A total of 306 patients were enrolled. Patients’ age ranged from 8 months to 18 years old. Nearly half of the patients (176, 57.5%) were on underlying disease therapy, including intensive chemotherapy (25.5%) and maintenance chemotherapy (32.0%). Eight patients received immunotherapy including blinatumomab, rituximab or CD19/CD22-directed CAR T therapy. In 52 patients who underwent transplantation, 23 patients were receiving immunosuppressive therapy. There was no patient actively undergoing allogeneic SCT within 1 month. The total demographic data of COVID-19 patients were showed in Table 1.

Severity was asymptomatic in 21 children (6.9%), mild in 262 (85.6%), and moderate in 17 (5.6%). 6 patients were reported as serious COVID-19 infection, severe (n = 5, 1.6%) and critical infections (n = 1, 0.3%) (Fig. 1B). There was no difference in the severity distribution of COVID-19 infection among different types of hematological and oncological diseases (Fig. 1C; Table 1). In 176 patients under chemotherapy, severe and critical severity was reported in 6 cases (3.4%) and 1 case died (0.6%). Twenty-five patients required admission to the hospital; and 1 child required mechanical ventilation and transferred to the intensive care unit. The most commonly presenting symptoms included fever (86.4%), cough (59.4%), fatigue (35.3%), and sore throat (31.0%) with vomiting/diarrhea, headache, and conjunctivitis being reported less frequently (Fig. 1A). The fever lasted from 1 day to 11 days with the highest temperature to 42℃. Other rare symptoms included rash, myalgias, and anorexia.

Symptoms and disease severity of patients with COVID-19 infection. A: symptoms of COVID-19; B: severity of COVID-19 infection in whole cohort; C: COVID-19 severity of underlying disease type; D: COVID-19 severity of underlying disease course

The data of lymphocyte subsets was available in 109 patients who recovered from COVID-19. Eighty-three (76.1%) patients were in underlying disease treatment stage, 12 patients (11.0%) have stopped therapy within one year, and 14 patients (12.8%) have stopped therapy more than one year. The possible risk factors of severity of COVID-19 infection were shown in Table 2. Univariable analysis revealed that low absolute CD3 + T cell count (P = 0.013), low CD8 + T cell count (P = 0.022), high percentage of CD4 + TCM cells (P = 0.016), and low percentage of transitional B cells (P = 0.045) were the risks for severe severity (Table 2). The absolute lymphocytes counts were variable and showed a mild relation to clinical severity (P = 0.095). Severity of COVID-19 were not associated with the absolute B cell counts and the subtype of T cells (P = 0.436, 0.893; Table 2). We also found active chemotherapy (P < 0.0001, Fig. 1D; Table 1) and long duration of symptom (P<0.0001; Table 1) were associated with the severity. Further Cox regression model analysis showed that the absolute lymphocytes count (P = 0.027) and long duration of symptom (P = 0.002) were independent prognostic factors for severity.

NK cells, Natural killer cells; TCM, central memory T cells; TEM, effector memory T cells; TEMRA, terminally differentiated effector memory T cells; SCT, stem cell transplantation; dom, dominant; dep, depletion; nor, normal.

Laboratory parameters of lymphocyte subtype were collected before COVID-19 diagnosis and at the resolution of the disease within 3 weeks after diagnosis. The absolute counts of CD3 + and CD8 + T cells were lower in recovery status compared with primary status, and CD4 + and B cells fluctuated mildly, but all with no significance (Fig. 2A-D). While NK cells were noticed significantly increasing during the infection (P = 0.026, Fig. 2I). Furthermore, CD4 + T cells proportion showed a lower frequency in recovery status than primary status (29.6% vs. 32.5%, P = 0.049, Fig. 2G). But mildly decreased CD3 + T proportion was seen in recovery status (P = 0.081, Fig. 2E). The proportion of CD8 + T and B cells fluctuated mildly during the infection (P = 0.329, 0.377; Fig. 2F, H). A lower CD4+/CD8 + ratio was seen in recovery patients than primary status (0.70 vs. 0.77, P = 0.021, Fig. 2L). NK cells proportion was higher in recovery status than primary status (11.6% vs. 6.0%, P = 0.002, Fig. 2J). The total lymphocyte was similar during the infection (P = 0.768, Fig. 2K).

Lymphocyte subtype in patients with hematological and oncological disease post COVID-19 infection compared with basic status. A-D: The absolute levels of CD3+, CD4+, CD8 + T cells, and CD19 + B cells; E-H: The frequency of CD3+, CD4+, CD8 + T cells, and CD19 + B cells; I-J: absolute and frequency of NK cells; K: absolute lymphocyte counts; L: CD4+/CD8 + ratio. *P < 0.05, * *P < 0.01 (Pair t test or Wilcoxon matched-pairs signed rank test)

According to the subtle subtype of T and B cells at resolution of COVID-19 infection, we further divided patients by CD8 + dominant and B cell depletion (CD8domBdep, 35.8%), CD8 + dominant and B cell positive (CD8domB+, 31.2%), B cell positive and a CD4-dominant (CD4domB+, 4.6%), normal CD4+/CD8 + T cell ratio and B cell depletion (CD4/CD8norBdep, 17.4%), or balanced CD4+/CD8 + and B cells positive (CD4/CD8norB+, 11.0%). The data of subtle subtypes were available in 90 patients. Comparison with groups, patients with CD8domBdep had very lower frequency of naive T cells but higher frequency of CD8+ TEM and TEMRA cells than CD4/CD8norBdep (Fig. 3A, C,D). TCM CD8 + cells expressed higher in CD8domBdep group than CD4/CD8norB+ (Fig. 3B). CD8domBdep had lower frequency of CD4+ naive T cells but higher frequency of CD4+ TEM and TEMRA cells comparing with CD4/CD8norBdep/+ (Fig. 3E, G, H). CD8domBdep had lower frequency of naïve B cells than CD8domB+, CD4/CD8norB+ and CD4domB+; but had higher frequency of memory B cells than CD8domB+ and CD4domB+. Transitional B cells expressed lower in CD8domBdep than in CD4domB+. CD8domBdep had very higher proportions of plasmablast than CD4/CD8norB+ and CD8domB+ (Fig. 3I-L). Other expression of T and B cells phenotypes were showed in Fig. 3.

T cell exhaustion and B cell response in COVID-19 patients with defined phenotype. A-D: CD8 + T cell differentiation of Naïve T, TCM, TEM, and TEMRA in 5 groups with CD8domBdep, CD8domB+, CD4domB+, CD4/CD8norBdep, and CD4/CD8norB+; E-H: CD4 + T cell differentiation of Naïve T, TCM, TEM, and TEMRA in 5 groups; I-L: B cell response of Naïve B, memory B, transitional B and plasmablast in 5 phenotypic groups. *P < 0.05, * *P < 0.01, ***P < 0.001 (Kruskal-Wallis test or Brown-Forsythe ANOVA test)

Most patients (85.6%) were in quarantine at home with antipyretic or antitussive therapy. 6 (2.0%) patients need supplemental oxygen requirement. Twenty-one (6.9%) patients received anti-virus therapy; intravenous gamma globulin (84.0%) and oseltamir (40.0%) were used frequently in inpatients. Paxlovid was used in 1 severe patient. Remdesivir or convalescent plasma were not used in this cohort. Reported complications from COVID-19 included bacterial or fungus superinfection in 7.5%, respiratory failure in 2.0%, and encephalitis in 0.7% of all cases. Klebsiella pneumoniae, stenotrophomonas maltophilia, or enterococcus faecalis has been detected from phlegm, bronchoalveolar lavage fluid or blood. Two cases were coinfected with mycoplasma. Patients combined with co-infection had a risk for severe disease (65.2% vs. 2.8%, P < 0.0001). Those patients usually under intensive chemotherapy with myelosuppression. Overall, one death was reported in the cohort with relapsed acute lymphoblastic leukemia in reinduction phase. The patient developed Klebsiella pneumoniae sepsis during the period of COVID-19 infection. Of the 158 patients who under anticancer therapy, there were interruptions in 86.7% of patients, with a median of 10 days delay ranging from 5 to 21 days. Reasons for delays included high fever, underlying myelosuppression, at myelosuppressive phase, or other complication.

Median time to symptom disappearance of COVID-19 was 4 days (range 1–18 days). Median time for COVID-19 clearance was 7 days (range 1–35). Patients with CD8domBdep subtype had a longer COVID-19 duration of 9.1±4.0 days, compared with 6.8±2.8 days in CD8domB+ (P = 0.045), 6.0±1.5 days in CD4/CD8norBdep (P = 0.001), and 4.8±2.3 days in CD4/CD8norB+ (P = 0.001). Only 5 patients were included in CD4domB+ subset and had a COVID-19 duration time of 13±12.9 days. Patients with CD4/CD8norB+ had the quickest clearance of COVID-19, but there was no deference with patients with CD8domB+, CD4/CD8norBdep, or CD4domB+ subset. The clearance day (≤ 7 days) was not related with B cell level (P = 0.876), CD8 + level (P = 0.803), CD3 + value (P = 0.873), and severity (P = 0.367).