In this issue:
Measures & Methods
PRO-CTCAE: Enhancing Reporting of Symptomatic Adverse Events
Treatment tolerability is an important endpoint when making conclusions about therapeutic effectiveness, and as such, accurately gauging the toxicity profile of new agents is essential to oncology drug development. Treatment toxicity in cancer clinical trials is evaluated using NCI's toxicity grading lexicon -- Common Terminology Criteria for Adverse Events (CTCAE). Many of the toxicities included in the CTCAE are symptomatic side effects that are most validly and efficiently collected by direct patient report.
Currently, research staff identify and grade adverse events (AEs) that are documented by clinical staff in the health record. However, studies show that health professionals may underestimate the onset, duration and severity of symptoms such as fatigue and pain, compared to patients' own accounts. Also, since staff-based AE reporting generally occurs at clinic visits, AEs that occur between visits may be missed. This incomplete picture of treatment-associated toxicities hinders the drug development process and may impede clinical management of toxicities for individual patients. Incorporating the patient's self-report into toxicity evaluation may also improve patients' satisfaction and communication with their clinicians and foster early detection and preemptive management of potentially serious AEs.
Leadership & Goals
In 2007, an NCI team of outcomes researchers, clinical trialists, informaticians, and colleagues in the clinical trials community conceived the Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). Contracts were awarded in 2008 and 2010 to Memorial Sloan-Kettering Cancer Center (Principal Investigator Ethan Basch, MD) to develop PRO-CTCAE in collaboration with NCI staff. Dr. Sandra Mitchell (Program Director in ARP's Outcomes Research Branch) provides NCI scientific direction for this project.
Representatives from NCI's Division of Cancer Control and Population Sciences, Division of Cancer Prevention, Division of Cancer Treatment and Diagnosis, and Center for Bioinformatics and Information Technology collaborate to advance the technical development and psychometric testing of PRO-CTCAE, and to promote its incorporation into the workflow and operations of cancer clinical trials in academic and community-based cancer centers. The overall goal is to create a valid and responsive system for patient self-reporting of adverse symptoms in cancer trials that is widely accepted, and that is interoperable with existing NCI AE reporting systems such Medidata-RAVE and caBIG Cancer Adverse Event Reporting System (caAERS). NCI is also working collaboratively with the US Food and Drug Administration to ensure that PRO-CTCAE is ultimately eligible to receive FDA qualification as a drug development tool.
Development & Testing
The PRO-CTCAE item bank comprises 124 questions that assess attributes (e.g. presence/frequency, severity, and interference) of 78 symptoms represented in both the CTCAE (v. 4) and Medical Dictionary for Regulatory Activities (MedDRA) AE lexicons. The questions have been evaluated using cognitive testing methods and found to be clear, comprehensible, and relevant to patients. An electronic web-based architecture to permit efficient patient self-reporting has also been developed, and formal usability testing is wrapping up. Hosted by NCI, the electronic interface establishes routines to permit the collection and interpretation of these patient-reported symptom and AE data. "When a patient's symptom report crosses a severity threshold," comments Dr. Mitchell, "the system also automatically sends a message to the treating clinicians and/or investigators to follow up with the patient."
A validation study conducted at five NCI-designated cancer centers and five NCI Community Cancer Centers Program (NCCCP) centers has demonstrated the favorable psychometric properties of PRO-CTCAE in a large, heterogeneous sample of patients undergoing cancer treatment. Most PRO-CTCAE items (116/124) were shown to be valid across one or more validity criteria (p<.05). All PRO-CTCAE items correlated in theoretically expected directions with the EORTC QLQ-C30, and 96 out of 124 PRO-CTCAE items distinguished subgroups based on performance status, disease site, and/or treatment characteristics. In addition, the PRO-CTCAE items demonstrated acceptable test-retest reliability (Median ICC 0.77).
As explained by Dr. Mitchell, "based on these encouraging findings, we are pursuing additional psychometric studies to examine the responsiveness of PRO-CTCAE as an outcome measure of treatment toxicity. We are also expanding the accessibility of PRO-CTCAE to accommodate a range of patient characteristics (including patients with lower linguistic or digital capabilities, and Spanish speakers), and gathering necessary data to ensure that PRO-CTCAE is scalable for efficient integration into the workflow of the cooperative groups."
Impact on Research & Clinical Care
According to Dr. Mitchell, "data collected in the PRO-CTCAE system can be used to report in real time the onset, severity, and resolution of adverse events during and following treatment, thereby providing a more efficiently gathered and comprehensive understanding of treatment outcomes. In addition to enhancing data accuracy in NCI's clinical trials, the data gathered via PRO-CTCAE may also inform comparative effectiveness research, health technology assessments, quality improvement, and patient and family decision-making about treatments, while improving symptom monitoring and clinical management at the individual patient level."