Antinuclear antibodies may predict the development of immune-related adverse events in asymptomatic patients treated with immune checkpoint inhibitors: results from a single-center cohort

We aim at investigating the association between subclinical autoimmunity and immune-related adverse events (irAEs) in a cohort of patients treated by immune checkpoint inhibitors for solid metastatic cancer. In the context of an oncology/rheumatology outpatient clinic, we evaluated patients treated with anti-PD-1 or anti-PD-L1. Before treatment, each patient underwent a physical evaluation and a blood sample to identify the presence of a set of autoantibodies. Indeed, all the patients were followed during treatment to identify irAEs and to assess the association with autoantibodies. Fifty-one patients (M/F 16/35; median age 70 years, IQR 16.5) were evaluated; 34.8% of patients showed ANA positivity, 6.5% ENA positivity (anti-SSA), 4.3% Ratest positivity, and 2.1% (one patient) ACPA positivity. During a median period of 21 months (IQR 38.75), 39.2% of patients developed irAEs. Musculo-skeletal manifestations, in particular arthritis, were the most frequent. We found a significant association between the positivity for ANA and the development of irAES (p = 0.03, RR 2.01, 95% CI 1.03–3.92). Furthermore, the progression-free survival was significantly longer in patients developing irAEs compared to those who are not experiencing these events (p = 0.007). This study underlines the potential role of ANA positivity as a predictive biomarker for the development of irAEs. Supplementary Information The online version contains supplementary material available at 10.1007/s10238-024-01317-z.


Introduction
In the past decade, immune checkpoint inhibitors (ICIs) have emerged as the predominant therapeutic approach in the field of oncology, leading to a major shift in the cancer management, with a significant impact on patients prognosis.The increasing use of ICIs in oncological clinical practice focused attention on ICIs-related side effects, the so-called immune-related adverse events (irAEs).These last encompass a spectrum ranging from benign, self-resolving conditions to severe, life-threatening clinical presentations, potentially affecting every organ/system, including musculoskeletal system [1,2].
When evaluated exclusively from an oncological perspective, irAEs could have been inadequately assessed and then underdiagnosed.Indeed, the application of a multidisci-plinary approach could provide more detailed information in terms of prevalence, phenotype, severity and management of irAEs.It is relevant that this shared approach plays an important role in the decision whether or not to continue ICIs treatment when these adverse events develop [3].
Taken all together, these observations opened new research fields focusing on the identification of biomarkers able to predict the development of irAEs and maximize the risk-benefit ratio of ICIs treatment [4].
If one can expect a relapse of a pre-existing autoimmune disease during ICIs administration [5], little is known about the evolution of the so-called "subclinical autoimmunity."This condition, consisting of seropositivity for autoantibodies without any clinical signs or symptoms of autoimmune diseases, may be triggered by ICIs treatment, which would therefore act as a "second hit," thus leading to the development of irAEs.
Building on these premises, the primary objective of our study was to investigate the association between subclinical autoimmunity and irAEs development in a cohort of patients treated by ICIs for solid metastatic cancer.

Materials and methods
We evaluated a cohort of patients diagnosed with solid metastatic cancer as a part of an oncology/rheumatology outpatient clinic.In detail, the patients were referred from the Oncology Department, Policlinico Umberto I, Sapienza University of Rome, to the Arthritis Center of Sapienza University of Rome.
For the present analysis, we included all the patients eligible for treatment with ICIs targeting PD-1 (pembrolizumab, nivolumab) or PD-L1 (cemiplimab, avelumab, durvalumab).Patients with previous autoimmune diseases were excluded from the study.
Before starting immunotherapy, each patient underwent a complete physical examination.Furthermore, a blood sample was collected to identify the presence of specific autoantibodies before treatment starting.This evaluation included: • Antinuclear antibodies (ANA) and anti-dsDNA, detected by means of indirect immunofluorescence [6,7]; • Rheumatoid factor (RF), anti-citrullinated protein antibodies (ACPA), and extractable nuclear antigen antibodies (ENA) detected by using commercial ELISA kits (results evaluated according to the manufacturers' instructions).
The study was conducted in compliance with the principles outlined in the Declaration of Helsinki, and informed consent was obtained from each patient.

Follow-up
Patients were educated about possible irAEs and were explained how to recognize them.According to oncological diagnosis, disease stage, and medication, each patient was clinically followed and each side effect related to ICIs treatment as well as chemotherapy was reported.
Furthermore, data about the overall survival (OS) and the progression-free survival (PFS) were registered.A rheumatological evaluation was performed each time a suspicious sign or symptom emerged.

Statistical analysis
The statistical analyses were performed using version 9.0 of the GraphPad statistical package.Normally distributed variables were summarized using the mean ± standard deviation (SD), and non-normally distributed variables by the median and interquartile range (IQR).Frequencies were expressed by percentage.Univariate comparisons between nominal variables were calculated using the chi-square test or Fisher's exact test, where appropriate.Two-tailed p values were reported; p values less than 0.05 were considered significant.In the follow-up analysis, the estimated risk of development of a specific event was determined with the relative risk (RR).

Longitudinal analysis
Patients were followed up for a median period of 21 months (IQR 38.75).The median overall survival (OS) was 18 months (IQR 18) and the median progression-free survival (PFS) was 6 months (IQR 11).
No differences were found in the irAEs prevalence according to administered drug.
No differences were found between ANA + and ANA − patients in terms of treatment response and PD-L1 expression (Supplementary Table 1-S1).Similarly, no differences were identified when comparing ANA + and ANA − patients in terms of PFS [9.5 months (IQR 19) versus 5 months (IQR 9.5); p = 0.08) and OS [18.5 months (13.75) versus 18 months (IQR 20); p = 0.39].When stratifying patients by the occurrence of irAEs, the PFS was found to be significantly longer in patients developing adverse events compared to those who did not experience these events [10 months (IQR 16.5) versus 4 months (IQR 9.5); p = 0.007].
Indeed, arthritis was found to be the most frequent "irAEs" and developed in six patients (30% of total irAEs).Three of them were treated with nivolumab and three with pembrolizumab.The serological features of these six patients are reported in Table 1: of note, 5 out of 6 (83.3%) were ANA positive, associated with other autoantibodies positivity in 4 patients.Of note, one patient positive for ANA, Ratest, and anti-SSA, developed lower limb purpura, xerostomia and lymphopenia associated with arthritis.Accordingly, a diagnosis of Sjögren Syndrome (SS) was made (patient 2).
All the patients were treated with a median glucocorticoids (GCs) dose of 10 mg (IQR 5) with fast tapering.Treatment with cDMARDs was added in three patients; specifically, hydroxychloroquine (HCQ) was added in the patient diagnosed with SS and in a second patient due to the persistence of arthralgia at GCs tapering (patient 4).Finally, one patient received methotrexate (10 mg weekly) due to persistent arthritis despite GCs treatment; this patient was diagnosed with undifferentiated arthritis (patient 6).

Discussion
In this single-center prospective study, we evaluated the association between a pre-existing condition of subclinical autoimmunity and the development of irAEs in patients treated by ICIs.Indeed, our investigation underlines the potential role of positivity for ANA as a predictive biomarker for the development of these immune-related side effects.This finding could contribute to a growing body of evidence highlighting the intricate relationship between autoimmunity and irAEs occurrence.In the last years, several studies have reported data about the development of immune-related conditions in patients treated by ICIs [3].However, in the majority of cases, no data are available about the autoantibodies status before starting the treatment.Thus, little is known about the role of pre-existing autoimmunity, and the available data reported contrasting results.We previously described the case of a 69-year-old Fig. 2 Incidence of immune-related Adverse Events (irAEs) in ANA + and ANA − patients male patient affected by advanced cutaneous squamous cell carcinoma in the head and neck with pre-existing ANA positivity, developing arthritis 6 weeks after the introduction of nivolumab treatment [8].
A recent paper published by Hsu and colleagues, including ninety-three liver cancer patients treated with anti-PD1, suggested the role of ANA positivity in predicting the development and severity of irAEs.Furthermore, in this cohort, the time to develop irAEs was significantly shorter in seropositive patients [9].Similar results were reported in a second larger cohort of NSCLC patients treated with nivolumab or pembrolizumab.Indeed, positivity for ANA, rheumatoid factor and anti-thyroid antibodies was an independent factor for the development of irAEs [10].
On the other side, other studies, prevalently evaluating cohorts of patients with NSCLC, did not find any predictive role for pre-existing ANA positivity in the development of irAEs [11][12][13].
These contrasting results could be related to different aspects.First, these studies are not comparable, due to the great heterogeneity of cancer types, ICIs therapy protocol, disease stage, comorbidities, and irAEs assessment.Not least, the cut off for autoantibodies positivity, especially ANA, was different in the published studies [10,13].
Besides the prediction of the irAEs itself, Campochiaro and colleagues showed that the autoantibodies may predict the disease severity and the need for more aggressive therapy.However, in this study, the serological evaluation was contextual to the development of the adverse event and no information was available about autoantibodies before immunotherapy [14].
In light of these divergent findings, it is worth mentioning the first EULAR recommendations released in 2021 [15].These guidelines advise against testing asymptomatic patients for autoantibodies, aiming to underscore that the decision to initiate ICIs therapy should not be hindered solely by the presence of a pre-existing autoimmune disease or sieropositivity for autoantibodies in asymptomatic patients.Consequently, our results should be interpreted in accordance with these recommendations, serving as a valuable contribution to the understanding of the relationship between autoimmunity and the likelihood of developing irAEs.Furthermore, in line with the same recommendations, irAEs could be viewed as potential markers of treatment efficacy.This is clearly reported by different studies, showing an improved PFS in ANA-positive NSCLC patients treated with anti-PD1 drugs [10,16].
In our cohort, we did not find the association between ANA positivity and better survival.Nevertheless, this analysis was not the main objective of our study and the results could be influenced by cancer types, treatment protocol before ICIs treatment and disease stages.Our study has some strengths and some limits.First, this is a longitudinal study performed in a dedicated oncological/rheumatological outpatient clinic with easy dedicated access for cancer patients.We thus evaluated each patient with any suspicious ICIs side effects, allowing us to early identify and treat rheumatological irAEs.Furthermore, we analyzed a homogenous cohort in terms of ICIs treatment, because all the patients were treated by anti-PD-1 or anti-PD-L1 drugs.The major limits were certainly the cohort size and the heterogeneity in cancer disease and stage.Indeed, various diagnoses could negatively influence the statistical power of survival analysis.Indeed.the statistical result needs to be confirmed on larger prospective studies.Moreover, the heterogeneity of oncological diagnosis and stage should be kept in consideration when evaluating our results.Finally, it could be interesting to evaluate the evolution of ANA-negative patients who became ANA-positive during ICIs treatment, but this was not the purpose of the present study and we do not have information about change in the serological status.
In conclusion, our results underline the potential role of ANA positivity as a predictive factor for irAEs development in oncological patients undergoing ICIs therapy.

Fig. 1 a
Fig. 1 a Prevalence of autoantibodies in the analyzed cohort before immune-checkpoint inhibitors treatment, b Kaplan-Meier analysis of the occurrence of immune-related Adverse Events in the overall patients and by dividing ANA + and ANA − patients (Log-rank Mantel-cox test, (62.5% versus 31.0%;p = 0.029).c Prevalence of immune-related Adverse Events (irAEs) developed during follow-up /die + HCQ 200 mg/die Introduction of HCQ due to persistence of arthralgias 5 /die and + MTX 10 mg/weekly Introduction of MTX due to persistence of arthritis

Table 1
Serological status and treatment of six patients experiencing arthritis during treatment with immune-checkpoints inhibitors (ICIs)