Hyaluronidase‐facilitated subcutaneous immunoglobulin 10% as maintenance therapy for chronic inflammatory demyelinating polyradiculoneuropathy: The ADVANCE‐CIDP 1 randomized controlled trial

ADVANCE‐CIDP 1 evaluated facilitated subcutaneous immunoglobulin (fSCIG; human immunoglobulin G 10% with recombinant human hyaluronidase) efficacy and safety in preventing chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) relapse.

Interpretation: fSCIG 10% more effectively prevented CIDP relapse than placebo, supporting its potential use as maintenance CIDP treatment.

K E Y W O R D S
ADVANCE-CIDP 1 randomized controlled trial, chronic inflammatory demyelinating polyradiculoneuropathy, efficacy, hyaluronidase-facilitated subcutaneous immunoglobulin 10%, safety

| INTRODUCTION
Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is a rare immune-mediated neurological disorder in which the immune system attacks the myelin sheath resulting in weakness, impaired sensation, pain, fatigue, and a significant impact on functional ability. [1][2][3][4] The European Academy of Neurology and Peripheral Nerve Society joint guidelines recommend systemic corticosteroids or intravenous immunoglobulin (IVIG) for first-line treatment of CIDP with disabling symptoms, both for induction and maintenance of response. 5 These guidelines also recommend subcutaneous immunoglobulin (SCIG) as an alternative maintenance therapy in IVIG-responsive patients with active disease. 5 In addition to the substantial physical, psychosocial, and economic burden associated with CIDP, 2,6 existing recommended treatments have limitations including tolerability issues and administration challenges. 7 HyQvia, a facilitated subcutaneous immunoglobulin (fSCIG) therapy (Baxalta US Inc., a Takeda company) comprises a dual-vial unit of immunoglobulin G (IgG) 10% (GAMMAGARD LIQUID, Baxalta US Inc.; KIOVIG, Takeda Manufacturing Austria AG) and recombinant human hyaluronidase (rHuPH20). rHuPH20 depolymerizes hyaluronan in the extracellular matrix, transiently increasing the permeability of subcutaneous tissue to IgG. This allows high-volume IgG administration (equivalent to volumes administered intravenously) into the subcutaneous tissue over a short time. fSCIG 10% is approved in the EU as immunoglobulin replacement therapy for adults and children (aged 0-18 years) with primary or secondary immunodeficiency diseases, 8 and in the USA for the treatment of primary immunodeficiency diseases in adults and children aged 2 years and above. 9 fSCIG 10% combines the benefits of IVIG and SCIG, may have fewer systemic adverse reactions than IVIG, [10][11][12] and offers the opportunity for selfadministration at home. Additionally, fSCIG 10% enables the administration of large infusion volumes with high infusion rates.
Hence, for equivalent monthly doses, less frequent infusions and fewer infusion sites (up to three per infusion) are required for fSCIG 10% compared with conventional SCIG. 13 Here we report results from the phase 3 placebo-controlled trial, ADVANCE-CIDP 1, which evaluated the efficacy, safety, and tolerability of fSCIG 10% as a maintenance therapy to prevent disease relapse leading to neuromuscular disability in patients with CIDP.  Table S1). Following a screening and baseline period (≤8 weeks), the study comprised 2 "epochs" (Epochs 1 and 2; Figure S1), with the results of Epoch 1 presented here (hereafter referred to as ADVANCE-CIDP 1).
In ADVANCE-CIDP 1, patients were randomized 1:1 to fSCIG 10% or placebo (0.25% albumin solution together with rHuPH20), for a period of 6 months or until relapse or withdrawal from the study. fSCIG 10% was administered at the same dose as the patient's prerandomization monthly equivalent IVIG dose, or matching infusion volume for those receiving placebo, and at the same interval as prior IVIG (maximum 4-weekly administration). Patients receiving placebo who did not experience relapse remained blinded to study treatment and were permitted to enter an ongoing, long-term, open-label extension study (ADVANCE-CIDP 3), during which all patients received fSCIG 10%. Further information regarding dosing, administration, and infusion rates is provided in Appendix S1, Table S2, and Appendix S2, respectively.
Additional details of study blinding are provided in Appendix S3, and details of protocol amendments are given in Appendix S4.  15 Adjusted INCAT disability scores (used for assessment of the primary endpoint) were identical to INCAT disability scores (scored 0-10, with higher values indicating increasing inability to make purposeful movements), with the exception that upper extremity score changes from 0 to 1 (normal to minor symptoms) or from 1 to 0 were excluded. Secondary efficacy outcomes included the proportion of patients experiencing functional worsening, time to relapse, and least-squares mean change from baseline in Rasch-built Overall Disability Scale (R-ODS) centile scores. 16 Functional worsening was a composite endpoint defined as the occurrence of at least 1 of the following: a ≥ 1-point increase relative to baseline in 2 consecutive adjusted INCAT disability scores; a ≥ 8 kPa decrease in hand grip strength (in the more affected hand) measured using the Vigorimeter (Martin), or a ≥ 4-point decrease in R-ODS raw summed scores. For hand grip strength and R-ODS scores, changes were relative to the baseline at withdrawal from fSCIG 10% treatment. The R-ODS score, used to assess activities of daily living, was a centile metric score with lower scores reflecting more severe limitations. 16 Aside from the composite endpoint, the other R-ODS score outcome used the centile score.

| Patient population
Pre-infusion R-ODS scores were recorded at least once weekly throughout the study in electronic diaries (DIARYpro) provided to patients during screening. Manuals containing detailed instructions for use were provided to study sites and patients.

| Infusion characteristics
The DIARYpro tool was also used to capture data on the characteristics of infusions, namely mean and median monthly dose equivalents, mean duration of infusion per dose, dosing interval, and the number of sites used per infusion. HRQoL. 17,18 The EQ-5D also includes a visual analog scale (EQ-VAS) to indicate general health status, scored between 0 and 100 (higher scores indicate better health). SF-36 and EQ-5D scores were measured during the baseline period and at the final treatment visit.
Patient treatment satisfaction and treatment preference were evaluated using the self-administered abbreviated 9-Item Treatment Satisfaction Questionnaire for Medication (TSQM-9) and a nonvalidated treatment preference questionnaire, respectively. The TSQM-9 was used to assess effectiveness, convenience, and global satisfaction, with higher scores representing increased satisfaction. 19,20 The treatment preference questionnaire was used to assess patient preference for various attributes of fSCIG 10% therapy, such as ease of administration, frequency and duration of administration, and convenience. The TSQM-9 and treatment preference questionnaire were administered during the baseline period and at the final treatment visit for patients receiving fSCIG 10% or placebo. Development of binding and neutralizing anti-rHuPH20 antibodies was assessed using anti-rHuPH20 antibody (GCL-612) and
Neutralizing antibodies were assessed following a binding antibody titer of ≥1:160. All AEs and related AEs in patients with titers ≥1:160 were summarized.

| Sample size
It was originally planned in 2015 to randomize 174 patients 1:1 to fSCIG 10% or placebo to detect a difference in relapse rates of 18% with a power of $80% at the two-sided 5% significance level. This was based on a 15% drop-out rate, and assumed relapse rates of 7% and 25% for fSCIG 10% and placebo, respectively. 21 Owing completely to external factors, such as the COVID-19 pandemic and slow recruitment, the trial was stopped on blinded data by the sponsor in 2022 with 138 patients randomized in the study (Appendix S4). This was supported by more recent scientific literature suggesting a larger treatment difference based on expected relapse rates of 10% and 39% for fSCIG 10% and placebo, respectively. 22,23 Prior to stopping, and under these modified assumptions, 120 randomized patients (i.e., 60 per group and assuming a 15% drop-out rate) were estimated to provide 90% power to detect a treatment difference of 29% at the two-sided 5% significance level.

| Analysis cohorts
The primary efficacy analysis was conducted in the modified intention-to-treat analysis set, which included all randomized patients who received any double-blind study medication. Analysis of additional efficacy and patient-reported outcomes was also performed in the modified intention-to-treat analysis set. The safety set included all patients who received any double-blind study medication, and was used for analysis of safety outcomes. The prespecified sensitivity perprotocol analysis for the primary endpoint included all randomized patients who received any double-blind study medication without major or critical protocol deviations during the study, which may have significantly affected the primary outcome measure.

| Methods of analysis
The primary efficacy analysis compared the relapse rates in the 2 treatment groups using a continuity-corrected χ 2 test conducted at the 5% level of statistical significance, with a missing relapse outcome imputed as no relapse. Estimated relapse rates in each treatment group and the difference in relapse rates between the treatment groups were determined, along with the 95% confidence intervals (CIs). 24,25 Several additional prespecified sensitivity analyses of the primary endpoint were also performed to assess the impact of imputing missing relapse data as "no relapse" and the requirement for a confirmatory INCAT assessment. Prespecified sensitivity analyses included comparisons of the relapse rates in: (1) the modified intent-to-treat (MITT) analysis set with missing or incomplete relapse data imputed as relapse; (2) a MITT observed cases analysis with missing relapse outcomes excluded; (3) the per protocol set with missing relapse outcomes imputed as no relapse; and (4) the MITT analysis set with missing relapse data imputed as no relapse, where relapse was alternatively defined as an increase in adjusted INCAT disability score of ≥1 point relative to the presubcutaneous treatment baseline score, but on a single INCAT assessment only. The fourth sensitivity analysis removed the requirement for the increase to be confirmed at a secondary confirmatory INCAT evaluation to classify a patient as having relapsed. All sensitivity analyses used similar statistical methods to the primary analysis. Two additional ad hoc sensitivity analyses of the primary endpoint were also performed to further assess the impact of missing relapse outcomes data: (1) a sensitivity analysis that imputed data for patients with a missing second confirmatory INCAT score in a setting of clinical deterioration as "relapse"; and (2) a multiple imputation analysis under the missing at random premise via multiple imputation chained equations. 26 Time to relapse was compared between treatment groups using the generalized Wilcoxon (i.e., Gehan's) survival test, 27 with survival functions estimated using the Kaplan-Meier method. An ad hoc time to relapse analysis with missing relapse outcomes imputed as relapse was also performed. The change in R-ODS centile scores from baseline to end of treatment was analyzed using an analysis of covariance (ANCOVA) model to test the treatment effect, with baseline R-ODS score as a covariate. No adjustments for multiplicity were performed for any efficacy evaluation. Change from baseline in adjusted INCAT disability scores, hand grip strength scores, and MRC sum scores were summarized using descriptive statistics. No statistical testing was performed for tertiary efficacy outcomes. SF-36, EQ-5D, and TSQM-9 scores and treatment preference responses were all summarized using descriptive statistics. AEs were coded using the Medical Dictionary for Regulatory Activities version 24.1 and were summarized descriptively.   analyses supported the findings of the primary analysis, showing consistent reductions in relapse with fSCIG 10% over placebo (Table 2A).

| Ethics statement
The two additional ad hoc sensitivity analyses also supported the primary analysis. When missing relapse outcomes were imputed as relapse for patients who had a missing confirmatory INCAT assessment in a setting of clinical deterioration in the MITT population, the treatment difference between fSCIG 10% and placebo was statistically significant (À16.9% [95% CI: À30.30%, À2.40%; p = .0373]).
The multiple imputation analysis with missing outcomes imputed

Analysis
F I G U R E 1 Study disposition. CIDP, chronic inflammatory demyelinating polyradiculoneuropathy; fSCIG, facilitated subcutaneous immunoglobulin; INCAT, Inflammatory Neuropathy Cause and Treatment; IVIG, intravenous immunoglobulin. a Patients were excluded for not meeting inclusion criteria (n = 21) for the following reasons: did not meet diagnosis criteria (n = 4); did not meet prior treatment criteria (n = 5); did not meet INCAT disability score criteria (n = 5); unwilling/unable to comply with protocol requirements (n = 7). Patients were also excluded for meeting exclusion criteria (n = 25) for the following reasons: presence of disease potentially affecting assessment (n = 2); presence/history of exclusionary conditions/infections (n = 14); hypersensitivity/allergy to study treatments (n = 3); abnormal laboratory values (n = 4); treatment with corticosteroids within 8 weeks of screening (n = 1); condition judged to impede participation/pose patient risk/confound study results (n = 1). b Patients who relapsed and entered Epoch 2 were not included in the number of patients who completed Epoch 1 nor in the number of patients who discontinued early from Epoch 1. c CIDP relapse per protocol definition was recorded as the reason for early discontinuation under the following categories: fSCIG 10% treatment armphysician decision (n = 2); placebo armphysician decision (n = 1), withdrawal by patient (n = 1), "other" (n = 3). d 2 patients relapsed per protocol definition and decided not to enroll in Epoch 2; 1 patient relapsed per protocol definition and did not enroll in Epoch 2 because of "sponsor decision: protocol deviation" (patients used IVIG outside of study).   under the missing at random premise using multiple imputation chained equations estimated a treatment difference of À17.8% in relapse rates between fSCIG 10% and placebo (95% CI: À31.24%, À2.80%), indicating a statistically significant (at the two-sided 5% level) lower relapse rate with fSCIG 10% than with placebo (as the CI did not contain zero).
The proportion of patients who experienced functional worsening (composite secondary efficacy endpoint) was 37.5% in the fSCIG 10% group and 54.4% in the placebo group, with a treatment difference of À16.9% (95% CI À33.02%, 0.69%; Table 2B). Probability of relapse was higher with placebo versus fSCIG 10% over time (Table 2B and

| Patient-reported outcomes
Patients receiving fSCIG 10% showed favorable changes from baseline in both SF-36 and EQ-5D scores compared with patients receiving placebo (

| Safety analysis
In total, 491 AEs were reported in 89 patients during the study (  fSCIG 10% Placebo F I G U R E 2 Kaplan-Meier curves for time to relapse. CIDP, chronic inflammatory demyelinating polyradiculoneuropathy; fSCIG, facilitated subcutaneous immunoglobulin; MITT, modified intention-to-treat. Curves estimated using the Kaplan-Meier method for the MITT population, with missing outcomes imputed as no relapse. Time to relapse was calculated as: date of relapse À date of initial dose of treatment +1. Patients who did not relapse were censored with time to censoring calculated as: date of discontinuation or completionÀdate of initial treatment +1. ADVANCE-CIDP 1 was a double-blind, placebo-controlled phase, in which patients were randomized 1:1 to receive either fSCIG 10% or placebo for a period of 6 months or until relapse. T A B L E 3 Patient-reported outcomes.  (Table S5). More patients receiving fSCIG 10% experienced AEs than those receiving placebo (79.0% vs. 57.1% [ Table 4]). However, the majority of AEs in the fSCIG 10% group were mild or moderate, local, did not require suspension of infusions, and resolved without sequelae (Table S5).
Causally related AEs occurred in 19 patients (27.1%) in the placebo group and 38 patients (61.3%) in the fSCIG 10% group. The most common (reported in >5% of patients) causally related AEs included headache and nausea, as well as local AEs including infusion site pain, erythema, pruritis, and edema. Overall, 7 patients (5.3%) reported serious AEs with a lower rate of occurrence in the fSCIG 10% group (3.2%) than in the placebo group (7.1%) ( Table 4). In total, 1247 infusions were administered during the study, of which 72 (5.8%) were interrupted, stopped, or had the infusion rate reduced, with <1% of infusions affected by intolerability and/or AEs (Table S5).

| DISCUSSION
A major goal of maintenance treatment in CIDP is to prevent a relapse and thereby maintain neuromuscular function and quality of life.
ADVANCE-CIDP 1 showed that fSCIG 10%, when administered at the same dose and interval as prior IVIG therapy, is more effective proportions observed in the placebo group can potentially be attributed to those patients that were likely to be in remission.
fSCIG 10% administration is a 2-step process using a subcutaneous needle/infusion set to infuse rHuPH20, followed by immunoglobulin 10%. rHuPH20 is a highly purified human hyaluronidase that modifies connective tissue permeability through the hydrolysis of hyaluronan, temporarily decreasing the viscosity of the extracellular matrix and promoting dispersion of infused fluids, facilitating their absorption. 34-36 ADVANCE-CIDP 1 examined the immunogenic potential of rHuPH20. A total of 8 patients developed nonneutralizing binding anti-rHuPH20 antibodies (≥1:160), which were not associated with an increased incidence of AEs, or local or systemic reactions. The lack of clinical relevance of binding anti-rHuPH20 antibodies in ADVANCE-CIDP 1 is consistent with an analysis of a study of fSCIG 10% for primary immunodeficiency diseases, as well as an extensive review of hyaluronidase-conjugated antibody therapeutics in a multitude of diseases. 37,38 This study was affected by the COVID-19 pandemic and was closed to recruitment before the target sample size was achieved. This might have limited the characterization of the safety profile. However, the long-term safety profile of fSCIG 10% in CIDP is currently being explored in the ongoing ADVANCE-CIDP 3 trial, which is the longest extension study of its kind, with up to 6 years of follow-up data for some patients. Other recruitment challenges were common to other CIDP trials, such as low disease prevalence, competition for recruitment, patient unwillingness to discontinue existing treatments, and the reluctance of physicians and patients to receive placebo given the existence of efficacious therapies. Relapse status was missing for 8 patients overall, with more missing data in those receiving active treatment. However, prespecified sensitivity analyses evaluating the impact of missing data generally supported the primary analysis. In addition, although treatment preference for fSCIG 10% was high, this finding may be biased given the clinical trial setting, and thus may limit the wider generalizability of this finding in real-world practice.
In conclusion, ADVANCE-CIDP 1 met its primary endpoint. It demonstrated that fSCIG 10%, used as a maintenance therapy in a patient population with stable CIDP receiving intravenous immunoglobulin treatment, was more effective than placebo in preventing relapse of neuromuscular disability. Both fSCIG 10% and placebo subcutaneous infusions were well tolerated and preferred by most patients over their previous intravenous treatment.

ACKNOWLEDGMENTS
The authors thank the patients who participated in this trial, their caregivers, study-site personnel, and the investigators.