Introduction

Increased risk of malignancy has been reported in various autoimmune diseases, often attributed to chronic inflammation, immune dysregulation, multi-organ involvement, and the use of immunosuppressive therapies1. However, IgG4-related disease (IgG4-RD) exhibits unique pathophysiological characteristics that distinguish it from other autoimmune conditions. While specific autoantibodies and well-defined mechanisms are lacking, the disease is marked by infiltration of IgG4-positive plasma cells, storiform fibrosis, and obliterative phlebitis. IgG4-RD can affect nearly any organ system - including the pancreas, salivary glands, biliary tree, kidneys, and lungs - and frequently presents as mass-like (“tumefactive”) lesions that respond well to glucocorticoid therapy2,3. These distinct features suggest that the pattern and magnitude of cancer risk in IgG4-RD may differ from those observed in other autoimmune diseases.

Although several studies have reported an increased risk of malignancies such as lymphoma, pancreatic cancer, and lung cancer in patients with IgG4-RD4,5,6,7, the findings remain inconclusive. Discrepancies across studies may reflect variations in diagnostic criteria, study design (e.g., retrospective vs. prospective), duration of follow-up, and population characteristics. Moreover, the rarity of IgG4-RD and its clinical similarity to cancer-like lesions can confound the interpretation of cancer risk. Some early studies were limited by small sample sizes and short follow-up durations, which may have resulted in imprecise or exaggerated estimates of malignancy risk.

To address these limitations, we conducted a large-scale, population-based cohort study using data from the Korean National Health Insurance Service (NHIS) to evaluate the risk of malignancy in patients diagnosed with IgG4-RD. By comparing cancer incidence in this cohort to that of the general population, we aimed to clarify the overall and site-specific cancer risks associated with IgG4-RD, with further stratification by age, sex, follow-up duration, and use of immunosuppressive agents.

Materials and methods

Data resource

We utilized a customized database from Korean National Health Insurance Service (NHIS) to identify patients diagnosed with IgG4-RD in Korea. Korean NHIS is a mandatory, nationwide health insurance system, which covers approximately 97% of the entire Korean population. This database includes extensive information on demographic characteristics, diagnostic codes, and prescription records for both inpatient and outpatient services. Relevant diagnostic codes were identified using the International Classification of Diseases, 10th Revision (ICD-10). Cancer incidence data for the general population was obtained from the Korean National Cancer Registry data for 20198. This registry provides annual cancer incidence rates categorized by age, sex, and cancer type for the entire Korean population.

Study population

We identified eligible patients diagnosed with IgG4-RD between January 1, 2012, and December 31, 2020, from the NHIS database. Patients with IgG4-RD were defined as those who had at least two claim with the diagnostic code for IgG4-RD (ICD-10 codes D89.8W1) within 6 months and were registered in the Rare and Intractable Disease (RID) program for IgG4-RD (RID code V900). The RID program, managed by the NHIS, provides financial support to patients with certain RID. To be enrolled in the RID program for IgG4-RD, patients have to meet the 2011 comprehensive diagnostic criteria for IgG4-related disease9. We included patients aged 20 years or older and excluded those with a previous history of cancer. The final study population consisted of patients diagnosed with IgG4-RD between January 1, 2012, and December 31, 2020 (Fig. 1). The index date was defined as the first registration date of the IgG4-RD diagnostic code. Patients were followed from the index date until the occurrence of the cancer, death, or the end of the study (December 31, 2021).

Fig. 1
Fig. 1
Full size image

Flow diagram of study population.

Baseline patient characteristics, including age, sex, and Charlson comorbidity index (CCI) scores, were collected10. CCI score, a standardized, validated method to quantify patient comorbidity burden, was calculated and used for risk adjustment. Prescription data for immunosuppressive agents, including glucocorticosteroid, hydroxychloroquine, methotrexate, azathioprine, mycophenolic acid, tacrolimus, and cyclophosphamide, were also collected after the index date.

Statistical analysis

We calculated the standardized incidence ratios (SIRs) to compare the cancer incidence between IgG4-RD patients and the general population. SIRs were estimated by dividing the observed number of cancer cases in IgG4-RD patients by the expected number of cancer cases. The expected number of cancer cases was calculated by multiplying the age- and sex-specific cancer incidence rates of the general population by the person-years of patients with IgG4-RD. The 95% confidence interval (CI) was calculated using Byar’s approximation of the Poisson distribution. Subgroup analyses were conducted by cancer type, age at diagnosis, follow-up duration. SIRs were calculated for 20 types of solid cancers and all types of hematologic cancers. Solid cancers were identified using ICD-10 codes C00–C80, and hematologic cancers included Hodgkin lymphoma (C81), non-Hodgkin lymphoma (C82–C86, C96), multiple myeloma (C90), leukemia (C91–C95), myelodysplastic syndrome (MDS) (D46), and myeloproliferative disease (D47.1). Age at diagnosis was categorized into 20–39 years, 40–59 years, 60–79 years, and ≥ 80 years. The diagnose period of malignancy was grouped into < 1 year, 1–2 years, 3–4 years, 5–6 years, and 7–8 years. SAS Enterprise Guide version 9.4 (SAS Institute Inc., Cary, NC, USA) were used for all statistical analyses.

Results

Baseline characteristics of patients with IgG4-RD

Between 2012 and 2020, a total of 2,150 patients diagnosed with IgG4-RD were identified from the NHIS database. After excluding 883 patients with a prior history of malignancy, 1,267 patients remained for further analysis (Fig. 1).

The mean age at diagnosis was 56.36 ± 13.76 years, and 42.46% of patients were female (n = 538). The mean follow-up duration was 3.22 ± 1.97 years. Glucocorticoids were the most frequently prescribed medication (92.34%), followed by azathioprine (22.49%), methotrexate (16.81%). Glucocorticoids were prescribed more than a year in 31.49% of patients (n = 399). A CCI score of ≥ 1 was observed in 81.29% of the patients (Table 1).

Table 1 Clinical characteristics of study population.
Full size table

Risk of cancer incidence in IgG4-RD patients compared to the general population

Compared to the general population and after adjustment for age and sex, the observed numbers and standardized incidence ratios (SIRs) of overall and site-specific cancers in patients with IgG4-RD are presented in Table 2 Overall SIR was in 4.12 (95% CI 3.48–4.85). Over 4,085.20 person-years of follow-up, 147 patients developed cancer, including 110 cases of solid cancers (SIR 3.33, 95% CI 2.74–4.02) and 28 hematologic cancers (SIR 15.31, 95% CI 10.17–22.13). Among solid cancers, based on incidence, pancreatic cancer was the most common site (n = 16), followed by colorectal cancer (n = 13), biliary tract cancer (n = 12), prostate cancer (n = 11), hepatocellular carcinoma (n = 10) and stomach cancer (n = 10). Based on SIR cancers, pancreatic cancer showed highest SIR (14.54, 95% CI 8.31–23.62), followed by central nervous system cancer (SIR 13.32, 95% CI 2.68–38.93), and biliary tract cancer (11.85, 95% CI 6.12–20.71). Among hematologic cancers, based on incidence, non-Hodgkin’s lymphoma was the most common cancer (n = 14), followed by MDS (n = 10), multiple myeloma (n = 3). Based on SIR, MDS showed the highest SIR (49.46, 95% CI 23.68–90.97), followed by non-Hodgkin’s lymphoma (19.95, 95% CI 10.90-33.47). According to sex, females showed a higher overall SIR (5.51, 95% CI 4.25–7.04) compared to males (SIR 3.45, 95% CI 2.75–4.28). The SIRs for hepatocellular carcinoma, bladder cancer, prostate cancer, and central nervous system cancer were significantly increased only in men, whereas those for stomach cancer, colon cancer, and breast cancer were elevated only in women (Table 2).

Table 2 Estimation of cancer risk in IgG4-RD patients.
Full size table

Risk of cancer in IgG4-RD patients by age at diagnosis and follow-up duration

We assessed the risk of cancer in individuals with IgG4-RD based on their age at diagnosis and the length of follow-up. When compared to the general population, the likelihood of developing cancer was notably higher across all age categories, with the most significant relative risk found in younger patients aged 20–39 years (SIR 12.21, 95% CI 6.09–21.86).

Analyzing the data by follow-up duration, the greatest cancer risk was observed within the first year following an IgG4-RD diagnosis, and it remained elevated in the years that followed. Although fewer patients were monitored beyond six years, the findings indicate that the heightened risk continues over time. These results emphasize that while younger patients face the highest relative risk, ongoing cancer monitoring is crucial for all age groups throughout the follow-up period (Table 3).

Table 3 Estimation of cancer risk in patients with IgG4RD by age at diagnosis and years of follow-up.
Full size table

Risk of cancer in IgG4-RD patients based on immunosuppressive agent use

We identified ever-users of azathioprine (total follow-up 285,943.78 person-years), methotrexate (total follow-up 213,759.54 person-years), hydroxychloroquine (total follow-up 64,195.66 person-years), cyclophosphamide (total follow-up 32,108.70 person-years), tacrolimus (total follow-up 48,172.53 person-years), and mycophenolic acid (total follow-up 45,158.52 person-years) during the follow-up period. The SIRs for both overall and site-specific cancers in comparison to the general population were calculated, stratified by immunosuppressive agent use. Overall cancer risk were elevated in azathioprine (SIR 2.24, 95% CI 1.37–3.46), methotrexate (SIR 2.71, 95% CI 1.58–4.34), hydroxychloroquine (SIR 7.59, 95% CI 4.04–12.98) and tacrolimus (SIR 5.67, 95% CI 2.44–11.18). No significant associations were observed among users of mycophenolic acid or cyclophosphamide. To be specific, in azathioprine users, bladder cancer and non-Hodgkin’s lymphoma risks were increased. In methotrexate users, MDS, non-Hodgkin’s lymphoma, and pancreatic cancer risks were increased. In hydroxychloroquine users, MDS, pancreatic cancer, and colorectal cancer risks were increased. In tacrolimus users, MDS, pancreatic cancer, and prostate cancer risks were increased. (Supplementary Table S1)

Discussion

This nationwide population-based study demonstrates that, compared to the general population, patients with IgG4-RD have a significantly increased risk of developing cancer. In particular, the SIRs were markedly elevated for MDS, non-Hodgkin’s lymphoma, pancreatic cancer, and biliary tract cancers. The cancer risk was especially pronounced in younger patients aged 20–39 years, and was highest during the first year following IgG4-RD diagnosis. Furthermore, compared to the general population, IgG4-RD patients who were treated with immunosuppressive agents such as azathioprine, methotrexate, and hydroxychloroquine exhibited a higher risk of malignancy.

Previous studies on the association between IgG4-RD and malignancy have reported inconsistent results, with SIRs ranging from no significant increase to values exceeding 20. For instance, studies by Yamamoto et al.4, Shiokawa et al.5, and Asano et al.6 reported SIRs of approximately 2–3, while Takahashi et al.11 and Ahn et al.12 reported much higher values (SIR 16.0 and 23.08, respectively). However, these differences may reflect methodological limitations, including small sample sizes and single-center designs. The meta-analysis by Yu et al., which pooled 10 studies, reported an overall SIR of 2.57 (95% CI, 1.72–3.84), supporting a modest but consistent increase in cancer risk among patients with IgG4-RD13. In comparison, our nationwide cohort study found a slightly higher SIR of 4.12, which may reflect the combination of improved detection in a large population-based dataset and the robust sample size. This direct comparison suggests that while the overall magnitude of cancer risk in IgG4-RD is modest, our findings provide more precise estimates and further reinforce that IgG4-RD is associated with an elevated risk of malignancy compared to the general population.

Interestingly, the spectrum of associated malignancies has shifted compared to earlier reports. While initial studies emphasized lung and gastric cancers5,6, our study found no significant association with lung cancer. Instead, the strongest associations were with hematologic malignancies (particularly non-Hodgkin’s lymphoma and MDS) and pancreaticobiliary cancers. Given that IgG4-RD frequently affects the pancreas and bile ducts, this finding is clinically coherent. Moreover, heightened surveillance and frequent imaging at diagnosis may contribute to earlier cancer detection in these organs, possibly exaggerating the observed association. This phenomenon, known as surveillance bias, may account for the especially high SIRs within the first year of follow-up. In addition, reverse causality - wherein subclinical malignancies may trigger immune dysregulation resembling IgG4-RD - cannot be excluded, as observed in other autoimmune paraneoplastic syndromes.

The strong association between IgG4-RD and lymphoid malignancies likely reflects the combined effects of chronic inflammation and underlying immune dysregulation inherent to the disease14. Persistent immune activation may foster a pro-oncogenic environment by promoting genetic instability, inducing epigenetic changes, and impairing immune surveillance, ultimately facilitating malignant transformation15,16. In our study, the markedly elevated risk of MDS—particularly among patients treated with methotrexate or hydroxychloroquine—raises concern for potential drug-related hematologic toxicity17. However, this may also reflect more severe or refractory disease requiring escalation to second-line immunosuppressive agents. Patients needing treatment beyond glucocorticoids likely represent a subset with more extensive immune dysregulation, which may itself confer an inherently higher malignancy risk.

From a clinical perspective, the use of immunosuppressive agents warrants careful consideration. While glucocorticoids remain the cornerstone of IgG4-RD treatment and are generally effective for inducing remission, second-line agents such as azathioprine and methotrexate are often used for maintenance or in steroid-refractory cases. In our study, azathioprine and methotrexate were associated with an increased risk of lymphoma, while methotrexate and hydroxychloroquine were linked to higher MDS risk. Whether these associations reflect a direct oncogenic effect of the drugs or are confounded by underlying disease severity remains uncertain. While our analysis indicated an increased cancer risk in patients treated with hydroxychloroquine, this finding should be interpreted with caution. Hydroxychloroquine is generally considered to have a favorable oncologic safety profile18. The observed association may reflect confounding by indication, as patients receiving hydroxychloroquine often have more severe or systemic disease, which itself could contribute to elevated cancer risk. Consequently, the increased risk may not represent a direct effect of hydroxychloroquine, but rather the underlying disease severity. Non-significant associations observed for mycophenolic acid and cyclophosphamide may reflect the limited number of patients exposed to these agents, resulting in reduced statistical power, and thus these findings should be interpreted cautiously; larger studies are warranted to clarify their potential impact on cancer risk. Nonetheless, these findings underscore the importance of judicious immunosuppressive use and suggest that effective disease control—even with low-dose glucocorticoids—may help mitigate cancer risk, as previously reported in patients with IgG4-sclerosing cholangitis19.

Notably, patients diagnosed with IgG4-RD at a younger age exhibited the highest relative risk of cancer. This is because SIRs are determined by comparing the number of cancer cases observed in IgG4-RD patients to the expected number in a general population matched for age and sex. In younger individuals, the baseline incidence of cancer is low, which can lead to a significantly high SIR even if the actual number of cancer cases is small20. Consequently, although younger patients seem to have the greatest relative risk, the overall cancer burden is still less than that seen in older age groups. This observation underscores the need to consider both relative and absolute risks when interpreting age-specific SIRs and devising surveillance strategies.

Our research revealed that the occurrence of cancer was most prevalent within the first year after an IgG4-RD diagnosis, with a SIR of 7.13 (95% CI 5.65–8.89). This aligns with earlier cohort studies and meta-analyses, which also identified a significantly increased cancer risk during the initial year post-diagnosis5,6. In our cohort, most cancer cases (n = 80) were detected within this first year. Contrary to previous findings that suggested cancer risk returns to normal after the first year13, we found that the elevated SIR persisted significantly for up to six years of follow-up. This ongoing risk might be attributed to the chronic inflammatory nature of IgG4-RD, extended immunosuppressive therapy, or delayed tumor development in some patients. Additionally, the large sample size and prolonged follow-up in our study may have enhanced the detection of long-term cancer risk. While the mean follow-up duration was 3.22 ± 1.97 years, a proportion of patients were followed for up to 6 years, enabling calculation of SIRs across extended intervals. These analyses suggest that cancer risk remains elevated beyond the first year, although interpretation should consider the smaller numbers of patients in later follow-up periods. Overall, these results highlight the importance of thorough cancer screening not only at the time of IgG4-RD diagnosis and during the first year but also through continued monitoring for several years afterward.

This study has several limitations. First, due to the nature of claims data, important confounding factors - most notably family history of cancer and lifestyle-related variables such as smoking, alcohol consumption, and obesity - were not available. While these limitations should be considered when interpreting our findings, the large sample size, nationwide scope, and population-based design of our study provide robust and reliable estimates of cancer risk in patients with IgG4-RD. Second, clinical details - including disease activity, extent and site of organ involvement, and serologic markers - could not be assessed, which limited our ability to evaluate potential correlations between the anatomical distribution of IgG4-RD and subsequent cancer development. Such information would provide valuable clinical insight, and future prospective studies with detailed phenotyping are needed to clarify these associations. Third, we used 2019 national cancer incidence data as the comparator, which may not fully reflect temporal trends throughout the study period and could lead to underestimation of SIRs. Finally, information on the dose, duration, and combination of immunosuppressive therapies was limited, potentially affecting the interpretation of cancer risk associated with specific agents. Nevertheless, the large sample size and nationwide scope of the data provide robust evidence on the association between IgG4-RD and cancer in a real-world population.

Conclusion

In summary, patients with IgG4-RD have a significantly increased risk of both hematologic and solid malignancies, particularly within the first year and in those with high disease burden requiring immunosuppressive therapy. Pancreaticobiliary and lymphoid cancers warrant particular attention given their strong associations. These findings highlight the need for tailored cancer surveillance strategies in IgG4-RD, ideally starting at diagnosis and continuing for several years. Future prospective cohort studies incorporating clinical, immunologic, and treatment data are essential to better characterize cancer risk and its pathophysiological underpinnings in this complex disease.