Background

Myelofibrosis (MF) is a malignant clonal disease characterized by progressive marrow failure, splenomegaly, and decreased longevity due to infections, bleeding, and leukemic transformation [1]. MF comprises numerous burdensome symptoms for patients, including fatigue, night sweats, upper left quadrant abdominal pain, bone pain, and, among others, unintentional weight loss. These symptoms, in turn, often have negative impacts on patients’ quality of life; patients with MF frequently complain of night sweats, severe itching, early fullness or satiety, and a variety of pains in their bones and muscles, under their ribs, and in their abdomen, which all contribute to a reduced health related quality of life (HRQoL) [2].

When assessing the full impact of disease and treatment on HRQoL, it is important to consider the patient’s perspective. Indeed, the European Medicines Agency (EMA) [3] and United States (US) Food and Drug Administration (FDA) [4] acknowledge the importance of the patient’s voice in clinical trials by outlining steps for good patient-reported outcome (PRO) instrument development. The only currently FDA-approved drug for first-line treatment of MF is Jakafi™ (ruxolitinib [RUX]) [5]. RUX is indicated for the treatment of patients with intermediate or high-risk MF, including primary MF (PMF), post-polycythemia vera (PPV-MF), and post-essential thrombocythemia (PET-MF), having been shown to reduce spleen volume and improve MF-associated symptoms in those populations.

The modified Myelofibrosis Symptom Assessment Form (MFSAF) v2.0 diary (a symptom assessment PRO tool) was used successfully to obtain PRO label claims from the FDA [6] and EMA in patients naïve to RUX inhibition (RUX-naïve). Fatigue, a hallmark symptom of MF, is missing from MFSAF v2.0, and it is unclear if this tool is appropriate for patients resistant or intolerant to RUX treatment (RUX-experienced). To overcome these issues, this study aimed to develop a new PRO instrument. Specifically, the goal of this study was to understand the patient perspective of living with MF and establish content validity of a new PRO instrument for administration in both RUX-experienced and RUX-naïve MF patients. Data was collected from three sources: patients, experts, and the published literature. The result of this work contributed to the development of the Myeloproliferative Neoplasm Symptom Assessment Diary (MPN-SD), a PRO instrument designed for the purpose for assessing key symptoms among RUX-experienced and RUX-naïve MF patients.

Methods

This study was conducted in accordance with the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) PRO Good Research Practices Task Force Report: Part 1 [7] and 2 [8] and FDA PRO Guidance [4], which each specify methods for development of a PRO tool.

Qualitative interviews with MF patients

Patient recruitment

All study documents were institutional review board-approved prior to study initiation. Interviews were scheduled after a potential participant confirmed his or her interest in participating, signed informed consent and the Health Insurance Portability and Accountability Act (HIPAA) Authorization, and was deemed eligible (see Fig. 1 displaying the recruitment process).

Fig. 1
figure 1

Agency and academic research site patient recruitment process

Full size image

Patient eligibility criteria

Eligible patients included those who were ≥ 18 years of age; fluent in US English; willing and able to participate and complete questionnaires; diagnosed with PMF or secondary MF (PET-MF and PPV-MF) as per World Health Organization 2008 criteria; had failed prior treatment with a RUX (RUX-experienced), or had no prior RUX treatment and Intermediate-1, Intermedate-2 or High-risk MF according to the Dynamic International Prognostic Scoring System (RUX-naïve); had a palpable spleen; and had reported current MF symptoms. A patient was not eligible if he or she met any of the following exclusion criteria: previous treatment with a licensed or experimental smoothened inhibitor; targeted anti-cancer therapy up to 14 days prior to enrollment; a history of drug or alcohol abuse within the last 6 months; cognitive impairment; the presence of a psychiatric condition; or any other clinically relevant concern that would interfere with the patient’s ability to provide written informed consent and/or participate.

Patient interview methods

A semi-structured interview guide was developed to capture open-ended and probing questions regarding symptoms and impacts of MF from the patients (concept elicitation interviews, CEIs); evaluate the MPN-SD for readability, comprehensibility, relevance, and comprehensiveness (cognitive debriefing interviews, CDIs); and assess the usability of the electronic patient-reported outcome device (ePRO usability testing). CEI, CDI, and ePRO usability testing were conducted during the same interview session. Interviews were audio recorded; conducted face to face, over the telephone, or virtually; and each lasted 60–90 min. Interviewers were trained via the National Institutes of Health Human Participant Protection.

Data collection

During CEIs, symptom expressions were tracked; participants ranked the top five symptoms they would like to have improve with treatment; and reported on the impact of MF symptoms. Concepts were tabulated for frequency (i.e., number and percent of spontaneous reports).

For CDIs, the MPN-SD was presented on an ePRO device and patients were asked to “think aloud” about the process they used to arrive at each answer. Data were collected on the words, terms, or concepts that were not understood or were interpreted differently than intended [9]. The MPN-SD was modified mid-way and a second wave of interviews was performed to evaluate the changes. The usability of the ePRO device was evaluated via questionnaire.

Audio-recordings of interviews were transcribed, anonymized, and entered into ATLAS.ti Version 7.0 (ATLAS.ti Scientific Software Development GmbH, Berlin) [10]. A code book was developed based on researchers identifying transcript text relevant to the research objectives and matching it to a code from the coding scheme that best characterized the data [11]. Analysis was performed separately for RUX-experienced and RUX-naïve patient groups. Saturation was evaluated; saturation characterizes the point at which no new information can be generated from additional interviews [7]. The demonstration of saturation is used as evidence for the adequacy of the sample size (if saturation is not observed, additional interviews may be considered).

Literature review

In order to provide further evidence of the relevance and importance of concepts identified during patient interviews, and to help guide revisions to the interview guide between interview waves, a conceptual literature review was conducted concurrent with the start of patient interviews. It was designed to capture key articles that identify, define, and substantiate the symptom-level concepts experienced by adults with MF. The search was conducted January 2015 via the OvidSP platform. Key words included: “myelofibrosis,” “osteomyelofibrosis,” “patient report,” “qol,” “quality,” “signs,” and “symptoms.” The search was limited to humans, US English, and journal publication type. No limits were placed on publication year. Concepts from the literature were categorized and prioritized (reported by ≥5 full text articles) according to symptoms, impacts, or MF treatment (see Fig. 2).

Fig. 2
figure 2

Literature review screening process and article flow

Full size image

Expert advice meetings

Interviews with four experts in MF substantiated symptom and impact level concepts of measurement. Experts were identified from a list of clinicians who specialized in hematology. Interviews were conducted over the phone by trained interviewers, followed a semi-structured interview guide, and were each 60 min. The interviews were analyzed using semi-quantitative and qualitative data analytic methods via ATLAS.ti Version 7.0 (ATLAS.ti Scientific Software Development GmbH, Berlin) [10].

Results

Qualitative interviews with MF patients

Twenty-three MF patients participated in interviews (RUX-experienced = 16, RUX-naïve = 7). Eleven were recruited and enrolled from Global Market Research Group, three from Mayo Clinic, six from Utah University, and three from the MPN Research Foundation. Interviews occurred between October 2014 and September 2015.

Demographics

Mean age for RUX-experienced patients was 67 years (range of 30 to 85, standard deviation [SD] ±15.5). Most were white (n = 14, 88%), and male (n = 9, 56%). Thirty-one percent of patients reported “fair” general health (n = 5) and 69% rated the severity of their condition as moderate (n = 11). Mean age for RUX-naïve patients was 59 years (range of 46 to 71, SD ±8.8). All were white/Caucasian (n = 7, 100%) and most were male (n = 4, 57%). Seventy-one percent of RUX-naïve patients (n = 5) reported being in “good” general health and rated the severity of their condition as mild. See Table 1 for additional demographic characteristics.

Table 1 Patient reported demographic and health information
Full size table

Concept elicitation interviews

Patient-reported symptoms of MF

A total of 29 symptoms were expressed by RUX-experienced patients, and 14 symptoms were reported by RUX-naïve patients. Eleven symptoms were reported by both groups of patients. Eighteen symptoms were reported by RUX-experienced patients only. Three symptoms (appetite loss, cold, and rashes) were reported by RUX-naïve patients only (see Table 2). Table 3 displays examples of MF patient quotes for the prioritized symptom expressions.

Table 2 Myelofibrosis primary symptom concepts
Full size table
Table 3 Example patient expressions
Full size table

All 16 (100%) RUX-experienced patients reported fatigue and/or tiredness (“…all of a sudden I started getting more lethargic and more tired, more fatigued” [05–02-M-74]), while 11 (69%) reported shortness of breath (“…I feel like I can’t even catch my breath. It’s, it’s getting worse. It’s not getting better at all.…” [01–02-M-73]), and nine (56%) reported pain below the ribs on the left side and/or stomach pain and/or abdominal pain (“I used to have a lot of spleen pain before I was on treatment.…it’s upper left … right under where my boob is … all the way to the left. Like on the side.” [03–09-F-29]), and enlarged spleen. The signs and symptoms most commonly reported by the RUX-naïve patients were fatigue and/or tiredness (reported by six patients, 86%; “I’d get really fatigued during the day. Come home, take a three-four-hour nap and, uh, never felt any different. Still fatigued …” [01–04-M-50]), pain below the ribs on the left side (reported by six patients, 86%; “…it is definitely on the, uh, left side of the ribcage.… It’s continual.…” [01–17-M-70]), and enlarged spleen, full quickly/filling up quickly, night sweats and/or general sweats, and itching (reported by four patients, 57%, each).

Patient-reported impacts of MF

Thirty-two impact concepts were expressed by RUX-experienced patients and 20 by RUX-naïve patients. The most frequently reported impacts for the RUX-experienced patients were impact of daily activities (n = 13, 81%), physical impact (n = 13, 81%), emotional and/or psychological impact (n = 11, 69%), social impact (n = 9, 56%), and impact on work and/or school (n = 8, 50%). The most commonly reported impacts for the RUX-naïve patients were social (n = 5, 71%), emotional and/or psychological (n = 4, 57%), and impact on work (n = 4, 57%).

Saturation

Concept saturation was achieved for both groups (see Table 4 for analysis of RUX-experienced patients, and Table 5 for analysis of RUX-naïve patients). Based on these saturation results, it was determined that an adequate number of interviews were performed.

Table 4 Saturation grid for subjects with MF who are treatment-experienced (n = 16)
Full size table
Table 5 Saturation grid for subjects with MF who are treatment naïve (n = 7)
Full size table

Cognitive debriefing interview and ePRO usability

The final version of the 10-item MPN-SD includes items on filling up quickly, abdominal discomfort, inactivity, itching, night sweats, pain below the ribs on the left side, bone pain, fatigue (tiredness), shortness of breath, and appetite (loss); see Table 6 for the final version of the MPN-SD. Participants interpreted the MPN-SD instructions, items, and response options as the developers intended, and the concepts were relevant to both RUX-experienced and RUX-naïve patients.

Table 6 MPN-SD
Full size table

The MPN-SD instructions were modified based on two waves of interviews. Eight MF patients provided feedback on the original MPN-SD instructions (five RUX-experienced and three RUX-naïve patients), and five MF patients provided feedback on the alternate instructions (one RUX-experienced and four RUX-naïve patients).

The original instructions “During the PAST 24 HOURS, at its WORST, how was each of the following myelofibrosis symptoms? Please tap one number.” were modified to “The following screens display questions about your myelofibrosis symptoms. Please rate each symptom at its WORST during the PAST 24 HOURS.” More patients interpreted the alternate instructions correctly (92% of RUX experienced [n = 13] and 100% of RUX-naïve patients [n = 7]) as compared to the original instructions (92% of RUX-experienced and 43% of RUX-naïve patients).

“Bone or muscle pain” in the original MPN-SD was modified to “bone pain” based on feedback from five RUX-experienced and seven RUX-naïve patients. RUX-experienced patients (n = 5) indicated that bone pain was more relevant to their experience, expressing that (05–03-M-79) “Yes, I prefer the other one, the bone pain.… I’m not aware of any muscle pain associated with myelofibrosis.” Further, separate items for fatigue and tiredness in the original MPN-SD were combined into “fatigue (tiredness)” based on feedback from 12 RUX-experienced and seven RUX-naïve patients. Sixty-one percent of the RUX-experienced and 43% of RUX-naïve patients reported that fatigue and tiredness were the same. Patients expressed that (05–02-M-74) “…all of the sudden I started getting more lethargic and, and more tired, more fatigued.”

All RUX-experienced patients spontaneously reported shortness as breath as relevant and important to their experience of MF, five patients mentioned it spontaneously and six confirmed its relevance when probed. Experts also prioritized shortness of breath. Based on these two lines of evidence a shortness of breath item was added to the MPN-SD.

An item for ‘appetite’ was added to the MPN-SD. Appetite loss was relevant and important for treatment naïve patients (14%), by experts, and in the literature. The concept did not come up spontaneously in the treatment experienced patient interviews, although weight loss did.

No revisions were made to either the MPN-SD response options or to the ePRO device. The final MPN-SD includes 10 items, with response options ranging from 0 to 10 (0 = absent, 10 = worst imaginable for nine items, and 0 = normal appetite to 10 = complete loss of appetite for one item). It has a 24-h recall period and is administered via ePRO device. The MPN-SD has a total symptom scoring algorithm. See Table 6 for the final version of the MPN-SD.

Literature review

Thirty-two full text articles were included in the literature review [1, 2, 12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41]. A total of 32 symptoms were identified from the literature (see Table 7). The symptom concepts most frequently attested in the reviewed literature were: night sweats, fatigue, itching, bone pain, weight loss, fever, abdominal discomfort, early satiety, pain under the left ribs, dyspnea, and appetite loss. All of these, with the exception of weight loss and fever (which are clinically assessed), were also included in the symptom concepts assessed in the MPN-SD.

Table 7 Concepts identified in the literature review
Full size table

Expert advice meetings

Four experts in hematology/oncology, hematology, and internal medicine, who all worked academic (teaching) hospitals, participated in the expert advice meetings. Twenty-four symptoms and 31 impacts were identified by experts. All experts reported that patients with MF generally experience the same set of symptoms pre- and post-RUX treatment failure. See Table 8 for the list of symptoms identified by the experts.

Table 8 Expert-reported symptoms of myelofibrosis
Full size table

Discussion

This work demonstrates the comprehensiveness of the MPN-SD in assessing MF symptoms in RUX-experienced and RUX-naïve patients. The MPN-SD is simple and easy for patients to understand and complete and was developed following the US FDA regulatory guidance on PRO instruments. The patient CEI and CDI, expert interviews, literature review, and usability testing have shown that the MPN-SD has robust content validation.

Concepts identified in the CEIs, CDIs, and usability testing indicated that the MPN-SD is suitable for use in both MF patient groups. Of importance is the applicability of the MPN-SD in the clinical setting and ensuring that the instrument captures the real world experience of MF patients. Experts provided support for the use of the MPN-SD to assess many relevant symptoms of MF in both RUX-naïve and RUX-experienced patients.

A recent internet survey study [2] of 1179 patients with myeloproliferative disorders identified fatigue as one of the most frequently reported symptoms (80.7%) and main contributor to poor quality of life. Indeed, findings from this work indicate that fatigue is a relevant component of MF and if it were improved it would greatly benefit patients’ lives. Given that patients talked about both fatigue and tiredness as the same experience, the two concepts were included in one item on the final MPN-SD.

Several limitations of this study should be considered. Only three of the 16 RUX-experienced patients participated in the ePRO usability interviews; this is a very small sample size to draw conclusions from. The study sample is predominately white, which limits the generalizability of the findings to other racial groups. The CEIs and CDIs were combined in this study; one of the main criticisms of this approach is respondent bias. To attempt to reduce bias, the interview order was randomized. The study reported herein was completed in 2015; thus it does not reflect literature published after 2015, and more recent information could be available at the time of this publication.

The next step in development of the MPN-SD is psychometric evaluation. Demonstration of adequate psychometric performance (e.g., score reliability, construct-related validity, and sensitivity to change) and the creation of score interpretation guidelines (e.g., using anchor-based analyses) could facilitate the use of the MPN-SD to support the development, regulatory approval, and expedited availability of novel medicines for MF patients.

Conclusion

Improved understanding and development of an appropriate measurement tool to assess patients’ experiences of MF expands a perspective that was previously based primarily on tools developed for RUX-naïve patients. This work may help identify critical target areas for evaluation in clinical studies and guide investigators in selecting outcome variables suitable for intervention for MF patients.