Severe Infections in Patients Treated with Tocilizumab for Systemic Diseases Other Than Rheumatoid Arthritis: A Retrospective Multicenter Observational Study

Abstract

Objective:

This study aimed to describe severe infections in patients treated with tocilizumab for systemic diseases other than rheumatoid arthritis.

Methods:

Data from patients receiving at least 2 doses of tocilizumab for systemic diseases other than rheumatoid arthritis between January 1, 2012, and July 1, 2020, in the region Poitou-Charentes (France) were retrospectively collected from medical records. Psoriatic arthritis and systemic juvenile idiopathic arthritis were also excluded as usually treated with similar modalities to rheumatoid arthritis.

Results:

Of 37 patients, mainly suffering from giant cell arteritis, 25 patients (68%) had at least 1 infectious event and 15 severe infections occurred in 6 patients (3.2/100 patient-years), mainly bacterial. Lower respiratory tract and skin were the main sites. Severe bacterial infections were associated with a marked biological inflammatory syndrome, even under a cycle of administration of tocilizumab. Two severe zonas and 1 severe diverticulitis occurred. No tuberculosis or viral hepatitis reactivation was observed.

Conclusion:

The incidence rate of severe infections was 3.2/100 patient-years and seems lower than that reported in rheumatoid arthritis. C-reactive protein dosage could be helpful for the diagnosis of bacterial infectious adverse events in patients on tocilizumab. Further larger studies are needed to confirm these results to assess potential risk factors for severe infections.

Main Points

• The ratio of severe infections in patients treated with tocilizumab for systemic diseases other than rheumatoid arthritis was 3.2/100 patient-years.

• The main sites of severe infections were the skin and lungs.

• A significant elevation of C-reactive protein was observed in majority of severe bacterial infections.

• The occurrence of severe zonas suggests possible interest in prophylaxis

Introduction

Tocilizumab (TCZ) is a humanized immunoglobulin (Ig) G1 monoclonal antibody that selectively neutralizes both the soluble and membrane-bound forms of interleukin (IL)-6 receptor. It was first approved in second-line treatment for rheumatoid arthritis (RA) and polyarticular and systemic juvenile idiopathic arthritis, then more recently for giant cell arteritis (GCA) and polymyalgia rheumatica.¹ Tocilizumab is also used in several other systemic diseases such as Castleman disease,² polychondritis,³ systemic sclerosis,⁴ Still disease,⁵ and Behçet disease.⁶

Some cases of severe and life-threatening bacterial infections with minimal fever and low biological inflammatory markers, especially C-reactive protein (CRP), were reported in patients treated by TCZ.^7,8 Moreover, several randomized clinical trials and real-life studies in RA suggested that TCZ increased the risk of opportunistic and serious bacterial infection to an extent similar to other biological disease-modifying anti-rheumatic drugs (DMARDs) such as anti-TNF alpha agents.^9-12 Data are however limited by various confounding factors such as comorbidities and the use of other immunosuppressive drugs.^13,14 To the best of our knowledge, except for RA and limited evidence for GCA, there are no available data about severe infections occurring in patients with other systemic diseases treated with TCZ.

The aim of this study was to describe severe infections occurring in patients treated with TCZ for systemic diseases other than RA.

Methods

Study Design and Patient Recruitment

We conducted a retrospective, multicenter observational study including all patients receiving at least 2 doses of TCZ (intravenous or subcutaneous) between January 1, 2012 (which was the date of first use of TCZ in the region regardless of indication) and July 1, 2020, in the former region Poitou-Charentes in France, an administrative region of 1.8 million inhabitants in south-western France, including 1 university hospital (Poitiers) and 4 secondary hospital centers (Niort, Angouleme, Rochefort, and La Rochelle). Patients were screened from the pharmacy of each hospital. The exclusion criteria were age under 18 years, non-systemic disorders (Devic’s Neuromyelitis, Graves’ orbitopathy), major lack of data (especially about dose of immunosuppressive agents used or patients lost to follow-up), and patients with RA. Psoriatic arthritis and systemic juvenile idiopathic arthritis have also been excluded from this study as usually treated by similar modalities to RA. Cytokine release syndrome associated with chimeric-antigen receptor T cells (CAR T cells) and severe COVID-19 disease were also excluded since a single injection was generally performed. Patients were informed with a booklet on TCZ administration, systematically given upon arrival in the hospital and orally expressed their non-opposition to data collection.

Data Collection

Biometrical, clinical, and biological data were collected from the patients’ medical records of each hospital. The cumulative dose of TCZ (with the protocol used) and glucocorticoids (GCs) was calculated. Use of other immunosuppressants before TCZ introduction was also recorded. Severe infections occurring before and up to at least 6 months after the last injection of TCZ were recorded. An infection was defined as severe if required hospitalization or intravenous treatment, was life-threatening, or led to disability or permanent damage or death. Type(s) of infections, severity, delay since the first TCZ injection, biological inflammatory markers, type and duration of the treatment received, and outcomes were recorded. C-reactive protein was considered increased if greater than 5 mg/L.

Outcomes

Primary outcome was the frequency of severe infections. Secondary outcomes were the characteristics of patients with severe infections and their description.

Ethics

This study complies with the Declaration of Helsinki and was approved by the local ethics committee of the University of Poitiers and registered with the French data protection agency (Commission Nationale de l’Informatique et des Libertés, CNIL) (No. CHU86-R2019-11-06). According to French regulations, a non-opposition was obtained from all participants who participated in this study.

Statistical Analysis

Descriptive analyses were done using the Statview software 5.0 and standardized methods to determine the mean and standard deviations or the median and quartile range value [Q1-Q3] depending on the distribution of the variables studied. No univariate or multivariate analysis was performed due to the small sample size.

Results

Population Study

Totally 229 patients received TCZ during the inclusion period, of which 192 were excluded according to the inclusion criteria. Finally, 37 patients were included in the study and received the first infusion of TCZ between January 1, 2015, and July 1, 2020 (Figure 1).

Figure 1.

Flow chart. TCZ, tocilizumab; SARS Cov2, severe acute respiratory syndrome due to COVID-19.

Patients’ characteristics are depicted in Table 1. Median age at inclusion was 66 years [61-74], and most of them were women (70%). Nearly a third of patients had a history of infection before any immunosuppressive therapy. Seven (19%) had a history of colonic diverticulosis, 2 had a positive QuantiFERON test, and 1 had anti-Hbc antibody positivity with negative viral load before TCZ initiation. Twenty-seven (73%) received vaccines before the first TCZ infusion, mostly against Streptococcus pneumoniae (n = 24). The main indication of TCZ was GCA (n = 17), then polymyalgia rheumatica, Takayasu arteritis, Still disease, polychondritis (n = 4 each), and Castleman disease (n = 2) or 2 other vasculitis (Behçet disease and unclassified vasculitis). The median time between the diagnosis of the disease and TCZ initiation was 2 years [1-4]. At TCZ initiation, all patients except 1 were treated by GCs (with a median cumulative dose of 13.7 g prednisone equivalent (p.e.) [6.4-34.4]) and 27 (73%) received other immunosuppressants, mostly methotrexate (n = 23). Mean duration of follow-up during TCZ therapy was 33 ± 19 months with an average cumulative dose of 9.2 ± 5.9 g.

Table 1.

Patients’ Characteristics

Characteristicsa	All Patients (n = 37)
Age (years), median [Q1-Q3]	66 [61-74]
Female sex	26 (70)
BMI (kg/m2), mean ± SD	25.5 ± 5
Diabetes mellitus	6 (16)
Chronic renal insufficiency	3 (8)
Medical history of cancer	11 (30)
Medical history of colic diverticulosis	7 (19)
History of vaccination against Streptococcus pneumoniae	24 (65)
History of infections before TCZ introductionb	15 (41)
Indication for TCZ
GCA	17 (46)
Takayasu	4 (11)
Polymyalgia rheumatica	4 (11)
Still disease	4 (11)
Polychondritis	4 (11)
Castleman disease	2 (5)
Other vasculitisc	2 (5)
Duration of disease before TCZ start (months), median [Q1-Q3]	19 [6-49]
Immunosuppressants used before TCZ start
GCs	36 (97)
Total cumulative dose (g, p.e.), median [Q1-Q3]	13.7 [6.4-34.4]
Daily dose (mg), median [Q1-Q3]	17.5 [8-30]
Methotrexate	23 (62)
Azathioprine	5 (14)
Anakinra	5 (14)
Anti-TFN𝛼	4 (11)
Othersd	12 (32)
Tocilizumab regimen
Intravenous	32 (89)
Subcutaneous	12 (33)
Total cumulative dose (g), mean ± SD	9.2 ± 5.9
Duration of TCZ (months), mean ± SD	33 ± 19

^aAll values are expressed in n (%), unless otherwise specified;

^bUrinary tract infections (n = 3), dermohypodermatitis (n = 3), tuberculosis (n = 2), bursitis and tenosynovitis (n = 2), chronic skin Alternaria inflectoria infection (n = 1), hepatitis C (n = 1) (treated with ribavirin and interferon and considered as cured), cholecystitis (n = 2), herpes zoster (n = 1), infectious mononucleosis (n = 1), bacteriemia (n = 1), and bacterial meningitis (n = 1);

^cBehçet disease and unclassified vasculitis;

^dDisulone (n = 3), salazopyrine (n = 3), leflunomide (n = 2), cyclophosphamide, tadekinig-𝛼, canakinumab, and rituximab (n = 1 each).

BMI, body mass index; TCZ, tocilizumab; GCA, giant cell arteritis; GCs, glucocorticoids; p.e., prednisone equivalent.

Among patients with GCA (n = 17), the duration before TCZ was 19 [7.5-24] months and GC daily dose at TCZ initiation was 10 [3.6-13.1] mg/day. Most of them (60%) had uncontrolled GCA with a history of failure of other immunosuppressants (e.g., methotrexate).

Severe infections

During TCZ therapy, 6 patients (17%), all over 65 years of age, had severe infections with a total of 15 severe infections and ratio of 3.2 events/100 patient-years of exposure (Table 2). All necessitated hospitalization and none resulted in death. Of these 6 patients, 2 (33%) had subcutaneous TCZ. Of 15 severe infections, 12 (80%) were currently treated with usual cycle of administration of TCZ (subcutaneous n = 3, intravenous n = 9).

Table 2.

Description of Serious Infections

Sex	Disease	Age at First Infection (Years)	Therapies Before TCZ	Infections Before TCZ (n)	TCZ Regimen	Time Since TCZ Start (Months)	Time Since Last TCZ (Days)	GC Dose (mg/day)	Infections
									Site	Micro organism	CRP (mg/L)	Outcomes
Ma	Unclassified vasculitis	82	GC, ANA, MTX, AZA, DIS	3	IV	16	12	5	ORL	No	140	Healing
						24	6	9	Urinary	E. coli	161	Healing
						31	28	15	Pulmonary	No	331	Healing
						34	23	15	Skin	No	340	Healing
						37	76	15	Skin	No	187	Healing
M	GCA	75	GC, ANA	0	SC	5	4	20	Pulmonary	No	368	Healing
M	Castelman	88	No	0	IV	1	9	0	Digestive (diverticulitis)	No	36	Healing
F	Relapsing polychondritis	73	GC, DIS, MTX, AZA	1	IV	3	5	30	Skin	VZV	NA	Post-zosterian pain
F	GCA	66	GC, MTX	1	SC	3	7	12	Digestive (oesophagitis)	Candida	2	Healing
						15	4	2	Skin	VZV	<1	Post-zosterian pain
Mb	GCA	65	GC, MTX	3	IV	1	13	50	Skin	E. cloacae	NA	Skin wound
						3	8	30	Skin	E. coli	61	Skin wound
						3	21	30	Bacteriemia	P. aeruginosa	NA	Healing
						6	117	8	Pulmonary	HRSV, E. cloacae	126	Healing
						7	160	6	Skin	P. aeruginosa, S. aureus, S. pyogenes	47	Healing

^aHad non-substituted hypogammaglobulinemia;

^bHad substituted hypogammaglobulinemia.

M, male; F, female; GCA, giant cell arteritis; TCZ, tocilizumab; GC, glucocorticoids; ANA, anakinra; MTX, methotrexate; AZA, azathioprine; DIS, disulone; IV, intravenous; SC, subcutaneous; E. coli, Escherichia coli; VZV, varicella zoster virus; E. cloacae, Enterobacter cloacae; P. aeruginosa, Pseudomonas aeruginosa; HRSV, human respiratory syncytial virus; S. aureus, Staphylococcus aureus; S. pyogenes, Streptococcus pyogenes; CRP, C-reactive protein; NA, not available.

Skin was the main infection site (n = 7) and resulted in permanent sequelae in 4 cases (2 disabling post-zosterian pain and 2 extensive skin wounds due to necrotizing fasciitis). One severe diverticulitis with digestive perforation was recorded. Most infections were documented or considered bacterial (80%) and associated with an increased biological inflammatory syndrome with a median value of CRP of 150 mg/L [61-331]. Among 6 patients, 2 presented several infections (patients 1 and 6), both presenting a history of severe infection on immunosuppressant before TCZ introduction (1 with severe urinary tract infection and 1 with chronic skin Alternaria inflectora infection requiring 10 months of antifungal treatment). Only 2 of these patients did not have any infection prior to initiation of TCZ. Median time to severe infection from the start of TCZ was 6 months [3-22]. Except 1, all patients received TCZ in association with GCs only. The median cumulative dose of GCs at TCZ initiation was 22.8 g [9.4-77.8] with a mean dose of 26 ± 17.8 mg per day p.e and a median cumulative dose of 3.8 g [3-4.5] under TCZ course. Two patients had hypogammaglobulinemia around 4 g/L including 1 receiving immunoglobulin supplementation.

Discussion

Tocilizumab-related infections were mostly evaluated in RA and more recently in GCA. The reported incidence of severe infections in RA varies from 3.9/100 patient-years of exposure to 4.7/100 patient-years of exposure depending on studies and remains the most frequent cause of death in the all-exposed population.^9,15-18 This risk of global infection seems comparable to other biologic agents, although the risk of serious infections seems lower than that observed with TNF antagonists¹¹ or Rituximab.¹⁹ In GCA, the GiACTA study (Trial Of Tocilizumab in Giant-cell Arteritis) study reported severe infections in 7% of the patients in the TCZ group¹ with corresponding incidence ratios at least 3-fold greater than that reported in patients with RA but may be more related to the disease itself and GCs regimen than TCZ treatment.²⁰ In our study, 3 patients (8%) with GCA presented serious infections which are in line with the results of GiACTA Study. Another national multicentric observational study reported severe infections in 11.9% (10.6/100 patient-years) of cases of GCA under TCZ,²¹ but although age was similar than our 17 patients with GCA (73 vs. 71 years), they had more frequently a history of DMARDs usage (73% vs. 60%) and a greater dose of GCs at TCZ initiation (21.5 vs. 10 mg/day).

In our study, severe infections occurred within 6 months to 2 years after TCZ initiation and affected only people aged over 65 years. The number of severe infections was greater after TCZ initiation than during the use of conventional immunosuppressive therapy. This result should be interpreted with caution because of the very small sample of patients and the biases related to other infectious risk factors that may be associated with the occurrence of such episodes after initiation of TCZ, especially concomitant treatment with GCs which is known to be associated with an increased risk of severe infections in patients treated with TCZ.²² In contrast to the literature, these infections occurred in patients without long-standing disease. History of infections seems to be an important factor for the occurrence of severe infections, since 4 out of 6 patients had such a history, 2 of whom had recurrent severe infections. Among the 4 patients with available data about gamma globulin level, 2 had hypogammaglobulinemia around 4 g/L including 1 receiving immunoglobulin supplementation.

In agreement with the literature on RA, most infections were respiratory and skin infections.^13,17,23 Importantly, contrary to previous clinical cases report,^7,8 our study shows that patients with various systemic diseases who develop severe bacterial infections on TCZ mostly had associated marked biological inflammatory syndrome, even under usual cycle of administration of TCZ (Table 2). This has only been described in one other study on 6/8 patients with RA and severe infections.¹³ This could be helpful for the diagnosis of bacterial infectious adverse events in patients on TCZ, but larger studies are needed to confirm this observation.

Our study found only one but severe case of diverticulitis with lower intestinal perforation. This is in line with the well-known increased risk of lower intestinal perforation related to diverticulitis in patients treated with TCZ compared with all other DMARDs.²⁴ Interestingly, while the risk of herpes zoster infection is similar to that of other immunosuppressive agents such as TNF alpha inhibitors,¹⁸ 2 severe de novo zonas occurred in our study resulting in disabling post-zosterian pain. No case of tuberculosis or reactivation of chronic viral hepatitis was found, which is in agreement with the literature on RA.^9,11,25

Except for partial data in GCA detailled above, our real-life multicentric study is the first to describe sevre infections in patients treated with TCZ for systemic diseases other than RA. The other strengths of our study are its multicentric and real-life design. In addition, the monitoring period offered a long-term follow-up under treatment, which allowed to have sufficient perspective about the occurrence of infections. Our study has several limitations. First, the small sample size limited the interpretability of the results and did not allow statistical analysis, especially for evaluation of the main clinical factors associated with severe infections and adjustment of confounding factors, especially concomitant GC exposure. Indeed, a real-world database multicentric study found a statistically significant increased risk associated with each 1-g increase in cumulative glucocorticoid exposure for glucocorticoid-related serious infections and for infectious adverse events related to TCZ.²² It is noteworthy that 9/15 severe infections occurred in patients receiving a daily dose of GC >10 mg p.e. in our study. Secondly, the retrospective design may have underestimated the number of infectious events, especially mild ones, and therefore, our study focused on severe infections only. Furthermore, we did not conduct an imputability investigation to conclude that TCZ is an independent risk factor for severe infections.

Conclusion

The incidence rate of severe infections was 3.2/100 patient-years and seemed lower than that reported in RA. Severe infections were mostly bacterial with a significant elevation of CRP in the majority of cases. The occurrence of severe zonas suggests the possible interest in prophylaxis in patients treated with TCZ. A larger study should be carried out to confirm these results and to assess potential risk factors of severe infections in patients with systemic diseases treated by TCZ.