Mixed methods analysis of Health-Related Quality of Life in ambulant individuals affected with RYR1-related myopathies pre-post-N-acetylcysteine therapy

In the era of precision medicine, the importance of patient-centered data is becoming increasingly recognized by regulators, clinicians, and the research community. Since the Food and Drug Administration’s (FDA’s) Guidance for Industry on Patient-Reported Outcome (PRO) Measures [1] and Patient-Focused Drug Development [2, 3], there has been a shift from traditional methods solely focused on clinical study goals to the incorporation of participant experience outcomes. Research methods are now including the patient perspective early in order to ensure (and validate) that any changes measured are meaningful to those patients [4]. PROs can provide important insights into disease symptoms or treatment side effects (e.g., pain, fatigue, or anxiety); functional outcomes (e.g., physical, social, emotional, etc.); and/or multidimensional assessment of the impact of both disease and treatment on physical, psychological, and social well-being, also known as Quality of Life (QoL) [5].

PROs can be used to generate both quantitative (e.g., questionnaires) and qualitative (e.g., interview, journals) data. Mixed methods research (MMR) employs quantitative and qualitative approaches to provide a more comprehensive understanding of findings. MMR has been proposed as a best practice to inform Patient-Centered Outcomes Measures in rare diseases where quantitative methods alone have found challenges [6]. Thus, MMR was employed during a combined natural history study and clinical trial of N-acetylcysteine (NAC) in Ryanodine 1-Related Myopathies (RYR1-RM) to facilitate an understanding of participants’ characterization of their condition and to identify any subjective improvements post-intervention. RYR1-RM comprise the most common of the non-dystrophic neuromuscular disorders with a pediatric incidence of > 1:90,000 in the United States, yet there are no FDA-approved treatments [7, 8].

Based on our qualitative results, we were able to identify and characterize major HRQoL concerns/issues in ambulant individuals affected with RYR1-RM. The Quality of life domain was the most relevant for participants when discussing their HRQoL. The most emphasized subdomain was social impacts, depicting specific perceptions, limitations, and deficits of understanding due to the disease. For example, mildly-affected participants were widely accused of being “lazy” or “liars”, which is consistent with previous reports in qualitative research on NMD [29, 30]. This finding highlights the importance of establishing positive interactions to enhance HRQoL. General assessment and positive aspects also stood out as notable subdomains. Participants were very optimistic in their appraisals of their QoL as a result of internal acceptance, focus on improvement, encouraging experiences, and support of family and friends. This depicts a high level of resilience in facing their condition, which is a well-documented response to adversity in physical and mental health [31, 32]. The remaining subdomains, however, received moderate-to-little attention, highlighting that, globally, they were less important to participants than the previous subdomains. For example, negative experiences were linked to specific economic-work situations, unhappy moments, the progression of the disorder, daily care and domestic activities, annoyance reactions, difficulties in being diagnosed, and uneasy feelings. Most of these have been described in RYR1-RM [12, 33, 34] and other NMD [35, 36]. However, it is surprising that self-esteem was not explored in depth in participant interviews. This indicates very little effect on self-esteem, which is in contrast to previous research in individuals with NMD [37, 38]. It is also worth noting that contrary results regarding anxiety have been reported in the RYR1-RM population [12]. All subdomains for quality of life revealed connections with the other identified areas of interest, especially with adaptive strategy and symptoms. This observation is based on the close interaction between QoL and HRQoL concepts [39].

The second domain was adaptive strategy, which gave insight into several coping strategies employed by participants. Most of these fell into the subdomains of physical behaviors and psychological strategy, which described the approaches taken to address motor restrictions (e.g., limiting physical activities, increasing rest periods, using external support, etc.) and psychological implications (e.g., planning ahead, scheduling energy, remaining positive, etc.). Implications in both areas were also reported recently for individuals with RYR1-RM [12] and HRQoL research in NMD [24]. Accommodation, equipment and help from others attained similar emphasis, disclosing the need for additional assistance from institutions, tools, and people, respectively. These factors revealed a lack of independence for participants with RYR1-RM, which is a key aspect for the NMD population in general [20]. Surprisingly, physical therapy was scarcely mentioned by participants, with minor mentions of therapeutic and occupational interventions like horseback riding. We posit this is due to the relative normality of this type of therapy in this population. However, this could be also due to differences in access to physical therapy services. This issue should be explored in depth in future research. Every adaptive strategy was linked to the specific symptoms of each participant, indicating the strong relationship between both domains, as might be expected.

Another important domain was symptoms. The most influential subdomain was fatigue, followed by weakness. This is consistent with documented clinical effects of RYR1-RM [8, 11, 12, 33, 40‐44]. Pain and impaired mobility were secondarily important in this study and were mainly described as consequences of fatigue and weakness. Likewise, there are several antecedents of pain and impaired mobility in affected individuals with RYR1-RM [8, 10, 12, 34, 43]. The remaining subdomains were less widely reported and included issues more specific to individuals such as additional complications, malignant hyperthermia experiences, respiratory issues, hand dexterity problems, and abnormal curvatures of the spine, aligning with manifestations of RYR1-RM [8, 11, 12, 33, 40, 42, 44‐46]. Rhabdomyolysis was not described in depth. However, it is possible that participants reported rhabdomyolysis involvement within the weakness subdomain, as previous research notes the importance of this symptom for individuals with RYR1-RM [10, 45, 46]. Finally, muscle spasms and numbness were touched on but not elaborated, suggesting they are not paramount. Despite this, these symptoms have been previously reported [43, 46]. Every aspect of the symptoms domain displayed a high influence on other subdomains, reinforcing the whole construct and emphasizing the use of the Symptom Science Model [15].

The domain future thoughts captured feelings about events to come. The Fears subdomain contained concerns about potential issues, including limitations on activities, worsening symptoms, or genetic implications for children. Treatment desires, a subdomain consistent with recent qualitative data in NMD [35], was essentially focused on improving HRQoL as a result of increasing physical performance. These findings could be interpreted as a dualistic display (i.e., positive vs negative) of treatment expectations [47]. This domain was closely related to other identified domains/subdomains such as symptoms and disease progression, bolstering their identification as areas of interest.

Finally, the post-intervention domain captured individual impressions of treatment effect after the NAC/placebo intervention, addressing our research hypothesis regarding whether or not these data captured additional information complementing PROMIS and Neuro-QoL instruments to detect changes in HRQoL in ambulant individuals with RYR1-RM post-treatment. These reactions were categorized as positive or no change. Participants’ feedback revealed valuable treatment benefits, most of which were related to increased levels of energy and strength, which are consistent with preclinical effects of NAC on ryr1 mutant zebrafish and mice [13, 14]. However, results from our quantitative analysis of Neuro-QoL and PROMIS T-scores were inconclusive. Social and psychological profits were also reported [39]. Only minor side effects were noted, providing further evidence that NAC is safe and well tolerated in this population. As in previous subdomains, both options were directly connected to the specific symptoms of each participant, which indicates a straightforward relationship between both domains.

Regarding data transformation, analysis by domain exposed a clear imbalance in future thoughts and post-intervention domains, recording more quotes for treatment desires and treatment benefits subdomains. This highlighted that participants with RYR1-RM tended to show more positive attitudes towards their HRQoL in comparison to their family, caregivers, or external observers, which is known as the “disability paradox” [48]. The “disability paradox” has been noted in previous research on NMD [49, 50]. Quality of life, adaptive strategy, and symptoms domains were relatively balanced. However, social impacts, psychological strategy, and fatigue subdomains scored higher number of comments for each case. These findings are consistent with the qualitative data, reinforcing the results obtained via both approaches.

The word cloud revealed the presence of three main groups of terms “people, family, friends, school, person, others”; “fatigue, energy, weakness, myopathy, affected, symptoms”; and “quality, little, problem, better, harder, difficult, without, describe”. Those words reemphasize the importance of several domains/subdomains such as social impacts; symptoms and physical behaviors; and quality of life and adaptive strategy, respectively. These results also coincide with our previous findings from qualitative analysis.

Global analysis, involving transformed data, found balanced percentages of all identified domains, considering the number of subdomains in each case and that post-intervention comments were only reported on post-intervention interviews. This outcome differed slightly from qualitative data, where quality of life, adaptive strategy and symptoms were more relevant, and suggested a similar influence of all five domains in HRQoL for participants with RYR1-RM. This slight difference highlights the value of the mixed methods approach because it allows for a more nuanced understanding of the data. Future thoughts and symptoms recorded lower percentage of comments than the remaining domains, suggesting that participants were more focused on current issues than upcoming problems, and that it was easier for them to accept physical limitations associated with their condition than social or psychological implications. A similar situation has previously supported this idea in individuals with ALS [51]. These results are similar to previous frequency analysis in NMD [24].

The analysis by visit displayed similar percentages for each domain, which is consistent with our findings for global analysis. Taking into account the number of quotes per visit, baseline interviews included more comments. This observation suggests that participants were inclined to provide more information at the beginning of the study as it was their first visit/interview. Perhaps this was also due to positive expectations of the trial [47]. Normalized percentages per visit reported similar results, revealing that participants provided comparable information in each visit. Finally, results for data transformation analysis by participant resulted in a similar trend to the global analysis findings.

As previously stated, RYR1-RM are slowly progressive diseases, which makes it necessary to capture disease progression when considering effects of treatments and HRQoL. This issue was captured though the identification of the domain and subdomain of post-intervention and disease progression. Both concepts were analyzed in qualitative and data transformation analyses.

Three domains from Neuro-QoL showed some evidence of a difference between the NAC and placebo groups, fatigue, anxiety, and stigma. The p values for these domains ranged between 0.014 and 0.034, which do not reflect strong evidence of a difference given that 20 tests were performed. The sample size was small given the rare nature of the disease, and therefore the statistical power to detect differences was also small. Likewise, although there was a trend on the subjective self-assessment, suggesting a potential positive effect of NAC on participants with RYR1-RM, our sample size may be too small to produce strong evidence of a causal relationship between subjective self-assessment and NAC treatment. Thus, there was no clear evidence that any of the three tools captured positive changes as a result of NAC treatment on quantitative analysis. Interestingly, the results of our analyses of both subjective quantitative and qualitative/transformed data were similar to our findings from our primary endpoints in that NAC did not significantly improve any endpoints [9].

As mentioned, the main limitation in this study (for the transformed and quantitative analyses) was the sample size, which was small. Another major limitation was that all enrolled participants were ambulatory due to 6MWT being a primary endpoint, which meant that only mild-to-moderately affected individuals were enrolled.