Brain metastases are a frequent, life-threatening, often debilitating and difficult to treat complication of HER2-positive breast cancer. Despite the frequency of brain metastases, relatively little research, both on their pathobiology and treatment strategies, has been conducted. The design of clinical trials on brain metastases is particularly challenging, as many patient and tumour variables (e.g. number and location of brain metastases, status of primary tumour and extracranial metastases, previous therapies, etc.) have to be taken into account. Endpoint definition needs special consideration as brain-focused endpoints (e.g. time to progression in the brain) or more general endpoints (e.g. overall survival) have to be selected depending on the trial context. For many studies, cognitive and quality of life measures may be of interest and should be included whenever possible. Furthermore, practical and logistic issues complicate brain metastases trials and well organised interdisciplinary networks are necessary for successful patient recruitment and study conduct.
Given these difficulties, Dr. Bachelot and his colleagues must be congratulated for being able to run and complete a prospective trial on a defined subgroup of patients with brain metastases. The trial builds on previous data from small studies indicating that the combination of capecitabine (a cytotoxic agent) and lapatinib (a dual kinase inhibitor of HER1 and HER2) might be active in patients with brain metastases of HER2-positive breast, particularly in those not previously treated with radiotherapy to the brain. If confirmed, these results would be highly relevant for patient treatment, as whole brain radiotherapy is associated with considerable neurotoxicity and upfront medical treatment could offer better preservation of neurocognitive function. Bachelot et al conducted an open-label, single-arm, multicentre phase 2 study and recruited 45 patients form 11 centres in France. Capecitabine was administered at 2000mg/m2 from day 1 to day 14 of every 21 day cycle and lapatinib was given daily at 1250mg. A Simon´s optimal two-stage design was used and the objective CNS response rate was the primary endpoint. This endpoint was met in that partial CNS response was seen in 29 (65.9%) of patients. Median time to CNS progression was 5.5 months (4.5 to 6.1) and median overall survival was 17 months (13.7 to 24.9).
There are several issues that need to be recognised with this study and that should be considered for the design of any follow-up trials. Unfortunately, no neurocognitive and quality of life data were reported and a comparison of such measures in patients treated with capacitabine/lapatinib and patients treated with whole-brain radiotherapy in the context of a randomised trial would be highly interesting. Another point of concern is treatment toxicity. In this trial, almost half of patients had at least one grade III or IV adverse event (most commonly diarrhoea or hand-foot syndrome) and dose reductions of lapatinib were necessary in 36% of all patients and of capecitabine in 58% of cases. Again, a randomised trial is urgently needed to compare the toxicity profile and its implications for quality of life, treatment compliance and outcome with radiotherapy. One interesting issue relates to patient inclusion and baseline parameters. It is surprising to note that more than 40% of all patients were neurologically asymptomatic at baseline. Radiological screening is usually not recommended practice in breast cancer patients and therefore the typical patient diagnosed with brain metastases in the clinical setting is symptomatic. It seems some sort of screening was performed in patients included in the LANDSCAPE trial and therefore the results of the study cannot directly be extrapolated to most patients encountered in clinical practice. The question of whether or not to screen for brain metastases in breast cancer patients is yet to be answered by adequate studies.
In any case, the LANDSCAPE trial is a valuable addition to the sparse literature on brain metastases and will hopefully motivate and facilitate more research groups to accept the challenge of conducting brain metastasis trials in breast cancer, but also in other tumour types with frequent CNS involvement such as lung cancer and melanoma. A randomised trial of lapatinib/capecitabin versus whole-brain radiotherapy in patients with brain-metastatic HER2-positive breast cancer is warranted and needed to validate the findings of the LANDSCAPE trial.