Patient characteristics
144 patients were included in this study. Patient and disease characteristics are detailed in Table 1. The median age of the cohort was 50 years (range, 10–70). Acute leukemia, myelodysplastic syndrome, lymphoproliferative disorders and myeloma accounted for 43%, 39% and 18% respectively. All patients had achieved a CR following the first line therapy and were on no active therapy when enrolled in the study. The only exception was those with myeloma, who had achieved at least a VGPR (n = 12) and were on oral maintenance therapy with lenalidomide or thalidomide. Based on DRI scoring, 58%, 35% and 7% of the patients had a high-risk or very-high risk (DRI-high) intermediate risk and low-risk disease (DRI-non-high) respectively. Of the 84 patients in the DRI-high cohort, 45 underwent an allogeneic hematopoietic cell transplantation (HCT), 45 from family donors (HLA-haploidentical- 41; HLA-matched-4) and 6 underwent an autologous HCT. Minimal residual disease was not considered for analysis as all patients with acute leukemia or MDS were MRD positive at the time of HCT.
Table 1
|
COVID-19 negative (N = 121)
|
COVID-19 positive (N = 23)
|
p-value
|
Age (median, range) years
|
51 (10–70)
|
46 (12–65)
|
0.3
|
Disease Type
AML/ ALL/MDS
Myeloma
Lymphoma
|
22/24/5
22
48
|
6/3/2
4
8
|
0.9
|
Disease Risk Index
Low
Intermediate
High/Very High
|
8
44
69
|
3
5
15
|
1.0
|
Gender (Male/Female)
|
73/48
|
15/8
|
0.3
|
HCT
Allogeneic (HFD/MFD)
Autologous
|
41
37
34/3
4
|
10
8
7/1
2
|
0.9
|
Relapse
|
31
|
14
|
0.0001
|
Abbreviations: AML- Acute Myeloid Leukemia; ALL-Acute Lymphoblastic Leukemia; HCT - Hematopoietic Cell Transplantation; HFD-Haploidentical Family Donor; MDS- Myelodysplastic Syndrome; MFD- Matched Family Donor
COVID-19 and Relapse
The incidence of COVID-19 was 18.9% (n = 23; 95% CI, 15.3–22.5). Moderate to Severe COVID-19 was documented in six of them (26%). The median duration of illness was 18 days (12–28). The overall incidence of relapse was 30.9% (45 patients). This was 60.9% in those with COVID-19, compared to 25.2% in those without (Fig. 1A, HR-3.6[1.9–6.8], p = 0.0001). Relapse was not influenced by severity of COVID-19. There were no deaths directly related to COVID-19.
Relapse tended to occur sooner in patients with COVID-19 (median 127 [22–280] days vs 259 [28–680] days, p = 0.001). Patients were stratified as DRI-high (includes high and very-high DRI) or DRI-non-high (includes low and intermediate DRI). The effect of COVID-19 on relapse was more striking in DRI-high group (64.3% vs 20.1%, p = 0.0001, Fig. 1B), but tended to be more in DRI-non-high group as well (55.6% vs 32.5%, p = 0.06, Fig. 1C). There was no influence of disease type on relapse with respect to COVID-19.
COVID-19 and Relapse in relation to HCT.
There was no difference in the incidence of COVID-19 between HCT and non-HCT patients (10/51 vs 13/93, p = 0.5). The incidence of relapse was 22% in the HCT cohort, compared to 36.9% in the non-HCT cohort (p = 0.06). There was no difference in relapse in the HCT group, stratified by COVID-19 (20.6% vs 21.0%, p = 1.0, Fig. 1D). However, in the non-HCT group, incidence of relapse was 27.8% (22/80) in those without COVID-19, compared to 92.3% (12/13) in the COVID-19 positive group (HR-8.9, 95% CI-4.2-18.9, p = 0.0001, Fig. 1E). There was no relation to DRI status or the disease-type on relapse incidence in the HCT group.
On multivariate analysis, COVID-19 was the only risk factor for relapse (HR-4.8, 95% CI-2.4-9.5; p = 0.0001). On the other hand, HCT was associated with a protective effect against relapse with COVID-19 in the model (HR-0.37, 95% CI-0.2-0.7; p = 0.007).
Survival
The overall survival in patients with COVID-19 at 2 years was 82.6%, compared to 94.2% in those without COVID-19 (p = 0.05, Fig. 2A). There was no difference in survival in the HCT cohort stratified by SARS-CoV2 exposure (Fig. 2B). However, in the non-HCT cohort, OS was significantly inferior in those with COVID-19 (69.2% VS 93.8%, P = 0.003, Fig. 2C).
Immune exhaustion following COVID-19 and Relapse
Longitudinal evaluation for PD1 expression on CD3 + T cells and NKG2A expression on NK cells was carried out in 16 unselected patients from COVID-19 exposed cohort at 30–45 days and 60–90 days post-infection; 8 patients each from HCT and non-HCT groups. This was also carried out during the same period in 10 COVID-19 non-exposed patients who received an allogeneic HCT, at 30 and 60 days post-HCT. Patients who relapsed within 60 days of either diagnosis of COVID-19 or the HCT procedure were not included in the analysis.
Upregulation of exhaustion receptors on T and NK cells was associated with Relapse
Relapse was documented in 8 of these 26 patients at a median of 86 days (range, 68–152). PD1 (CD279) was significantly upregulated in CD3 + T cells in those with relapse at day 30 (65.3 ± 17.5 vs 22.8 ± 17.2 p = 0.0001, Fig. 3A) and day 60 (54.8 ± 12.8 vs 31 ± 12.8, p = 0.001, Fig. 3A). A similar trend was noted in NK cells as well (Fig. 3A), with regard to expression of the inhibitory NKG2A receptor at day 30 (74.3 ± 14.5 vs 57.5 ± 15.4 p = 0.01) and day 60 (66 ± 16.5 vs 46 ± 16.9, p = 0.01).
Exploring the reason for higher relapse in non-HCT cohort in relation to Immune exhaustion
PD1
While analyzing the propensity for higher incidence of relapse in COVID-19 exposed non-HCT cohort, it was noted that PD1 expression was significantly upregulated in in CD3 + T cells in COVID-19 + non-HCT cohort at day 30 (60.6 ± 11.2% vs 39.8 ± 12.1% in COVID-19 + HCT cohort, p = 0.01). PD1 was persistently upregulated in the non-HCT cohort at day + 60, compared to the COVID-19 + HCT cohort, where it showed a reduction in expression of PD1(Fig. 3B, p = 0.002).
On the other hand, the COVID-19 negative HCT cohort showed a lower expression of exhaustion markers on both subsets of T cells at both day + 30 and + 60. The expression of PD1 was significantly lower, compared to COVID-19 exposed patients from both HCT and non-HCT cohorts at the day 30 assessment. This trend was sustained when compared with COVID-19 + non-HCT cohort at day 60 (p < 0.001), but not for COVID-19 + HCT cohort (Fig. 3B, p = 0.2).
NKG2A
The median NKG2A expression at day 30 for COVID-19 negative HCT cohort was 49.6%, compared to 64.4% (p = 0.03) and 80.5% (p < 0.0001) in COVID-19 positive HCT and non-HCT cohorts respectively (Fig. 3C). NKG2A expression reduced at day 60 in both HCT cohorts, 38% in COVID-19 negative (p = 0.02) and 46.1% in COVID-19 positive (p = 0.004) cohorts. However, high NKG2A expression was sustained at a median of 79.5% in the non-HCT cohort at 60 days. Thus, the significant difference in NKG2A expression was sustained between HCT and non-HCT groups (Fig. 3C).