JAK1/2 inhibition with baricitinib in the treatment of autoinflammatory interferonopathies

BACKGROUND. Monogenic IFN–mediated autoinflammatory diseases present in infancy with systemic inflammation, an IFN response gene signature, inflammatory organ damage, and high mortality. We used the JAK inhibitor baricitinib, with IFN-blocking activity in vitro, to ameliorate disease. METHODS. Between October 2011 and February 2017, 10 patients with CANDLE (chronic atypical neutrophilic dermatosis with lipodystrophy and elevated temperatures), 4 patients with SAVI (stimulator of IFN genes–associated [STING-associated] vasculopathy with onset in infancy), and 4 patients with other interferonopathies were enrolled in an expanded access program. The patients underwent dose escalation, and the benefit was assessed by reductions in daily disease symptoms and corticosteroid requirement. Quality of life, organ inflammation, changes in IFN-induced biomarkers, and safety were longitudinally assessed. RESULTS. Eighteen patients were treated for a mean duration of 3.0 years (1.5–4.9 years). The median daily symptom score decreased from 1.3 (interquartile range [IQR], 0.93–1.78) to 0.25 (IQR, 0.1–0.63) (P < 0.0001). In 14 patients receiving corticosteroids at baseline, daily prednisone doses decreased from 0.44 mg/kg/day (IQR, 0.31–1.09) to 0.11 mg/kg/day (IQR, 0.02–0.24) (P < 0.01), and 5 of 10 patients with CANDLE achieved lasting clinical remission. The patients’ quality of life and height and bone mineral density Z-scores significantly improved, and their IFN biomarkers decreased. Three patients, two of whom had genetically undefined conditions, discontinued treatment because of lack of efficacy, and one CANDLE patient discontinued treatment because of BK viremia and azotemia. The most common adverse events were upper respiratory infections, gastroenteritis, and BK viruria and viremia. CONCLUSION. Upon baricitinib treatment, clinical manifestations and inflammatory and IFN biomarkers improved in patients with the monogenic interferonopathies CANDLE, SAVI, and other interferonopathies. Monitoring safety and efficacy is important in benefit-risk assessment. TRIAL REGISTRATION. ClinicalTrials.gov NCT01724580 and NCT02974595. FUNDING. This research was supported by the Intramural Research Program of the NIH, NIAID, and NIAMS. Baricitinib was provided by Eli Lilly and Company, which is the sponsor of the expanded access program for this drug.

the last dose of investigational product. 5. Have had symptomatic herpes zoster infection within 12 weeks prior to entry or during the screening period. 6. Have a history of disseminated/complicated herpes zoster. 7. Have evidence of active infection, at the time of entry or during the screening period, that, in the opinion of the investigator, would pose an unacceptable risk for participating in the program. 8. Have a history of active hepatitis B, hepatitis C, or human immunodeficiency virus (HIV). 9. Have documented high titer autoantibodies suggestive clinically of autoimmune diseases other than severe JDM. 10. Are immunocompromised and, in the opinion of the investigator, are at an unacceptable risk for participating in the program. 11. Have had a serious systemic or local infection within 12 weeks prior to entry or during the screening period. Exceptions include SAVI patients with infected ulcerative skin lesions, which in the opinion of the investigator would not pose an unacceptable risk for participating in the program. 12. Have been exposed to a live vaccine within 12 weeks prior to entry or are expected to need/receive a live vaccine during the course of the program. 13. Have had household contact with a person with active tuberculosis (TB) and did not receive appropriate and documented prophylaxis for TB. 14. Have a serious and/or unstable illness that, in the opinion of the investigator, poses an unacceptable risk for the patient's participation in the program. 15. Have an estimated glomerular filtration rate (eGFR) based on the most recent available serum creatinine of <40 mL/min/1.73 m2. 16. Have or have had a history of lymphoproliferative disease; or signs or symptoms suggestive of possible lymphoproliferative disease, or active primary or recurrent malignant disease; or been in remission from clinically significant malignancy for <5 years. 17. Have a history of chronic alcohol abuse or intravenous drug abuse within the 2 years prior to entry. 18. Are unable or unwilling to make themselves available for the duration of the program and/or are unwilling to follow protocol restrictions/procedures. 19. Are investigator site personnel directly affiliated with this program and/or their immediate families. Immediate family is defined as a spouse, parent, child, or sibling, whether biological or legally adopted. 20. Are currently enrolled in, or discontinued within the last 30 days from, a clinical trial involving an investigational product or non-approved use of a drug or device (other than the investigational product used in this program), or concurrently enrolled in any other type of medical research judged not to be scientifically or medically compatible with this program.
Entered patients were ineligible for enrollment (that is, ineligible to receive baricitinib) and were discontinued from the program if they met any of the following criteria: 21. Have screening laboratory test values outside the reference range for the population or investigative site that, in the opinion of the investigator, pose an unacceptable risk for the patient's participation in the program. 22. Have any of the following specific abnormalities on screening laboratory tests: • AST or ALT >2 × ULN unless the hepatitis is confirmed as resulting from the autoinflammatory condition. Even if inflammatory myositis is considered present, AST or ALT cannot exceed 5 × upper limit of normal (ULN). • Hemoglobin <10 g/dL (100 g/L). Patients may be enrolled with hemoglobin <10 g/dL if the anemia is considered a result of the underlying disease (see below). • Total WBC count <2500 cells/μL. Patients may be enrolled with WBC count <2500 cells/μL if the low WBC count is considered a result of the underlying disease (see below). • Neutropenia (absolute neutrophil count [ANC] <1200 cells/μL). Patients may be enrolled with an ANC <1200 cells/μL if the low ANC is considered a result of the underlying disease (see below). • Thrombocytopenia (platelets <100,000/μL). Patients may be enrolled with a platelet count <100,000/μL if the low platelet count is considered a result of the underlying disease (see below). • eGFR <40 mL/min/1.73 m2 Note: A patient with CANDLE, CANDLE-related condition, or SAVI may be enrolled with any of the above specific abnormalities on screening laboratory tests if these laboratory abnormalities are considered a feature of the disease. An expert independent of the principal investigator (preferably a hematologist) must evaluate the patient and, in conjunction with the principal investigator, document that the laboratory abnormality is a feature of the underlying CANDLE, CANDLE-related condition, or SAVI condition; the investigator must also consult with the Sponsor before the patient can be enrolled. 23. Have screening thyroid-stimulating hormone and/or thyroxine values outside of the laboratory's reference range and are assessed to be clinically significant. Patients who are receiving thyroxine as replacement therapy may participate in the program, provided stable therapy has been administered for ≥3 months and thyroid-stimulating hormone is within the laboratory's reference range. 24. Have screening electrocardiogram (ECG) abnormalities that, in the opinion of the investigator, are clinically significant and indicate an unacceptable risk for the patient's participation in the program (for example, Bazett's corrected QT interval >450 msec for males and >470 msec for females). 25. Have evidence of active or latent TB as documented by a positive purified protein derivative (PPD) test. Exceptions include patients with a history of latent TB who have documented evidence of completing a course of appropriate treatment: 26. Have a positive test for hepatitis B defined as (1) positive for hepatitis B surface antigen, or (2) positive for anti-hepatitis B core antibody, but negative for hepatitis B surface antibody unless the anti-hepatitis B core antibody is thought to be a false positive result.
In the latter case, confirmation of the presence of hepatitis B virus (HBV) by DNA testing is required. An HBV DNA indeterminate result is considered HBV infection. 27. Have hepatitis C virus (positive for anti-hepatitis C antibody with confirmed presence of hepatitis C virus); have evidence of HIV infection, and/or positive HIV antibodies.

B. CLINICAL BENEFIT ASSESMENT
The outcomes and measurements marked with an asterisk (*) were collected under the natural history protocol or are part of routine care of patient with chronic inflammation who are or are not receiving corticosteroids.
Daily diary score (DDS) assessment: Disease-specific patient diaries (for CANDLE and SAVI patients) were provided for daily collection of information on patients' signs and symptoms. CANDLE and other interferonopathy patients or their parents recorded daily symptoms of fever, rash, musculoskeletal pain, headaches and fatigue; and SAVI patients recorded, fever, rash, musculoskeletal pain, fatigue, respiratory symptoms and severity of ulcers/ischemic lesions. Each symptom was rated on a scale of 0 to 4, with 0=no symptoms, 1=mild symptoms, 2=moderate symptoms, 3=more severe symptoms, and 4=severe symptoms (possible range 0-20 or 24, for CANDLE and other interferonopathy patients, and SAVI respectively). At each visit, the diary score was calculated as follows: a. Average score of each symptom was calculated using data entered since the previous visit and correcting for any day for which diary scores were not recorded. b. The calculated average score for each symptom was summed up and divided by the number of assessed symptoms (i.e. 5 symptoms for CANDLE, 6 symptoms for SAVI) to calculate the average score for each patient.
Steroid reduction: An average diary score <0.5 (CANDLE diary) or <1.0 (SAVI diary) was indicative of a response to treatment and was one criterion used to initiate steroid weaning (if the patient was receiving steroids). Additionally, if the patient was responding to treatment, but did not meet the average diary score threshold to begin steroid weaning, but was experiencing new or worsening clinically significant adverse effects from steroids (including, but not limited to, cataracts, vertebral fractures due to osteoporosis, Cushingoid habitus, substantial weight gain, avascular necrosis, dyslipidemia, hypertension, opportunistic infections or stunted growth), the steroid weaning was permitted. Mean corticosteroid doses were calculated as prednisone equivalents in milligram (mg) per kilogram (kg) per day for each patient at each visit (http://www.medcalc.com/steroid.html) Assessment of Remission*: The duration of remission for patients achieving remission criteria (DDS<0.15, off steroids, and CRP<5mg/L) was assessed by comparing the number of visits that patients fulfilled remission criteria for the visits before and after they first achieved remission criteria.
Disability and Quality of Life Assessments*: Questionnaires assessing disability and quality of life including Childhood Health Assessment Questionnaire (CHAQ), Pediatric Quality of Life Inventory (PedsQL), and physician and patient or parent global assessment (visual analog scale) were all completed at previous NIH visits or at the baseline visit, except for two patients (patient S2 and C10), who completed the questionnaires within 1 or 8 months of baricitinib initiation respectively. Questionnaires were obtained at most follow-up visits.
Height, weight, body mass index (BMI)*, bone age*, bone mineral density by Dualenergy x-ray absorptiometry (DEXA) measurements*, and Z-score calculations: Z-scores for height, weight, BMI and bone age were calculated as indicated in the respective table and figure legends. Patients with open growth plates at entry were considered to have growth potential. These patients were followed longitudinally. Z-scores for bone mineral density were calculated using the Bone Mineral Density in Childhood Study (BMDCS) calculator (https://bmdcs.nichd.nih.gov/zscore.htm) for patients between 2 and 19 years at the time of enrollment, Z-scores were adjusted for height. For patients older than 19 years, the NIH reported Z-scores were used. All patients had DEXA scans at the baseline visit or within 6 months of treatment initiation (except for two patients, patients C1 and C6 who obtained DEXAs 1 and 2.7 years post baricitinib respectively) that were used as "baseline". DEXA scans were repeated yearly as clinically indicated.

CANDLE-specific outcomes:
The outcomes and measurements marked with an asterisk (*) were collected under the natural history protocol or are part of routine care of patient with chronic inflammation who are or are not receiving corticosteroids. Assessment of Lipid Profile*. Total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and triglycerides (TG) were obtained at baseline and at every visit (see Table S8). Hyperlipidemia is defined as high LDL and/or high TGs. NIH laboratory references were used for normality https://ccinternal.cc.nih.gov/LTGRA/UL/public_labtest_detail.

Assessment of Hepatic Steatosis by conventional In-Phase and Opposed-Phase MRI*:
Liver fat was detected based on the relative signal loss on opposed-phase ("out-of-phase") images. Abdominal magnetic resonance imaging (MRI) were done every 6 to 12 months after treatment initiation. Assessment of Myositis by MRI*: Six CANDLE patients had MRIs within 13 months of baricitinib initiation that confirmed clinical suspicion of myositis with characteristic muscle enhancement on MRI.

SAVI-specific outcomes:
Chest Computed tomography (CT) scoring*: 20 Chest CT images from 4 SAVI patients collected at baseline and throughout the duration of treatment were scored by one radiologist (LF), who was blinded to the clinical data and the order of the scans. Inflammatory and lung damage findings were scored based on severity, with absent=0, mild or barely perceptible=1, moderate or obvious=2, severe or striking=3. Inflammatory findings that were scored included: 1. presence of ground glass opacities, 2. intralobular septal thickening, 3. pulmonary nodules, 4. consolidation, 5. atelectasis, 6. pleural effusion and 7. lymph nodes. Lung damage findings that were scored include: 1. presence of pneumothorax, 2. bronchiectasis, 3. parenchymal cyst(s), 4. subpleural cyst, and 5. vascular calcifications. Both lungs were scored together, the scores for the 7 inflammatory categories were added, the possible range of the inflammatory score is 0 (no abnormalities) to 21 (most severe score for all seven categories). The possible range of the summary damage score is 0 (no abnormalities) to 15 (most severe score for all 5 categories). For statistical analyses, the baseline CT scores were compared to the CT scores obtained at last visit included in the analysis up to February 2017.

D. IMMUNOLOGICAL ASSESSMENTS
1. Inflammatory markers*. Acute phase reactants including high sensitivity (hs) C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), were collected at each visit under the NIH Natural History Protocol and compared to pre-treatment values.
2. Hematologic values. Hemoglobin (Hgb), white blood cell count (WBC), platelet count (Plts), and absolute lymphocyte count (ALC) were assessed at each visit. In patients with cytopenias at baseline, improvement of the cytopenias was an outcome measure. These parameters were also monitored for drug safety in all patients. 5. 25-gene IFN score determination*. RNA was prepared from peripheral blood collected in PAXgene tubes (Qiagen). A nanostring assay with 25-interferon response gene score (manuscript in press) was assessed at baseline and at most follow up visits until October 31, 2016. Based on control data, a normal IFN score was defined as below 44 (cut off is 95%ile in healthy controls).
6. Cytokine Analysis by Luminex*. A 27-plex assay that included the IP-10 analyte was performed on serum samples from all patients that were collected until December 2015 (Thermo Fisher, Raleigh, NC).

E. SAFETY ASSESSMENTS
1. Assessment of lipid levels and lipid ratios. Total/HDL and LDL/HDL cholesterol ratios are risk indicators with greater predictive value than isolated parameters used independently, particularly LDL. To assess cardiac risk, the TC/HDL-C ratio, the TG/ HDL-C ratio, and the LDL-C/H LDL-C ratios were calculated for each patient. (see Supplemental  Table 8).
2. BK titer assessment in blood and urine by polymerase chain reaction (PCR) assay* were first assessed in the context of clinical care in the first identified patient (C7) in June 2015.
When BK viral copies were elevated in the blood and urine, BK assessments were included in the routine safety evaluations under the Expanded Access program. BK titers in blood were retrospectively assessed in serum samples collected prior to baseline, at enrollment, and the different protocol visits, baseline urine samples were not collected and therefore baseline BK urine titers could not be obtained. Serial prospective Haufen testing as described below was also performed after identification of the first patient.

F. STATISTICAL ANALYSES
Per protocol, if more than 5 patients were enrolled into this program, the protocol defined that "…2-dimensional plots of various data may be utilized to explore the relationship between variables of interest including final dose level versus efficacy measures to explore recommended dosing guidelines, and plots of efficacy measures versus laboratory measures to explore risk/benefit relationships." Assessment of the duration of remission. 2x2 tables were calculated for the number of visits that each measure (steroid dose = 0 mg, DDS < 0.15 and hsCRP < 5mg/dL) fulfilled remission criteria and were compared before and after patients first reached remission criteria (two periods). The Cochran-Mantel-Haenszel test was used to analyze the association between these binary outcomes across patients.

I. Supplemental Tables
Supplemental Table 1: Baricitinib initial dosing, time to reach optimal tolerated doses Individual listings of weight, eGFR, initial dosing, time on the program, date when optimal tolerated dosing was achieved and final dosing regimen. eGFR -estimated glomerular filtration rate. QD -once a day dosing, BID -twice a day dosing and TID -three times a day dosing. Duration indicates the number of days a patient has been on the program at time of analysis.

Supplemental
Complete Count indicates the number of days when diary entries were made.
Complete% denotes the percentage of days a diary entry was made.
Some Missing Count indicates the number of days when diary entries were missing one or more components of the DDS.

Some Missing % indicates the percentage of incomplete DDS entries.
All Missing Counts indicates the number of days without any diary entry.
All Missing % indicates the percentage of days without any diary data.

Supplemental Table 3: Summary of primary outcomes (diary score and steroid dose), and on achieving clinical remission
A Diary score reduction criteria are a mean daily diary score of <0.5 for CANDLE and other IFNopathy, or <1 for SAVI. B Prednisone reduction criteria is at least 50% decrease from baseline or < 0.15mg/kg/day. All CANDLE and other IFNopathy patients, except one (C10) were on oral corticosteroids at baseline, 1 SAVI pt. (S1) was on prednisone as well. C Remission Criteria: 5 CANDLE patients (C2, C4, C5, C9, and C10) achieved remission criteria at their last visit with a mean diary score < 0.15, no prednisone, and a CRP < 5 mg/L. All CANDLE patients who achieved remission had mutations in PSMB8 the inducible proteasome components that are constitutively expressed in blood cells but are inducible in non-hematopoietic cells. The other 3 CANDLE patients (C3, C6 and C8) with digenic disease, fulfilled criteria for improvement in diary scores and had decreased steroid doses. Patient C1, who is compound heterozygous for PSMB4 did not fulfilled criteria for improvement in diary score, he continues to be on prednisone doses of 0.27 mg/kg/day which was decreased from 0.84mg/kg/d at baseline. Pt. C7 developed BK viremia and azotemia and was discontinued from the program. D 2 patients (pts.) (O1 and O3) discontinued after 244 and 98 days on the program (77 and 56 days on optimal tolerated dose). Pt. O1 discontinued due to lack of efficacy and Pt. O3 discontinued due to osteonecrosis; he also had an unsatisfactory response to treatment. No Yes S1

Supplemental Table 4: Stable and continued remission in 5 patients with CANDLE who achieved remission criteria on baricitinib treatment
* P-value based on Cochran-Mantel-Haenszel statistical test comparing proportions before and after remission with the 5 patients combined. All patients were in optimal tolerated doses of baricitinib at the time of remission. The cut off for a normal 25-gene IFN score was 42 at the 95 th % for healthy controls.
No patient restarted corticosteroids and fulfilled remission criteria at 82.1% of their follow up visits.

Supplemental Table 5: Changes in height Z-scores on baricitinib
Pre-baricitinib refers to baseline values that were collected prior to starting baricitinib. At the time of the last measurement, patients were on optimal tolerated doses of baricitinib for at least 1.5 years (excluding patients O1 and O3). Clinical significant improvement in the height Z-scores and percentiles of patients with growth potential (n=13) was seen, when comparing pre-baricitinib to last visit on baricitinib data. Mean height Z-scores improved from -4.03 ± 2.64 to -3.19 ± 2.33; with "catch up growth" observed in 9 patients, their improvement translates into a mean height percentile increase from the 1.4 th percentile to 7.2 th percentile. When 4 patients, 2 who discontinued from the program (O1 and C7) and 2 who were not able to wean steroids to below 0.16 mg/kg/day (C1 and S1) were excluded, the change in Z-score became more significant. A Indicates comparison between baseline to last clinic values, 2-sided p-value are denoted B Analysis of data excluding two patients (O1 and C7) who discontinued from the program. C Analysis of data excluding four patients (O1, C7 who discontinued and C1, and S1); all were unable to wean steroids to below 0.16 mg/kg/day

Supplemental Table 6: Changes in Dual-energy x-ray absorptiometry (DEXA) Zscores on baricitinib treatment
A At least 2 DEXA scans were available for 15 patients, a second DEXA scan was missing on the 3 patients who discontinued treatment, C7, O1 and O3. For 11 patients DEXA scans were obtained prior to or within 4 weeks of baricitinib initiation. Patients C2, C3, S1 and S2 had DEXAs within 5 months prior to baricitinib initiation, and pt. O4, 9 months prior to baricitinib initiation. Two patients (C4 and C9) had their first DEXA scan within 6 months of treatment, and two other patients (C1 and C6), at 1 and 2.7 years after treatment initiation respectively. Height adjusted Z-scores were used for all except for pt. C8, a developmentally severely delayed pre-pubertal girl with a chronological age of 14.3 years and a mean height and bone age of 2 years, in her, we used the bone age for Z-score calculation. The site of the worst Z-score was used for comparisons (n=8 left femoral neck, n=5 AP spine (L1-L4) and n=2 total hip). At the time of their first DEXA scan, 8 out of 15 patients (53%) had osteoporosis (Z-score less than -2.5) with a mean Z-score of -4.64 ± 1.67; 2 out of 15 patients (13%) were osteopenic (Z-scores between -2.5 and -2.0) with a mean Z-score of -2.15 ± 0.21 and 5 patients (33%) had normal Z-scores. Three patients (C1, S1 and O4) with osteoporosis were on bisphosphonates, pt. O4 was started 30 days prior to baricitinib, patients S1 and C1, 11 and 24 months after starting baricitinib treatment. B Comparison between first available DEXA vs. last visit, patients with osteoporosis and osteopenia were grouped for this analysis. p-values were not adjusted.

Supplemental Table 9: Changes in muscle enzymes in CANDLE patients
Myositis is a common feature in CANDLE patients; 8 of the 10 CANDLE patients had a history of myositis prior to enrollment. In 6 patients, myositis was demonstrated on MRI (4 patients had an MRI at baseline, and 2 patients within 13 months after baricitinib initiation). Muscle enzymes including creatinine kinase (CK), lactate dehydrogenase (LDH), aldolase, alanine transferase (ALT) and aspartate aminotransferase (AST), were assessed at each visit. ALT and AST improved in all CANDLE patients, except two, patient C4 who developed hepatic steatosis and patient C9 who had transient elevation of liver enzymes at the last visit, probable related to the use of protein supplements at that visit. After discontinuation of supplements LFTs normalized. Aldolase and LDH levels also decreased. CK levels however increased. At the last visit, 6 out of 10 patients had an elevation of the creatinine kinase without clinical evidence of myositis. The cause for CK elevations is likely multifactorial, CK elevations are also seen in SAVI patients who do not have myositis, and may represent an increase in muscle mass due to improved growth and increase in physical activity, or as an effect of baricitinib treatment (see in safety section). A Pre-baricitinib values were compared to post baricitinib values (last visit), p-value were not adjusted.  ), p-value were not adjusted. C Patients S3 and S4 were non-ambulatory and unable to complete the 6MWT at baseline. After 3 months on baricitinib, patient S4 was able to do his first 6MWT, this was considered baseline for this analysis. D Damage score was based on the presence of pneumothorax, bronchiectasis, parenchymal cyst, subpleural cyst, and vascular calcifications, each category was scored as absent=0, mild or barely perceptible=1, moderate or obvious=2, severe or striking=3. Damage score range from 0-15. E Inflammatory score was based on the presence of ground glass opacities, intralobular septal thickening, pulmonary nodules, consolidation, atelectasis, pleural effusion and lymph nodes. The inflammatory score ranges from 0-21.

Supplemental
The "worsening in post-dyspnea scores" likely reflects the fact that patients are now able to exert themselves and walk further, which they have not been able to do before treatment.

Supplemental Table 12: Association of IFN biomarkers (IP-10 and 25-gene IFN score) with Clinical Outcomes and with historical biomarkers (ESR and CRP)
A correlation values are shown B p-values of the slope from the linear mixed model analysis (see Statistical analysis of the Supplementary Methods) are denoted. Correlations between IP-10, 25-IFN Score, the inflammatory disease markers (erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP)) and the clinical outcomes (average daily diary score (DDS), and daily dose of steroids) were assessed. While IFN biomarkers and conventional biomarkers correlated similarly with patients' symptoms, the IFN biomarkers correlated better with the ability to wean corticosteroids.  C2, O3 A : each patient met criteria on 1 occasion post-baseline which resolved to above criteria without interruption of baricitinib S1: patient met criteria on 1 occasion postbaseline which resolved to above criteria with dose reduction Lymphopenia (lymphocyte count <500 cells/µL) 2 C7 A : met criteria on 4 occasions postbaseline which resolved to above criteria without interruption of baricitinib S3: met criteria on 2 occasions post-baseline which resolved to above criteria without interruption of baricitinib Thrombocytosis (platelet count >600,000/µL) 4 O2: elevated platelets at baseline that increased over time S1: met criteria on 2 occasions post-baseline which resolved to below criteria with dose reduction S4: met criteria on 1 occasion post-baseline which resolved to below criteria without interruption of baricitinib O4: met criteria at last 2 post-baseline observations at time of analysis Thrombocytopenia (platelet count <75,000/µL) 2 O3 A : had pre-existing thrombocytopenia with low platelet count (36,000/µL) at time of discontinuation due to osteonecrosis C7 A : met criteria on 1 post-baseline occasion which resolved to above criteria with dose reduction prior to discontinuation due to acute kidney injury    BK virus is ubiquitous and typically acquired in childhood. 1 The infection is typically asymptomatic and viral reactivation manifesting as asymptomatic viral shedding in the urine is not uncommon in immunocompetent individuals but increased in immunocompromised patients. [2][3][4] Symptomatic reactivation and the development of BK-associated nephropathy are seen in renal and hematopoietic stem cell transplant populations. 5 The presence of viral reactivation in our patients with immune dysregulation and exposure to long-term immunosuppression since childhood might therefore be expected but should be monitored until more data are available.

II. Supplementary Figures
Supplemental Figure 1

. Patient enrollment and disposition
Three patients discontinued the program, one due to non-response to treatment, one due to osteonecrosis; he also had an unsatisfactory response, and one patient due to the development of renal insufficiency and BK viremia and BK viruria.

Supplemental Figure 3. Delay in Bone Age at baseline
Bone Age by Greulich and Pyle was compared to chronological age at time of x-ray. Mean delayed bone age was 3.49 ± 3.99 years (p <0.05). All patients with growth potential were included except for patient O1, no bone age was available for this patient (n=12). X-rays were obtained within 6 months of baricitinib initiation, except for two patients (C3 and C7), their bone age was assessed 3 years and 17 months after program initiation respectively. (C) CANDLE patients (C1, C7 and C8) had heights and weights below the 3 rd percentile, patient (C6) was also below the 3 rd percentile for height, however, her weight was normal. Upon treatment with baricitinib, patients C6 and C8 were able to decrease daily steroids by more than 50% from prednisone dose at baseline. They both showed significant improvement in linear growth and body mass indices. Patient C1 (compound heterozygous for PSMB4) remains steroid dependent at doses > than 0.25 mg/kg/day, despite growth hormone initiation, he has a slow improvement in linear height. Patient C7 was discontinued from the program after the development of azotemia and BK viremia. He expired 4 weeks after discontinuation of baricitinib. DDS denotes daily diary score, Pred, prednisone; APR, acute phase reactants. (D) All SAVI patients except one (S2) were < 18.1 years of age at enrollment and considered to have growth potential. SAVI patients (S3 and S4) had heights and weights below the 3 rd percentile, patient (S1) was below 3 rd percentile for height but weight was normal. Upon treatment with baricitinib, patients S3 and S4 improved linear growth and body mass indexes, however, both continue to be underweight with BMIs < 18.5. Patient (S1) had an improvement in linear height that correlated with a transient decrease in prednisone. Growth velocity slow down and weight increase, after prednisone dose was increase due to subjective reports of respiratory symptoms (PFTs and Chest CT were stable). Patient S2 continued to gain weight on baricitinib, BMI as of last visit was >30 (obese range). A SAVI patients S3 and S4 were not on steroids at the beginning of the program.

Pre-treatment
(E) Patient O4 (homozygous for SAMDH1 deletion) and patient O2 (likely novel mutation) were able to decrease steroids and have an improvement in their linear height. Patient O4 normalized her BMI. Patient O1 discontinued from the program due to lack of efficacy, and patient O3 who also had a poor response, developed multifocal avascular necrosis, and was discontinued due to osteonecrosis. DDS denotes daily diary score, Pred, prednisone; APR, acute phase reactants.

Supplemental Figure 6. Cell subsets in CANDLE, SAVI and other IFNopathies
There were no significant changes on TBNK cell subsets in CANDLE, SAVI and patients with other interferon mediated autoinflammatory diseases treated with baricitinib (n=16). Patients O1 and O3 (discontinued from the study) were not included. Data is presented by disease with CANDLE in red, SAVI in blue and other interferonopathies in green. Darker shades indicate pretreatment and lighter shades last included visit on baricitinib treatment. Means and standard errors are depicted.

Supplemental Figure 7. Quantitative immunoglobulins in CANDLE, SAVI and other IFNopathies
There were no significant changes on quantitative immunoglobulins in CANDLE, SAVI and patients with other interferon mediated autoinflammatory diseases treated with baricitinib (n=16). Patients O1 and O3 (discontinued from the study) are not included. Data is presented by disease with CANDLE in red, SAVI in blue and other interferonopathies in green. Darker shades indicate (B) STAT1 Phosphorylation with IFNa stimulation before baricitinib treatment (historical data) or on low pre-optimal doses of baricitinib (left panels) compared to optimal tolerated baricitinib doses (right panels). As previously reported (Liu et al. NEJM 2014), STAT-1 phosphorylation in SAVI patients before treatment with baricitinib was maximally up-regulated and was not further induced with IFN stimulation (MFI ratios of stimulated over unstimulated cells equaled 1). Post baricitinib, STAT-1 phosphorylation was assessed in 8 CANDLE and 4 SAVI patients. Blood samples were obtained in the morning, before baricitinib dose administration (at the baricitinib trough level). Two healthy controls with repeated measurements (HC1, HC2) and healthy controls with one blood draw only (n=10) were compared with CANDLE (n=8) and SAVI (n=4) patients' samples. Red arrowheads indicate increase in MFI ratio in SAVI patients on treatment which is up compared to pre-baricitnib (left panels). *patients were on optimal tolerated doses of baricitinib at the time of blood draw.
As different methods for the STAT phosphorylation were used, no formal statistical comparisons between pre-/baricitinib of per-efficacious doses of baricitinib and optimal doses of baricitinib were performed. TOP: The heatmap depicts serum cytokine levels at baseline (BSL) and at the last visit (end of study, EOS). Samples were collected up to December 2015. The cytokines are sorted in the heatmap according to the level of statistical significance of the BSL vs. EOS comparisons and cytokines marked with an asterisk were significantly higher at baseline. The reduction of serum IP-10 levels, a downstream marker of IFN signaling was most pronounced. Other serum cytokine levels that significantly dropped on treatment included the myeloid growth and differentiation factor, granulocyte-macrophage colonystimulating factor (GM-CSF), and the GM-CSF induced chemokine MCP-1 (CCL2) that is dependent on JAK2-STAT5 signaling through the GM-CSF receptor. 31 Both modulate monocyte and macrophage differentiation 30 and lead to a proinflammatory environment. The treatment effect on the reduction of serum levels of the proinflammatory cytokine, IL-15, that is secreted by mononuclear phagocytes (and some other cells), and the eosinophil and B cell growth and differentiation factor, IL-5, were more pronounced than the effect on IL-6, which was not statistically significant. IL-15 and IL-5 signal through their respective receptors through recruiting JAK2-STATs, and amplify their production through an autocrine loop, IFNa and IFNb could not be reliably measured.
BOTTOM: The table depicts the comparisons of mean and standard deviations of cytokine serum levels between baseline and end of study (EOS) visit for each group of diseases (CANDLE, SAVI and undifferentiated interferonopathies, UNDIFF), separately.