乌鸦传媒

Rates of new-onset autoimmune inflammatory rheumatic disease (AIRD) such as rheumatoid arthritis and systemic lupus erythematosus were significantly increased following bouts of COVID-19 in South Korea and Japan, researchers found.

With data from large repositories in the two countries, AIRD rates were 25% higher in South Korea (95% CI 18-31) and 79% greater in Japan (95% CI 77-82) among COVID-19 patients versus uninfected controls from the general population, according to Dong Keon Yon, MD, PhD, of Kyung Hee University in Seoul, and colleagues.

Absolute rates after COVID were 1.15% in Korea and 3.87% in Japan.

However, vaccination against SARS-CoV-2 reduced the likelihood of developing AIRD following breakthrough infection, except when those infections became severe, the researchers . In fact, severity of COVID-19 increased AIRD risks across the board.

Yon and colleagues stopped short of calling AIRD a form of "long COVID," in which fatigue, malaise, and respiratory symptoms typically predominate. What they did conclude was that AIRD appears to qualify as a long-term COVID-19 complication: AIRD development rates remained strongly elevated in both countries up to a year after infection, and beyond that in the Japanese data (HR 1.57 vs general population, 95% CI 1.50-1.64) though not in Korea.

The risk increase wasn't an artifact of COVID-19 patients having more contact with healthcare systems and thus heightened observation, the researchers emphasized. Their analyses also included comparisons between COVID-19 patients and people treated for influenza; AIRD rates were significantly greater in the COVID groups, by 30% in the Korean data and by 14% in Japan.

Other studies had examined rheumatologic disease rates following COVID-19, which also found increases in risk. But those studies didn't account for possible ascertainment bias, Yon and colleagues pointed out, nor did they look at vaccination's potential influence.

The new analysis used a case-control design. Korean national data included some 394,000 people with documented COVID-19 infection. Yon and colleagues selected about 177,000 for propensity matching; each was coupled with four uninfected (neither COVID nor influenza) individuals in the general population, for a total of 676,000 controls. The comparison between COVID and flu patients included 95,000 in each group. Matching covered numerous health-related and sociodemographic parameters at baseline. The Japanese cohorts included 961,000 COVID-19 patients to be compared with 1.6 million uninfected people; for COVID versus influenza, the group sizes were 115,000 and 110,000, respectively.

Severity of COVID-19 was established through records of treatments such as intensive care admission and extracorporeal membrane oxygenation. Vaccination status, including the number of vaccine doses, was also contained in the data. Yon and colleagues identified AIRD cases in two categories: inflammatory arthritis, comprising rheumatoid and psoriatic arthritis and spondyloarthritis; and connective tissue diseases, which included lupus, Sjögren's syndrome, systemic sclerosis, polymyalgia rheumatica, mixed connective tissue disease, dermatomyositis, polymyositis, polyarteritis nodosa, and vasculitis.

Not all the results were similar between countries or in the comparisons between COVID and influenza. For example, no elevation in rates of inflammatory arthritis was seen in Korea with respect to the general population (HR 0.90, 95% CI 0.65-1.24), yet in Japan the risk was doubled (HR 2.02, 95% CI 1.96-2.07). On the other hand, Koreans with COVID went on to face substantially greater risk for inflammatory arthritis than did those with flu, albeit without statistical significance (HR 1.92, 95% CI 0.34-3.65), while in Japan the rates hardly differed (HR 1.07, 95% CI 1.03-1.13).

Patterns were more consistent for connective tissue diseases, with rates significantly higher after COVID-19 in both types of comparison in both countries.

Elevations in AIRD rates were apparent early, by 26% in Korea and 87% in Japan relative to general population controls within the first 6 months after infection, and by 33% and 44% in Korea and Japan, respectively, relative to influenza patients. The elevations shrank a little during the period 6 to 12 months after infection. After 1 year, there was no difference in either country between previous COVID and flu patients, and only in Japan did the elevation relative to the general population persist.

Yon and colleagues also looked at the influence of COVID severity on AIRD rates, both overall and with respect to vaccination status. Relative to uninfected controls, AIRD rates in Korea were 22% greater in patients with mild COVID, but 42% higher in moderate-severe cases. AIRD was also less frequent in COVID survivors who had been vaccinated, with hazard ratios of 0.59 after one dose and 0.42 after two (both P<0.05).

Vaccination did not, however, protect fully against AIRD when COVID was rated as moderate-severe (HR 1.30 vs uninfected controls).

The researchers cited a number of limitations to the analysis, including the reliance on administrative data, the possibility of unmeasured confounders, and the restriction to East Asian populations. Also, the data were recorded prior to 2022 and thus predated the broad spread of the Omicron SARS-CoV-2 variant. Sample sizes for some types of AIRD were probably too small to yield meaningful results.

Comment