Sustained elevation of soluble B- and T- lymphocyte attenuator predicts long-term mortality in patients with bacteremia and sepsis.

Soluble B and T lymphocyte attenuator (sBTLA) has been shown to be associated with severity and outcome, in critically ill septic patients. We aimed to assess the dynamic expression of sBTLA, as a prognostic biomarker of long-term mortality in patients with bloodstream infection (BSI) and sepsis, and to evaluate its association with biomarkers indicative of inflammation and immune dysregulation. Secondarily, sBTLA was evaluated in association with severity and bacterial etiology. Patients with BSI (n = 108) were prospectively included, and serially sampled from admission to day 28. Blood and plasma donors (n = 31), sampled twice 28 days apart, served as controls. sBTLA concentration in plasma was determined with enzyme-linked immunosorbent assay. Associations between sBTLA on day 1-2 and 7, and mortality at 90 days and 1 year, were determined with unadjusted, and adjusted Cox regression. Differences related to severity was assessed with linear regression. Mixed model was used to assess sBTLA dynamics over time, and sBTLA associations with bacterial etiology and other biomarkers. sBTLA on day 1-2 and 7 was associated with mortality, in particular failure to normalize sBTLA by day 7 was associated with an increased risk of death before day 90, adjusted HR 17 (95% CI 1.8-160), and one year, adjusted HR 15 (95% CI 2.8-76). sBTLA was positively associated with CRP, and negatively with lymphocyte count. sBTLA on day 1-2 was not linearly associated with baseline SOFA score increase. High SOFA (≥4) was however associated with higher mean sBTLA than SOFA ≤3. sBTLA was not associated with bacterial etiology. We show that sustained elevation of sBTLA one week after hospital admission is associated with late mortality in patients with BSI and sepsis, and that sBTLA concentration is associated with CRP and decreased lymphocyte count. This suggests that sBTLA might be an indicator of sustained immune-dysregulation, and a prognostic tool in sepsis.

Introduction

There is growing interest in long-term sepsis outcomes. While pre-existing comorbidities contribute to late deaths, epidemiological studies indicate that sepsis alone is associated with increased mortality in the subsequent months to years [1, 2].

Evidence suggests that a dysregulated immune response, with persistent inflammation and immunosuppression, may increase the risk of late complications and mortality in sepsis [3-5]. In sepsis survivors, persistent inflammation and immunosuppression can be detected as long as one year after the original sepsis episode [5, 6]. Advances in the understanding of the sepsis immune response has contributed to the new international definition of sepsis, and motivated interventional trials aiming to restore immune homeostasis, but much remains to be outlined in terms of the mechanisms behind immune dysregulation, and how and when it should be assessed [7-10].

Checkpoint inhibitor immunotherapy has transformed cancer treatment, by amplifying ant-tumor responses, and has been found to trigger immune-related adverse reactions, such as colitis and skin rash [11, 12]. PD-1, CTLA-4, and their ligands, have been shown to be upregulated on T cells during sepsis, which has led to investigations on immune checkpoint inhibition also as immunomodulatory treatment in infections [13, 14]. Pre-clinical studies provide evidence that blockade of PD-1 and CTLA-4 could reverse immune dysfunction and improve survival in sepsis, and three recently published phase-1 studies showed that PD-1 and PD-L1 antibody administered to patients with severe sepsis could restore immune cell function in septic patients with low lymphocyte count [9, 10, 15].

B and T lymphocyte attenuator (BTLA) is an important membrane-associated immune-regulatory protein, which is expressed on antigen-presenting cells, T and B lymphocytes and innate immune cells, and is up-regulated after T cell activation [16]. It shares its counter-receptor, herpes virus entry mediator (HVEM), which is also widely expressed by immune cells (T cells, B cells, NK cells, DC and myeloid cells), with other regulatory proteins, and is therefore considered a more fine-tuned immune modulator than PD-1 and CTLA-4 [17]. The BTLA-HVEM pathway has gained recent interest in cancer immunotherapy and autoimmunity [18, 19]. Interestingly, an anti-BTLA-antibody (TAB004) was approved for clinical trials in 2019, and is currently under investigation for treatment of advanced malignancy (NCT04137900).

Co-stimulatory molecules exist in both membrane-bound and soluble forms, which lack the transmembrane domain, and are produced either by shedding of the membrane form, or through alternative mRNA splicing [20]. Transcripts coding for soluble BTLA has been identified in B cells, CD4 and CD8 T cells [21]. We have previously shown that soluble BTLA (sBTLA) is associated with disease severity and short-term mortality, in severe sepsis and septic shock in the intensive care unit (ICU) setting, and others have further demonstrated that sBTLA can discriminate sepsis from other critical illness [22, 23]. Furthermore it was recently shown that sBTLA is associated with an increased risk of death in cancer patients [24-26].

In this study of patients with community-onset BSI, we set out to establish whether sBTLA is associated with mortality at 90 days and one year, and with markers of inflammation and immunosuppression over time. Additionally, we aimed to study sBTLA dynamics over 4 weeks in BSI compared to healthy controls. Finally, we wanted to address whether sBTLA is associated with disease severity, and whether there is a difference in sBTLA production with respect to bacterial etiology.

We found that sustained elevation of sBTLA one week after hospital admission is associated with late mortality in patients with BSI and sepsis, and that sBTLA concentration is associated with decreased lymphocyte count.

Materials and methods

Setting, study population and definitions

A prospective single-center observational study of patients with community-onset bloodstream infection was conducted at Örebro University Hospital, Örebro, Sweden, between November 2010 and May 2014. Patients >18 years of age, admitted to the Departments of Infectious Diseases and Internal Medicine with a suspected infection, and in whom a blood culture drawn on hospital admission (day 0) showed growth of clinically significant bacteria within 3 days after hospital admission, were eligible for inclusion, and asked for study participation by a medical doctor. Exclusion criteria were infection with HIV, Hepatitis B and C or previous inclusion in the study. The prevalence of the diagnoses in exclusion criteria are low in the studied population, therefore the sample was considered to be representative of the studied population. The study was approved by the regional ethical review board in Uppsala, Sweden. A written informed consent was obtained from patients or, if unable to sign, a relative.

The increase from each patient’s baseline Sequential Organ Failure Assessment (SOFA) score was calculated for the first 24 hours after hospital admission, and is referred to as day 1 Δ-SOFA. Blood samples were drawn from patients on day(s) 0, 1–2 (day of enrolment), 3, 7±1, 14±2, and 28±4 after admission, for biochemical tests, and analyzed with accredited routine laboratory methods. Blood was also collected for processing of plasma for future analyses. Monocytic HLA-DR (mHLA-DR) expression on monocytes was measured from day 1–2 with flow cytometry. HLA-DR data is described in detail elsewhere [27]. Patient data and information on outcome was abstracted from medical records. Healthy blood and plasma donors (n = 31) served as control subjects and were sampled twice, 4 weeks apart. Plasma was kept at -80°C pending analysis.

Blood and other cultures

Two sets of blood cultures, each consisting of 20 ml blood distributed equally between one aerobic and one anaerobic bottle, were incubated in a Bactec blood culturing system (Becton Dickinson, Franklin Lakes, NJ, USA). The bacterial species was determined by routine laboratory diagnostic procedures. Other cultures (urine, expectorated sputum or nasopharyngeal aspirate, cerebrospinal fluid, wound swab, joint aspirate) were taken depending on clinical suspicion of diagnosis, according to clinical routine.

Definitions

Sepsis was defined as an increase of the SOFA score of 2 points or more from baseline, according to The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [7].

ELISA analyses

BTLA in plasma was analysed in duplicate at 1:2 dilution with Human B- and T Lymphocyte Attenuator (BTLA) ELISA kit, Cusabio, detection range 0.45–30 ng/mL.

Statistics

IBM SPSS Statistics for Windows, Version 25.0 (IBM corp. Armonk, NY, USA) was used for statistical calculations. Due to the high number of missing samples on admission, the focus of analyses was on measurements day 1–2 on. Missing data analysis was made, using chi-square, Fisher, unpaired t, or Mann-Whitney test, according to variable distribution, to compare demographic and clinical characteristics in subjects with ≤1, or 2 or more missing plasma samples.

Non-normally distributed variables (sBTLA, CRP, and mHLA-DR) were log-transformed in further analyses. The sBTLA concentrations referred to in the results section are log10 sBTLA.

Cox regression was used to assess time to 90-day and 1-year mortality, in relation to sBTLA on day 1–2 and day 7, with complete follow up for all patients. sBTLA was evaluated on a continuous scale, and as a binary marker (high vs. low), where the threshold concentration was defined as the nearest even concentration above the highest concentration in the control group. Adjustments were made for age, sex, comorbidity on a categorical scale (0, 1–2, ≥3), and severity (day 1 Δ SOFA), on a continuous scale.

We used logistic regression to estimate the ability of sBTLA, on day 1–2, to differentiate sepsis from non-septic BSI, and on day 1–2 and 7 to predict survival on day 90 and 1 year, by calculating the area under the Receiver Operating Characteristic (ROC) curve (AUC) of the sensitivity and 1-specificity.

Linear Mixed Model with random intercept, and sBTLA as outcome, was used to evaluate sBTLA dynamics in patients over days 0–28, as well as to study age covariate-adjusted associations with inflammatory and immunosuppression markers from day 1–2 to 28, followed by stratified analyses, at individual time-points. Days after hospital admission was a fixed factor. Association coefficients are presented as mean difference (in text), and B (in figures), with 95% CI.

Unadjusted and age-adjusted linear regression with sBTLA as outcome was used to assess differences in sBTLA in patients and controls on day 0 and 28 and on day 1–2 to assess differences depending on day 1 ΔSOFA categorized as ≤3 and ≥4.

Patients were classified according to bacterial etiology: 1) E. coli, 2) S. aureus, 3) S. pneumoniae, and 4) other. Linear Mixed Model with random intercept, and sBTLA as outcome, was used to evaluate age-adjusted mean differences of sBTLA over days 1–2 to 28, between etiology groups 1–3. Fixed factors were days after hospital admission, etiology, and their interaction term. Autoregressive heterogeneous correlation structure was selected from best fit regarding Akaike information criteria. P-values comparing etiology groups were Bonferroni-corrected.

Statistical significance was defined as p<0.05.

Results

Study population

Altogether, 116 patients were enrolled. Eight were excluded due to growth of non-pathogenic bacteria (n = 5), or no available plasma samples (n = 3), leaving 108 valid subjects. Plasma was available for sBTLA analysis from 46 patients on admission, 96 patients on day 1–2, 70 patients on day 3, 85 patients on day 7, 77 patients on day 14, and 70 patients on day 28. Subjects with 2 or more missing plasma samples (n = 60) were older (median age 73 vs. 69, p = 0.03) but not statistically significantly differing with regard to other baseline characteristics, bacterial etiology, disease severity, or short or long term mortality from those with ≤1 missing samples (n = 48), S1 Table.

Patient characteristics and control demographics are detailed in Table 1. The main sources of infection were pneumonia, urinary tract infection, skin/soft tissue, and joint infection. Fifty-four subjects (50%) met criteria for sepsis. Fourteen (13%) were treated in the ICU at some point during their hospital stay, due to need of organ supportive therapy, such as vasoactive drugs, or mechanical ventilation, as a result of sepsis. Ten out of 108 (9.3%) patients were immunocompromised at baseline, as defined by APACHE II criteria [28]. The frequency of antibiotic resistance was low: one subject had growth of ESBL-producing E. coli, and none had MRSA. The controls had a mean age of 53 years (SD 8.3, range 41–73) and 7 (23%) were female.

Table 1

Patient and control characteristics.

	BSI (N = 108)	Controls (N = 31)
Demographics
Age, mean years (SD) range	70 (14.6) 24–93	53 (47–59)
Sex, female	49 (45)	7 (23)
Comorbidities
No comorbidities	37 (34)
Ischemic heart disease	21 (19)
Congestive heart failure	16 (15)
Peripheral vascular disease	5 (5)
Diabetes with complications	14 (13)
Chronic lung disease	8 (7)
Connective tissue disease	9 (8)
Moderate/severe kidney disease	7 (6)
Chronic liver disease	0
Malignancy the last 5 years	5 (5)
Immunosuppression	10 (9)
Day 1 Δ-SOFA1
0–1	54 (50)
2	22 (20)
3–4	19 (18)
≥5	13 (12)
Pathogen
E. coli	25 (23)
S. aureus	27 (25)
S. pneumoniae	29 (27)
Beta-hemolytic streptococci	7 (6)
Other gram negatives	11 (10)
Other gram-positives	5 (5)
Polymicrobial	4 (4)
Site of infection
Lungs/airways	29 (27)
Urinary tract	26 (24)
Abdominal	7 (7)
Skin/soft tissue	10 (9)
Joint/bone	11 (10)
Endocarditis	8 (7)
Primary BSI	2 (2)
Graft infection	2 (2)
Meningitis	3 (3)
Unknown	10 (9)

Data in presented as N (%), unless otherwise stated.

1The increase from the patients baseline SOFA score.

sBTLA in BSI and controls

In patients, sBTLA was statistically significantly negatively associated with age on day 1–2, a decrease of -0.004 (95% CI 0.009 to 0.0, p = 0.045) log10 sBTLA units per year, but not at other time-points. In controls, there was no significant association between sBTLA and age.

Mean log10 sBTLA concentrations were statistically significantly higher in subjects with bacteremia, compared to controls. On day zero, unadjusted mean difference 0.43, 95% CI 0.24–0.62, p<0.001, and age-adjusted mean difference 0.47 (95% CI 0.25–0.70) p<0.001. On day 28, unadjusted mean difference 0.14 (95% CI 0.04–0.25) p<0.008, age-adjusted mean difference 0.13 (95% CI 0.01–0.25) p = 0.033, Fig 1. The median peak day of sBTLA was day 1–2. The linear mixed model showed that mean sBTLA decreased by 34.1% (95% CI 13.7 to 49.6) from admission to day 28 in subjects with BSI, with a mean decrease by 7.6% (95% CI 4.7 to 10.5) per time point.

Fig 1

sBTLA concentrations in BSI and controls from baseline to day 28.

Dotted line at 11 ng/mL. Outliers >80 ng/mL have been excluded from the figure to increase clarity (day 0: n = 3, day 3: n = 1, day 7 n = 1, day 14: n = 1).

Disease severity

The association between day 1–2 sBTLA and day 1 ΔSOFA was non-linear, Fig 2. High ΔSOFA-scores (≥4, n = 20) were associated with higher mean log10 sBTLA than low scores (≤3, n = 76), mean difference 0.35 (95% CI 0.22–0.48, p<0.001. However, sBTLA did not perform well in differentiating sepsis (ΔSOFA score ≥2) from non-septic BSI at this time-point, AUC 0.54 (95% CI 0.43–0.66).

Fig 2

Day 1–2 sBTLA concentrations and day 1 delta-SOFA.

ΔSOFA: 0 (n = 20), 1 (n = 27), 2 (n = 18), 3 (n = 11), 4 (n = 7), 5 (n = 6), 6 (n = 3), and 7 (n = 4).

Mortality

Among all 108 study patients, death from any cause was 4.6% (n = 5) on day 28, 11.1% (n = 12) on day 90, and 20.4% (n = 22) at one year. 7 patients died as a result of uncontrolled infection during hospital care for the index BSI, out of which 3 were not admitted to intensive care due to treatment limitations.

On day 1–2, unadjusted HR, for time to 90-day mortality, was 4.6 (95% CI 0.6–34) p = 0.138, per log sBTLA unit, and adjusted (age, sex, ΔSOFA and comorbidity) HR was 114 (3.8–3409) p = 0.006. Unadjusted HR for sBTLA ≥11 (compared to <11) was 2.2 (95% CI 0.6 to 8.1) p = 0.236, and adjusted HR was 5.7 (95% CI 1.1–29) p = 0.034, Table 2.

Table 2

Relative risk for 90-day and 1-year mortality.

		90-day mortality						1-year mortality
	BSI n = 96	Events1	Rate2	Unadjusted		Adjusted3		Events1	Rate2	Unadjusted		Adjusted3
Day 1–2				HR (95% CI)	p	HR (95% CI)	p			HR (95% CI)	p	HR (95% CI)	p
log10 sBTLA, per unit	96	12		4.6 (0.6–34)	0.14	114 (3.8–3409)	<0.01	20		2.3 (0.5–11)	0.30	10 (0.9–111)	0.06
sBTLA<11	40	3	32	Ref		Ref		5	14	Ref		Ref
sBTLA≥11	56	9	70	2.2 (0.6–8.1)	0.24	5.7 (1.1–29)	0.03	15	33	2.3 (0.8–6.2)	0.11	4.3 (1.3–14)	0.02
	BSI n = 85	Events 1	Rate 2	Unadjusted		Adjusted 3		Events	Rate 2	Unadjusted		Adjusted 3
Day 7				HR (95% CI)	p	HR (95% CI)	p			HR (95% CI)	p	HR (95% CI)	p
log10 sBTLA, per unit	85	10		5.4 (0.7–40)	0.10	22 (1.3–371)	0.03	17		6.1 (1.4–30)	0.02	24 (2.8–208)	<0.01
sBTLA<11	42	1	11	Ref		Ref		2	5	Ref		Ref
sBTLA≥11	43	9	102	9.7 (1.2–77)	0.03	17 (1.8–160)	0.01	15	47	8.8 (2.0–39)	<0.01	15 (2.8–76)	<0.01

1 Number of deaths.

2 Mortality rate defined as deaths per 100 person years at risk.

3 For age, sex, delta-SOFA, and comorbidity (Charlson score (0, 1–2, ≥3).

On day 1–2, unadjusted HR, for time to 1-year mortality, was 2.3 (95% CI 0.5–11) p = 0.300 per log sBTLA unit, and adjusted HR was 10 (0.9–111) p = 0.061. Unadjusted HR for sBTLA≥11 (compared to <11) was 2.3 (95% CI 0.8 to 6.2) p = 0.114, and adjusted HR was 4.3 (95% CI 1.3–14) p = 0.016, Table 2.

On day 7, unadjusted HR, for time to 90-day mortality, was 5.4 (95% CI 0.7–40) p = 0.098 per log sBTLA unit, and adjusted HR was 22 (1.3–371) p = 0.034. Unadjusted HR for sBTLA ≥11 (compared to <11) was 9.7 (95% CI 1.2 to 77) p = 0.031, and adjusted HR was 17 (95% CI 1.8–160) p = 0.013, Table 2 and S1 Fig. AUC for day 7 sBTLA to predict mortality at 90 days was 0.72 (95% CI 0.59–0.84), Fig 3A.

Fig 3

Prognostic capacity of sBTLA on day 7.

(A) ROC for sBTLA on day 7 to predict 90-day mortality (B) ROC for sBTLA on day 7 to predict 1-year mortality.

On day 7, unadjusted HR, for time to 1-year mortality, was 6.1 (95% CI 1.4–30) p = 0.018 = per log sBTLA unit, and adjusted HR was 24 (2.8–208) p = 0.004. Unadjusted HR for sBTLA ≥11 (compared to <11) was 8.8 (95% CI 2.0–39) p<0.004, and adjusted HR was 15 (95% CI 2.8–76) p = 0.001, Table 2 and Fig 4. AUC for day 7 sBTLA to predict mortality at one year, was 0.75 (95% CI 0.64–0.86), Fig 3B.

Fig 4

Survival curves for high and low sBTLA on day 7.

sBTLA <11 ng/mL (n = 42), and sBTLA ≥11 ng/mL (n = 43). The p-value shown in the figure is after adjustment for age, sex, day 1 ΔSOFA, and comorbidity.

Significant associations between day 1–2 and day 7 sBTLA and mortality at 90 days and 1 year remained after adjustment for the presence of immunosuppression at baseline, S2 Table.

Associations with markers of inflammation and immunosuppression

The linear mixed model analysis, with sBTLA as outcome, and adjustment for time, showed a significant positive association with CRP, on average an increase of 0.15 (95% CI 0.09 to 0.20) p<0.001 sBTLA log units per log CRP increase, and a significant negative association with lymphocyte count, i.e. a decrease of -0.03 (95% CI -0.05 to -0.008) p<0.008 sBTLA log units per lymphocyte count increase. There was no significant association with neutrophil count (p = 0.876), or mHLA-DR (p = 0.314), Fig 5. Associations at individual time-points are presented in the figure. In addition to the above, there was no association between sBTLA and neutrophil to lymphocyte count ratio (p = 0.26).

Fig 5

Associations between sBTLA concentrations and immunological markers.

There was an overall significant association with CRP (A), and an overall negative association with lymphocyte count (B). There was no statistically significant association with neutrophil count (C), or with mHLA-DR (D). Association coefficients, and 95% CI (in parentheses), are shown for each individual time-point.

Bacterial etiology

The linear mixed model analyses showed no statistically significant interaction effect, that is, no significant overall difference in mean BTLA dynamics over time between the etiology groups (p = 0.41). When the etiology groups were compared at different time-points, no major differences in sBTLA concentrations were observed between S. aureus, E. coli and S. pneumoniae etiology, Fig 6.

Fig 6

BTLA concentrations in main etiology groups day 1–2 to day 28 post hospital admission.

Day 1–2: E. coli N = 20, S. aureus N = 24, S. pneumoniae N = 28 (E. coli vs S. aureus p = 1.0, E. coli vs S. pneumonia p = 0.4, S. aureus vs S. pneumonia p = 0.2). Day 3 E. coli N = 18, S. aureus N = 20, S. pneumoniae N = 17 (E. coli vs S. aureus p = 1.0, E. coli vs S. pneumonia p = 0.1). Day 7: E. coli N = 17, S. aureus N = 24, S. pneumoniae N = 25 (E. coli vs S. aureus p = 1.0, E. coli vs S. pneumonia p = 0.1, S. aureus vs S. pneumonia p = 0.2). Day 14: E. coli N = 14, S. aureus N = 19, S. pneumoniae N = 23 (E. coli vs S. aureus p = 1.0, E. coli vs S. pneumonia p = 0.9, S. aureus vs S. pneumonia p = 1.0). Day 28: E. coli N = 15, S. aureus N = 16, S. pneumoniae N = 21 (E. coli vs S. aureus p = 0.7, E. coli vs S. pneumonia p = 0.2, S. aureus vs S. pneumonia p = 1.0).

Discussion

This study assessed sBTLA over 28 days after admission for bloodstream infection. The main finding is that sBTLA is associated with increased mortality at 90 days and 1 year. Additionally, we show for the first time that sBTLA is negatively associated with lymphocyte count.

We find that high sBTLA concentrations measured on day 1–2 and 7 days after hospital admission are associated with an increased risk of death before 90 days and one year. The association is independent of disease severity, age, and baseline comorbidities, and is most notable for failure to normalize sBTLA after 1 week’s hospitalization. We previously demonstrated an association with early mortality, for high sBTLA concentration measured ≤24 hours after a diagnosis of severe sepsis or septic shock, in ICU-treated patients. This association was stronger in time-varying analysis, also indicating that the trajectory of sBTLA concentrations is of importance [22]. While one third of deaths in the studied cohort could be directly related to failure to clear the infection associated with the BSI, we did not have information on the cause of death in the majority of cases. The association between sBTLA and mortality may be due to immunological changes that occurred as a result from the index BSI. However, it could also reflect a pre-existing (i. e. before the infection) immune pathology that is associated with premature death. Experimental sepsis models have shown that BTLA deficient mice have a survival benefit, but treating wild type mice with anti-BTLA monoclonal antibody is unfavorable [29, 30]. It is possible that interference by sBTLA can account for these conflicting results. A recent study on circulating levels of 14 immune-checkpoint related proteins in patients with clear cell renal cancer, showed that high levels of sBTLA alone, were associated with an increased death risk [25]. Soluble BTLA has also been shown to have prognostic significance in other solid malignancies [24, 26, 31].

This study finds the strongest association with mortality for the failure to normalize sBTLA by day 7. Recent studies support dynamic assessment of inflammation and immunosuppression markers. In a retrospective study of patients with bacteremia and sepsis, Drewry et al. found that lymphopenia on day 4, but not on admission, was associated with 28-day mortality [32]. Stortz et al. studied critically ill septic patients with different clinical trajectories. They compared absolute lymphocyte count and levels of mHLA-DR and sPD-L1, and found that there were no differences at early assessment points (< 4 days), but that those with protracted critical illness failed to return to normal levels from day 7 onwards [33]. In another study, patients were followed until one year after sepsis, and the authors identified a patient phenotype with sustained inflammation and immunosuppression (elevated C-reactive protein (CRP) and sPD-L1), that was associated with increased long-term mortality and risk of hospital readmission [5].

This study shows an association between sBTLA and decreased lymphocyte count, which might indicate that sBTLA is linked to the process of immunosuppression in sepsis. There was however no significant overall association with mHLA-DR, another established marker of an immunosuppressed phenotype in sepsis [34]. We also find that sBTLA concentrations are positively associated with CRP, i e. with ongoing inflammation. In a previous study, we found no correlation between sBTLA and CRP, measured early in severe sepsis or septic shock. This discrepancy might be accounted for by the different analytic methods, where we in the present study took into account all measurements of sBTLA and CRP, which increases statistical power. Our findings support the latest model describing sepsis pahophyisiology beyond the acute phase, i e. the Persistent Inflammation, Immunosuppression, and Catabolism Syndrome (PICS), in which sustained elevation of CRP along with lymphopenia and low serum albumin is associated with poor outcome, and suggest that sBTLA might be an interesting addition to these established markers for immune monitoring [35].

We find that sBTLA concentrations peak early in BSI, decrease gradually, but on a group level remain elevated compared to controls 28 days post hospital admission. We do not find an overall age association for sBTLA in patients with BSI or controls. sBTLA was previously shown to exist at low concentrations, but increasing with age, in serum from healthy subjects [36]. A recent study by Gorgulho et al. showed increased levels of sBTLA in patients with advanced cancer compared to healthy controls, but no significant difference in in older (>67) compared to younger patients [31]. We previously reported that sBTLA is higher in sepsis and other critical illnesses, compared with healthy controls [22]. Monaghan et al. recently made similar findings, in human sepsis, as well as in a murine sepsis model, and they also showed that sBTLA is an alternative mRNA splice product, that is produced to a higher degree relative to the full length receptor, in critical illness in mice [23]. As to studies on cell-bound BTLA expression, results are conflicting, ranging from higher to lower CD4+ BTLA expression in sepsis compared to healthy volunteers [13, 37, 38]. The target cells and immunoregulatory effect of sBTLA remains to be fully clarified. A study by Han et al. showed that sBTLA can block BTLA-HVEM interaction, and in an anti-tumor vaccine model, they found that co-treatment of splenocytes with sBTLA resulted in increased IL-2 and IFN-γ and T cell cytotoxic activity [39]. Similarly, the Monaghan study also demonstrated that sBTLA increased splenocyte proliferation in vitro [23].

Among BSI patients, of which 50% fulfilled sepsis criteria, we find no linear association between sBTLA and severity, and that sBTLA cannot differentiate sepsis from non-septic BSI. sBTLA is however notably increased at higher ΔSOFA scores. In a previous study of ICU-treated patients with sepsis and septic shock, who had a mean enrolment SOFA of 10, we saw that sBTLA was correlated with SOFA score, and the strongest correlation was found at 24 hours after study enrolment [22]. A study by Shao et al. showed that BTLA expression on CD4+ cells is high in healthy controls compared to sepsis, with a gradual decrease of expression with increasing severity [38]. The alternative mRNA splice mechanism proposed by Monaghan et al., as described above, could account for an inverse relation between findings regarding soluble and membrane-associated BTLA in sepsis. On the other hand, Gorgulho et. al recently assessed the association between sBTLA and BTLA cell expression in a small number of cancer patients, and found a strong correlation between serum levels of sBTLA and the percentage of peripheral CD8+T cells expressing BTLA [31].

One limitation of this study is the size of the BSI and control cohorts, which were also unmatched with regard to age and comorbidities. Another, is the low number of samples at baseline, and the loss of sampling over time, which we tried to account for with missing samples analyses. The number of deaths at 90 days and 1 year is low, resulting in wide confidence intervals of hazard ratio estimates. An evaluation of the prognostic capacity of sBTLA in larger studies is therefore warranted. Furthermore, the association we report is with all-cause mortality. Finally, we chose a cut-off for sBTLA-concentration that was based on the upper limit concentration in the control group, and it is possible that the normal range in different populations differs from that in this sample. For reference, there is however one study that supports this choice of cut-off as the above “normal” [36].

Strengths of this study include the clear etiologic definition of the study cohort. There was a low prevalence of antibiotic resistance, thus treatment failure due to ineffective antibiotics is less likely to have affected the outcome. The range of severity from uncomplicated bloodstream infection to septic shock also reflects the protean presentations of bacterial infections.

It has been suggested that future adjuvant sepsis strategies must be individualized, which requires readily available clinical and biomarker-guided phenotyping. This was recently addressed in two retrospective cohort studies [40, 41]. The present study further supports that sBTLA may hold potential as a marker of disturbed immune homeostasis, and a prognostic indicator in sepsis. Together with the increasing number of reports promoting BTLA as a prognostic marker in cancer, mechanistic studies on how, and to what extent sBTLA modifies the immune response are motivated. This can also shed light on whether the BTLA-HVEM pathway might be a target for immune intervention. It is however possible that the observed association between increased levels of sBTLA and mortality reflects an immunological vulnerability, or anergy, in some individuals, which increases the risk of a life-threatening infection. To address that question, we need studies on sBTLA dynamics with relevant control groups, i e. non-infected individuals with increased risk for BSI and sepsis, and we need longer follow-up times with documentation of the cause of death, to better understand the association with mortality. The utility of sBTLA, alone, or in combination with for example CRP and lymphocyte count, in immune profile assessment, must also be validated in larger studies on septic and other dysregulated immunological conditions.

Conclusions

We show that soluble B and T lymphocyte attenuator increases in community-acquired bloodstream infection and remains elevated, on a group level, compared with healthy controls up to 28 days. Sustained elevation of sBTLA one week after hospital admission is notably associated with an increased risk of 90-day and 1-year mortality, and sBTLA is associated with decreased lymphocyte count. This suggests that sBTLA might be a useful indicator of sustained immune dysregulation, and a prognostic tool in sepsis.

Missing sample analysis.

Baseline and clinical characteristics in patients with ≤1 and ≥2 missing samples.

(PDF)

Click here for additional data file.

Survival analyses with adjustment for immunosuppression at baseline.

(PDF)

Click here for additional data file.

sBTLA on day 7 in survivors and non-survivors at 90 days and 1 year.

(A) sBTLA on day 7 in patients alive at, or dead within 90 days post hospital admission (B) sBTLA on day 7 in patients alive at or dead within 1 year post hospital admission.

(PDF)

Click here for additional data file.

(XLSX)

Click here for additional data file.

11 Oct 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Muhammad Adrish, MD, MBA, FCCP, FCCM

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Methods section, please provide additional information about the participant recruitment method and the demographic details of your participants. Please ensure you have provided sufficient details to replicate the analyses such as:

a) the recruitment date range (month and year),

b) a statement as to whether your sample can be considered representative of a larger population, and

c) a description of how participants were recruited.

3. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

4. Please include your full ethics statement in the ‘Methods’ section of your manuscript file. In your statement, please include the full name of the IRB or ethics committee who approved or waived your study, as well as whether or not you obtained informed written or verbal consent. If consent was waived for your study, please include this information in your statement as well.

5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This manuscript technically sound, and the data support the conclusions. The sample size seems acceptable .Data analysis was appropriate also it seems that all data are available. Various groups of patients with different types of infection was included in study , but the main limitation is the number of patients and it seems that we need more patients and also an study with a larger sample size,

Reviewer #2: Thank you for the invitation to review this manuscript.

The authors present data on serial sBTLA levels in a 2011-14 patient cohort admitted with acute illness due to confirmed bloodstream infection. Their principal finding was of an association between persistently elevated sBTLA levels at day 7 and mortality at 90 days and 12 months.

Comments

1. Could the authors please include a concise description of what is known about the biology of sBTLA? Are lymphocytes / specific lymphocyte subsets the origin? What differentiates cell surface bound from soluble BTLA? What is the fate and intravascular half-life of sBTLA? What effector cells does sBTLA act upon and to what effect? How is individual age known to affect the production and kinetics of sBTLA?

2. Could the authors please account for the apparent slow recruitment of 116 patients over 4 years and the 7 year gap between the close of recruitment and the submission of this paper?

3. Only 14 (13%) of patients were admitted to ICU during their acute hospital admission. Where any patients not admitted to ICU due to poor prognosis or for any other reason despite apparent indications for such an admission? What were the reasons for the ICU admission - specifically, how many were the consequence of the development of septic shock or acute severe organ(s) dysfunction due to the BSI that had precipitated hospital admission? How did ICU admission affect short and long term mortality risk?

4. I am surprised that the distribution of included patient ages is parametric, I would have expected it to be negatively skewed and reported as a median [IQR] - could the authors comment?

5. In Figure 1, the median value of sBTLA does appear to fall progressively over the 28 day period but by a very small increment compared to the range of measured values. If the authors analysed the delta sBTLA, and / or delta sBTLA as a percentage of the peak value, for each individual, does this provide a more insightful picture of the evolution of sBTLA levels? For example, percentage drop from peak value of both CRP and PCT are emerging as viable stopping criteria for antimicrobial therapy - hence the question.

6. The mortality data is fascinating, but requires more information. I presume that cause of death / death certification information is a matter of public record in Sweden? As such, could the authors please provide this data and factor it into their discussion. What does this data tell us about the apparent association between prolonged immune dysregulation and the pathology that causes apparent "premature" death? Or, does this data suggest that the index BSI itself occurred because of an occult pathology that would go on to cause death in the succeeding 12 months?

7. Given the association / correlation with widely used existing biomarkers, can the authors comment or speculate on the value of measuring sBTLA levels in addition to, or instead of CRP, for example. Furthermore, could the authors comment on the correlation between sBTLA levels and neutrophil : lymphocyte ratio, which is emerging as a more useful marker than absolute values of either?

8. Could the authors provide a figure that shows the differences in peak, day 7, and day 7 as a percentage of peak in sBTLA levels in survivors verses non-survivors at 28 days, 90 days and 12 months?

9. Could the authors please add more detail to their suggestions regarding what implications their findings should have in terms of the next questions / hypotheses research into sBTLA and other markers of immunosuppression should aim to address?

Reviewer #3: Dear Sirs,

I read both your papers with great interest.

Regarding the paper in review process: the new research actually expends your previous results, regarding sBTLA and sepsis/septic shock patients to the BSI patents. Were some of the patients in both studies? If so, please acknowledge that in the new study (same 31 patients in the control, or just a coincidence?)

I have some concerns regarding the fact that the statistics is too complex, and I think it should be validated by a statistician.

Row 93-95- in introduction do you state your previous research, or did you include a conclusion statement in introduction?

You have a lot of missing data, not only at admission time, but also during the data collecting process. How do you explain this fact?

Do you have a table with the demographic data of the controls? It would be interesting to compare their data with patients' data.

You have more than 100 patients in the patients' group and only 31 in the controls. Why did you decide do compare uneven groups?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Mohsen Rajaeinejad

Reviewer #2: Yes: Jonathan Ball

Reviewer #3: Yes: Oana Antal

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

2 Dec 2021

See Response to reviewer (attached document).

Submitted filename: Response to reviewers.docx

Click here for additional data file.

9 Mar 2022

Sustained elevation of soluble B- and T- lymphocyte attenuator predicts long-term mortality in patients with bacteremia and sepsis

PONE-D-21-08617R1

Dear Dr. Lange,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

George Vousden

Deputy Editor-in-Chief

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: (No Response)

Reviewer #3: Thank you for replying to all my concerns. I have no further issues to raise, as you replayed to all I have asked for.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Jonathan Ball

Reviewer #3: No

11 Mar 2022

PONE-D-21-08617R1

Sustained elevation of soluble B- and T- lymphocyte attenuator predicts long-term mortality in patients with bacteremia and sepsis

Dear Dr. Lange:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. George Vousden

Staff Editor

PLOS ONE