FormalPara Key Summary Points

Why carry out this study?

Complement factor 5 (C5) inhibitors are the standard of care for paroxysmal nocturnal hemoglobinuria (PNH), which is a rare, acquired blood disorder characterized by complement-mediated hemolysis and thrombosis.

Real-world dosing studies found that some patients treated with the first-approved C5 inhibitor eculizumab benefit from increased dosing to control their PNH symptoms, while data about ravulizumab dosing are currently lacking.

This retrospective study was conducted to describe dosing patterns of ravulizumab, a newer, long-acting C5 inhibitor, among insurance claims from patients with PNH in the USA.

What was learned from the study?

As with eculizumab, potential deviations from label-recommended dosing regimens with ravulizumab were identified in this study, with a substantial proportion of patients receiving above label-recommended ravulizumab loading and maintenance doses.

Deviations from label-recommended ravulizumab dosing regimens occurring in real-world clinical practice suggest the need for further investigations of response to complement inhibitor therapy in patients with PNH.

Introduction

Paroxysmal nocturnal hemoglobinuria (PNH) is a rare, acquired blood disorder characterized by the specific degradation of blood cells that become vulnerable to complement-mediated hemolysis [1, 2]. While the exact prevalence of this rare disease is difficult to ascertain, global incidence is approximately 1–1.5 cases per million [1]. In an analysis of insurance claims from the USA in 2016 to 2017, the estimated prevalence was 12–13 per million, with an incidence rate of 5.7 per million person-years [3]. This estimate, along with other studies, suggests that there are approximately 5000–6000 patients with PNH in the USA [2, 4]. Most patients with PNH are aged 30–59 years and are often diagnosed at 30–40 years [1, 2, 5]. Clinical manifestations of PNH include hemolytic anemia, bone marrow failure, and thrombophilia, which may be life-threatening [1, 6]. Common symptoms include fatigue, dyspnea, abdominal pain, erectile dysfunction, and dysphagia, which together represent a significant disease burden [5, 7].

In 2007, the complement factor 5 (C5) inhibitor eculizumab became the standard of care for PNH in the USA [8,9,10]. Although effective at reducing C5-dependent intravascular hemolysis of red blood cells and improving overall symptoms [11,12,13,14], eculizumab treatment is associated with variable patient response. Approximately 11–27% of patients treated with the label-recommended eculizumab dosage continue to experience signs of ongoing extravascular hemolysis [14,15,16], and 87–90% may experience partial or suboptimal symptom control [17, 18]. Consequently, modified dosing regimens, including increases in dose or frequency of infusions, are not uncommon [13, 14, 16].

Ravulizumab, a long-acting C5 inhibitor with a reduced dosing frequency compared with eculizumab, was approved by the US Food and Drug Administration in 2018 for the treatment of adults with PNH and in 2021 for pediatric patients with PNH 1 month of age and older [19]. The approved dosing regimen of ravulizumab is based on the patient’s body weight and consists of a single loading dose, with maintenance doses starting 2 weeks after the loading dose and given every 8 weeks thereafter in patients weighing ≥ 20 kg. Compared with eculizumab, ravulizumab has been associated with a less burdensome dosing schedule [20,21,22]. Consequently, patients are more willing to initiate ravulizumab or to switch from eculizumab [23].

While the use of ravulizumab has been well documented in the setting of clinical trials [20,21,22, 24,25,26], knowledge regarding its use and dosing pattern in routine clinical practice remains limited [7, 27, 28]. This retrospective study was conducted to describe ravulizumab dosing regimens administered in real-world clinical practice among US patients with PNH.

Methods

Data Source

This study used provider-based claims data from the Symphony Health Integrated Dataverse (IDV®) database, which represents large proportions of US pharmacy, medical, and hospital claims and covers approximately 280 million beneficiaries each year. Symphony Health IDV is an ideal database to generate longitudinal summaries of patient treatment patterns across all payer types while providing a significantly large sample size to study a rare disease such as PNH. The study was performed in accordance with the Helsinki Declaration of 1964 and its later amendments. Institutional review board approval and patient consent to participate were not required for this study because all data used were de-identified and compliant with the US patient confidentiality requirements of the Health Insurance Portability and Accountability Act.

Study Design

This retrospective, longitudinal, cohort study of ravulizumab-treated patients with PNH included records from October 1, 2014 to May 6, 2021. The index date was defined as the first observed medical claim for ravulizumab infusion occurring on or after June 21, 2019, with a prior baseline period of ≥ 6 months of continuous clinical activity (defined as having ≥ 1 medical or prescription claim per quarter). Given the US approval of ravulizumab on December 21, 2018, the earliest index date that allowed for ≥ 6 months of baseline period was June 21, 2019. The observation period for continuous ravulizumab treatment started from the index date and ended at the earliest date of the last ravulizumab infusion prior to treatment discontinuation, end of continuous clinical activity, or end of data availability (May 6, 2021). Treatment discontinuation was defined as having a gap of > 126 days (representative of missing two consecutive 8-week periods, each with a 1-week grace period) between two ravulizumab infusions or between the last ravulizumab infusion and end of the observation period.

Sample Selection

Eligible patients had ≥ 2 medical claims for infusion of ravulizumab (identified using the Healthcare Common Procedure Coding System codes J1303 and C9052) between June 21, 2019 and May 6, 2021; ≥ 6 months of continuous clinical activity prior to the index date; and ≥ 12 years of age on the index date. Patients were excluded if they either had ≥ 1 medical claim for ravulizumab infusion before June 21, 2019, or ≥ 1 diagnosis code of atypical hemolytic uremic syndrome (International Classification of Diseases, Ninth Revision, Clinical Modification code 283.11 or International Classification of Diseases, Tenth Revision, Clinical Modification code D59.3) during the baseline period or on the index date. As atypical hemolytic uremic syndrome was another indication for ravulizumab at the time of the study, patients with a diagnosis code for this condition had to be excluded since patients with PNH were selected on the basis of receiving ravulizumab rather than on a specific diagnosis code of PNH, which was not available during the study period. This inclusion-by-exclusion principle has been applied in previous claim-based analyses of patients with PNH [29]. Since ravulizumab was approved for myasthenia gravis after the study period, it was not necessary to exclude patients with a diagnosis code for this condition.

All patients who met the eligibility criteria were included in the overall cohort. Patients initiating ravulizumab without prior eculizumab treatment during the baseline period were categorized into the eculizumab-naive cohort, whereas those who had received any eculizumab treatment during the same period were categorized into the eculizumab-experienced cohort.

Study Measures

Demographic characteristics evaluated at the index date included age, sex, US geographic region, and type of insurance plan. Clinical characteristics were assessed during the baseline period up to the day before index, including the Quan-Charlson Comorbidity Index (Quan-CCI) [30]; PNH-related comorbidities of aplastic anemia and myelodysplastic syndrome; PNH-related symptoms of anemia other than aplastic anemia, viral and bacterial infections, fatigue, chronic kidney disease, abdominal pain, dyspnea, thrombosis, pulmonary hypertension, dysphagia, and erectile dysfunction; and PNH- and anemia-related treatments (eculizumab, anticoagulants, corticosteroids, blood transfusions, immunosuppressants, iron supplementation, androgens, and iron-chelating agents). Specific diagnosis, prescription, and procedure codes used to identify these characteristics are listed in Supplementary Tables 1 and 2.

Infusion doses administered by health care providers and recorded precisely in insurance claims during the continuous ravulizumab treatment period were reported for the initial loading dose and the average of all maintenance doses administered 2 weeks after the loading dose and every 8 weeks thereafter. The label-recommended dosing of ravulizumab is based on patients’ body weights, with 2400 mg loading and 3000 mg maintenance doses for patients weighing ≥ 40 to < 60 kg, 2700 mg loading and 3300 mg maintenance doses for those weighing ≥ 60 to < 100 kg, and 3000 mg loading and 3600 mg maintenance doses for those weighing ≥ 100 kg [9]. The average number of days between maintenance doses was reported for patients with ≥ 2 ravulizumab infusions during the maintenance phase (continuous treatment period starting 2 weeks after the index date), with shorter intervals between infusions representing a proxy measure of dose increase [31]. Percentages of patients with dose escalation, defined as either an increase in dose or an increase in the frequency of infusions, were estimated.

In addition to the descriptive analysis using observed values in the claims data, three sets of scenario-based analyses were performed to estimate the overall percentages of patients who received the loading and average maintenance doses below, at, and above label-recommendations, with statistical adjustments applied according to the 2015–2018 weight distribution among men and women between the ages of 40 and 49 years from the US general population since patients’ body weights were not available in the Symphony Health IDV database. First, results across the overall and eculizumab cohorts were generated assuming all the patients in the cohort belonged to the same label-defined weight category of either ≥ 40 to < 60 kg (low body weight, scenario 1), ≥ 60 to < 100 kg (medium body weight, scenario 2), or ≥ 100 kg (high body weight, scenario 3). These three empirical scenarios were then aggregated according to gender and body weight distribution data of the US general population reported by the National Center for Health Statistics [32] and the National Health and Nutrition Examination Survey [33], respectively, to generate statistically adjusted estimates, with 0% men and 15% women in the low weight category, 75% men and 70% women in the medium weight category, and 25% men and 15% women in the high weight category [32, 33].

Statistical Analysis

Descriptive summary statistics (mean and standard deviation [SD], median and interquartile range [IQR] for continuous variables, and numbers and percentages of patients for categorical variables) were generated for baseline patient characteristics and ravulizumab infusion doses. Patient characteristics during the baseline period were compared between the eculizumab-naive and eculizumab-experienced cohorts using standardized differences, with values greater than 0.2 indicating small but significant effect size [34]. All analyses were conducted using SAS Enterprise Guide Version 7.1 (Cary, NC).

Results

Baseline Characteristics

Of 798 patients with ≥ 2 medical claims of ravulizumab infusion between June 21, 2019 and May 6, 2021, 433 (54.3%) met all subsequent criteria for inclusion and were included in the overall cohort (Fig. 1). Among the overall cohort, 198 patients (45.7%) were eculizumab-naive, and 235 (54.3%) were eculizumab-experienced.

Fig. 1
figure 1

Patient disposition. HCPCS Healthcare Common Procedure Coding System, ICD-9-CM International Classification of Diseases, Ninth Revision, Clinical Modification, ICD-10-CM International Classification of Diseases, Tenth Revision, Clinical Modification, PNH paroxysmal nocturnal hemoglobinuria. aMedical claims for ravulizumab infusion were identified using the HCPCS procedure codes J1303 and C9052. bRavulizumab received approval from the US Food and Drug Administration on December 21, 2018 for the treatment of PNH. The start date of June 21, 2019 was selected to allow for ≥ 6 months of baseline period; the end date of May 6, 2021 represents the end of data availability. cClinical activity was measured in increments of calendar year quarters and was used as a proxy for health plan enrollment given the lack of eligibility files in the Symphony Health Integrated Dataverse (IDV®) database; a patient was considered active for the quarter if there was ≥ 1 medical or prescription claim during the quarter. dAt the time of the study, ravulizumab was also indicated for atypical hemolytic uremic syndrome (identified by ICD-9-CM code 283.11 and ICD-10-CM code D59.3) in adults and children 1 month of age or older. eMedical claims for eculizumab infusion were identified using the HCPCS procedure code J1300

Full size image

Demographic and clinical characteristics evaluated at the index date (first recorded ravulizumab infusion) are summarized in Table 1. The overall mean (SD) age was 47.4 (17.0) years, with 46.5 (16.0) years among eculizumab-naive and 48.3 (17.9) years among eculizumab-experienced patients. Across all cohorts, more than half of all patients were female and had their index date during the year 2019. The overall mean (SD) Quan-CCI was 0.4 (1.1) and was slightly lower in the eculizumab-naive cohort than in the eculizumab-experienced cohort (0.3 [1.0] vs. 0.5 [1.1], standardized difference 0.109). Aplastic anemia and myelodysplastic syndrome were observed at a respective frequency of 21.7% and 4.8% in the overall cohort, and at similar frequencies across both eculizumab cohorts. The most common sign of PNH was anemia (39.7% in the overall cohort), with a greater frequency among eculizumab-experienced patients (43.8%) than eculizumab-naive patients (34.8%, standardized difference 0.185). Fatigue, observed among 10.2% of the overall cohort, also occurred at a greater frequency among eculizumab-experienced patients than eculizumab-naive patients (12.3% vs. 7.6%, standardized difference 0.160). Anticoagulants were used by 20.8% of the overall cohort, and more frequently by eculizumab-experienced patients than eculizumab-naive patients (25.1% vs. 15.7%, standardized difference 0.236), whereas corticosteroids were used frequently and equally across cohorts (17.8% overall, 17.9% for eculizumab-experienced vs. 17.7% for eculizumab-naive, standardized difference 0.005). Blood transfusions were also common (15.5% overall, 17.4% vs. 13.1% in the eculizumab-experienced and eculizumab-naive cohorts, standardized difference 0.120).

Table 1 Baseline demographic and clinical characteristics
Full size table

Observed Dosing Patterns

The observed ravulizumab dosing patterns in the overall and eculizumab cohorts are summarized in Table 2. The overall mean (SD) duration of continuous ravulizumab treatment was 11.8 (6.9) months. The mean (SD) loading dose in the overall cohort was 3316.3 (2931.7) mg, which exceeds the maximal label-recommended loading dose of 3000 mg for patients weighing ≥ 100 kg by slightly more than 10%. The median (IQR) loading dose was 3000 (2700, 3300) mg. Overall, 48.0% patients received a loading dose greater than the maximal label-recommended loading dose (> 3000 mg) and 6.5% received a loading dose inferior to the minimal label-recommended loading dose (< 2400 mg). Mean (SD) duration of continuous ravulizumab treatment among the eculizumab-naive and eculizumab-experienced patients were 11.3 (6.7) and 12.3 (7.0) months, respectively. The mean and median loading doses were greater among eculizumab-naive patients than eculizumab-experienced patients, as were the percentages of those receiving loading doses > 3000 mg (63.6% vs. 34.9%, respectively).

Table 2 Ravulizumab observed dosing patterns
Full size table

Among patients with ≥ 1 maintenance dose during the continuous ravulizumab treatment period (n = 423), the mean (SD) number of infusions was 5.8 (3.2), with a mean (SD) number of days between infusions of 57.6 (6.4) days (Table 2). The number of infusions and time between infusions corresponded to the label-recommended 8-week period between maintenance infusions over 12 months. On average over the continuous ravulizumab treatment period, the mean (SD) average maintenance dose per infusion was 3403.7 (1024.4) mg, falling between the label recommendations of 3300 mg for patients weighing ≥ 60 to < 100 kg and 3600 mg for those weighing ≥ 100 kg; the median (IQR) of the average maintenance dose per infusion was 3300.0 (3300, 3506) mg. Slightly more than half of patients (53.0%) also received an average maintenance dose of 3300 mg over the continuous ravulizumab treatment period, although 33.8% had at least one maintenance dose escalation. Among eculizumab-naive (n = 190) and eculizumab-experienced patients (n = 233) with ≥ 1 maintenance dose during the continuous ravulizumab treatment period, the number of infusions and the number of days between infusions were comparable. The mean average maintenance doses were slightly lower in those who were eculizumab-naive compared with eculizumab-experienced patients.

Scenarios of Estimated Dosing Patterns

After adjustments based on distribution of body weights in the general US population, 23.1% of patients overall (n = 433) were estimated to have received the label-recommended loading dose, whereas 17.7% were estimated to have received less than the label-recommended loading dose and 59.1% were estimated to have received more than the label-recommended loading dose (Table 3). During the maintenance phase, 39.2% of patients (n = 433) were estimated to have received the label-recommended maintenance dose on average, whereas 32.4% were estimated to have received below and 28.4% above label-recommended doses on average. At least one episode of dose escalation was observed in 33.8% (n = 143) of patients, among which large percentages of patients were estimated to have received above label-recommended loading (80.5%) and maintenance (84.7%) doses on average.

Table 3 Statistically adjusted ravulizumab dosing patterns based on scenario analyses and reported body weights for the general US population
Full size table

The statistically adjusted results of ravulizumab dosing patterns across the eculizumab cohorts are summarized in Table 4. An estimated 73.9% of eculizumab-naive and 46.7% of eculizumab-experienced patients were estimated to have received an above label-recommended loading dose. The differences between eculizumab cohorts in above label-recommended average maintenance doses were smaller (24.8% and 31.4% among eculizumab-naive and eculizumab-experienced patients, respectively) than for the loading doses, as were the estimated percentages of patients with ≥ 1 episode of dose escalation (30.6% and 36.5%, respectively).

Table 4 Statistically adjusted ravulizumab dosing patterns by eculizumab experience cohort based on scenarios of body weights for the general US population
Full size table

Discussion

This study explored ravulizumab dosing patterns among patients with PNH treated in US real-world clinical practice, including those with and without prior eculizumab treatment. Findings suggest that a substantial proportion of patients may have received a greater than label-recommended ravulizumab loading dose regardless of label-defined weight categories, as reflected in almost half of patients receiving loading doses greater than the maximal label-recommended loading dose for patients weighing ≥ 100 kg. In comparison, the observed average maintenance doses were within the range of label-recommended maintenance doses for the medium label-defined weight category, which corresponds to the reported average body weight (69–78 kg) of patients with PNH enrolled in clinical trials of ravulizumab [20, 21, 35].

Above label-recommended dosing has been described in clinical trials of eculizumab to manage breakthrough hemolysis events, and anemia-related laboratory signs or symptoms (such as fatigue) [13, 14, 16]. A retrospective claims-based analysis (2015–2019) assessing real-world eculizumab treatment patterns found that 68% and 43% of 707 patients received a greater than label-recommended average loading dose (600 mg) and average maintenance dose (900 mg), respectively [36]. In a 2021 retrospective chart review of 100 US patients with PNH, almost 21% of those treated with eculizumab (n = 53) had their last maintenance dose at 1200 mg, 300 mg above the label-recommended maintenance dose [9, 28].

With its longer half-life and reduced dosing frequency compared with eculizumab, ravulizumab has the potential to achieve more sustained C5 inhibition through weight-based dosing and reduce the risk of breakthrough hemolysis associated with suboptimal C5 inhibition [10, 37, 38]. Although limited evidence is available regarding ravulizumab dosing in real-world clinical settings, it is possible that increased eculizumab dosing to treat patients with PNH may have carried over to the treatment of patients with PNH using ravulizumab. In contrast to the literature, our findings suggest that some providers may opt to use a loading dose equivalent to the label-recommended maintenance dose of ravulizumab when treating patients who are receiving high doses of eculizumab to deliver enough circulating antibodies up-front and reduce the risk of breakthrough hemolysis between the ravulizumab loading and maintenance doses [27].

Our findings of ravulizumab dosing patterns are in line with those from a 2020 cross-sectional survey investigating the impact of eculizumab and ravulizumab treatment in 122 US patients with PNH [39]. Among the 87 patients treated with ravulizumab, 33% in the 40–60 kg (low body weight) category and 18% in the 60–100 kg (medium body weight) category reported receiving greater than label-recommended maintenance doses for their weight category. Similarly, 31% of the 35 patients treated with eculizumab reported receiving greater than label-recommended doses (> 900 mg) [39]. Despite treatment with eculizumab or ravulizumab, many of these patients continued to experience signs and symptoms of PNH such as fatigue, anemia, and thrombotic events, and rely on blood transfusions, suggesting that their disease remained uncontrolled [39]. The reasons behind the ravulizumab dosing patterns observed in our study could not be ascertained because of the unavailability of patient clinical data. For example, patients with clinical conditions inducing strong complement activation, such as infections or major surgeries, which could potentially lead to breakthrough extravascular hemolysis, might benefit from receiving increased doses of a C5 inhibitor [38]. On the other hand, pilot pharmacokinetic studies showed that eculizumab label-recommended dosing may be sufficient when used for other indications (namely atypical hemolytic uremic syndrome), suggesting the possibility to administer below label-recommended eculizumab dosing to reduce costs or improve dosing schedule convenience [40,41,42].

The greater than label-recommended loading doses observed in this study were driven by greater loading doses administered to eculizumab-naive patients compared with eculizumab-experienced patients. The reason for the difference in loading doses could not be ascertained in this study, as the baseline demographics and clinical characteristics were similar between the two cohorts. In a 2021 retrospective chart review study [28], reasons for C5 inhibitor treatment initiation were explored among patients receiving only eculizumab (n = 53) or ravulizumab (n = 32), and among those who switched from eculizumab to ravulizumab (n = 15). Slightly greater percentages of patients initiated ravulizumab compared with eculizumab as a result of a hemolytic crisis requiring hospitalization (22% vs. 19%), or to prevent complications (3% vs. 2%) or thrombosis (3% vs. 0). Among 15 patients switching from eculizumab to ravulizumab, the most frequent reasons for switching were related to treatment preference and convenience (53% to 73%), but also as a result of increased markers of disease (13% to 27%), improved quality of life (27%), symptoms of hemolysis (13%), improved symptom control (7%), and previous thromboembolism (7%) [28]. Therefore, it is possible that eculizumab-naive patients observed in this study were administered greater than label-recommended loading doses to overcome any of these potentially serious events.

Limitations

The lack of patient body weight in Symphony Health IDV posed a methodological challenge given that ravulizumab dosing regimen is based on patient’s body weight. Using various scenarios allowed results to be statistically adjusted on the basis of body weight distribution among the general US population, an approach that has been used in a cost-utility analysis [43]. Although patients with PNH who frequently experience episodes of hemolysis would be expected to have lower body weights than the general US population, other conditions impacting body weight, such as cancer or inflammatory bowel diseases, were also present in the general population used as a reference in this study (i.e., the National Health and Nutrition Examination Survey) [32, 33]. The average body weight of patients with PNH initiating ravulizumab in clinical trials (69–78 kg) [20, 21, 35] and in real-world clinical settings (77 kg) [39] is similar to the average of the US population for the same age range (80 kg for women and 94 kg for men) [32]. Furthermore, in our modeling analysis, adjustment may have overestimated the proportion of individuals with high body weights in our study, which would result in an underestimation of patients receiving greater than label-recommended doses.

Other inherent limitations to administrative claims databases include the lack of clinical and behavioral predictors of continued use, dose escalations, or reason for discontinuations. Patients may have received dose escalation because of vial size availability, increasing body weight secondary to corticosteroid use, clinical conditions activating the complement (e.g., infections, surgery, pregnancy), or lack of treatment response; however, this distinction could not be captured in this analysis owing to the unavailability of patient clinical data. The impact of clinical conditions on dosing patterns and longitudinal variations in dose over time could be addressed in future studies. Specific limitations of the Symphony Health IDV are that it does not include eligibility files, so quarterly claims activity was used as a proxy for health plan enrollment, and the provider-based structure (vs. insurance-based) may have resulted in some patients being considered more than once, particularly if they received care from different doctors in the same office. Finally, omissions, misclassifications, and/or other coding errors with diagnosis or procedural codes may have occurred as in any claims-based analysis.

Conclusions

This study aimed to assess ravulizumab dosing patterns in real-world clinical practice. Similar to treatment with eculizumab, potential deviations from label-recommended dosing regimens were identified on the basis of estimation of patients’ body weights using body weight distribution reported for the US general population. Findings suggest that deviations from label-recommended dosing regimens occur with ravulizumab dosing patterns in real-world clinical practice, highlighting the need for further investigations of treatment response in patients with PNH who receive complement inhibitor therapy.