Unfolding the enigma of familial Hodgkin lymphoma: Current insights

Abstract

Hodgkin lymphoma (HL) is a heterogenous lymphoproliferative disorder of B-cell origin and represents one of the most common malignancies in children and young adults. In addition to well-known underlying factors - such as Epstein-Barr virus infection - the familial aggregation demonstrated in large population studies suggested a genetic predisposition. First-degree relatives of patients with HL have an approximately threefold increased risk of developing the disease compared to the general population. These observations have recently prompted several whole-genome studies in affected families, identifying variants possibly implicated in lymphomagenesis, including alterations in DICER1 (a member of the ribonuclease III family), POT1 (protection of telomeres 1), KDR (kinase insert domain receptor), KLHDC8B (kelch domain-containing protein 8B), PAX5 (paired box protein 5), GATA3 (GATA binding protein 3), IRF7 (interferon regulatory factor 7), EEF2KMT (eukaryotic elongation factor 2 lysine methyltransferase), and POLR1E (RNA polymerase I subunit E). In this article, we review current insights into the etiopathogenesis and risks of familial HL, and present case reports involving two sisters diagnosed with HL nearly 17 years apart. Recognizing the risk for first-degree relatives may potentially increase awareness of early symptoms among family members of HL patients, leading to earlier diagnosis and better outcomes. Conversely, understanding that the hereditary risk, though higher than in the general population, remains relatively low may provide reassurance for affected families.

Key Words: Hodgkin lymphoma; Familial Hodgkin lymphoma; Genetic analysis; Whole genome sequencing; Pathogenic variants

Core Tip: Hodgkin lymphoma (HL), although rare, is one of the most common pediatric malignancies. Familial clustering points to genetic susceptibility, especially in young-onset cases. Recently, whole-genome sequencing studies of affected families have identified several predisposing genetic variants, contributing to unraveling the complex etiopathogenesis of HL. New insights into tumor biology and the identification of germline mutations may facilitate genetic counselling, raising awareness among first-degree relatives of HL patients about potential symptoms and leading to earlier diagnosis and better outcomes. Emphasizing that the hereditary risk of HL remains low can provide significant psychological relief for affected families.

INTRODUCTION

Hodgkin lymphoma (HL) is a biologically and pathologically heterogenous malignancy arising from the clonal proliferation of aberrant B lymphocytes[1]. According to the World Health Organization Classification of Hematolymphoid Tumors (5^th edition), HL can be divided into two distinct immunohistochemical variants: Classic HL (cHL) and nodular lymphocyte-predominant HL[2]. HL is considered a rare malignancy, with a global age-standardized incidence rate estimated at 0.98 per 100000 individuals. Higher incidence rates are reported in high-income countries; these are largely attributed to greater exposure to environmental risk factors and comorbidities, such as metabolic syndrome, obesity, and hypertension. In contrast, a lack of diagnostic resources in low-income regions contributes to underdiagnosis and underreporting, while mortality rates remain disproportionately high[3].

Although rare, HL is among the most common pediatric malignancies and is the most frequent cancer in adolescents aged 15-19[4]. Moreover, a rising trend of HL among younger individuals and females has recently been observed, likely linked to a rise in westernized lifestyles and metabolic risk factors within these groups[3]. Despite the increasing incidence, survival rates continue to improve, driven by advances in early diagnosis, non-invasive staging techniques, and new treatment modalities[3,5,6]. Brentuximab vedotin (BV) and immune checkpoint inhibitors are increasingly being adopted as front-line therapies, while novel targeted agents and cellular therapies are emerging as promising options for patients with relapsed or refractory HL[6]. Although current five-year survival rates exceed 90%, late-treatment-related toxicities, secondary malignancies, and decreased life expectancies remain major concerns for HL survivors[7-10]. The adoption of risk-based and response-adapted treatment strategies provides an optimal balance between maintaining favorable outcomes and minimizing long-term sequalae, which may persist or occur even decades after the end of treatment. A deeper understanding of the underlying biology of HL - particularly the role of hereditary factors - may help identify new therapeutic subgroups, paving the way for targeted therapies and appropriate surveillance strategies for individuals at risk.

FAMILIAL HL IN TWO SISTERS

In March 2008, an 18-year-old female (Patient A) was admitted to the Department of Pediatrics at a Croatian tertiary care center due to a firm, rubbery, painless mass in the left supraclavicular region. Ultrasound revealed a 46 mm × 22 mm × 23 mm lesion suspicious for a lymphoproliferative disorder. The mass had first been noted by the patient two months prior, but had grown rapidly during the two weeks preceding admission. She also reported several months of fatigue, recurrent unexplained fever, and night sweats. The patient was previously healthy, a first-year medical student, a non-smoker, and lived in good socioeconomic and hygienic conditions in an urban area. Her 12-year-old sister was healthy, and there was no family history of malignancy among first-degree relatives or any known cases of lymphoma in the extended family. On examination, she was in good general condition (Karnofsky score: 80%), with neck lymph nodes enlarged bilaterally along the sternocleidomastoid muscles, each measuring up to 2 cm in diameter, and a liver and spleen marginally palpable below the costal margin. Laboratory tests were unremarkable, except for an elevated erythrocyte sedimentation rate. Serological testing confirmed past Epstein-Barr virus (EBV) infection. Histopathological examination of the excised left supraclavicular lymph node established the diagnosis of cHL, with the subtype of mixed cellularity. In situ hybridization for EBV was negative. Positron emission tomography-computed tomography (PET-CT) revealed increased fluorodeoxyglucose uptake in the carinal, pulmonary hilar, interaortocaval, and para-aortic lymph nodes (Figure 1A-C), consistent with stage IIIB. The patient received six cycles of doxorubicin, bleomycin, vinblastine, and dacarbazine chemotherapy, which were complicated by severe nausea, vomiting, and myelosuppression. Following the first three cycles, PET-CT confirmed complete remission (Figure 1D). At present, the patient is a healthy 35-year-old woman with no late-treatment-related sequelae. She is the mother of a healthy two-year-old girl and a two-week-old boy. She works in the field of pediatric hematology and oncology.

Figure 1 Patient A’s initial positron emission tomography-computed tomography scan. A-C: Demonstrating increased fluorodeoxyglucose accumulation in the carinal, pulmonary hilar, left interaortocaval, and para-aortic lymph nodes. Increased metabolic activity is also evident in the left supraclavicular region, corresponding to the site of lymph node excision; D: After the first three cycles of doxorubicin, bleomycin, vinblastine, and dacarbazine chemotherapy, showing no evidence of pathological metabolic activity. Diffuse, reactive osteomedullary fluorodeoxyglucose uptake is present.

In November 2024, 16.5 years after the diagnosis of patient A, her younger sister - now a 28-year-old woman (Patient B) - underwent a chest X-ray due to an acute-onset cough and fever, initially suggestive of pneumonia. Chest X-ray revealed significant widening of the upper mediastinum, a nodular lesion in the right lower lobe of the lung (32 mm × 23 mm), and bilateral pleural effusion. The patient reported several months of fatigue, loss of appetite, and night sweats, which she attributed to work-related stress. She was employed in an office setting, lived in adequate socioeconomic and hygienic conditions, was a smoker, was physically active, and owned a cat. Her medical history was notable for infectious mononucleosis, which was serologically confirmed in her mid-twenties. At admission, she was febrile, with no palpable lymph nodes, liver, or spleen, but presented with bilaterally diminished breath sounds and distant, muffled heart sounds. Laboratory tests revealed elevated inflammatory markers - an erythrocyte sedimentation rate of 52 mm/hour (normal: < 20 mm/hour) and a C-reactive protein level of 92.6 mg/L (normal: < 5 mg/L) - while other parameters were within normal limits. Echocardiography showed a 13 mm pericardial effusion, for which corticosteroid therapy was promptly initiated. Histopathological and immunohistochemical analysis of mediastinal lymphoid tissue, obtained via CT-guided biopsy, confirmed the diagnosis of cHL. Due to the limited tissue sample, the histological subtype could not be determined. Initial CT staging revealed lymphoma involving the anterior superior and middle mediastinum (88 mm × 97 mm × 105 mm) and the anterior inferior mediastinum (47 mm × 14 mm) (Figure 2). Additionally, peribronchial infiltrates were observed in the right middle and lower pulmonary lobes, consistent with the primary diagnosis. No evidence of malignancy was observed below the diaphragm, and bone marrow biopsy ruled out marrow infiltration. The patient was classified as stage IIEB and commenced treatment according to the BV, etoposide, cyclophosphamide, doxorubicin, dacarbazine, and dexamethasone protocol. Treatment was complicated by episodes of febrile neutropenia and mucositis. After two cycles of chemotherapy, PET-CT confirmed remission. The patient completed a total of four cycles and is currently three months post-treatment and under close follow-up; she is experiencing mild peripheral neuropathy, which is attributed to BV.

Figure 2 Patient B’s initial chest computed tomography scan demonstrating a bulky mass in the anterior mediastinum.

Given the fact that in these two cases, exposure to the same environmental carcinogen and an underlying disease predisposing to HL can be excluded, it is likely that the sisters harbor a genetic polymorphism that increases their susceptibility to lymphoma. Young age at onset and a favorable treatment response, as observed in our patients, are features frequently reported in familial HL. As patient B has only recently completed treatment, and patient A has recently undergone pregnancy and delivery, germline testing has not been performed at the time of the manuscript submission. However, both sisters plan to undergo genetic testing in the near future.

DISCUSSION

Although the overall lifetime risk of developing HL is relatively low, it remains one of the most common malignancies in children and young adults[4]. Despite extensive research, the etiology of HL remains largely enigmatic. Most studies have focused on dysregulated cell-mediated and humoral immune responses to viral infections, primarily EBV, as the most plausible trigger. It is postulated that chronic antigenic stimulation by EBV mutates gene expression in malignant cells. Serological studies have shown altered immune responses to EBV in individuals subsequently diagnosed with HL, including elevated antibody titers against viral capsid antigens and early antigens[11,12]. Immunohistochemical analyses of affected tissues have demonstrated the presence of EBV in neoplastic Reed-Sternberg cells, which express clonal latent viral genes - findings that strongly support the role of EBV in HL pathogenesis[12]. In addition, patients who have undergone bone marrow transplantation (BMT) have a sixfold higher risk of developing HL compared to the general population[13]. This association was particularly evident in recipients of allogeneic BMT, although several HL cases have also been reported following autologous BMT[13-15]. In allogeneic BMT recipients, EBV-positive tumors and graft-versus-host disease were present in the majority of secondary HL cases. These findings further support the hypothesis that dysregulated cell-mediated immunity and exposure to EBV are key contributors to the development of HL[13].

A significantly increased standardized incidence ratio for all cHL subtypes has also been observed in individuals with various autoimmune conditions, including autoimmune hemolytic anemia, immune thrombocytopenia, sarcoidosis, systemic lupus erythematosus, polyarteritis nodosa, Sjögren’s syndrome, and rheumatoid arthritis, with a slightly stronger positive correlation detected in men. Possible mechanisms underlying this association may involve the overstimulation and defective apoptosis of B lymphocytes, chronic inflammation, the use of immunosuppressants, and shared genetic alterations, which may lead to a predisposition to both autoimmunity and lymphomagenesis[16]. Furthermore, a small proportion of HL cases are associated with rare cancer predisposition syndromes, such as ataxia-telangiectasia and other primary immunodeficiencies[17,18]. However, given the fact that the EBV genome is detected in only 20%-40% of neoplastic tissues, and that most patients do not have an underlying autoimmune disorder, the etiology of HL, similar to that of many other malignancies, is thought to arise from a complex and multifactorial interaction between genetic and environmental factors[12,19,20]. The characteristic bimodal age distribution of HL, with one incidence peak in adolescence/young adulthood and another in adults over the age of 55, suggests that different pathogenic mechanisms may be involved depending on the age of onset, with hereditary factors playing a more prominent role at a younger age[21].

The first reports of familial HL date back to the end of the 19^th century[22]. However, observations of familial clustering became more frequent in the early 1970s, primarily through case reports involving multiple affected family members[23-27]. Early findings indicated that HL tended to occur in sex-concordant siblings under the age of 45, suggesting that the disease might be triggered by environmental factors - presumably the transmission of an infectious agent through close and prolonged contact between genetically susceptible individuals[25,27]. A landmark study published in 1995 investigated 187 pairs of dizygotic and 179 pairs of monozygotic twins affected by HL. Over an average follow-up period of 14 years, no concordant HL cases were observed among dizygotic twins, whereas 10 pairs of monozygotic twins - all under the age of 50 (mean age: 25.5 years) - were concordant for the disease. These findings indicated a much stronger genetic influence on HL pathophysiology, particularly in cases with onset during young adulthood[28].

Over the years, the rarity of familial HL and a lack of adequate analytical methods have hampered research on genetic susceptibility, contributing to wide variations in risk estimates for family members and leaving potential causative genetic variants unidentified. In 2015, a large Nordic study (Denmark, Finland, Iceland, Norway, and Sweden) provided robust statistical estimates of familial HL risk using data from national registries. The study reported an overall cumulative risk of 0.6% for HL among first-degree relatives of affected individuals, representing a threefold increase compared to the general population. The risk was higher among siblings compared to parent-offspring pairs, and it was highest in families with multiple affected members and in same-sex twins. These findings suggested a recessive inheritance pattern, while the observed gender concordance pointed to pseudoautosomal regions of the sex chromosomes as potential loci for HL susceptibility genes. Furthermore, familial risk was highest for the lymphocyte-rich cHL subtype and for early-onset disease occurring before the age of 30[29]. Since the early 2000s, increasing evidence has emerged linking specific genetic polymorphisms to lymphomagenesis, with particular emphasis on variants promoting the dysregulation of immune response and B-cell survival and proliferation[30].

Interleukin (IL)-6 plays a key role in HL pathogenesis. Serum IL-6 levels are markedly elevated in untreated patients and show a positive correlation with the presence of “B” symptoms, advanced disease, and poorer prognosis. Hodgkin/Reed-Sternberg cells secrete IL-6 (along with other cytokines) and express IL-6 receptors, suggesting that IL-6 may drive malignant cell proliferation through an autocrine feedback loop[31,32]. Notably, evidence exists that genetic variants associated with lower IL-6 levels provide protection against HL. Analysis of IL-6 174G>C promoter polymorphism showed that the risk of HL decreased with an increase in the number of C alleles, with the CC genotype corresponding to a low-secretion phenotype[32,33]. IL-12 is another cytokine implicated in HL development, acting through the activation of cell-mediated immunity via the stimulation of cytotoxic lymphocytes and natural killer cells. Reduced IL-12 levels lead to a diminished Th1 cytotoxic response to common childhood viral infections, a feature frequently observed in young-adult forms of HL. The presence of certain variants in the regulatory regions of the IL-12 gene has been associated with a significantly increased risk of HL, a finding further supported by studies in twin pairs[34].

The remarkable diversity of human leukocyte antigen (HLA) alleles impacts the affinity and specificity of antigen binding, and this is responsible for the existence of varying and individualized antiviral and antitumor immune responses. It is therefore unsurprising that different HLA polymorphisms have been described for various cancers, particularly for those of viral origin, including HL. Individuals carrying the HLA-A*02 allele have a reduced risk of developing EBV-positive HL, whereas the presence of the HLA-A*01 allele is linked to an increased risk. The protective safeguarding element of the HLA-A*02 allele lies in its ability to present EBV-derived peptides, resulting in a more effective immune response[35,36]. Genome-wide association studies have further identified additional loci within the HLA class I, II, an regions associated with HL[37-39]. For example, HLA-B5 is defined as a risk allele, while HLA-DR7 appears to be protective against cHL, regardless of EBV status. HLA-DR2 and HLA-DR5 have been linked to tumor susceptibility in EBV-negative cases, whereas HLA-B37 and HLA-DR10 confer increased risk in EBV-positive patients[38].

Recent whole genome sequencing studies have identified novel non-HLA predisposing variants in HL-affected families[21,40]. A missense mutation in DICER1 (a member of the ribonuclease III family), causing impaired expression of tumor-suppressing microRNAs, was identified in a case of mother-daughter HL, suggesting a potential mechanism of lymphomagenesis in that family[41]. Alterations in POT1 (protection of telomeres 1), leading to excessively long telomeres and increased DNA fragility, may also play a role in HL tumorigenesis[42]. Other genetic alterations, including those in KDR (kinase insert domain receptor), KLHDC8B (kelch domain-containing protein 8B) and ACAN (aggrecan), were identified in HL-affected families and are therefore implicated in disease development[43-45]. In 2023, Flerlage et al[21] reported results from the largest whole genome sequencing study of familial HL to date. They identified 44 HL-associated variants - 33 coding and 11 non-coding - with the most recurrent variants affecting KDR, KLHDC8B, PAX5 (paired box protein 5), and GATA3 (GATA binding protein 3) genes. Novel loci were also described in PAX5, GATA3, IRF7 (interferon regulatory factor 7), EEF2KMT (eukaryotic elongation factor 2 lysine methyltransferase), and POLR1E (RNA polymerase I subunit E). The study was performed on germline DNA from 234 individuals from 36 unrelated families, mostly of European ancestry, each with two or more first-degree relatives affected by HL, including at least one case diagnosed before the age of 21. Building on these pivotal findings, additional germline predisposition variants have since been identified, such as a homozygous alteration in the PXR (pregnane X receptor) ligand binding domain[46]. Table 1 provides an outline of the genetic alterations associated with HL described in this section[21,41-46].

Table 1 Candidate genes associated with Hodgkin lymphoma.

Candidate gene	Full name/role in tumorigenesis	Ref.	Number of mutation-segregating families	Number of affected individuals/family relation/age of HL onset
KDR	Kinase insert domain receptor/encodes type III receptor tyrosine kinase of vascular endothelial growth factor; mutations cause uncontrolled cell survival and growth	Rotunno et al[43], 2016	2	5; family 1 (two brothers, 13 and 16 yo, and a first cousin, 18 yo M); family 2 (father, 40 yo, and daughter, 14 yo)
		Flerlage et al[21], 2023	3	7; family 1 (two brothers and a first cousin, M); family 2 (father and daughter) - previously reported by Rotunno et al[43]; family 3 (brother and sister, ages unknown)
DICER1	Ribonuclease type III/endoribonuclease required for microRNA biogenesis; microRNAs act as tumor suppressors by inhibiting oncogenic signaling pathways	Bandapalli et al[41], 2018	1	2; (mother, 34 yo, and daughter, 7 yo)
POT1	Protection of telomeres 1/mutations lead to excessive telomere lengthening and DNA fragility	McMaster et al[42], 2018	2	5; family 1 (mother, 29 yo, and 3 siblings: 18 yo F, 19 yo F, and 31 yo M); family 2 (two siblings, 16 yo F and 22 yo M)
		Flerlage et al[21], 2023	1	3; (mother, son, and daughter, ages unknown)
KLHDC8B	Kelch domain-containing protein B/tumor suppressor gene; mutations lead to errors in cytokinesis, binucleation, and chromosomal instability	Salipante et al[44], 2009	1	3; (two sisters, 39 yo, and 42 yo, and a brother, 28 yo - possibly also the mother as a fourth member; she died of a mediastinal tumor but declined further diagnosis)
		Flerlage et al[21], 2023	3	8; family 1 (brother and sister, ages unknown); family 2 (grandfather, his two grandsons – brothers - and one great grandson, ages unknown); family 3 (mother and daughter, ages unknown)
PAX5	Paired box protein 5/codes a protein crucial for the differentiation and maturation of lymphoid progenitor cells into B cells; mutations cause the loss of CD20 and formation of Reed-Sternberg cells	Flerlage et al[21], 2023	2	5; family 1 (mother and two daughters, ages unknown); family 2 (two brothers, ages unknown)
GATA3	GATA binding protein 3/tumor suppressor specifically in B-cell lymphomagenesis; deficiency promotes B-cell differentiation and proliferation	Flerlage et al[21], 2023	2	5; family 1 (brother and sister, ages unknown); family 2 (mother, son, and daughter, ages unknown)
IRF7	Interferon regulatory factor 7/encodes a protein engaged in antiviral response and immune system regulation; involved in the regulation of EBV latency	Flerlage et al[21], 2023	1	4; (sister and brother and the brother’s two sons, ages unknown)
EEF2KMT	Eukaryotic factor 2 lysine methyltransferase/role in lymphomagenesis is unclear	Flerlage et al[21], 2023	1	3 (brother, sister, and their first cousin, F, ages unknown)
POLR1A	RNA polymerase I subunit E/encodes a subunit of RNA polymerase I; responsible for transcribing ribosomal RNA; abnormal ribosomal biogenesis can lead to uncontrolled cell growth	Flerlage et al[21], 2023	2	4; family 1 (brother and sister, ages unknown); family 2 (mother and daughter, ages unknown)
ACAN	Aggrecan/encodes proteoglycan; alterations may promote the oncogenic effects of EBV	Ristolainen et al[45], 2015	1	3; (3 siblings: 5 yo M, 12 yo M, and 12 yo F)
PXR	Pregnane X receptor/member of the nuclear receptor superfamily; the role in lymphomagenesis is still uncertain	Khodzhaev et al[46], 2024	1	2; (2 sisters, 10 and 24 yo)

M: Male; F: Female; HL: Hodgkin lymphoma; yo: Years old.

The growing interest in genetic research on affected families will certainly identify additional polymorphisms relevant to the etiology of HL. However, while these studies provide valuable preliminary insights, their conclusions must be interpreted with caution due to two major limitations. First, the statistical power of most analyses remains limited due to the relatively small number of affected families studied. Second, follow-up functional studies are required to validate these findings and elucidate their biological relevance. As with other genomic studies of cancer predisposition, significant ethical considerations arise - particularly because testing often reveals uncertain or probabilistic genetic risks that may place an emotional burden on carriers. To mitigate this emotional distress, clinicians must provide comprehensive pre- and post-test genetic counseling. This should include a clear explanation of the scope and limitations of genetic testing. When predisposing variants are identified, it is crucial to distinguish between increased risk and definitive diagnosis. Furthermore, identifying genetic predisposition may have a benefit by raising awareness, facilitating early detection, and improving outcomes, given the excellent prognosis of HL when diagnosed early. Moving forward, successful research and clinical translation involving familial HL will depend on an individualized approach to at-risk families, combining scientific investigation with tailored psychological and genetic counseling support.

CONCLUSION

The well-documented familial aggregation of HL highlights the role of genetic susceptibility as a significant factor in disease etiopathogenesis, particularly in younger patients. Recognizing the potential risk for first-degree relatives can influence clinical practice by increasing awareness of early symptoms, thereby facilitating prompt diagnosis and improving prognosis. At the same time, understanding that the absolute risk remains relatively low - close to that of the general population - can provide relief to the affected-families. Further research aimed at identifying causative genetic variations will improve genetic counseling and may ultimately support targeted genetic screening in selected individuals. Deeper insights into the molecular biology of HL could pave the way for novel personalized therapies that could replace conventional regimens based on chemotherapy and radiotherapy.

ACKNOWLEDGEMENTS

We express our sincere gratitude to Zinaida Peric, MD, Professor, and Neven Franjic, MD, Assistant Professor, Department of Hematology, Clinical Hospital Center Rijeka, Croatia, for their outstanding work in the field of hematology and for providing access to valuable data and insightful comments during the writing of the manuscript.