Subcutaneous Ocrelizumab in Patients With Multiple Sclerosis

Abstract

Background and Objectives

IV-administered ocrelizumab (OCR) is approved for the treatment of relapsing and primary progressive multiple sclerosis (RMS/PPMS). OCARINA II (NCT05232825) was designed to demonstrate noninferiority in drug exposure of OCR subcutaneous (SC) vs IV administration.

Methods

This phase 3, randomized, open-label study enrolled OCR-naive patients aged 18–65 years with RMS/PPMS and an Expanded Disability Status Scale score of 0–6.5. Patients received OCR IV 600 mg or OCR SC 920 mg (controlled period), followed by OCR SC 920 mg every 24 weeks, up to week 96 (OCR IV/SC and OCR SC/SC). The primary end point was OCR area under the serum concentration–time curve from day 1 to week 12 (AUC_W1‒12); other end points included clinical, biomarker, and pharmacodynamic outcomes and safety data.

Results

Baseline demographics were balanced across OCR IV/SC and OCR SC/SC arms (N = 118/118, 40.0 ± 11.9/39.9 ± 11.4 years, 59.3%/65.3% female, 89.0%/89.0% with RMS). The study demonstrated noninferiority of OCR SC 920 mg to OCR IV 600 mg for the primary end point AUC_W1–12 and also over the dosing interval for AUC_W1‒24 (geometric mean ratios [90% CI] 1.29 [1.23–1.35] and 1.27 [1.21–1.34], respectively). At week 48, 111 of 118 (OCR IV/SC) and 114 of 118 (OCR SC/SC) had received OCR SC. A near-complete suppression of MRI activity was reported in OCR IV/SC and OCR SC/SC: 0 of 113 and 0 of 113 patients had T1 lesions while 1 of 114 and 1 of 113 had 2 and 1 new/enlarging T2 lesions, respectively. Two patients (1.9%) in each arm had 1 relapse, and 1 patient (0.9%; OCR SC/SC) had 2 relapses. In both arms, rapid and sustained B-cell depletion was observed and serum neurofilament light chain reduction was comparable. Patients receiving at least 1 dose of OCR SC 920 mg in the OCR IV/SC and OCR SC/SC arms reported adverse events (AEs): 75.4% and 86.4%, and serious AEs: 5.9% and 2.5%. The most frequently reported AEs were injection reactions (IRs, 51.5%); local and systemic IRs were experienced by 117 of 233 patients (50.2%) and 27 of 233 patients (11.6%), respectively. All IRs were mild/moderate; intensity and duration decreased with subsequent injections.

Discussion

The OCR SC formulation demonstrated noninferiority to OCR IV formulation regarding drug exposure, providing comparable efficacy and safety and an additional treatment option for patients with multiple sclerosis.

Classification of Evidence

This study provides Class II evidence that a single SC injection of 920 mg of OCR achieves a noninferior 12-week area under serum concentration–time curve to that of 2 IV infusions of 300-mg OCR administered 2 weeks apart.

Trial Registration Information

ClinicalTrials.gov Identifier NCT05232825; submitted: January 27, 2022; first patient enrolled: May 3, 2022; available at: clinicaltrials.gov/study/NCT05232825?term=NCT05232825&rank=1.

Introduction

Ocrelizumab (OCR) is an anti-CD20 monoclonal antibody and is the only high-efficacy disease-modifying therapy (DMT) approved for the treatment of patients with relapsing and primary progressive multiple sclerosis (PwRMS/PwPPMS).^1,2 As of March 2024, more than 350,000 patients had started OCR (resulting in more than 1,000,000 patient-years of follow-up), with over 10 years of safety and efficacy data from clinical trials.^2,3 OCR is administered as an initial dose of two 300-mg IV infusions 2 weeks apart, each lasting approximately 2.5 hours, with subsequent doses administered every 6 months as single 600-mg infusions over 2 or 3.5 hours.^1,2

Accessing IV administration may be challenging for some patients, if they do not live close to an infusion center; some multiple sclerosis (MS) centers have limited or no IV infrastructure. For some patients, IV administration is not an option, and others may favor subcutaneous (SC) administration.^4-6 To deliver the clinical benefit of OCR and provide treatment flexibility, a new OCR formulation was developed for SC administration with the same twice-yearly schedule.

OCR SC dose is coformulated with recombinant human hyaluronidase PH20 (rHuPH20; Halozyme Therapeutics, Inc., San Diego, CA), which allows for increased dispersion and absorption of SC-injected drugs by degrading hyaluronan at the local injection site, temporarily removing the barrier to bulk fluid flow, thereby permitting the rapid delivery of greater fluid volumes.⁷ rHuPH20 is currently used in SC coformulations with other approved monoclonal antibodies.^8,9 The OCR SC dose tested in this trial was selected based on safety, tolerability, and pharmacokinetic (PK) results of a phase 1b, dose-escalation, open-label study (OCARINA I, NCT03972306).¹⁰

OCARINA II is an ongoing phase 3, multicenter, randomized, open-label, parallel-group study with the primary objective of demonstrating PK noninferiority of the SC formulation of OCR compared with the currently approved OCR IV 600 mg dose in PwRMS and PwPPMS. In this study, we report the primary results from the controlled period of the study, efficacy data up to week 48, and safety data until the latest clinical cutoff.

Methods

Patients

Key eligibility criteria included age 18–65 years, a diagnosis of RMS or PPMS per the 2017 revised McDonald criteria,¹¹ an Expanded Disability Status Scale (EDSS) score of 0.0–6.5 at screening, and disease duration from onset of MS symptoms of less than 15 years for patients with an EDSS score less than 2.0 at screening. Key exclusion criteria were previous treatment with anti-CD20 drugs including OCR, if the last treatment was within 2 years before screening, and/or if the B-cell count was below the lower limit of normal (LLN) at screening (80 cells/μL), and contraindications to mandatory premedication (i.e., corticosteroids and antihistamines).

Trial Design

OCARINA II is a phase 3 trial comprising a 24-week controlled treatment period, followed by a 72-week noncontrolled treatment phase and a 24-week safety follow-up period for patients completing the treatment phase or discontinuing early (maximum duration 120 weeks). Eligible patients were randomized 1:1 to receive a single OCR IV dose administered as two 300-mg infusions, 2 weeks apart, or OCR SC dose administered as a single 920-mg injection at day 1. At week 24, patients randomized to the OCR IV/SC arm were switched to OCR SC 920 mg every 24 weeks until week 96 (Figure 1); patients randomized to OCR SC/SC remained on this until week 96. OCR SC dose (23-mL solution) was administered as an injection in the abdomen of the patient by trained site staff, either manually or with a pump, over a recommended period of approximately 10 minutes. OCR IV dose was administered according to product guidelines.^1,2

Figure 1. OCARINA II Study Design.

ADA = antidrug antibody; EDSS = Expanded Disability Status Scale; OCR = ocrelizumab; PD = pharmacodynamic; PK = pharmacokinetic; PPMS = primary progressive multiple sclerosis; RMS = relapsing multiple sclerosis; SC = subcutaneous. ^aThe 920 mg OCR SC dose was selected based on the OCARINA I study data (NCT03972306). ^bThe first OCR IV dose was administered as two 300-mg IV infusions given 2 weeks apart. ^cThe screening phase in patients with RMS and PPMS took place before baseline MRI readings, and patients were randomized 1:1 between the 2 arms.

Patients treated with the OCR SC dose received mandatory premedication (20 mg of dexamethasone and 5 mg of desloratadine; if contraindicated or not available, alternatives could be used) orally 1–2 hours before each OCR SC administration, to minimize the risk of injection reactions (IRs). The protocol was subsequently amended, allowing premedication to be given shortly before (non-US sites) or 30–60 minutes before (US sites) the injection. Premedication was given before OCR IV administration to minimize infusion-related reactions (IRRs) per product guidelines. Patients in both arms could receive optional oral analgesics. A 1-hour postinjection monitoring (PIM) was initially required; a protocol amendment removed this requirement for patients receiving OCR SC dose during the treatment phase (non-US sites).

Trial Procedures and Study End Points

The primary objective was to establish noninferiority of OCR SC 920 mg compared with OCR IV 600 mg, with the corresponding end point the OCR area under the serum concentration–time curve from day 1 to week 12 (AUC_W1–12). The 12-week cutoff was selected based on the observed time of the peak concentration in OCARINA I. Absorption was expected to be complete by week 12, and the AUC_W1–12 covers >90% of the total AUC over the 6-month dosing interval and is, therefore, representative of the overall OCR exposure.¹⁰ The total AUC from day 1 to week 24 (AUC_W1–24), determined from all available samples up to week 24, was also calculated. OCR serum concentration was measured using an ELISA-based method. Serum samples for AUC determination were collected from patients after SC administration on days 1, 2, 3, 5, 7, 10, and 14; for IV administration on day 1 and day 14 before and 30 minutes after the infusion; and for both on weeks 4, 8, 12, 16, and 24.

Key secondary end points included maximum serum concentration and the proportion of patients achieving CD19⁺ B-cell level ≤5 cells/µL at weeks 12 and 24 (exploratory at weeks 48 and/or 96). In both arms, samples were drawn during screening and at weeks 1, 2, 4, 12, 16, 24, 48, and 96 and the absolute counts of CD19⁺ cells in blood were measured using the BD Multitest 6-color TBNK reagent and BD Trucount tube on FACSCanto II cytometer. Immunogenicity was assessed by determining the incidence of treatment-emergent antidrug antibodies (ADAs) to OCR after IV or SC administration and the formation of antibodies to rHuPH20 after OCR SC administration. Samples for analysis of ADAs to OCR and rHuPH20 were collected from patients at weeks 1, 24, 48, 72, and 96 in both arms (at week 1, ADA to rHuPH20 sample was collected for the OCR SC/SC arm only).

Secondary radiologic end points were as follows: total number of T1 gadolinium-enhancing (Gd+) lesions at weeks 8 and 24 and total number of new/enlarging T2-weighted lesions (N/E T2w-L) at weeks 12 and 24; exploratory end points were T1 Gd+ lesions at weeks 48 and 96 and N/E T2w-L at weeks 8, 48, and 96. Other exploratory end points included annualized relapse rate in PwRMS, calculated as the total number of protocol-defined relapses for all patients divided by the total patient-years (PYs) of exposure, at weeks 24, 48, and 96, and change in the EDSS score from baseline to weeks 48, 72, and 96.

The biomarker exploratory end point included the levels of serum neurofilament light chain (sNfL) compared with baseline, measured using the high-sensitivity single-molecule array (Simoa) NF-Light V2 Advantage kit on HD-X (Quanterix, Billerica, MA) at weeks 12, 24, 48, and 96. The safety profile of OCR SC administration was compared with OCR IV administration during the 24-week controlled period, and evaluation at subsequent visits was continued during the treatment phase. Nature and incidence of adverse events (AEs), classified according to National Cancer Institute's Common Terminology Criteria for Adverse Events v5.0, and serious AEs (SAEs), namely IRs for OCR SC dose, IRRs for OCR IV dose, and infections and serious infections (SIs), were reported. IRs are any AEs defined per protocol as occurring during the OCR SC injection or within 24 hours after the end of the injection and judged by the investigator to be related to the OCR SC injection. IRs were categorized as either local or systemic based on symptoms. Serum levels of immunoglobulins G, M, and A (IgG, IgM, and IgA) were assessed at weeks 24, 48, 72, and 96 and compared with baseline. Patient satisfaction and experience scores with IV infusion or SC injection were measured using the Treatment Administration Satisfaction Questionnaire (TASQ) after doses 1, 2, and 3.

Pharmacokinetic and Statistical Analyses

For the primary end point, the geometric mean ratio (GMR) of the AUC SC dose vs the AUC IV dose was determined, along with the 2-sided 90% CI. Based on an unequal variance analysis, assuming a coefficient of variation for SC and IV doses of 0.75 and 0.50, respectively, with a true GMR of 1.0, 232 patients (116 patients per treatment group) needed to be enrolled to achieve 90% power with a 1-sided α of 0.05 for the noninferiority analysis. The noninferiority was established if the lower end of the 2-sided 90% CI of the GMR of AUC was >0.8; the noninferiority limit of 0.8 corresponds to a maximal 20% loss in AUC for the SC administration compared with IV dose. Additional information on PK and statistical analysis is given in eMethods.

Other Analyses

For MRI end points, the unadjusted estimated lesion rates (i.e., total number of lesions divided by the number of patients with a readable MRI scan at each visit) are reported and were used to assess whether OCR SC vs IV treatment would result in similarly low rates of lesions. NfL levels were analyzed using log-transformed concentrations and reported as geometric means and percentage change from baseline. Descriptive summary statistics were used for demographics and other outcomes.

The cutoff date for the PK data set was March 10, 2023 (when the last patient reached week 12). All available PK data beyond week 12 were also included in the PK analysis. Pharmacodynamic, clinical, radiologic, fluid biomarker, and laboratory data are reported up to week 48, and the total exposure time up to the data cutoff for these analyses (December 4, 2023) was used to report safety data in 3 groups: patients randomized to IV and switched to SC at week 24 (OCR IV/SC), patients randomized to SC and continued to receive OCR SC during the treatment phase (OCR SC/SC), and patients who received at least 1 OCR SC dose, which combines patients from the first 2 arms.

Standard Protocol Approvals, Registrations, and Patient Consents

The trial (ClinicalTrials.gov identifier number NCT05232825) was approved by a local or central institutional review board at each participating institution and conducted in accordance with the International Conference on Harmonization guidelines, applicable regulations, and the Declaration of Helsinki. All patients provided written informed consent.

Data Availability

For up-to-date details on Roche's Global Policy on the sharing of clinical information and how to request access to related clinical study documents, readers can visit go.roche.com/data_sharing. Anonymized records for individual patients across more than 1 data source external to Roche cannot, and should not, be linked, because of a potential increase in risk of patient reidentification. For eligible studies, qualified researchers may request access to individual patient-level clinical data through a data request platform. At the time of writing, this request platform is Vivli (vivli.org/ourmember/roche/). The study protocol and statistical analysis plan are available in eSAP1 and eSAP2, respectively.

Results

Patient Characteristics and Patient Disposition

Overall, 236 patients from 37 centers in 8 participating countries underwent 1:1 randomization, with 118 patients allocated to the OCR IV/SC and OCR SC/SC arms, each. Baseline demographics and disease characteristics were broadly reflective of the overall MS population and were, overall, comparable between the 2 groups (Table 1). A total of 233 of 236 patients (98.7%) completed the 24-week controlled period and entered the treatment phase; 111 of 118 patients (94.1%) and 114 of 118 patients (96.6%) completed week 48 in the OCR IV/SC and OCR SC/SC arms, respectively. Reasons for discontinuation are shown in Figure 2. At the cutoff date, the median treatment duration was 64.1 weeks (range 24.1–82.1) and 63.6 weeks (range 40.3–83.0) for patients in the OCR IV/SC and OCR SC/SC arms, respectively.

Table 1.

Baseline Demographics and Disease Characteristics

Variable	OCR IV/SC (N = 118)	OCR SC/SC (N = 118)
Age, y, mean ± SD	40.0 ± 11.9	39.9 ± 11.4
Female sex, n (%)	70 (59.3)	77 (65.3)
Weight, kg, mean ± SD	76.1 ± 22.7	75.4 ± 16.6
Time since symptom onset, y, mean ± SD	6.8 ± 7.1	7.7 ± 8.3
MS subtype, n (%)
RMS	105 (89.0)	105 (89.0)
PPMS	12 (10.2)	11 (9.3)
aSPMS	1 (0.8)	2 (1.7)
Patients with T1 Gd+ lesions, n (%)	25 (24.3)	22 (21.2)
T2 lesion count, mean ± SD	49.8 (34.6)	44.5 (32.3)
EDSS score, median (range)	3.0 (0–6.5)	2.5 (0–6.5)
Patients with at least 1 previous DMT exposure, n (%)a,b	59c (50.0)	65c (55.1)
Interferon β-1a	17 (14.4)	21 (17.8)
Interferon β-1b	2 (1.7)	5 (4.2)
Peginterferon β-1a	7 (5.9)	5 (4.2)
Glatiramer acetate	15 (12.7)	14 (11.9)
Teriflunomide	15 (12.7)	15 (12.7)
Fingolimod	6 (5.1)	2 (1.7)
Natalizumab	6 (5.1)	5 (4.2)

Abbreviations: aSPMS = active secondary progressive multiple sclerosis; DMT = disease-modifying therapy; EDSS = Expanded Disability Status Scale; Gd+ = gadolinium-enhancing; MS = multiple sclerosis; OCR = ocrelizumab; PPMS = primary progressive multiple sclerosis; RMS = relapsing multiple sclerosis; SC = subcutaneous.

^aOnly DMTs that started before randomization and have been used in >4% of patients in either arm are reported here.

^bMultiple uses of a specific medication for a patient were counted once in the frequency for the medication.

^cFor frequency counts in the total number of patients with at least 1 treatment, multiple uses of the same medication for a patient were counted separately.

Figure 2. Patient Disposition.

OCR = ocrelizumab; SC = subcutaneous; SFU = safety follow-up. ^aOne patient who entered the SFU from the OCR IV/SC arm discontinued the study because of physician decision.

Pharmacokinetic Analysis

The estimated mean exposure over the first 12 weeks after OCR administration (AUC_W1–12) was 2,750 μg/mL·d for OCR IV 600 mg and 3,500 μg/mL·d for OCR SC 920 mg. The GMR (90% CI) for OCR SC AUC_W1–12 vs IV was 1.29 (1.23–1.35), demonstrating noninferior exposure of OCR SC 920 mg to OCR IV 600 mg. SC bioavailability was estimated at 81%. The GMR (90% CI) for the AUC over the first 24 weeks was 1.27 (1.21–1.34), confirming the PK noninferiority of the OCR SC formulation over the total dosing interval. The mean peak concentration and the mean concentration at the end of a dosing interval (i.e., just before next drug administration) were 137 μg/mL and 0.58 μg/mL for OCR 600 mg IV, respectively, and 132 μg/mL and 0.63 μg/mL for OCR 920 mg SC, respectively. Peak concentrations for OCR SC administration were reached at 4 days after injection (range 2–13 days; Figure 3).

Figure 3. Observed Individual OCR Serum Concentration vs Time Profiles by Treatment (A: Linear Scale; B: Log Scale).

OCR = ocrelizumab; SC = subcutaneous.

Pharmacodynamics: Impact on B-Cell Counts

The onset and extent of B-cell depletion in blood were similar between the OCR IV/SC and OCR SC/SC arms. At day 14, the median B-cell count was 0 cells/µL (eFigure 1A). More than 90% and 95% of patients achieved and maintained CD19⁺ B-cell counts ≤5 cells/µL (a common literature-reported threshold¹²; Figure 4A) and <10 cells/µL (the empirically established cutoff¹³; eFigure 1B), respectively, by day 14. During the SC treatment phase, >85% and >88% of patients displayed B-cell counts ≤5 cells/µL and <10 cells/µL, respectively, at week 48 in both arms (Figure 4A and eFigure 1B).

Figure 4. (A) Percentage of Patients With ≤5 Cells/µL of CD19+ B-Cell Absolute Counts, (B) MRI Activity Over Time: Gd+ Lesion Rate, and (C) N/E T2w-L Rate.

Gd+ = gadolinium-enhancing; N/E = new/enlarging; OCR = ocrelizumab; SC = subcutaneous; T2w-L = T2-weighted lesion. ^aThe n/N values refer to the number of patients with CD19⁺ B-cell count ≤5 cells/μL divided by the number of patients with a valid CD19⁺ B-cell count result at the visit. ^bThe lesion rate is the total number of lesions divided by the number of patients with a readable MRI assessment at the visit. ^cN/E T2 lesion count measurement is performed regarding the previous scheduled available visit. ^dAt baseline for T1 Gd+ lesions, 78 of 103 patients (75.7%) had no lesions and 11 of 103 (10.7%) had ≥4 lesions. ^eAt baseline for T1 Gd+ lesions, 82 of 104 patients (78.8%) had no lesions and 5 of 104 (4.8%) had ≥4 lesions.

Impact on Radiologic End Points

At baseline, patients in the OCR IV/SC arm had a greater rate of T1 Gd+ lesions compared with patients in the OCR SC/SC arm (lesions per scan [SD] 0.98 [2.52] vs 0.54 [1.66]). By week 8, the rate of T1 Gd+ lesions had decreased by 78.6% in the OCR IV/SC arm (0.21 [0.98]) and 72.2% in the OCR SC/SC arm (0.15 [0.45]). By the end of the controlled period (week 24), an almost complete suppression of T1 Gd+ lesions was observed, with 1 patient in the OCR SC/SC arm having a single detectable T1 Gd+ lesion. At week 48, no T1 Gd+ lesions were detected in either arm (Figure 4B).

At week 12 in both arms, an almost complete suppression of N/E T2w-L was observed that was maintained at weeks 24 and 48 (lesion rate [SD], IV/SC vs SC/SC: 0.06 [0.35] vs 0.03 [0.18]; 0.01 [0.09] vs 0.03 [0.18]; 0.02 [0.19] vs 0.01 [0.09] respectively; Figure 4C).

Impact on Neurofilament Levels

Notable reductions in sNfL levels were observed in both OCR IV/SC and OCR SC/SC arms by week 24, decreasing further at week 48 (all patients receiving OCR SC from week 24) (percentage change from baseline [95% CI], IV/SC vs SC/SC: week 24, –15.4% [–23.2 to −6.7] vs −20.3% [–27.7 to −12.2]; week 48, –25.8% [–33.1 to −17.7] vs −23.6% [–31.1 to −15.3], respectively; eFigure 2A). At week 48, the geometric mean sNfL levels were 7.5 pg/mL (CI 6.9–8.1) in the OCR IV/SC arm and 7.5 pg/mL (CI 6.8–8.3) in the OCR SC/SC arm (eFigure 2B), values that are similar to healthy control levels.¹⁴

Impact on Clinical End Points

In the RMS population, 98.1% of patients (104/106) in the OCR IV/SC arm and 97.2% of patients (104/107) in the OCR SC/SC arm were relapse-free up to week 48. Two patients (1.9%) in each arm had 1 relapse and 1 patient (0.9%) in the OCR SC/SC arm had 2 relapses at week 48, resulting in an unadjusted relapse rate per year of 0.02 in the OCR IV/SC arm and 0.04 in the OCR SC/SC arm during a total follow-up time of 97.36 PYs and 99.04 PYs, respectively.

The mean (±SD) EDSS scores at baseline were 3.0 (±1.7) in the OCR IV/SC arm and 2.8 (±1.6) in the OCR SC/SC arm; EDSS scores remained stable up to week 48 (change from baseline in the EDSS score of −0.11 points [±0.74] in the OCR IV/SC arm and −0.07 points [±0.68] in the OCR SC/SC arm).

Patient-Reported Outcomes

Responses to the TASQ showed that, at the first OCR SC injection, 95.8% of patients (113/118) randomized to the OCR SC/SC arm were either satisfied or very satisfied with the SC procedure, 87.3% (103/118) felt that the SC procedure was convenient or very convenient, and 91.5% (108/118) felt that the time taken to get the injection was just right. Patients switching from OCR IV to OCR SC at week 24 reported similar satisfaction results (eTable 1). A similar proportion of patients from the OCR SC/SC arm reported being satisfied or very satisfied with the SC procedure at injections 2 (97.4%) and 3 (97.3%). At week 48, most patients in both the OCR IV/SC and OCR SC/SC arms, respectively, either did not experience or experienced mild pain (96/109 [88.1%] vs 106/113 [93.8%]), swelling (97/109 [89.0%] vs 101/113 [89.4%]), and redness at the site of the injection (90/108 [83.3%] vs 93/113 [82.3%]). Most patients in the OCR SC/SC arm responded that they would definitely recommend OCR SC administration at injection 1 (78/118, 66.1%), injection 2 (94/116, 81.0%), and injection 3 (87/113, 77.0%).

Safety

Adverse Events

A total of 89 of 118 patients (75.4%) randomized to OCR IV/SC and 102 of 118 patients (86.4%) randomized to OCR SC/SC reported an AE over a median (range) treatment phase of 64.1 weeks (24.1–82.1) and 63.6 weeks (40.3–83.0), respectively. The difference between the 2 arms was mostly driven by a greater number of patients experiencing IRs in the group randomized to OCR SC/SC (OCR IV/SC: 42.4% vs OCR SC/SC: 59.3%), which is explained by the greater number of SC injections (Table 2). The most common AEs were IRs (for OCR SC), IRRs (for OCR IV), infections (mostly upper respiratory tract infections, followed by coronavirus disease 2019 [COVID-19] and urinary tract infection), and headache (Table 2). Among patients who received at least 1 OCR SC injection (OCR SC all-exposure, n = 233), 175 patients (75.1%) reported ≥1 AE. Most patients reported mild (grade 1) or moderate (grade 2) AEs (169/175, 96.6%), and 6 of 175 patients (3.4%) reported grade 3 events; no life-threatening (grade 4) or fatal events (grade 5) were reported. A total of 12 SAEs were reported in 6 patients (6/233, 2.6%) treated with OCR SC dose, all of which resolved. No treatment discontinuation was reported due to AEs.

Table 2.

Adverse Events

Variablea	Controlled period		Treatment phase		Patients who received at least 1 dose of OCR SC 920 mgc (N = 233)
	OCR IV/SC (N = 118)	OCR SC/SC (N = 118)	OCR IV/SCb (N = 118)	OCR SC/SCb (N = 118)
Time on treatment, wk, median (range)	24.29 (0.1–35.6)	24.14 (22.4–39.7)	64.1 (24.1–82.1)	63.6 (40.3–83.0)	53.71 (1.0–83.0)
Total no. of IV infusions/SC injections, n	118	118	259	380	639
No. of patients with at least 1 AE,d n (%)	56 (47.5%)	92 (78.0%)	89 (75.4)	102 (86.4)	175 (75.1)
IRe	0	57 (48.3%)	50 (42.4)	70 (59.3)	120 (51.5)
Local IRs	0	54 (45.8%)	50 (42.4)	67 (56.8)	117 (50.2)
Systemic IRs	0	13 (11.0%)	9 (7.6)	18 (15.3)	27 (11.6)
Infusion-related reaction	20 (16.9%)	0	20 (16.9)	NA	NA
Upper respiratory tract infection	9 (7.6%)	8 (6.8%)	14 (11.9)	11 (9.3)	18 (7.7)
COVID-19	5 (4.2%)	8 (6.8%)	6 (5.1)	13 (11.0)	14 (6.0)
UTI	5 (4.2%)	4 (3.4%)	9 (7.6)	7 (5.9)	10 (4.3)
Headache	3 (2.5%)	12 (10.2%)	3 (2.5)	12 (10.2)	12 (5.2)
Patients with at least 1 SAE, n (%)	3 (2.5%)	3 (2.5%)	7 (5.9)	3 (2.5)	6 (2.6)
AEs leading to withdrawal from treatment, n (%)	0	0	0	0	0
AEs leading to modificationf/interruption, n (%)	6 (5.1%)	0	4 (3.4)	1 (0.8)	1 (0.4)
Patients with an infection, n (%)	33 (28.0%)	44 (37.3%)	53 (44.9)	59 (50.0)	89 (38.2)

Abbreviations: AE = adverse event; COVID-19 = coronavirus disease 2019; IR = injection reaction; MedDRA = Medical Dictionary for Regulatory Activities; OCR = ocrelizumab; NA = not applicable; NCI CTCAE = National Cancer Institute's Common Terminology Criteria for Adverse Events; SAE = serious adverse event; SC = subcutaneous; UTI = urinary tract infection Wk = week..

AEs were reported in over 5% of patients. Additional information related to SAEs are given in eAppendix.

^aOCR IV/SC = patients randomized to OCR IV 600 mg and who switched to OCR SC 920 mg at week 24; OCR SC/SC = patients randomized to OCR SC 920 mg and who continued to receive this during the treatment phase. Patients who received at least 1 dose of OCR SC 920 mg = patients who received at least 1 dose of OCR SC 920 mg, that is, a combination of patients from OCR IV/SC and OCR SC/SC.

^bTime on treatment is computed as the date of the last study drug administration minus the date of the first study drug administration plus 1 day.

^cTime on treatment is computed as the date of the last OCR SC dose administration minus the date of the first OCR SC dose administration plus 1 day.

^dAEs were encoded using MedDRA v26.1.

^eIRs were graded based on the NCI CTCAE (v5.0) as per protocol and ranged from grades 1–5 (see eMethods).

^fDose modification involves interrupting, temporarily discontinuing, or reducing the rate of drug administration.

Infections

In the controlled period, the proportion of patients with an infection was 28.0% (33/118) in the OCR IV/SC arm and 37.3% (44/118) in the OCR SC/SC arm. All patients recovered, and all infections resolved. In the controlled period, 3.4% of patients (4/118) in the OCR IV/SC arm experienced SIs of appendicitis, cellulitis staphylococcal, pneumonia, SC abscess, and upper respiratory tract infection. No patients in the OCR SC/SC arm experienced SIs during the controlled period.

Among patients receiving at least 1 OCR SC dose, 38.2% (89/233) experienced an infection. The most frequent infections were upper respiratory tract infection (7.7%) and COVID-19 (6.0%); no infections at the site of injection were reported. One patient reported a urinary tract infection that was considered to be serious but resolved. All SIs resolved in fewer than 22 days.

Injection Reactions

The median treatment duration in patients who received OCR SC 920 mg from the time of their first OCR SC injection was 53.71 weeks (range 1.0–83.0); 51.5% of these patients (120/233) experienced at least 1 IR (n = 230 IRs and n = 395 IR symptoms). 50.2% of patients (117/233) reported local IRs, and the most common local IR symptoms were erythema (34.8%), pain (17.2%), swelling (9.4%), and pruritus (5.6%), all of which were grades 1–2. Most (81.9%) of the local IRs occurred within 1 hour of SC administration, and most (160/219 events [73.1%]) resolved within 3 days; the median duration was 2 days (range 1–16 days). A total of 32 systemic IRs occurred in 11.6% of patients (27/233), and a total of 41 symptoms were reported. The most common systemic IR symptoms were headache (2.1%), flushing (1.3%), and nausea (1.3%). Symptoms were nonserious and mild (17/32, 53.1%) or moderate (15/32, 46.9%). Systemic IR symptoms mostly occurred within 1 hour of SC administration, with most (20/32 events [62.5%]) resolving within 3 days; the median duration was 3 days (range 1–16 days). IRs recovered or resolved in 99.1% of cases (228/230); in 2 patients from the OCR SC/SC arm, symptoms of bruising were still recovering or resolving at the cutoff date. Of the patients who experienced at least 1 IR, 20.8% of patients (25/120) required standard-of-care treatment, such as analgesic, and oral or topical antihistamines. All IRs were mild or moderate, and the incidence was highest at the first injection (Figure 5); severity grade and duration of IR decreased with subsequent injections. In addition, fewer patients required symptomatic treatment with repeated injections.

Figure 5. Percentage of Patients With Injection Reactions by Injection and Severity: (A) Injection Reactions by Treatment Arm; (B) Local Injection Reactions for the All-Exposure Population; (C) Systemic Injection Reactions for the All-Exposure Population.

IR = injection reaction; NCI CTCAE = National Cancer Institute's Common Terminology Criteria for Adverse Events; OCR = ocrelizumab; SC = subcutaneous. The percentage of patients with at least 1 IR for the selected injection is calculated as the number of patients with IRs at this injection (n) divided by the number of patients who received the injection per treatment group (N). The reported most extreme NCI CTCAE grade of IRs is used to display the severity of IRs. If patients experience multiple IRs for 1 injection, the maximum most extreme NCI CTCAE grade is displayed per patient and injection. Injection 1 corresponds to dose 2 for patients randomized to the OCR IV/SC arm and dose 1 for patients randomized to the OCR SC/SC arm. Subsequent injections correspond to subsequent doses.

Postinjection Monitoring and Premedication

Of 233 patients who received OCR SC administration, 80 patients (34.3%) did not undergo PIM. Among these, 8 patients reported an IR at Injections 2 and 3 while 5 patients reported an IR at Injection 4. The reported IRs were mostly grade 1 local reactions and resolved; no grade 3 or higher IRs were reported. Making the 1-hour PIM optional (at the treating physician's discretion) had no major impact on the reported IR characteristics and did not reveal any new safety signals.

After implementing a shorter premedication interval, the proportion of patients with a shorter premedication time increased with subsequent injections. Overall, 53.2% (124/233) of patients had at least 1 injection with premedication in the period ≤30 minutes. Similar incidence of IRs was observed after the implementation of a shorter premedication interval, including patients who received premedication ≤30 minutes before an injection (eTable 2).

Immunoglobulin Levels

A small decrease from baseline in mean IgA, IgG, and IgM levels was observed in patients receiving at least 1 OCR SC dose up to week 48. The proportion of patients with no episodes of Ig concentrations below the LLN was 97.0% for IgA, 97.9% for IgG, and 82.4% for IgM (data not shown).

The proportion of patients who had IgA and IgG concentrations below the LLN remained low and consistent over time, whereas the proportion of patients with IgM concentrations below the LLN increased over time.

Antidrug Antibodies

At baseline, the prevalence of antibodies to OCR was 0.9% (2/234 patients) and anti-rHuPH20 antibodies was 5.6% (10/178 patients). At the latest clinical cutoff, 0 of 235 patients had treatment-emergent ADAs to OCR and 1 of 230 patients (0.4%) had treatment-emergent anti-rHuPH20 antibodies.

Classification of Evidence

This study provides Class II evidence that a single SC injection of 920 mg of OCR achieves a noninferior 12-week area under serum concentration–time curve to that of 2 IV infusions of 300 mg OCR administered 2 weeks apart.

Discussion

We report that the SC formulation of OCR (OCR SC 920 mg) showed PK noninferiority compared with the IV formulation of OCR 600 mg (primary end point), indicating that similar systemic exposure to OCR is achieved with the 2 formulations. This resulted in a similarly rapid and near-complete depletion of peripheral CD19⁺ B cells in the first 24 weeks in both arms, which was sustained at week 48. These results are consistent with the well-known pharmacodynamic effect previously reported in the pivotal trials.^15,16

In addition, both treatment arms showed a rapid and pronounced reduction in new MRI activity by week 8, with near-complete suppression of new lesion activity by week 12, sustained at subsequent time points. Minimal clinical activity was observed with over 97% of patients free of relapses up to week 48 and disability levels remaining largely unchanged over the same period. These findings support the conclusion of similar efficacy for IV and SC routes of administration, as expected according to the principle of PK bridging.^17-19 A remarkable decrease in sNfL levels after treatment with the 2 formulations was also observed, with values at week 48 similar to those reported in healthy controls,¹⁴ indicating consistent reduction of neuroaxonal injury. The rate of decline of sNfL levels was slower than that observed for MRI measures; this prolonged elevation of sNfL relative to MRI lesions has consistently been reported^20-23 and aligns with findings from OCR IV pivotal trials.²⁴ Patients reported very high levels of satisfaction with the SC formulation, for convenience and short administration time, in line with the results from the OCARINA I trial.¹⁰

The safety profile of OCR in both arms was similar and consistent with the rate and nature of AEs reported in previous studies,^15,16,25,26 except for AEs typically associated with a SC route of administration. Overall, OCR SC dose was well tolerated. The difference in the proportion of patients reporting AEs between both treatment arms was driven by the reporting of IRs with the OCR SC formulation. Reported IRs were all mild or moderate and consisted primarily of local symptoms such as erythema, pain, swelling, and pruritus and, to a lesser extent, of systemic symptoms such as headache, flushing, and nausea. Most IRs required no symptomatic treatment, and none were considered serious or led to treatment withdrawal. The incidence, intensity, and duration of IRs and the need for symptomatic treatment decreased with repeated injections. No treatment-emergent ADAs to OCR were reported in either treatment arm; anti-rHuPH20 antibodies were only reported in 1 patient at the clinical cutoff date. In this study, different premedication intervals and PIM were evaluated. Making the PIM optional after the second injection and implementing a shorter premedication interval (shortly or 30–60 minutes before an injection) had no major impact on the IR characteristics and did not reveal any new safety signals.

This study has some limitations. Although the AUC is usually reported over 1 dosing interval, that is, 24 weeks for OCR, this study used a cutoff of 12 weeks. Absorption from the SC injection site is expected to be complete within these 12 weeks, and the elimination of a molecule is the same after IV and SC administration. Therefore, 12 weeks were considered sufficient to assess potential exposure differences between IV and SC administration. The sensitivity analysis conducted for patients enrolled at the beginning of the study and with data available up to 24 weeks after dose administration confirmed this. Although open to all PwMS, this study enrolled a higher percentage of PwRMS (almost 90%) vs PwPPMS compared with the expected distribution of MS phenotypes. However, it has previously been shown that the PK properties of OCR are not different in PwRMS and PwPPMS,¹² making the results of OCARINA II generalizable to different clinical phenotypes. The median treatment durations of 64.1 weeks (range 24.1–82.1, OCR IV/SC) and 63.57 weeks (range 40.3–83.0, OCR SC/SC) reported in this article were relatively short for analyses of disability and safety; however, no differences regarding these parameters were observed between IV and SC administration, as expected given the similar PK exposure. Furthermore, the data presented here are consistent with the well-characterized profile of OCR.³ The completion of this 2-year study will provide additional insights.

OCARINA II confirms that the SC formulation, administered twice-yearly in around 10 minutes, delivers similar OCR exposure as the OCR IV formulation. As anticipated, owing to the similar PK exposure, we observed analogous benefits on efficacy-related parameters for OCR SC and IV administration. In turn, the SC formulation provides an additional treatment choice and flexibility, including the option for patients being treated with OCR IV dose to transition to OCR SC dose.² This novel formulation could lessen treatment time and treatment burden^11,12 and provide greater access to a highly effective DMT.

Glossary

ADA

antidrug antibody

AE

adverse event

AUC

area under the serum concentration–time curve

COVID-19

coronavirus disease 2019

DMT

disease-modifying therapy

EDSS

Expanded Disability Status Scale

Gd+

gadolinium-enhancing

GMR

geometric mean ratio

Ig

immunoglobulin

IR

injection reaction

IRR

infusion-related reaction

LLN

lower limit of normal

MS

multiple sclerosis

N/E T2w-L

new/enlarging T2-weighted lesion

OCR

ocrelizumab

PIM

postinjection monitoring

PK

pharmacokinetic

PwMS

people with MS

PwPPMS

patients with primary progressive multiple sclerosis

PwRMS

patients with relapsing multiple sclerosis

PY

patient-year

rHuPH20

recombinant human hyaluronidase PH20

SAE

serious adverse event

SC

subcutaneous

SI

serious infection

sNfL

serum neurofilament light chain

TASQ

Treatment Administration Satisfaction Questionnaire

Author Contributions

S.D. Newsome: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. E. Krzystanek: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. K.W. Selmaj: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. M. Dufek: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. L. Goldstick: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. C. Pozzilli: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. C. Figueiredo: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. B. Townsend: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. H. Kletzl: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. O. Bortolami: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. D. Zecevic: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. C. Giacobino: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. S. Clinch: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. Y.-A. Shen: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. G.D. Bhullar: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. H.-M. Schneble: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. D. Centonze: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

This work was supported by financial support from F. Hoffmann-La Roche Ltd., Basel, Switzerland, for the study and publication of the manuscript. Preclinical and clinical research was supported by grants from Bayer Schering, Bristol Myers Squibb, Biogen, Celgene, Lundbeck, Merck-Serono, Novartis, Roche, Sanofi-Genzyme, and Teva.

Disclosure

S.D. Newsome received consultancy fees for scientific advisory boards from Biogen, Genentech Inc., Bristol Myers Squibb, EMD Serono, Greenwich Biosciences, Horizon Therapeutics, Novartis, and TG Therapeutics; is a study lead PI for a Roche clinical trial program; and received research funding (paid directly to institution) from Biogen, Lundbeck, Roche, Genentech Inc., the National MS Society, the Stiff Person Syndrome Research Foundation, the Department of Defense, and the Patient-Centered Outcomes Research Institute. E. Krzystanek received consultancy fees for scientific advisory boards from Biogen, Merck-Serono, Bayer, Roche, Novartis, and the Polish Multiple Sclerosis Society; is a study lead PI for Roche, TG Therapeutics, Merck, Biogen, Lundbeck, and Janssen clinical trial programs; and received compensation for speaking services from Biogen, Bayer, Novartis, UCB, Roche, Merck-Serono, Teva, Lundbeck, Pfizer, Sandoz, and Sanofi-Genzyme. K.W. Selmaj received honoraria for speaking, consulting, and serving for advisory boards for AstraZeneca, Merck, Novartis, Roche, Biogen, Celgene, Bristol Myers Squibb, and TG Therapeutics. M. Dufek received compensation for travel, speaker honoraria, and consultancy fees from Biogen Idec., Novartis, Merck-Serono, Roche, Sanofi-Genzyme, Teva, and Janssen-Cilag. L. Goldstick received consultancy fees from EMD Serono, Bristol Myers Squibb, Biogen, Sanofi-Genzyme, and Roche/Genentech; and has received research support from Biogen, Roche/Genentech, and Sanofi-Genzyme. C. Pozzilli has served on scientific advisory boards and consultancy and/or received speaking fees from Almirall, Alexion Pharmaceutical Inc., Biogen, Janssen, Roche, Merck, and Novartis; and has received research support from Almirall, Biogen, Roche, Merck, and Novartis. C. Figueiredo, B. Townsend, and H. Kletzl are employees of and shareholders in F. Hoffmann-La Roche Ltd. O. Bortolami was a contractor for F. Hoffmann-La Roche Ltd. until December 2023; his current affiliation is IQVIA RDS. D. Zecevic, C. Giacobino, and S. Clinch are employees of and shareholders in F. Hoffmann-La Roche Ltd. Y.-A. Shen is an employee of Genentech Inc., and a shareholder in F. Hoffmann-La Roche Ltd. G.D. Bhullar and H.-M. Schneble are employees of and shareholders in F. Hoffmann-La Roche Ltd. D. Centonze acted as an advisory board member and received honoraria for speaking or consultancy fees from Alexion, Almirall, Amicus, Bayer, Biogen, Bristol Myers Squibb, Celgene, Chiesi, GW Pharmaceuticals, Horizon, Janssen, Lundbeck, Merck-Serono, Novartis, Roche, Sandoz, Sanofi-Genzyme, Viatris, and Teva; and is the PI in clinical trials of Biogen, Bristol Myers Squibb, Merck-Serono, Mitsubishi, Novartis, Roche, Sanofi-Genzyme, and Actelion. Go to Neurology.org/N for full disclosures.