Improved outcome of HSCT in STAT1 gain-of-function disease following JAK inhibition bridging

Autosomal dominant gain-of-function (GOF) mutations in signal transducer and activator of transcription 1 (STAT1) were identified in 2011 as a cause of chronic mucocutaneous candidiasis (1, 2). Since then, it has been recognized that STAT1 GOF disease is an inborn error of immunity with severe immune dysregulation and various effects on humoral and adaptive immunity that lead to increased susceptibility to a variety of bacterial, viral, and fungal pathogens. The occurrence of vascular aneurysms and carcinoma are associated with particularly poor outcome (3). Hyperinflammation and organ-specific and systemic autoimmunity are also a key feature of STAT1 GOF disease. This is mediated through increased signaling through the Janus kinase (JAK)/STAT pathway and an enhanced response to cytokine signaling, including type I interferons (IFNs) (4). Emerging reports show promising results with the use of JAK inhibitors to treat STAT1 GOF disease (5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16). However, it requires careful balancing to prevent severe infectious complications (17). In addition, not all patients have complete remission of symptoms, and it is not yet clear if prolonged treatment with these agents will result in sustained long-term disease control and better outcome (7).

Early reports of hematopoietic stem cell transplantation (HSCT) in STAT1 GOF disease, which was often performed in patients with severe disease manifestations, had mixed outcomes (6, 16, 18, 19, 20, 21, 22). In the largest series of transplanted patients with STAT1 GOF disease reported thus far (N = 15) (18), there was an abysmal overall survival (OS) of 40% and an event-free survival (EFS) of only 10%, mainly related to high rates of secondary graft failure. Although not completely understood, the hyperinflammatory state associated with the enhanced response to circulating cytokines, such as IFNs, is thought to contribute to the poor outcome of HSCT for STAT1 GOF disease, as is also seen in other IFN-driven diseases (23). Because JAK inhibition for treatment of STAT1 GOF disease is effective in controlling immune dysregulation, treatment with JAK inhibitors has been increasingly used as a “bridge” to HSCT to optimize clinical condition prior to HSCT. In a recent report on JAK inhibitor treatment in patients with STAT1 or STAT3 GOF disease, transplanted patients pre-treated with JAK inhibitors had an excellent OS of 91% (7).

Here, we report the latest outcome data of patients transplanted for STAT1 GOF disease from an international cohort representing 14 countries and give an update on the historical cohort in this joint Primary Immune Deficiency Treatment Consortium (PIDTC)/European Society for Blood and Marrow Transplantation – Inborn Errors Working Party (EBMT-IEWP) retrospective study.

36 patients with STAT1 GOF disease who underwent HSCT since 2010 were included (Table 1). Range of follow-up of surviving patients is 114 days–8 years. 15 patients (42%) were female and 21 (58%) were male. Data of 7 patients previously reported by Leiding et al. (18) were included (with follow-up clinical data if still alive) in the new eCRF to allow for an integral analysis. Patients reported in other manuscripts are also shown in Table 1. 20 patients had defects in the DNA-binding domain, 14 in the coiled-coil domain (CCD), and one each in the SH2 or TAD domain. The median age at first symptoms of STAT1 GOF disease was 0.8 years (interquartile range [IQR] 0.4–1 year), the median age at diagnosis was 7 years (IQR 2.25–9.3 years), and the median age at first HSCT was 11 years (IQR 6.25–16 years) (Fig. S1).

All patients had infections prior to first HSCT, with a median age of first infection of 0.8 years (Table S2 and Fig. S2). Mucocutaneous fungal infections were most common, in 34/36 (94%) patients. Nine of 36 patients (25%) had invasive fungal infections. Other common infections were skin and lung infections, often caused by Staphylococcusaureus, in 11/36 patients (30.5%). Autoimmune disease was present in 25/36 (69%) patients, with a median age of onset of 3.7 years (Table S3), most commonly thyroid disease and autoimmune cytopenia. Immunosuppressive medication was prescribed at any time prior to first HSCT in 26/36 (72%) patients, most commonly steroids and/or a JAK inhibitor. Two patients had hemophagocytic lymphohistiocytosis, both at 10 years of age. Two-thirds of patients had an antibody deficiency (Table S4), and immunoglobulin replacement therapy was prescribed prior to first HSCT in 25/36 (69%) patients. Two-thirds of patients (22/36, 61%) had a history of failure to thrive. Many patients had chronic lung disease, including bronchiectasis (15/36, 42%). Four patients (11%) had aneurysms prior to first HSCT, at a median age of 12.5 years (range 0.9–35 years): three cerebral and one extracerebral aneurysm. There were no patients with malignancies in this cohort. Most patients (69%) had an hematopoietic cell transplantation-specific comorbidity index (HCT-CI) score of 1 or higher prior to first HSCT, all but two patients had been hospitalized at least once, and only 8/36 (22%) had a Karnofsky (adults) or Lansky (children) performance status of 100 (Table S5).

20 (56%) patients were treated with ruxolitinib leading up to their first HSCT. One patient received baricitinib in the past, but not directly prior to HSCT. Ruxolitinib was started a median of 6 mo prior to the first HSCT (IQR 4–18 mo). A wide range of doses was administered, with a median of 15 mg/m²/day (IQR 9–26 mg/m²/day). 13 of the 20 ruxolitinib-pre-treated patients (65%) received concomitant fluconazole, itraconazole, voriconazole, or posaconazole, which are presumed to increase exposure to ruxolitinib by about twofold. Ruxolitinib was stopped in 18 patients directly prior to start of the conditioning regimen. One patient continued ruxolitinib for graft versus host disease (GVHD) prophylaxis after the infusion of the stem cells, and time of discontinuation is unknown for one patient.

There were 40 transplant procedures in 36 patients (Table 1). Indications for first HSCT were combined immunodeficiency in 14/36 (39%), autoimmunity and/or IPEX-like disease in 10/36 (28%), and severe infections in 12/36 (33%) of patients. Combined immunodeficiency was characterized as quantitative or functional defects in humoral and adaptive immunity; IPEX-like disease was a predominant presence of autoimmunity and hyperinflammation. Indications for re-transplant were graft failure (n = 3, one with primary graft failure and two with secondary graft failure) or relapse (n = 1, decreasing donor chimerism with recurrence of STAT1 GOF symptoms). Most patients received myeloablative conditioning (24/40 or 60% of transplant procedures) or myeloablative reduced-intensity conditioning (10/40 transplant procedures or 25%). In 6/40 transplant procedures, non-myeloablative conditioning was given because of poor clinical condition of the patient, recent transplant with graft failure, and/or very young age. Serotherapy was administered as part of the conditioning regimen in 36/40 (90%) HSCT procedures (ATG n = 23, alemtuzumab n = 13). Graft source was unmanipulated bone marrow (BM) in 20/40 (50%) HSCT procedures, PBSC in 15/40 (38%), of which 4 were CD3+TCR alphabeta depleted, CB in 3/40 (8%), and manipulated BM in 2/40 (5%). A calcineurin inhibitor plus mycophenolate mofetil (13/40) or methotrexate (12/40), with or without other agents, were the most commonly used GVHD prophylaxis regimens (Table S6). In four HSCT procedures, posttransplant cyclophosphamide was used, all in a mismatched donor setting.

Probability of achieving engraftment by 100 days after HSCT were 87.2% for absolute neutrophil count (ANC) 0.5, 74.3% for platelet count (PLC) 20, and 66.8% for PLC 50 (Fig. S3). After the first HSCT, 24/36 (64%) of patients achieved durable full donor chimerism (Table S7). Overall, 27/40 (68%) transplant procedures resulted in full donor chimerism at last follow-up (Table S8). Acute GVHD occurred after 18/40 (45%) HSCT procedures, but only 6/40 (15%) were grade III or IV. Chronic GVHD (liver) occurred in one patient.

OS of this cohort was 72.2%, and EFS was 55.6% (Fig. S4). Causes of death were sepsis with HLH in one patient, pulmonary infection in one patient, likely pulmonary transplant-associated thrombotic microangiopathy in one patient, intracranial hemorrhage due to cerebral aneurysm in one patient, infection with inflammatory pneumonitis in one patient, immune cytopenia with infection in one patient, graft failure with infection in two patients, and acute GVHD with multi-organ failure in two patients. In Table 2, the outcomes of all individual patients are listed. There was no significant difference in OS according to age at transplantation (above or below the median of 11 years).

In univariate analyses, pre-treatment with ruxolitinib was associated with a trend for better OS (P 0.0671) and was significantly associated with better EFS (P 0.0044) (Fig. 1). OS was 84% in the ruxolitinib-treated group compared with 54% in the untreated group, and event survival was 73% versus 30%, respectively. Similarly, there was a trend for patients transplanted after 2018 to have a better OS (P 0.0645) and a significantly better EFS (P 0.0284). In a multivariate Cox regression analysis with both ruxolitinib and alemtuzumab and HSCT before or after 2018 as covariates, only ruxolitinib pre-treatment was significantly associated with better EFS (HR 0.286, 95% CI 0.092–0.894, P = 0.031). There was no significant difference in time to engraftment of neutrophils or platelets and no difference in the number achieving full donor first chimerism between those pre-treated with ruxolitinib and those not treated.

There were no differences in OS or EFS for patients with HLA-identical versus unrelated matched or mismatched donors, primary indication for HSCT (autoimmunity/IPEX-like disease, CID, or severe infections), or STAT1 mutation domain (BDB domain, CCD, or other) (Fig. S5 A). Patients with a HCT-CI score of 1 or higher had worse EFS (P 0.0454) and a trend for worse OS (P 0.10) (Fig. S5 B). Patients receiving myeloablative conditioning regimens had similar outcome as patients receiving reduced-intensity myeloablative conditioning regimens. Patients receiving non-myeloablative conditioning had worse event-free and OS. 2 of the 18 patients in the ATG group received non-myeloablative conditioning versus 0/12 in the alemtuzumab group, but this difference was not statistically significant (Fisher’s exact test P value 0.50). Patients receiving alemtuzumab had better OS and EFS as compared with patients receiving ATG or no serotherapy (P 0.0070 and 0.0737, respectively).

Allogeneic HSCT is a curative treatment for STAT1 GOF disease, but experience is still limited, in part because historical data on outcome of HSCT for this condition were discouraging. In the previous cohort study from Leiding et al. (18), many patients suffered from secondary graft failure, and OS was only 40%. Here, we analyzed data from patients with STAT1 GOF disease transplanted from 2010–2023. Data from patients previously reported were also reentered, to allow for an integral analysis. Most patients in the current study were severely affected. All patients had infections, the majority had autoimmunity, many had chronic lung disease and failure to thrive, and all but two patients had been hospitalized at least once prior to HSCT. Despite this high burden of disease, outcome after HSCT was markedly improved as compared with the previous report, with 72.2% OS and 55.6% EFS.

Targeted therapies such as JAK inhibitors are now accessible in many countries worldwide and allow for better control of inflammatory symptoms with few medication side effects when modulating the JAK-STAT pathway in patients with STAT1 GOF disease (5, 7). Despite these therapeutic advances, long-term experience with JAK inhibition in this rare inborn error of immunity is still missing. Transplanted patients that were pre-treated with the JAK inhibitor ruxolitinib had significantly better EFS than patients not receiving ruxolitinib. Also, there was a trend for better OS.

In this cohort, posttransplant infections, with or without cytopenia or graft failure, were an important cause of death, in 6/10 deceased patients. Two patients died because of complications of severe acute GVHD. Six of 40 HSCT procedures (15%) resulted in graft failure. Although patients receiving a transplant after 2018 and those receiving alemtuzumab also had better outcome, multivariate analysis, including these possible confounders as covariates, showed that only ruxolitinib pre-treatment was significantly associated with better EFS. We hypothesize that the better EFS in patients pre-treated with JAK inhibitors is due to a better control of immune dysregulation prior to transplant, resulting in less graft failure and less inflammatory complications posttransplant in patients treated with a JAK inhibitor. Perhaps this is at least partly mediated through a reduction in IFN-γ levels peri-transplant, thereby preventing IFN-γ-mediated graft failure (23). Another way to reduce IFN-γ levels is by adding the anti-IFNγ antibody emapalumab to the conditioning regimen, which was done in one patient in the current cohort (patient 22). This patient is currently doing well, over a year post-HSCT (20).

At last follow-up, 21/36 patients (58%) were alive and well, and 5/36 (14%) were alive with complications: poor immune reconstitution in one, kidney failure in one, bronchiectasis in two (previously present in 1, but not the other patient), and chronic GVHD with graft failure in one. Most surviving patients had full donor chimerism. One of the surviving patients with mixed chimerism was alive and well at last follow-up; the other surviving patient with mixed chimerism was alive with bronchiectasis. Because most patients in this study had full donor chimerism at last follow-up, we were not able to assess if mixed donor chimerism is sufficient to control STAT1 GOF disease symptoms.

Patients receiving non-myeloablative conditioning regimens had worse outcome than patients receiving (reduced intensity) myeloablative regimens. This is probably confounded by the fact that patients with the highest morbidity were not in a clinical condition to receive more intensive conditioning. Patients receiving alemtuzumab had better OS and EFS than patients receiving ATG or no serotherapy, perhaps because of better control of peri-transplant inflammation. Donor match grade was not associated with OS or EFS, although the numbers of patients in each donor group were small. Similarly, there were no differences in outcome depending on the primary indication for HSCT or which protein domain was affected by the STAT1 mutation.

Due to improved and accessible genetic diagnostics in clinical practice, patients with inborn errors of immunity, such as STAT1 GOF disease, can be diagnosed earlier following symptom onset. Here, we show that those STAT1 GOF patients who had most disease-related complications and therefore higher scores on the HCT-CI had worse EFS and a trend for worse OS. However, it remains difficult to determine the optimal timing for HSCT in these patients. As in other inborn errors of immunity, the choice for or against curative cell therapy depends on the clinical condition of each patient with the aim not to miss a window of opportunity in which this therapeutic option can be safely offered. Here lies the dilemma for patients and treating physicians: the outcome of HSCT is usually better when the disease is not yet advanced, while transplant-associated complications increase and are more difficult to predict in patients with permanent or progressive organ dysfunction, such as chronic lung or liver disease or chronic infections. Also, our data show that pretransplant control of immune dysregulation improves the\ outcome of patients with STAT1 GOF disease undergoing HSCT, which may also be true for other inborn errors of immunity with immune dysregulation.

In conclusion, allogeneic HSCT is a curative and safe treatment option for patients with STAT1 GOF disease, particularly for those patients without significant comorbidities. Pre-HSCT use of JAK inhibitors is recommended to control disease-related symptoms, reduce the possibility of graft failure, and improve survival.

We included 36 patients with STAT1 GOF disease who underwent HSCT regardless of outcome after 2010 from PIDTC and EBMT-IEWP centers. Inclusion was also solicited by announcement through the EBMT-IEWP and the PIDTC as well as by contact to centers of all published cases. Informed consent of participants was obtained according to the local regulations of the participating centers, in accordance with the Declaration of Helsinki.

Data were collected via Castor Electronic Data Capture (26), using a shared eCRF. Basic demographic data included age, sex, genetic diagnosis, age at diagnosis, age at HSCT, time from diagnosis to HSCT, comorbidities, and year of HSCT. Transplant-specific demographics included donor type, cell source, conditioning intensity, and use of serotherapy. We also collected data on prior usage of immunosuppressive treatment, including treatment with JAK inhibitors. We collected data on HSCT complications, outcome including clinical condition at last follow-up, and donor chimerism. An attempt was made to obtain follow-up clinical data from the patients reported by Leiding et al. (18) and for those transplanted after 2010, a request was made to reenter data in the current eCRF (Table S1).

For all patients, pretransplant disease burden was assessed by calculating the HCT-CI (27). The conditioning regimen was classified as myeloablative (regimen A or B from the EBMT/ESID IEWP guidelines paper [28]), myeloablative reduced intensity (regimen C or D), or reduced intensity/non-myeloablative (regimen E or F). Neutrophil engraftment was defined as the first of three consecutive days with an ANC of 0.5*10⁶/L or higher (ANC 0.5). Thrombocyte engraftment was defined as the first of three consecutive days with PLCs of 20*10⁹/L (PLC 20) or 50*10⁹/L (PLC 50), without transfusion support for the last 7 days. Chimerism was reported as either whole-blood or lineage-specific analysis as per center protocol. Full donor chimerism was defined as >95% donor chimerism, mixed chimerism as 5–95% donor chimerism, and complete autologous reconstitution as >95% patient chimerism. For grading of acute GVHD, the modified Glucksberg criteria were used (29). Chronic GvHD was assessed according to the 2014 NIH consensus statement (30).

Categorical variables were expressed as the number and percentage of the group. Continuous variables were expressed as median with IQRs. Primary endpoints were OS and EFS since date of first HSCT. For OS, an event was defined as death of any cause. For EFS, an event was defined as graft failure, re-transplant, acute GVHD grade III or IV, moderate to severe chronic GVHD, or death of any cause, whichever occurred first. Patients without an event were censored at last follow-up. OS and EFS were determined using the Kaplan–Meier estimator, and subgroups were compared using the log-rank test. A multivariate Cox regression analysis was done to evaluate the association of pre-treatment JAK inhibitor use, alemtuzumab use, and HSCT before or after 2018 with survival. 2018 was the median year of transplant in our series. Statistical analyses were done using IBM SPSS Statistics version 29.0.0.0 and GraphPad Prism version 9.3.1.

In Table S1, the follow-up information on the patients reported by Leiding et al. (18) is given. In Tables S2, S3, S4, and S5 and in Figs. S1 and S2, detailed information on the pretransplant features of the patients is given. Tables S6, S7, and S8; and Figs. S3, S4, and S5 have additional information on transplant procedures and outcome.

Data are available upon request due to legal/ethical reasons.

Data are available upon request due to legal/ethical reasons. The authors wish to thank all participating PIDTC and EBMT-IEWP transplant centers for their cooperation in this study and Estefanía Laney of the Willem-Alexander Children’s Hospital, LUMC, Leiden, for help in designing the Castor eCRF.

This study was supported by the National Institutes of Health (NIH), Division of Allergy, Immunology and Transplantation, National Institute of Allergy and Infectious Diseases (NIAID), NIH, Office of Rare Diseases Research (ORDR), National Center for Advancing Translational Sciences (NCATS), Bethesda, MD: NIH NIAID grant U54AI082973, and flow cytometer LSR FORTESSA grant number “DFG INST 269/866-1 FUGG project number 446167311.” The PIDTC is a part of the Rare Diseases Clinical Research Network of ORDR, NCATS.

Author contributions: E.P. Buddingh: conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, supervision, validation, visualization, and writing—original draft, review, and editing. M. Slatter: conceptualization, visualization, and writing—original draft, review, and editing. J.C. Aldave Becerra: resources and writing—review and editing. L. Alonso Garcia: investigation and validation. E.v. Asmuth: formal analysis and writing—review and editing. S. Baris: resources and writing—review and editing. O.d.l. Calle-Martin: resources. A.Y. Chan: data curation and writing—review and editing. S.-W.B. Chan: resources and writing—review and editing. S. Chandrakasan: investigation and resources. D. Chellapandian: resources and writing—review and editing. J. Dara: writing—review and editing. S. Farmand: investigation and writing—review and editing. A. Fasth: conceptualization, data curation, investigation, resources, validation, and writing—review and editing. L.F. Satter: data curation, investigation, and writing—review and editing. R. Formankova: data curation. E. Grunebaum: data curation, investigation, resources, and writing—review and editing. S.J. Keogh: resources and writing—review and editing. A. Kiykim: data curation, resources, validation, and visualization. J.-S. Kuhl: data curation, investigation, resources, validation, and writing—review and editing. A. Laberko: resources and writing—review and editing. T.R. Leahy: data curation and writing—review and editing. C. Lindemans: resources and writing—review and editing. C. Martinez: conceptualization, data curation, resources, and writing—review and editing. L. Martinez-Martinez: resources and writing—review and editing. W.G. Mitchell: conceptualization, data curation, and writing—review and editing. E. Morris: investigation, project administration, and writing—review and editing. J.H. Oved: data curation and writing—review and editing. M. Polacik: resources. J.G. Riviere: data curation, investigation, and writing—review and editing. C.M. Roifman: investigation, resources, and writing—review and editing. S.S. Kilic: investigation and resources. P. Sedlacek: data curation and investigation. A.J. Shah: writing—review and editing. L. Vong: investigation and writing—review and editing. A.C. Lankester: resources and writing—review and editing. M.H Albert: project administration, resources, supervision, and writing—review and editing. B. Neven: conceptualization and resources. T. Torgerson: conceptualization and writing—review and editing. J. Leiding: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, supervision, validation, visualization, and writing—original draft, review, and editing. C. Schuetz: conceptualization, data curation, investigation, project administration, supervision, and writing—original draft, review, and editing.