Abstract S4-05: Impact of intrinsic subtype by PAM50 and other gene signatures on pathologic complete response (pCR) rates in triple-negative breast cancer (TNBC) after neoadjuvant chemotherapy (NACT) +/- carboplatin (Cb) or bevacizumab (Bev): CALGB 40603/150709 (Allianc

Background: Adding either Cb or Bev to standard NACT significantly increases pCR rates in TNBC (Sikov et al, SABCS 2013). Genomic analysis may help us to identify determinants of response within this clinical phenotype.

Methods: Patients (pts) with clinical stage II-III TNBC received weekly paclitaxel x 12 followed by ddAC x 4 +/- Cb and/or Bev. Pre-treatment biopsies were collected in formalin, RNAlater and OCT; residual disease at surgery was biopsied when possible. Illumina mRNA sequencing (RNAseq) was performed. Gene expression values were normalized to a TCGA subset of clinically TNBC samples prior to downstream analysis. pCR was defined as the absence of residual invasive cancer in the breast (ypT0/is). For each molecular signature, prognostic (effect on pCR in the overall study population) and predictive (effect of the addition of Cb or Bev, separately, on pCR) relationships were explored with logistic regression models.

Results: PAM50 subtype analysis was performed on 367 pre-treatment samples (of 443 pts who started NACT); pCR results were available for 360, comprising the analysis subset. 87% of these displayed a basal-like gene expression pattern, 2% claudin-low, 4% HER2-enriched, <1% luminal A and 7% normal-like. In pts with basal-like tumors, pCRs rose from 47% to 61% with the addition of Cb (p=0.014), an increment which did not differ significantly from the overall study population (adding in the small number of non-basal-like tumors) (interaction p=0.93). In contrast, the addition of Bev increased pCRs in basal-like tumors from 45% to 64% (p=0.0009), while reducing pCRs in non-basal-likes from 60% to 43% (interaction p=0.024); thus, a basal-like gene expression pattern was predictive of benefit from Bev. Expression of a variety of immune signatures (B cell, T cell, IgG) was positively associated with pCR, but not predictive of increased benefit from either Cb or Bev. High expression of the HER2 amplicon signature was uncommon and not prognostic for pCR overall but was associated with reduced benefit from Cb (interaction p = 0.025). High proliferation, high p53 mutation and low IE (estrogen signaling) signatures were prognostic for higher pCR rates and predictive of benefit from Bev (interaction p=0.031, 0.0017, 0.0002, respectively). In basal-like pts with residual disease, surgical samples often (52%) displayed a normal-like PAM50 pattern, though this might be due to ‘contamination’ in low volume residual disease.

Conclusions: Selection criteria led to accrual of a high % of pts with basal-like tumors, limiting our ability to assess prognostic or predictive impact of intrinsic subtype on pCR. Given that limitation, the magnitude of pCR benefit with Cb was consistent across subtypes, while a basal-like pattern was predictive of greater pCR increment with Bev. Ongoing studies will test a large number of other gene signatures and biomarkers, including the Lehmann et al subtypes. Recognition of clinically relevant subpopulations within TNBC may distinguish pts likely to achieve a pCR from those for whom an investigational approach might be considered.

Citation Format: William M Sikov, William T Barry, Katherine A Hoadley, Brandelyn N Pitcher, Baljit Singh, Sara M Tolaney, Charles S Kuzma, Timothy J Pluard, George Somlo, Elisa R Port, Mehra Golshan, Donald A Berry, Olwen M Hahn, Lisa A Carey, Charles M Perou, Clifford A Hudis, Eric P Winer. Impact of intrinsic subtype by PAM50 and other gene signatures on pathologic complete response (pCR) rates in triple-negative breast cancer (TNBC) after neoadjuvant chemotherapy (NACT) +/- carboplatin (Cb) or bevacizumab (Bev): CALGB 40603/150709 (Allianc [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr S4-05.

Abstract S3-06: Mutational analysis of CALGB 40601 (Alliance), a neoadjuvant phase III trial of weekly paclitaxel (T) and trastuzumab (H) with or without lapatinib (L) for HER2-positive breast cancer

Background: In CALGB 40601, the HER2-Enriched (HER2-E) molecular subtype had significantly higher pathologic complete response (pCR) rates regardless of treatment arm (single HER2-targeting with T+L or T+H, dual targeting with T+H+L) (Carey et al, ASCO 2014). A TP53 mutation gene expression signature was significant in a multivariable analysis as were treatment, molecular subtype, proliferation, and an immune cell genomic signature. Mutational analysis is now available for this sample set.

Methods: 265 of 305 enrolled patients (pts) had RNA sequencing (RNAseq) of pre-treatment biospecimens; 181/265 had whole exome sequencing (WES) available from tumor and matched normal blood. Somatic mutations were detected by the program UNCeqR, which integrates WES and RNAseq. We examined the association of mutations with in-breast pCR, molecular subtypes, and gene expression signatures.

Results: In this subset, there were 57 HER2-E, 58 Luminal A, 51 Luminal B, 9 Basal-like, 4 Normal-like, and 2 Claudin-low. The pCR rate was 45% (51% THL, 47% TH and 34% TL), consistent with the entire study population. TP53 was the most frequently mutated gene (56%); frequency varied by molecular subtype (Fisher p<0.0001) with the highest in the HER2-E (88%). Type of mutation also varied by molecular subtype: 34% of TP53 mutations in HER2-E pts were nonsense or frame shift mutations compared to 20% in Luminal B and 11% in Luminal A. The presence of a TP53 mutation was significantly associated with achieving pCR (59% compared to 28% in wildtype; odds ratio=3.7, p<0.0001) which did not vary by treatment arm. TP53 mutation status by WES was strongly associated with a gene-expression based predictor (AUC=0.85, p<0.001), suggesting the RNAseq-based signature could be used as a surrogate measure of genotype. PIK3CA mutations were present in 36 pts (20%); 33/36 (92%) were in exons 9 and 20. PIK3CA mutations varied moderately among subtypes and were most prevalent in Luminal B (31%) and HER2-E (25%). Rates of pCR did not vary by PIK3CA mutation status (39% vs 47% in wildtype, p=0.46). GATA3 mutations were identified in 7 pts (4 Luminal A, 3 Luminal B), but only 1 pt achieved pCR. ERBB2 mutations were found in 7 pts: 2 HER2-E, 2 Luminal A, 3 Luminal B. Two were previously identified (the lapatinib-sensitive activating mutation V777L and the lapatinib-resistant mutation L755S), both were trastuzumab resistant in experimental models (Bose et al, Cancer Discovery 2013). Consistent with these results, the V777L pt achieved pCR on the THL arm; the L755S pt did not achieve pCR on the TL arm.

Conclusions: TP53 mutation is a frequent, clinically important event in HER2-positive disease and predicts pCR to chemotherapy plus HER2-targeting. Frequency and type of mutation was dependent on molecular subtype within this clinically HER2-positive cohort. Ongoing analyses are comparing WES data between pre- and post-treatment samples as well as investigating copy number and clonality. This research is supported in part by funds from GlaxoSmithKline and grants from the Breast Cancer Research Foundation.

Citation Format: Katherine A Hoadley, William T Barry, Brandelyn N Pitcher, Joel S Parker, Matthew D Wilkerson, William Irvin Jr, Norah Lynn Henry, Sara M Tolaney, Chau Dang, Ian E Krop, Donald A Berry, Elaine R Mardis, Charles M Perou, Eric P Winer, Clifford A Hudis, Lisa A Carey. Mutational analysis of CALGB 40601 (Alliance), a neoadjuvant phase III trial of weekly paclitaxel (T) and trastuzumab (H) with or without lapatinib (L) for HER2-positive breast cancer [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr S3-06.

The Neoadjuvant Model Is Still the Future for Drug Development in Breast Cancer

The many improvements in breast cancer therapy in recent years have so lowered rates of recurrence that it is now difficult or impossible to conduct adequately powered adjuvant clinical trials. Given the many new drugs and potential synergistic combinations, the neoadjuvant approach has been used to test benefit of drug combinations in clinical trials of primary breast cancer. A recent FDA-led meta-analysis showed that pathologic complete response (pCR) predicts disease-free survival (DFS) within patients who have specific breast cancer subtypes. This meta-analysis motivated the FDA's draft guidance for using pCR as a surrogate endpoint in accelerated drug approval. Using pCR as a registration endpoint was challenged at ASCO 2014 Annual Meeting with the presentation of ALTTO, an adjuvant trial in HER2-positive breast cancer that showed a nonsignificant reduction in DFS hazard rate for adding lapatinib, a HER-family tyrosine kinase inhibitor, to trastuzumab and chemotherapy. This conclusion seemed to be inconsistent with the results of NeoALTTO, a neoadjuvant trial that found a statistical improvement in pCR rate for the identical lapatinib-containing regimen. We address differences in the two trials that may account for discordant conclusions. However, we use the FDA meta-analysis to show that there is no discordance at all between the observed pCR difference in NeoALTTO and the observed HR in ALTTO. This underscores the importance of appropriately modeling the two endpoints when designing clinical trials. The I-SPY 2/3 neoadjuvant trials exemplify this approach.

Combining progression-free survival and overall survival as a novel composite endpoint for glioblastoma trials

BACKGROUND

The use of auxiliary endpoints may provide efficiencies for clinical trial design, but such endpoints may not have intrinsic clinical relevance or clear linkage to more meaningful endpoints. The purpose of this study was to generate a novel endpoint that considers both overall survival (OS) and earlier events such as progression-free survival (PFS) and determine whether such an endpoint could increase efficiency in the design of glioblastoma clinical trials.

METHODS

Recognizing that the association between PFS and OS varies depending on therapy and tumor type, we developed a statistical model to predict OS based on PFS as the trial progresses. We then evaluated the efficiency of our model using simulations of adaptively randomized trials incorporating PFS and OS distributions from prior published trials in neuro-oncology.

RESULTS

When treatment effects on PFS and OS are concordant, our proposed approach results in efficiency gains compared with randomization based on OS alone while sacrificing minimal efficiency compared with using PFS as the primary endpoint. When treatment effects are limited to PFS, our approach provides randomization probabilities that are close to those based on OS alone.

CONCLUSION

Use of OS as the primary endpoint, combined with statistical modeling of the relationship between OS and PFS during the course of the trial, results in more robust and efficient trial designs than using either endpoint alone.

Effects of screening and systemic adjuvant therapy on ER-specific US breast cancer mortality

BACKGROUND

Molecular characterization of breast cancer allows subtype-directed interventions. Estrogen receptor (ER) is the longest-established molecular marker.

METHODS

We used six established population models with ER-specific input parameters on age-specific incidence, disease natural history, mammography characteristics, and treatment effects to quantify the impact of screening and adjuvant therapy on age-adjusted US breast cancer mortality by ER status from 1975 to 2000. Outcomes included stage-shifts and absolute and relative reductions in mortality; sensitivity analyses evaluated the impact of varying screening frequency or accuracy.

RESULTS

In the year 2000, actual screening and adjuvant treatment reduced breast cancer mortality by a median of 17 per 100000 women (model range = 13-21) and 5 per 100000 women (model range = 3-6) for ER-positive and ER-negative cases, respectively, relative to no screening and no adjuvant treatment. For ER-positive cases, adjuvant treatment made a higher relative contribution to breast cancer mortality reduction than screening, whereas for ER-negative cases the relative contributions were similar for screening and adjuvant treatment. ER-negative cases were less likely to be screen-detected than ER-positive cases (35.1% vs 51.2%), but when screen-detected yielded a greater survival gain (five-year breast cancer survival = 35.6% vs 30.7%). Screening biennially would have captured a lower proportion of mortality reduction than annual screening for ER-negative vs ER-positive cases (model range = 80.2%-87.8% vs 85.7%-96.5%).

CONCLUSION

As advances in risk assessment facilitate identification of women with increased risk of ER-negative breast cancer, additional mortality reductions could be realized through more frequent targeted screening, provided these benefits are balanced against screening harms.

Brain Malignancy Steering Committee clinical trials planning workshop: report from the Targeted Therapies Working Group

Glioblastoma is the most common primary brain malignancy and is associated with poor prognosis despite aggressive local and systemic therapy, which is related to a paucity of viable treatment options in both the newly diagnosed and recurrent settings. Even so, the rapidly increasing number of targeted therapies being evaluated in oncology clinical trials offers hope for the future. Given the broad range of possibilities for future trials, the Brain Malignancy Steering Committee convened a clinical trials planning meeting that was held at the Udvar-Hazy Center in Chantilly, Virginia, on September 19 and 20, 2013. This manuscript reports the deliberations leading up to the event from the Targeted Therapies Working Group and the results of the meeting.

Impact of the addition of carboplatin and/or bevacizumab to neoadjuvant once-per-week paclitaxel followed by dose-dense doxorubicin and cyclophosphamide on pathologic complete response rates in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance)

PURPOSE

One third of patients with triple-negative breast cancer (TNBC) achieve pathologic complete response (pCR) with standard neoadjuvant chemotherapy (NACT). CALGB 40603 (Alliance), a 2 × 2 factorial, open-label, randomized phase II trial, evaluated the impact of adding carboplatin and/or bevacizumab.

PATIENTS AND METHODS

Patients (N = 443) with stage II to III TNBC received paclitaxel 80 mg/m(2) once per week (wP) for 12 weeks, followed by doxorubicin plus cyclophosphamide once every 2 weeks (ddAC) for four cycles, and were randomly assigned to concurrent carboplatin (area under curve 6) once every 3 weeks for four cycles and/or bevacizumab 10 mg/kg once every 2 weeks for nine cycles. Effects of adding these agents on pCR breast (ypT0/is), pCR breast/axilla (ypT0/isN0), treatment delivery, and toxicities were analyzed.

RESULTS

Patients assigned to either carboplatin or bevacizumab were less likely to complete wP and ddAC without skipped doses, dose modification, or early discontinuation resulting from toxicity. Grade ≥ 3 neutropenia and thrombocytopenia were more common with carboplatin, as were hypertension, infection, thromboembolic events, bleeding, and postoperative complications with bevacizumab. Employing one-sided P values, addition of either carboplatin (60% v 44%; P = .0018) or bevacizumab (59% v 48%; P = .0089) significantly increased pCR breast, whereas only carboplatin (54% v 41%; P = .0029) significantly raised pCR breast/axilla. More-than-additive interactions between the two agents could not be demonstrated.

CONCLUSION

In stage II to III TNBC, addition of either carboplatin or bevacizumab to NACT increased pCR rates, but whether this will improve relapse-free or overall survival is unknown. Given results from recently reported adjuvant trials, further investigation of bevacizumab in this setting is unlikely, but the role of carboplatin could be evaluated in definitive studies, ideally limited to biologically defined patient subsets most likely to benefit from this agent.

Addressing overdiagnosis and overtreatment in cancer: a prescription for change

A vast range of disorders--from indolent to fast-growing lesions--are labelled as cancer. Therefore, we believe that several changes should be made to the approach to cancer screening and care, such as use of new terminology for indolent and precancerous disorders. We propose the term indolent lesion of epithelial origin, or IDLE, for those lesions (currently labelled as cancers) and their precursors that are unlikely to cause harm if they are left untreated. Furthermore, precursors of cancer or high-risk disorders should not have the term cancer in them. The rationale for this change in approach is that indolent lesions with low malignant potential are common, and screening brings indolent lesions and their precursors to clinical attention, which leads to overdiagnosis and, if unrecognised, possible overtreatment. To minimise that potential, new strategies should be adopted to better define and manage IDLEs. Screening guidelines should be revised to lower the chance of detection of minimal-risk IDLEs and inconsequential cancers with the same energy traditionally used to increase the sensitivity of screening tests. Changing the terminology for some of the lesions currently referred to as cancer will allow physicians to shift medicolegal notions and perceived risk to reflect the evolving understanding of biology, be more judicious about when a biopsy should be done, and organise studies and registries that offer observation or less invasive approaches for indolent disease. Emphasis on avoidance of harm while assuring benefit will improve screening and treatment of patients and will be equally effective in the prevention of death from cancer.

Comparison of doxorubicin and cyclophosphamide versus single-agent paclitaxel as adjuvant therapy for breast cancer in women with 0 to 3 positive axillary nodes: CALGB 40101 (Alliance)

PURPOSE

Optimal adjuvant chemotherapy for early-stage breast cancer balances efficacy and toxicity. We sought to determine whether single-agent paclitaxel (T) was inferior to doxorubicin and cyclophosphamide (AC), when each was administered for four or six cycles of therapy, and whether it offered less toxicity.

PATIENTS AND METHODS

Patients with operable breast cancer with 0 to 3 positive nodes were enrolled onto the study to address the noninferiority of single-agent T to AC, defined as the one-sided 95% upper-bound CI (UCB) of hazard ratio (HR) of T versus AC less than 1.30 for the primary end point of relapse-free survival (RFS). As a 2 × 2 factorial design, duration of therapy was also addressed and was previously reported.

RESULTS

With 3,871 patients enrolled onto the trial, a median follow-up period of 6.1 years, and 437 RFS events, we achieved an HR of 1.26 (one sided 95% UCB, 1.48; favoring AC does not allow a conclusion of noninferiority of T with AC; UCB > 1.3). With 266 patient deaths, the HR for overall survival (OS) was 1.27 favoring AC (UCB, 1.56). The estimated absolute advantage of AC at 5 years is 3% for RFS (91 v 88%) and 1% for OS (95 v 94%). All nine treatment-related deaths were patients receiving AC and are included in the analyses of both RFS and OS. Hematologic toxicity was more common in patients treated with AC, and neuropathy was more common in patients treated with T.

CONCLUSION

This trial did not show noninferiority of T to AC, a conclusion that is unlikely to change with additional events and follow-up. T was less toxic than AC.

Predicting clinical trial results based on announcements of interim analyses

Background

Announcements of interim analyses of a clinical trial convey information about the results beyond the trial’s Data Safety Monitoring Board (DSMB). The amount of information conveyed may be minimal, but the fact that none of the trial’s stopping boundaries has been crossed implies that the experimental therapy is neither extremely effective nor hopeless. Predicting success of the ongoing trial is of interest to the trial’s sponsor, the medical community, pharmaceutical companies, and investors. We determine the probability of trial success by quantifying only the publicly available information from interim analyses of an ongoing trial. We illustrate our method in the context of the National Surgical Adjuvant Breast and Bowel (NSABP) trial, C-08.

Methods

We simulated trials based on the specifics of the NSABP C-08 protocol that were publicly available. We quantified the uncertainty around the treatment effect using prior weights for the various possibilities in light of other colon cancer studies and other studies of the investigational agent, bevacizumab. We considered alternative prior distributions.

Results

Subsequent to the trial’s third interim analysis, our predictive probabilities were: that the trial would eventually be successful, 48.0%; would stop for futility, 7.4%; and would continue to completion without statistical significance, 44.5%. The actual trial continued to completion without statistical significance.

Conclusions

Announcements of interim analyses provide information outside the DSMB’s sphere of confidentiality. This information is potentially helpful to clinical trial prognosticators. ‘Information leakage’ from standard interim analyses such as in NSABP C-08 is conventionally viewed as acceptable even though it may be quite revealing. Whether leakage from more aggressive types of adaptations is acceptable should be assessed at the design stage.

Body mass index and breast cancer risk according to postmenopausal estrogen-progestin use and hormone receptor status

To assess the joint relationships among body mass index, menopausal status, and breast cancer according to breast cancer subtype and estrogen-progestin medication use, we conducted a meta-analysis of 89 epidemiologic reports published in English during 1980-2012 identified through a systematic search of bibliographic databases. Pooled analysis yielded a summary risk ratio of 0.78 (95% confidence interval (CI): 0.67, 0.92) for hormone receptor-positive premenopausal breast cancer associated with obesity (body mass index (weight (kg)/height (m)(2)) ≥30 compared with <25). Obesity was associated with a summary risk ratio of 1.39 (95% CI: 1.14, 1.70) for receptor-positive postmenopausal breast cancer. For receptor-negative breast cancer, the summary risk ratios of 1.06 (95% CI: 0.70, 1.60) and 0.98 (95% CI: 0.78, 1.22) associated with obesity were null for both premenopausal and postmenopausal women, respectively. Elevated postmenopausal breast cancer risk ratios associated with obesity were limited to women who never took estrogen-progestin therapy, with risk ratios of 1.42 (95% CI: 1.30, 1.55) among never users and 1.18 (95% CI: 0.98, 1.42) among users; too few studies were available to examine this relationship according to receptor subtype. Future research is needed to confirm whether obesity is unrelated to receptor-negative breast cancer in populations of postmenopausal women with low prevalence of hormone medication use.

Bayesian hierarchical modeling of patient subpopulations: efficient designs of Phase II oncology clinical trials

BACKGROUND

In oncology, the treatment paradigm is shifting toward personalized medicine, where the goal is to match patients to the treatments most likely to deliver benefit. Treatment effects in various subpopulations may provide some information about treatment effects in other subpopulations.

PURPOSE

We compare different approaches to Phase II trial design where a new treatment is being investigated in several groups of patients. We compare considering each group in an independent trial to a single trial with hierarchical modeling of the patient groups.

METHODS

We assume four patient groups with different background response rates and simulate operating characteristics of three trial designs, Simon's Optimal Two-Stage design, a Bayesian adaptive design with frequent interim analyses, and a Bayesian adaptive design with frequent interim analyses and hierarchical modeling across patient groups.

RESULTS

Simon's designs are based on 10% Type I and Type II error rates. The independent Bayesian designs are tuned to have similar error rates, but may have a slightly smaller mean sample size due to more frequent interim analyses. Under the null, the mean sample size is 2-4 patients smaller. A hierarchical model across patient groups can provide additional power and a further reduction in mean sample size. Under the null, the addition of the hierarchical model decreases the mean sample size an additional 4-7 patients in each group. Under the alternative hypothesis, power is increased to at least 98% in all groups.

LIMITATIONS

Hierarchical borrowing can make finding a single group in which the treatment is promising, if there is only one, more difficult. In a scenario where the treatment is uninteresting in all but one group, power for that one group is reduced to 65%. When the drug appears promising in some groups and not in others, there is potential for borrowing to inflate the Type I error rate.

CONCLUSIONS

The Bayesian hierarchical design is more likely to correctly conclude efficacy or futility than the other two designs in many scenarios. The Bayesian hierarchical design is a strong design for addressing possibly differential effects in different groups.