Germline defects of familial haemophagocytic lymphohistiocytosis—Related genes may represent a predisposing factor for mature T‐ and natural killer‐cell lymphoma

Summary

Peripheral T‐cell lymphoma (PTCL) is relatively prevalent in Asian populations. Previous studies suggest that germline mutations in familial haemophagocytic lymphohistiocytosis (FHL)‐related genes may predispose individuals to lymphoproliferative disorders. To investigate the underlying molecular mechanisms, we analysed paired tumour and germline deoxyribonucleic acid from 74 patients with T‐ and natural killer‐cell lymphomas. Germline variants in FHL‐related genes (UNC13D, PRF1, STXBP2, STX11, SH2D1A and XIAP) were assessed by whole‐exome sequencing, while somatic mutations were analysed by targeted sequencing. A total of 21 germline mutations in FHL‐related genes were detected in 14 of 74 patients (18.9%), including mutations in UNC13D (N = 11), STXBP2 (N = 6), PRF1 (N = 3) and STX11 (N = 1). The most frequent mutation was UNC13D c.2588G>A (p.G863D), which was significantly enriched in PTCL patients compared to the general Chinese Han population (allele frequency: 4.7% vs. 0.7%, OR = 6.785, p = 0.002). In line with established PTCL mutation profiles, somatic mutations were frequently detected in TET2, RHOA, DNMT3A and IDH2. Patients with FHL‐related germline mutations exhibited a trend towards better overall survival. In conclusion, germline mutations in FHL‐related genes, particularly UNC13D, may contribute to PTCL susceptibility in Chinese patients and are associated with clonal somatic mutations.

INTRODUCTION

Peripheral T‐cell lymphoma (PTCL) is a heterogeneous group of aggressive non‐Hodgkin lymphomas (NHLs) derived from mature T and natural killer (NK) cells. It accounts for 25%–30% of all NHL cases in Asia, compared to 5%–10% in Western countries. ¹ , ² , ³ Most types of aggressive PTCLs exhibit an aggressive clinical course and poor prognosis, with long‐term survival of around 30%–40%. ³ , ⁴ , ⁵ The molecular mechanisms underlying PTCL are complex and diverse, with common alterations observed in genes involved in epigenetic regulation (TET2, DNMT3A, IDH2), T‐cell receptor signalling pathways (CD28, VAV1, PLCG1, FYN), as well as mutations in RHOA associated with T‐follicular helper (TFH) cell phenotype. ⁶ , ⁷ , ⁸ Despite these findings, the specific molecular drivers and genetic susceptibility factors that predispose individuals to PTCL remain incompletely understood, presenting a challenge for early intervention, precision diagnosis and targeted therapies.

Familial haemophagocytic lymphohistiocytosis (FHL) is a rare life‐threatening immune dysregulation characterized by marked hyperinflammation from uncontrolled activation of T cells and macrophages. Germline mutations in genes involved in perforin–granzyme‐mediated cytotoxicity such as PRF1, UNC13D, STX11 and STXBP2 are known to cause FHL. ⁹ Deficiencies in these genes compromise cytolytic granule exocytosis at the immunological synapse, which impairs the cytotoxic activity of T/NK cells. ¹⁰ , ¹¹ Emerging studies and case reports have highlighted a potential relationship between defects of FHL‐related genes and lymphoma predisposition. In the study conducted by Chen et al., variant UNC13D c.2588G>A was reported to be a founder mutation for lymphoma in the Chinese population. ¹² The carrier frequency of UNC13D c.2588G>A mutation in lymphoma patients was significantly higher than in the healthy controls (10% vs. 1.9%), suggesting that haploinsufficiency of this gene may predispose patients to lymphomas. Moreover, in the study conducted by Guan et al., germline mutations in UNC13D, LYST and PRF1 were reported to be linked to EBV‐associated T/NK‐cell lymphoproliferative diseases in Chinese patients. ¹³ Interestingly, our group recently reported a unique group of three patients with germline mutations of UNC13D and STX11 genes who presented as adult‐onset PTCL with cytotoxic T‐cell phenotype. ¹⁴

Based on this emerging evidence, germline defects of FLH‐related genes may represent a novel predisposing factor for PTCLs. In this study, we investigated the potential role of germline mutations in FHL‐related genes in conferring susceptibility to PTCL in the Chinese population. Both the somatic tumour mutations and germline mutations of FHL‐related genes were analysed using paired samples in a cohort of patients with PTCL.

MATERIALS AND METHODS

Patient enrolment

This study enrolled 74 patients diagnosed with PTCL and treated at Peking Union Medical College Hospital (PUMCH) between January 2020 and December 2023. Diagnoses were made based on the 2016 World Health Organization classification. ¹ The study was approved by the institutional review board of PUMCH, and all samples were collected with informed consent.

DNA extraction

Formalin‐fixed, paraffin‐embedded (FFPE) tumour tissue samples and peripheral blood were collected from each patient prior to treatment. Germline mutations were analysed using whole‐exome sequencing (WES) on peripheral blood samples, while somatic mutations in tumour samples were analysed using a custom panel of 413 lymphoma‐associated genes (Oncolym Panel, Geneplus‐Beijing, Beijing, China) targeted next‐generation sequencing (NGS) panel. Genomic deoxyribonucleic acid (gDNA) from FFPE tissues and peripheral blood was extracted with the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany). DNA concentration was measured with a Qubit fluorometer and the Qubit dsDNA HS Assay Kit (Invitrogen, Carlsbad, CA, USA). The plasma DNA size distribution was assessed using an Agilent 2100 BioAnalyzer (Agilent Technologies, Santa Clara, CA, USA).

Library construction and sequencing

Libraries were prepared from 400 to 800 ng of gDNA extracted from tumour tissue samples and 200 ng from blood samples using the KAPA DNA Library Preparation Kit (Kapa Biosystems, Boston, MA, USA). For WES, whole‐exome capture was performed with the Target Cap® Core Exome Panel v3.0. For targeted NGS of tumour gDNA, the capture probe was designed based on ~1.5 Mb genomic regions of 413 genes frequently mutated in lymphoma with the SeqCap EZ Library (Roche NimbleGen, Madison, WI, USA). gDNA libraries for WES were sequenced on an MGISEQ T7 system, while gDNA libraries for targeted NGS were sequenced on a Geneplus Seq‐2000 instrument (Geneplus‐Beijing, China).

Data processing and analyses

Reads were aligned to the hg19 reference genome assembly using the Burrows‐Wheeler Aligner (BWA) aligner (version 0.7.17‐r1188). ¹⁵ Single‐nucleotide variants (SNVs) and insertions/deletions (indels) were called using the GATK4 HaplotypeCaller (version 4.1.4.01). ¹⁶ For the WES data, transcripts of the FHL‐related genes, including UNC13D, PRF1, STXBP2, STX11, SH2D1A and XIAP, were analysed. SNVs and indels were annotated using the Exome Aggregation Consortium (ExAC) and 1000 Genomes databases. Variants with minor allele frequencies >1% in the 1000 Genomes and/or ExAC databases were regarded and filtered as SNPs rather than mutations. Somatic SNVs and indels in hotspot and non‐hotspot regions were reported with supporting reads of ≥4 and ≥8 respectively. The variant allele frequency (VAF) detection threshold was set at 1.0%.

Statistical analysis

Chi‐squared test and Fisher's exact test were used to analyse categorical data, while the independent t‐test was applied for comparing continuous data between groups. Progression‐free survival (PFS) was calculated from the date of diagnosis to the date of disease progression, death from any cause or last follow‐up. Overall survival (OS) was measured from the date of diagnosis to the date of death from any cause, with censoring at the last follow‐up. Survival analyses for PFS and OS were performed using the Kaplan–Meier method, and survival rate comparisons between groups were made using the log‐rank test. Cox regression analysis was conducted to identify risk factors for OS. A p‐value of <0.05 was considered statistically significant. All data analyses were performed using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism version 9.0 (GraphPad Software, La Jolla, CA, USA).

RESULTS

Patient characteristics of the study cohort

The baseline characteristics of the patients are summarized in Table 1. The median age at diagnosis was 55 years (range: 17–80), with a male‐to‐female ratio of 2.1:1. At diagnosis, 85.1% (63/74) of the patients presented with advanced‐stage disease. Elevated LDH levels were observed in 64.9% of patients, and positive plasma EBV‐DNA was detected in 29.7%. The histological subtypes included angioimmunoblastic T‐cell lymphoma (AITL, N = 21), nodal TFH‐cell lymphoma, not otherwise specified (nTFHL‐NOS, N = 5), PTCL, not otherwise specified (PTCL‐NOS, N = 29), extranodal NK/T‐cell lymphoma (ENKTL, N = 10), monomorphic epitheliotropic intestinal T‐cell lymphoma (MEITL, N = 5) and anaplastic large‐cell lymphoma (ALCL, N = 4).

TABLE 1.

Clinical characteristics of the patients in the two groups.

Characteristic	Entire cohort (N = 74)	Pts with FHL‐related mutations (N = 14)	Pts without FHL‐related mutations (N = 60)	p‐value
Age, years
Median (range)	55 (17–80)	42 (19–79)	57 (17–80)	0.273
>60	29 (39.2%)	3 (21.4%)	26 (43.3%)	0.131
Sex, male	50 (67.5%)	12 (85.7%)	38 (63.3%)	0.107
Ann Arbor stage III/IV	63 (85.1%)	12 (85.7%)	51 (85.0%)	0.946
Elevated LDH level	48 (64.9%)	8 (57.1%)	40 (66.7%)	0.501
EBV‐DNA ≥500 copies/mL	22 (29.7%)	5 (35.7%)	17 (28.3%)	0.586
Presentation of HLH	5 (6.8%)	2 (14.3%)	3 (5.0%)	0.213
Past history of autoimmune conditions	3 (4.1%)	0 (0.0%)	3 (5.0%)	0.393
Tumour history in first‐degree family members	9 (12.2%)	2 (14.3%)	7 (11.6%)	0.787
Histological subtypes
PTCL‐NOS	29 (39.2%)	7 (50.0%)	22 (36.7%)	0.357
nTFHL	26 (35.1%)	3 (21.4%)	23 (38.3%)	0.233
AITL	21 (28.4%)	2 (14.3%)	19 (31.7%)
nTFHL‐NOS	5 (6.8%)	1 (7.1%)	4 (6.7%)
ENKTL	10 (13.5%)	3 (21.4%)	7 (11.7%)	0.336
MEITL	5 (6.8%)	0 (0%)	5 (8.3%)	0.263
ALCL	4 (5.4%)	1 (7.1%)	3 (5.0%)	0.576
Treatment regimens
CHOP/CHOEP	8 (10.8%)	1 (7.1%)	7 (11.7%)	0.624
Chidamide + CHOEP	3 (4.1%)	1 (7.1%)	2 (3.3%)	0.515
Azacitidine + chidamide + CHOP	40 (54.1%)	5 (35.7%)	35 (58.3%)	0.126
Brentuximab vedotin + CHP	3 (4.1%)	1 (7.1%)	2 (3.3%)	0.515
GDP‐ML/SMILE	8 (10.8%)	3 (21.4%)	5 (8.3%)	0.155
Bendamustine	5 (8.1%)	2 (14.3%)	3 (5.0%)	0.247
Others	15 (20.3%)	1 (14.3%)	14 (23.3%)	0.175

Abbreviations: AITL, angioimmunoblastic T‐cell lymphoma; ALCL, anaplastic large cell lymphoma; CHOEP, cyclophosphamide, doxorubicin, vincristine, etoposide and prednisone; CHOP, cyclophosphamide, doxorubicin, vincristine and prednisone; DNA, deoxyribonucleic acid; EBV, Epstein–Barr virus; ENKTL, extranodal natural killer/T‐cell lymphoma; FHL, familial haemophagocytic lymphohistiocytosis; GDP‐ML, gemcitabine, cisplatin, dexamethasone, methotrexate and pegaspargase; HLH, haemophagocytic lymphohistiocytosis; LDH, lactate dehydrogenase; MEITL, monomorphic epitheliotropic intestinal T‐cell lymphoma; nTFHL, nodal peripheral T‐cell lymphoma with TFH phenotype; PTCL‐NOS, peripheral T‐cell lymphoma, not otherwise specified; SMILE, dexamethasone, methotrexate, ifosfamide, L‐asparaginase and etoposide; TFH, T‐follicular helper.

A total of 14 patients were identified with germline mutations in FHL‐related genes. No significant differences were observed in baseline characteristics between patients with or without FHL‐related germline mutations. Among the 14 patients carrying germline mutations in FHL‐related genes, none had a known history of FHL or autoimmune disease, nor did they have documented autoimmune diseases or recurrent infections prior to the lymphoma diagnosis. All 14 patients carrying FHL‐related germline mutations were clinically asymptomatic with respect to immune dysfunction prior to lymphoma onset. Five patients developed haemophagocytic lymphohistiocytosis (HLH) syndrome at the time of lymphoma diagnosis, including two patients with FHL‐related germline mutations and three patients without FHL‐related germline mutations. Based on the available medical records, a family history of malignancy in first‐degree relatives was documented in 2/14 (14.3%) patients with FHL‐related germline mutations and in 7/60 (11.6%) patients without such mutations. There was no statistically significant difference in tumour family history between the two groups.

Owing to the differences in pathological subtypes, treatment regimens of this cohort were relatively heterogeneous. Most patients with PTCL‐NOS received cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP)‐based chemotherapy, whereas those with extranodal NK/T‐cell lymphoma were treated with regimens containing methotrexate and asparaginase. Patients with ALCL predominantly received therapies incorporating brentuximab vedotin. No significant differences in treatment modalities were observed between the patients with or without FHL‐related germline mutations.

Germline mutations of FHL‐related genes in PTCL

In WES, the median deduplicated sequence depth was 137.2. Among the 74 patients, a total of 21 germline mutations in FHL‐related genes were identified in 14 unrelated patients (18.9%) (Figure 1A). These included 11 mutations in the UNC13D gene, six in the STXBP2 gene, three in the PRF1 gene and one in the STX11 gene (Figure 2A). The schematic locations of mutations in the UNC13D, STXBP2, PRF1 and STX11 genes are illustrated in Figure 1B–E. The primary type of mutation in these FHL‐related genes was missense mutations (16/21), with the remaining mutations including three indels, one nonsense mutation and one splicing mutation (Figure 2A).

FIGURE 1.

Distribution of germline mutation in familial haemophagocytic lymphohistiocytosis (FHL)‐related genes. (A) FHL‐related genes were identified in 14 unrelated patients. Light blue indicates samples having a single UNC13D mutation and dark blue indicates samples having multiple UNC13D mutations. Light red indicates samples having a single STXBP11 mutation and dark red indicates samples having multiple STXBP11 mutations. Schematic illustrations of the locations of mutations in UNC13D (B), STXBP2 (C), STX11 (D) and PRF1 (E) are shown. AITL, angioimmunoblastic T‐cell lymphoma; ALCL, anaplastic large cell lymphoma; ENKTL, extranodal natural killer/T‐cell lymphoma; nTFHL, nodal peripheral T‐cell lymphoma with TFH phenotype; PTCL‐NOS, peripheral T‐cell lymphoma, not otherwise specified; TFH, T‐follicular helper.

FIGURE 2.

Germline mutation in familial haemophagocytic lymphohistiocytosis (FHL)‐related genes of patients with PTCL. (A) The variant counts and types of mutations in UNC13D, STXBP2, PRF1 and STX11 genes, along with the carrier frequency of patients with these FHL‐related mutations. (B‐D) Pie chart depicting the proportions of patients with or without germline FHL‐related mutations across different pathological subtypes of PTCL‐NOS (B), nTFHL (C), and ENKTL (D). (E) Enrichment of germline UNC13D p.G863D mutations in PTCL. Allele frequencies of UNC13D p.G863D are shown in PTCL cohort from our study, as well as from Chinese Han population, East Asian, South Asian and the global populations from the gnomAD dataset. *p < 0.05, **p < 0.01, ***p < 0.001. ENKTL, extranodal natural killer/T‐cell lymphoma; nTFHL, nodal peripheral T‐cell lymphoma with TFH phenotype; PTCL‐NOS, peripheral T‐cell lymphoma, not otherwise specified; TFH, T‐follicular helper.

Eight patients (10.8%) carried mutations in UNC13D, including five with heterozygous mutations, one with a homozygous mutation and two with compound heterozygous mutations. Five patients (6.8%) carried STXBP2 mutations, comprising four with heterozygous mutations and one with a compound heterozygous mutation. Additionally, three patients carried heterozygous mutations in PRF1, and one patient carried a heterozygous mutation in STX11. No mutations were detected in SH2D1A or XIAP. Germline FHL‐related mutations were identified in 7 of 29 patients (24.1%) with PTCL‐NOS, 3 of 26 patients (11.5%) with nTFHL and 3 of 10 patients (30.0%) with ENKTL (Figure 2B–D).

Individuals with germline UNC13D mutations are highly susceptible to PTCL

The UNC13D c.2588G>A (p.G863D) mutation was the most frequent mutation identified in this cohort, observed in six patients (8.1%) with an allele frequency of 4.7%, including one homozygous and five heterozygous cases. The UNC13D p.G863D variant (rs140184929) is classified as a variant of uncertain significance for FHL type 3 according to ACMG guidelines in ClinVar. In silico analysis suggests that the amino acid position of UNC13D p.G863D was predicted to be probably damaging and deleterious in PolyPhen‐2 and SIFT respectively.

According to the gnomAD dataset, the global minor allele frequency (MAF) of UNC13D p.G863D is 1.7 × 10⁻⁴ (N = 152 295) in the general population, and the variant is more prevalent in the East Asian population (MAF = 0.5%, N = 5146) (Figure 1E). Data from the 1000 Genomes Project also confirm that the p.G863D allele is predominantly found in East Asians, with frequencies of 0.7% (N = 413) in the Chinese Han population and 0.5% (N = 208) in the Japanese population. This variant was not observed in 1006 individuals from Europe or 694 individuals of European ancestry in the America. A chi‐squared test revealed that the allele frequency of the UNC13D p.G863D in PTCL patients was significantly higher than in the Chinese Han population (4.7% vs. 0.7%, OR = 6.785, 95% CI: 1.731–26.594, p = 0.002), indicating a notable enrichment of the UNC13D p.G863D allele in PTCL patients.

Somatic tumour mutations of PTCL

Using NGS with a 413‐gene panel on tumour samples, we identified a total of 475 somatic mutations (range: 0–26; mean: 6.2) in 71 of 74 (95.9%) patients with PTCL. Two patients showed negative results, and one patient was excluded due to insufficient gDNA for library construction. The mean deduplicated sequencing depth was 1910 reads (range: 546–4160), with a mean VAF of 0.14. The mutational landscape of both the somatic mutations and paired germline FHL‐related gene mutations is depicted in Figure 3, along with essential patient characteristics.

FIGURE 3.

Mutational profile of both the germline and somatic mutations of the 74 patients with PTCL. Each column represents one case and each row represents one gene. Only genes that were mutated in more than two cases are shown in the somatic oncoplot. AITL, angioimmunoblastic T‐cell lymphoma; DNA, deoxyribonucleic acid; EBV, Epstein–Barr virus; ENKTL, extranodal natural killer/T‐cell lymphoma; nTFHL, nodal peripheral T‐cell lymphoma with TFH phenotype; PTCL‐NOS, peripheral T‐cell lymphoma, not otherwise specified; TFH, T‐follicular helper.

The most frequently mutated genes are primarily involved in epigenetic modification, including TET2 (24/74, 32.4%), RHOA (15/74, 20.3%), DNMT3A (12/74, 16.2%), IDH2 (9/74, 12.2%) and JAK3 (9/74, 12.2%) (Figure 4B). Mutations in TET2 and DNMT3A were distributed across various loci, with no clear hotspots (Figure 4C,D). In contrast, all IDH2 mutations occurred at the hotspot p.R172G/S/K, and most RHOA mutations were located at the c.50G>T (p.G17V) hotspot (Figure 4E,F). These results suggest that additional somatic mutations might be associated with PTCL genesis, as well as germline predisposition.

FIGURE 4.

Somatic mutations of patients with peripheral T‐cell lymphoma. (A) The variant count and types of the top 10 somatic mutations, along with the carrier frequency of patients harbouring these mutations. (B–E) Schematic representations of the locations of somatic mutation in TET2 (B), DNMT3A (C), RHOA (D) and IDH2 (E).

For patients with germline mutations in FHL‐related genes, fewer somatic mutations were detected compared to those without germline mutations (4.2 per sample vs. 7.0 per sample, p = 0.019). The top five most frequently mutated genes in patients without germline FHL‐related mutations were TET2 (20/60, 33.3%), RHOA (13/60, 21.7%), DNMT3A (12/60, 20.0%), IDH2 (9/60, 15.0%) and JAK3 (9/60, 15.0%), which were consistent with that in the entire cohort. In contrast, among patients with germline FHL‐related mutations, the top five most frequently mutated genes were TET2 (4/14, 28.6%), MGAM (3/14, 21.4%), RHOA (2/14, 14.3%), FAS (2/14, 14.3%) and LRP1B (2/14, 14.3%).

Prognostic value of the germline and somatic mutations

As of December 2024, the median follow‐up time was 31.5 months (range: 1.2–96.0). The OS for different subtypes of mature T‐ and NK‐cell lymphomas is illustrated in Figure 5. The 3‐year OS rates for PTCL‐NOS, nTFHL (including nTFHL‐NOS and AITL) and ENKTL were 59.7%, 54.0% and 57.1% respectively. ALCL demonstrated the highest 3‐year OS of 66.7%, whereas MEITL had the poorest 3‐year OS of 30.0%, with a median OS of 10 months (Figure 5A).

FIGURE 5.

Survival analysis of the patients with different pathological subtypes and mutation status. (A) Overall survival of patients with different pathological subtypes. (B) Overall survival of patients with mature T‐ and natural killer (NK)‐cell lymphomas in this cohort, stratified by the presence or absence of germline mutations in FHL‐related genes. (C) Overall survival of patients with PTCL‐NOS, stratified by the presence or absence of germline mutations in FHL‐related genes. (D) Overall survival of patients with or without somatic TET2 mutations. ALCL, anaplastic large cell lymphoma; ENKTL, extranodal NK/T‐cell lymphoma; FHL, familial haemophagocytic lymphohistiocytosis; MEITL, monomorphic epitheliotropic intestinal T‐cell lymphoma; nTFHL, nodal peripheral T‐cell lymphoma with TFH phenotype; PTCL‐NOS, peripheral T‐cell lymphoma, not otherwise specified; TFH, T‐follicular helper.

Patients with mature T‐ and NK‐cell lymphoma who harboured germline mutations in FHL‐related genes exhibited a trend towards improved survival compared to those without these mutations (Figure 5B,C), though the difference did not reach statistical significance. Additionally, no significant differences in survival were observed between patients with or without somatic TET2, DNMT3A, RHOA or IDH2 mutations (Figure 5D).

DISCUSSION

The prevalence of PTCL is higher in Asia than in Western countries, suggesting potential ethnic or geographic predispositions to these malignancies. In recent years, accumulating evidence has highlighted the molecular mechanisms of PTCL, characterized by epigenetic dysregulation with mutations in TET2, DNMT3A and IDH2. ⁶ , ⁷ However, germline molecular susceptibility factors in PTCL lymphomagenesis remain largely unexplored. Germline homozygous mutations in PRF1, UNC13D, STX11 and STXBP2 are responsible for FHL types 2–5 respectively, ⁹ with PRF1 and UNC13D being the most commonly mutated genes in adolescence‐ or adulthood‐onset FHL. ¹⁷ , ¹⁸ Emerging studies and case reports have suggested a potential link between defects in FHL‐related genes and lymphoma predisposition. Germline defects of UNC13D and STX11 genes were detected in three patients with PTCL in our centre. ¹⁴ In this study, we investigated the potential role of germline variants in FHL‐related genes in conferring susceptibility to mature T‐ and NK‐cell lymphoma.

Our findings revealed that 14/74 (18.9%) of the patients carried germline mutations in at least one of these genes, supporting a possible predisposing effect of FHL gene defects in PTCL lymphomagenesis. Notably, the UNC13D c.2588G>A (p.G863D) variant was identified in 8.1% of the PTCL patients with an allele frequency of 4.7%. This variant was significantly enriched in PTCL patients compared to the Chinese Han population, suggesting that it may represent a founder variant contributing to PTCL susceptibility in East Asian populations. These results align with earlier studies, including those by Chen et al., Guan et al. and Löfstedt, which pointed to a strong association between germline defects in FHL‐related genes and lymphoproliferative diseases. ¹² , ¹³ , ¹⁹ , ²⁰ From the perspective of population‐based screening, the UNC13D c.2588G>A variant may serve as a biomarker for genetic screening, especially among high‐risk populations. First‐degree relatives of mutation carriers could undergo genetic testing to determine the carrier status, which may facilitate early tumour monitoring. However, these strategies will require further validation in larger cohorts and prospective studies to assess their clinical utility and cost‐effectiveness.

Interestingly, the prevalence of germline FHL‐related mutations appeared to differ among PTCL subtypes in our cohort, with higher frequencies observed in PTCL‐NOS (24.1%) and ENKTL (30%) compared to nTFHL (11.5%). This discrepancy may reflect underlying differences in the tumour cell of origin and dominant oncogenic mechanisms across subtypes. PTCL‐NOS and ENKTL are often derived from cytotoxic T cells or NK cells, in which impaired cytolytic function caused by germline mutations in FHL‐related genes could directly contribute to lymphomagenesis. In contrast, AITL and other TFH‐derived lymphomas are typically driven by recurrent somatic mutations in epigenetic regulators such as TET2, DNMT3A and IDH2, as well as the RHOA G17V mutation. These mutations may act independently of cytotoxic pathways, which may explain the lower incidence of germline defects in FHL‐related genes observed in this subgroup. Further studies in larger, subtype‐specific cohorts are warranted to validate these observations and clarify their biological significance.

The role of germline FHL‐related gene mutations in the pathogenesis of PTCL remains incompletely understood. FHL‐related genes encode proteins critical for perforin–granzyme‐mediated cytotoxicity. Haploinsufficiency or loss‐of‐function variants in these genes can impair T/NK‐cell cytotoxic function, potentially compromising immune surveillance. This functional impairment may permit the survival and expansion of aberrant lymphoid clones, allowing the stepwise accumulation of additional somatic driver mutations. Recurrent somatic mutations affecting epigenetic regulators, T‐cell receptor signalling components and key signal transduction pathways are well‐established contributors to PTCL pathogenesis. Notably, in our cohort, patients with germline FHL‐related mutations exhibited significantly fewer somatic mutations compared to those without such mutations. This observation supports a ‘two‐hit’ model, in which inherited immune dysfunction may impair tumour surveillance or alter the lymphoid microenvironment, thereby facilitating clonal evolution upon acquisition of critical somatic drivers. Moreover, the landscape of somatic mutations differed between the two groups. In patients with germline FHL‐related mutations, somatic variants were more frequently enriched in immune‐related and apoptotic pathways and were less frequently involved in canonical epigenetic regulators. These findings suggest distinct molecular mechanisms underlying lymphomagenesis in patients with germline predisposition. Further mechanistic studies are warranted to validate these hypothesis.

In terms of prognosis, patients with FHL‐related germline mutations exhibited a trend towards better survival. However, this difference did not reach statistical significance, potentially due to the limited sample size. The observed survival difference may reflect distinct tumour biology in this specific cohort. First, while defective cytotoxic function predisposes individuals to lymphomagenesis, partial or heterozygous defects may create a unique tumour–host interaction, where persistent low‐grade immune activation paradoxically maintains some level of immune surveillance over tumour cells. Second, patients with FHL‐related germline variants accumulated fewer additional somatic alterations compared to those without germline variants (4.2 per sample vs. 7.0 per sample). Third, a higher proportion of localized stage ENKTL, which is associated with a relatively favourable prognosis, may have contributed to the survival advantage observed in patients with FHL‐related germline mutations. Nevertheless, larger multi‐centre cohorts with extended follow‐up periods are needed to better define the prognostic significance of these germline mutations.

Despite these insights, our study has several limitations. First, due to the rarity of T‐ and NK‐cell lymphomas, our sample size is relatively small. Large‐scale multicentre and database studies are needed to validate our findings. Second, we relied on predictive in silico tools (e.g. SIFT, PolyPhen‐2) to assess the potential impact of the UNC13D c.2588G>A variant. Future studies should incorporate in vitro and in vivo functional assays to elucidate the precise mechanisms by which this mutation impairs cytotoxic function. Third, given the heterogeneity of PTCL subtypes, certain germline variants may preferentially predispose individuals to specific disease subsets. Larger studies focusing on subtype‐specific patterns could help clarify these associations. Finally, the follow‐up duration in this cohort was relatively short, and long‐term outcomes, including tumour prognosis, risk of secondary malignancies and OS, warrant validation in larger studies with extended follow‐up periods.

In conclusion, by integrating germline genetic analysis with somatic mutational profiling, our findings support the hypothesis that germline variants in FHL‐related genes, particularly UNC13D c.2588G>A (p.G863D), may confer increased susceptibility to PTCL in East Asian populations. This study also highlights the complex interplay between inherited innate immunodeficiency and acquired driver mutations in PTCL. Findings in this study underscore the need for comprehensive genetic screening and personalized treatment approaches, aiming ultimately to improve patient outcomes in these aggressive lymphomas.

AUTHOR CONTRIBUTIONS

CW, WZ and DBZ designed the study. CW, DQZ, YZ, WZ and DBZ analysed all the data. CW wrote the main manuscript. All authors approved the final manuscript.

FUNDING INFORMATION

National High Level Hospital Clinical Research Funding (2022‐PUMCH‐B‐134). Natural Science Foundation of Beijing Municipality (No. 7244384). National High Level Hospital Clinical Research Funding (2022‐PUMCH‐C‐056).

CONFLICT OF INTEREST STATEMENT

The authors declare no potential conflicts of interest.

ETHICS STATEMENT

The study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of each participating centre.

CONSENT TO PARTICIPATE

The requirement for informed consent was waived because of the use of anonymized data.

Wei C, Zhang Y, Zhao D, Zhang W, Zhou D. Germline defects of familial haemophagocytic lymphohistiocytosis—Related genes may represent a predisposing factor for mature T‐ and natural killer‐cell lymphoma. Br J Haematol. 2025;207(3):842–850. 10.1111/bjh.20231

DATA AVAILABILITY STATEMENT

The data generated in this study are available upon request from the corresponding author.