PracticeUpdate: Haematology & Oncology

CONFERENCE COVERAGE 26

Low level of preexisting stromal cytotoxic-cell infiltration predicts womenwho will benefit most from trastuzumab over lapatinib A low level of preexisting stromal cyto- toxic T-cell infiltration has been shown to predict women who would benefit

This differential effect was confirmed in multivariate analysis (interaction test P = 0.042). The other tested biomarkers did not demonstrate significant predictive effects. Dr Nielsen concluded that, in this correlative study of metastatic HER2-positive breast cancer, a low level of pre-existing stromal cytotoxic T-cell infiltration predicted women with metastatic breast cancer who would benefit most from trastuzumab over lapatinib. Overall tumour infiltrating lymphocyte counts were neither prognostic nor predictive. He continued, “The answer to the question we posed, whether the superiority of anti- body-based therapy over non-antibody-based therapy aimed at HER2 could be predicted by the tumour’s immune status, was yes. But this is only the case for the CD8 marker and not for the eight other factors we examined. Breast cancers without CD8-positive tumour infiltrating lymphocytes responded better to trastuzumab than to lapatinib. This may have been because trastuzumab woke up an immune response that helps improve survival in metastatic patients. Lapatinib cannot do this. In patients who have demonstrated a CD8 immune response, the two agents work almost as well.” “This study was performed as part of a phase 3 randomised trial, using formal prespecified evidence and independent statistical analysis. Nevertheless, for the results to be definitive, results need to be repeated in a second, similar study.”

(95% CI 1.13–1.65). Though both agents block HER2 signalling, trastuzumab possesses additional mecha- nisms of action via the immune system. Dr Nielsen and colleagues hypothesised that tumour-infiltrating lymphocyte levels may predict response to HER2-targeted therapy (trastuzumab vs lapatinib). MA.31 included patients with HER2-posi- tive metastatic breast cancer, median age 55 years. Median follow-up was 21.5 months. Overall tumour-infiltrating lymphocytes were counted in the original haematoxylin and eosin-stained sections taken at study entry. Immunohistochemistry was performed on unstained sections from tissue microarrays or individual formalin-fixed paraffin-embedded blocks to test expression of lymphocyte bio- markers CD8, FOXP3, CD56, and PD-1 on stromal and intratumoural tumour-infiltrating lymphocytes. Statistical analysis was conducted by the Canada Clinical Trials Group for a total of nine prespecified biomarker tests. Associations of tumour-infiltrating lymphocytes with progression-free survival were evaluated by univariate stratified log-rank test with graphic Kaplan-Meier curves, and by stratified multivariate Cox proportional hazards regression analysis. Predictive effect was examined with a test of interaction between treatment allocation and biomarker classification (high vs low, using pre-established cut points). Of 652 cases, slides were available for overall tumour-infiltrating lymphocyte assessment in 614 and immunohistochemical biomarker assessment in 427. In this correlative study set, the superiority of trastuzumab over lapatinib for progression-free survival was confirmed in multivariate analysis (lapatinib/ taxane vs trastuzumab/taxane/lapatinib: HR 2.55, 95% CI 1.43–4.55, P = 0.001). Tumour-infiltrating lymphocyte counts by haematoxylin and eosin staining were nei- ther prognostic nor predictive in this set of metastatic HER2-positive breast cancers. Lymphocyte immunohistochemical markers were not prognostic. Prespecified stratified univariate analysis, however, detected a significantly higher risk for lapatinib over trastuzumab (HR 2.94, 95% CI 1.40–6.17, P = 0.003) in patients with low CD8-positive stromal tumour infiltrating lymphocytes (fewer than three) than observed among those with high CD8-positive stromal tumour infiltrating lymphocytes (HR1.36, 95% CI 1.05–1.75, P = 0.019).

most from trastuzumab over lapatinib. Torsten O. Nielsen, MD, PhD, of the Univer- sity of British Columbia, Vancouver, Canada, explained that the presence of tumour infil- trating lymphocytes, particularly CD8-pos- itive cytotoxic T-cells, has been associated with improved prognosis in patients with HER2-positive breast cancer.

This may have been because trastuzumab woke up an immune response that helps improve survival in metastatic patients. Lapatinib cannot do this.

In patients who have demonstrated a CD8

immune response, the two agents work almost as well.

Increasing levels of tumour-infiltrating lymphocytes also appear to predict a response to adjuvant trastuzumab in early breast cancer, though they did not predict benefit of combined trastuzumab-lapatinib neoadjuvant dual therapy over monotherapy in Neo Adjuvant Lapatinib And/or Trastuzumab treatment Optimisation study (NeoAALTO). “We performed this study,” Dr Nielsen said, “to determine whether the evident superiority of an antibody-based therapy (which can in- duce or augment an immune response), over a nonantibody drug aimed at the same target (HER2), could be predicted by the tumour’s immune status as assessed by the presence of tumour infiltrating lymphocytes in general, and/ or of specific populations of lymphocytes (for example, cytotoxic T-cells, as marked by CD8).” National Cancer Institute of Canada Clinical Trials Group MA.31 randomised 652 wom- en with HER2-positive metastatic breast cancer to trastuzumab or lapatinib, in com- bination with taxane chemotherapy for 24 weeks, followed by the same HER2-targeted monotherapy. Final results fromMA.31 found trastuzumab was superior to lapatinib for the primary end- point of progression-free survival: the hazard ratio for lapatinib to trastuzumab was 1.37

PracticeUpdate Editorial Team

PRACTICEUPDATE HAEMATOLOGY & ONCOLOGY