Outcomes and inflammation changes in different types of immunocompromised patients with critically ill COVID-19 admitted to ICU: a national multicenter study

Abstract

Background

Immunocompromised patients face higher risks of Severe Acute Respiratory Syndrome Coronavirus 2 infection and co-infections, leading to a possibility of high disease severity and poor outcomes. Conversely, immunosuppression can mitigate the excessive inflammatory response induced by the virus, potentially reducing disease severity. This study aims to investigate the prognostic differences and early inflammatory response characteristics in various types of immunocompromised patients with severe coronavirus disease 2019 (COVID-19) admitted to intensive care unit (ICU), summarize their clinical features, and explore potential mechanisms.

Methods

A retrospective analysis was conducted on critically ill COVID-19 patients admitted to the ICU of 59 medical centers in mainland China during the Omicron outbreak from November 2022 to February 2023. Patients were categorized into two groups based on their immunosuppression status: immunocompromised and immunocompetent. Immunocompromised patients were further subdivided by etiology into cancer patients, solid organ transplant (SOT) patients, and other immunocompromised groups, with immunocompetent patients serving as controls. The mortality rates, respiratory support, complications, and early inflammatory cytokine dynamics upon ICU admission among different populations were analyzed.

Results

A total of 2030 critically ill COVID-19 patients admitted to ICU were included, with 242 in the immunocompromised group and 1788 in the immunocompetent group. Cancer patients had a higher median age of 69 years (IQR 59, 77), while SOT patients were generally younger and had less severe illness upon ICU admission, with a median APACHE II score of 12.0 (IQR 8.0, 20.0). Cancer patients had a twofold increased risk of death (OR = 2.02, 95% CI 1.18–3.46, P = 0.010) compared to immunocompetent patients. SOT and cancer patients exhibited higher C-reactive protein and serum ferritin levels than the immunocompetent group in their early days of ICU admission. The CD8⁺ T cells dynamics were inversely correlated in cancer and SOT patients, with Interleukin-6 levels consistently lower in the SOT group compared to both immunocompetent and cancer patients.

Conclusion

Critically ill COVID-19 patients admitted to the ICU exhibit distinct clinical outcomes based on their immunosuppression status, with cancer patients facing the highest mortality rate due to variations in inflammatory responses linked to their immunosuppression mechanisms. Monitoring dynamic changes in inflammatory markers and immune cells, particularly CD8⁺ T lymphocytes and IL-6, may offer valuable prognostic insights for these patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12890-024-03362-6.

Background

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has triggered a global pandemic of Coronavirus Disease 2019 (COVID-19), posing a significant threat to global healthcare systems and human health [1]. As of May 26, 2024, the global confirmed cases of COVID-19 exceeded 776 million, with over 7.1 million deaths [2]. Compared to immunocompetent patients, immunocompromised individuals, due to underlying conditions and treatment-induced immunosuppression, face a heightened risk of SARS-CoV-2 infection, potentially accompanied by increased incidence of bacterial and fungal co-infections [3, 4].

In the context of COVID-19, there is a correlation between excessive host inflammatory response and disease severity, known as the “cytokine storm” hypothesis [5–7]. This theory posits that overactivation of the immune system may lead to excessive inflammatory responses, exacerbating clinical manifestations of the disease and posing a serious threat of death. However, studies have shown that immunosuppression may partially inhibit cytokine release, thereby alleviating disease severity in certain cases [8, 9]. This raises questions about whether immunocompromised patients exhibit atypical clinical features and whether they are more prone to avoiding cytokine-mediated inflammatory responses compared to immunocompetent patients, potentially reducing the incidence of severe disease.

Previous research indicates that immunocompromised patients, particularly those with cancer and solid organ transplants (SOT), face a higher risk of COVID-19 [10, 11]. These patients are classified as immunocompromised due to primary or secondary immunological abnormalities [12]. Although some studies have explored the relationship between immunocompromised patients and COVID-19, the correlation between different types of immunosuppression and prognosis in immunocompromised patients with critically ill COVID-19 requiring an intensive care units (ICU) admission remains understudied [13–16]. We conducted a nationwide multicenter study aimed at assessing the mortality rate, respiratory support, and complications in patients with different types of immunosuppression compared to immunocompetent individuals following critically ill COVID-19 infection admitted to ICU, as well as the differences in inflammatory cytokine dynamics during the initial ICU period. This in-depth analysis is expected to provide a more comprehensive understanding and aid in optimizing management and treatment strategies for this vulnerable critically ill population.

Methods

Study design and patient enrollment

This multicenter retrospective observational cohort study was conducted across 28 provinces in Chinese mainland, involving 59 ICU, from November 1, 2022, to February 11, 2023. Inclusion criteria were age > 18 years old, diagnosed with SARS-CoV-2 pneumonia via polymerase chain reaction (PCR) or SARS-CoV-2 antigen testing on nasopharyngeal swabs or lower respiratory tract specimens (the sputum, transtracheal aspirates, bronchoalveolar lavage fluid), and classified as severe or critical, requiring ICU care. Severe COVID-19 was defined by any of the following criteria: (1) shortness of breath with a respiratory rate exceeding 30 breaths per minute; (2) pulse oxygen saturation of 93% or lower at rest; or (3) a partial pressure of oxygen to fraction of inspired oxygen ratio (PaO₂/FiO₂) of 300 mmHg or lower. Critical COVID-19 was characterized by any of the following: (1) respiratory failure requiring invasive ventilation; (2) shock; or (3) organ failure that necessitated ICU care [17]. The exclusion criteria included individuals with insufficient basic information (such as gender and age), inadequate prognostic information, unclear respiratory support modalities, or those who were pregnant. ICU mortality was the primary endpoint. The study protocol was approved by the Research Ethics Committee of the China-Japan Friendship Hospital (2019-79-K51-1). Informed consent from patients or their legal guardians was waived due to the retrospective nature of the study.

Data collection

Following patient enrollment, electronic medical records for each patient were extracted using standardized data collection forms and analyzed by two independent researchers. Demographic and baseline characteristics (including age, sex, comorbidities, and smoking history) of immunocompromised and immunocompetent patients, as well as laboratory results at admission (including blood cell counts, biochemical parameters and coagulation indices), and etiological examinations were obtained and analyzed. Clinical outcomes for each patient, such as ICU mortality or secondary infection were also recorded.

Definitions

To facilitate a clear understanding of the patient groups, patients were categorized into two groups based on their immune status: immunocompromised and immunocompetent [18]. Immunocompromised patients were further categorized into groups based on the cause of immunosuppression: cancer patients, SOT recipients, and other immunocompromised patients (the composition of other immunocompromised patients is illustrated in Supplementary Fig. 1). Secondary infections were defined as bacterial, viral, or fungal infections occurring in the blood, respiratory tract, urinary system, or other sterile sites more than 48 h after ICU admission. ICU mortality was defined as the death of patients within the ICU, including those expected to die following the discontinuation of treatment. Acute physiology and chronic health evaluation II (APACHE II) [19], and sequential organ failure assessment score (SOFA) scores [20] were the worst conditions or values for the first 24 h of ICU admission. Cancer patients were defined as individuals with active malignancy or those diagnosed with malignancy within one year of developing community-acquired pneumonia (CAP), excluding localized skin cancers and early-stage cancers. CAP [21] refers to inflammation of the lung parenchyma (including the alveolar walls, i.e., the broader lung interstitium) that occurs outside the hospital setting, including pneumonia with documented pathogen infection that develops during the incubation period after hospital admission.

Statistical analysis

Baseline information and logistic multivariate regression analyses were performed using SPSS software version 27.0 (IBM). Bar charts and Kaplan-Meier survival curves were created using R software package (version 4.3.2). Categorical variables were described using frequencies and percentages, with group comparisons conducted via chi-square tests. Continuous variables following a normal distribution were described by means and standard deviations, with group comparisons performed using independent sample t-tests. For variables not following a normal distribution, medians and quartiles were reported, and group comparisons were conducted using nonparametric rank-sum tests. Potential confounders were input into multivariate logistic regression analysis to determine the risks of different outcomes in the subgroups. To compare outcomes between two groups of patients who had similar baseline characteristics except for different immune status (immunocompetent, cancer patients, and SOT), we performed a three-to-one propensity score (PS) matching analysis estimated by logistic regression to account for potential confounding factors. Variables involved in the PS estimation included age and APACHE II score. Matching was based on the logit of the PS using nearest-neighbor matching (greedy-type matching) with a caliper width of 0.25. A p-value < 0.05 was considered statistically significant.

Results

A total of 2030 critically ill COVID-19 patients were enrolled from November 1, 2022, to February 11, 2023, across 28 provinces in mainland China. These patients were admitted to 59 ICUs. Among them, 1788 patients were immunocompetent, while 242 patients were immunocompromised. Of the immunocompromised patients, 42 had undergone SOT, 77 had cancer, and 123 belonged to other immunocompromised group.

Baseline characteristics

Table 1 summarizes the demographic and clinical characteristics of critically ill COVID-19 patients with different immune statuses upon ICU admission. (1) Age: immunocompetent patients had a median age of 76 years (IQR, 68–84), while SOT recipients were significantly younger, with a median age of 54 years (IQR, 48–60). (2) Duration from Symptom Onset: the immunocompromised group, particularly SOT recipients, experienced significantly prolonged durations from symptom onset to dyspnea (20 days [14, 24]) and from symptom onset to ICU admission (15 days [11, 19]) compared to the immunocompetent group. (3) APACHE II Scores: within 24 h of ICU admission, SOT patients exhibited significantly lower APACHE II scores (12.0 [8.0, 20.0]) compared to other groups.(4) Chronic Conditions: notably, nearly half of the SOT group had chronic renal disease (47.6%). (5) Laboratory Parameters: upon ICU admission, cancer patients exhibited higher levels of IL-6 and serum ferritin. In contrast, SOT recipients showed significantly lower lymphocyte counts and IL-6 levels but had higher levels of creatinine, C-reactive protein (CRP), and serum ferritin. Other immunocompromised patients had the lowest serum ferritin levels. An additional file includes a detailed comparison of significance between each pair of groups (Supplemental Table 1) and provides the baseline characteristics of the other immunocompromised group (Supplemental Table 2) for a more comprehensive understanding .

Table 1.

Baseline demographic characteristics, medical history and laboratory findings of critically ill COVID-19 patients

	Immunocompetent	Immunocompromised	SOT	Cancer	Other	p value
n	1788	242	42	77	123
Demographics
Age (years), median [IQR]	76[68, 84]	67[56,76]	54[48,60]	69[59,77]	69[59,77]	< 0.001
Male sex (%)	1296 (72.5)	167 (69.0)	34(81.0)	57(74.0)	76(61.8)	0.292
BMI (kg/m2)	23.4[20.8, 25.8]	23.7[21.1,26.2]	24.5[21.7,27.2]	24.2[20.6,25.4]	23.4[20.9,26.3]	0.437
Smoking history (%)	473 (26.5)	71 (29.3)	13(31.0)	19(24.7)	39(31.7)	0.342
Time from onset to dyspnea (days), median [IQR]	10[5, 16]	14[9, 23]	20[14,24]	11[7,22]	14 [9,21]	< 0.001
Time from onset to ICU admission (days), median [IQR]	9[5, 14]	12[6, 18]	15[11, 19]	8[4,15]	12[7,18]	< 0.001
PaO2/FiO2 (mmHg), median [IQR]	103.0[56.0, 188.0]	114.0[58.8, 207.0]	102.0[62.3,203.0]	109.0[59.5,188.0]	120.0[51.7,213.0]	0.564
APACHE II score, median [IQR]	16.0[11.0, 23.0]	14.0[11.0, 21.0]	12.0[8.0,20.0]	17.0[13.0,21.0]	14.0[10.0,19.3]	0.045
SOFA score, median [IQR]	5.0[3.0, 8.0]	5.0[3.0, 8.0]	5.0[3.0,7.0]	6.0[3.0,8.0]	4.0[3.0,7.0]	0.312
Medical History
Chronic heart disease (%)	501 (28.0)	42(17.4)	3(7.1)	17(22.1)	22(17.9)	< 0.001
Chronic liver disease (%)	17 (1.0)	4 (1.7)	0(0.0)	2(2.6)	2(1.6)	0.500
Chronic lung disease (%)	221 (12.4)	38 (15.7)	1(2.4)	10(13.0)	27(22.0)	0.174
Chronic renal disease (%)	113 (6.3)	44 (18.2)	20(47.6)	5(6.5)	19(15.4)	< 0.001
Diabetes (%)	535 (29.9)	64 (26.4)	17(40.5)	17(22.1)	30(24.4)	0.300
Hypertension (%)	1000 (55.9)	118 (48.8)	24(57.1)	31(40.3)	63(51.2)	0.042
Laboratory Parameters
White blood cell count (× 109 /L), median [IQR]	9.5 [6.4, 13.8]	8.2 [5.6,12.2]	8.4[5.8,13.2]	8.2[6.5,12.3]	7.9[5.3,11.7]	0.004
Neutrophil count (× 109 /L), median [IQR]	8.3[5.3, 12.5]	7.2[4.7,10.7]	8.0 [5.2,10.9]	7.0[5.2,11.4]	6.8[4.2,10.4]	0.007
Lymphocyte count (× 109 /L), median [IQR]	0.6 [0.3, 0.9]	0.5 [0.2,0.7]	0.3[0.2,0.5]	0.5[0.4,0.9]	0.5 [0.3,0.7]	< 0.001
Hemoglobin (g/L), median [IQR]	122.0[105.0, 138.0]	111.0[93.5, 130.0]	112.0[91.0,128.0]	113.0[94.8,127.0]	110.0[93.0,131.0]	< 0.001
Platelet count (× 109 /L), median [IQR]	173.0[120.0, 245.0]	164.0[100.0, 208.0]	140.0[105.0,202.0]	170.0[113.0,201.0]	155.0[93.8,214.0]	0.001
D-dimer (μg/mL), median [IQR]	2.2[1.0, 6.2]	2.1 [0.9, 6.0]	2.2 [1.1,4.4]	2.1 [1.0,5.8]	2.1 [0.8,6.7]	0.647
Aspartate aminotransferase (IU/L), median [IQR]	38.7[25.6, 64.0]	32.4[22.5, 52.3]	21.5[17.0,33.0]	43[30.0,73.0]	32[23.9,51.0]	0.002
Creatinine (μmol/L), median [IQR]	87.6[63.2, 152.0]	88.8[58.3, 144.0]	164.0[120.0,219.0]	68.8[54.8,97.1]	73.0 [54.0,113.0]	0.146
Albumin (g/L), median [IQR]	31.0[27.9, 34.2]	31.4[28.5, 35.0]	29.9[28.6,35.0]	33.6[29.8,35.8]	30.6[27.6,34.4]	0.353
Lactate dehydrogenase (U/L), median [IQR]	396.0[287.0, 548.0]	419.0[302.0, 601.0]	419.0 [357.0,521.0]	481.0 [261.0,650.0]	389.0 [304.0,584.0]	0.256
C-reactive protein (mg/L), median [IQR]	14.9[6.8, 55.4]	22.8[7.7,89.8]	46.9[14.2,110.0]	22.3[7.4,140.0]	20.8[7.2,65.1]	0.034
Interleukin-6 (pg/mL), median [IQR]	69.3[19.6, 272.0]	50.8[12.9, 136.0]	25.5[11.8,83.8]	92.8[39.2,204.0]	42.8[9.6,134.0]	0.009
Serum ferritin (μg/L), median [IQR]	802.0[394.0, 1458.0]	748.0[445.0, 1016.0]	890.0[801.0,1754.0]	890.0[589.0,1016.0]	445.0[282.0,654.0]	0.912

COVID-19: coronavirus disease 2019; SOT: Solid organ transplantation; Other: persons with other causes of immunosuppression; BMI: body mass index; ICU: intensive care unit; PaO₂/FiO₂: partial pressure of oxygen/fraction of inspiration oxygen; APACHE: acute physiology and chronic health evaluation; SOFA: sequential organ failure assessment; The p value was the difference between the non-immunosuppression and immunosuppression group

Organ support and clinical outcomes

Table 2 presents a comparative analysis of organ support and clinical outcomes among subgroups of patients. For SOT patients, the treatment methods utilized included corticosteroids (37, 88.1%), Tocilizumab (5, 11.9%), antiviral drugs (17, 40.5%), and intravenous immunoglobulin (12, 28.6%). And 28.6% of SOT patients received high-flow nasal cannula oxygen therapy. For cancer patients, 24.7% (19 patients) received high-flow nasal cannula oxygen therapy, and they had a higher frequency of invasive mechanical ventilation, with 40.3% (31 patients) undergoing this procedure. Additionally, secondary fungal infections were more frequently detected in SOT patients (5, 11.9%). Treatments, respiratory support, and outcomes for other immunocompromised patient groups can be found in Supplemental Table 3.

Table 2.

Treatments, respiratory support and outcomes of critically ill COVID-19 patients

	Immunocompetent	Immunocompromised	SOT	Cancer	Other	p value
n	1788	242	42	77	123
Anti-inflammatory treatment
Corticosteroids (%)	1291 (73.1)	199 (83.6)	37(88.1)	57(76.0)	105(86.8)	< 0.001
Tocilizumab (%)	77 (4.3)	19 (7.9)	5(11.9)	6(7.8)	8(6.5)	0.023
Baricitinib (%)	114 (6.4)	18 (7.4)	3(7.1)	3(3.9)	12(9.8)	0.624
Antiviral treatment
Paxlovid (%)	377 (21.1)	78 (32.2)	17(40.5)	24(31.2)	37(30.1)	< 0.001
Intravenous immunoglobin (%)	218 (12.2)	41 (16.9)	12(28.6)	12(15.6)	17(13.8)	0.048
Anticoagulation (%)	1116 (62.4)	148 (61.2)	29(69.0)	47(61.0)	72(58.5)	0.705
Respiratory support
High flow nasal cannula (%)	258(14.4)	55 (22.7)	12(28.6)	19(24.7)	24(19.5)	< 0.001
Non-invasive mechanical ventilation (%)	348(19.5)	41 (16.9)	11(26.2)	7(9.1)	23(18.7)	0.350
Invasive mechanical ventilation (%)	630(35.2)	79 (32.6)	13(31.0)	31(40.3)	35(28.5)	0.428
ECMO (%)	32(1.8)	6 (2.5)	3(7.1)	2(2.6)	1(0.8)	0.458
Secondary infection (%)	403 (22.5)	61 (25.2)	11 (26.2)	20 (26.0)	30 (24.4)	0.354
Bacteria (%)	353 (19.7)	56 (23.1)	10 (23.8)	18 (23.4)	28 (22.8)	0.216
Fungus (%)	35 (2.0)	10 (4.1)	5 (11.9)	3 (3.9)	2 (1.6)	0.031
Virus (%)	15 (0.8)	3 (1.2)	2 (4.8)	1 (1.3)	0 (0.0)	0.533
ICU mortality (%)	1106 (61.9)	148 (61.2)	19(45.2)	59(76.6)	70(56.9)	0.889
Duration of ICU stay (days), median [IQR]	8[4, 14]	9[4, 16]	12 [4, 20]	11[4,18]	8[5,14]	0.068

COVID-19: coronavirus disease 2019; SOT: Solid organ transplantation; Other: persons with other causes of immunosuppression; ECMO: extracorporeal membrane oxygenation; ICU: intensive care unit; The p value was the difference between the non-immunosuppression and immunosuppression group

Our study revealed a 61.9% mortality rate among immunocompetent patients, a 45.2% mortality rate for SOT patients, a 76.6% mortality rate among cancer patients, and a 56.9% mortality rate among other immunocompromised patients. After eliminating the effects of factors such as age and APACHE II score at ICU admission through PS matching, the ICU mortality rate in cancer patients remained significantly higher compared to immunocompetent patients (74.6% vs. 58.7%, P = 0.020) (Supplemental Table 4).

Survival analysis

To further explore these findings, we illustrated the survival analysis of immunocompetent and immunocompromised subgroups within the cohort (Fig. 1). The survival duration for cancer patients was notably shorter, while SOT recipients consistently exhibited higher survival rates, significantly surpassing those of other groups.

Fig. 1.

Kaplan-Meier survival curve for all-cause mortality after critically ill COVID-19 diagnosis for selected populations. ICU: intensive care unit; SOT: Solid organ transplantation; Other: persons with other causes of immunosuppression

Relationship between immune status, ICU mortality, invasive ventilation, ECMO, and secondary infection

For SOT patients, the incidence of secondary fungal infections was significantly higher than in the immunocompetent group (11.9% vs. 2.0%, P < 0.01) (Fig. 2). Multivariate analysis indicated that SOT patients admitted to the ICU due to critically ill COVID-19 had a risk of death that was half that of immunocompetent patients (OR = 0.51, 95% CI 0.28–0.94; P = 0.032).

Fig. 2.

Proportion of ICU mortality, ICU length, invasive ventilation, ECMO, secondary infection of patients with critically ill COVID-19 infection. ICU, intensive care unit; SOT, Solid organ transplantation; Other, persons with other causes of immunosuppression; *, p < 0.05; **, p < 0.01; Unmarked bars indicate non-significant results in terms of statistical significance

For cancer patients, the risk of death was found to be twice that of immunocompetent patients (OR = 2.02, 95% CI 1.18–3.46; P = 0.010). Additionally, there were no statistically significant differences in the risks of invasive ventilation, ECMO use, or secondary bacterial infections between SOT and cancer patients compared to other immunocompromised subgroups (Table 3).

Table 3.

Association of different immunocompromised patients with ICU mortality, invasive ventilation, ECMO, secondary infection in multivariable analysis

Group	ICU mortality		Invasive ventilation		ECMO		Secondary bacterial infection		Secondary fungus infection
	OR (95% CI)	p value	OR (95% CI)	p value	OR (95% CI)	p value	OR (95% CI)	p value	OR (95% CI)	p value
Immunocompetent	Reference		Reference		Reference		Reference		Reference
Immunocompromised	0.97	0.833	0.90	0.421	1.29	0.641	1.22	0.217	2.16	0.035
	(0.74–1.28)		(0.69–1.17)		(0.44–3.76)		(0.89–1.69)		(1.06–4.42)
SOT	0.51	0.032	0.79	0.445	3.84	0.075	1.27	0.515	6.77	<0.001
	(0.28–0.94)		(0.43–1.45)		(0.88–16.83)		(0.62–2.61)		(2.51–18.25)
Cancer	2.02	0.010	1.43	0.133	1.01	0.992	1.24	0.435	2.03	0.248
	(1.18–3.46)		(0.90–2.30)		(0.14–7.58)		(0.72–2.13)		(0.61–6.75)
Other	0.97	0.880	0.66	0.067	0.67	0.702	1.40	0.153	0.73	0.663
	(0.63–1.48)		(0.43–1.03)		(0.09–5.11)		(0.88–2.22)		(0.17–3.08)

ICU: intensive care unit; ECMO: extracorporeal membrane oxygenation; SOT: Solid organ transplantation; Other: persons with other causes of immunosuppression

Dynamics of lymphocytes and serum inflammatory markers

In light of these findings, we performed a dynamic observation of inflammatory markers at 24 h, 72 h, and the seventh day post-ICU admission for immunocompetent and immunocompromised patients and the subgroups respectively (Fig. 3).

Fig. 3.

Proportion of inflammation indicators 24 h, 72 h and 7 days after ICU admission. ICU, intensive care unit; SOT, Solid organ transplantation; Other, persons with other causes of immunosuppression; LYM, lymphocyte count; CD4, CD4⁺ T cells; CD8, CD8⁺ T cells; IL-6, interleukin-6; CRP, C-reactive protein; Fet, serum ferritin

For SOT patients, the lymphocyte count was lower than that of the immunocompetent group on the first day, but it exhibited a gradual upward trend by the third and seventh days. Additionally, the counts of both CD4⁺ and CD8⁺ T cells showed a marked increase over this period. The SOT group also demonstrated elevated levels of CRP and serum ferritin, which were higher than those in the immunocompetent group. Notably, the IL-6 level in the SOT patients was significantly lower than in both the immunocompetent and cancer groups, with a gradual decrease from the first to the seventh day.

For cancer patients, the lymphocyte count on the first day was similar to that of the immunocompetent group. In this group, the CD4⁺ T cells increased, while the CD8⁺ T cells decreased. Like the SOT group, the cancer patients also presented high levels of CRP and serum ferritin.

Discussion

This study is the largest clinical investigation of critically ill COVID-19 patients in ICU, focusing on immunocompromised individuals categorized by different causes of immunosuppression. Key findings include: (1) Cancer patients faced twice the mortality risk from critically ill COVID-19 compared to immunocompetent patients and had a significantly higher mortality rate than SOT patients. (2) In cancer patients with critically ill COVID-19, a decline in CD8⁺ T lymphocytes shortly after ICU admission suggests a poor prognosis. In contrast, SOT patients showed an increase in CD8⁺ T lymphocytes and consistently low IL-6 levels, indicating a favorable prognosis.

In our study, the ICU mortality rate for immunocompetent patients was slightly higher than in similar research [22]. Additionally, patients with an APACHE II score of 16 upon ICU admission had a significantly higher actual mortality rate than predicted [19, 23], primarily due to being older and having more comorbidities. The APACHE II score, which mainly assesses bacterial infections like sepsis [24], is less accurate for predicting mortality in severe viral pneumonia cases, potentially leading to an underestimation [25]. Besides, we hypothesize that increased mortality in cancer patients may be linked to their treatment, as prior research identified chemotherapy within four weeks before COVID-19 infection as a risk factor for in-hospital death [26]. We categorized cancer patients based on chemotherapy status and found minimal impact on prognosis, possibly due to significant age differences in our sample (Supplemental Table 5). However, the limited sample size hindered PS matching for age.

Cytokine release syndrome has been identified as a key pathophysiological mechanism in COVID-19 since the early reports [27]. Cytokine storms significantly impact clinical outcomes in both immunocompetent and immunocompromised patients. Our data reveal that immunocompetent patients with fatal outcomes exhibited higher levels of CRP, serum ferritin, IL-6, IL-8, IL-10, and PCT compared to survivors, highlighting the critical role of cytokines (Supplemental Table 6). In immunocompromised individuals, excessive cytokine release leads to severe systemic inflammation, significant organ dysfunction, and heightened vulnerability to complications [28]. These patients often present with severe symptoms, such as persistent fever, hypotension, and multi-organ failure, which jeopardize their survival [29]. While immunocompetent patients can also experience cytokine storms, their immune systems are generally more resilient [30], though they still face risks from extreme cytokine dysregulation [28].

Understanding the different clinical outcomes in these populations is essential for optimizing treatment strategies during cytokine storms. Previous studies [31] have shown that during the acute phase of COVID-19, immune cell activation leads to the release of CRP, ESR, procalcitonin, IL-6, and ferritin. In severe cases, significant alterations in IL-1β, IL-6, and TNF-α levels, along with a decline in CD8⁺ T lymphocytes, correlate strongly with disease severity [27, 32]. These biomarkers are likely involved in the inflammatory response and immune regulation during severe COVID-19 infections [33, 34]. Therefore, we conducted analyses on these aforementioned markers and included them as references (Supplementary Fig. 2). Our analyses indicate significant changes in lymphocytes, particularly CD8⁺ T lymphocytes, IL-6, CRP, and ferritin after severe COVID-19 infection in cancer and SOT patients, which we will discuss further.

Our study found that in cancer patients diagnosed with critically ill COVID-19, there was an increase in CD4⁺ T lymphocytes and a significant decrease in CD8⁺ T lymphocytes one week after diagnosis, potentially indicating a poorer prognosis. While both CD4⁺ and CD8⁺ T cells are lymphocytes, they have distinct roles in the cytokine storm associated with critically ill COVID-19 [35]. CD4⁺ T cells may enhance and modulate immune responses, leading to increased inflammation [36, 37], whereas CD8⁺ T cells are crucial for directly killing infected cells and providing protection against secondary infections [38]. The observed increase in CD4⁺ T cells and decrease in CD8⁺ T cells by day 7 suggests that although inflammation is activated, the reduction of CD8⁺ T cells—vital for pathogen clearance—could contribute to higher mortality rates. Previous studies [39, 40] have reported persistently reduced counts of lymphocyte subsets, especially CD8⁺ T cells, in critically ill COVID-19 patients, identifying them as independent predictors of disease severity and treatment efficacy. Lymphopenia is a known biomarker for viral infections, affecting 63–85% of SARS-CoV-2 infected patients, with around 20% of deceased patients exhibiting severe lymphopenia—a predictive indicator of COVID-19 mortality in immunocompromised individuals [41, 42]. However, this may not apply to cancer patients, as our findings show that they can have normal baseline lymphocyte counts yet still experience high mortality rates. Therefore, baseline lymphocyte counts may not accurately predict mortality in this population. Instead, a persistent decline in lymphocyte counts, particularly CD8⁺ T cells, could serve as a more reliable predictor of poor prognosis in cancer patients with COVID-19.

Dynamic monitoring of SOT patients revealed low baseline lymphocyte counts; however, there was a significant early increase in CD8⁺ T lymphocytes following diagnosis of critically ill COVID-19, accompanied by low IL-6 levels. These characteristics suggest a potentially favorable prognosis. Previous studies have also found that SOT patients admitted to the ICU for critically ill COVID-19 had lower IL-6 levels compared to non-SOT patients [43]. Elevated serum IL-6 levels have been linked to increased mortality, while decreased IL-6 may contribute to secondary fungal infections [36, 44–47]. Our findings indicate that SOT patients experience a higher incidence of these infections alongside notably lower IL-6 levels compared to other groups. Thus, we propose that monitoring IL-6 could serve as a potential biomarker for assessing the risk of secondary fungal infections in these patients, though further research is needed to validate this hypothesis.

Our study has several limitations. Firstly, as a retrospective observational cohort study involving 242 immunocompromised patients admitted to the ICU, it may not fully represent the broader immunocompromised population. Secondly, the waning of the COVID-19 pandemic makes it challenging to establish a larger external cohort for validating our predicted outcomes. Thirdly, due to the complex origins of immunosuppression in cancer patients—whether from their primary diseases or secondary effects of chemotherapy—our study could not provide detailed classifications based on etiology or specific treatments. This limits our ability to investigate the types of cancers related to high mortality rates and to guide prognosis for these patients following severe COVID-19. Lastly, as a retrospective study, it contains inevitable missing data, including details on immunosuppressive drug usage prior to infection. We hope future studies will gather more comprehensive data on this population to further validate our findings.

Conclusion

Critically ill COVID-19 patients exhibit varied clinical outcomes based on their immunosuppression status, with cancer patients facing the highest mortality rates due to differing inflammatory responses. Our findings indicate that while cancer patients may have normal initial lymphocyte counts, they experience a decline in CD8⁺ T lymphocytes shortly after ICU admission, suggesting a potentially unfavorable prognosis. Conversely, SOT patients, despite lower baseline lymphocyte counts, show an increase in CD8⁺ T lymphocytes and consistently low IL-6 levels, indicating a more favorable prognosis. This study underscores the importance of monitoring the dynamic changes within a combination model of both CD8⁺ T lymphocytes and IL-6 as a prognostic tool for critically ill COVID-19 patients, particularly those with cancer and SOT, emphasizing the need for further research in this area.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

APACHE

Acute physiology and chronic health evaluation

BMI

Body mass index

CAP

Community-acquired pneumonia

COVID-19

Coronavirus disease 2019

CRP

C-reactive protein

CTD

Connective tissue diseases

DMARDs

Disease-modifying antirheumatic drugs

ICU

Intensive care unit

IL

Interleukin

PCR

Polymerase chain reaction

PS

Propensity score

SARS-CoV-2

Respiratory Syndrome Coronavirus 2

SOFA

Sequential organ failure assessment

SOT

Solid organ transplant

Author contributions

CYL, HYH, XJW, QYZ designed the study. CYL drafted the first manuscript. CYL and YQW performed the analyses. LNH, ZYC, QZ, YC, TSZ conducted data collection and organization. HYH, XJW, QYZ requested major revisions and provided comments on the manuscript. All authors read and approved the final manuscript.

Funding

The funding sources of the study [National High Level Hospital Clinical Research Funding & Elite Medical Professionals Project of China-Japan Friendship Hospital (NO. ZRJY2023-GG20)], [Key Research and Development Program of Jiangxi province (NO. 20232BBG7002)] and [CAMS Innovation Fund for Medical Sciences (NO. 2022-I2M-JB-016)] are academic non-profit organisations that played no role in the design of the study, collection, analysis, and interpretation of data, and in writing the manuscript.

Data availability

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Research Ethics Commission of China-Japan Friendship Hospital (2019-79-K51-1). Due to the study’s retrospective nature, the need for informed consent from the patients or their legal guardians was waived. All methods were carried out in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.