Important Concerns Over SARS-CoV-2 Infection in Children With Cancer—Reply

In Reply We thank Zhang et al for their interest in our Research Letter regarding coronavirus disease 2019 (COVID-19) in children with cancer in New York City¹ and for their comments. In terms of the first concern, we agree that different types of pediatric cancer are pathophysiologically different, are treated differently, and may have different degrees of immune suppression. Since publication of the article,¹ the MSK Kids pediatric program at Memorial Sloan Kettering Cancer Center has now cared for 34 patients who tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), of 465 unique patients tested. Their diagnoses were as follows: sarcoma (n = 14), leukemia/lymphoma (n = 12), neuroblastoma (n = 3), Wilms tumor (n = 1), astrocytoma (n = 1), and nonmalignant hematologic and immune disorders (n = 3). Four patients were posttransplant. To the specific inquiry from Zhang et al about COVID-19 mortality, there were no deaths due to COVID-19 in the 34 patients. Unfortunately, with such small numerators, further subgroup analyses are not statistically rigorous.

With regard to the second concern regarding sex differences, we agree that this comparison was not preplanned; however, our analysis adequately controlled for type I error. Small sample size most negatively affects the power of a comparison, whereas statistical significance tends to increase with an increasing sample size.² To this point, reassessing the updated patient numbers referenced above (nearly double the initial published numbers), 76% of the SAR-CoV-2–positive patients were male (26 of 34), compared with 52% of the patients who tested negative (223 of 431; Fisher exact test P = .007), reinforcing the observed sex difference in a larger sample and with even higher statistical significance (originally P = .02).¹ We also note that the articles referenced by Zhang et al focused on adult patients with cancer³ or did not provide primary data on pediatric patients.⁴ We note that studies specifically focusing on children with SAR-CoV-2 infection found a preponderance of males with 64% (23 of 36) in one case⁵ and 61% (104 of 171) in the other.⁶ We concur that data in pediatric patients with cancer do not represent children in general, but we do believe that data in pediatric patients with cancer are biologically more representative of children than extrapolation from adult data and agree that these results should be validated in larger population-level studies.

Corresponding Author: Andrew L. Kung, MD, PhD, Department of Pediatrics, Memorial Sloan Kettering Cancer Center, 1275 York Ave, New York, NY 10065 (kunga@mskcc.org).

Published Online: October 15, 2020. doi:10.1001/jamaoncol.2020.4767

Conflict of Interest Disclosures: None reported.