In this study, we analyzed immunoglobulin levels in a cohort of patients with severe COPD who were admitted to hospital for an AECOPD and found that low IgG and IgM levels were common among these patients. Moreover, IgM levels were associated with a longer hospital stay during the index admission for COPD exacerbation, even after adjustment for other variables. In addition, IgM levels in the whole sample showed a negative correlation with hospital stay during the index admission supporting the observation above. Conversely, IgG levels were not associated with any of the clinical outcomes of the study.
Fifty-one percentage of our cohort had below the normal range levels for at least one immunoglobulin class (27% had low IgG, 31.4% of had low IgM and 2% had low IgA, while few subjects had more than one class low) at presentation. Unfortunately, we do not have longitudinal data on the cohort, so we do not know if this decrease represents an acute drop preceding the exacerbation, or a stable characteristic of these subjects with COPD. Previous studies have shown variable rates of IgG decrease in patients with COPD ranging from 11% [8] up to 20%. Low IgG levels were associated with lower BMI and this association suggests that the nutritional status of the patient may be linked to their IgG levels. The exact reasons, however, for low immunoglobulin levels in patients with COPD are not clear. Many factors may be at play in addition to poor nutrition, including chronic disease or chronic oral corticosteroid use. A previous study has shown that patients with COPD on oral corticosteroids have lower IgG levels than those who were not on corticosteroids [11]; whether this association depends on the severity of the disease, or is linked directly to corticosteroid use, is not clear.
Subjects with low IgM levels at presentation had higher length of stay for the index admission compared to those with normal IgM levels. This finding was further confirmed adjusting with age and biological sex. There are multiple potential explanations for this observation. Since the most frequent reason for COPD exacerbations is infections, low IgM may mean that the immune system is not able to mount an immediate response against the infectious agent and therefore these patients present with increased severity of infection that would require longer hospital stay. It would be beneficial to design a longitudinal study to understand whether there is a specific phenotype of patients presenting with AECOPD who have low IgM levels and whether this translates into an endotype with pathophysiological significance. If this were the case, this may be a population where focused efforts, such as close medical follow up, optimized medical treatment including prophylactic antibiotics, appropriate vaccinations and other preventive measures are needed to prevent subsequent infections. It would also have been interesting to have data on the bacteriology of infections leading to AECOPD, but we do not have such data in our study.
When the sample was sub classified into those with low and with normal IgG levels, we found a trend towards a negative correlation between the number of AECOPD-related hospital readmissions and serum IgG only in the subjects with decreased IgG levels. One limitation regarding this observation is that we had a very small number of patients with low IgG in our population. This observation may indicate that the degree of IgG level decrease is associated with an increased risk of infections. This observation is concordant with observations from a previous study that showed worse outcomes with decreasing IgG levels, although in that case the difference could be seen across the whole spectrum of IgG levels [10]. Larger prospective studies in similar populations will provide more robust evidence of the associations between IgG levels and the course of COPD.
The study has also a number of limitations. First, the population is small, and the observation needs to be validated in a larger study of patients presenting with COPD exacerbations. Second, the measurement of immunoglobulins in our study was performed in serum collected during the presentation with AECOPD. Therefore, it cannot be determined whether immunoglobulin levels change acutely around the time of an AECOPD and whether immunoglobulin levels during convalescent periods would also correlate with the same outcomes. Third, there was limited granular data on the exacerbation, management, and previous health, which all impact the primary outcomes of length of stay. Fourth, there is no data demonstrating acute alterations of immunoglobulin levels during an exacerbation. This may be a possibility due to the likely presence of a concurrent infection and the increased dose of mediations, especially oral steroids, subjects may be taking during exacerbations. Finally, the study was conducted in Canada, where access to health care and drugs is federally regulated through the Canada Health Act. The study should be reproduced in other settings.