Introduction

Chronic spontaneous urticaria (CSU) is a mast-cell driven disease characterized by the presence of recurring hives and wheals and/or angioedema lasting a minimum of six weeks [1]. It affects approximately 0.5 to 1% of the global population [2,3,4]. Current guidelines recommend that second-generation H1-antihistamines be used as first-line treatment for CSU, with escalation up to four times the approved dose as second-line treatment if symptoms persist [1, 5]. Refractory cases of CSU should then receive adjunctive therapy with omalizumab (Xolair®), a monoclonal antibody directed against Immunoglobulin E (IgE) [6], which has demonstrated high efficacy and safety for the treatment of chronic urticaria (CU) [6,7,8,9,10,11].

Drug survival refers to the length of time a patient receives a particular medication until discontinuation due to any reason. This parameter not only allows for the assessment of long-term impact, safety and effectiveness, but also of several influencing factors, such as tolerability and patient and doctor preferences [12, 13]. In previous studies, average omalizumab drug survival was described to be 443.1 days (= 14.6 months) in patients with CSU [14]. However, several predictors of omalizumab response and drug survival have been described in recent years. Chronic inducible urticaria (CindU) was found to be an independent predictor of overall longer drug survival in CU patients [15], whereas other studies found pre-treatment IgE levels to robustly predict response to omalizumab [16, 17].

The systemic immune-inflammation index (SII) and the systemic inflammatory response index (SIRI) have been linked to the overall prognosis of variety of diseases like cancer, stroke or pneumonia [18,19,20]. It has been postulated that these indices could potentially also act as predictive markers for drug survival in CSU [21].

The aim of this study was to evaluate omalizumab drug survival in a large daily practice cohort from Switzerland and to assess predictive values of SII and SIRI indices on this survival rate.

Methods

Patient characteristics and data acquisition

Following ethical approval (BASEC-Nr.: 2021-01715), the electronic medical database at the University Hospital Zurich’s Department of Dermatology was searched for adult patients (≥ 18 years) with CSU treated with omalizumab between January 2018 and May 2021. The primary database search yielded a total of 181 cases. The following inclusion criteria were then applied: patients must be at least 18 years old, have a definitive diagnosis of CSU made by a board-certified allergologist or dermatologist, be treated with oral antihistamines for at least 4 weeks, have treatment with omalizumab 300 mg per month, have a documented outcome of either good response or no response, and have written general consent. Any patients not meeting inclusion criteria, who classified as partial responders and/or without accepted general consent were excluded from this study, as demonstrated in Fig. 1.

Fig. 1
figure 1

Patient characteristics and data acquisition

Full size image

The following parameters were then extracted from the electronic medical database: age, sex, pre-existing illnesses including history of atopy, clinical characteristics, presence of angioedema, presence of chronic inducible urticaria (CindU), disease duration before treatment initiation, prior therapy, pre-treatment blood values (IgE, tryptase, complete blood count), omalizumab injection dates, concurrent immunosuppressive treatments and medication, clinical response at data lock, timing of first clinical improvement, treatment termination date, reason for discontinuation and side effects.

Calculation of SII and SIRI

SII and SIRI were calculated using the complete blood count as follows [22]:

SII = platelet count [G/L] × neutrophil count [G/L]/lymphocyte count [G/L].

SIRI = monocyte count [G/L] × neutrophil count [G/L]/lymphocyte count [G/L].

Procedures

Based on recent CSU guidelines, participants who remained symptomatic after 4 weeks of antihistamine therapy with four times the approved dose were eligible for treatment with 300 mg subcutaneous omalizumab every 4 weeks [1, 23]. Blood samples were drawn prior to the first administration to establish baseline IgE levels. First patient follow-ups and efficacy assessments took place three months after treatment initiation. Clinical responses were rated into three categories based on to the patients’ reported outcomes as well as their dermatologists’ judgments: good, partial and no response. Good responders were defined as patients whose symptoms completely vanished during omalizumab treatment. Partial responders demonstrated slight improvement only and, hence, experienced recurrence of wheals either less frequently or less intensively under administration of omalizumab. Patients who demonstrated insufficient clinical improvement were considered for higher omalizumab doses as per recent literature [24,25,26]. This led to exclusion, as previous studies demonstrated the dose-effect-relationship to influence response [9]. In accordance with current guidelines, treatment was regularly interrupted every five to six months to assess whether urticaria persisted [1, 27]. During these interruptions, the spontaneous course of the disease was evaluated, and treatment was only restarted if patients experienced recurrence of urticaria symptoms. Post-withdrawal relapse led to a deduction of injection free months from total treatment time and re-administration of the last effective dose [28]. If treatment was discontinued, either due to resolution of symptoms or a change in therapeutic strategy, the last injection date was recorded as termination date. Partial responders were excluded, resulting in a clear responder and a non-responder group. Drug survival was defined as treatment duration, with an event being defined as the definitive discontinuation of omalizumab. This study was conducted in accordance with good clinical practice.

Statistical analysis

Statistical analysis was conducted using R software (version 4.2.2) with p-values ≤ 0.05 indicating significance. Kaplan-Meier survival curves were used to visualize drug survival. Initial screening of associations with drug survival was performed via univariate Cox regression analysis. The proportional hazard assumption was verified by examining individual Schoenfeld residuals. Previous studies used p-values ≤ 0.2 for inclusion in the multivariate Cox regression model [15]. Due to only IgE passing this screening, the minimum p-value was raised to ≤ 0.3, allowing both IgE and sex to be included in the multivariate Cox regression curves. Three multivariate models with baseline variables alone, baseline variables and SII, and with baseline variables and SIRI were evaluated. The 66 (60.6%) patients who were still on continuous treatment with omalizumab at data lock were not excluded due to group size.

The Cox regression model describes the hazard function as a function of time and covariates to assess event risk (e.g. the discontinuation of omalizumab) at a given time, assuming survival (e.g. regular omalizumab administration) up to that point. The hazard ratio, represented by Exp(coeff), represents omalizumab discontinuation risk change per unit increase in the covariate, with all other model covariates held constant.

Results

Sample size

From January 2018 to May 2021, 181 patients were assessed for eligibility. A total of 109 patients were enrolled in this study, as they were ≥ 18 years old as of index date, had a definitive diagnosis of CSU made by a board-certified allergologist or dermatologist, had ≥ 4 weeks of treatment with oral antihistamines, were treated with 300 mg omalizumab per 4 weeks, were not partial responders, and had given written general consent.

Demographic characteristics

As shown in Table 1, the overall population, consisting of 63 (57.8%) female and 46 (42.2%) male patients showed a mean (± SD) age of 40.6 ± 14.7 years. All patients had hives and wheals, whereas 46 (42.2%) patients had concomitant angioedema. 41 (37.6%) patients were identified as having an inducible component, classified as CSU with an inducible component (CSU-CindU). Ten (9.1%) patients demonstrated concomitant autoimmune disease, defined as the presence of any autoimmune disease such as Hashimoto’s thyroiditis at the beginning of omalizumab therapy. Atopy, which was defined as a clinical history of allergic rhinitis, asthma, or atopic dermatitis, was positive in 39 (35.7%) patients. The medical records of 4 (3.7%) patients did not contain a conclusive start of symptoms date. The remaining 105 (96.3%) patients showed an average disease duration prior to initiation of omalizumab treatment of 49.2 ± 74.7 months.

Table 1 Demographics and individual co-medication of all CSU cases analyzed (n = 109)
Full size table

Immunomodulatory medication was administered to 33 (30.3%) patients, either as an adjunctive to omalizumab treatment, or for pre-existing concomitant autoimmune disease. Prednisone was administered to 30 (27.5%) patients, of which 9 (30.0%) on a continuous basis and 21 (70.0%) on an as-needed basis, 2 (1.8%) patients received cyclosporine and 3 (2.8%) with others (1 nivaquine, 1 plaquenil, 1 azathioprine). All the patients included in this study were co-medicated with oral antihistamines during their omalizumab treatment. Additionally, 10 (9.2%) patients were treated with montelukast. One patient was treated with colchicine and dapsone.

Laboratory evaluations

As depicted in Table 2, 90 (82.56%) patients underwent a differential blood count before initiating omalizumab treatment, showing an average SII of 796.1 ± 961.3 and average SIRI of 2.1 ± 3.1. The average plasma levels for various blood components were as follows: neutrophils (5.2 ± 2.4 G/L), lymphocytes (2.3 ± 1.1 G/L), monocytes (0.8 ± 1.4 G/L), eosinophils (0.2 ± 0.1 G/L) and basophils (0.03 ± 0.02 G/L). Regarding specific IgE measurements, pre-treatment data was available for 69 (63.3%) patients, with an average baseline IgE level of 205 ± 250.2 IU/mL. Non-responders demonstrated an average baseline IgE level of 103.3 ± 118.5 IU/mL. Additionally, 73 (70.0%) patients were tested for tryptase, showing an average baseline value of 5.4 ± 3.3 ug/L.

Table 2 Blood values in patients with CSU
Full size table

Response at data lock

Table 3 shows the clinical responses of all patients at data lock. Of the 109 patients included in this study, 95 (87.2%) responded positively to treatment, while 14 (12.8%) were classified as non-responders. Among the 95 responders, 53 (55.8%) demonstrated an early response (prior to the second administration) and 35 (36.8%) showed a delayed response (between the second and fourth administration). The exact timing of improvement was not documented for the remaining 7 (7.3%) responders.

Table 3 Responses of CSU cases analyzed (n = 109)
Full size table

Drug survival

Amongst all 109 patients included in this study, mean treatment duration was 25.7 ± 19.6 months at time of data lock. Of the 43 (39.4%) patients who had discontinued omalizumab treatment, average treatment duration was 13.6 ± 10.9 months, as seen in Fig. 2. More than half of these patients discontinued treatment due to remission, as seen in Table 4.

Fig. 2
figure 2

Omalizumab drug survival in patients with CSU (n = 43)

Full size image
Table 4 Reasons for omalizumab discontinuation
Full size table

Table 5 presents the univariate model representing sex, age, concomitant autoimmune disease, baseline IgE, tryptase, SII and SIRI’s effects on omalizumab drug survival of omalizumab. None of the variables had significant Schoenfeld residuals, confirming survival testing assumptions. Three variables (sex [p-value = 0.281], concomitant autoimmune disease [p-value = 0.106] and baseline serum IgE [p-value = 0.195]) showed p-values < 0.3, warranting inclusion in a primary multivariate Cox regression model to assess for drug survival. However, concomitant autoimmune disease did not correlate with drug survival in the multivariate model (p-value = 0.915). After exclusion of concomitant autoimmune disease, SII and SIRI demonstrated a p-value of 0.367 and 0.098 respectively in the multivariate model including sex and baseline IgE, as demonstrated in Table 6. The correlation between SIRI and the drug survival of omalizumab is shown in Fig. 3.

Table 5 Results of univariate Cox regression analysis
Full size table
Table 6 Results of multivariate Cox regression analysis
Full size table
Fig. 3
figure 3

Correlation between SIRI and omalizumab drug survival

Full size image

Discussion

Among 109 patients with chronic spontaneous urticaria (CSU) enrolled in this retrospective single center study, 43 (39.4%) patients discontinued omalizumab treatment after a mean treatment duration of 13.6 ± 10.9 months, which is comparable to the findings of previous studies, two of which postulated a mean drug survival between 13.3 and 14.6 months, respectively [14, 29]. However, both studies considered temporary pauses as an event in determining drug survival. A third study described a drug survival of 8.5 months only [30].

Accumulating evidence supports the systemic immune-inflammation index (SII) to be a precise predictor of outcome in tumors and other diseases [18, 31, 32]. Similar to SII, the systemic inflammatory response index (SIRI) also emerged as an independent prognostic factor for comparable diseases. [20, 33, 34]. SII and SIRI, identified as reliable inflammatory markers, could be associated to CSU and may predict omalizumab drug survival. In our study, average SII and SIRI were 796.1 ± 961.3 and 2.1 ± 3.1 respectively, which is slightly higher than 552 (137–2270) and 1.4 ± 0.7 reported by Coşansu [21].

In this study, the univariate analysis did not demonstrate significant p-values for neither SII (p-value = 0.635) nor SIRI (p-value = 0.763), suggesting that neither factor in isolation had a strong correlation with omalizumab drug survival. While not reaching statistical significance, the multivariate analysis revealed pre-treatment SIRI (p-value = 0.098) as a more reliable omalizumab drug survival predictor than pre-treatment SII (p-value = 0.393). This suggests that SIRI might have some bearing on omalizumab drug survival and that a high baseline SIRI could improve omalizumab drug survival. These results are consistent with previous studies such as Coşansu et al., who established a correlation between both SIRI and SII with omalizumab response at three months in CSU [21]. However, further investigation is needed to establish their potential predictive value.

Current literature suggests a reduction of mast cell releasability and basopenia reversion under omalizumab therapy [35, 36]. This result is supported by current literature, which underlines the important role of mast-cell activation in the pathogenesis of chronic spontaneous urticaria [37]. Although the significance of basophil levels in CSU has already been documented in previous studies [38,39,40], basophil levels are included similarly in the calculation of SII and SIRI and can therefore not explain the difference in prognostic factor of omalizumab drug survival.

CSU is considered a risk factor for autoimmune diseases [41], with prevalence up to 28% in patients with CSU [42]. However, only 10 (9.1%) patients in this study had concomitant autoimmune disease. The multivariate analysis failed to demonstrate a significant effect of concomitant autoimmune disease on omalizumab drug survival (p-value = 0.915). This is an important finding which reflects results of a recent study, showing that thyroid autoimmunity, formerly considered a potential marker of type IIb CSU, is unable to predict omalizumab response [43].

In spite of prior research indicating better and faster response to omalizumab treatment in patients with elevated baseline serum IgE, this study found limited prognostic value for omalizumab drug survival from baseline IgE levels (p-values of 0.367 and 0.5421 in multivariate analysis with SII and SIRI, respectively) [16, 17, 44]. However, mean baseline IgE levels were higher than other studies, suggesting that the majority of the cohort presented with type I CSU, a condition associated with a prompt response to omalizumab [45]. Consistent with prior research, poor responders demonstrated an average baseline IgE level of 103.3 ± 118.5 IU/mL only [35, 45, 46].

Therefore, baseline IgE, is a predictor of omalizumab response, but might not always be a reliable indicator of duration of the disease and, hence, of omalizumab drug survival. Individual omalizumab dosing was adjusted in accordance with clinical results, as described in recent literature [24,25,26]—patients with little to no clinical improvement qualified for higher omalizumab doses, typically 600 mg every 4 weeks or 300 mg every 2 weeks. This led to exclusion in our study due to the demonstrated dose-effect-relationship influencing response [9].

This study provides valuable insights into omalizumabs’ effectiveness in treating CSU. One significant strength is the centralized patient care at the University Hospital Zurich, at one of the largest university hospitals in Switzerland, enabling standardized diagnosis, management, treatment, improved patient follow-up, minimized confounders, and comprehensive data collection. A limitation, however, is that no validated numerical instrument such as the Urticaria Activity Score [47] was used to measure symptom reduction more accurately with omalizumab treatment. Lastly, co-medication usage might have influenced the clinical response and therefore the drug survival of omalizumab.

In conclusion, this study found limited predictive value for omalizumab drug survival from concomitant autoimmune disease. The systemic inflammatory response index (SIRI) proved to be a more reliable predictor than the systemic immune-inflammation index (SII). However, neither of both indices reached statistical significance and can therefore be promoted as a useful predictive tool. Nevertheless, these findings contribute to ongoing efforts to optimize omalizumab treatment for CSU.