Practice Points
- Nearly 25% of people with multiple sclerosis (MS) in our study reported wearing off prior to their next ocrelizumab infusion while on the standard 6-month dosing interval.
- The most common wearing-off manifestations were worsening of chronic residual MS symptoms and worsening fatigue.
- Higher disability severity, presence of fatigue, and obesity were identified as risk factors for experiencing ocrelizumab wearing off.
- In most people with MS with wearing off, shortening ocrelizumab infusion intervals or in-class disease-modifying therapy switches improved or resolved this phenomenon.
Multiple sclerosis (MS) is a chronic neurological condition characterized by inflammation, demyelination, gliosis, and neuronal loss within the central nervous system (CNS). Though it remains a leading cause of neurologic disability, the emergence of B cell–depleting therapies has revolutionized care for people with MS. These monoclonal antibodies (mABs) target CD20, a cell surface marker, on the surface of B cells and dampen the aberrant immune response seen in people with MS; by suppressing this inflammatory response, mABs have been shown to be highly effective in preventing disease activity in people with relapsing MS.1-3
Ocrelizumab (OCR) is an anti-CD20 mAB that is typically administered via intravenous infusion every 6 months. OCR was approved by the US Food and Drug Administration in 2017 for treatment of adults with relapsing forms of MS and primary progressive MS based on results from pivotal phase 3 trials,1,4 and has since become one of the most widely prescribed disease-modifying therapies (DMTs) in the United States.5
Wearing off describes subjective worsening of residual symptoms or transient emergence of generalized systemic symptoms in people with MS occurring shortly before the next scheduled dose of their DMT, with return to baseline after the next dose. This phenomenon appears to occur independently of the therapeutic effect of these DMTs on suppressing clinical relapse activity. Wearing off was first noted in people with MS using natalizumab; studies reported it in over half of people with MS who were taking the drug, and it was found that higher body mass index (BMI) and lower natalizumab receptor occupancy were associated with wearing off.6,7 In addition, wearing off has been reported with other DMTs as more mABs have been introduced. One study found that 141 of 248 people with MS receiving treatment with mABs (including ocrelizumab, rituximab, natalizumab, and ofatumumab) reported wearing off, and the phenomenon was associated with reduced treatment satisfaction.8
Multiple studies have investigated the prevalence of wearing off, specifically in those receiving OCR. One study identified wearing off in 61% (71 of 117) of participants using OCR, with fatigue identified as the most common symptom. Toorop et al also found a 2.7-times higher risk of wearing off with OCR in those with a BMI greater than or equal to 25.9 The results of another study showed wearing off in 57.6% of 106 people with MS occurring 2 weeks prior to their next scheduled OCR infusion, with cognitive decline the most commonly reported effect.10 However, another study found that, despite 54% of the 103 people with MS surveyed describing symptoms of wearing off, clinically meaningful worsening in quality-of-life and patient-reported outcome measures (PROMs) was not observed in most people prior to their next infusion.11 Also, in contrast to Toorop et al, BMI was not shown to predict wearing off, but younger age and a higher degree of disability did predict the tendency to self-report this phenomenon.11
A pathophysiologic rationale for wearing off in people with MS receiving OCR remains unknown. Prior studies have not shown an association with pharmacodynamic changes in CD19+ B-cell counts.8,9 Interestingly, a study did identify an association with elevated CD3+, CD8+, and CD3+/CD27+ T lymphocytes (T cells).10 Clinical relapses, subclinical disease activity on MRI, and serum levels of neurofilament light chain (NfL), a marker of axonal injury, were not shown to be associated with wearing off in 2 studies,9,11 though a separate study did link higher serum NfL levels with wearing off.10
The competing conclusions from prior studies highlight the need to expand our understanding of wearing off with OCR. This study, utilizing a large cohort of people with MS receiving OCR, further investigates this phenomenon to clarify its prevalence, predictive factors, and the impact of interventions used to counteract the wearing-off effect.
Methods
Participants
This retrospective, observational, noninterventional study was conducted at a single outpatient tertiary care MS subspecialty center at the University of South Florida. Participants were included under a waiver of consent granted by the University of South Florida Institutional Review Board, study protocol number STUDY003662. All people with MS from our center who were treated with OCR between September 1, 2021, and August 31, 2023, were identified by their electronic medical records. They were excluded if they had been lost to follow-up, only seen in consultation, not seen in an adequate time frame after starting OCR (precluding assessment of tolerability/wearing off), or if their MS diagnosis could not be confirmed.
Data Collection
Demographic data, including age, sex, race/ethnicity, height, and weight, were recorded. All clinical variables were collected by chart review of documentation from providers with MS subspecialty training, including MS subtype (defined as relapsing versus progressive) and tolerability of OCR (rated as good vs poor, with the latter defined as having any reported adverse effects during an infusion). PROMs routinely obtained during in-person clinic visits were also captured, including the Patient-Determined Disease Steps (PDDS) scale, which has been shown to correlate with the Expanded Disability Status Scale scores as a marker of MS-related disability12; the Fatigue Severity Score (FSS), a validated and well-recognized measure of fatigue in people with MS correlating to quality-of-life measures13,14; and the Patient Health Questionnaire-9 (PHQ-9), which has been validated as a screening tool for major depressive disorder in people with MS.15 Depression was chosen as a particular comorbidity of interest given prior literature associating it with OCR wearing off.11 Additional surrogate variables for fatigue were also captured, including whether a fatigue diagnosis code had been listed and whether the participant had been treated with a medication for fatigue, including amantadine, modafinil, armodafinil, amphetamine/dextroamphetamine, lisdexamfetamine, and/or methylphenidate. Serum biomarkers, including lymphocyte subsets, quantitative immunoglobulins, and NfL, are not routinely obtained in our practice due to their limited impact on clinical management and, thus, were not captured in this study.8,9
The presence of wearing off was also captured by a review of clinical documentation. Wearing off was defined as subjective worsening of fatigue, cognition, or residual focal neurologic deficits toward the end of a participant's OCR infusion dosing interval, absent any exam or imaging changes, with return to baseline after receiving their subsequent OCR infusion. When an individual did experience wearing off with OCR, additional clinical variables were obtained, including their characterization of wearing-off symptoms, when they began to note symptomatic worsening relative to their next infusion date, and how long they had been treated with OCR before they first noted wearing off. Management changes in response to wearing off, such as shortening OCR infusion intervals and switching DMTs, and the impacts of these changes on wearing off were also recorded.
Data Analysis
Descriptive statistics are used to report sample demographics, proportion of people with MS with wearing off, symptoms of wearing off, timing of wearing off relative to the next infusion due date, onset of wearing off relative to OCR start, and interventions to wearing off and their outcomes. Multivariable logistic regression modeling was used to assess various predictors of wearing off, including age, sex, race, ethnicity, BMI, MS subtype, tolerance of OCR, disease severity, fatigue, and comorbid depression. Multiple continuous variables were converted to categorical variables to highlight their impacts in regression modeling and promote ease of communication with patients, ie, age was categorized by decade of life; BMI was categorized into normal (< 25), overweight (25-29.9), and obese (30+); PDDS was dichotomized into 0 to 3 vs 4 and above, the latter category requiring use of an assistive device for ambulation or greater degree of disability to improve objectivity of this scale; and PHQ-9 was converted into generally accepted categories of no/minimal depression (0-4), mild depression (5-9), moderate depression (10-14), moderately severe depression (15-19), and severe depression (20-27). Multiple imputation by chained equations was used to handle missing data, such as missing PROMs for telemedicine visits.
Results
The record review found 578 people with MS treated with OCR. Of that number, 45 had not had a subsequent clinic visit since starting OCR, either due to having started OCR only recently or being lost to follow-up, precluding assessment of tolerability/wearing off. Later, an additional 5 people were thought to have an alternative neurologic condition (2 myelin oligodendrocyte glycoprotein antibody disease, 1 stroke, 1 microvascular disease, and 1 neuropathy).
Sample demographics of the remaining 528 individuals are shown in Table 1. The average age was 47.5 years, and 72.9% were women. The majority of the study population was White (72.9%) and non-Hispanic (79.0%). Most were diagnosed with a relapsing form of MS, though 27.7% had clinically progressive disease, and individuals most commonly and on average noted a moderate level of disability (median PDDS score = 3). Fatigue and depression were common in people with MS at all levels of disability. Most participants subjectively tolerated OCR well (91.3%).
A wearing-off phenomenon was noted by 24.8% of our sample population (Table 2). In most cases, it was characterized by transient and reversible perceived worsening of residual focal MS symptoms (79.4%) and worsening of fatigue (67.2%). When it occurred, wearing off was most frequently noted to begin 2 to 4 weeks before the next OCR infusion (59.5%). In those who experienced wearing off, onset was uncommon after the initial split-dose infusions or first full-dose infusion (20.6%); most people with MS who developed wearing off did so within the first 3 years of treatment (80.2%).
Regression modeling revealed multiple protective and risk factors associated with wearing off (Figure). Individual univariable regression models determined that, though fatigue by any measure (diagnosis code, FSS, or fatigue medication) was significantly associated with wearing off, fatigue by diagnosis code was most highly correlated and, thus, was the measure included in the multivariable regression model. In the final multivariable regression model, controlling for all other variables, statistically significant risk factors for wearing off included race reported as other (OR, 2.13 compared with White; P = .028), obese BMI (OR, 1.93 compared with normal/underweight BMI; P = .013), presence of fatigue (OR, 2.22 compared with no fatigue; P < .001), and requiring an assistive device for ambulation (PDDS ≥ 4; OR, 2.16 compared with no assistive device; P = .004). Statistically significant protective factors against wearing off included male sex (OR, 0.52 compared with female sex; P = .014) and Hispanic ethnicity (OR, 0.45 compared with non-Hispanic ethnicity; P = .022).
Interventions made in response to wearing off and their impacts are shown in Table 3. In most individuals with MS who experienced wearing off, a change in DMT was made (73.3%), most frequently by shortening OCR dosing intervals (47.3%) or switching DMTs within class to ofatumumab or ublituximab (22.9%). Of the 131 people with wearing off, 24 had had an intervention within the preceding 6 months, and a response could not be assessed. In 2 others, treatment changes were made for other reasons (1 for recurrent infection, 1 for unintentional pregnancy); thus, response to intervention on wearing off was assessed in 105 people with MS. When OCR dosing intervals were shortened, wearing off typically improved or completely resolved (77.1%). A similar benefit was observed when an in-class DMT switch was made to ofatumumab or ublituximab (85.7%). Conversely, when no intervention was made, wearing off tended to persist or worsen (88.2%). Overall, any intervention made in response to wearing off was associated with a positive outcome (80.3%) compared with continuing the current management of MS symptoms (11.8%; P < .001).
Discussion
Our results demonstrate wearing off in 24.8% of people with MS receiving OCR. Though prior studies have identified this phenomenon at a higher rate,9-11 our study cohort is large, and the population of mostly relapsing MS (72.3%) and moderate disability (median PDDS score = 3) is broadly reflective of the real-world population. Prior work has questioned the veracity of this phenomenon with OCR,11 and it is indeed challenging to objectively measure because it is, by definition, subjective. Despite this, similar findings of patient-reported worsening of residual symptoms and fatigue have been reported across multiple cohorts of people with MS receiving OCR,8,9,11 and our work aligns with this body of evidence. The authors agree that rates of wearing off in general may be artificially inflated by attribution bias on the part of both patients and providers; however, the strict definition of wearing off in our study, excluding any objective exam or imaging changes, and our finding that this phenomenon was ameliorated by interventions such as shortening OCR infusion intervals, support the existence of a true wearing-off effect with OCR. Clinicians should proactively address this phenomenon in people with MS prior to starting DMTs and while they are receiving OCR because, when present, wearing off may be associated with depression and reduced treatment satisfaction.8
The wearing-off phenomenon with OCR may have implications for proposed extended interval dosing (EID) strategies. EID has been shown to mitigate the risk of progressive multifocal leukoencephalopathy in people with MS receiving natalizumab, but also increases the prevalence and severity of wearing off.16,17 Extending OCR infusion intervals beyond the standard 6-month schedule alters B cell and immunoglobulin kinetics, but does not appear to impact clinical or radiologic markers of MS disease activity.18-21 However, symptomatic wearing off has not been investigated specifically in groups treated with OCR EID.
Clinicians should be aware of factors that may impact the risk of OCR wearing off. Leveraging our large sample size and regression modeling to control for confounding, we found that a diagnosis code for fatigue increased the risk (OR, 2.22 compared with no fatigue), and that obesity (BMI ≥ 30; OR, 1.93 compared with normal/underweight BMI) nearly doubled the risk of experiencing wearing off with OCR, similar to prior work.9 Also aligning with prior research, increased degree of disability (ie, use of an assistive device for ambulation) also correlated with higher risk of wearing off (OR, 2.16 compared with no assistive device).11 It is possible that people with obesity and MS and those with higher overall clinical disability may, in effect, be underdosed: Obesity and high clinical severity of disease have been associated with markers of increased CNS inflammation, suggesting that wearing off may be dependent on the degree of OCR exposure.22 Post hoc analysis of the pivotal phase 3 trials of OCR showed greater reduction in risk of disability progression in those with higher OCR serum levels.23 The prevalence of wearing off should be assessed in ongoing clinical trials utilizing higher doses of OCR based on body weight (ie, NCT04544436, NCT04548999).
Although no definitive physiologic explanation for wearing off has yet been determined, additional prognostic factors uncovered by our study offer new avenues for further investigation. In particular, previously unreported demographic factors impacting risk of wearing off, including sex, race, and ethnicity, may point toward a biologic etiology, including specific predisposing phenotypes or genetic factors. Symptom worsening has been temporally linked with menstrual cycles in women with MS,24 potentially explaining the protective effect of male sex seen in our study. Alternatively, effects of male sex and Hispanic ethnicity may be explained by these individuals underreporting wearing off; prior research has shown lower rates of symptom reporting and health care utilization in this population.25-27
Wearing off most commonly began 2 to 4 weeks prior to the next OCR infusion due date, roughly 8% to 15% of the standard 6-month infusion dosing interval. This timing of onset differs from that reported with natalizumab: 4 to 7 days prior to the next infusion due date, or roughly 15% to 25% of the standard 4-week dosing interval,17 suggesting different underlying pathophysiologies. In people with MS with wearing off on natalizumab, a direct measure of natalizumab’s pharmacodynamic effect (lower natalizumab receptor occupancy) has correlated with the risk of wearing off.7 However, similar associations with direct measures of OCR’s pharmacodynamic effect, namely CD19+ B-cell counts, have not shown such a correlation.8,9 Recent research has highlighted other potential biologic etiologies for wearing off with OCR, including elevated T-cell counts and levels of NfL, potentially indicative of dysregulated immune responses and subclinical neuroaxonal damage.10 Our study does not endeavor to find a physiologic correlate for wearing off with OCR, but these associations should be further investigated in future prospective work.
If an as-yet-unrecognized physiologic change occurring toward the end of an OCR dosing cycle is responsible for wearing off, it stands to reason that shortening the dosing interval would mitigate this phenomenon, which was indeed shown in our sample population. Of 48 individuals experiencing wearing off in whom OCR dosing intervals were shortened (either to every 5 months or every 24 weeks), 37 (77.1%) noted improvement or complete resolution of this phenomenon. Such a physiologic change may be unique to OCR among anti-CD20 mABs—although the number of cases is lower, an in-class DMT switch to ofatumumab or ublituximab showed a similar benefit (85.7%). Though a burgeoning body of evidence suggests that subjective wearing off also occurs with ofatumumab,8,28,29 assessment of wearing off after switches from OCR has not yet been reported. Ublituximab has been shown to have a significant impact on fatigue in earlier studies, but rates of wearing off remain under investigation in the ongoing ENHANCE study (NCT05877963). The significant rates of response to interventions made after wearing off contrast with the low rate of self-resolution without intervention (11.8%), highlighting that when wearing off with OCR is identified, clinicians should consider changes in management.
Though we assessed wearing off with OCR in a large cohort, our study is limited by factors inherent to its retrospective chart review design. We followed a strict definition of the wearing-off phenomenon, but the determination of wearing off and subsequent response to intervention were ultimately made by the reviewer based on clinical documentation. Patients receiving OCR are not systematically surveyed regarding wearing off during their clinic visits, and the design of this retrospective study likely underestimates the true rate of OCR wearing off. Given that wearing off was uncommon within the first year of treatment with OCR (20.6% of cases with wearing off), a longer observational period may uncover more cases. Participants were collected from a single MS subspecialty clinic, facilitating standardization but potentially introducing unknown cognitive bias. As mentioned earlier, serum biomarkers were not collected in clinical practice in a regimented fashion that is conducive to retrospective analysis. Other important clinical considerations in the use of B cell–depleting mABs, such as infection rates (particularly with shortened infusion intervals), were also not captured in our study.
Future studies should further investigate rates and predictors of wearing off with OCR; other B cell–depleting mABs, including rituximab, ofatumumab, ublituximab, and subcutaneous OCR; and other interval-dosed DMTs, including natalizumab, alemtuzumab, and cladribine. Retrospective analyses may expand data collection to include PROMs, diagnosis codes, and/or medications associated with obstructive sleep apnea, depression, and other medical comorbidities. Longitudinal studies should include prospectively captured biomarkers beyond those solely related to CD19+ B-cell pharmacokinetics, such as T-cell subsets, quantitative immunoglobulins, and NfL, to search for a physiologic rationale for wearing off and to assess responses to interventions. These studies should also extend for longer durations to capture accurate rates of wearing off with OCR.
Conclusions
Wearing off was observed in nearly a quarter of our study population of people with MS receiving OCR. Clinicians should be aware of several factors that may increase the risk of wearing off with OCR, including obesity, premorbid fatigue, and use of an assistive device for ambulation. Wearing off tended not to self-resolve, and interventions were generally effective at mitigating this phenomenon, which may encourage neurologists to consider changes in disease management in response to wearing off with OCR.
