISS A, DIÉRAS V, IWATA H, BARRIOS CH, NECHAEVA M, MOLINERO L, NGUYEN DUC A, FUNKE R, CHUI SY, HUSAIN A, WINER EP, ADAMS S, SCHMID P BLOOMBERG?KIMMEL | DECEMBER 05, 2018  

44. Summary Although, results of IMpassion130 are certainly practice changing, it is essential to note no role of atezolizumab in patients with PD-L1-negative tumors (around 60% of all mTNBC) Adequate PD-L1 testing exclusively by the SP 142 assay is extremely important , rather than any other commercially available assays, which were recently found to have an inferior capability to predict atezolizumab benefit in mTNBC

45. Targeting Androgen Receptors

46. Luminal Androgen Receptor (LAR) subtype LAR subtype observed in 15% of all TNBC patients LAR has specific gene expression subgroup which is recognized by both Vanderbilt and Baylor classification LAR cells have gene expression features similar to ER-positive luminal breast LAR subtype is enriched for expression of AR and its downstream gene targets Breast J.2020;26:69–80

47. LAR Tumors Metastatic Behaviors LAR tumors predominantly tend to spread to the Lymph nodes Soft tissues Bones LAR tumors have A pocrine features AR expression Less proliferation markers Modest response to chemotherapy Breast J.2020;26:69–80 Whereas BL TNBCs are known to preferentially metastasize to visceral organs, especially the lungs and brain

48. Bicalutamide :Early evidence in TNBC Gucalp et al reported 1 ST clinical trial on use of bicalutamide (150 mg daily) in 26 patients with AR-positive TNBC (defined as >10% AR nuclear staining by IHC)

49. Of 165 patients screened, 118 (72%) (intent-to-treat (ITT) population) were AR+, of whom 89 had AR staining ⩾10%. Of the patients with AR IHC ⩾ 10% and who had a post-baseline tumor assessment, 75 patients constituted the ‘evaluable population’. CBR16 :was - 25% (95% CI: 17, 33) CBR24 :was - 20% (95% CI: 14, 29) CR/PR: 6%, Median PFS :13 weeks for ITT patient population Enzalutamide in TNBC Predominant toxicities were fatigue (40% (5% G3)), nausea (32%), decreased appetite (19%). Grade 3 toxicities were observed in 10%.

50. Is There An Assay to Predict : Androgen Sensitivity In an exploratory analysis, that used a novel genomic assay (PREDICT AR) to quantify AR-related RNA expression Patients with PREDICT AR-positive tumors (nearly half of the ITT population) had more favorable outcomes than those with PREDICT AR-negative tumors, CBR at 24 weeks (36% vs. 6%) mPFS (16.1 weeks vs. 8.1 weeks) PREDICT AR genomic assay may look promising, however, it requires further validation in a larger data set before its widespread utility as a predictive biomarker for anti-AR response

51. Further directions Antiandrogen + PI3K inhibitors Frequency of PIK3CA mutations is much higher in AR-positive tumors compared to AR-negative tumors (40%vs. 4%, respectively). Promising results of combinations in preclinical trials have lead to studies using AR antagonists in combination with PI3K inhibitors in AR-positive/ PI3K mutant mTNBC Ongoing phase I/II trial is investigating efficacy and safety of alpelisib plus enzalutamide in these patients Antiandrogen + CDK4/6 i Several pre-clinical studies have shown a crosstalk between AR signaling and RB protein function that might cause resistance to antiandrogen therapy. So ,proof of concept studies are planned to test combined use of Antiandrogen agents with CDK4/6 inhibitors Breast J.2020;26:69–80

52. PI3K/AKT/mTOR Pathway in TNBC

53. PI3K/AKT/mTOR pathway in TNBC According to TCGA incidence of PI3K/AKT/mTOR pathway aberrations is quite high in BL tumors. Since all of the above molecular abnormalities would result in AKT hyper activation, hence the interest in selective AKT inhibitors to treat these tumors. Intriguingly, preclinical studies have shown that taxane treatment can induce an immediate AKT hyper activation associated with resistance to these agents Interestingly, in these models Co administration of AKT inhibitors and taxanes resulted in increased tumor cell death compared to their monotherapy use. Ipatasertib and Capivasertib are two highly selective oral pan-AKT inhibitors , which were recently tested in combination with paclitaxel phase II trials [PAKT and LOTUS

54. PAKT Trial Paclitaxel plus pan- AKT inhibitor Capivasertib Capivasertib (4 days on, 3 days off) was chosen over continuous daily dosing on the basis of data from phase I trials Paclitaxel Alone 140 Women with untreated Metastatic TNBC randomized (1:1)

55. Capivasertib plus Paclitaxel versus Placebo plus Paclitaxel as first-line therapy for metastatic triple-negative breast cancer (PAKT): A randomised, double-blind, placebo-controlled, phase II trial PIK3CA/AKT1/PTEN-non altered A) intent-to-treat population (B) PIK3CA/AKT1/PTEN-altered subgroup (mPFS) duration was 5.9 months with capivasertib versus 4.2 months with placebo (1-sided P = 0.06,meeting the predefined significance level of 1-sided P = 0.10) Median overall survival (mOS) not reached for Capivasertib Non significant improvement in OS

56. Capivasertib plus Paclitaxel versus Placebo plus Paclitaxel as first-line therapy for metastatic triple-negative breast cancer (PAKT): A randomised, double-blind, placebo-controlled, phase II trial PIK3CA/AKT1/PTEN-non altered A) intent-to-treat population (B) PIK3CA/AKT1/PTEN-altered subgroup Median overall survival (mOS) was prolonged by 6.5 months with capivasertib 19.1 months vs 12.6 months; 2-sided P = 0.04) Median overall survival (mOS) was prolonged by 6.5 months with capivasertib 19.1 months vs 12.6 months; 2-sided P = 0.04) Median overall survival (mOS) not reached for Capivasertib Non significant improvement in OS

57. Lotus Clinical Trial

58. Final OS in the ITT population

59. Final Safety Most Common Adverse Events

60. Conclusions

61. AKT Inhibition – Summary LOTUS PAKT study provides important evidence that AKT inhibition might be able to improve patient outcomes in this difficult-to-treat subtype of breast cancer Confirmatory and adequately powered phase III trials are required. On the basis of the current data, paclitaxel should be considered as the chemotherapy backbone for future studies, but other chemotherapy combinations may also be explored.

62. Targeting Androgen Receptors

63. Sacituzumab Govitecan (Immu 132) SN38 – Analogue of Irinotecan

64. Ascent Phase 3 Study

65. Median PFS Sacituzumab 5.6 months vs. 1.7 months for TPC HR = 0.41; P < .0001) HR = 0.48; 95% CI, 38-59 Significant improvement in Median OS Significant improvement in Median PFS Most common grade 3 treatment-related adverse events included n eutropenia (51%), diarrhea (10.5%), anemia (8%) and febrile neutropenia (6%), whereas the most common of these in the chemotherapy group were neutropenia (33%), diarrhea (< 1%), anemia (5%) and febrile neutropenia (2%). Results

66. The FDA has accelerated approval to Sacituzumab govitecan for the treatment of adult patients with metastatic triple-negative breast cancer (mTNBC) who have received at least 2 prior therapies for metastatic disease

67. Halaven (Eribulin Mesylate ) in TNBC

68. Eribulin Significantly Improved OS vs Capecitabine In TNBC Patients Median OS was significantly longer for Halaven vs Capecitabine in TNBC patients 14.4 vs 9.4 months, respectively (HR: 0.70; 95% CI: 0.54–0.91; P=0.006) Twelves C, et al. Breast Cancer: Basic and Clinical Research 2016;10:77–84. Prespecified analysis of 301 clinical trial of Eribulin vs Capecitabine

69. Large Real World Study confirms that Eribulin used earlier (2 nd line) can lead to higher ORR than when used in later lines New Real World Study

70. TNBC accounts for 15% - 20% of all breast cancers diagnosed in United States HR-positive & HER2-positive tumors average survival exceeds 50 months Survival for metastatic (mTNBC) is considerably shorter ranging from 11 to 17.8 months Sequential chemotherapy remains the mainstay of treatment for patients with mTNBC TNBC in united states

71. 1 st retrospective large community-based study of Eribulin in mTNBC patients in US Patients from Oncology Providers Extended Network treated with Eribulin Biopsy confirmed diagnosis of mBC Triple negative defined as ER <1%, PR <1%, and HER2-negative according to current American Society of Clinical Oncology and College of American Pathologists guidelines Methods Exclusion Criteria: Patients participating in any interventional clinical trials prior to start of eribulin therapy Patients receiving treatment for a second primary malignancy while under the care of responding physician

72. ECOG-PS Drug regimens received up to initiation (including neoadjuvant/adjuvant) treatment Eribulin by line of therapy (LOT) Disease response (per treating physician) Adverse events (AEs) during eribulin treatment Date of death (or last follow-up) e-CRF data Data validation : All submitted eCRFs were reviewed by Cardinal Health clinical research staff for quality control Random sample of 10% of all submitted eCRFs was validated through provider follow-up. Data which could not be validated were not included in the final study analysis

73. Study Design End Points : Response Rate (CR + PR) Overall Survival

74. Results Cohort Response Rate Median OS EU (1 st or 2 nd line) 69.9 % 23 Months (95% CI:18.7-27.3) LU (3 rd or 4 th line 48.8 % 14.7 Mont hs (95% CI:12.6-16.9) Median OS in all patients 17.6 months (95% CI:15.3-19.9) Safety Most common AEs reported among EU patients were as follows: fatigue(64.8%), neutropenia (36.8%), weakness (36.8%), leukopenia (32.0%), and peripheral neuropathy (25.6%) Most frequently diagnosed AEs in LU patients were: fatigue (65.4%),weakness (43.3%), decreased appetite (40.2%), peripheral neuropathy (37.0%), and alopecia (36.2%).

75. Why Eribulin works so well in TNBC 1.Converting aggressive Mesenchymal Phenotype to Epithelial Phenotype 2.Immunomodulatory MOA

76. Image reference https://blogs.scientificamerican.com/guest-blog/the-hallmarks-of-cancer-6-tissue-invasion-and-metastasis/ EMT : Key Metastasis Promoting Step in Any Cancer Mesenchymal Phenotype is predominant phenotype in TNBC

77. Correlation between the expression of EMT-TFs and prognostic value in MBC patients 3,218 MBC patients from 14 eligible studies were evaluated High EMT-TF expression was significantly associated with poor prognosis in MBC patients (HRs = 1.72; 95% CI= 1.53–1.93; P = 0.001) EMT-TF expression levels significantly increased the risk of MBC in the Asian population (HR = 2.11, 95% CI = 1.70–2.62) Reference: Emani, Saber et al. “Prognostic Value of EMT-Inducing Transcription Factors (EMT-TFs) in Metastatic Breast Cancer: A Systematic Review and Meta-Analysis.”  Scientific Reports  6 (2016): 28587.  PMC . Web. 4 Feb. 2018. EMT TFs = EMT -Inducing Transcription Factors Impact of tumor microenvironment related factors on tumor progression Expression on mesenchymal markers is often associated with poor prognosis & dev of metastases

78. Eribulin may render residual tumors less aggressive & less likely to metastasize 2 Effect of Eribulin on EMT in preclinical studies 2 Phenotypic switch from mesenchymal cells to less aggressive epithelial cells Observed decrease in migration & invasiveness Suppression of further metastasis Up regulation of epithelial markers Down regulation of mesenchymal markers Down regulation of TGF ß/ Smad pathway    MOA is also substantiated by studies in actual patients

79. Why Eribulin works so well in TNBC 2.Immunomodulatory MOA

80. Effect To Antitumor A utoimmune Response of Eribulin 52 patients with locally advanced or metastatic breast cancer treated with eribulin Before/after biopsies from 10 patients Correlated response rate (RR) with immunohistochemical evaluation: TILS: PD-1, CD8, FOXP3 Cancer cells: PD-L1, PD-L2 EMT markers: E-cadherin, N-cadherin, vimentin, CA9 W Goto et al.: AACR2016 Abst #5127 MOA data is now supported by studies in patients also

81. 100 before E-cadherin 60 20 (%) after 80 before Vimentin 60 40 20 (%) after 100 before CA9 (%) after 80 40 before N-cadherin (%) after 100 60 20 80 40 80 60 40 20 100 84 W Goto et al.: AACR2016 Abst #5127 Comparison of the EMT and hypoxia marker express ion of before and after treatment in 5 patients(responders). Increase in expression of epithelial markers E –Cadherin after Eribulin treatment Decrease in expression of Mesenchymal markers Vimentin after Eribulin Treatment Decrease in expression of Mesenchymal markers N Cadherin after Eribulin Treatment Decrease in expression of Hypoxia markers CA-9 after Eribulin Treatment

82. Comparison of the immune environment marker expression of before and after treatment in 5 patients(responders). 8 before PD-1 6 4 3 2 1 (%) after 20 before PD-L1 15 10 5 (%) after before PD-L2 after 60 50 40 30 20 10 (%) 30 before CD8 25 20 15 10 5 (%) after 35 7 5 85 W Goto et al.: AACR2016 Abst #5127 Increase in expression of CD8 markers Decrease in PD1,PDL1,PDL2 epxression post Eribulin treatment

83. Statistically significant correlations between clinical RR and change in marker status for PD-L1 (p = 0.024) FOXP3 (p = 0.004) E-cadherin (p = 0.004) CA9 (p = 0.024) Immune suppression markers (PD-L1, FOXP3) going negative correlated with Clinical RR Reversal of EMT (E-cadherin going up) Loss of hypoxia (CA9 going down) correlated with clinical RR Results

84. Eribulin’s Complex Effects on Tumor Biology Now Recognized in the FDA-Approved Prescribing Information after comprehensive preclinical and clinical data 87 Halaven Package Insert, Woodcliff Lake, NJ. Eisai Inc. Jan 2016

85. Trimodal Molecular Mechanism Of Eribulin In Metastatic Breast Cancer Vascular remodeling & increase in tumor MVD Clinical implications: Increase subsequently administered drug delivery Decrease in hypoxia induced consequences Mesenchymal to Epithelial Phenotype Switch: Action mediated by : TGF b inhibition Clinical implication : Renders residual tumor less aggressive and less likely to metastasize by changing Mesenchymal to epithelial phenotype Facilitate favorable immune microenvironment Clinical implication: Inhibits evasion from immune mechanisms by inhibiting PD-1 & PD L-1 Local infiltration of antitumor effector cells (CD8) Anticancer Research 38,2929-2938;2018.

86. Eribulin has Very Effective rating from NCCN Gemcitabine, Capecitabine & Vinorelbine has moderately effective rating

87. Let Us Simplify Metastatic TNBC Through A Treatment Algorithm

88. TNBC Probable Treatment Algorithm in 2020 Breast J.2020;26:69–80

89. Conclusion Breakthrough genomic and molecular data, has lead development of targeted therapies for different subsets of mTNBC Few promising therapies have already changed treatment landscape & many others are undergoing clinical development Identification of appropriate predictive biomarkers is key challenge for new targeted agents Prospective validation of biomarkers to identify the best responders to these drugs is required to realize promise of molecularly targeted therapy Chemoterapy with Eribulin in 2 nd line & Carboplatin based chemotherapy in 1 st line are still effective treatment options for many patients