Major comorbidities of idiopathic inflammatory myositis: A population-based study using 10 years of follow up from the national health insurance in Korea

Jeong Seok Lee¹, Min Jung Kim¹, Hee Young Lee², So Yeon Ahn³, Yeong Wook Song¹, Eun Bong Lee¹, Eun Young Lee¹, Yun Jong Lee⁴, Eun Ha Kang⁴.

¹Division of Rheumatology, Department of Internal Medicine, Seoul National University Hospital, Seoul, Korea. ²Center for Preventive Medicine and Public health, Seoul National University Bundang Hospital, Seongnam, Gyeonggi-do, Korea. ³Medical Research Collaborating Center, Seoul National University Bundang Hospital, Seongnam, Gyeonggi-do, Korea. ⁴Division of Rheumatology, Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam, Gyeonggi-do, Korea.

Background/Purpose: Patients with idiopathic inflammatory myositis (IIM) suffer from comorbidities such as interstitial lung disease (ILD), cancer, and infections related to immunosuppressive agents. We evaluated incidence rate ratio (IRRs) and characteristics of major comorbidities in IIM compared to non-IIM Koreans.

Methods: A retrospective cohort study was performed using the 2005-2014 National Health Insurance Service database covering more than 99% of Korean population. Incident IIM (n = 4088) was identified by at least two visits (first visit date = index date) to the tertiary hospitals under ICD9 codes of IIM with at least 1 year devoid of such code preceding the index date. Age- and sex-matched subjects were enrolled as the unexposed at 10:1 ratio. The occurrence of ILD, cancer, herpes zoster, and tuberculosis, and death was captured using ICD9 code, disease-specific medication, and/or government-approved qualification for co-payment reduction on these outcomes.

Results: Newly diagnosed IIM patients was 4,088 (male 39.3%, age at diagnosis 51.0¡¾15.3), in 2005-2014. Most patients (> 90%) defined as such were found to have muscle biopsy, electromyogram, or multiple laboratory examinations on muscle enzymes. Their incidence rate of the above outcomes was significantly elevated; IRRs were 36.8 [95% confidence interval: 33.9-39.9] for ILD, 2.1 [2.0-2.2] for cancer, 2.2 [2.1-2.3] for herpes zoster, 3.5 [3.3-3.7] for tuberculosis, and 4.9 [4.7-5.1] for mortality. ILD and cancers were associated with shortened survival while herpes zoster and tuberculosis were not. Among cancers whose incidence rate more than 1.0 case per 1000 person-year, the most common type of cancer was metastasis of unknown origin (MUO) (6.3 [3.9-10.1]) followed by lymphoma (4.9 [3.0-7.9]) (Table 1). The temporal relationship between cancer and IIM was highest in stomach cancer showing the majority occurring within one year of IIM diagnosis (82.8%) (Fig1). Among cancers whose incidence rate more than 10.0 cases per 1000 person-year, lung cancer and metastasis of known primary cancers showed the highest incidence rate ratio and strong temporal relationship with IIM.

Conclusion: This is the first report of the nation-wide population based evaluation of IIM in Korea. The incidence of ILD, cancers, zoster, and tuberculosis was exceptionally higher in IIM than non-IIM. ILD and cancers were associated with worse survival. Advanced cancers with metastasis were strongly associated with IIM, showing high IRR in MUO and metastasis of known primary cancers.  

Table 1. Common type of cancers in IIM patents with incidence rate higher than 1.0 case per 1000 person-year.

MUO: metastasis of unknown origin (ICD9 code: C80). Metastasis: secondary malignant neoplasm of known primary sites (ICD9 code: C78, C79).

Figure 1. Time interval between IIM and cancer diagnosis among cancers with incidence rate of higher than 1.0 case per 1000 person-year.

MUO: metastasis of unknown origin (ICD9 code: C80). Metastasis: secondary malignant neoplasm of known primary sites (ICD9 code: C78, C79). *Cancers with incidence rate higher than 10.0 cases per 1000 person-year