Phenotype of BTK‐lacking myeloid cells during prolonged COVID‐19 and upon convalescent plasma

Abstract

XLA patient with 7‐month course of COVID‐19 with persistent plasma SARS‐CoV‐2 load revealed a sustained non‐inflammatory profile of myeloid cells in association with contained severity of disease, arguing in favor of the use of BTK inhibitors in SARS‐COV‐2 infection.

Patients with X‐linked agammaglobulinemia (XLA) may feature a severe course of SARS‐CoV‐2 infection that has been mainly attributed to the lack of B cells, supporting the use of plasma from convalescent individuals to replace the missing humoral response. ¹

XLA is caused by mutations in the Bruton tyrosine kinase (BTK), which in addition to its essential role in B‐cell development, impacts in many pathways in monocytes, namely TLR signaling, cytokine production, and modulation of M1‐like pro‐inflammatory profile and M2‐like immuno‐regulatory phenotype. ²

Monocyte/macrophages are key determinants of the evolution of SARS‐CoV‐2 infection, contributing to the lung disruption associated with pneumonia, as well as to the systemic pro‐inflammatory state and the cytokine storm associated with worst prognosis. ³ On the other hand, we found an M2‐like shift of circulating monocytes during the recovery of severe COVID‐19 that may be linked to tissue repair. ³

Notably, BTK has been emerging as a possible therapeutic target, based on the positive association found between BTK activity and severity of SARS‐CoV‐2 infection, and the observed decrease in inflammatory markers and improved clinical outcomes in COVID‐19 patients under treatment with BTK inhibitors in the context of other concomitant diseases. ⁴

These findings prompted us to investigate the myeloid phenotype in a 35‐year‐old XLA patient that featured an extremely prolonged disease course with persistent SARS‐CoV‐2 viral load for more than 7 months (Table 1), documented both in nasopharyngeal swabs, assessed by RT‐PCR, and in the plasma, quantified by ddPCR. ³ Data were compared with 20 COVID‐19 patients evaluated at hospital admission and 13 healthy subjects recruited in parallel, that have been included in recently published studies, ³ , ⁵ and whose clinical and epidemiological data are summarized in Table 1. As expected, the hospitalized Covid‐19 patients were older than the XLA patient, but were used as illustrative of the immunological alterations associated with severe COVID‐19. All samples were processed immediately after collection, and the staining performed in whole blood allowed the analysis of a large number of cells by flow cytometry using both unsupervised and manual approaches, as previously described. ³ , ⁵

TABLE 1.

Clinical and epidemiological data from patients and healthy controls

	Patient with X‐linked agammaglobulinemia (study timepoints)				COVID‐19 Cohort a	Healthy controls
	1	2	3	4
Number (male)	1 (1)				20 (17)	11 (9)
Age (years)	35				55.5 (39–65)	58 (39–65)
Time from symptom start (days)	75	145	212	217 b	8.5 (5–11)	NA
CRP (mg/dl)	0.7	8.9	6.1	1.6	8.8 (4.85–25.5)	ND
PCT (ng/ml)	0.03	0.03	0.05	<0.02	0.16 (0.11–0.38)	ND
Ferritin (ng/ml)	1827	1600	1430	1539	939 (402–1906)	ND
Interleukin‐6 (pg/ml)	6.3	ND	33.5	1.8	18 (4.5–36)	0.85 (0.24–1.6)
Total serum IgG (mg/dl)	744	655	735	1518
Lymphocytes/μl	2556	1113	1697	2165	920 (845–1662)	1940 (1423–2200)
CD4+ T cells/μl	1140	402	304	854	247 (133–392)	768 (544–998)
CD8+ T cells/μl	1007	306	853	894	145 (81.2–262)	414 (158–577)
CD4/CD8 ratio	1.13	1.11	0.36	0.95	1.78 (0.93–2.50)	1.95 (1.53–4.17)
Neutrophils/μl	2916	7186	14 515	3339	4251 (2413–6917)	3228 (2521–6390)
Lymphocytes/neutrophils ratio	0.88	0.15	0.12	0.65	0.23 (0.15–0.50)	0.51 (0.47–0.61)
Monocytes/μl	480	356	487	348	349 (223–537)	398 (275–733)
Basophils/μl	12	10	17	12	20 (10–35)	32 (16–63)
Eosinophils/μl	36	17	50	35	13 (7–56)	115 (96–297)
SARS‐CoV‐2 plasma viral load (RNA cps/ml)	23	201	116	ULoD	112 (24–498) c	NA

Note: Values expressed as medians (interquartile range) unless otherwise specified.

Abbreviations: CRP, C reactive protein; NA, not applicable; ND, not done; PCT, procalcitonin; ULoD, under limit of detection.

^a COVID‐19 associated co‐morbidities: arterial hypertension 9 (45%); Diabetes type II 6 (30%); Obesity 6 (30%); Lung emphysema 2 (10%).

^b Convalescent plasma was administered at Days 213 and 214 after starting of symptoms.

^c Quantified in the 13 (65%) COVID‐19 patients with detectable plasma viremia.

The XLA patient features a c.1559G>A mutation in the exon 15 of the BTK gene. Intravenous IgG (IVIG) replacement therapy was started at the age of 4 after septic hip and recurrent respiratory infections, maintaining IgG serum level above 800 mg/dl, without major infections besides occasional sinusitis. At the age of 31, he was diagnosed with type 1 diabetes, with difficult metabolic control due to poor compliance with diet (overweigh since adolescence, BMI 29 kg/m²) and treatment (glycosylated hemoglobin A1c around 11%).

SARS‐CoV‐2 infection manifestations were fever with cough and dyspnea. Chest CT scan performed on the 8th day of disease revealed multifocal lung ground glass opacities that slowly progressed over more than 7 months leading to involvement of up to 70% of both lungs, with a radiological pattern of organizing pneumonia (Figure 1A), despite steroid therapy started after day 35 of disease (mean 0.4 mg/kg). The imaging alterations could not be ascribed to other infections, given the lack of microbial isolation and the absence of response to broad‐spectrum antibiotics. Peripheral O₂ saturation was never below 85% and therefore there was only need for low‐flow nasal cannula oxygen and no need for any form of non‐invasive/invasive ventilation or ICU admission.

FIGURE 1.

Longitudinal lung imaging and monocyte profile in SARS‐CoV‐2‐infected XLA patient treated with convalescent plasma. (A) Chest CT. (B) Unsupervised analysis, UMAP of CD45⁺Lineage^Neg cells, marker expression, and manual annotation; relative subset distribution in patient timepoints and representative COVID‐19 and healthy individuals. (C) Monocyte manual analysis showing CD16/CD14 within CD45⁺Lineage^Neg cells (top); HLA‐DR/PD‐L1 in classical (middle); and Slan/CD163 in non‐classical (bottom) monocytes.

It is therefore likely that the BTK defect contributed to this smoldering relatively protected course despite the presence of co‐morbidities associated with an adverse prognosis. Although the loss of function of BTK may impact the biology of many cell populations, monocytes are probable main culprits given their role in COVID‐19. ² , ³ We evaluated longitudinally the myeloid compartment by flow cytometry, as illustrated in Figure S1A, and found lower expression of several activation markers like HLA‐DR, PD‐L1, CD86, and CD80 as compared with patients with moderate to severe COVID‐19 and healthy individuals, in both the unsupervised approach to total monocytes (Figure 1B), and the manual analysis of monocyte subsets (Figure 1C). Additionally, there was marked reduction in Slan+ non‐classical monocytes (Figure 1C), a sub‐population involved in inflammatory conditions but shown to be reduced in COVID‐19. ³ , ⁵ The levels of HLA‐DR were also reduced in plasmacytoid dendritic cells (DCs), as well as in CD141+ and CD1c+ DCs (Figure S1B). These findings contrasted with the progressive increase observed in serum IL‐6 levels and CRP (Table 1), as well as the increase in neutrophils and decline in lymphocyte counts, which have been associated with adverse prognosis of COVID‐19. ³ , ⁵ Regarding T cells, the flow cytometry analysis revealed the expected expansion of memory‐effector CD4 and CD8 T‐cell subsets (Figure S2). Therefore, our data support a contribution of the reduced myeloid activation to the slow disease progression throughout the 7‐month course of the SARS‐CoV‐2 infection.

Remdesivir (Day 64, 5‐day course) had no impact. A clear reduction in inflammatory markers was only observed upon treatment with convalescent plasma (Days 213 and 214, Table 1). The plasma viral load was undetectable after 3 days (Table 1), although SARS‐Cov‐2 was found in a control bronchoalveolar lavage performed at Day 221 (643 382 RNA cps/ml by ddPCR) and the nasopharyngeal swabs remained positive for 60 days more. The impact of convalescent plasma on the myeloid profile was more evident on the recovery of intermediate and non‐classical monocyte populations and DCs (Figure 1). There was a progressive clinical/laboratorial improvement, allowing steroid tapering (stopped on Day 292) and gradual resolution of chest CT lung opacities (Figure 1A). Functional recovery also supports a potential contribution of the BTK defect for the prevention of lung fibrosis.

There are limited data on the phenotype of myeloid cells in patients with germline loss‐of‐function mutations in BTK. ² Our detailed immunological study provides evidence of a non‐inflammatory profile of myeloid cells in XLA that was sustained upon persistent SARS‐CoV‐2 infection, providing a possible explanation for the protracted course of our case and others previously reported. ¹ Patients with inborn errors of immunity offer unique opportunities to better understand the host‐pathogen interactions, despite the limitations imposed by their rarity. The ability to limit the disease severity in this patient with acknowledged risk factors, adds evidence in favor of early use of BTK inhibitors to treat COVID‐19, as a strategy to ameliorate the hyper‐inflammatory response, improve survival and limit inflammatory complications.

AUTHOR CONTRIBUTIONS

Amelia C. Trombetta, Ana E. Sousa, and Susana L. Silva designed and supervised the study. Inês Parreira, Hélder Diogo Gonçalves, Mariana Lessa Simões, Patrício Aguiar, Maria Manuel Deveza, João Inácio, and Susana L. Silva collected clinical data. André M. C. Gomes, Guilherme B. Farias, Amelia C. Trombetta, and Ana Godinho‐Santos performed the experiments. André M. C. Gomes, Guilherme B. Farias, Ana E. Sousa, and Susana L. Silva wrote the paper.

FUNDING INFORMATION

This work was funded by the following grants from Fundação para a Ciência e a Tecnologia (FCT), Portugal, through “Apoio Especial Research4COVID‐19,” project numbers 125 and 803. André M. C. Gomes and Guilherme B. Farias received Fellowships funded by FCT (Doctorates4COVID‐19, 2020.10202.BD), and Janssen‐Cilag Farmacêutica, respectively.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

Figure S1. Myeloid analysis. (A) Flow cytometry gating strategy. Dot plots showing data from a representative healthy control illustrating the gating strategy used to identify monocyte and dendritic cells from whole blood, starting from 10 million leukocytes. (B) Circulating dendritic cells in a SARS‐CoV‐2‐infected patient with X‐linked agammaglobulinemia treated with convalescent plasma. Contour‐plots showing the analysis within a CD45⁺lineage⁻ gate of HLA‐DR staining against: CD123 to identify the plasmacytoid dendritic cells (top), CD1c to identify the CD1c⁺ myeloid dendritic cells (middle), and CD141 to identify the CD141⁺ myeloid dendritic cells (bottom), in the four timepoint evaluations of the patient, as well as in representative individuals from the COVID‐19 and healthy control groups. Viremic and non‐viremic status was added to represent the timepoints when the patient had detectable or non‐detectable SARS‐CoV‐2 plasma viral load, respectively.

Figure S2. Circulating T cells in a SARS‐CoV‐2‐infected patient with X‐linked agammaglobulinemia treated with convalescent plasma. (A) Unsupervised analysis of T cells using UMAP of the CD3⁺ cells with manual annotation of the main cell subsets namely: naïve (CCR7⁺CD45RO⁻); central memory (CCR7⁺CD45RO⁺); effector memory (CCR7‐CD45RO⁺); terminally differentiated (CCR7⁻CD45RO⁻); follicular helper T cells (Tfh; CCR7⁺CD45RO⁺CXCR5⁺); regulatory T cells (Treg; CD25⁺CD127^low). (B) Relative event distribution in the four evaluations of the patient, as well as in representative individuals from the COVID‐19 and healthy control groups. Viremic and non‐viremic status was added to represent the timepoints when the patient had detectable or non‐detectable SARS‐CoV‐2 plasma viral load, respectively.

Gomes AMC, Farias GB, Trombetta AC, et al. Phenotype of BTK‐lacking myeloid cells during prolonged COVID‐19 and upon convalescent plasma. Eur J Haematol. 2023;110(2):209‐212. doi: 10.1111/ejh.13881

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request to the corresponding author.