AMSTERDAM – Two-fifths of patients with refractory recurrent or metastatic cancers were successfully matched to an appropriate targeted treatment using molecular profiling, according to research presented at the multidisciplinary European cancer congresses.
Feasibility findings from the phase II, proof-of-concept SHIVA study showed that a molecular abnormality leading to targeted treatment was present in 40% patients.
Dr. Christophe Le Tourneau
This is the first randomized trial to consider treating patients purely on the basis of their molecular profile rather than the type of tumor that they have. It is also the first trial designed to determine what the outcome of such an approach will be, with the primary endpoint being progression-free survival.
Currently, the prescription of approved molecularly targeted agents relies on the primary tumor location and histological subtype. For example, a woman with breast cancer overexpressing HER2 may be treated with a HER2-targeting agent such as trastuzumab or lapatinib, or with endocrine therapies such as tamoxifen and letrozole, if there is overexpression of the estrogen or progestogen receptor (ER/PR).
However, it is not clear if this is the best means of using these therapies, as their molecular targets may not be specific to a single tumor type. Indeed, HER2 is overexpressed in several tumor types other than breast cancer, including ovarian, gastric, colorectal, pancreatic, and endometrial cancers.
The idea of giving patients treatment based on their molecular profiles rather than their tumor types represents a new paradigm and requires prospective validation before being implemented in clinical practice, Dr. Christophe Le Tourneau of Institut Curie, Paris, said in an interview.
The SHIVA trial was therefore initiated to compare molecularly targeted therapy based on tumor molecular profiling versus conventional therapy in patients with refractory cancer (Target Oncol. 2012;7:253-65). Although two previous studies have suggested that this approach is possible and could improve patient outcomes (J. Clin. Oncol. 2010;4877-83; Clin. Cancer Res. 2012;18:6373-83), these studies were not randomized against the standard of care and so the conclusions that can be drawn from their results are not robust.
Patients are eligible for inclusion in the ongoing SHIVA trial if they have recurrent or metastatic disease despite standard treatment, an ECOG performance status of 0 or 1, adequate organ function, and measurable disease that can be biopsied not involving the bone or brain.
After recruitment, all patients will have a biopsy taken of one metastasis that is then assessed for gene mutations and amplifications, and for the expression of hormone (estrogen, progestogen and androgen) receptors. If a molecular abnormality is identified for which an approved targeted agent is available, then patients are randomized to receive either the matching targeted therapy or the conventional therapy that they would have received had they not be identified as having a possible molecular target. At progression, patients in the conventional treatment arm may crossover to targeted therapy.
Patients are only randomized if they have a molecular target that is not already known for their disease. So a patient with breast cancer found to have amplified HER2 or a patient with non–small-cell lung cancer and overexpression of EGFR would not be included, Dr. Le Tourneau said.
The SHIVA investigators have developed an algorithm or Molecular Biology Board (MBB) to help guide treatment in the experimental arm. The following targeted agents may be used if an appropriate molecular target is found: KIT or ABL for imatinib; PTEN or mTOR, among others, for everolimus; BRAF for vemurafenib; PDGFRA/B or FLT-3 for sorafenib; EGFR for erlotinib; HER2 for lapatinib in combination with trastuzumab; SRC, and others, for dasatinib; ER and PR for tamoxifen or letrozole; and the androgen receptor for abiraterone.
The trial is using high throughput sequencing, so processing the biopsies involves only two assays – one for gene mutation analysis and one for gene copy number alterations – "so we look at everything at the same time," Dr. Le Tourneau said.
As of September 2013, 350 patients had been included in the study and feasibility of the molecular profiling approach was tested in the first 95 patients. Of these, a complete molecular profile could be obtained in 61%. The median time to obtaining the biopsy sample and performing the MBB was 26 days, ranging from 2 weeks to 42 days. Mutations could be assessed in two-thirds of samples, with gene copy alterations analyzed in 68%, and immunohistochemistry for hormone receptors possible in 92%.
In the experimental arm, 14 patients (15%) with a variety of tumor types are now being treated with abiraterone, 13 (14%) with everolimus, nine (9%) with endocrine treatment, and two (2%) with anti-HER2 treatment or sorafenib.