Pozelimab for CHAPLE disease: results from in-trial interviews and clinical outcome assessments

Abstract

Background

CD55 deficiency with hyper-activation of complement, angiopathic thrombosis, and protein-losing enteropathy (CHAPLE) disease is ultra-rare (< 100 children or young adults worldwide) and potentially fatal. The study used mixed-methods approaches to assess how pozelimab impacts the signs and symptoms of CHAPLE disease from the patient perspective by combining within-trial interviews and clinical outcome assessments (COAs) (ClinicalTrials.gov, NCT04209634).

Methods

Interviews conducted with patients/caregivers at screening and week 24 assessed the signs and symptoms of CHAPLE disease, including those which were most bothersome, and evaluated the change. Patients/caregivers and clinicians completed the COAs; interview data contextualized the meaningfulness of change.

Results

Ten patients (aged 3–19 years) were enrolled; caregivers contributed to nine interviews. Abdominal pain, diarrhea, facial and peripheral edema, nausea, and vomiting are the core signs and symptoms of CHAPLE disease (≥ 90% patients experienced pre-treatment); the most bothersome signs and symptoms were abdominal pain (n = 9) and facial edema (n = 1). All core signs and symptoms were reported as resolved at week 24 interviews. Severity on global assessments changed from “mild” to “very severe” at baseline to “no signs or symptoms” at week 24. Interview results were generally consistent with sign- or symptom-specific COA scores.

Conclusions

Patients with CHAPLE disease treated with pozelimab for 24 weeks experienced complete resolution of core signs and symptoms. Mixed-methods approaches can contextualize the patient experience (how patients feel and function) in rare disease trials.

Trial registration

Clinicaltrials.gov, NCT04209634, registered December 24, 2019, https://classic.clinicaltrials.gov/ct2/show/NCT04209634.

Introduction

CD55 deficiency with hyper-activation of complement, angiopathic thrombosis, and protein-losing enteropathy (CHAPLE) disease is an ultra-rare, potentially fatal condition caused by mutations in the CD55 gene. These mutations prevent normal cell surface expression of the CD55 protein, which regulates the complement system, by accelerating the inactivation of C3 and C5 convertases [1]. Loss of CD55 causes primary intestinal lymphangiectasia and protein-losing enteropathy, which can result in severe abdominal pain, chronic/recurrent diarrhea, vomiting, and edema that can require frequent hospitalizations and medical interventions [1]. Patients experience malabsorption, resulting in impaired growth, anemia, and vitamin and nutrient deficiencies [1]. This newly described disease is associated with high mortality [2, 3]. Estimates suggest there are fewer than 100 patients worldwide [4] and, prior to the August 2023 approval of pozelimab (ClinicalTrials.gov: NCT04209634) [5] by the US Food and Drug Administration (FDA), management for the condition was supportive and used a similar therapeutic approach to that for acquired or familial forms of protein-losing enteropathy, e.g. corticosteroids, biologics, immunomodulators, intravenous (IV) or subcutaneous (SC) immunoglobulin, IV albumin, micronutrients, and restricted diets [1, 3, 6]. Additionally, off-label IV administration of eculizumab has been shown to ameliorate CHAPLE disease [2, 7]. Pozelimab is an SC-administered fully human monoclonal immunoglobulin G4P antibody directed against the terminal complement protein C5, a key protein for activation of the terminal pathway of the complement system. The SC administration of pozelimab may be less burdensome and costly to patients compared to treatments requiring IV administration [8, 9].

The FDA has advanced patient-focused drug development guidance documents and dual-methodological approaches (i.e. “mixed methods” using qualitative and quantitative information) as a means to incorporate the patient perspective into clinical trials for rare diseases [10, 11]. Given the rarity of the condition and the trial’s small sample size, and the potential heterogeneity of the symptomology, within-trial interviews and clinical outcome assessment (COA) questionnaires were used to evaluate the signs and symptoms of CHAPLE disease, including each participant’s most bothersome sign/symptom (MBS) [12], from the perspective of the patient before and during treatment with pozelimab. Here, we report on the treatment benefit of pozelimab on the signs and symptoms of CHAPLE disease, an ultra-rare condition, using a novel mixed-methods approach combining within-trial interviews and COA data to better contextualize the patients’ experience of the medication’s effects.

Methods

Trial design

NCT04209634 is an ongoing, open-label, single-arm, multicenter, phase 2/3 study evaluating the efficacy and safety of pozelimab in pediatric and adult patients with CHAPLE disease. Methods and results of the trial, including details on the study design, inclusion/exclusion criteria, and the efficacy and safety results, are described elsewhere [13, 14], while the research described herein focuses on the patient experience of signs and symptoms of the condition. Patients aged ≥ 1 year with a clinical and genetically confirmed CHAPLE disease diagnosis and active or inactive/controlled disease were eligible. At baseline (day 1), patients received a single IV loading dose of pozelimab 30 mg/kg, followed by SC dosing based on body weight once weekly over the treatment period. During clinical trial planning, potential core sign and symptom concepts were identified by reviewing the published literature and through discussions with treating clinicians. Six signs and symptoms (abdominal pain, facial edema/swelling, peripheral edema/swelling, nausea, diarrhea, and vomiting) were selected for the COA measurement strategy, as key features of the disease experience which were expected to improve with treatment [15]. Given the potential for heterogeneity in the experience of disease between patients and the possibility that patients could select a sign/symptom that could not ameliorate with treatment (e.g., speech disability following a stroke), each participant selected the sign/symptom that was most bothersome before starting pozelimab from a list of potential core signs and symptoms identified through clinician interviews [12, 16, 17].

Interview methods

Two 60-minute interviews were conducted at screening and week 24 with all trial participants and/or their caregivers. Screening interviews were conducted prior to initiating pozelimab and were used to confirm the core signs and symptoms as well as to document each patient’s MBS; week 24 interviews were conducted at the week 24 clinical trial visit to understand the impact of treatment with pozelimab on the patient’s experience of their disease. Patients ≥ 8 years of age were the primary respondents of the interview, with input from their caregiver as appropriate; for patients < 8 years old, or those with cognitive impairments, the caregiver was the primary or sole respondent of the interview, although the patient was also invited to contribute.

Each interview was conducted by a trained interviewer in person or via telephone (an option for the week 24 interviews due to the COVID-19 pandemic) in the native language of the participant. Except for one participant whose interviews were conducted by a research nurse at the site, all other interviews were conducted by persons independent from the clinical site staff. All interviewers were trained in qualitative interview methodology. The interviews were audio-recorded, transcribed, anonymized, and translated into US English by an independent transcription company. Transcripts were anonymized by removing any potentially identifying information (e.g. names of people and places) from the transcripts.

Clinical outcome assessments

Given the inherent constraints of studying an ultra-rare disease group, COAs included in the clinical trial were selected with the goal of following best measurement practices summarized in regulatory guidance documents [10]. Depending on the age of the patient, the primary respondent of the COAs was either the patient themselves (if aged ≥ 12 years) or the caregiver (if aged < 12 years).

Global assessments of symptom/disease severity and change were completed by patients or caregivers and clinicians. Specifically, the patient/caregiver global impression of severity (GIS) questions assessed overall sign and symptom severity using a five-point response scale ranging from absent (0) to very severe (4) [18, 19]. A similar approach was used for the patient/caregiver global impression of change (GIC) questions, except that the change in the disease severity was assessed on a seven-point response scale ranging from much worse (–3) to no change (0) to much better (3). Clinicians also rated global disease severity and change using the same five-point and seven-point scales for clinician GIS and GIC, respectively.

Several COAs were administered to assess potential core signs and symptoms: the Pediatric Quality of Life Inventory (PedsQL™) gastrointestinal symptom scales (pain and hurt, diarrhea, and nausea/vomiting subscales) [20], facial edema physician assessment, and peripheral edema physician assessment. These are summarized in Table 1 and Additional file 1.

Table 1 Sign- and symptom-specific COAs by core signs and symptoms

Analysis

Interview data elicited from patients and caregivers from the screening and week 24 interviews were coded for each patient, with specific attention to any changes in the signs and symptoms identified as key during trial planning, along with the MBS that was reported at the screening interview. Coding and analysis were performed on the translated transcripts, using qualitative data analytic methods and following a pre-specified qualitative data analysis plan. Interview data were pooled across participants and were analyzed for concept frequency and concept descriptions.

To evaluate overall improvement by week 24 for each participant, both information from the week 24 interviews about signs/symptoms and scores on the global assessments of severity and change (reported by patient/caregiver and clinician) are presented per patient.

Additionally, changes in COA scores are reported alongside example interview quotes from participants describing their experience of the sign or symptom and how it changed during treatment. Specifically, to understand the response to pozelimab treatment and whether it was meaningful to patients, key quotes from the screening and week 24 interviews were used to contextualize COA change scores in specific concepts and the overall disease experience gathered during the trial, and these were compared to scores on the GIS. Change on the COAs from baseline to week 24 (mean, median, minimum, and maximum) was evaluated at the patient level to assess whether the participant’s sign or symptom improved, did not change, or worsened by week 24.

Finally, for each patient’s MBS, the associated sign- or symptom-specific COA scores and patient- or caregiver-reported GIS/GIC scores are summarized, along with patient/caregiver quotes about the MBS at screening and week 24 interviews as well as the meaningfulness of the change.

Results

Patient characteristics

Ten patients were enrolled in the clinical trial at three sites in Türkiye (n = 7), Thailand (n = 2), and the USA (n = 1). The mean age of patients was 9.3 years (standard deviation = 4.9), and more than half were female (n = 6; 60.0%). Additional demographic and health information, as well as trial efficacy and safety results, are presented elsewhere [14].

One patient completed both the screening and week 24 interviews independently, while the remaining interviews were conducted either as dyads (n = 5 at screening and n = 6 at week 24) or with the caregiver only (n = 4 at screening and n = 3 at week 24). Similarly, the COAs were completed by caregivers for most patients (n = 8). Following pozelimab treatment, all 10 patients met the composite primary endpoint of albumin normalization and improvement (or no worsening) in clinical signs and symptoms at week 24 (reported elsewhere). Table 2 summarizes the demographic information of the participants, as well as the respondent information specific to the interviews and COAs.

Table 2 Site-reported patient demographic information (n = 10)

Screening interview results

Prior to initiating treatment, nine patients (90.0%) reported experiencing all of the following signs and symptoms: abdominal pain, diarrhea, facial edema/swelling, nausea, peripheral edema/swelling, and vomiting; one patient (10.0%) reported experiencing all these except for nausea. Other signs and symptoms were reported by ≤ 3 patients. Abdominal pain, diarrhea, facial edema/swelling, nausea, peripheral edema/swelling, and vomiting were the signs and symptoms identified by the literature and clinicians at the trial planning stage as being central to the patient’s disease experience. Therefore, the results of the interviews confirmed that the six signs and symptoms selected for the COA measurement strategy were central to the patient’s disease experience, and were the core signs/symptoms of CHAPLE disease [15].

Abdominal pain was reported to be the MBS for nine out of 10 patients (90.0%); for the remaining patient (n = 1, 10.0%), facial edema was the MBS. Furthermore, before starting treatment most patients (n = 7, 70.0%) reported that abdominal pain was the most important symptom that they would like to see improved with treatment.

Week 24 interview results

All patients experienced a complete resolution of core signs and symptoms at the week 24 interview (Table 3). Moreover, 19 signs and symptoms were discussed in the context of change (including the six core signs and symptoms) [15], and nine participants (90.0%) had improvement in all signs and symptoms; the remaining patient (n = 1, 10.0%) experienced an improvement in all signs and symptoms except for one, the inability to gain weight, which did not change following treatment. Of note, no patients reported worsening of any signs and symptoms.

Table 3 Example quotes of overall improvement over the clinical trial

Overall improvement in signs and symptoms/disease activity at week 24

The interview findings align with the scores on the global assessments (Table 4). Specifically, all patients (N = 10, 100.0%) had no signs or symptoms at week 24 (as reported by both patients or caregivers and their clinicians), and nearly all patients indicated that their change from baseline for overall disease severity on the GIC was “much better” (n = 9, 90.0%). One patient (10.0%) had a GIC (reported by the caregiver and the clinician) of “a little better,” but it should be noted that the caregiver and clinician reported the patient’s baseline severity as mild. Of note, for a small number of patients there were some slight discrepancies at baseline between patient/caregiver assessments and clinician assessments. However, at Week 24, participant and clinician global ratings post-treatment were aligned for all patients; all patients or caregivers and clinicians reported a positive impact of treatment on disease activity.

Table 4 GIS and GIC for overall CHAPLE disease signs and symptoms

Table 5 presents the median change in scores on the COAs summarized across participants for each core sign or symptom of CHAPLE disease, along with example quotes. For abdominal pain, responses to the PedsQL™ Pain and Hurt subscale (i.e. abdominal pain) from baseline to week 24 indicated a median improvement of 43.8 points (N = 10; range, 0.0–91.7). For facial edema, the physician assessments from baseline to week 24 indicated a median improvement of one point (n = 10; range, 0–3). Eight patients (80.0%) had improved scores on the facial edema physician assessment by week 24; the two patients (20.0%) who did not have a change in their scores had baseline scores of “mild edema” and “no edema.” For the COA assessments for the other core signs and symptoms (diarrhea, peripheral edema/swelling, nausea, and vomiting), most patients (n ≥ 8) had scores that indicated improvement and/or resolution of the symptom from baseline at week 24. Those who did not report improvement either were not experiencing the sign or symptom at baseline or had reported the severity at baseline to be mild.

Table 5 Clinical outcome assessment response from baseline to week 24, and accompanying example quotes from the week 24 interviews

Resolution of MBS for each patient

Table 6 presents quotes from the interviews and patient-level COA scores associated with MBS from screening and week 24. These results demonstrate that, based on both the data from the interviews and the COA scores, all patients’ MBS improved or resolved following treatment. For nine patients (90.0%), data from the interviews and the COA scores demonstrate complete resolution of their MBS (i.e. a score of 100 on the PedsQL™ Pain and Hurt subscale and GIS rated as “absent”). For one patient, the caregiver reported during the interview that the MBS (abdominal pain) had resolved, indicating on the GIS that the child’s symptoms were “absent”; this patient’s PedsQL™ Pain and Hurt subscale score improved from 0 to 91.7, a nearly perfect score.

Table 6 Change in MBS over the clinical trial as assessed by the PedsQL™ GI Stomach Pain and Hurt subscale for abdominal pain, facial edema physician assessment, global assessments, and participant interview quotes

Discussion

CHAPLE disease is an ultra-rare condition [1, 4] for which pozelimab treatment was recently approved by the FDA [6]. Interviews with patients and/or caregivers were conducted as part of the pozelimab clinical trial to further elucidate the patient experience by contextualizing the change in COA scores and meaningfulness of the change, including signs and symptoms not explicitly measured in the trial. This is critical for drug evaluation because the FDA standard for a clinically positive effect of a new drug, i.e. efficacy, is whether it improves how a patient feels, functions, or survives [21]. Interviews completed at the screening visit confirmed that the signs and symptoms identified as core through a review of the literature [1, 3, 7] and discussion with treating clinicians (i.e. abdominal pain, facial and peripheral edema, diarrhea, nausea, and vomiting) were indeed central to the disease experience. Further, abdominal pain was identified by most patients as the MBS that they wanted to see improve with treatment. The week 24 interviews demonstrated that all but one sign improved following 24 weeks of pozelimab treatment, no sign/symptom worsened, and that all core signs/symptoms had completely resolved.

The COA measurement strategy for the trial was based on the concepts identified in the literature and discussed in meetings with treating clinicians, and there was therefore a risk that these concepts would not be confirmed during the within-trial interviews. However, concepts covered by the COAs included in the trial were identified as core signs and symptoms during the screening interviews. Consequently, the COA measurement strategy comprehensively captured CHAPLE disease. Further, an added benefit of the data gathered from the interviews was the opportunity to assess change in signs and symptoms that were not targeted for assessment via COAs, which helped to further contextualize the pozelimab treatment effects on the disease experience holistically. Specifically, the interview data contextualized meaningful within-person change on COA scores, which would have not been possible to evaluate quantitively given the small sample size.

The mixed method analyses specific to each participants’ MBS also demonstrated resolution of the single sign/symptom that was noted before treatment as being the most bothersome. While the use of MBS has been reported in the literature [12, 16, 22, 23] and in an FDA regulatory guidance document for migraine [17], to our knowledge this is the first use of the MBS approach in an ultra-rare disease trial. Nearly all participants experienced all six core signs/symptoms of disease before starting pozelimab and nine participants had abdominal pain as their MBS, suggestive of a homogenous experience of disease. Given the paucity of information on the patient experience during protocol development, use of the MBS approach could have been helpful in interpreting the results of the trial had the CHAPLE disease experience been more heterogenous.

Overall, interview results aligned with the changes in scores on the COAs and global assessments of CHAPLE disease activity from the perspective of the participants and treating physicians. When looking at the responses to the COAs in the clinical trial, findings are largely consistent with those from the interviews and provided further evidence of the meaningfulness of the change experienced after treatment.

The results of this study must be interpreted in the context of certain limitations. The COAs for the clinical trial were selected to overlap conceptually with the core signs and symptoms; however, there may be differences between patient- or caregiver-reported concepts and the concepts in the COAs, as well as the way that the concepts were reported and measured (e.g. different time frames/recall periods). Given the wide age range of patients, not all patients could be the primary respondent and questionnaires appropriate for caregiver report of symptoms were selected. During the qualitative interviews, however, caregivers were asked to explain how they were aware of the symptom and how they gauged severity. Caregivers reported both verbal and nonverbal ways the child communicated their symptoms and the severity of their symptoms [15]. Depending on age, global assessments were completed by either by caregivers or by patients (depending on age), as well as clinicians, and while caregivers/patients and clinicians rating of baseline assessments differed for some patients, all scores at week 24 demonstrated the resolution of signs/symptoms, including patients who had scores of “mild” before treatment. It should be noted that all patients enrolled in the clinical trial had active disease and hypoalbuminemia [14]. Finally, the trial design (single-arm) and sample size might be viewed as a limitation; however, it was considered adequate given the overall prevalence of CHAPLE disease (< 100 patients worldwide) and the life-threatening nature of the condition [14].

Conclusion

Findings from the within-trial interviews align with the change in COA scores and the patient or caregiver and clinical global assessments of CHAPLE disease activity. The mixed-methods approach allowed for a more comprehensive assessment of the patient experience in a trial evaluating a treatment for an ultra-rare and newly discovered disease by assessing the efficacy of treatment for signs and symptoms in the trial, including those that were not measured by COAs. Together, the results from the interviews and COAs comprehensively contextualized the patient experience and demonstrated the efficacy of pozelimab on the signs and symptoms of CHAPLE disease, as well as the transformative impact treatment had on the lives of patients. This approach may be especially useful for meeting the FDA requirements for drug efficacy, especially for rare and severe diseases.