Although COVID-19 has been effectively controlled, it can still cause severe pneumonia and death, especially in immunocompromised patients and elderly patients. Respiratory virus infections can increase susceptibility to secondary bacterial or fungal infections, and coinfections can have an adverse effect on prognosis(6, 7, 19, 20). Previous studies reported that the probability of COVID-19 coinfection was 8%-14.5%(5, 6, 21). In a study of all hospitals or outpatient patients with malignancies, the incidence of coinfections was 16.6%(22). In another study of patients with malignancies or who underwent organ transplantation in the intensive care unit, the incidence of coinfections was 27%, whereas it was as high as 46.7% in patients with HMs(23). However, the main microbiological detection methods used in previous studies were traditional methods, and their sensitivity remains to be evaluated. To the best of our knowledge, this is the first study to describe coinfections in HM patients with SARS-CoV-2-caused pneumonia by detecting the BALF of patients using the highly sensitive NGS method.
Our study showed that the coinfection rates of patients with HMs and those without HMs were 80.0% and 85.7%, respectively, which were significantly greater than those previously reported. The NGS method we used in this study was highly more sensitive than traditional microbiological detection methods used in previous studies, which may account for the greater rate of coinfection in our study. Pneumocystis jirovecii was the most common coinfected pathogen, with a coinfection rate of 33.3%. Pneumocystis jirovecii is a common opportunistic infection pathogen in immunocompromised patients. Pneumocystis jiroveci pneumonia may also present as diffuse pulmonary interstitial infiltration(24–26), which is sometimes difficult to distinguish from SARS-CoV-2 pneumonia. The traditional detection methods for Pneumocystis jiroveci infection have poor sensitivity, but NGS has been proven to be an effective method for detecting this disease(27–30) (9). In our previous study of lymphoma patients with chemotherapy-related IP, Pneumocystis jirovecii was detected in the BALF of 12 of 15 patients by NGS(29). In this study, all the patients with HMs had previously received chemotherapy, and 13 of 15 (86.7%) patients had received anti-CD20 mAbs, which may have resulted in severe immunodeficiency and increased susceptibility to Pneumocystis jirovecii. These data suggest that identifying Pneumocystis jirovecii coinfected with COVID-19 is necessary in HM patients after chemotherapy. In patients with a long course of SARS-CoV-2 pneumonia, NGS testing of BALF and anti-pneumocystis therapy may be considered.
Previous studies have reported that the probability of bacterial coinfection in patients with COVID-19 is approximately 8%-15%, while the incidence is relatively high in critically ill patients (approximately 20%-30%)(5, 6, 31). Our study showed that the probability of bacterial coinfection in patients with HMs was significantly lower than that in patients without HMs. This may be related to the differences in baseline characteristics between the two groups. Patients in the non-HM group had worse performance status and more comorbidities. Moreover, half of the patients in the non-HM group had severe disease. This selection bias may be due to the differences between hematologists and respiratory physicians in deciding which patients to perform bronchoscopy and NGS. For COVID-19 patients without HMs, respiratory physicians may suggest bronchoscopy for more critically ill patients. Multivariate analysis in our study also showed that bacterial coinfection was associated with severe disease. Notably, according to previous reports, the majority of hospitalized COVID-19 patients received antibiotics, despite the low incidence of bacterial coinfection(6, 32). The overuse of antibiotics can increase the risk of multidrug-resistant infections and lead to poor prognosis(33). Therefore, we should carefully evaluate the use of antibiotics in HM patients with mild COVID-19.
The incidence of viral coinfection reported in previous literature was 2.1%(22), which was significantly lower than that in our study. This may be due to the poor sensitivity of traditional virus detection methods. In our study, coinfection with herpesviruses occurred frequently in the two groups. Previous studies showed that herpesviruses, such as Epstein-Barr virus and cytomegalovirus, are common in critically ill patients, patients with hematologic disorders, and patients treated with immunosuppressive agents(34–37). Moreover, the reactivation of herpesviruses is associated with the severity and length of COVID-19 symptoms(38, 39). Gold et al. suggested that long COVID-19 symptoms may not be a direct result of the SARS-CoV-2 virus but may be the result of COVID-19-induced Epstein-Barr virus reactivation(40). Furthermore, anti-herpesvirus therapy with ganciclovir may reduce the risk of death in patients with severe COVID-19(41). Therefore, coinfection with herpesviruses may affect the prognosis of patients with COVID-19. The high detection rate of herpesviruses in our study suggested that we need to pay attention to coinfections caused by these viruses and provide effective treatment.
There are several limitations of our study. First, this was a single-center study, and the results only represent coinfections around that center. Second, because this was a retrospective study, the baseline characteristics of patients in the HM group and non-HM group were not completely compared. Patients in the non-HM group had worse performance status and more comorbidities and seemed to have more severe disease. Finally, the sample size was small, resulting in no significant differences in the comparison of some outcomes between the two groups.