Publication
Article
Author(s):
The field of measurement in Mycosis Fungoides/Sézary syndrome type cutaneous T-cell lymphoma lacked a specific quality-of-life measure to describe patient experience or guide treatment decisions. Using an online platform with an engaged patient community, we developed and psychometrically validated a new measure in just under a year.
Mycosis fungoides and its leukemic variant Sézary syndrome (MF/SS), comprise the most common forms of cutaneous T-cell lymphoma (CTCL), a class of Non-Hodgkin’s lymphomas.1 Approximately 3000 new cases of this rare disease are reported in the United States every year.1 While CTCL typically affects older adults (median age at diagnosis: 55 to 60 years2) around 7% of cases are diagnosed before age 30.3 CTCL also affects more males than females (male to female ratio is approximately 1.6:1) and is more prevalent within the black population (black-to-white ratio is approximately 1.3:1) although both these trends have been stabilizing relative to historic norms.4
Measuring the impact of MF/SS CTCL experience for research purposes, or even clinical management with a healthcare provider, is made more challenging by the absence of disease-specific validated measures.5 Clinicians currently use patient-reported outcome (PRO) tools such as the Itchy-QOL,6 DLQI,7 Skindex-29,8 VAS itch, or cancer PROs such as the EORTC family of tools9 or the FACT-G.10 In order to capture quality-of-life (QoL) information, both a cancer-specific (eg, FACT-G) and skin-specific (eg, Skindex-29) measures are often administered in trials, which is shown to be responsive to change, but, at 56 questions, these measures can be burdensome for patients.11
There is a dearth of literature on aspects of QoL impairment by specific MF/SS CTCL subgroups selected by gender, age, clinical type, treatment, or clinical status.12 Traditional clinical assessments (such as extent of lesion, lymph node involvement, or flow cytometry) do not capture patient perspective on how disease manifestation can have a major impact on many domains related to QoL,13 such as role function or psychological distress. As a result, there is little data on the relationship between MF/SS CTCL symptom burden and health-related QoL, and researchers have sought to answer whether the connotation of having a poorly understood form of cancer causes distress, over and above the consequences of cutaneous manifestations of disease.13
Addressing these gaps in the literature is challenging, because new condition-specific measures have traditionally been expensive and time consuming to develop. According to some estimates, the time taken may be as much as 2 to 4 years, and the costs range from $725,000 to $2.1 million, depending on the extent of rigor required for developing them.14,15 A robust research program requires:
Even when resources are available, PROs, developed by researchers in isolation, risk measuring things that don’t really matter to patients.16 Despite the fact that the perspective of patients is increasingly recognized as crucial across many aspects of medical research,17 several legacy measures in widespread use today were developed with little or no patient input. These are significant challenges that are amplified when the condition of interest is as rare as MF/SS CTCL.
Early experiences have shown that patients using the internet to share information about their condition and to connect with peers can contribute their experiences to PRO developers in a rigorous fashion that helps overcome some of the aforementioned challenges.18-22 Methodology studies undertaken to evaluate the quality and representativeness of data gathered online has found that it is similar in nature to data gathered through more traditional means such as guided interview.23 Legitimate questions remain as to how online tools can verify patients identity, obtain objective data of their diagnosis and status, and overcome biases inherent in the use of online tools.24
Funded by a grant from the Robert Wood Johnson Foundation, PatientsLikeMe, built the Open Research Exchange (ORE),25 an online software platform with tools that simplify the process of developing, testing, and sharing new PROs. The instrument development process is based on a sequential multi-step, iterative process that complies with widely acceptable scientific recommendations 26-28 and FDA guidelines.29
The process consists of the following steps:
ORE is connected to an online platform where more than 400,000 patients currently track their health, connect with others for information and support, and contribute data for research; consequently, the challenges of recruitment, data collection, and follow-up are rendered more manageable. To date, more than 20 new instruments have been developed using the platform, including published measures of treatment burden,30 extracampine hallucinations, in Parkinson’s disease,31 and hypertension.32
Recognizing the need for an MF-CTCL—specific HRQoL instrument, Actelion Pharmaceuticals, and PatientsLikeMe, sought to harness new technology to rapidly develop, prototype, and psychometrically validate a new instrument with patient input throughout. Our objective was to create an instrument that could be used to better understand patient unmet need (eg, disease burden, treatment satisfaction, and health-related QoL), and help characterize the experience of patients living with MF/SS CTCL. We also sought to contribute the tool to the wider community in a Creative Commons ShareAlike 3.0 license, so that it could be used, expanded upon, and improved by other researchers.
Developing the Instrument
While a full description of the study methods is beyond the scope of this article (for more, see Towers, et al33), we will summarize the approach taken. Following a literature review and interviews with clinical experts, 21 patients reporting a physician-confirmed diagnosis of MF/SS CTCL were invited to complete open-text concept elicitation items (eg, “How is your physical well-being affected by MF/SS CTCL?”). In addition, 10 patients out of the 21 were invited for further telephone interview to probe their answers. Qualitative analysis suggested that content saturation was reached after 15 patients and thematic content analysis identified 6 major themes that were subsequently used to generate a preliminary version of the 31-item questionnaire. This long-form questionnaire was administered to 42 patients for their item-level feedback, where multiple-choice and open text survey items allowed patients to identify issues with any items they found lacked clarity or relevance. After removing redundant items, an abbreviated 14-item scale was administered to a sample of 126 patients for psychometric validation, including comparator measures. Sixty-six patients completed test-retest performance, 5 days later, with known-group validity analysis.
Psychometric analysis showed good internal consistency, test-retest reliability, and that the items were appropriately ordered in terms of severity. We also found that the MF/SS CTCL QoL is capable of discriminating between individuals with low and high levels of interference in their QoL, and that the items adequately covered the varying levels of interference with QoL due to MF/SS CTCL. Further work detailing the validation process and final instrument is under preparation.
Discussion
The use of the ORE platform enabled patient feedback to be easily incorporated throughout every step of the development process. Additionally, thanks in large part to a relationship with the Cutaneous Lymphoma Foundation, recruitment was faster and more cost effective than traditional methods of sourcing patients through clinical centers. Information collected from key opinion leaders and the literature helped to inform a scientifically grounded and relevant measure that accurately depicts the patient experience with MF/SS CTCL, particularly given the variability that patients often encounter, as they cope with the daily demands of the condition. Online feedback replaced more time and resource-expensive interviewer-led cognitive debriefing interviews. In all, the process took less than a year, and the total cost was a fraction of what traditional instrument development costs.
Limitations
The greatest limitation of the study is the sample size, due in part to the rarity of the condition. A number of psychometric validation measures were, therefore, unable to be completed and could be addressed in future research. Additionally, while it is impossible to guarantee that those who registered on the website actually have MF/SS CTCL, a recent study in other conditions, on PatientsLikeMe, found that around 95% of patients could be identified via IMS Health medical and pharmacy claims (Eicher et al, In Press). Further, the sample of patients who participated in this study were largely in Stage 1 of mycosis fungoides, or Sézary syndrome, with few patients reporting more advanced stages of the syndrome (Stage III, n = 2; Stage IVa, n = 4; Stage IVb, n = 1). Inclusion of more individuals at later stages of the syndrome would allow for empirical testing of differential item function by later stages of the syndrome.
Conclusion
EBO
The MF/SS-CTCL PRO tool project represented a unique collaboration between a pharmaceutical company and a patient network, using an innovative platform to work on a new measure for a rare disease that would otherwise have never been developed. We hope use of the MF/SS CTCL-QoL tool, which is freely available, will enable better communications about QoL between patients and their care teams, and improve and inform clinical management and treatment decision making related to MF/SS-CTCL.
Author information
Paul Wicks, PhD, is vice president of innovation, PatientsLikeMe, Inc.
Marjan Sepassi, PharmD, is associate director, Medical Affairs, Actelion Pharmaceuticals, Inc.
Gaurav Sharma, PharmD, is associate director, Medical Affairs, Actelion Pharmaceuticals, Inc.
Margot Carlson Delogne, is vice president of communications, PatientsLikeMe, Inc.
Address for correspondence
Paul Wicks, PhD
PatientsLikeMe
160 2nd Street,
Cambridge, MA
References
E-mail: pwicks@patientslikeme.com