Autoimmune lymphoproliferative syndrome in Boerboel dogs: clinicopathologic, diagnostic, and genetic characterization

Abstract

Background

Autoimmune lymphoproliferative syndrome (ALPS) is a rare genetic disorder characterized by Fas cell surface death receptor (FAS)-mediated defective lymphocyte apoptosis, leading to chronic lymphadenomegaly, splenomegaly, and autoimmune cytopenias. Although documented in humans and British Shorthair cats, ALPS has not been reported in dogs.

Hypothesis/Objectives

Characterize the clinical presentation, laboratory findings and genetic basis of ALPS in Boerboel dogs.

Animals

Four affected Boerboel pups from a prospectively studied litter, 2 additional retrospectively identified littermates from a different litter, and 17 unaffected relatives (including littermates, dam, sire).

Methods

Prospective case series, with retrospective review of 2 additional affected dogs. Medical histories, clinicopathologic results and imaging findings were analyzed. Whole genome sequencing was conducted on 2 affected pups, with variant evaluation against 3023 canine whole genome sequences. Relatives were genotyped for the putative causal variant.

Results

Affected pups presented with lymphadenomegaly, splenomegaly, and variable cytopenias. Lymph node and splenic cytology showed reactive lymphoid hyperplasia and expanded large lymphocytes, with molecular clonality PCR indicating polyclonal lymphoproliferation. Lymph node flow cytometry and immunocytochemistry identified CD3⁺/CD4^–/CD8^– (double-negative) T cell proliferation. Whole genome sequencing of 2 affected pups identified a homozygous 14-base pair duplication in exon 2 of the FAS gene in both, predicted to result in a premature stop codon, absent from 3023 database dogs. All affected dogs were homozygous for the variant, whereas unaffected littermates, parents, and other relatives were heterozygous or clear of the variant.

Conclusions and clinical importance

We describe the features of ALPS in dogs and emphasize the importance of considering non-neoplastic lymphoproliferative disorders in young dogs with lymphadenomegaly and splenomegaly.

Introduction

Lymphoproliferation occurs in neoplastic conditions, particularly lymphoma, and as an immunological response to infections and immune-mediated disorders. Inherited defects in lymphocyte apoptosis represent another category of lymphoproliferation. Fas cell surface death receptor (FAS) (CD95) is a receptor that triggers programmed cell death upon binding with its ligand (FasL), a mechanism crucial for lymphocyte homeostasis and self-tolerance (Figure 1).¹ Autoimmune lymphoproliferative syndrome (ALPS) arises from inherited defects in FAS-mediated apoptosis, resulting in chronic lymphoproliferation and autoimmunity.¹ The disorder is rare in humans, but well-documented as a genetically defined immunodeficiency,² with FAS gene variants accounting for over 90% of cases^2,3 and additional cases are linked to FASLG and FAS-associated via death domain (FADD) gene variants.² Hallmarks of ALPS include non-neoplastic lymphadenomegaly, splenomegaly, and immune-mediated cytopenias,¹ and immunophenotypic analysis identifies an abnormal polyclonal expansion of mature T cells expressing the T-cell receptor (TCR) α/β chains but lacking both CD4 and CD8 co-receptors (TCRαβ⁺ double-negative [DN] T cells).⁴

Figure 1.

The FAS pathway’s crucial role in lymphocyte apoptosis. The process begins when FAS ligand (FasL) binds FAS, triggering FAS clustering. The clustered FAS receptors then recruit FAS-associated via death domain (FADD) protein and pro-caspases 8 and 10. Together, these components form the death-inducing signaling complex (DISC). Once assembled, DISC initiates a pro-apoptotic cascade by activating terminal caspases, ultimately leading to programmed cell death. Modified from Matson DR and Yang DT “Autoimmune Lymphoproliferative Syndrome: An Overview” (Arch Pathol Lab Med. 2020;144(2):245-251) with permission from Archives of Pathology & Laboratory Medicine. Copyright 2020. College of American Pathologists.¹

Despite recent advances in understanding the pathophysiology, diagnosis, and genomics of ALPS in human medicine, major knowledge gaps remain in veterinary medicine. In cats, ALPS has been described in British shorthairs,⁵ where a homozygous adenine insertion in exon 3 of the FASLG gene caused a frameshift, likely resulting in a truncated, non-functional protein.⁶ Naturally-occurring ALPS has been documented in mice with variants in FAS or FASLG genes.⁷ To date, no equivalent in dogs has been described, and the genetic and clinical features of ALPS in this species remain undescribed. This deficit limits our ability to diagnose and manage the range of ALPS-like conditions in dogs and restricts opportunities for comparative research.

We investigated 6 Boerboel dogs with non-neoplastic lymphoproliferation and cytopenias. Objectives were to (1) characterize the clinical, hematologic, cytologic, and immunologic features, and (2) investigate the genetic basis. We hypothesized affected dogs would show ALPS-consistent features and carry a pathogenic variant in a FAS apoptosis gene.

Materials and methods

Animals

Four Boerboel littermates (designated Pup 1 through Pup 4) suspected of having ALPS, were included in the study. All affected dogs presented with marked generalized lymphadenomegaly, splenomegaly, and cytopenias. Pups were presented to 4 veterinary facilities across Australia, including one referral center (index case: Pup 1), which performed most of the diagnostic evaluation. Clinicopathologic data from an unaffected littermate (Pup 5) was analyzed. The 4 affected pups, Pup 5, and the phenotypically normal dam and sire were included in the initial genomic analysis. Fourteen additional relatives were subsequently recruited for genotyping to assess variant segregation with disease.

Two additional Boerboel pups (referred to as “Additional case A and Additional case B”), identified later, were studied retrospectively. Both were littermates from a second affected litter in which the dam was a full sibling to the sire of the prospectively studied index litter, and the sire was distantly related to the dam of the index litter. Both Case A and Case B had the identical ALPS variant on genetic testing.

Diagnostic procedures

All 4 prospectively studied pups received serial physical examinations, CBCs and serum biochemistry profiles. The CBCs and biochemistry profiles for Pup 1 were performed by Nova Vet, with some testing conducted by IDEXX Laboratories, both accredited laboratories. Pup 2 had CBCs performed using a Vetscan HM5 (version 2.7) and biochemistry using a Vetscan VS2. Pup 3 underwent CBCs and biochemistry testing using a Zoetis Vetscan Fuse. Pup 4 had CBCs performed using the IDEXX ProCyte One and biochemistry on the IDEXX Catalyst One analyzer. The CBC for pup 5 also was performed by IDEXX.

Additional diagnostic tests were performed (Table S1).

Abdominal ultrasonography, and digital abdominal and chest radiography were performed at the attending veterinary hospital.

Cytology, immunocytochemistry, and molecular clonality PCR on lymph node and spleen aspirates were performed by the Department of Pathology, Microbiology and Immunology, University of California, Davis School of Veterinary Medicine. For molecular clonality testing by PCR for antigen receptor rearrangements (PARR), assays were performed to detect rearrangements of the T cell receptor gamma (TRG), immunoglobulin heavy chain (IgH), and kappa deleting element (Kde) genes in lymph node and splenic aspirates to evaluate clonality. Additional cytology, free abdominal fluid analysis, and a serum vitamin B12 concentration were performed by Nova Vet Diagnostics.

Flow cytometric analysis was performed on lymph node and blood samples from Pup 2. The antibody panel included CD3 and CD5 (pan T cell markers), CD4 and CD 8 (T cell subsets), CD 21 (B cell marker), CD45 (leukocyte common antigen), and major histocompatibility complex (MHC) Class II, allowing immunophenotyping of lymphocyte populations in both tissue aspirates and blood. Double-negative (CD4⁻/CD8⁻) T cells were identified as a subset within the CD3⁺ T cell population. Canine-specific TCR α/β and γ/δ antibodies were not available at the time of analysis. Reference intervals (RI) were applied, when available (Tables 2 and 3). Flow cytometry analyses were performed by Vetnostics.

Table 2.

Flow cytometry of lymph nodes in Boerboel puppies with autoimmune lymphoproliferative syndrome (pup 2 and additional case A).

Marker	Result (% of total lymphocytes) Pup 2 (4.5 months)	Result (% of total lymphocytes) Additional case A (21 months)	RI (%)
CD3 (Pan T cell)	62	75	60-80
CD5 (Pan T cell)	61	72	60-80
CD 4 (T cell subset)	7	12	35-50
CD8 (T cell subset)	13	29	15-25
CD21 (B cell subset)	10	12	5-20
CD4/CD8 double negative T cells	61 (of T cells)	40 (of T cells)	9-2519
CD45+ lymphocytes	81	88	N/A
MHC Class II	99	99	N/A

Table 3.

Flow cytometry of blood in Boerboel puppies with autoimmune lymphoproliferative syndrome (pup 2 and additional case A).

Marker	Result (% of total lymphocytes) Pup 2 (4.5 months)	Result (% of total lymphocytes) Additional case A (21 months)	Absolute Count (×10 9 /L)	RI
CD3 (Pan T cell)	16	67	0.2	1.5-2.2 × 109/L
CD5 (Pan T cell)	13	60	0.2	1.4-2.0 × 109/L
CD 4 (T cell subset)	5	23	0.1	0.6-1.9 × 109/L
CD8 (T cell subset)	8	27	0.1	0.45-1.0 × 109/L
CD21 (B cell subset)	79	9	1.2	0.085-0.35 × 109/L
CD4/CD8 double negative T cells	16.5 (of T cells)	35 (of T cells)	N/A	Up to 15%20
CD45+ lymphocytes	10	3	N/A	N/A
MHC Class II	97	82	N/A	N/A

Infectious disease screening was performed on blood samples using PCR assays targeting tick-borne pathogens including Bartonella spp., Babesia spp., Anaplasma platys, Ehrlichia spp., Mycoplasma haemocanis, and Candidatus Mycoplasma haematoparvum. These tests, along with fecal flotation and multiplex PCR testing for gastrointestinal pathogens, including common protozoal, viral, and bacterial agents, were conducted by IDEXX Laboratories. Infectious disease serology was performed using the IDEXX SNAP 4 Dx Plus Test.

For Additional case A and Additional case B, available clinical records and pathology data were reviewed retrospectively. Case A had CBC, serum biochemistry, lymph node cytology, lymph node histopathology with B- and T-cell marker immunohistochemistry, and flow cytometry of lymph node and blood (as per the index litter), with all analyses performed by Vetnostics. Case B was evaluated by lymph node histopathology and B- and T-cell marker immunohistochemistry at Gribbles Veterinary Pathology.

Molecular genetic analysis

Whole genome sequencing and genotyping

Whole genome sequencing and genotyping of affected puppies and family members were conducted by the Canine Genetics Lab, Minnesota. Blood samples anticoagulated with EDTA were collected from affected dogs (Pups 1-4), one unaffected littermate (Pup 5) and their dam and sire. Genomic DNA was isolated using a Puregene blood kit (Qiagen, Germantown, MD) following the manufacturer’s recommended protocol for mammalian blood. Short-read whole genome sequencing of DNA from Pup 1 and Pup 2 was performed at Azenta Life Sciences (South Plainfield, NJ). The DNA libraries were prepared using an Illumina TruSeq PCR-Free Kit and 150 base-pair, paired-end reads were generated using an Illumina NovaSeq 6000 system. Raw sequence reads are publicly available from the National Center for Biotechnology Information Short Read Archive (SRR33639172 and SRR33639173 under BioProject PRJNA937381). Reads were mapped against the dog reference genome UU_Cfam_GSD_1.0^8,9 concatenated with the Y chromosome (NC 051844.1) from the ROS Cfam 1.0 assembly (GCF 014441545.1) and processed for each dog using the OnlyWAG pipeline.¹⁰ Variants in each affected dog were compared to an internal database of whole genome sequences from 3023 dogs, wolves, and coyotes of 402 diverse breeds, including 4 Boerboels; this database includes 1971 dogs, wolves, and coyotes released by the Dog10K consortium.⁹ Variants unique to each affected dog were prioritized by predicted consequence and impact using Variant Effect Predictor.¹¹ High- and moderate-impact variants were searched for genes that have been previously implicated in causing ALPS in other species. These genes include FAS, FASLG, and FADD.^{2,3,6,7,12–16}

Based on whole genome sequencing (WGS) findings, a high-priority variant in the FAS gene was selected for genotyping by Sanger sequencing in Pups 3 and 4, the sire, dam, and a clinically unaffected littermate (Pup 5). Genomic DNA from each dog was amplified by PCR using Qiagen HotStar Taq (Qiagen, Germantown, MD) according to the manufacturer’s recommended reaction setup (scaled down to 20 μL) with 35 cycles at an annealing temperature of 58 °C utilizing forward primer 5′-CTGCAGCTCTCCTGCTTTTC-3′ and reverse primer 5′-AGAGAAAGCCTCTCCCTCCA-3′ to produce 403 or 417 base pair (bp) amplicons.

For screening additional related Boerboels, cheek swab samples from 14 additional Boerboel dogs were collected, and genomic DNA was isolated using the Puregene blood kit according to the manufacturer’s instructions for buccal brushes. Polymerase chain reaction using Qiagen HotStar Taq according to the manufacturer’s recommended reaction setup (scaled down to 20 μL) with 35 cycles at an annealing temperature of 58 °C with forward primer 5′-GTCAGGTGCTGGGAAGACC-3′ and reverse primer 5′-CCTAACGCAGGAGGACCAAG-3′ was utilized to produce 147 or 161 bp amplicons resolved by electrophoresis on a 3% agarose gel. Cheek swab samples from Additional case A and Additional case B were processed as above.

Results

Four Boerboel littermates with clinicopathologic abnormalities and genetic analyses consistent with ALPS were studied from 8 weeks to 14 months of age. Phenotypically normal littermates and other relatives were included in the study.

Clinical signs

All 4 prospectively studied pups presented with marked peripheral lymphadenomegaly and abdominal distension caused by splenomegaly at 8 (Pups 1 and 4), 10 (Pup 2), and 11 weeks of age (Pup 3; Figures 2 and 3). Lymphadenopathy was non-painful and generalized, with most pronounced enlargement in the mandibular lymph nodes. Pup 5, the dam, and the sire had normal physical examination findings.

Figure 2.

Peripheral lymph node enlargement in Boerboel puppies (pups 1-4) with autoimmune lymphoproliferative syndrome. Panels (A and B, C and D, E and F, and G and H) correspond to pups 1-4, respectively. Marked mandibular lymphadenomegaly in (A, C, E, G). Additional findings include inguinal (B), popliteal (D and F), and superficial cervical lymphadenomegaly (H). Enlarged lymph nodes are indicated by black arrows.

Figure 3.

Abdominal distension in Boerboel puppies (pups 1-4) with autoimmune lymphoproliferative syndrome. panels (A-D) correspond to pups 1-4, respectively, demonstrating marked abdominal enlargement caused by extreme splenomegaly.

During the study period, Pups 1-4 intermittently had secondary signs, including diarrhea (4/4), lethargy (4/4), inappetence (3/4), pyrexia (3/4), stertor (2/4), and coughing (2/4).

Additional dogs (cases A and B), identified retrospectively from a second affected litter, both developed marked peripheral lymphadenomegaly and abdominal distension at approximately 15 and 10-12 weeks of age, respectively. An additional dog (case A) developed recurrent diarrhea, inappetence, and pyrexia.

Diagnostic results

Hematology

Three of the 4 clinically affected pups (Pups 1, 3, and 4) presented with anemia. Pups 1 and 3 had mild or borderline anemia, with hematocrits (Hct) of 26% (reticulocytosis of 182 × 10⁶/μL) and 27%, respectively (RI for Hct in immature dogs, 27%-35%).¹⁷ The RI for reticulocyte counts was 0-90 × 10⁹/L. Pup 4 had more severe anemia, with Hct of 15% (reticulocytosis of 147 × 10⁶/μL). Pup 2’s first Hct, measured at 4 months of age after starting prednisolone, was 32%. The carrier (Pup 5) had Hct within the RI (initial Hct of 32% at 11 weeks of age). Total plasma protein concentrations were within laboratory RIs for all pups. Additional case A, identified retrospectively, had moderate normocytic, hypochromic non-regenerative anemia at 15 weeks of age (Hct 28%; reticulocyte count, 44 × 10⁹/L).

Platelet counts were obtained for all 4 affected pups. Pup 1 had moderate thrombocytopenia with a manual platelet count of 87 × 10⁹/L (RI, 143-448 × 10⁹/L) at 9 weeks of age. Pup 4 had a single manual count of 90 × 10⁹/L at 14 weeks of age. Additional case A had moderate thrombocytopenia on manual count (72 × 10⁹/L, confirmed by smear evaluation). For Pups 2, 3, 4, and 5, only automated platelet counts were available and considered minimum values because of potential underestimation caused by platelet clumping. Initial platelet counts for Pups 2-4 were: 114 × 10⁹/L, 37 × 10⁹/L, and 145 × 10⁹/L, respectively. The phenotypically normal carrier (Pup 5) had platelet counts within the RI, with an initial platelet count of 163 × 10⁹/L.

Serum biochemistry results

Routine serum biochemistry profiles for Pups 1-4 were normal at baseline, except for a moderately increased C-reactive protein concentration of 36 mg/L (RI, ≤ 10 mg/L) in Pup 1 and mild hyperglobulinemia 6.0 g/dL (RI, 2.3-5.2 g/dL) in Pup 2. A single mildly increased serum vitamin B12 concentration was noted in Pup 1 (681 pmol/L; RI, 185-670 pmol/L). Baseline serum biochemistry for Additional case A was unremarkable.

Infectious disease screening

Polymerase chain reaction testing of blood for vector-borne infectious diseases including Bartonella spp., Babesia spp., A. platys, Ehrlichia spp., M. haemocanis, Candidatus Mycoplasma haematoparvum was negative (Pup 1 and Additional case A). An in-house serologic test for Ehrlichia was negative (Pup 4).

Fecal screening for infectious disease

Pup 1 underwent fecal evaluation for pathogens at 3 months of age because of intermittent small and large bowel diarrhea. Fecal flotation and multiplex PCR testing were negative for parasitic ova, cysts, and enteric pathogens including Cryptosporidium spp., although a high quantity of Clostridium perfringens Enterotoxin A (CPEA) gene was detected. At 6 months of age, after diarrhea recurrence, repeat fecal evaluation was positive for Cryptosporidium spp. and continued to show high CPEA gene abundance. Pup 1 was treated for Cryptosporidium infection with azithromycin (15 mg/kg PO q24 h for 15 days). Despite treatment, intermittent diarrhea persisted, before diagnosis and treatment of ALPS.

Pup 5, the littermate without clinical suspicion for ALPS, presented with diarrhea (predominantly large bowel, with possible small bowel involvement) and tested positive for both Cryptosporidium and Campylobacter jejuni. Diarrhea in Pup 5 resolved spontaneously by 3 months of age, without antimicrobial treatment. All pups received routine anthelmintics.

Diagnostic imaging

Abdominal ultrasonography of Pups 1 and 3 identified marked multicentric lymphadenomegaly and severe splenomegaly. Lymph node measurements in Pup 1 ranged from 24 × 9 mm (left medial iliac) to 54 × 25 mm (jejunal; Figure 4A-D). Lymph nodes in both pups were sonographically heterogenous, characterized by a hypoechoic, often nodular periphery and a more echogenic center.

Figure 4.

Multicentric lymphadenomegaly in a 3-month-old Boerboel puppy with autoimmune lymphoproliferative syndrome (pup 1, body weight 11 kg). Enlarged lymph nodes include (A) gastric, (B) mesenteric, and (C) medial iliac. (D) Very large jejunal lymph node (54 mm × 25 mm); scale bar (right-hand side, panel D only) = 10 mm.

The spleens of both pups were diffusely mottled with small hypoechoic nodules (often elongated), approximately 2 × 5 mm in diameter, many of which contained a linear hyperechoic center (Figure 5A,B,D). In Pup 1, the combined volume of the enlarged lymph nodes and spleen was estimated to occupy more than 50% of the abdominal cavity. Additionally, a mild to moderate volume of anechoic free peritoneal fluid was observed (Figure 5C). Pup 5 (carrier) had an unremarkable abdominal ultrasound examination.

Figure 5.

Ultrasonographic findings in 3-month-old Boerboel puppies with autoimmune lymphoproliferative syndrome. (A) Extreme splenomegaly with diffuse hypoechoic nodules in pup 1. (B) Magnified view of splenic nodules showing linear hyperechoic centers (short arrows) in pup 1. (C) Anechoic free peritoneal fluid between spleen and left kidney (long arrows) in pup 1. (D) Enlarged spleen with diffuse small hypoechoic nodules in pup 3, similar to pup 1, and enlarged splenic lymph nodes.

Abdominal radiographs were obtained when Pup 1 was 4.5 months old, 4 weeks after starting sirolimus (proprietary formulation: Rapamune®, Pfizer Australia Pty Ltd) and 5 weeks after commencement of prednisolone treatment, because of suspected foreign body ingestion. Radiographs indicated moderate splenomegaly with smooth margins and large body habitus (Figure 6A and B).

Figure 6.

Radiographs of a Boerboel puppy with autoimmune lymphoproliferative syndrome (A) Right lateral and (B) Ventrodorsal abdominal views of pup 1 (4.5 months of age), demonstrating moderate splenomegaly.

To investigate coughing, thoracic radiographs were performed in Pup 3 at 4.4 months of age, 3 weeks after starting sirolimus and 4 weeks after starting prednisolone. A mild generalized interstitial lung pattern was observed. The visible portion of the liver appeared normal in size. No enlarged intrathoracic lymph nodes were observed. The cause of the reported cough was not apparent, and it resolved without intervention.

Cytology findings

Cytological examination of smears made from lymph node aspirates of Pups 1, 3, and additional case A stained with Wright-Giemsa identified a markedly heterogeneous lymphocyte population with moderately to markedly increased intermediate and large lymphocytes (Figure 7A). Increased small and intermediate lymphocytes with uropod formation (unipolar cytoplasmic extensions) were observed, indicating an activated phenotype, and histiocytic hyperplasia.

Figure 7.

Right popliteal lymph node and spleen fine needle aspirate cytology from a Boerboel puppy with autoimmune lymphoproliferative syndrome (pup 1): (A) Right popliteal lymph node aspirate cytology has a heterogeneous population of lymphocytes with a moderate to marked expansion of intermediate and large lymphocytes with round nuclei, multiple variably prominent nucleoli and moderate volumes of mid to dark blue cytoplasm. Scattered mitoses are observed (center right). Wright-Giemsa stain, 50× objective. (B) Splenic aspirate cytology is hypercellular with a heterogeneous population of lymphocytes that includes a markedly expanded population of large lymphocytes with round nuclei, multiple variably prominent nucleoli and moderate volumes of dark blue cytoplasm. Many lymphocytes have uropod formation, unipolar cytoplasmic extensions that are morphologic evidence of activation. Wright-Giemsa stain, 50× objective. (C) Splenic aspirate cytology. In addition to being hypercellular with a markedly expanded population of large lymphocytes, there is also histiocytic hyperplasia. Histiocytes have oval or indented nuclei and moderate volumes of pale gray cytoplasm (center and upper). Wright-Giemsa stain, 50× objective.

Cytological examination of splenic aspirates from Pup 1 had similar findings with a heterogeneous population of lymphocytes including markedly expanded intermediate and large lymphocytes, many with uropod formation, and more pronounced histiocytic hyperplasia than observed in the lymph nodes (Figure 7B and C). Abdominal fluid from Pup 1 contained mixed nucleated cells, consisting of small lymphocytes and neutrophils, with fewer large mononuclear cells and eosinophils. Low numbers of intermediate to large lymphocytes were present.

Immunocytochemistry

Lymph node immunocytochemistry (Pup 1) identified expanded CD3⁺CD4^–CD8^– DN T cells of approximately 20%, based on single labelling of multiple lymph nodes. This inference was based on findings that approximately 70% of lymphocytes were CD3⁺, of which about 50% were CD8⁺ cells (Figure 8A and B), with almost complete absence of CD4⁺ cells. Normal lymph nodes typically contain 60-70% T cells with CD4:CD8 ratio ranging from 2:1 to 3:1.^18,19 Results for CD3, CD4, CD8, and other markers are shown in Table 1.

Figure 8.

Left superficial cervical lymph node immunocytochemistry from a Boerboel puppy (pup 1) with autoimmune lymphoproliferative syndrome. (A) CD3 and (B) CD8a. Approximately 70% of the cells, mostly small lymphocytes, are strongly CD3⁺, and approximately 50% of the cells are strongly CD8⁺. Only rare, scattered lymphocytes were CD4⁺ (not shown), inferring increased, approximately 20%, DN (CD4^–CD8^–) T cells. Streptavidin-peroxidase with vector NovaRED substrate and hematoxylin counter stain, 50× objective.

Table 1.

Immunocytochemical profile of lymph nodes in a Boerboel puppy with autoimmune lymphoproliferative syndrome (pup 1).

Marker	Positivity	Lymph node	Result (%)
CD21 (B cells)	Strong	Right popliteal	20-30
CD3 (T cells)	Strong	Right popliteal, left superficial cervical	70
CD4 (helper T cells)	Rare scattered	Left superficial cervical, left popliteal	<1
CD8a (cytotoxic T cells)	Strong	Left popliteal, left superficial cervical	50
CD8b (cytotoxic T cells)	Strong	Left superficial cervical	50
TCRγδ (γδ T cells)	Strong (few scattered)	Left popliteal	<1
TCRαβ (αβ T cells)	Strong	Right popliteal	70

Immunocytochemistry of splenic aspirates from Pup 1 showed a mixed lymphocyte population, consisting of 70% CD3 positive T cells and 30% Pax5-positive B cells.

Immunohistochemistry

Histopathology of a right inguinal lymph node biopsy sample at 5 months of age (Additional case A) showed interfollicular expansion by intermediate and large lymphocytes, plasma cells, Mott cells, and neutrophils, with frequent multifocal aggregates of multinucleated giant cells in the medulla. The mitotic rate in monomorphic regions of intermediate to large sized lymphocytes, was 17 per high-power field (HPF). Immunohistochemistry for CD3 and Pax5 identified appropriate labelling for T and B cell markers.

Histopathology of a mesenteric lymph node biopsy sample (Additional case B) showed diffuse expansion of intermediate lymphocytes, with a mitotic rate of 68 per 10 HPF and numerous tingible body macrophages. Immunohistochemistry indicated that >95% of lymphocytes were CD3-positive, with small, condensed follicles containing Pax5-positive cells.

Flow cytometry

Flow cytometric analysis of lymph node and blood samples from Pup 2 indicated marked expansion of CD3⁺CD4^–CD8^– DN T cells in the lymph node. In lymph node samples, DN T cells accounted for approximately 61% of the CD3⁺ T cell population (Table 2, Figure 9A). In peripheral blood, DN T cells accounted for 16.5% of the CD3⁺ T cell population, indicative of mild expansion (Table 3, Figure 9B). These findings are consistent with the lymph node immunocytochemistry results obtained in Pup 1.

Figure 9.

Representative flow cytometry scatter plots showing CD4 and CD8 expression among gated CD3⁺ T cells in a Boerboel puppies with autoimmune lymphoproliferative syndrome. (A) Lymph node (pup 2): DN T cells comprise approximately 61% of the CD3⁺ T cell population (J3 quadrant; red dots), indicating marked expansion. (B) Peripheral blood (pup 2): DN T cells account for 16.5% of the CD3⁺ T cell population (K3 quadrant; red squares), indicating mild expansion. (C) Lymph node (additional case A): DN T cells comprise 40% of the CD3⁺ T cell population (I3 quadrant; red dots). (D) Peripheral blood (additional case A): DN T cells account for 35% of the CD3⁺ T cell population (I3 quadrant; red dots). Quadrant percentages represent the proportions of cells within the gated CD3⁺ T cell population.

In Additional case A, DN T cells accounted for 40% of the CD3⁺ T cell population in lymph node (Table 2, Figure 9C) and 35% in peripheral blood (Table 3, Figure 9D). Both results are above published RIs for normal dogs (9-25% for lymph node,¹⁹ up to approximately 15% for peripheral blood²⁰). These samples were collected while the dog was receiving prednisolone.

Molecular clonality

Molecular clonality PCR of lymph node and spleen aspirates from Pup 1 showed polyclonal TRG, IgH, and Kde gene rearrangements.

Molecular genetic findings

Whole genome sequencing and genotyping of affected Boerboels and family members

Whole genome sequencing of Pup 1 identified 222,264 unique variant positions and 315 679 unique variant effects (Table S2); WGS of pup 2 identified 248 609 unique variant positions and 349 366 unique variant effects (Table S2). Across the 2 pups, 144 858 variant positions were shared, and included 206 297 variant effects (33 high, 133 moderate, 335 low, 205 796 modifier impact) in Pup 1 and 207 351 variant effects (33 high, 133 moderate, 337 low, and 206 848 modifier impact) in Pup 2. Only 1 high or moderate impact variant was present in both pups among the identified candidate genes. Both affected pups were homozygous for a 14-base-pair duplication in exon 2 of the FAS gene (XM_038576079.1:c.179_192dup) at chr26:39,040,587. This duplication is predicted to cause a frameshift mutation resulting in a downstream premature stop codon (XP_038432007.1:p.Pro65ValfsTer54) in the gene transcript. This variant was absent from all 3,023 dogs in our in-house WGS database. FAS encodes the FAS receptor, also known as CD95, a “death receptor” that leads to programmed cell death.^21,22 No high or moderate impact variants shared by both probands and absent from our WGS database of 3023 dogs were identified in any of the other candidate genes. Mapped alignments of all candidate genes were scanned visually to search for structural variants; none were identified. The remaining unique variants with predicted high or moderate impact common to both affected pups were reviewed; none were located in genes known to be associated with any lymphoproliferative disorder.

Sanger sequencing subsequently was used for genotyping of the FAS 14-base pair insertion in exon 2 in Pups 3, 4, and 5, and their dam and sire. Pups 3 and 4 were homozygous for the variant. Pup 5, the dam and the sire were heterozygous (Figure 10). Subsequent gel electrophoresis fragment length genotyping of 14 relatives identified 8 additional heterozygous dogs, and 6 dogs clear of the FAS insertion. Additional case A and Additional case B, identified subsequently, were genotyped retrospectively and were homozygous for the same ALPS variant.

Figure 10.

Sanger sequence chromatograms of affected pups 1-4 and unaffected pup 5, sire and dam showing the FAS 14 bp duplication within exon 2 (highlighted region), all affected dogs are homozygous for the 14 bp FAS duplication, while the unaffected dogs are heterozygous for the duplication (XM_038576079.1:c.179_192dup) at chr26:39,040,587.

Treatment and clinical outcomes

All 4 prospectively studied pups were treated with sirolimus and prednisolone, based on approaches described for human patients. All 4 pups had a favorable response to treatment. Detailed information regarding treatment administration, monitoring, and outcomes are reported separately.²³

Additional cases A and B, included retrospectively, received prednisolone (case B also received mycophenolate mofetil). Both cases had incomplete responses and substantial adverse effects attributed to prolonged corticosteroid treatment.

Pedigree analysis for inheritance pattern and shared common ancestry

A partial pedigree (Figure 11) illustrates the relatedness of the affected dogs. Both sets of parents from the 2 affected litters were confirmed heterozygous carriers of the causal variant, and all clinically affected pups were homozygous for the variant, consistent with autosomal recessive inheritance. Among the genotyped relatives, 2 siblings from 1 litter and other more distantly related dogs were either heterozygous or clear of the variant. The specific ancestral origin of the variant could not be determined from the available data.

Figure 11.

Partial pedigree illustrating inheritance pattern in Boerboel puppies affected with autoimmune lymphoproliferative syndrome. Six clinically affected dogs that were all demonstrated to be homozygous for a FAS variant; these dogs are shown in solid black, including the 4 prospectively studied littermates and 2 retrospectively identified affected littermates (additional case A and additional case B). The dam and sire, both confirmed carriers (denoted with vertical stripes). Circles represent females, squares represent males, and diamonds represent dogs of unknown sex; open circles and open squares represent females and males of unknown ALPS or carrier status, respectively.

Discussion

Here, we report ALPS in 4 Boerboel littermates based on clinicopathologic findings, family history, and a pathogenic FAS variant. These criteria align with both historical and revised ALPS classifications in humans.^2,24 Immunocytochemistry and flow cytometry findings showed the accumulation of abnormal DN T cells, a hallmark of ALPS in humans.²⁵ The description of these cases facilitates the clinical recognition of this disorder in dogs. Two additional affected littermates from a separate litter were later identified retrospectively, further supporting the consistency of the clinical phenotype.

The 14-base-pair duplication in exon 2 of the canine FAS gene (c.179_192dup) is predicted to cause a frameshift mutation beginning at amino acid residue 65 (p.Pro65ValfsTer54), and leading to a premature stop codon 54 amino acids downstream. This mutation is expected to truncate the FAS receptor protein, approximately 319 amino acids in dogs, resulting in loss of most of the extracellular domain, the transmembrane domain, and the intracellular death domain critical for initiating apoptosis. Given the early position of the premature stop codon, the mutant transcript may be subject to nonsense-mediated mRNA decay, decreasing protein expression.²⁶

The domain structure and function of the canine FAS receptor is conserved with that of humans and other mammals, supporting the functional impact of this truncation.^27,28 Frameshift mutations introducing premature stop codons have well-established effects of generating nonfunctional proteins or triggering mRNA degradation.²⁶ Studies using whole genome and exome sequencing in dogs have demonstrated the utility of these approaches in predicting variant effects consistent with known protein structures and functions.²⁹ Therefore, this predicted truncation likely impairs FAS receptor function, consistent with the ALPS phenotype observed in the affected Boerboel puppies.

The different outcomes between FAS-variant puppies and FASLG-variant kittens may reflect how these variants impact apoptosis. Complete loss of FasL theoretically abolishes the initiation of FAS-mediated cell death, leading to complete loss of lymphocyte regulation, whereas some FAS variants may permit residual receptor activity and thus partial apoptotic function, potentially allowing survival but with marked disease. This interpretation is mechanistic and speculative, because direct comparative data for survival differences between biallelic FAS or FASLG variants are not available in people or animals.^3,30–34

The most common laboratory abnormalities in ALPS are cytopenias, particularly anemia, which in humans can result from autoimmune destruction of blood cells or splenic sequestration.^35–37 The initial Hct in Pup 4 was moderately to markedly decreased, and borderline decreased in Pups 1 and 3, consistent with both of the cytopenias typically described in ALPS and the physiologic expectation of lower Hct in young, growing animals. Without direct evidence of immune-mediated red blood cell destruction, such as autoagglutination at 4 or 37 °C or Coombs test positivity, the precise cause of the anemia remains uncertain. Furthermore, Additional case A demonstrated mild, non-regenerative anemia, inconsistent with hemolysis as the primary underlying mechanism. In contrast, platelet counts are not typically lower in healthy puppies compared to adults, and the observed thrombocytopenia in Pup 1, Pup 4, and Additional case A is likely attributable to disease rather than age-related variation.

The mild to moderate free peritoneal fluid in Pup 1 may be a consequence of lymphatic or venous obstruction secondary to marked abdominal lymphadenomegaly and splenomegaly. These conditions can impair both lymphatic drainage and portal venous return, increasing intra-abdominal hydrostatic pressure and resulting in fluid accumulation.

Increased serum vitamin B12 concentrations are a characteristic finding in humans with ALPS, resulting from increased production of haptocorrin by lymphocytes, particularly DN T cells, which binds and accumulates B12 in circulation.³⁸ In ALPS patients, serum vitamin B12 concentrations can be up to 15 times the upper limit of the RI as a result of this mechanism. In our study, vitamin B12 concentration was measured only in Pup 1 and was mildly increased. Because vitamin B12 was not assessed in the other affected pups, no conclusions can be drawn regarding its consistency or diagnostic value.

The lymphoproliferation with expanded large lymphocytes on cytology initially raised consideration of lymphoma. The young age of the pups and the involvement of multiple littermates was, however, much more consistent with an inherited, non-neoplastic cause. PARR testing more definitively excluded lymphoma. Our report emphasizes the importance of recognizing ALPS as a potential cause of lymphadenomegaly in young animals and raises the possibility that it may have been missed in previous cases, as in the additional cases detected retrospectively here.

It is recommended that all young dogs and cats presenting with lymphoma-like pathology undergo PARR testing, flow cytometry, immunocytochemistry, or some combination of these, including assessment of DN T cells to evaluate for ALPS and differentiate it from lymphoma. Genomic analyses to identify variants should be considered, but specific variant sites may vary among breeds, emphasizing the need for breed-specific molecular investigations.

Infectious diseases such as ehrlichiosis and leishmaniosis are potential differential diagnoses in young dogs presenting with lymphadenomegaly and lymphocyte proliferation. These infections can induce both monoclonal and polyclonal lymphocyte expansions, which clinically and cytologically may resemble lymphoproliferative disorders such as ALPS. Consequently, comprehensive infectious disease screening should be considered to exclude these conditions, particularly in geographic regions where such pathogens are endemic.

A hallmark of ALPS in humans is the proliferation of DN T cells. Limited data is available regarding the percentage of DN T cells in the lymph nodes of normal dogs. In a study of popliteal lymph nodes in 20 dogs, DN T cells ranged from 9% to 25% of CD3⁺ T cells.¹⁹ In Pup 1 that underwent flow cytometry of lymph node aspirates, the DN T cell population was markedly higher at 61% of total T cells, whereas flow cytometry of peripheral blood identified 16.5% DN T cells. Previous studies have reported that DN T cells can make up to approximately 15% of T cells in peripheral venous blood of dogs.²⁰ In Additional case A, flow cytometry during prednisolone treatment showed increased DN T cells (40% in lymph nodes, 35% in blood), exceeding RIs. This finding indicates expanded DN T cells as a diagnostic marker for ALPS in dogs, although corticosteroid effects and published data remain limited. Because only a few dogs were assessed, it is not possible to determine whether flow cytometry of peripheral blood for enumeration of DN T cells would be a useful diagnostic marker. Studies in humans show the percentage of DN T cells decreases with age, with younger individuals having higher proportions in their lymph nodes compared with older individuals.³⁹ A similar age-related trend may occur in dogs, with puppies naturally having higher percentages of DN T cells in lymph nodes compared with adult dogs.

The diagnostic criteria for ALPS in humans have evolved with genetic advancements and improved differentiation of disorders that mimic this syndrome. The 2010 criteria required lymphoproliferation, expansion of DN T cells, and evidence of defective FAS-mediated apoptosis, with genetic confirmation considered supportive but not mandatory.²⁴ The revised 2024 criteria now require a proven genetic basis for ALPS, including variants in FAS, FASLG, or FADD, along with evidence of a functional FAS signaling defect, and key clinical abnormalities to establish a definitive diagnosis.² Functional confirmation of a FAS signaling defect may include abnormal FAS-mediated apoptosis assays, increased biomarkers highly sensitive and specific for ALPS (eg, sFasL, IL-10, vitamin B12), or a positive family history, in addition to genetic findings.²

Although the presence of DN T cells supports a diagnosis of ALPS, genetic testing remains the definitive diagnostic method. Flow cytometry and genetic testing are available through commercial and academic laboratories, but access is limited in some countries and regions, particularly outside North America. Collaboration with laboratories in other countries can help overcome some of these accessibility barriers. Advances in WGS and diagnostic technologies are improving accessibility, and are expected to enhance global availability of these important diagnostic tools in veterinary practice.

Two additional affected littermates from a separate litter were identified retrospectively. Additional case A had hematology and flow cytometry performed, showing a comparable phenotype to the index litter with marked expansion of DN T cells. Additional case B exhibited similar clinical signs, and lymph node histopathology identified expanded intermediate lymphocytes. Another pup from this litter was euthanized because of suspected lymphoma. These additional cases broaden the observed clinical phenotype and emphasize that ALPS can closely resemble lymphoma based on routine histopathological assessment, reinforcing the importance of integrating immunophenotyping, genetic testing, and clinical context, particularly in young, related animals, to achieve an accurate diagnosis.

In this family of dogs, ALPS exhibited an autosomal recessive inheritance pattern, with apparent complete penetrance. Among the index litter of 15 pups, 4 (approximately 25%) were affected, consistent with Mendelian expectations for autosomal recessive traits, although not all littermates were available for genotyping. In a second litter, 2 additional affected pups (Additional cases A and B) were confirmed to be homozygous for the FAS variant; the total number of pups in this litter was not determined. In humans, ALPS most commonly results from heterozygous variants in the FAS gene and displays autosomal dominant inheritance.³ Rarely, autosomal recessive inheritance is observed, mainly in consanguineous families with biallelic FAS variants causing severe early-onset disease.^34,40–44

Conclusions

We have described 4 littermate Boerboel pups diagnosed with ALPS, a previously undocumented genetic disorder of dogs. Two additional affected littermates were identified later, both homozygous for the causal FAS variant. Autoimmune lymphoproliferative syndrome should be considered in the differential diagnosis of lymphoproliferation, particularly lymphadenomegaly and splenomegaly in young dogs, with or without cytopenias. Expanded DN T cell populations were observed in affected dogs, supporting their diagnostic value, especially when several related animals show similar signs. Because large lymphocytes are present on cytology or histopathology, care must be taken to avoid misdiagnosing ALPS as lymphoma.

Abbreviations

ALPS

autoimmune lymphoproliferative syndrome

bp

base pair

DISC

death-inducing signaling complex

DN

double-negative

FADD

Fas-associated via death domain (gene and protein symbol)

FAS

Fas cell surface death receptor (gene and protein symbol)

FasL

Fas ligand (protein symbol)

FASLG

Fas ligand (gene symbol)

Hct

hematocrit

HPF

high-power field

IgH

immunoglobulin heavy chain

Kde

kappa deleting element

MHC

major histocompatibility complex

PARR

PCR for antigen receptor rearrangement

Pax5

paired box gene 5

RI

reference interval

TCR

T cell receptor

TRG

T cell receptor gamma

WGS

whole genome sequencing

Author contributions

Linda Jane Tong (Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing—original draft, Writing—review & editing), Steven Friedenberg (Conceptualization, Formal analysis, Methodology, Resources, Supervision, Writing—original draft, Writing—review & editing), Jonah Cullen (Formal analysis, Methodology, Resources, Writing—review & editing), Hayden Hamsher (Formal analysis, Methodology, Resources, Writing—review & editing), Katie M Minor (Conceptualization, Formal analysis, Methodology, Resources, Writing—review & editing), Eva Furrow (Conceptualization, Methodology, Resources, Writing—review & editing), William Vernau (Investigation, Methodology, Resources, Visualization, Writing—original draft, Writing—review & editing), Lisa Horgan (Supervision, Writing—review & editing), Jason Stayt (Investigation, Resources, Writing—review & editing), George Reppas (Investigation, Resources, Writing—review & editing), Adrien-Maxence Hespel (Investigation, Visualization), and Richard Malik (Conceptualization, Funding acquisition, Methodology, Supervision, Visualization, Writing—original draft, Writing—review & editing)

Conflicts of interest

Authors declare no conflicts of interest.

Funding

Authors received no specific funding for this work.

Off-label antimicrobial declaration

Authors declare no off-label use of antimicrobials.

Institutional animal care and use committee or other approval declaration

Authors declare no institutional animal care and use committee or other approval was needed.

Human ethics approval declaration

Authors declare human ethics approval was not needed.