Imaging and Clinical End Points in Brain Metastases Trials

Abstract

When designing a clinical trial on brain metastases the choice of end points is critical. The Response Assessment in Neuro-Oncology (RANO) International Group has recently critically reviewed the different end points used in the clinical trials on brain metastases [1,2].Some factors heavily influence the choice of end points. First, patients with solid tumors may differ in prognosis and competing risk of extracranial progression. Second, the trial setting (Phase II vs Phase III trials) and type of intervention (CNS-directed vs systemic therapies) also require different end points.The knowledge of prognostic factors, in other words, of those factors influencing the outcome regardless of treatments, is critical in order to identify subgroups of patients with different outcomes (e.g., Recursive Partitioning Analysis [RPA] and Graded Prognostic Assessment [GPA] classes) [3][4][5]. In clinical trials, these subgroups are essential either as a stratification factor or inclusion criteria or even for post hoc analyses.There are key issues in imaging brain metastases in clinical trials: modalities and frequency of assessment, type and magnitude of change used to define response or progression of disease, and incorporation of steroid use and neurological symptoms/signs with current imaging definitions of response and progression. Last but not least, the ability to differentiate tumor-related and treatment-related changes is another critical point.The objective response is used to screen the activity of novel compounds and represents a primary end point in Phase II trials. It may be a surrogate for other markers of clinical benefit, such as neurological symptoms, neurocognitive function or survival. The interlesional variability of response to treatment can render problematic the evaluation of response in multiple metastases. A methodological open issue is whether it is preferable a randomized Phase II trial or a single arm Phase II trial with historical control in large datasets (e.g., Radiaition Therapy Oncology Group [RTOG] datasets) when larger Phase III trials are not feasible.The criteria of response used in brain metastasis trials in the past were quite heterogeneous. The use of MRI for response assessment was not always a standard procedure. There were major areas of difference across trials on brain metastases regarding the evaluation of response to treatment: definition of a target lesion (not defined, ≥1 cm); number of lesions (variable); type of measurement (unidimensional, bidimensional, volumetric); degree of tumor shrinkage required for response (≥30%, ≥50%); requirement for confirmatory scans (more commonly not required); inclusion of steroids and neurological symptoms (more commonly not done); evaluation of extracranial disease (more commonly not included). Overall, none of the Standard Response Criteria (Response Evaluation Criteria In Solid Tumors [RECIST], WHO, MacDonald and RANO for high-grade gliomas) were designed specifically to evaluate brain metastases, thus investigators have not consistently ch...

Prognostic analysis of patients with non-small cell lung cancer harboring exon 19 or 21 mutation in the epidermal growth factor gene and brain metastases

“…Overall survival after BM diagnosis (mOS) was the time from the diagnosis of BM to death or last follow-up. We defined intracranial progression-free survival (iPFS) as the interval between the diagnosis of BM and intracranial progression or mortality from any cause [8,9]. The therapeutic effects were evaluated at 3 months using the RECIST criteria [10].…”

Neratinib: the emergence of a new player in the management of HER2+ breast cancer brain metastasis

“…Note that while RECIST 1.1 criteria considers partial response to be an at least 30% reduction in the sum of diameters in target lesions [31], a 30% unidimensional reduction in a perfect sphere would correspond to a 65% volumetric reduction. Across multiple studies, however, a 20% volumetric reduction is considered clinically meaningful [32,33]. An extension of median overall survival (OS) to 27.9 months from time of brain metastasis development in patients treated with lapatinib and capecitabine was separately reported, as compared with 16.7 months in patients continued on trastuzumab based therapies [34].…”

An update on the epidemiology, imaging and therapy of brain metastases

“…However, the molecular and histological features of the tumors have a big impact on survival rates. For instance, patients with breast cancer brain metastases and positive HER2 status can survive over five years with multimodal therapy and good control of the systemic disease [41]. In melanoma or NSCLC, the median survival is poor, and even young patients with good life quality generally survive less than a year [35].…”

“…In melanoma or NSCLC, the median survival is poor, and even young patients with good life quality generally survive less than a year [35]. Management should consider tumor histology where possible, available elective treatment, the patient`s age, Karnofsky performance status, the volume of brain metastases, and extracranial disease activity [41]. The main treatment methods include surgery, whole brain radiotherapy (WBRT), stereotactic radiosurgery, and chemotherapy.…”

“…Whole brain radiotherapy alone has been compared to surgery plus WBRT for the treatment of newly-diagnosed brain metastases [42,43], with good quality evidence of the benefits from the application of the combined regimen (median survival 4-6 and 10 months, respectively). Furthermore, disease-free survival and CNS recurrence rates are improved with multimodal treatment [41].…”