This year’s annual European Society for Medical Oncology (ESMO) 2019 Congress took place from 27-30 September. Researchers, clinicians, cancer nurses, industry partners and patient advocates from all over the world gathered down in Barcelona, Spain. With the tagline ‘Translating science into better cancer patient care’, ESMO created the perfect place to exchange knowledge and expertise on the latest advances in oncology.Many practice-changing data were presented, and are summarized in this blog for you.
Continued advances with PD(L)1 inhibitors
Several presentations showed how PD(L)1 inhibitors keep on improving clinical outcomes for several tumor types. One example was the Impower110 trial presented by Dr. David Spigel of the Sarah Cannon Research Institute in Tennessee, analyzing atezolizumab vs platinum-based chemotherapy in first-line treatment of PD-L1 selected NSCLC. The interim results showed improvement in overall survival (OS) at 6 months 76.3% (68.2-84.4) vs 70.1% (60.8-79.4) as well as at 12 months 64.9% (55.4-74.4) vs 50.6%, with a HR of 0.59 (0.40-0.89) in the group investigating Roche’s PD-L1 inhibitor. These results are similar to the results of the KEYNOTE-189, in which the addition of pembrolizumab to chemotherapy in first-line metastatic NSCLC. These results confirm that combining PD-L1 inhibition with platinum-based chemotherapy will become an important part of clinical practice for NSCLC.
Another interesting example was the ATTRACTION-3 study presented by B.C. Cho from the Yonsei University, Seoul, of which results were actually published in The Lancet Oncology during the conference. Here, nivolumab was compared to chemotherapy in advanced oesophageal squamous cell carcinoma. The trial showed significant improvement in OS for patients receiving world’s first-ever approved PD1 inhibitor vs chemotherapy: 10.9 months vs 8.4 months with a HR of 0.77 (0.62-0.96).
PARP inhibitors improve PFS in ovarian cancer
Beyond IO, PARP inhibitors were a key focus. A challenge with PARP inhibition is the associated toxicity, which has limited efforts to combine treatment with regular chemotherapy. Veliparib is a milder PARP inhibitor compared to the four that have already been EMA and/or FDA approved (olaparib, rucaparib, niraparib, talazoparib). In the VELIA trial presented by R. Coleman from the M.D Anderson Cancer Center, veliparib was tested in 3 combinations: platinum chemo plus veliparib as induction and maintenance, platinum plus veliparib as induction with placebo as maintenance, and platinum plus placebo as induction and maintenance. The trial showed that the addition of veliparib as induction and continued as maintenance significantly extended progression-free survival (PFS) across the entire population compared to chemotherapy plus placebo (median PFS 23.5 months vs 17.3 months, HR 0.68 (0.56-0.83)). Invited as the discussant, Dr. M.R. Mirza from Copenhagen University Hospital noted that the addition of veliparib caused increased toxicity, reduced use of potentially curative chemotherapy and did not provide a benefit over what is observed in other PARP first-line trials. According to him, the absence of a comparator arm in maintenance phase made it also unclear if the addition of veliparib to induction chemotherapy is necessary, or whether a PARP inhibitor during maintenance as monotherapy is sufficient.
Interestingly, A. Gonzales Martin from the Clinica Universidad de Navarra presented that niraparib showed activity regardless of BRCA status when compared to placebo in the PRIMA study. PFS in the overall population treated was 13.8 months vs 8.2 months with niraparib and placebo retrospectively, by a HR of 0.62 (0.50-0.76). Also the PAOLA study, in which bevacizumab was combined with olaparib vs bevacizumab plus placebo after first-line platinum based chemotherapy, showed significantly improved PFS (median 22.1 months vs 16.6 months, HR 0.59 (0.49-0.72)), regardless of BRCA mutation status. They also assessed if there was benefit based upon homologous recombination deficiency (HRD) positive or negative tumor; an important therapeutic target with shown efficacy for PARP inhibitors and platinum analogues. Less benefit was seen in non-BRCA mutated tumors, but no substantial benefit was shown in HRD- disease or unknown status (16.9 vs 16.0 months, HR 0.92 (0.72-1.17)). These results in HRD+ without BRCA mutation identify a new population in ovarian cancer, which significantly benefits from treatment with olaparib plus bevacizumab and niraparib.
An interesting session was on the evaluable results of precision medicine trials, which focused on umbrella and basket trials. The latter are seen as a well-established methodology to explore a single targeted therapy in multiple diseases, which are relevant for biomarker defined diseases, or subsets of cancer sharing a common molecular target. According to one of the presenters, Professor P. Schöffski from UZ Leuven, these studies are however very demanding, as they require a high level of trial coordination, logistics, and costs, while also being challenging from an intellectual point of view. Results from the EORTC 90101 CREATE study showed that counterintuitive results can be attained in biomarker negative participants (an ALK negative patient responding to treatment with crizotinib). Although these studies tend to be signal seeking, it can have major regulatory implications in the era of accelerated approvals. An example of such is the tissue independent approval for pembrolizumab. Professor T. Maughan of the University of Oxford presented the various challenges of umbrella trials, in which 1 type of cancer is investigated with various drugs for different genetic mutations – where precisely in the disease pathway should the treatment target be? Should patients be randomized with various drugs for that particular pathway? Should the genetic aberrational test be done in an unselected or in a selected population? Do you test archival or contemporaneous tissue? How is the supportive data for the predictive marker for the specific targeted therapy? Is the biomarker assay reproducible in real time? How should we take into account tumor heterogeneity and evolution?
Related to achieving personalized cancer medicine, the Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination (BATTLE)-1 trial was the first completed, prospective biopsy-mandated, biomarker-based, adaptive randomized clinical trial for advanced NSCLC. Data of this umbrella BATTLE-1 trial showed that NSCLC patients with an EGFR mutation had a better disease control rate (DCR), and that those with KRAS/BRAF mutations benefited from treatment with sorafenib (with an 8 week disease control rate of 79%, compared to 14% with erlotinib). However this hypothesis was disproven in multiple latter studies (including a Dutch study, in which a 8 week DCR of 52.6% was found and median PFS of 2.3 months). In the end, Professor Maughan concluded that there needs to be rigorous justification that the proposed target is disease modifying and not just an “actionable” mutation.
Professor C. Le Tourneau from the Institut Curie Paris & St. Cloud, presented an example of a randomized precision medicine trial using an algorithm, the SHIVA01 trial (of which he was an investigator). The SHIVA01 trial was done in a large variety of tumor types, molecular alterations and drugs. This trial showed no benefit of a molecularly targeted agent vs treatment at physicians’ choice. There are individual patients with success stories which drive the promise of precision medicine. Ultimately, one should take into account that there is clinical benefit in less than 5% of patients (as described by Tannock et al, NEJM 2016)
Kidney function and drug-drug interaction
One important presentation by V. Launay-Vacher of the Service ICAR, Groupe Hospitalier Pitié Salpetriere in Paris concerned drug-drug interaction and kidney function. At presentation, between 37% of patients have a reduced glomerular filtration rate (GFR), which is below 60 ml/min. Reduced GFR is associated with increased cancer mortality, due to over dosage (impaired renal excretion of the anticancer drugs and/or metabolites) with resulting dose depended toxicities. However, adjustment of the anticancer drug to renal function resulted in similar cancer mortality. Dr. Launay-Vacher, and Dr. B. Sprangers of the UZ Leuven, indicated during the discussion that the cut-off of 60 ml/min is arbitrary. When writing protocols, kidney function should not be assessed using a creatinine value, but using clearance, and the best test for clearance would be the CKD-EPI collaboration equation, instead of the Cockcroft-Gault method. The presenters also indicated that, especially for trials in which the drug is not cleared via the kidney, having less stringent kidney function criteria would be beneficial (up to a clearance of > 30 ml/min).
Drug-drug interactions are important to consider, but often not given the necessary attention in daily clinical practice. Equally, patients buy over the counter alternative drugs, herbs and spices which can also have an interaction with other drugs. For instance, a case was discussed in which the usage of formulated curcuma pills in combination with everolimus decreased the level of everolimus to below the therapeutic range in a patient. Stopping with the curcuma pills increased the everolimus level to therapeutic levels. Garlic, aloe vera, ginseng, grapefruit, licorice, medicinal sage and soy are all known examples of plants/herbs which affect either the CYP3A4 or CYP2D6 enzymes and subsequently have potential for drug-drug interactions. This stresses the importance of capturing information on alternative herbs and spices, especially for treatments with cytochrome interactions.
CDK4/6 inhibition in HR+ breast cancer
CDK4/6 inhibition were the topic of the presidential symposium, in which the results from the MONARCH 2 trial and the MONALEESA-3 trial were presented. The MONARCH 2 trial presented by G. Sledge of the Indiana University Simon Cancer Center in Indianapolis showed an OS benefit of 9.4 months with pretreatment of abemaciclib added to fulvestrant compared to fulvestrant alone in pre- and post-menopausal patients with hormone receptor (HR) positive tumors (46.7 months vs 37.3 months respectively). The MONALEESA-3 trial, presented by D. Slamon of the UCLA / David Geffen School of Medicine, Los Angeles, compared ribociclib added to fulvestrant vs fulvestrant alone as first- or second-line treatment in post-menopausal women showed an obvious benefit (HR 0.724, CI 0.568-0.924) both in first-line and in patients with early relapse or second-line treatment. Based on these findings, the three approved CDK4/6 inhibitors to date (palbociclib, abemaciclib and ribociclib) seem to benefit patients when added to endocrine therapy and probably will become part of standard of care for breast cancer. The right inhibitor and optimal treatment sequencing needs further investigation.
Patient advocates were also present and stressed the various needs of patients: proper referral to specialized centers, proper testing for genetic material with possible therapeutic implications, and the need to solve problems with performing (due to reimbursement) of diagnostic tests. Patient advocacy groups can be seen as a partner during clinical trials for their specific diseases, especially for recruiting patients with rare subtypes of common cancers. These groups are well aware of the specific trials that are running for which they are eligible, are part of large Facebook groups in which they inform each other of disease specific relevant news, and can aid with referring patients to the proper hospitals and to the right trials.
Cancer is becoming more and more a chronic disease. The various types and genetic subtypes necessitate specialized treatment. The future is looking bright. Combining big data consisting of demographics, genetic aberrational screening and treatment outcome into an artificial intelligence system will become a big part of routine cancer care. At the same time, it will also be important to select the right target for the right drug as soon as feasible during your clinical trial.