Comparing [³H] thymidine LPT and CFSE assay to assess lymphocyte proliferation in beryllium-exposed sarcoidosis patients

Abstract

The detection of antigen specific lymphocyte responses plays a vital role in the diagnosis of various diseases. Beryllium-specific [³H] thymidine lymphocyte proliferation test (LPT) is regarded as a gold standard in identifying chronic beryllium disease (CBD) cases. Alternatively, flow cytometric based carboxyfluorescein succinimidyl ester (CFSE) assay, has several benefits as opposed to LPT, since it further permits both phenotypical characterization and functional analysis of proliferating lymphocyte subsets.

The suitability of both LPT and CFSE assay to therefore detect beryllium sensitivity in a group of Be-exposed sarcoidosis patients with suspected beryllium exposure, was evaluated in this study. The clinical relevance of the test responses, expressed as stimulation indices (SI), were additionally compared on a group and individual level.

Agreement in clinical interpretation of the test responses between both methods was observed in 4 out of 5 recruited patients, when considering total lymphocyte population i.e., CD3⁺ and CD19⁺-cells combined, on day 7 and with CFSE-SI >1.5, when compared with LPT-SI >2.5. Variability in responses to beryllium was additionally evaluated in Be-exposed sarcoidosis patients and compared with healthy controls.

To conclude, both LPT and CFSE assay are suitable assays to detect Be sensitivity in Be-exposed sarcoidosis patients. At the same time, flow cytometric based CFSE assay has the edge over LPT in identifying the relevant proliferating lymphocyte populations. As such, when comparing two or more methods, factors that contribute to assay variability such as timepoints, lymphocyte subsets and number of replicates should always be accounted for.

1. Introduction

The in vitro assessment of antigen-induced lymphocytic proliferative responses is one of the tools to study biological processes and mechanisms that contribute towards the dysregulation of the immune system. The relative accessibility of peripheral blood mononuclear cells (PBMCs) [1,2], allows the use of this heterogenous population to evaluate antigen stimulation responses. Monitoring lymphocytes’ responses to antigen stimulation [3] is, however, still challenging due to the possible low frequency [4] of responder cells in PBMCs. Several assays have been developed to quantify the extent of stimulation to a specific antigen, including analyzing the production of cytokines by intracellular cytokine staining or enzyme-linked immunosorbent assay (ELISA), assessing the induction of activation markers by flow cytometry and assessing proliferation with the incorporation of [³H] thymidine [also known as the lymphocyte proliferation test (LPT)].

LPT remains the most commonly used method to measure lymphocyte proliferation, although it does not provide information on the phenotype [5] or specific lymphocyte subsets of proliferating cells in isolated PBMCs. Moreover, the use of radioactive material and the subsequent waste is a limitation in current laboratory operating standards.

Developments in the field of flow cytometry, has since allowed more detailed analysis based on the characterization of cell subsets, using fluorescent cell permeant, cytoplasmic proliferation dyes such as violet proliferation dye 450 (VPD-450) [1] and carboxyfluorescein diacetate succinimidyl ester (CFSE) [6]. The number and extent of cell divisions can be further enumerated since the fluorescence intensity of these dyes are divided equally per generation of daughter cells [6]. These dividing cells can also be characterized using antibodies specific for cell surface or clusters of differentiation (CD) markers. This in turn, expands the information that can be obtained from such single functional assays [1] by leveraging on the advantage of parallel lymphocyte lineage analysis [7].

With the advent of increasing alternatives proposed, comes the necessity to verify correspondence between CFSE assay and LPT. Likewise, several comparative studies have focused on investigating whether non-radioactive assays like CFSE assay, are reliable enough to quantitatively detect lymphocyte proliferation [8]. Strong and in most cases, significant correlations between both assays have been reported with mitogens, such as Concavalin A (ConA) [8], phytohemagglutinin (PHA) [[8], [9], [10], [11], [12]], pokeweed mitogen (PWM) [8], with antigens such as Bacillus Calmette–Guérin (BCG) vaccine [8], Candida albicans [8,13], autologous regulatory T cells (Tregs) [14] and with metals like beryllium [15].

Metals are known to cause a variety of pathological conditions and the inhalation of metal dust is known to cause granulomatous lung disorders, among others [16]. Increased industrial use of beryllium (Be) and subsequent Be exposure, resulted in a rise in patients suffering from beryllium sensitivity (BeS) or Chronic Beryllium Disease (CBD), a granulomatous disease caused by cell-mediated immune response to Be, mainly affecting the lung [16]. Beryllium-induced immune response is triggered when beryllium ions interact with antigen-presenting cells (APCs) expressing major histocompatibility complex (MHC) class II molecules, before being presented to T lymphocytes [17]. Hyperproliferation of CD3⁺ T cells upon Be exposure marks the basis for Beryllium Lymphocyte Proliferation Test (BeLPT) [18]. As such, workplace screening with the BeLPT has since identified workers with BeS or CBD [19].

In the clinical context, metal-specific LPT has thus far been used to diagnose CBD via BeLPT [20] or to identify nickel hypersensitivity in patients undergoing joint replacement surgery via nickel LPT (NiLPT) [21]. Similarly, Be [15] and Ni [22]-induced lymphoproliferative responses have been evaluated with CFSE assay.

Regardless of the antigen of interest, proliferation of T-cells with the use of a CD3⁺ marker [23] or its subsets, CD4⁺ (T helper cells) [3,5] and CD8⁺ (cytotoxic T cells) [1,24], has remained the cornerstone of LPT and related assays. As such, there is limited number of studies that focus on the proliferation of CD19⁺ or CD20⁺ B-cells [25,26]. Accounting for such a population of cells that make up to 10–25% of PBMC population can enrich the interpretation of lymphocyte proliferation tests [27,28].

The extent of antigen-stimulated in vitro proliferation with LPT or lymphocyte subset(s)-specific CFSE assay can be quantified and expressed as stimulation index (SI). SI is a ratio determined based on the mean proliferative response in stimulated conditions compared to the mean proliferative response in an unstimulated condition. In this regard, a positive test is identified by two or more SIs across a range of concentrations, exceeding a threshold of abnormal (SI threshold) [29].

However, SI thresholds do not account for possible assay variation observed across replicates, since only the mean of unstimulated and stimulated wells are used. Therefore, to determine a more objective cut-off, Di Blasi and colleagues, recently defined a positive response by calculating the percentage of stimulated wells with an increased proliferative response when compared to the mean of the proliferative responses in unstimulated wells +2 standard deviations (mean of control + 2SD) or more [3]. In doing so, they demonstrated the necessity to consider variability of unstimulated wells to distinguish between true positive and false positive responses.

In this study, we hence, determine the extent of concordance between SI values of CD3⁺ and CD19⁺ cells combined, CD3⁺ T cell subset or CD19⁺ B cell subset from the CFSE assay (CFSE-SI), with the SI values of the LPT (LPT-SI) in a beryllium-exposed group of sarcoidosis patients. In addition, the proliferative response of PBMCs from healthy controls and beryllium-exposed patients is compared across the two cut-off methods, SI threshold [20,30] and mean of control + 2SD [3].

2. Materials and methods

2.1. Human blood samples

Whole blood (40–50 ml) was obtained from 5 patients who were recruited based on their occupational history, which was suggestive for past or ongoing beryllium exposure (Table 1), and 4 healthy control subjects, upon informed consent. Subjects were recruited via the outpatient clinic for occupational and environmental diseases of the University Hospitals Leuven (Belgium). The diagnosis of sarcoidosis had been established according to the criteria of the American Thoracic Society/European Respiratory Society [31]. The study was approved by the Ethics Committee Research UZ/KU Leuven (S61777).

Table 1.

Patient characteristics.

Patients	Patient 1	Patient 2	Patient 3	Patient 4	Patient 5
Industry	Metal recycling	Metal machining	Metal recycling	Chemical industry	Automotive parts production
Years of exposure	1/2 y	40 y	36 y	31 y	3 y
Smoking status	Never smoked	Never smoked	Never smoked	Never smoked	Former smoker
Age at blood acquisition	38 y	59 y	57 y	56 y	45 y
Immunosuppressive medication	Methotrexate 15 mg/week	None	None	None	None
Initial diagnosis	Sarcoidosis	Sarcoidosis	Sarcoidosis	Sarcoidosis	Sarcoidosis
Healthy Controls	Control 1	Control 2	Control 3	Control 4
Industry	Desk-bound	Desk-bound	Desk-bound	Desk-bound
Years of exposure	0	0	0	0
Smoking status	Never smoked	Never smoked	Never smoked	Never smoked
Pack years (PY)	0	0	0	0
Age at blood acquisition	58 y	42 y	35 y	39 y
Immunosuppressive medication	None	None	None	None
Diagnosis	No	No	No	No

2.2. Compounds and chemicals

2.2.1. Pokeweed mitogen (PWM)

0.5 μg/ml of PWM was prepared from a stock solution (prepared in-house) of 500 μg/ml by diluting in Roswell Park Memorial Institute (RPMI) 1640 medium (ref. 52400025) supplemented with 10% fetal bovine serum (FBS).

2.2.2. Beryllium sulfate tetrahydrate (BeSO₄.4H₂O)

29.9 mg of BeSO₄ (purchased from ThermoFisher Scientific [ref. 16104]) was diluted in 12 ml of distilled H₂O (dH₂O) to prepare a stock solution of 2.49 mg/ml (1 x 10⁻² M). The stock solution was serially diluted by ten-folds to prepare a starting concentration of 1 x 10⁻⁴ M.

2.3. Culture and lymphoproliferative evaluation of stimulated PBMCs

2.3.1. Isolation of PBMCs

Human PBMCs were isolated by Lymphoprep™ (STEMCELL) density gradient centrifugation. Whole blood was first diluted 1:2 with phosphate-buffered saline (PBS) and poured into a 12 ml Leucosep™ tube (Greiner Bio-One) containing 3 ml of Lymphoprep™ at a ratio of 1.5:1 and centrifuged at 20 °C for 15 min at 400 g, with centrifuge brakes off. PBMCs were removed carefully and washed 3 times with PBS for 10 min at 400 g, 20 °C. Viability and cell counting of freshly acquired PBMCs was performed with trypan blue exclusion assay. A final PBMC concentration of 1 x 10⁶ cells/ml was then prepared by resuspending cells in serum-free PBS.

2.3.1.1. PBMCs incubation

1.8 x 10⁵ PBMCs/well were incubated in RPMI 1640 medium with 5% autologous plasma, 1% penicillin-streptomycin (P–S) (100 U/ml), 1% l-glutamine (2 mM) and 1% fungizone (2.5 g/ml) in U-bottomed 96-well culture plates at 37 °C in a 5% CO₂ environment. These cells were stimulated for 7 days by several concentrations of BeSO₄ at 1:10 dilution: 10⁻⁴, 10⁻⁵, 10⁻⁶, 10⁻⁷,10⁻⁸, 10⁻⁹ and 10⁻¹⁰ M. Positive control (stimulated with pokeweed mitogen [PWM]) and negative control (unstimulated condition supplemented with RPMI media) were consistently tested across all samples. All stimulation conditions were performed as three to four replicates.

2.3.2. Lymphocyte proliferation via tritiated thymidine test (classic lymphocyte proliferation test)

Blood samples drawn on the same day, were simultaneously sent to the Central Laboratory of the University Hospitals Leuven, Belgium, where LPT was performed according to the protocol described in the ATS official statement on diagnosis and management of beryllium sensitivity and chronic beryllium disease [32].

On day 7, 50 μl (1:50 diluted) of [³H] thymidine was added per well. Cells were collected 6 h later on glass-fiber filters using a cell harvester (PerkinElmer Filtermate Cell Harvester). The incorporated radioactivity was measured using a β-counter (PerkinElmer Tricarb 2910 TR Liquid Scintillator Analyzer), to determine extent of cell proliferation. Lymphocyte proliferation results are expressed as Stimulation Index (SI) which is calculated by dividing the mean counts/min (cpm) in stimulated conditions by the mean cpm in unstimulated control wells. [³H] thymidine LPT derived SI (LPT-SI) of >2.5 was considered as an abnormal or positive response [33].

2.3.3. Lymphocyte proliferation via CFSE assay

Fluorescein isothiocyanate (FITC)-Carboxyfluorescein succinimidyl ester (CFSE; CellTrace™ CFSE proliferation kit, ref. C34554, ThermoFisher Scientific) labelling of PBMCs was performed at a 1:1 ratio with a final concentration of CFSE of 2.5 μg/ml and allowed to incubate at room temperature (RT) for 8 min while protected from light. CFSE labelled PBMCs were then washed twice before incubation. In case PBMC count was insufficient, experiments were still performed in triplicates with the omission of lower concentrations of BeSO₄ (10⁻⁷ and 10⁻⁸ M).

Following the incubation period, experiments were halted on the 6th and 7th day and the cell suspension was stained with FVD-780 for 30 min at RT. Cell suspension was then resuspended in an antibody cocktail with combinations of Alexa Fluor 700 (AF700) anti-CD3 (UCHT1) and anti-CD19 (HIB19). This suspension was incubated for 25 min at 4 °C while protected from light. Following the washing step, cell suspension was then resuspended in 1% paraformaldehyde (PFA) and incubated for 20 min at 4 °C while protected from light. After 20 min, cell suspension was washed in PBS and resuspended in 100 μl of 0.05% BSA in PBS and processed within 1–3 days.

The median fluorescence intensity (MFI) levels of CFSE labelled PBMCs were measured using BD LSR Fortessa and twenty thousand events were collected. Data acquired was analyzed using FlowJo software. Compensation matrix was always performed for each experiment on FlowJo, prior to gating strategy (see supplementary methodology, Fig. S1).

CFSE assay derived SI (CFSE-SI) was calculated based on the percentage of the cells that have proliferated in a stimulated condition divided by the percentage of proliferated cells in the unstimulated condition, on Days 6 and 7. To determine the CFSE-SI threshold, mean SI + two standard deviations [2SD, [20,34]] and SI + three SDs [3SD, [30]] of healthy controls across all beryllium concentrations, were calculated. As such, a CFSE-SI threshold of 1.5–2, was established for CD3⁺ and CD19⁺ combined subpopulations.

2.3.4. Measurement of cellular viability

Cellular viability of PBMC cultures were determined by fixable viability dye (FVD-780) in the total population (%) of lymphocytes that were determined by side scatter (SSC-A)/forward scatter (FSC-A) gatings using FlowJo software (FlowJo™ v10.4). A viability threshold of 75% (with reference to control) was applied across all concentrations [35].

2.4. Statistical analysis

Statistical analysis was performed using GraphPad Prism 8.4.3. One-way and two-way ANOVA were performed to determine the statistical significance of the differences between concentrations of each exposure condition, and between LPT-SI and CFSE-SI values, respectively. Spearman's correlation coefficients were determined for correlation analyses. P-values <0.05 were considered statistically significant, *.

In line with the study of Di Blasi and colleagues' 2020 [3] the percentage of wells with a positive response was calculated. For each individual included, the unstimulated condition was considered as control value and a mean and standard deviation (SD) were calculated for the different replicates. For each concentration of the stimulated conditions, the percentage of wells (replicates) with an outcome higher than the mean unstimulated condition (control) plus two SD (2SD) was calculated. According to Di Blasi and colleagues’ proposal, a response in a stimulated condition is thus denoted positive based on an increased percentage (>0%) of wells exhibiting proliferative response values higher than unstimulated conditions [3].

3. Results

3.1. Association between LPT-derived values and CFSE assay derived values in PWM stimulated lymphocytes

Viablilty of the ungated lymphocyte population of PBMCs from patients was initially assessed by FVD-780, at the end of exposure period to beryllium (BeSO₄). The viability of Be-exposed cells remained above the threshold at any beryllium concentration applied, with no significant change in viablity observed between concentrations (Fig. S2).

Recruited individuals’ cells were stimulated with positive control, pokeweed mitogen (PWM), and evaluated with [³H] thymidine LPT and CFSE assay. The PWM-stimulated LPT-SI values were compared with PWM-stimulated CFSE-SI values based on the extent of proliferation observed in T- (CD3⁺) and B- (CD19⁺) cells combined on day 7. A table of CFSE-SI values of each participant at two timepoints (day 6 and 7), with respect to CD3⁺ only, and CD3⁺ and CD19⁺ combined is included in the supplementary material (Table S1).

No significant differences were noted between the PWM-stimulated LPT-SI and PWM-stimulated CFSE-SI values per individual, despite the differences in SI values observed between the 2 methods (Fig. 1). PWM-stimulated SI values were always >10, for all patients regardless of the method used, confirming the efficacy of and concordance between both tests.

Fig. 1.

Comparing lymphoproliferative capacity of positive control, PWM-stimulated cells based on LPT and CFSE assay derived SI values.

The sensitivity of the two tests, LPT and CFSE assay was evaluated by comparing LPT-SI with CD3⁺ and CD19⁺ combined CFSE-SI values on day 7, across 5 patients that were stimulated with positive control, PWM. Legend: LPT: lymphocyte proliferation test, CFSE: Carboxyfluorescein diacetate succinimidyl ester assay, PWM: pokeweed mitogen.

3.2. Determining the association between LPT-SI and CFSE-SI and identifying trends based on SI threshold in beryllium stimulated lymphocytes

Individual LPT-SI and CFSE-SI values were also calculated per concentration of BeSO₄ (see Table S1), across 5 beryllium exposed (Be-exposed) sarcoidosis patients.

To determine a clinically relevant positive test response, two or more SI values across the concentrations per method, must be above the individually calculated SI threshold for beryllium response, in order to be considered positive (++). Additionally, if only one or none of the concentrations are above the threshold, it is then regarded as a borderline (+) and negative test response (−), respectively.

SI values derived across Be-exposed patients were compared on day 7, across a maximum of 7 and 5 concentrations of BeSO₄ with [³H] thymidine LPT and CFSE assay, respectively. Assessing the similarities between LPT-SI and CD3⁺ and CD19⁺ combined CFSE-SI values on day 7 was specifically selected, due to the similar timepoint of comparison with positive control, PWM. No significant differences between both LPT-SI and CFSE-SI values were noted in all patients (Fig. 2A).

Fig. 2.

(A–C): Comparing lymphoproliferative capacity of beryllium-stimulated cells based on LPT-SI and CFSE-SI.

The number of patients with a positive, borderline, and negative test response per timepoint and per lymphocyte subset(s) of interest, were then compiled to identify for similarities across the two tests on a population or group level (Fig. 2B) and on an individual level (Fig. 2C). Additional comparison of individual CFSE-SI test responses with LPT-SI test responses at timepoints, day 6 and 7, with consideration of CD3⁺ only as well as CD3⁺ and CD19⁺ combined subsets are included in supplementary (Table S2). LPT-SI >2.5 was specifically compared with CFSE-SI > 1.5, 2 and 2.5, since LPT-SI >2.5 is the current threshold applied in BeLPT routine testing.

On a group level, exact concordance in patients’ test responses derived from both [³H] thymidine LPT and CFSE assay, was observed with CD3⁺ and CD19⁺ combined, on days 6 and 7, when a CFSE-SI >2 threshold was applied (Fig. 2B). However, when test responses per individual was compared on day 6, with LPT-SI > 2.5 and CFSE-SI >2 with CD3⁺ and CD19⁺ combined, agreement was only observed within patient 4.

On an individual level, the highest number of concurrences within individuals was instead noted on day 7 and with consideration of CD3⁺ and CD19⁺ combined subsets, wherein LPT-SI test responses >2.5 corresponded with the CFSE-SI test responses of 4 (patients 1, 2, 3 and 4), 3 (patients 2, 3 and 4) and 2 (patients 3 and 4) patients per respective CFSE-SI threshold of 1.5, 2 and 2.5 (Fig. 2C).

Individual LPT-SI and CD3⁺ and CD19⁺ combined CFSE-SI values on day 7 were compared in Be-stimulated cells of patients at relevant concentrations from 10⁻¹⁰ to 10⁻⁴ M (A). The number of patients with a positive, borderline and negative test response based on LPT-SI >2.5 and CFSE-SI thresholds 1.5, 2 and 2.5 were determined on a group level (B) and on an individual level (C). “Green” indicates perfect agreement between LPT-SI and CFSE-SI test responses; “Yellow” indicates high levels of agreement between LPT-SI and CFSE-SI test responses. Two-way ANOVA was performed to determine statistical significance between LPT-SI and CFSE-SI values. Legend: P: patient, LPT: lymphocyte proliferation test, CFSE: Carboxyfluorescein diacetate succinimidyl ester assay, SI: stimulation index, ++: positive, +: borderline, -: negative.

3.3. Assessing proliferation of T and B cells subsets with a second cut-off: mean of control + 2SD

To determine a cut-off that is specific to CFSE assay and is capable of distinguishing healthy controls from exposed patients; the percentage of wells with a proliferative response above the mean of proliferative response in the unstimulated replicates (Mean of control) + two times standard deviation (SD) was calculated – indicated as M2SD [3].

Among the healthy controls, most wells (replicates) did not show levels of proliferation that are higher than M2SD at all concentrations of BeSO₄ when evaluating CD3⁺ and CD19⁺ combined (Fig. 3A) or CD3⁺ and CD19⁺ individually (Fig. 3B & C). Considering the patients, all except for patient 3 (with limited number of concentrations used) and 4, had in most wells a response higher than M2SD at all concentrations of BeSO₄ (Fig. 3D–F).

Fig. 3.

(A–F): Selection of cut-off based on mean of control + 2SD in Be-stimulated cells of healthy controls and patients, on day 7.

With the CD19⁺ only subset, a similar trend was observed between healthy controls where 1 to 3 healthy controls had 0% positive wells i.e., lower than M2SD, for BeSO₄ concentrations 10⁻⁸ to 10⁻⁵ M, while all healthy controls had 0% positive wells at the highest concentration of 10⁻⁴ M (Fig. 3C). In patients, lower concentrations of BeSO₄ elicited levels of proliferation higher than M2SD across most wells, as opposed to the higher concentrations of 10⁻⁵ and 10⁻⁴ M (Fig. 3C). Additional comparison of M2SD between healthy controls and sarcoidosis patients on day 6 are included in supplementary (Fig. S3).

In general, we observed variability in the response to beryllium in both sarcoidosis patients and healthy controls. While Di Blasi and colleagues’ propose that a response in a stimulated condition is denoted positive based on wells exhibiting proliferative response values higher than M2SD [3], we still observed some wells with increased proliferation in samples of control subjects. Thus, we propose to impose a higher threshold of M2SD at 50%.

As a result, for both CD3⁺ and CD19⁺ combined and CD3⁺ only subset(s), only 2 out of 5 patients had a positive response at all concentrations (patients 2 and 5). In the case of CD19⁺, only patient 3 had a positive response at all concentrations of BeSO₄, while no clear trend was noted across the rest of the patients.

Overall, patients 1,2,3 and 5 had ≥50% of positive wells with at least 1 concentration of BeSO₄, when considering the CD3⁺ and CD19⁺ combined and CD3⁺ only subset(s) (Fig. 3D and E).

Percentages of positive wells in unstimulated and BeSO₄ concentrations of 10⁻⁸ to 10⁻⁴ M that are above mean of control + 2SD in healthy controls (A-C) and patients (D-F) were compared based on CD3⁺ and CD19⁺ combined (A and D), CD3⁺ (B and E) and CD19⁺ (C and F) CFSE assay values on day 7. Additionally implemented percentage threshold at ≥50% is denoted by the dotted line across all figures. Legend: US- Unstimulated, CFSE: Carboxyfluorescein diacetate succinimidyl ester assay.

4. Discussion

The accurate detection of antigen-stimulated lymphocyte responses plays a vital role in the diagnosis of diseases. A strong association between selection of relevant antigen to determine proliferative capacity and disease presentation is in fact a requisite. For instance, a positive test response with [³H] thymidine LPT has mostly been reported across diseased patients with clear evidence of direct association between a singular causative agent and respective occupational exposure history, such as beryllium-exposed workers in the context of CBD [29]. Although LPT has proven its use, it provides only a limited amount of information. More recently, other methods have been introduced, such as CFSE assay, that allows the identification and characterization of proliferating cells. In this study, we therefore chose to verify the reproducibility of results between two methods-the radioactive [³H] thymidine LPT and flow-cytometric CFSE assay, by focusing on a group of sarcoidosis patients with a suspicion of occupational Be-exposure, based on the provided job histories. The association of CFSE assay derived proliferative responses observed at two timepoints and across three lymphocyte subsets encompassing T and B cells (CD3⁺ and CD19⁺ combined), T cells only (CD3⁺) or B cells only (CD19⁺) with [³H] thymidine LPT, was additionally addressed.

Firstly, we chose to assess the sensitivity of the two methods in response to a positive control like pokeweed mitogen (PWM), on day 7 and in consideration of CD3⁺ and CD19⁺ combined. PWM, a plant-derived polysaccharide or lectin, is known for its potent immunostimulatory properties in, in vitro T and B cell assays [36]. Day 7 was also selected as the optimal timepoint at which maximal levels of PWM-stimulated proliferation would be observed, based on our preliminary observations in healthy controls (data not shown). No significant differences in LPT-SI and CFSE-SI within individuals was noted, despite the discordance in SI values reported between both methods.

Subsequently, we compared 4 healthy controls (with no know history of Be-exposure) with 5 sarcoidosis patients that were suspected of occupational Be-exposure, by calculating the SI values derived from LPT and CFSE assay. Thus far, there is no concrete evidence on why peak levels of proliferation should be expected on the same day with both [³H] thymidine LPT and CFSE assay. Therefore, we focused on evaluating CFSE assay derived lymphoproliferative responses on both days 6 and 7 before comparing with LPT-SI values. Due to a lack of pre-defined CFSE-SI threshold for Be-stimulation, a range of CFSE-SI thresholds of 1.5–2.5 were implemented and cross-compared with a validated and biologically relevant BeLPT-SI threshold of 2.5 [33].

When the extent of proliferation observed on a group level was explored, CFSE-SI >2 in the Be-exposed group on days 6 and 7 with CD3⁺ and CD19⁺ combined, was proposed as a viable and comparable SI threshold with LPT-SI >2.5. On an individual level however, similar test responses between [³H] thymidine LPT and CFSE assay was only observed in patient 4 on day 6 with CFSE-SI >2, when considering CD3⁺ and CD19⁺ combined. However, concordance between the two methods based on observations made on an individual level is improved if CD3⁺ and CD19⁺ combined and a CFSE-SI threshold >1.5 or 2 on day 7, is considered instead. With these alternate conditions, analogy in 4 and 3 out of 5 patients respectively, rather than 1 out of 5 patients when considering CD3⁺ and CD19⁺ combined with CFSE-SI >2 on day 6, was noted. The proposed CFSE-SI thresholds based on an individual level assessment, at which the highest number of patients are observed, also coincides with our laboratory determined estimation of CFSE-SI thresholds of 1.5 and 2.

At the same time, this observation contests the reliability of identifying positive responders based on a cut-off such as SI threshold, that is conventionally determined on a group level which is also then rarely re-validated on an individual level. Ways to determine SI thresholds also vary across studies such as the use of mean + 3 times SD in unexposed subjects [30], 2 times SD of differences between stimulated and unstimulated values [20], use of receiver operating characteristic (ROC) curves [37], and lastly and more frequently with an empirically selected cut-off between 1.5 and 3 [3].

Recently, Di Blasi and colleagues’ 2020 [3] suggested to evaluate proliferation based on the number/percentage of positive responding wells (value higher than M2SD compared to unstimulated). They found that their selected cut-off was suitable to detect lymphocyte responses in healthy vaccinated donors, when stimulated with hemagglutinin (HA) peptide pool [3]. By relying on M2SD, we noticed that 3 out of 5 Be-exposed patients (patients 1, 2 and 5) had a high number of positive wells, in line with the derived CFSE-SI values. The positive response, with at least 3 BeSO₄ concentrations was obtained when considering CD3⁺ only and CD3⁺ and CD19⁺ combined subsets. Hence, we postulate that the proliferative function of CD3⁺ T cells have an active role in contributing towards the overall proliferative response of CD3⁺ and CD19⁺ combined in Be-exposed patients. This concurs with previous studies that identified that CD3⁺ T cells mediate beryllium-specific immune responses in CBD [23,38]. CFSE assay therefore has a distinct advantage over [³H] thymidine LPT as it allowed us to identify which lymphocyte subsets are stimulated and proliferating to a larger extent, in response to beryllium; an aspect that is in fact lost with the use of LPT.

With reference to Di Blasi and colleagues’ original definitions of positive response higher than M2SD, we also noticed that some of the healthy controls in our study, had >0% positive wells at certain concentrations of the stimulated conditions. Based on these observations, the faint response noted in healthy controls is probably a reflection of the variability in the model used with a relatively low number of cells. Thus, we suggest to introduce an additional threshold at 50%, to distinguish between responders in CFSE assay. Based on this newly implemented percentage threshold, we postulate that all but one sarcoidosis patient could have either a borderline (1 concentration above cut-off) or positive test response (≥2 concentrations above cut-off). Specifically with this cut-off, patients 1, 2 and 5 reportedly have positive test responses, patient 3 had a borderline test response while patient 4 had a negative test response.

In the current context, the use of M2SD as a second cut-off, identified and validated that a CFSE-SI threshold >1.5 would be most appropriate to elucidate for the same set of patients with similar test responses, when compared with LPT-SI threshold >2.5. Therefore, we postulate that a CFSE-SI threshold >1.5 can serve as an analogous alternative to LPT-SI threshold >2.5 in our group of Be-exposed sarcoidosis patients.

4.1. Strengths and limitations of our study

The initial objective of this study was to compare the applicability of the two most preferred lymphocyte proliferation tests: [³H] thymidine LPT and CFSE assay, to detect lymphoproliferative responses to beryllium in a group of sarcoidosis patients. We noted similarities in the extent of proliferation observed with [³H] thymidine LPT and CFSE assay specifically on day 7 when considering CD3⁺ and CD19⁺ combined population with the latter method. In the process of determining the extent of concordance between the two methods, we however recognized that SI as a cut-off does not account for assay variability. To address this limitation, we therefore applied a second cut-off- M2SD, based on CFSE assay-derived values and compared it with the original LPT-SI threshold. Additionally, we addressed how the observations made with either of the methods and/or either of the cut-offs can be translated in the clinical context. We recognize that it is generally advised for lymphocyte proliferation tests to be performed on two separate occasions before interpreting results, especially in cases for which a borderline result was observed. However, repeat analyses was not performed in any of these patients, which could limit the accuracy of our current deductions. Additionally, we only chose to compare CFSE-SI against LPT-SI values on days 6 and 7, due to the methods’ similarity in length of exposure and incubation period. It would, however, be more ideal if CFSE-SI values were calculated at 24-h intervals leading up to the 7th day, in order to accurately identify an optimal timepoint for peak CFSE-SI values to be observed with BeSO₄ stimulation. At the same time, inclusion of more replicates is necessary to improve our understanding on the variability observed across stimulated conditions. In line with that, lower PBMC count observed in 1 of the patients, also prevented an accurate clinical interpretation of test responses, since only 3 concentrations were tested for (patient 3). At the current standpoint, this patient has a borderline test response, and this observation would be more precise if both methods were performed across 4–7 concentrations instead. It is also necessary to note that [³H] thymidine LPT was not performed within healthy controls, resulting in a lack of direct comparison to be made with LPT-SI values between healthy controls and sarcoidosis patients. Despite these setbacks, similarity between both methods was noted independently in 3 out of 5 patients when LPT-SI >2.5 was compared with CFSE-SI >1.5 and with ≥50% positive wells above M2SD.

Due to practical and administrative constraints we were not able to perform additional tests with the use of blood cells of confirmed BeLPT positive individuals. In addition, we could not use Be-exposed but not diseased control samples. Both additions would add significant information to the presented experiments.

In general, we would also like to point out that only a small percentage of T cells amidst the large repertoire of lymphocytes will proliferate in response to any particular antigen [39]. Moreover, in each sample PBMCs, varying population of proliferating responder T cells will be present which can potentially result in significant variation in proliferation observed per cell culture in a 96 well plate (41). In line with this consideration, no clear dose-dependent relationship, as noted in LPT-SI and CFSE-SI values with increasing BeSO₄ concentrations, is obtained. Accounting for contributors to assay variability such as timepoints, lymphocyte subsets and even number of replicates should therefore, always be considered when comparing two or more methods. In addition, it is beneficial to consider the proliferative capacity of both T and B cells combined as well as estimate the variability in responding cells across wells, to avoid misinterpretation or inadequacy in the number of true positive responders reported within a patient cohort. In due course, overall proliferative responses of lymphocytes i.e., T cells, B cells and even natural killer (NK) cells, as well as attention to clinical interpretation of results per individual, should also be prioritized.

Although we did attempt to holistically address several pitfalls in this area of research, we however noticed that there are still some knowledge gaps that ought to be also addressed in future studies.

Knowledge gaps to address.

• 1)Would we always expect an equally strong concordance or correlation between both methods if we focused instead on the use of inorganic antigens like metals and silica, as opposed to positive controls?
• 2)Should we always assume that LPT values need to be compared with CD3⁺ CFSE-stained T cells only or consider the combination of CFSE-stained CD3⁺ T and CD19⁺ B cells?
• 3)Would it be feasible to test for overall lymphocyte proliferative responses with the use of a lymphocyte marker like CD45, when comparing LPT with flow-cytometric alternatives?
• 4)How can we also address clinical interpretation beyond the confirmation that both methods are comparable or in correspondence?

5. Conclusion

The results described in this study, highlight the suitability of both [³H] thymidine LPT and CFSE assay in detecting Be-stimulated responses in a group of sarcoidosis patients. Agreement in clinical interpretation of the test responses tabulated across both methods, was observed within 4 out of 5 recruited patients, when considering total lymphocyte population or CD3⁺ and CD19⁺ combined and on day 7 with a CFSE-SI threshold >1.5.

Additionally, we also identified the advantages of relying on a flow cytometric based CFSE assay instead of LPT, since it permits both phenotypical characterization and functional analysis of proliferating lymphocyte subsets.

CFSE assay is therefore a reliable alternative to LPT, due to the similar outcomes and extent of variability observed with both methods in response to a stimulus like Be.

Author contribution statement

Nirosha Ganesan: Steven Ronsmans: Peter Hoet: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

This research was funded by KU Leuven Internal Funding C2 (C24/18/085).

Data availability statement

Data included in article/supp. material/referenced in article.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: