Immune Checkpoint Inhibitor Therapy Is Linked to COVID-19 Severity

Tobias M. Hohl, MD, PhD

Immunotherapy use is associated with a higher risk for hospitalization and severe coronavirus disease 2019 (COVID-19) infection in patients with cancer and the novel respiratory virus, according to findings from an observational study of 423 patients with hematologic malignancies or solid tumors and COVID-19.

In a univariate analysis of patients with cancer who presented with COVID-19 symptoms at Memorial Sloan Kettering Cancer Center (MSK) between March 10, 2020, and April 7, 2020, checkpoint inhibitor use was the single highest risk factor for hospitalization among 411 patients (OR, 2.53; 95% CI, 1.18-5.67; P = .017). Twelve patients were hospitalized for other reasons before testing positive for severe acute respiratory syndrome coronavirus 2 (SARSCoV-2) and were consequently excluded from this evaluation.

Treatment with immuno-oncology agents was also the primary predictor of severe respiratory illness, defined as the requirement for high-flow oxygen supplementation or mechanical ventilation, in the total population (OR, 2.38; 95% CI, 1.29-4.38; P = .005; Table).¹

Table. Most Common Predictors for COVID-19 Hospitalization, Respiratory Illness Severity¹

Thirty-one patients received immuno-oncology agents within 90 days of receiving a diagnosis of COVID-19, 18 of whom were treated with pembrolizumab (Keytruda), 5 with nivolumab (Opdivo), 3 with atezolizumab (Tecentriq), and 1 each treated with avelumab (Bavencio), durvalumab (Imfinzi), ipilimumab (Yervoy), nivolumab in combination with ipilimumab, and pembrolizumab followed by nivolumab.

Notably, immune checkpoint inhibitor (ICI) therapy was a risk factor for severe outcomes in immunotherapy-receiving patients independent of age, cancer type, and other comorbidities. The trend seen in the univariate analysis was also observed in the multivariate analysis, where the ORs for ICI use and hospitalization and severe COVID-19 were 2.84 (95% CI, 1.24-6.72; P = .013) and 2.74 (95% CI, 1.37-5.46; P = .004), respectively.

Study author Tobias M. Hohl, MD, PhD, said these results were unexpected. “Our finding that immune checkpoint inhibitors correlated with more severe disease, in particular the need for oxygen support, was a little surprising,” Hohl explained in an interview with OncLive^®.

Lung Cancer Risks

Data indicated that COVID-19 outcomes were more severe in ICI recipients with lung cancer compared with any other malignancy. “This was expected based on the organ that COVID-19 primarily affects,” said Hohl, chief of Infectious Diseases Service at MSK in New York, New York.

In a stratum-specific point estimate of outcomes by immunotherapy treatment and underlying solid cancer stratified by lung versus nonlung cancer, investigators observed higher rates of hospitalization and worsened COVID-19 infection in patients with lung cancer. Data indicated that 83% (10 of 12) and 58% (7 of 12) of ICI-treated patients with lung cancer were hospitalized and had severe respiratory illness, respectively, versus 47% (8 of 17) and 26% (5 of 19) of patients with other solid tumors. Investigators said these phenomena could possibly be attributed to ICI-triggered immune dysregulation through T-cell activation resulting in ICI-related lung injury. Subsequently, this could lead to acute respiratory distress syndrome, which is recognized as a complication of COVID-19.

“One hypothesis that comes out of our observations is that an overly robust immune response can be damaging in patients with COVID-19 and cancer and can make it more challenging for the lung to perform its oxygen exchange function,” Hohl said. “This need for more oxygen in these patients might be linked to a more robust or overly robust immune response triggered by this class of agents. We don’t have data to test this hypothesis because at the peak of the epidemic, we were not able to collect the kinds of biological samples that would allow us to address this question more mechanistically, but this analysis will certainly come later.”

Impact of Recent Therapy

In addition to ICI therapy, age greater than 65 years also correlated with increased hospitalization and severe COVID-19. Of note, neither metastatic disease nor recent treatment with chemotherapy or major surgery undergone within 30 days of receiving a diagnosis of COVID-19 were significantly associated with higher hospitalization rates or worsened COVID-19.

When metastatic disease, systemic chemotherapy, and major surgery were evaluated as predictors for hospitalization, the univariate ORs were 0.89 (95% CI, 0.53-1.50; P = .647), 1.04 (95% CI, 0.70-1.54; P = .845), and 1.24 (95% CI, 0.53-2.84; P = .612), respectively. In a univariate analysis of the predictors for severe respiratory illness, the univariate ORs for these measures were 0.87 (95% CI, 0.48-1.59; P = .658), 1.19 (95% CI, 0.78- 1.82; P = .407), and 1.31 (95% CI, 0.63-2.71; P = .464), respectively.

“This is probably the most important finding of our study, and the most reassuring in terms of oncologic practice,” Hohl said. “One of our huge concerns was whether patients could undergo surgery or chemotherapy during the pandemic, so it was very reassuring to us that patients who had very recently received these interventions within 30 days of developing disease had similar COVID-19 outcomes compared with patients who had not received a major surgery or active chemotherapy.”

Hohl said these data affirm that clinicians and patients should neither stop nor postpone cancer treatment. “This study makes it clear that cancer does not wait. I think the important message is patients are diagnosed with new cancers every day and patients with cancer can get sick with COVID-19, but in almost every case, a diagnosis of cancer is worse than a diagnosis of COVID-19, so it is incumbent upon us to make sure that patients get treated and we don’t lose people because of treatment delays.”

Patient Characteristics

Most of the 423 patients included in the final study population were adults over the age of 60 (56%, n = 234) who were followed for a minimum of 30 days after their initial COVID-19 diagnosis. Nearly half of the population (n = 184) had solid tumors, with breast, colorectal, and lung cancers proving to be the most common nonhematologic malignancies, accounting for 20%, 9%, and 8% of the solid cancers observed in the study, respectively. There were 102 patients with hematologic malignancies. Specifically, 8%, 11%, and 5% of patients were diagnosed with leukemia, lymphoma, and myeloma.

Among the 423 patients in the general population, 40% (168) were hospitalized for COVID-19, 11% (47) required high-flow oxygen, and 9%, mechanical ventilation. Twenty percent developed severe respiratory illness and 12% died within 30 days of their COVID-19 diagnosis. The fatality rates for the 168 patients admitted to the hospital and the 48 patients admitted to the intensive care unit were 24% and 35%, respectively.

In both univariate and multivariate analyses, patients with hematologic cancers had an increased risk of hospitalization and severe respiratory illness. In the univariate analysis, hematologic malignancies were the second highest predictor of hospitalization (OR, 2.24; 95% CI, 1.25-4.06; P = .007) and the fifth highest predictor of worsened respiratory illness (OR, 1.69; 95% CI, 0.92-3.10; P = .092). In the multivariate analysis, hematologic cancers were the second most influential predictor of both hospitalization (OR, 2.49; 95% CI, 1.35- 4.67; P = .003) and severe respiratory illness (OR, 1.79; 95% CI, 0.97-3.32; P = .063).

Clinical Implications

The investigators said the study reports the “most extensive single-institution experience in patients with cancer from the epicenter of the United States outbreak” published to date and noted that the association seen between ICI use and COVID-19 outcomes will require further study in tumor-specific cohorts. Further, a comparative analysis of COVID-19 outcomes between patients with cancer and patients without cancer should also be conducted.

Moving forward, authors said results from the MSK study should be interpreted with caution due to the limitations of the ICI data set in evaluating tumor-specific risk. They also recommended increased vigilance with SARS-CoV-2 testing in patients who are either initiating or continuing immunotherapy, as certain questions remain regarding the clinical implications of ICIs in patients with cancer and COVID-19. Beyond adjusting for comorbidities and patients with cancer on active therapy, future analyses should implement a uniform testing strategy and enable the assessment of the clinical effect of interrupting cancer care, they added.

“One of the areas we need to study further is whether there is a time dependency to immunotherapy use. We looked at patients who received immune checkpoint inhibitors within 90 days of developing COVID-19… but if immunotherapy was administered 6 months prior, it may no longer matter. Similarly, there might be a window during which patients receive chemotherapy that is critical to outcomes,” Hohl said.

Hohl added that MSK is currently collaborating with its oncology colleagues to conduct more “in-depth” studies on COVID-19 in patients with cancer.

Reference:

1. Robilotti EV, Babady NE, Mead PA, et al. Determinants of COVID-19 disease severity in patients with cancer. Nat Med. Published online June 24, 2020. doi:10.1039/s41591-020-0979-0