{"entity":{"id":"paper-shiva-off-label-targeted-therapy-lancet-oncol-2015","kind":"paper","name":"SHIVA: molecularly targeted therapy based on tumour molecular profiling versus conventional therapy for advanced cancer","aka":[],"tldr":"The one randomised test of giving people a targeted drug matched to their tumour's mutation but outside its licence found they did no better than with their doctor's choice of treatment.","summary":"SHIVA was an open-label randomised phase 2 trial at eight French academic centres. Adults with any metastatic solid tumour that had progressed on standard treatment had a mandatory biopsy of a metastasis and large-scale genomic testing. Those with an alteration in one of three pathways (hormone receptor, PI3K/AKT/mTOR, RAF/MEK) that could be matched to one of ten regimens of 11 marketed targeted drugs were randomised 1:1 to the matched drug, used outside its indication, or to treatment at the physician's choice.\n\nOf 741 patients screened between October 2012 and July 2014, 293 (40 per cent) had a matchable alteration and 195 (26 per cent) were randomised, 99 to matched treatment and 96 to control. Median progression-free survival was 2.3 months with matched treatment and 2.0 months with physician's choice (hazard ratio 0.88, 95 per cent confidence interval 0.65 to 1.19, p 0.41). Grade 3 to 4 adverse events occurred in 43 per cent of patients on targeted drugs and 35 per cent on chemotherapy. The authors concluded that off-label use of targeted drugs should be discouraged and enrolment in trials encouraged.","asOf":"2026-10-20","links":[{"label":"Le Tourneau et al., Lancet Oncol 2015","url":"https://doi.org/10.1016/S1470-2045(15)00188-6"},{"label":"PubMed","url":"https://pubmed.ncbi.nlm.nih.gov/26342236/"},{"label":"ClinicalTrials.gov NCT01771458","url":"https://clinicaltrials.gov/study/NCT01771458"}],"tags":["precision","failure"],"related":[],"cancers":[],"sections":["diagnostics"],"technologies":["cgp"],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":["christophe-le-tourneau"],"bottlenecks":[],"keyPapers":[],"journals":["lancet-oncology"],"dependsOn":[],"notes":[],"journal":"The Lancet Oncology","year":2015,"doi":"10.1016/S1470-2045(15)00188-6","pmid":"26342236","authors":"Le Tourneau C, Delord JP, Gonçalves A, et al.","paperType":"rct","findings":["293 of 741 screened patients (40 per cent) had an alteration matching one of the ten regimens; 195 (26 per cent) were randomised.","Median progression-free survival 2.3 months with matched off-label treatment against 2.0 months with physician's choice (hazard ratio 0.88, 95 per cent confidence interval 0.65 to 1.19, p 0.41).","Grade 3 to 4 adverse events in 43 of 100 patients on targeted drugs and 32 of 91 on cytotoxic chemotherapy (p 0.30)."],"whatItMeans":"A mutation found on a profiling report is a reason to look for a trial, not proof that a drug licensed for another cancer will work. Matching to drugs approved for that alteration in that cancer is a separate question that SHIVA did not test.","caveats":["Patients were heavily pretreated, and the drugs were the targeted agents marketed in France in 2012.","Progression-free survival was not assessed by independent central review, and clinicians and patients knew the allocation.","Three pathways and ten regimens cover a fraction of the alterations matched today."],"changedPractice":false,"participants":195},"route":"/key-papers/paper-shiva-off-label-targeted-therapy-lancet-oncol-2015/","neighbours":{"section":[{"id":"diagnostics","kind":"section","name":"Diagnostics & Biomarkers","route":"/fronts/diagnostics/"}],"technology":[{"id":"cgp","kind":"technology","name":"Comprehensive genomic profiling","route":"/technologies/cgp/"}],"person":[{"id":"christophe-le-tourneau","kind":"person","name":"Christophe Le Tourneau","route":"/people/christophe-le-tourneau/"}],"journal":[{"id":"lancet-oncology","kind":"journal","name":"The Lancet Oncology","route":"/journals/lancet-oncology/"}]}}