Introduction

Immunotherapy (IT) is a well‑established treatment for allergic diseases, with over 100 years of continuous advancements enhancing its safety and efficacy.¹ Allergen IT (AIT) with inhalant allergens is primarily used for allergic rhinitis, while venom IT (VIT) significantly lowers the risk of severe reactions to Hymenoptera stings. As the only disease‑modifying treatments for allergic diseases, both AIT and VIT remain the cornerstone of modern allergology practice.^1,2 Ongoing improvements in IT formulations and accumulating clinical evidence, demonstrating enhanced tolerability and proven long‑term preventive effects, steadily broaden the range of conditions potentially responsive to this treatment. Despite its benefits, concerns persist regarding IT’s potential to induce or exacerbate autoimmune diseases (AIDs). Early apprehensions were largely based on historical case reports from the 1970s and 1980s, which described instances of autoimmune conditions, particularly vasculitis, emerging shortly after initiation of IT​.³ Many of these cases were associated with the use of older, poorly purified allergen extracts, and the proposed mechanisms often involved type III hypersensitivity reactions mediated by immune complex formation rather than classical autoimmune pathways.

These early observations were further reinforced by theoretical models based on the type 1 / type 2 T‑helper cell (Th1/Th2) paradigm, suggesting that suppressing Th2‑mediated allergic responses through IT could inadvertently promote Th1‑driven autoimmunity. However, subsequent analyses have questioned the validity of this hypothesis, noting that AID pathogenesis is far more complex and not uniformly Th1‑dependent. Although IT is now generally considered safe, rare case reports and small case series have continued to raise concerns about possible temporal associations between IT and the onset or worsening of AIDs.^4-7 Due to inconclusive and often outdated evidence,^3,8,9 national guidelines vary considerably in their approach to AIDs, frequently adopting a conservative stance and listing autoimmune conditions as contraindications to IT.^10,11 Such caution may unnecessarily limit access to IT for patients who could otherwise benefit from its disease‑modifying potential. Given the increasing prevalence of allergic and autoimmune conditions,¹² a better understanding of their relationship is essential for optimizing IT use in clinical practice. This real‑life study aimed to evaluate whether patients undergoing IT have an increased risk of developing AID or experience exacerbation of pre‑existing AID, as compared with allergic individuals not receiving IT.

Patients and methods

Our study included all patients diagnosed with allergic rhinitis (confirmed through clinical evaluation and allergy testing using either skin prick tests or serum‑specific immunoglobulin E assays) or a documented history of anaphylaxis who were treated at our tertiary allergy center between January 2014 and December 2024. Participants were retrospectively identified by searching the hospital’s electronic medical records using the following International Statistical Classification of Diseases and Related Health Problems, Tenth Revision codes: T78.2 (anaphylactic shock, unspecified), J30 (vasomotor and allergic rhinitis), and Z51.6 (encounter for desensitization and hyposensitization procedures to allergens). The study population was subsequently divided into groups based on AID status (patients without pre‑existing AID vs patients with documented AID) and IT exposure (patients receiving AIT or VIT vs those not undergoing IT). Data were collected through a combination of structured patient questionnaires and a comprehensive review of electronic and paper‑based medical records. Recorded variables included sociodemographic characteristics (age, sex, place of residence), personal and family history of AID, as well as other coexisting chronic conditions, particularly those with an immunologic or inflammatory component. Clinical data included symptom onset, date of diagnosis, disease control status, treatment details, and severity assessment. In the patients with allergic rhinitis, additional information on confirmed allergen sensitizations was recorded, whereas for the individuals with a history of anaphylaxis, data regarding the first anaphylactic episode and identified triggers were collected. Among the patients who underwent IT, specific information was gathered regarding the type of vaccine, treatment initiation date, total number of administered doses, duration of treatment, and occurrence of local or systemic adverse events. Self‑reported information was cross‑verified with the medical records. The patients were categorized into 2 primary groups based on the presence or absence of AID: those with no prior history of AID (AID‑negative group) and those with a documented diagnosis of AID prior to the observation period (AID‑positive group). Subsequently, within each primary group, the patients were stratified according to the IT status: those who had received allergen‑specific treatment (either AIT or VIT) and those who had not undergone IT. To account for differences in treatment types, a subgroup analysis was conducted, comparing the patients receiving AIT with those undergoing VIT, in order to evaluate whether the nature of IT or the underlying allergic condition influenced autoimmune risk.

For the patients undergoing IT, the observation period was defined as the time from the first vaccine dose to the last available clinical record. In the non‑IT group, the observation period started on the date of allergic disease diagnosis, and a fixed follow‑up length equal to the mean follow‑up of the IT group was assigned to ensure comparable observation times across both cohorts. The primary outcomes were cases of new‑onset AID and exacerbations of pre‑existing AID.

Results

Out of the 1466 patients initially identified as potentially eligible for inclusion, 1198 met the study criteria and had complete, analyzable clinical data. Of those, 1053 were classified as AID‑negative, while 145 had a documented diagnosis of AID prior to the observation period (Table 1). Within the AID‑negative group, 536 individuals received IT, while 517 did not. Median (IQR) follow‑up was 5.5 (3.4–7.8) years with a mean (SD) of 6 (2.7) years. During the observation period, new‑onset AID was diagnosed in 30 IT recipients (5.6%), as compared with 47 in the non‑IT group (9.1%; P = 0.03), indicating a lower incidence of new AID among the patients receiving IT. The most common diagnoses in this group included Hashimoto thyroiditis, rheumatoid arthritis, and psoriasis vulgaris. In the AID‑positive group, 52 patients underwent IT and 93 did not. Exacerbation of autoimmune symptoms was observed in 6 IT recipients (11.5%) and 20 non‑IT patients (21.5%), which was insignificant, indicating no demonstrable association between IT and disease exacerbation. Further subgroup analysis showed no differences between the patients receiving AIT vs those undergoing VIT, either in terms of new‑onset AID (AIT, 5.6% vs VIT, 5.6%; P = 0.86) or in AID exacerbations (AIT, 11.5% vs VIT, 11.5%; P >0.99), suggesting that the type of IT or the underlying allergic condition had no measurable impact on autoimmune risk.

Discussion

Our study indicates that in a real‑world setting, IT is not associated with an increased incidence of new AID cases, nor with the exacerbation of pre‑existing autoimmune conditions. These findings stand in contrast to earlier concerns, which were largely based on isolated case series and theoretical models suggesting a potential link between IT and immune dysregulation.^3-8 Much of this apprehension arose from the hypothesis that modulating Th2‑driven allergic responses through IT could inadvertently disturb the balance between Th1 and Th2 lymphocyte subsets—an axis often implicated in both allergic and autoimmune pathogenesis. Specifically, it was proposed that downregulating Th2‑driven allergic responses might indirectly enhance Th1‑mediated autoimmunity. However, more recent evidence, including our findings, supports a more nuanced immunological understanding. Instead of causing immune imbalance, IT appears to exert immunomodulatory effects. This process involves the expansion of regulatory T cells, a shift away from Th2 dominance, and an increase in interleukin 10 and transforming growth factor β—key anti‑inflammatory cytokines that maintain immune balance and prevent excessive immunce cell activation.¹³ Moreover, IT has been shown to decrease mast cell sensitivity and activation thresholds, a factor increasingly recognized not only in allergy but also in the pathophysiology of certain autoimmune conditions, such as multiple sclerosis, lupus, and rheumatoid arthritis.¹⁴ Given the complex nature of immune regulation, AIDs represent a heterogeneous group of disorders with diverse immunopathologic mechanisms, not all of which are driven by Th1 predominance.¹⁵ Building on this improved understanding of immunologic mechanisms, it is important to revisit earlier safety concerns. Initial concerns about the unfavorable association between IT and AID were largely based on outdated studies that used unrefined allergen extracts.³ Advances in IT formulations and administration protocols have likely mitigated these previously suspected risks. Our evaluation of a well‑characterized patient group in routine clinical practice offers valuable insights into the long‑term effects of IT, while the large sample size further strengthens the reliability and clinical relevance of our findings. Nonetheless, our study has some important limitations. The retrospective nature of our analysis limits causal inference, and reliance on patient‑reported outcomes introduces potential recall bias. Additionally, differences in health care access and physician decision‑making may have influenced patient selection for IT, introducing another source of bias. Finally, the broad variability in the pathophysiological mechanisms underlying different AID subtypes may limit the generalizability of our findings. Nevertheless, in view of the large sample size and long observation period, we believe that the presented data provide valuable insights into the safety of IT a in real‑world setting. Future large‑scale prospective cohort studies are warranted to further investigate mechanisms linking IT and autoimmunity, and to determine whether specific AID subtypes exhibit differential responses to immunomodulation. A more comprehensive understanding of these interactions is essential for objectively evaluating the of impact of IT on AID progression and risk.