Abstract P6-07-09: Biomarker analysis from the neoadjuvant KRISTINE study in HER2-positive early breast cancer (EBC)

Introduction KRISTINE (NCT02131064) is an open-label, phase 3 study of neoadjuvant trastuzumab emtansine + pertuzumab (T-DM1+P) vs docetaxel + carboplatin + trastuzumab + pertuzumab (TCH+P) in pts with HER2+ EBC (all pts, Table). In exploratory analyses we assessed whether HER2 expression and gene amplification levels and PIK3CA mutation were associated with pathologic complete response (pCR) in KRISTINE.

Methods Treatment-naive pts with stage II–IIIC HER2+ EBC were randomized to receive 6 cycles of T-DM1+P or TCH+P and assessed for the primary endpoint, pCR (ypT0/is, ypN0). HER2 status was confirmed centrally as immunohistochemistry (IHC) status of IHC3+ and/or HER2/CEP17 gene ratio ≥2 by in situ hybridization (ISH) before study entry. Baseline HER2 mRNA, PIK3CA mutation, and hormone receptor (HR) status were evaluated. Rates of pCR were compared across biomarker subgroups, including: HER2 mRNA, HER2 IHC staining percentage, HER2 gene ratio, and PIK3CA mutation. Incidence of overlap between biomarkers was assessed. All analyses were descriptive.

Results KRISTINE randomized 444 pts (data cutoff, Dec 3, 2015; TCH+P=221; T-DM1+P=223). The biomarker population was representative of the ITT population. Biomarker values and distribution were balanced across treatment arms. Lower pCR rates were seen in both treatment arms across biomarker subgroups with lower vs higher HER2 expression levels (Table). PIK3CA mutation was associated with numerically lower pCR with T-DM1+P but not TCH+P. Evaluation of 15 pts who progressed during T-DM1+P neoadjuvant therapy suggests that nonresponse to therapy in this arm may be associated with low HER2 levels (mRNA, IHC, and ISH).

PIK3CA mutation rate was higher in tumors with ≤median HER2 mRNA (38%) vs >median (16%), and in focal (53%) and variable (41%) HER2 expression vs homogeneous (24%). A numerically higher proportion of HR positivity was seen in the HER2 IHC2+ (81%) subgroup vs IHC3+ (55%), and in the focal (83%) and variable (62%) HER2 subgroups vs homogeneous (53%).

Table TCH+PT-DM1+P  BiomarkernpCR (%)npCR (%)Difference in pCR rates (%)95% CIAll pts22155.722344.4-11.3-20.5, -2.0HER2      IHC3+19460.819549.7-11.1-20.9, -1.3IHC2+2520.0287.1-12.9-31.2, 5.5HER2 IHC 2+/3+ fraction*      Focal (<30%)1533.3160-33.3-57.2, -9.5Variable (30–79%)2729.62722.2-7.4-30.7, 15.9Homogeneous (≥80%)17961.518051.7-9.8-20.0, 0.4HER2 mRNA expression      ≤median10649.110937.6-11.4-24.6, 1.7>median10763.610851.9-11.7-24.8, 1.4HER2/CEP17 gene ratio      ≥2 to <44533.35219.2-14.1-31.6, 3.4≥416662.015853.2-8.9-19.6, 1.8PIK3CA mutation      Mutated5354.76131.1-23.6-41.3, -5.8Nonmutated16056.315151.0-5.3-16.3, 5.8*Categorization based on sum of IHC2+ and IHC3+ staining percentage.

Conclusions None of the evaluated biomarker subgroups showed superior pCR rates for T-DM1+P vs TCH+P. Lower HER2 levels based on mRNA and protein expression and gene amplification were associated with numerically lower pCR in both treatment arms. PIK3CA mutation results differ from prior observations. Unfavorable biomarkers for pCR, like low HER2, PIK3CA mutation, and HR positivity, showed a tendency for co-expression, which should be considered when planning future trials in HER2+ BC.

Citation Format: de Haas SL, Hurvitz SA, Martin M, Kiermaier A, Lewis Phillips G, Xu J, Helms H-J, Slamon D, Press MF. Biomarker analysis from the neoadjuvant KRISTINE study in HER2-positive early breast cancer (EBC) [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P6-07-09.

Point and Interval Estimators of the Target Dose in Clinical Dose-Finding Studies with Active Control

In a clinical dose finding study with active control a new drug with several dose levels is compared with an active comparator drug. The main focus of such studies often lies on the estimation of a target dose that leads to the same efficacy as the control. This article investigates the finite sample properties of the maximum likelihood estimation of the target dose and compares several approaches for constructing corresponding confidence intervals under the assumption of a linear dose-response curve and normal error terms. Furthermore, the impact of deviations from the model assumptions regarding the error distribution is explored.

Discrepancies between medical oncologists and surgeons in assessment of resectability and indication for chemotherapy in patients with colorectal liver metastases

BACKGROUND

Multidisciplinary discussion of the treatment of patients with colorectal liver metastases (CRLM) is advocated currently. The aim of this study was to investigate medical oncologists' and surgeons' assessment of resectability and indication for chemotherapy, and the effect of an educational intervention on such assessment.

METHODS

Medical histories of 30 patients with CRLM were presented to ten experienced medical oncologists and 11 surgeons at an initial virtual tumour board meeting (TB1). Treatment recommendations were obtained from each participant by voting for standardized answers. Following lectures on the potential of chemotherapy and surgery, assessment was repeated at a second virtual tumour board meeting (TB2), using the same patients and participants.

RESULTS

Overall, 630 answers (21 × 30) were obtained per tumour board meeting. At TB1, resectability was expected more frequently by surgeons. Participants changed 56·8 per cent of their individual answers at TB2. Assessment shifted from potentially resectable to resectable CRLM in 81 of 161 and from unresectable to (potentially) resectable CRLM in 29 of 36 answers. Preoperative chemotherapy was indicated more often by medical oncologists, and overall was included in 260 answers (41·3 per cent) at TB1, compared with only 171 answers (27·1 per cent) at TB2. Medical oncologists more often changed their decision to primary resection in resectable patients (P = 0·006). Postoperative chemotherapy was included in 51·9 and 52·4 per cent of all answers at TB1 and TB2 respectively, with no difference in changes between medical oncologists and surgeons (P = 0·980).

CONCLUSION

Resectability and indication for preoperative chemotherapy were assessed differently by medical oncologists and surgeons. The educational intervention resulted in more patients deemed resectable by both oncologists and surgeons, and less frequent indication for chemotherapy.

An appraisal of the current and potential value of web 2.0 contributions to continuing education in oral implantology

OBJECTIVE

To systematically assess the informational value, quality, intention, source and bias of web 2.0 footage whose aim is peer-to-peer education about oral implantology.

METHODS

YouTube (http://www.youtube.com) was scanned on 15 October 2010 for oral implantology-related videos using an adequately pre-defined search query. Search results were filtered with the system-generated category 'education' and the additional criterion 'most viewed'. Only those videos with at least 1000 views were included (total 124, of which 27 were excluded because they were not related to implantology). Filtered videos were discussed and rated with particular regard to the educational needs of potential groups of addressees [(i) undergraduates and laymen, (ii) dentists without or currently undergoing a specialisation in oral implantology and (iii) dentists who have completed a specialisation in the field of oral implantology] by a jury consisting of (i) an accredited post-graduate university instructor with 22 years of professional teaching experience in the field of implantology, (ii) a university lecturer in dentistry/orthodontics with 10 years teaching experience and (iii) a university haematologist/oncologist. They were required to fill out a questionnaire for each video. The data were statistically analysed using non-parametric ANOVA (α = 5%) and a sign test (α = 0.05/3 = 0.017).

RESULTS

The YouTube scan produced 1710 results in the category 'EDU'. The analysis revealed that there is a wide range of instructional footage on this topic, but with highly variable range in quality and informational value. Footage intention was to large proportions (47.4%) a mixture of education and advertisement. Its usefulness differed significantly for the three groups of addressees, offering greater novelty to undergraduates and post-graduates.

CONCLUSION

YouTube and similar social media websites may have a potential capacity and value in complementing continuing education in the technique of oral implantology. As a means of achieving an acceptable level of knowledge about the topic when used alone, it should not be considered to be suitable at this point in time.