Perforin Expression and Natural Killer-Cell Proportion as Biomarkers in Secondary Hemophagocytic Lymphohistiocytosis

Introduction

Hemophagocytic lymphohistiocytosis (HLH) is a hyperinflammatory clinical syndrome characterized by excessive activation of CD8+ T lymphocytes and monocyte-macrophage systems and the production of several inflammatory factors, which are associated with complex clinical manifestations and very high mortality [1]. The etiology of HLH is broadly categorized into primary HLH (pHLH), which is caused by genetic mutations, and secondary HLH (sHLH), which is triggered by various underlying conditions, such as infections, malignancies, or autoimmune diseases. While pathologically linked, the underlying conditions can significantly influence immune profiles, including perforin expression and cytotoxic lymphocyte populations [2].

Cytotoxic lymphocytes, including natural killer (NK) cells and CD8+ T cells, are critical in the immune system’s response to infected or transformed cells. The functional impairment or dysregulation of these cells is a hallmark of HLH pathogenesis [3]. Perforin, also known as perforin-forming protein, is expressed and secreted by cytotoxic lymphocytes and leads to lytic damage to target cells through the formation of pores on target cell membranes [4]. It is essential for cytotoxic CD8 cell and NK-cell function, including the killing of transformed cells [5]. This paradigm is best understood in pHLH, in which mutations in genes such as PRF1, which encodes the perforin protein, account for approximately half of all cases [6]. This genetic defect results in a significant decrease in the expression of the perforin protein and a decrease in the killing activity of NK and cytotoxic CD8 cells. Patients with pHLH lack expression of the intracellular perforin protein in cytotoxic lymphocytes, whereas their heterozygous parents also have abnormal expression of the intracellular perforin protein [7].

However, the role of perforin in sHLH, a condition driven by overwhelming inflammatory triggers rather than by primary genetic defects, is more complex and remains debated. It is unclear whether the immune signature of sHLH mirrors the perforin deficiency of pHLH, or if, conversely, perforin is upregulated as part of a widespread, albeit ineffective, immune activation. This is particularly relevant for under-investigated populations, such as CD56+ T cells, which share characteristics of NK and T cells. Therefore, a detailed characterization of perforin expression within key cytotoxic subsets is crucial for advancing our understanding of sHLH pathophysiology.

In light of this, in the present study, we aimed to characterize the proportions of CD8+ T, NK, and the increasingly recognized CD56+ T cells, as well as their intracellular perforin expression, in patients with newly diagnosed sHLH using flow cytometry. Furthermore, we sought to determine if a correlation exists between these cellular profiles and key clinical parameters, with the hypothesis that specific immunological signatures could serve as potential biomarkers for disease activity in sHLH.

Material and Methods

STUDY DESIGN AND PARTICIPANTS:

This was a single-center, cross-sectional observational study with a longitudinal sub-cohort. Patients were prospectively enrolled between June 2016 and June 2017 at the Department of Hematology and Department of Geriatric Hematology, The First Affiliated Hospital with Nanjing Medical University (Jiangsu Province Hospital), Nanjing, China. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Ethical Committee of The First Affiliated Hospital with Nanjing Medical University (approval No. 2019-SR-446). Written informed consent was obtained from all participants or their legal guardians prior to enrollment.

PATIENT COHORT:

A total of 19 patients with newly diagnosed sHLH were included in this study, including 14 men and 5 women, with a median age of 52 (39–59) years. The diagnosis of HLH was established according to the HLH-2004 diagnostic criteria [8]. The inclusion criteria included age ≥18 years and fulfillment of 5 or more of the HLH-2004 diagnostic criteria. The exclusion criteria included age <18 years, history of cirrhosis or other severe liver disease, and pregnancy. To rigorously exclude primary forms of the disease, all patients underwent next-generation sequencing analysis, which confirmed the absence of pathogenic mutations in genes commonly associated with familial HLH.

The underlying etiologies of sHLH were determined through a comprehensive diagnostic workup, which included clinical manifestations, bone marrow smears and biopsies, imaging studies (computed tomography, positron emission tomography-computed tomography, and ultrasound), and extensive etiological tests as clinically indicated (eg, virology, microbiology, and rheumatology screenings). For patients with lymphoma-associated HLH, detailed immunophenotyping and morphological examinations were performed on samples to meticulously exclude contamination with circulating lymphoma cells from the analytical gate.

TREATMENT RESPONSE EVALUATION AND CLINICAL DATA COLLECTION:

Treatment response was evaluated based on a modification of previously reported criteria [9–12]. All patients were evaluated for therapeutic efficacy at 8 weeks after treatment. Complete response (CR) was defined as the normalization of all 6 of the following parameters: (1) blood cell counts, (2) soluble interleukin-2 receptor (sCD25) levels, (3) serum ferritin levels, (4) absence of hemophagocytosis on follow-up bone marrow aspirates, (5) triglyceride levels, and (6) level of consciousness (for patients with central nervous system involvement). Partial response (PR) was defined as an improvement of ≥25% in at least 2 laboratory or clinical parameters without meeting CR criteria. Failure to achieve PR was defined as no response (NR). Following these criteria, follow-up blood samples were collected from 6 patients who achieved CR, to assess immunological changes during remission.

A control group consisting of 10 healthy volunteers (5 men, 5 women; median age 34.5 years, range 22–61 years) was recruited from our hospital. All controls were confirmed to be free of any acute or chronic inflammatory, infectious, autoimmune, and neoplastic diseases.

Clinical and laboratory data were collected from the electronic medical records. These included: white blood cell counts, hemoglobin levels, platelet counts, alanine aminotransferase (ALT) levels, aspartate aminotransferase levels, triglyceride levels, lactate dehydrogenase levels, albumin levels, erythrocyte sedimentation rate, C-reactive protein levels, fibrinogen levels, sCD25 levels, and serum ferritin levels. Overall survival (OS) was defined as the time from the date of diagnosis until death from any cause or the last follow-up date, with the data censored on June 30, 2024.

FLOW CYTOMETRIC ANALYSIS OF INTRACELLULAR PERFORIN STAINING:

Peripheral whole blood samples were collected in EDTA-anticoagulant tubes. For each sample, 120 μL of whole blood was first surface-stained with a cocktail of the following antibodies: T-cell receptor αβ (TCRαβ)-fluorescein isothiocyanate (BD Biosciences, WT-31, #347773), CD8-PerCP (BD, SK1, #347314), and CD56-APC (BD, B159, #555518) for 20 min at room temperature. Red blood cells were then lysed for 15 min with 2 mL of Cell Lysis Buffer (BD, #555899) and washed. The resulting leukocyte pellet was then permeabilized using Fixation/Permeabilization Solution (BD, #554722) and stained with either phycoerythrin-conjugated anti-perforin (eBioscience, dG9, #12-9994-42) or a phycoerythrin-conjugated mouse IgG2b kappa isotype control (eBioscience, eBMG2b, #12-4732-42) for 30 min at room temperature. After being washed, the cells were resuspended in 1% paraformaldehyde and stored at 4°C prior to analysis by flow cytometry.

DATA ACQUISITION AND GATING STRATEGY:

We analyzed samples using a FACSCalibur flow cytometer (Beckman Coulter, Gallios, AU39633). Instrument settings and fluorescence compensation were established using single-stained controls before each batch of experiments. At least 50 000 lymphocyte events were acquired per sample to ensure robust analysis of all subsets. The perforin-positive gate was strictly set according to the isotype control staining for each sample, to minimize artifacts. Within the primary lymphocyte gate identified by forward and side scatter properties, the following gates were used to distinguish the 3 populations of interest: CD8⁺ T cells were defined as TCRαβ⁺, CD8⁺, and CD56⁻; NK cells were defined as TCRαβ⁻ and CD56⁺; and CD56⁺ T cells were defined as TCRαβ⁺ and CD56⁺. The percentage of positive cells in each defined subset was reported. Data analysis was performed using Kaluza v2.1.

STATISTICAL ANALYSIS:

Statistical analysis was performed using SPSS (version 23.0; IBM Corp, Armonk, NY, USA). The normality of data distribution was assessed using the Shapiro-Wilk test. As most quantitative data were not normally distributed, data were presented as median (interquartile range [IQR]) and were compared between independent groups using the Mann-Whitney U test. For the longitudinal analysis, the Wilcoxon signed-rank test for paired samples was used to compare pre- and post-treatment values. Correlations between variables were analyzed with Spearman rank correlation analysis. A 2-sided P value <0.05 was considered statistically significant.

Results

PATIENT CHARACTERISTICS:

A total of 19 patients with newly diagnosed sHLH were enrolled in this study. Among these patients, 11 (57.9%) had lymphoma-associated HLH (LHLH), 6 (31.6%) had infection-associated HLH (IHLH), and 2 (10.5%) had rheumatic immune-associated HLH (AHLH). A total of 7 patients achieved CR, 7 achieved PR, 4 achieved NR, and 1 was lost to follow-up and could not be evaluated. Among the 7 patients with CR, 1 patient (No. 16) improved after 4 weeks of treatment and was discharged from our hospital without further treatment. Subsequent data were obtained through telephone follow-up. The patient was evaluated as CR, but was not included in the CR group due to the lack of post-treatment blood samples.

The clinical data and the evaluation of efficacy are presented in Table 1. Kaplan-Meier survival curves of the 19 patients with sHLH are shown in Figure 1.

ALTERATIONS IN CYTOTOXIC LYMPHOCYTE SUBSETS AND PERFORIN EXPRESSION IN NEWLY DIAGNOSED SHLH AND SHLH WITH CR:

Compared with healthy controls, patients with newly diagnosed sHLH showed a significantly decreased proportion of NK cells, while no significant difference was observed in CD8+ T cells or CD56+ T cells (Figure 2A, Table 2). The proportion of perforin-positive CD56+ T cells was significantly elevated in patients with newly diagnosed sHLH, compared with in controls. However, there were no significant differences in perforin expression in CD8+ T cells and NK cells between these 2 groups (Figure 2B, Table 2).

Patients with sHLH who achieved CR exhibited a significantly lower proportion of CD8+ T cells than did patients with newly diagnosed sHLH. No significant differences were observed in the proportions of NK cells or CD56+ T cells between the 2 groups (Figure 2C, Table 2). Additionally, perforin expression in CD8+ T cells, NK cells, and CD56+ T cells showed no significant difference between sHLH patients with CR and newly diagnosed sHLH patients (Figure 2D, Table 2). Representative flow cytometry plots illustrating these differences are presented in Figure 3.

LONGITUDINAL CHANGES IN LYMPHOCYTE SUBSETS AFTER COMPLETE RESPONSE:

For the 6 patients who achieved CR, a longitudinal analysis was performed. The proportion of CD8+ T cells was significantly lower after achieving CR than at initial diagnosis (P=0.0028). No significant differences were observed in the proportions of NK cells or CD56+ T cells, nor in the perforin expression within any of the 3 cytotoxic subsets, between the 2 time points (P>0.05 for all other comparisons). The results of paired analysis are presented in Figure 4. Representative flow cytometry plots demonstrating this change for 1 patient are illustrated in Figure 5.

SUBANALYSIS OF IMMUNOLOGICAL PROFILES BY SHLH ETIOLOGY:

To investigate whether the immunological profile differed by the underlying disease trigger, a subanalysis was performed between patients with LHLH (n=11) and those with non-lymphoma-associated HLH (non-LHLH) (n=8). The proportion of NK cells was significantly lower in LHLH than in non-LHLH (P=0.0425). No statistically significant differences were found in the proportions of CD8+ T cells or CD56+ T cells, or in the perforin expression within any of the 3 cytotoxic subsets between these 2 subgroups (P>0.05 for all comparisons). Results are shown in Figure 6.

CORRELATIONS BETWEEN IMMUNOLOGICAL AND CLINICAL PARAMETERS:

In patients with newly diagnosed sHLH, correlation analysis revealed that the proportion of CD56+ T cells was positively correlated with hemoglobin levels (r=0.464, P<0.05) and negatively correlated with age (r=−0.456, P<0.05). The proportion of NK cells showed a positive correlation with fibrinogen levels (r=0.531, P<0.05) (Figure 7A). Furthermore, perforin expression in CD56+ T cells was negatively correlated with ALT levels (r=−0.516, P<0.05), while perforin expression in CD8+ T cells was positively correlated with serum sCD25 levels (r=0.498, P<0.05) (Figure 7B).

Discussion

In this study, we investigated the immunological landscape of sHLH and report 2 key findings. First, there was a selective and significant increase in intracellular perforin expression within CD56+ T cells of sHLH patients, compared with healthy controls. Second, there was a significant reduction in the proportion of peripheral NK cells. These observations challenge the simple paradigm of global perforin deficiency in HLH and suggest a more complex, subset-specific immune dysregulation in the secondary form of the disease.

Our most novel finding is the elevated perforin expression in CD56+ T cells, which is consistent with a previous report on Epstein-Barr (EBV)-associated HLH [6]. This suggests that unlike in pHLH, in which perforin production is genetically impaired [5], the cytotoxic machinery in sHLH can be intact and even hyperactivated in specific lymphocyte subsets as part of an overwhelming immune response. We acknowledge that the baseline perforin levels in our healthy control group appeared lower than those reported in some historical studies [13,14]. This variation could be attributable to technical differences, such as antibody clones, staining protocols, or flow cytometry platforms. However, since all patient and control samples in our study were processed and analyzed identically, the significant difference observed between our patient and internal control groups represents a robust and biologically relevant finding. Multiple studies have confirmed that perforin expression is corelated with age [15–17]. However, no correlation between perforin expression and age was observed in our research, which may be related to the relatively narrow age distribution and smaller sample. Although no pHLH-related gene mutations were confirmed in all sHLH patients, the possibility of undiscovered genetic predispositions cannot be fully excluded. Therefore, cohort-specific factors, such as age and genetic background, still need further study. The negative correlation we observed between perforin expression in CD56+ T cells and ALT levels further supports its clinical relevance, hinting at a potential, albeit not fully understood, regulatory role in mitigating liver injury.

We also observed a significantly reduced proportion of circulating NK cells in patients with sHLH, which is in accordance with a recent study [18]. Importantly, our subgroup analysis provided further nuance, demonstrating that this NK cell depletion was particularly severe in the LHLH subgroup. This finding likely reflects the direct effects of the lymphoma on the hematopoietic and immune systems, such as systemic immunosuppression or nutritional competition, which can impair NK cell development or survival. Lymphoma cells highly express PD-L1, which, upon binding to PD-1 on NK cells, transmits a strong inhibitory signal, leading to the exhaustion of NK cells. This might explain the large number of HLH cases related to immune checkpoint inhibitors that have been reported [19,20]. Tumor cells have an extremely high metabolic rate and consume large amounts of glucose and amino acids. However, the activation and function of NK cells are highly dependent on these nutrients. In the tumor microenvironment of LHLH, NK cells may be unable to perform their normal functions [21]. Despite this profound reduction in cell numbers, we found no decrease in perforin expression on a per-cell basis within the remaining NK cell population. Carvelli et al examined 68 cases of adult sHLH and found that the number of lymphocytes, including NK cells, was reduced; however, the expression of perforin in NK cells was almost normal, and no NK cell cytotoxic dysfunction was observed [22]. This finding pertains to a relative decrease and not necessarily a reduction in absolute cell numbers, a key distinction from many pHLH cases [23]. This proportional shift could reflect several non-mutually exclusive mechanisms, such as increased trafficking and sequestration of NK cells into inflamed tissues, such as the liver or spleen [24,25], or heightened activation-induced cell death within the hyperinflammatory cytokine milieu of sHLH [26]. Crucially, our data do not allow us to directly equate this reduced proportion with impaired overall cytotoxicity, which would require dedicated functional assays, including K562-cell direct lysis [27,28] and CD107a surface expression [29].

We found that the proportion of CD8+ T cells was significantly decreased in patients who achieved CR. CD8+ T cell was a key driver of the cytokine storm, which has been reported in COVID-19 [30], CAR-T cell-induced cytokine release syndrome [31], and acute and chronic viral infection [32]. The core of hemophagocytic syndrome treatment lies in eliminating overly activated lymphocytes, just as the key drug etoposide in the HLH-94/HLH-2004 protocol effectively reduces the number of activated T cells [33]. In successfully treated EBV-HLH cases, there is a significant decrease in cytotoxic T lymphocytes (mainly CD8+ T cells) [34]. This aligns with the understanding that successful therapy dampens the overactivated cytotoxic T cell response and suggests that dynamic monitoring of the CD8+ T-cell proportion could serve as a useful biomarker for therapeutic efficacy.

This study’s strength lies in its comprehensive, multi-parameter flow cytometric analysis of a well-characterized sHLH cohort, including longitudinal data. However, several limitations must be considered when interpreting our results. First, the small sample size limits the statistical power of our correlation and sub-group analyses and restricts the generalizability of our findings. This was evident in our subanalysis of patients with LHLH, which, while showing no difference from other sHLH etiologies, may have been underpowered to detect subtle distinctions. Second, our study lacked a non-HLH inflammatory disease control group, making it difficult to ascertain if the observed immune signature was entirely specific to sHLH or a feature of general hyperinflammation. We have taken this limitation into account and included disease control groups, such as systemic lupus erythematosus, sepsis and lymphoma (without HLH), in our subsequent larger-scale studies [35–37]. Third, as discussed, our analysis was based on lymphocyte proportions, and absolute cell counts were not determined. Fourth, we did not perform functional assays, such as NK cell degranulation or cytotoxicity assays; therefore, conclusions about cellular function based on perforin expression remain inferential. Finally, while we screened for common pHLH mutations, the possibility of rare or undiscovered genetic predispositions cannot be fully excluded. These limitations underscore the need for larger, multi-center prospective studies that incorporate functional analyses to validate our findings.

Conclusions

In this small cohort, we identified a distinct immunological signature in sHLH characterized by a reduced proportion of circulating NK cells and, concurrently, heightened perforin expression in CD56+ T cells. This profile suggests a complex pattern of immune dysregulation that differs from that of classical pHLH. While these findings require confirmation in larger, prospective studies, they indicate that longitudinal monitoring of cytotoxic lymphocyte subsets, including CD8+ T cells, could be a valuable tool for evaluating disease activity and therapeutic response in sHLH.