Tocilizumab for relapsing and remitting giant cell arteritis: a case series

This case series discusses 16 consecutive patients started on tocilizumab for relapsed or refractory GCA between August 2018 and April 2021, after the establishment of a new hub and spoke model. The central assessing hub was St George’s University Hospitals NHS Foundation Trust, a large tertiary rheumatology center in the south of England.

A new referral form for tocilizumab was utilized and all cases were discussed in a fortnightly multidisciplinary team meeting. Information was gathered for this analysis retrospectively through these referral forms and electronic patient records, with additional correspondence from referring clinicians and corroboration with patients themselves. The central dispensing pharmacy was at St George’s Hospital, who validated patient numbers and details.

A total of 16 patients were identified (Table 1): 11 female and 5 male, with an average age of 72.4 (range 61–82) years. Seven patients were internal referrals from clinicians at the tertiary center, with the remainder from local district general hospitals.

The new hub and spoke clinical pathway established at this regional London center for the use of tocilizumab in relapsed and refractory GCA, facilitated the increased uptake of the newly licensed drug. This model has been already been well established at the hospital-level in the UK, with district general hospitals linked to centralized tertiary hospitals for treatment and specialist clinics [14], and here we have incorporated this organizational structure for the referral pathway. The referral pathway was simple to establish with the resources present, as it utilizes existing national health system (NHS) frameworks and systems, and the majority of patients referred to the pathway were accepted for tocilizumab therapy.

The efficacy of the pathway, however, relies on familiarity that the pathway itself exists by both registrars and consultants, knowledge of how to refer with the referral form required, and the administrative responsibilities of the centralized coordinating hub and pharmacy. The fortnightly multidisciplinary teams (MDTs) upon which the pathway hinged allowed for the additional opportunity for senior clinicians to discuss complex cases, and their experiences with and side effects of a relatively new treatment.

Analysis of 15 cases in this series demonstrates an average daily dose reduction of 20.4 mg prednisolone [95% CI 13.0–27.8 mg], and three cases were weaned from steroid treatment completely. This represents a considerable reduction in daily dosing and therefore overall steroid burden for those initiated on tocilizumab treatment. Current studies suggest that the average duration of treatment with steroids in GCA is 3 years, with a long steroid tapering period between 1 and 9 years [5]; while another study identified that only 24% of GCA patients were able to stop steroids within 2 years [15]. Studies in steroid use in polymyalgia rheumatica (PMR), GCA’s counterpart disease, have also highlighted that it is duration of steroid treatment rather than the initial prednisolone dose or maintenance dose, that contributes to greater side effects [16], signifying the impact of protracted courses of steroids and the danger of continuing courses for longer than necessary. This series therefore supports that a reduction and/or complete cessation of steroids is more possible with the addition of tocilizumab.

Consideration of the balancing measures mean it remains important however to highlight that glucocorticoids are and do remain the mainstay of acute management of GCA, as outlined in the introduction. Qualitative work has demonstrated that rheumatology clinicians are “often under pressure from patients and their primary care physicians to taper corticosteroids” and that this may not always be what is appropriate [17].

Both clinic letters and personal correspondence with referring clinicians highlighted that side effects from steroids were common and present in 11 of 16 cases, with weight gain in five cases and the serious osteoporotic complication of avascular necrosis in two cases. Existing larger prospective studies of steroid use in GCA and PMR have identified the most common side effects to be fractures, peptic ulcer perforations, and diabetes mellitus in this population cohort [19], and it is therefore important to try to rationalize steroids as much as is possible in an ageing population with existing comorbidities. Women are particularly affected by steroid side effects due to their hormonal and genetic predisposition to poorer bone health [20], and with GCA more common in women this is also pertinent to consider.

The concurrent prescription of gastroprotective medication with steroid use is well established, and the majority of our cases (15 out of 16) were on such an agent—13 on a proton pump inhibitor and a further 2 on ranitidine. This contrasted however to the medications for the reduction of cardiovascular risk, which were prescribed to a lesser extent—9 of 16 cases were on the recommended daily aspirin (with one additional case on warfarin), and only 25% were on a statin. This highlights room for improvement and a need for greater focus on the significant association of GCA with cardiovascular disease.

The inflammatory markers monitored were erythrocyte sedimentation ratio (ESR) and C-reactive protein (CRP), both relevant markers monitored in GCA [21], with ESR also forming part of the diagnostic criteria as outlined previously. All the cases showed a biochemical response with ESR and CRP, with differences in both markers after treatment with tocilizumab (respective means of 10.3 and 7.7). The greater difference, however, was the change in these markers at presentation and with treatment with steroids, with a mean ESR drop of 22.3 and CRP of 68.2.

The pathology of vasculitis is one of inflammation, and it is noted in the spectrum of patients suffering from GCA that those with a greater inflammatory response at diagnosis (looking at parameters of fever, weight loss, ESR) were more likely to have greater and longer steroid requirements [22]. Therefore, these markers may be more important at the stage of diagnosis of GCA and impact of initial steroid therapy, rather than with monitoring the disease due to the impact of tocilizumab, which itself blunts a CRP/ESR response.

Fourteen of the 16 cases completed 12 months treatment with tocilizumab, and out of the two who did not, case O had ocular syphilis and case J had multifold side effects. The longest duration of treatment break was 3 months (case G), due to a routine hip operation.

As the use of tocilizumab becomes more common, its side effects have been better characterized in literature. They are summarised below as per the Summary of Product Characteristics (SmPC) on the Electronic Medicines Compendium (EMC) [23], and we relate them to our case series in Table 4. The most commonly reported adverse drug reactions (ADRs) in the SmPC data are upper respiratory tract infections, nasopharyngitis, headache, and hypertension.

These scenarios highlight the importance of close biochemical and clinical monitoring in those on this potent biologic. The consultant clinicians in the series approached tocilizumab prescribing with caution and care, and regular review and blood tests meant that dosing was very closely monitored. However, in the context of the new reliance of virtual care during the coronavirus pandemic, this level of monitoring is not always possible at present. The use of tocilizumab should therefore continue to be considered at a case-by-case level.

Further contraindications to tocilizumab include latent tuberculosis (TB) and hepatitis B and C, and all patients must be screened before starting tocilizumab [23]. If latent TB were to be identified, this must be fully treated before tocilizumab is started. Analysis of documentation in this case series identified avoidable delay between authorization of tocilizumab and initiation of treatment, due to the completion required for screening blood tests and chest radiographs for TB. Although not explicitly outlined in the guidance, other diseases can manifest upon initiation of tocilizumab, and case O was an unusual example of ocular syphilis being revealed and where treatment had to be stopped.

Of note, intravenous tocilizumab is to be stopped at least 4 weeks before any surgery, and subcutaneous tocilizumab at least 2 weeks before any surgery—as was the case for case G (hip replacement surgery) and N (operation for trapped nerve).

This case series is bound by the same limitations of any case series, including being retrospective and descriptive in nature. By having no control arm for the experiences of patients not taking tocilizumab, the conclusions formulated only suggest association and not causation. Selection bias was limited, however, by inclusion of all accepted MDT discussed cases within the specified time frame, and by corroborating cases with the pharmacy.

In regards to our data specifically, due to the nature of the hub and spoke referral pathway, it is perhaps understandable that a large proportion cases were referred from the hub hospital itself (7 of 16). Data from the spoke hospitals and referring district general hospitals were sometimes more limited due to not being electronic.

Half of cases in this series continue tocilizumab treatment beyond the initial NICE guidelines limit of 12 months. Studies are currently being carried out to investigate this, including an extension to the original GiACTA trial—where patients not in full remission after a year were randomized to tocilizumab and/or glucocorticoids again [26]. This trial extension identified that re-treatment with tocilizumab was more associated with complete remission, and that overall glucocorticoid dosing over 3 years was lower in those who received tocilizumab rather than placebo in the first part of the trial. The NICE guidelines on tocilizumab in GCA are to be reviewed shortly, and we anticipate that the extension of the use of tocilizumab to be discussed.