Study Offers New Ways to Report AEs in Era of Molecularly Targeted Therapies

Traditional reporting of adverse events (AEs) may not be sufficient in the era of molecularly targeted therapies. In a new analysis, researchers suggested two alternative methods for AE reporting that incorporate the recurrent nature of the events, as well as a weighted measure of the time spent in a given health state.

For cytotoxic drugs, traditional methods of reporting allow “only to summarize the number of patients entering into a grade of toxicity at least once, and the time until first occurrence of the toxicity of interest,” wrote study authors led by Thomas Filleron, PhD, of the Institut Claudius Regaud in Toulouse, France. “These methods are too rudimentary to properly describe the severity, the duration, and the repetitions of toxicity for cancer treatment delivered along several years,” as many targeted therapies are.

The authors presented two new methods for AE reporting: prevalence and Quality-Adjusted Time Without Symptoms or Toxicity (Q-TWiST). Their methods and analyses of one phase II trial and a hypothetical simulated trial were published in Annals of Oncology.

The first measure, prevalence, involves calculating the probability of a patient being alive and relapse-free with the toxicity in question at a certain time. Q-TWiST, meanwhile, was initially developed to incorporate quality-of-life considerations into treatment comparisons, but the authors wrote that it can be used to evaluate the “trade-off between toxicity and survival.”

Calculating Q-TWiST involves the total time before disease progression spent with various toxicities, time spent before progression with no toxicities, and the time period spent following progression or relapse and ending with death or censoring from a trial. It is possible with this method to incorporate multiple toxicities at once, and to weigh toxicities in varying ways.

The authors applied these methods to one phase II study (SUPAP) that evaluated sunitinib as first-line therapy for papillary renal cell carcinoma, and to a simulated hypothetical randomized two-arm trial. In the SUPAP study, 25 of 46 patients presented with at least one mucositis/stomatitis episode before progression. Using the prevalence reporting method, the prevalence of AEs increased to 37.7% during the first month on treatment and then decreased over time, to 22.5% at 2 months, and 10.7% at 4 months.

In the large simulated trial, traditional reporting methods would have found no significant difference in AEs between the two groups. However, with the Q-TWiST method, it became clear that patients in one of the groups spent more time with an AE than the other group, and spent less time “in TWiST,” or before disease progression and with no toxicities.

“Patients treated with molecularly targeted therapies are first at risk of progression or death but they can also suffer recurrent toxicity of different types,” the authors wrote. “Future work needs to compare these different methodologies on randomized clinical trials data” to determine optimal ways of studying these newer types of drugs.