This study describes the development of the PNH-SQ, a content-valid questionnaire developed in accordance with regulatory guidelines [9] and scientific best practices [14]. A literature review of instruments used to assess PNH symptoms in a clinical trial context informed the need for the development of a daily symptom diary that assesses the most salient symptoms of PNH from the perspective of the patient. The PNH-SQ was developed by combining the findings from the literature review, as well as expert clinician advice, and was further refined through patient interviews.
CE interviews with 15 PNH patients confirmed that most patients (i.e., > 90%) experience symptoms due to their PNH and that they experience a wide variety of symptoms [7, 8, 13]. Even while many (7/15) participants were receiving treatment for their PNH, study participants reported experiencing over 27 signs and symptoms due to PNH with the most common being fatigue (87%), abdominal pain (60%), dysphagia (47%), sexual difficulties (40%), and back pain (40%); and one-third (33%) of patients reported cognitive difficulties, headache, muscle weakness, and shortness of breath. A small qualitative study of 29 patients by Weitz et al. [7] reported similarly high rates of fatigue (97%) abdominal pain (59%), dysphagia (41%), and sexual difficulties (47%) [8]. Likewise, similar rates of erectile dysfunction (38%) and fatigue (80%) were reported in the International PNH registry. However, both in the International PNH Registry and in the Weitz study, patients reported substantially higher rates of headache (33% vs. 63% and 76%, respectively) and shortness of breath (33% vs. 64% and 66%, respectively) [7, 8]. Further, participants in our study also reported experiencing leg/back pain (53%), muscle weakness (33%), and sleep problems (27%), which were not reported by the patients in the International PNH Registry or in the qualitative study [7, 8]. Differences in the experience of symptoms may be driven by case mix, including severity of PNH (most patients in our study had self-reported mild disease but severity was not reported in the registry [7] or Weitz study [8]) and treatment status (none of the patients in the Weitz study [8] were treated with eculizumab vs. 25% in the registry [7] and > 45% in our study).
The significance of fatigue in the PNH experience is underscored by the fact that fatigue was the most commonly experienced symptom (13/15; 86.7%) and was considered the most bothersome symptom by the majority (8/15) of patients in this study. The proportion of individuals with PNH experiencing fatigue corresponds with findings from the International PNH Registry and the Weitz study; both of which found that fatigue was the most frequently reported symptom (80% and 97%, respectively) [8]. Moreover, in our study, nearly one-half of participants experienced fatigue in the prior 24-h (7/15) and more than half (4/7) rated the severity as moderate to severe. This is consistent with a cross-sectional survey including 74 participants with PNH in which mean levels of fatigue were severe [15]. Our findings are also consistent with the Weitz study, in which fatigue was reported to occur frequently or almost constantly in 52% of the study participants. Overall, the evidence from this and other studies demonstrate that fatigue is a frequently experienced symptom of PNH.
Fatigue has been positioned as a key secondary endpoint in 2 phase-3 PNH clinical trials [16, 17], both using the FACIT-Fatigue to measure the symptom. Although, the FACIT-Fatigue has been shown to comprehensively capture the fatigue that PNH patients with experience [8], it uses a 7-day recall period. As previously discussed, most patients with PNH experience fatigue frequently and the use of a 7-day recall period does not allow for the measurement of day-to-day variability in fatigue that participants may experience due to changes in disease activity or as a result of the treatment cycle. Although we acknowledge that there is some support for longer recall periods [18], the FDA’s PRO guidance suggests that a shorter recall period is preferable for instruments used in clinical trials due to concerns about recall bias [10]. As has been demonstrated in other chronic diseases, using a 7-day recall period may introduce a recall bias with patients recalling higher symptom intensity than when a daily recall period is used [19, 20]. For example, a study examining fatigue among 97 rheumatology patients confirmed that recall bias was not significant when comparing daily recalls of fatigue to momentary ratings, suggesting that a 24-recall period is suitable for fatigue [21]. The use of a daily symptom questionnaire, such as the PNH-SQ, which can capture day-to-day changes in fatigue and other symptoms can provide a more nuanced representation of the patient’s experience of these symptoms. Additionally, in measuring the daily variability in fatigue, the PNH-SQ may be a more sensitive measure than the FACIT-Fatigue in detecting changes in this symptom.
A widely used PRO measure in PNH clinical trials [16, 17], the EORTC QLQ-C30 was found to inadequately capture symptoms experienced by PNH patients [8]. The study found that while the measure was clear and easy to understand, it contained several items that were of no or low relevance to US patients with PNH, including items such as vomiting, needing help with eating and dressing, nausea, and diarrhea. In our study, none of these signs or symptoms emerged as concepts important to patients with PNH. Importantly, the most common symptoms of PNH such as headache, dysphagia, and abdominal pain reported by participants in that study and in our study are not directly assessed by the EORTC QLQ-C30. In addition, the most commonly experienced and bothersome symptom of PNH, fatigue, is not directly assessed in the EORTC-QLQ-30, which includes a question about “tiredness” rather than fatigue. In fact, the conclusion from the study evaluating the EORTC QLQ-30 as a PRO for patients with PNH was to add items regarding common symptoms of PNH, namely, abdominal pain, headache, and shortness of breath [8]. Evidence from qualitative studies, including our own, suggests that the EORTC QLQ-C30 does not assess all symptoms that are meaningful to patients with PNH [8], an aspect that is critical in the use of PROs in clinical trials [10].
In contrast, to ensure that the items of the PNH-SQ were relevant and comprehensively captured, participants in our study were asked to comment on the relevance of each item. Item relevance on the PNH-SQ was > 50% for all items (range 54.5–100%). Furthermore, when asked about measure comprehensiveness, no patient suggested adding further symptoms. These findings support the content validity and comprehensiveness of the PNH-SQ.
A new PRO instrument—the Quality of Life Questionnaire for patients with Aplastic Anemia and/or Paroxysmal Nocturnal Hemoglobinuria (QLQ-AA/PNH)—was recently developed to measure quality of life in patients with aplastic anemia and PNH [6, 22]. The instrument was developed with significant patient and clinician input, has 54 items with a 14-day recall period for most items and a 6-month recall period for 2 items. While the measure does assess some symptoms, the focus of the items are quality of life impacts and healthcare experience [22]. Indeed, the instrument addresses a significant need for a quality of life tool specific to patients with PNH but it does not comprehensively assess PNH symptoms, their frequency, or their severity. Further, as previously discussed, there is day-to-day variation in experience of PNH symptoms and the 2-week recall period may introduce recall bias in the measurement of symptoms [19,20,21]. The QLQ-AA/PNH may be a complementary instrument to the PNH-SQ; together these instruments can comprehensively assess both symptom burden and quality of life impact of PNH.
Recently, the patient-reported outcome questionnaire for aplastic anemia and paroxysmal nocturnal hemoglobinuria (PRO-AA/PNH) was developed [23]. However, upon comparison of the measures, only some of the 10 PNH-SQ items are also measured in the PRO AA/PNH (e.g., fatigue, shortness of breath, difficulty concentrating, and dysphagia). For instance, the PRO AA/PNH captures pain in one general question, whilst the PNH-SQ asks about specific issues with pain because this was a significant area of concern raised by patients (items include back, leg, and stomach pain). Additionally, the PRO-PNH/AA captures some signs of PNH (eg, hemoglobinuria, jaundice) and impacts (mood, trouble doing strenuous activities). The PNH-SQ focuses primarily on symptoms of PNH rather than signs or impacts of PNH, and was developed to capture the most relevant symptoms that only patients discussed (e.g., erectile dysfunction was not included because on a day-to-day basis, this was not reported as an issue). Thus, the PNH-SQ may be a more sensitive measure of change in symptoms. Furthermore, the impact of PNH is multi-faceted and arguably warrants a more comprehensive approach to assessment by using something like the QLQ-AA/PNH [6].
The findings from this study must be understood in the context of several limitations. The study included 15 patients with PNH, all of whom were from the US and all interviews were conducted in English which may impact its cross-cultural validity. However, a translatability assessment was conducted to ensure that the instructions, items, and response options could be translated across a variety of different languages, as individuals who speak different languages may express their symptom experience differently. Moreover, in order to obtain a diverse group of participants and best capture the wide range of PNH experiences, recruitment targets were set by age group, ethnicity, race, educational attainment, treatment status, and self-reported disease severity. While there is no evidence to suggest that PNH symptom experience would differ across race or ethnicity, no participants in this study were African American and only 1 Hispanic individual participated in the study. Nevertheless, since saturation was reached, the conduct of additional interviews is unlikely to result in the elicitation of additional, key symptoms that are relevant to patients with PNH.
Further, clinical confirmation of PNH diagnosis was only possible for 4 patients and thus the study relied primarily on self-reported PNH diagnosis for most participants. Considering the nature of the study, the detailed disease-related questions, and the fact that participants were recruited from a PNH or rare disease advocacy organization, it is unlikely that those participants who self-reported their PNH diagnosis did not in fact have PNH. Furthermore, PNH patients with AA were not included in this study; this measure may therefore not comprehensively capture the experience of patients who have both diseases.
In conclusion, the PNH-SQ is a new, content-valid PRO measure suitable for assessing daily symptoms experienced by patients with PNH in a clinical trial context. The development of the PNH-SQ was informed by the literature and by clinician and patient input. While further research is necessary to determine the psychometric properties of the PNH-SQ (i.e., reliability, construct validity, ability to detect change, meaningful change threshold, and scoring algorithm), this study demonstrates that the PNH-SQ is a clear and easy to understand questionnaire that has comprehensive coverage of relevant PNH symptoms. Lastly, because the PNH-SQ was developed for daily administration, it can capture daily variations in symptoms of PNH patients, providing a more granular assessment of symptom changes among patients enrolled in clinical trials.