Lineage-Specific Chimerism and Outcome After Hematopoietic Stem Cell Transplantation for DOCK8 Deficiency

The inborn error of immunity (IEI) caused by bi-allelic dedicator of cytokinesis 8 (DOCK8) deficiency was first described in 2009 by Zhang et al. [1]. The resulting combined T- and B-cell immunodeficiency in DOCK8 deficiency presents with food allergies, eczema, eosinophilia, and elevated IgE, explaining why it was historically categorized as an autosomal-recessive hyper-IgE syndrome (HIES) [2‐4]. Recurrent sinopulmonary infections with bronchiectasis, skin infections, lymphopenia, and hypogammaglobulinemia are other frequent findings [2, 3, 5]. Common infectious agents are bacteria and fungi, DNA viruses, or molluscum contagiosum. A T_H1-deficient phenotype and cytokine abnormalities with T_H2 activation may explain eosinophilia and IgE overproduction in DOCK8 deficiency [6, 7]. Diagnosis may be delayed or missed in patients with milder or varying phenotypes or near-normal flow cytometric DOCK8 expression caused by missense mutations or somatic reversions [3, 5, 8, 9]. The long-term natural disease outcome is poor, given infectious or vascular complications and an increased susceptibility to malignancy, predominantly of hematological or epithelial origin. Median survival is less than 20 years and almost all patients experience a life-threatening complication by the age of 25, despite prophylactic treatment [10].

DOCK8 is an atypical guanine nucleotide exchange factor of the DOCK 180 superfamily, which interact with Rho GTPases regulating cytoskeletal rearrangements [1, 5, 10, 11], partially explaining phenotypical overlap with other IEIs caused by cytoskeletal dysregulation, especially Wiskott-Aldrich syndrome (WAS) and CARMIL2 deficiency [12, 13]. Interaction between DOCK8 and WASP in NK-cells has been reported [2, 14, 15]. In comparison, the IEI caused by STAT3 deficiency, autosomal-dominant HIES, presents with different phenotypical features, such as connective tissue, dental, and skeletal abnormalities, and typically exhibits a less severe immunodeficiency [2, 3, 5, 16‐18].

Previous reports suggest that allogeneic hematopoietic stem cell transplantation (HSCT) offers a curative therapeutic option for DOCK8 deficiency [2, 12, 19‐22]. A recent retrospective analysis of a large cohort confirmed favorable outcomes after HSCT and suggested good safety and efficacy of reduced toxicity conditioning regimens. A detailed analysis of donor chimerism and immunological reconstitution was however not possible in this multi-center retrospective analysis [23]. Treosulfan or reduced-dose busulfan-based reduced toxicity regimens offer less short-term and possibly long-term toxicity [24]. However, they do not always result in complete donor chimerism of all cell lineages, which—depending on the underlying condition—may or may not result in reversal of the disease phenotype [25]. For example, in familial hemophagocytic lymphohistiocytosis, a donor T-cell chimerism of about 20–30% is deemed sufficient [26], regardless of the chimerism in other cell lineages, while in WAS complete donor chimerism is necessary for full disease correction [27]. Reports on immunological outcome of DOCK8 patients after HSCT are relatively scarce, especially for patients with mixed chimerism [19, 28, 29].

Given the fact that DOCK8 deficiency affects many hematopoietic lineages, we reasoned that reversal of disease phenotype after HSCT would depend on the level of cell lineage-specific donor chimerism. We here report detailed analysis of long-term clinical outcome, lineage-specific chimerism, laboratory parameters, and pulmonary function tests of nine patients who underwent HSCT for DOCK8 deficiency between 2004 and 2017, four of whom with ensuing variable degrees of mixed chimerism.

The IEI resulting from bi-allelic deficiency of DOCK8 presents with combined T- and B-cell immunodeficiency. Hallmarks include frequent respiratory and skin infections, food allergies, eczema, eosinophilia, and elevated IgE. Subsequent organ damage, susceptibility to malignancies, and vascular complications entail high mortality and morbidity, justifying HSCT as a curative treatment option. Limited data on long-term outcomes for patients with DOCK8 deficiency and mixed chimerism post-HSCT are available, warranting this report on nine patients with a median follow-up of 78 months (total 727 patient months) post-HSCT and in-depth analysis of cell lineage-specific chimerism. Overall, we report successful HSCT after reduced toxicity conditioning in all patients with improvement of susceptibility to infections and allergies and resolution of eczema. Specifically, patients with a mixed chimerism present nearly complete T-cell chimerism, suggesting a selective advantage for wild-type donor T-cells, but lower donor B-cell chimerism with a possible tendency to lower IgG values and immunoglobulin substitution. No increased allergic, respiratory, or infectious complications were observed in patients with mixed chimerism. Secondary thyroid cancer in P2 was attributed to TBI conditioning and was not interpreted as an increased risk for oncogenesis in the context of mixed chimerism.

Aydin et al. reported on a large cohort of 81 patients with DOCK8 deficiency and HSCT on behalf of the inborn errors working party of EBMT and ESID with promising overall outcomes of HSCT [23]. More recently, Haskologlu et al. reported on a Turkish cohort of 20 patients with DOCK8 deficiency of whom 11 patients underwent HSCT [28]. However, these and other reports have focused on survival, conditioning regimens, and transplant-related outcomes, rather than improvement of clinical aspects and immunologic reconstitution, particularly in patients with mixed chimerism. Al-Herz et al. reported a retrospective review of 11 patients with ameliorated infectious and atopic symptoms post-HSCT, including an analysis of mixed lineage-specific chimerism in three patients, albeit with significantly shorter follow-up than our cohort [19]. Overall survival was 100% in our study, compared to 91% (Haskologlu et al.), 91% (Al-Herz et al.), and 84% (Aydin et al.), and none of our patients suffered from cGvHD. In patients with mixed chimerism, lineage-specific analysis showed a median of 97% (87–99%) of donor T-cells and 41% (30–85%) of donor B-cells, comparable to results from Al-Herz et al. (donor T-cells 82–97%, donor B-cells 0–46%, donor myeloid 0–7%) [19]. In line with our findings, Al-Herz et al. conclude that donor-derived T cells have a selective advantage. The role of DOCK8 in T-cell survival may provide a selective advantage for DOCK8 wild-type donor T-cells after HSCT, resulting in more complete T-cell donor chimerism [30, 31]. The extensive lymphocyte phenotype of DOCK8 deficiency and its correction post-HSCT have been catalogued by Pillay et al., although only in patients with complete donor T-cell chimerism [29]. Due to the retrospective nature of our study and lacking consent for biobanking, no further T-cell phenotyping or cytokine profiles are available for our patients for further assessment of T-cell reconstitution in DOCK8 deficiency with mixed chimerism. P2–P5 present with near complete T-cell chimerism with proper overall T-cell subset reconstitution, and therefore, we expect comparable T-cell phenotypes to patients with complete chimerism as shown by Pillay et al. [29]. P1, however, with 87% CD3⁺ donor T-cells has reduced thymopoiesis after partial thymectomy, weakening overall interpretation of T-cell subsets for patients with mixed chimerism. Further analysis of T-cell subsets for P1 would therefore not add meaningful knowledge to T-cell reconstitution in mixed chimerism for DOCK8 deficiency. Therefore, we believe our cohort may add additional information for T-cell reconstitution post-HSCT in patients with mixed chimerism.

Other studies found lower proliferative rates of DOCK8-deficient B-cells [8, 11]. Conversely, we observed no relevant selective advantage for wild-type B-cells in vivo as evidenced by an overall lower but stable B-cell chimerism over time after HSCT, but subtle effects cannot be excluded due to the low number of patients. Myeloid chimerism was not available for our patients, but a selective advantage for wild-type DOCK8 in lymphoid cells, rather than myeloid cells, was previously shown [32].

DOCK8 deficiency has been shown to impair immunological synapses through ICAM-1 in mouse models and reduced persistence in germinal centers and affinity maturation or differentiation into marginal zone cells in DOCK8-deficient B-cells [33]. Additionally, decreased generation of memory B-cells with diminished long-lasting antibody production has been related to impaired B-cell signaling [2, 11] and reduced B-cell proliferation was seen in patients with DOCK8 deficiency [11, 29]. DOCK8 deficiency has been shown to disrupt B-cell responses to signals via TLR, BCR, CD40, and cytokines, especially IL-21 [29]. In one patient, Al-Herz et al. found increased donor chimerism in the switched memory B-cell compartment, in agreement with recent findings of DOCK8 enrichment in the memory B-cells [32]. However, we did not analyze this in our patients but reduced switched memory B-cell numbers were not clearly associated with mixed chimerism in our cohort. The latter is exemplified by P5 with hypogammaglobulinemia and a higher CD19⁺ donor B-cell chimerism, but the lowest CD19⁺ B-cell count and reduced memory B-cells as compared to other patients with mixed chimerism. B-cell reconstitution post-HSCT is variable and influenced by many factors in pediatric patients [34]. Additionally, DOCK8-mutant B cells are unable to form marginal zone B cells or to persist in germinal centers and to undergo affinity maturation [33]. Split chimerism, with allogeneic T-cells but persistent autologous B-cells, impairs reconstitution of humoral immunity and has been implied with defective IL-21 receptor signaling in T-cell-dependent B-cell differentiation post-HSCT in patients with X-SCID and JAK3-SCID [35]. Our data show a possible tendency to hypogammaglobulinemia and necessity for immunoglobulin substitution in patients with mixed chimerism, but with the limited data available, no causal claim for DOCK8-specific effects can be made. Further research is required to elucidate the effect of split chimerism on B-cell function and adaptive humoral immunity reconstitution in general and particularly in DOCK8 deficiency.

We observed no correlation between persistent airway obstruction post-HSCT and mixed chimerism. Persistent food allergies post-HSCT are commonly seen in DOCK8-deficient patients without correlation with chimerism and are slow to improve post-HSCT [12], presumably due to persistence of long-lived IgE-producing plasma cells [36] and tissue-resident memory B-cells. Happel et al. report similar results for 12 patients and detected allergen-specific mast cells through skin prick tests after HSCT [36]. Only few of our patients reported challenging prior food allergies post-HSCT; however, reintroduction of a diversified diet post-HSCT has been successfully reported for numerous other patients [28, 36‐38]. The low incidence of allergic events post-HSCT, despite overall stable allergen reactivity on immunoblot, may be attributed to higher T-cell chimerism and a speculative role of regulatory T-cell (T_reg) functions. Janssen et al. demonstrated reduced suppressive activity of T_reg in DOCK8 deficiency [39]. T_reg activity may contribute to improvement of eczema in our patients and previous reports [23, 37, 38]. Unfortunately, no evaluation of T_reg was available for our patients. A possible role for T_reg suppression of T_H2-cells in atopic dermatitis is reviewed by Agrawal et al. [40]. Interestingly, T_reg express an affinity towards skin-homing through CCR4, CCR6, and cutaneous lymphocyte-associated antigen (CLA), a similar mechanism as seen in IPEX syndrome, caused by a T_reg defect [40]. Further studies are necessary to elucidate the role of T_reg in DOCK8 deficiency. Unfortunately, the importance of NK-cell immunity in defense against virus infections cannot be properly addressed by this report, due to missing appropriate functional assays [14].

We believe that early diagnosis and timely curative management are crucial in DOCK8 deficiency as non-reversible organ damage may occur due to infectious or auto-inflammatory complications, but may be delayed by varying phenotypes or near-normal DOCK8 expression caused by hypomorphic variants or somatic reversions [5]. Somatic reversions were present in two of our patients and both present with complete donor chimerism post-HSCT. The importance of early detection is exemplified by P8, who was diagnosed at a later age than her sister P3 and suffers from structural pre-HSCT lung damage with frequent respiratory infections despite complete donor chimerism. In comparison, P3 presents as a healthy and active young girl, despite mixed chimerism. Somatic reversion in general may ameliorate disease course, but patients still do experience fatal complications [8]. Rare constellations such as somatic reversion of a hypomorphic DOCK8 allele leading to an atypical and relatively milder phenotype have been reported [9] and Pillay et al. recently reported in-depth analysis of partial somatic reversion in three patients with compound heterozygous mutations in DOCK8 with clinical improvement over time [32]. Overall, T-cells showed highest levels of somatic reversion, in agreement with our findings regarding selective advantages for donor T-cells post-HSCT. Therefore, indication for HSCT in patients with somatic mosaicism should be evaluated individually, but we currently recommend HSCT with reduced conditioning for those with infectious complications to prevent organ damage and fatal complications. Hypothetically there could be long-term complications of clonal selection by somatic reversion, but Pillay et al. have not observed that [32]. Metabolomic biomarkers may support clinicians in future differential diagnosis of atopic dermatitis in order to achieve early diagnosis of DOCK8 deficiency [41], but currently, the diagnosis demands experienced clinicians and intracellular flow cytometry for DOCK8, most successfully detected in B-cells, which show minimal reversion [8], and confirmation by Sanger and/or exome sequencing.

With this report, we hope to increase our understanding of long-term immunologic outcomes post-HSCT for DOCK8 deficiency and support clinicians, patients, and families facing this debilitating disease. We highlight the importance of intracellular flow cytometry for DOCK8 for diagnosis and chimerism monitoring at the single cell level. A review of conditioning regimens is beyond the scope of this report, but overall, promising outcomes for reduced toxicity conditioning reported here and elsewhere [23] support the notion that such regimens are preferable for this disease, despite a higher frequency of mixed chimerism. Aware of the relatively small cohort analyzed, we could not demonstrate a consistent detrimental effect of mixed chimerism on immunological and clinical outcomes, especially the frequency of infections. Still, further research is needed to investigate the effect of mixed chimerism in DOCK8 deficiency, especially on cell lineages not covered in this report. Given this uncertainty, we advocate aiming for complete donor chimerism in treating DOCK8 deficiency, but suggest aiming at reduced toxicity over myeloablation, especially for patients with pre-existing organ damage.