How Do You Develop, and Successfully Incorporate, PROs in Hematology? Experts Chime in at ASH

A patient-reported outcome (PRO) is defined as the measurement of a patient’s perception of a health condition and its treatment. A session on PROs and the importance of embracing the patient perspective in healthcare delivery witnessed a healthy discussion between a developer of these tools, a representative of the FDA who reviews these tools when used by drug developers, and a clinician who is the end-user of PROs in clinical practice. These presenters were participants in an education session on Patient-Reported Outcomes in Hematology on the second day of the annual meeting of the American Society of Hematology, being held December 5-8, in Orlando, Florida.

David Cella, PhD, from the Feinberg School of Medicine, Northwestern University, began the session, introducing The Science behind Developing a Patient-Reported Outcome Measure. Cella has been a pioneer in developing The Patient Reported Outcomes Measurement Information System (PROMIS), which has seen wide-scale adoption in clinical practice. PROMIS is a National Insitutes of Health (NIH)-funded measurement system that uses item response theory and provides researchers and clinicians the ability to determine PROs using precise and reliable tools.

Although Cella is involved in developing these outcomes measures across a range of therapeutic areas, he geared his talk to present the hematology context, outlining fatigue as an indexed symptom for patient-reported outcomes measures (PROMs), the promise of PROMs, and where he sees the future for these measures.

“There’s value in asking systematic questions on patient outcomes. The series of endpoints, such as tumor response, progression-free survival (PFS), disease progression, which are commonly used as outcomes measures, may not always relate with overall survival (OS),” Cella said. “However, if we place a value on tumor response and PFS that makes sense to people, including patients’ treatment experience, it can significantly contribute to the value proposition of a regimen,” he explained.

Providers can gather outcomes such as disease symptoms, side-effect burden, and tolerability throughout the treatment. These measures may be strongly associated with patient preference. However, in order to help patients make informed treatment decisions, there’s a need to assess these measures through well-controlled studies, Cella said. It is important to realize that “Treatment benefit can be measured as a combination of treatment efficacy and treatment toxicity,” he added.

Explaining fatigue as a therapeutic index, Cella informed the audience that there are 3 international prognostic scoring systems that are well established: WHO-classification-based Prognostic Scoring System (WPSS), the International Prognostic Scoring System (IPSS), and the IPSS-revised. He then shared data from a recent paper in Lancet Oncology that evaluated self-reported fatigue in patients with myelodysplastic syndrome as an outcomes measure for OS, beyond IPSS. The researchers found that self-reported fatigue indeed provides prognostic information for survival, independent of the god-standard classifications, creating a case for fatigue to be included in routine diagnostic investigation in randomized clinical trials.

“However, what we are currently lacking is a single metric to standardize measures like fatigue—1 gold standard like blood pressure,” said Cella.

That’s how the PROMIS trial came about, he explained—PROMIS is domain-specific, not disease specific. A domain is the specific function, feeling, or perception that you want to measure, and it includes physical, social, and mental health measures. Each domain, such as fatigue, is measured using multiple “item banks,” explained Cella. The data can be collected through survey questions or computerized adaptive testing or CAT, which is an adaptive tool that is response-based.

Another patient-reported self-assessment tool that Cella introduced is HealthMeasures; a trans-NIH cooperative agreement that includes PROMIS, ASCQ-Me, NIH Toolbox, and Neuro-QoL.

“The future lies in developing common standardized metrics that can help map out instruments like PROMIS, such as the PROsetta stone, which can help link scores on 2 different measures,” concluded Cella.

So do PROs find a place in the regulatory world? Can they contribute to the drug approval process? To answer these questions, Virginia E. Kwitkowski, MS, ACNP-BC, from the Division of Hematology Products, FDA, took to the podium next. With a focus on hematology, Kwitkowski described the advances in the assessment of PROs in clinical trials and provided an overview of regulatory issues to consider when using PROs in drug development.

Kwitkowski said, “An improvement in the way patients feel and function can also be considered evidence of their performance and be used as an outcome measure. Regulators do want PROs!” However, with respect to supportive care products that are developed to mitigate toxicity, there needs to be evidence to support that patients are not being harmed, like protecting tumor from responding to standard treatment, she said.

Pointing to the difference between the regulatory authorities in Europe versus in the United States, Kwitkowski explained that while payers in Europe want PROs and regulators do consider cost implications of a drug approval to the healthcare system as a whole, the FDA cannot do so when reviewing a new drug application.

In 2009, the FDA generated guidance for the industry on using PROs to support labeling claims, said Kwitkowski, “But flexibility and judgement are vital to meet regulatory questions.”

In her opinion, existing issues with PROs in hematology/oncology include:

• Historical assessment with static-global health related quality0of-life instruments
• Rare, randomized blinded trials
• Need for a modular approach

She went on to direct the audience to a roadmap developed by the FDA for patient-focused outcome measurement in clinical trials. The salient features of the roadmap include:

• Understanding the disease or condition
• Conceptualizing treatment benefit
• Selecting/developing the outcome measure

Kwitkowski informed the audience that the FDA is very open to an exchange of ideas with drug developers on developing these PRO tool measurements, outside of a drug application.

“We are encouraged that so many stakeholders are interested in PROs in drug development. The FDA staff are actively working with stakeholders to help integrate PRO-CTCAE [Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events] in drug development. But a modular approach is recommended here,” Kwitkowski said.

The end-user of these very informative tools are oncologists and other healthcare providers, who interact with patients on a regular basis and help them make informed decisions on treatment choices. A big proponent of the use of PROs in clinical decisions is Julie A. Panepinto, MD, MSPH, an oncologist at The Medical College of Wisconsin/Children's Research Institute of the Children's Hospital of Wisconsin, Milwaukee.

As she took to the stage next, Panepinto told the audience, “We don’t know what patients are really doing outside of the clinic or the office visit,” underscoring the importance of understanding patient behavior and functionality one they leave the clinic. She pointed to several studies that have shown that providers are not adept at estimating how patients function, especially psychosocial functions and pain management. “Additionally, there is evidence to show that there’s significant discordance in not just the patient’s PRO assessment, but also in the clinical documentation of these PROs,” Panepinto said. “So we don’t just not understand patient outcomes, we are not good at documenting it either.”

In their adaptation of gathering PROs in the clinic, practices have come up with their own unique ways, Panepinto explained, showing examples of digital devices that are used. Computer kiosks, iPhones, or tabs can all be tools that can be used to collect PROs. Databases like MyCHart and REDCap can then help assemble this data, she said.

The underlying question remains, “Why collect PROs?” PROs can help a care team understand how patients are functioning. They can also help patients understand how they are performing relative to other patients like them who may be receiving similar treatment, Panepinto explained. “This information can significantly impact decision support in disease and therapy,” she said. “While promoting patient-centered care, PROs can improve patient functioning by helping us tailor and targeting therapy to meet their individual needs.”

The field is now moving towards real-time incorporation of PROs in a clinical setting to inform and adapt provider decisions on care. One such study examined the feasibility of PRO assessment via the Patient Reported Outcomes Measurement Information System-Computerized Adaptive Testing within pediatric hematology, oncology, and bone marrow transplant clinic settings. Despite some technical barriers, the study saw it feasible to integrate PROs in the clinical setting in real time.

“PROs can greatly contribute to performance improvement. Healthcare systems can use PROs to track and improve delivery system performance, and more importantly, they can have a significant impact on improving population health,” Panepinto said.