Pharmacokinetics, effectiveness, tolerability and effect on quality of life of open-label tofacitinib for the treatment of moderately active mucocutaneous manifestations of SLE: results of a 76-week phase II study

Abstract

Objective

To investigate the pharmacokinetics, effectiveness, safety and tolerability of tofacitinib in young adults with active mucocutaneous (MC) SLE manifestations.

Methods

Patients with SLE and Cutaneous Lupus Erythematosus Disease Area and Severity Index (CLASI) Activity (CLASI-A) Scores >8 newly received open-label tofacitinib from baseline to week (week) 72. Intensive pharmacokinetics sampling was done at the end of week 1, and safety was assessed throughout the study. The primary effectiveness endpoint was partial response of MC activity defined as <20% improvement of the CLASI-A Score from baseline (CLASI-PR) at week 8 in intention-to-treat analysis; secondary and other endpoints included MC complete response (CLASI-CR; CLASI-A=0) and quality of life (QoL) as measured by Skindex-29.

Results

Subjects were 11 patients (female patients: 10, mean±SD age: 23.0±5.87 years) with moderate to severe MC manifestations (CLASI-A=16.6±8.03). The pharmacokinetics of tofacitinib was comparable to that seen in juvenile and adult arthritis. During 629 patient-weeks of follow-up, 73 adverse events (AEs) were reported, none of which were considered severe or led to study discontinuation. There were three serious AEs (pelvic inflammatory disease, appendicitis and migraine with aura) but no major cardiovascular events, malignancies or death. In intention-to-treat analysis, at week 8 CLASI-PR was achieved in 73% (8/11) of subjects, in 82% at week 24 and 100% at week 72, respectively. Although CLASI-A Scores significantly decreased from baseline (baseline/week 8/week 24/week 72=16.55±8.03/8.64±7.8/8.82+7 .9/7+9.7; p<0.001), CLASI-CR was achieved by only <10% (1/11), starting week 8. Skindex-29 Scores improved significantly as early as week 4 (baseline/week 4/week 24/week 72= 37.6±24.16/26.1±23.58/24.3±28.67/25.5±30.71; p=0.009).

Conclusion

Tofacitinib in patients with SLE with active MC manifestations showed good effectiveness by week 4 as well as tolerability at exposure comparable to those with other rheumatic diseases. Tofacitinib was associated with significant improvement of QoL due to rapid improvement of MC inflammation. Observed safety and pharmacokinetics were comparable to observations in juvenile and adult arthritis.

WHAT IS ALREADY KNOWN ON THIS TOPIC.

• Janus kinase (JAK) inhibitors block signalling through signal transducer and activator of transcription 4, resulting in the suppressed production of type 1 interferons, the key cytokines in SLE.

• Tofacitinib is a JAK inhibitor and indicated for the treatment of various paediatric and adult diseases, including moderate to severe rheumatoid arthritis, active psoriatic arthritis, active ankylosing spondylitis, moderate to severe ulcerative colitis, and polyarticular course of juvenile idiopathic arthritis. Some JAK inhibitors have wider indications, including cutaneous inflammatory diseases such as alopecia areata. Tofacitinib is not approved for the treatment of SLE.

• Mucocutaneous (MC) manifestations in SLE are common, including alopecia, scarring and non-scarring rashes, as well as mucosal ulcerations.

WHAT THIS STUDY ADDS

• The pharmacokinetics of tofacitinib in patients with SLE demonstrates that drug exposure levels are within the range observed in previously approved indications, including juvenile idiopathic arthritis.

• Tofacitinib was well tolerated, and safety observed in SLE was consistent with the known safety profile.

• Tofacitinib was associated with at least partial remission of SLE skin manifestations, including alopecia, as early as week 4 of use when used at dosages approved for rheumatoid arthritis.

• Tofacitinib led to significant improvement in skin-related quality of life (QoL) measured by the Skindex-29, especially emotional distress related to SLE-associated MC manifestations.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• At the dosage approved for rheumatoid arthritis, open-label tofacitinib markedly and quickly improved MC manifestations of SLE and was associated with significantly improved QoL. However, complete remission of skin manifestations was not commonly achieved at the dosage of 5mg two times per day.

• Any off-label usage of tofacitinib needs to carefully considered the known propensity of tofacitinib for increasing the risk of major adverse cardiovascular events.

Introduction

SLE is a multisystemic autoimmune disease characterised by the presence of autoantibodies towards nuclear antigens, immune complex deposition and chronic inflammation in classic target organs such as skin, joints and kidneys. Mucocutaneous (MC) manifestations of SLE are very common and include cutaneous rashes, such as the classic malar erythema and discoid lupus lesions, mucosal ulcerations, and alopecia.¹ SLE-associated MC manifestations can result in MC damage such as scarring, hyperpigmentation and/or hypopigmentation. The degree of MC manifestations in terms of active inflammatory lesions and chronic MC damage is commonly measured using the Cutaneous Lupus Erythematosus Disease Area and Severity Index (CLASI).^{2 3} The histology of SLE-associated MC lesions includes lymphocytic infiltrates, hyperkeratosis and vacuolar interface dermatitis. There is an increased presence of plasmacytoid dendritic cells with the downstream production of type I interferons.⁴ Further, gene expression analyses support the importance of type 1 interferons in the pathophysiology of MC manifestations.^5–8

Janus kinases (JAKs) are intracellular tyrosine kinases and a part of the signal transducer and activator of transcription signalling cascade that mediates the downstream immunological effects of type I interferons. Indeed, JAKs play a key role in the regulation of proinflammatory genes activity.^8–10 Tofacitinib is a first-generation JAK inhibitor (JAKi) that has been shown to be safe and efficacious in controlling inflammation in various rheumatic conditions, including rheumatoid arthritis and juvenile idiopathic arthritis (JIA).^{9 11} Notably, tofacitinib pharmacokinetics are influenced by cytochrome P450 enzymes. This is the first time that the pharmacokinetics of tofacitinib is evaluated in patients with SLE. Clearance mechanisms for tofacitinib are approximately 70% via hepatic metabolism and 30% by renal excretion of the parent drug. The metabolism of tofacitinib is primarily mediated by cytochrome (CYP)3A4, with a minor contribution from CYP2C19. Inflammation can significantly impact the activity of these enzymes, leading to changes in drug exposure. Proinflammatory cytokines, like interleukin 6, can downregulate CYP enzyme expression.^{12 13} Therefore, assessment of tofacitinib exposure is warranted in this SLE disease population.

The aim of our research was to study the pharmacokinetics, safety and effectiveness of tofacitinib for the treatment of young adults with SLE with active MC involvement, including its impact on health-related quality of life (QoL).

Patients and methods

Study design and subjects

This is an open-label 76-week, three-part trial of tofacitinib for the treatment of SLE with moderately to severely active MC involvement (NCT 03288324). Study subjects were recruited and studied at the Cincinnati Children’s Hospital Medical Center (CCHMC), Cincinnati, USA, and the MetroHealth System, Cleveland, USA. All study subjects received a stable dose of tofacitinib 5 mg orally twice daily starting at baseline (week 1) until the end of week 72, followed by a 4-week post-treatment observation period. Safety reporting was completed in accordance with the guidelines of CCHMC institutional review board (IRB) and US Food and Drug Administration (FDA) requirements (21 Code of Federal Regulation (CFR) Part 312 compliance) for studies performed under Investigational New Drug (IND) 132786. The study protocol, amendments, and study subject informed consent were all reviewed by the CCHMC Institutional Review Board Federalwide Assurance (FWA) #00002988 and the MetroHealth Institutional Review Board FWA #00003983.

There was no Data Safety Board, given the small sample size and the implementation of procedures to minimise the risk to study subjects, as established under both the Department of Health and Human Services and the Office for Human Research Protections. There were no patients or other public involvement in the design, conduct and/or reporting of this study. The ethical principles of the Declaration of Helsinki were followed, as were Good Clinical Practices and applicable regulatory requirements. Written informed consent was provided by the study subjects before the commencement of study-related procedures.

Eligible patients had at least moderately active MC involvement with SLE despite standard care therapy at baseline. During Part A (baseline to week 8), background therapy was kept stable, and intensive sampling for tofacitinib pharmacokinetics was done on day 5 of the study. During Part B (week 8 to week 24), tapering of corticosteroids was allowed, provided CLASI Activity (CLASI-A) Scores had improved by ≥50% from baseline. Study subjects were allowed to enter Part B of the study as early as week 4. During Part C (week 24 to week 72), tofacitinib was continued, but all SLE background medications could be discontinued, provided MC manifestations of SLE were inactive (CLASI-A Score=0). A protocol synopsis is provided as online supplemental document 1.

Study procedures

Eligible patients were 18–45 years of age, fulfilled the 1997 Classification Criteria for SLE¹⁴ and had at least moderately active MC manifestations as measured by a CLASI-A Score of ≥8 at baseline. Allowed stable background treatments included hydroxychloroquine (HCQ), conventional disease-modifying antirheumatic drugs and oral prednisone of up to 20 mg/day. Key exclusion criteria were active SLE likely requiring escalation of therapy within 2 weeks of baseline, treatment with other investigational agents within the last 6 months, or five half‐lives, whichever was longer, estimated glomerular filtration rate <60 mL/min/1.73 m², use of rituximab within 1 year and/or cyclophosphamide, cyclosporine or tacrolimus within 12 weeks before baseline and presence or history of antiphospholipid syndrome. The full list of eligibility criteria is shown in the online supplemental document 1 (Study synopsis). Study visits were scheduled at baseline (week 1), then weekly until week 8, followed by visits at weeks 10, 12, 16, 24, 32, 40, 48, 56, 64, 72, and a safety follow-up at week 76. An intensive pharmacokinetics sampling was done on day 5 of the study. Safety and tolerability were assessed throughout the study for all subjects who received at least one dose of tofacitinib. At each study visit, effectiveness was assessed with a focus on MC manifestations of SLE and related patient-reported outcomes.

Study endpoints

The primary endpoint was pharmacokinetics characterisation at the end of week 1 when steady-state exposure of tofacitinib was achieved. Comparison was made to the plasma exposures with JIA.^{15 16} Safety endpoints were the frequency of adverse events (AEs), and serious adverse events (SAEs). Effectiveness endpoints included the proportion of subjects who achieved at least a CLASI partial response (CLASI-PR), that is, a >20% improvement of CLASI-A Score at week 8 compared with baseline, and those who achieved CLASI complete response (CLASI-CR), that is, a CLASI-A Score of 0 at week 8. Measuring the primary endpoint at week 8 was done to allow escape of non-responders who might need other therapies. Non-response was defined as a change of CLASI-A Score from baseline of <20% of improvement at week 8. Other effectiveness outcomes measured throughout the study were the scores of the CLASI-A, CLASI damage (CLASI-D),^{2 3} the MC domain of the British Isles Lupus Assessment Group (BILAG) Index,¹⁷ SLE Disease Activity Index (SLEDAI) and its MC domain. Patient-reported outcomes included the Skindex-29 domains and summary scores,^18–20 patient well-being and change of background medications in Part B and Part C of the study.

Details on assessments

Intensive pharmacokinetics sampling was done at the end of week 1, with plasma samples collected at predose, and 0.5, 1, 2, 4 and 8 hours after the morning dose of tofacitinib. Pharmacokinetics samples were analysed at PPD Laboratory Services (Richmond, Virginia, USA) using a fully validated high-performance liquid chromatography (HPLC)/mass spectrometry (MS)-MS method following ICH M10 guidelines. The pharmacokinetics parameters of tofacitinib were derived from the concentration-time profiles using non-compartmental analysis. The observed predose concentration served as the 12-hour concentration in the analysis, assuming a steady state. Actual pharmacokinetics sampling times were used in the derivation of pharmacokinetics parameters from non-compartmental modelling. The t_½ calculation was based on the terminal phase pharmacokinetics included data points from the time required for tofacitinib to reach the maximum concentration (T_max). The percentage of area under the curve (AUC) extrapolation for all subjects was ≤30%. Plasma concentrations were summarised descriptively by pharmacokinetics sampling time. Individual subject concentration‐time data were plotted using actual pharmacokinetics sampling times. Summary profiles (mean and median plots) of the concentration‐time data were plotted as well. The pharmacokinetics parameters, such as geometric mean apparent clearance (CL/F) and volume to availability (Vz/F), were formally compared with historical data in children with JIA,^{15 16} using a mixed-effects modelling approach.

Safety assessments included the entire safety population, that is, all subjects who received at least one dose of tofacitinib. AEs were recorded and coded as per the Medical Dictionary for Regulatory Activities,²¹ with severity (mild, moderate, severe) assessed according to Common Terminology Criteria for Adverse Events.²² SAEs were also recorded, that is, AEs that can lead to life-threatening conditions, hospitalisation, disability or permanent damage, congenital anomalies, or birth defects, and are critical enough to require immediate medical intervention or even death. Events of special interest, such as cerebrovascular accident, myocardial infarction/acute coronary disease, pulmonary embolism, malignancy and lymphoproliferative disease, were also assessed.^{23 24}

For assessment of effectiveness, trained rheumatologists and dermatologists completed the CLASI, the MC domain of the BILAG Index,¹⁷ the SLEDAI,²⁵ and the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SDI).²⁶ Whenever possible, the same investigator assessed the same subject throughout the study.

The CLASI is a validated tool for the evaluation of MC activity and damage with SLE. A CLASI-A Score (range: 0–70) of 0–9 indicates mild MC activity, 10–20 moderate MC activity and 21–70 severe MC activity in SLE, respectively.² The CLASI is responsive to changes in MC activity with SLE.^{2 3 27} CLASI-D Scores (range: 0–56) of 0–5 indicate mild MC damage, 6–16 moderate MC damage and 17–56 severe MC damage, respectively.²⁸ The SLEDAI (V.2 k) was completed, and its MC domain (range: 0–6) and extra-MC (range: 0–99) Scores were calculated. SLEDAI summary scores can be interpreted as follows: 1–5 mild disease activity, 6–10 moderate disease activity, and 11 or more severe disease activity (range: 0–105),²⁹ with higher scores indicating higher activity.²⁵ Likewise, the MC domain of the BILAG Index was completed (A: very active MC disease, score=12; B: moderately active MC disease, score=8; C: mild MC activity, score=1; and D/E: no current/previous MC activity, score=0).¹⁷ The SDI (range: 0–47; 0=no damage)²⁶ was completed annually, starting at baseline and week 76.

Patient-reported outcomes included patient well-being (0=very well; 10=very poor) and pain (0=no pain; 10=very severe pain) were both measured on a 10 cm Visual Analogue Scale. The Skindex-29 is a skin-specific QoL measure that has been validated for use in SLE.^{18 19} Skindex-29 summary scores are the unweighted averages of the three domain scores and can be interpreted as no/mild/moderate/severe/very severe impairment of QoL for values of 0–10/11–25/26–50/51–70/71–100, respectively.¹⁹ For interpretation of Skindex-29 domain scores, the following threshold-off scores were used to define severe impairment: ≥52 points for the symptoms domain, ≥37 for the functional domain and ≥39 for the emotional distress domain.¹⁸

Statistical analysis

AEs, AEs by severity and SAEs were all summarised using frequencies and with consideration of the number of affected subjects. In the primary effectiveness analysis, we used the intention-to-treat approach with the last observation carried forward. In the secondary effectiveness analysis, we used an as-observed data approach instead. Descriptive analysis and contingency table analyses were done to determine the proportion of subjects with CLASI-PR or CLASI-CR, with linear trends during the study (baseline to week 72) evaluated using the Cochran-Armitage trend test. Linear mixed models were used to evaluate trends in other measures (CLASI-A, CLASI-D, SLEDAI-MC, BILAG-MC, Skindex-29) during the study (baseline to week 72). To assess the prednisone-sparing properties of tofacitinib, we determined the proportion of subjects who discontinued prednisone at the end of Parts B and C, respectively. We also compared the change from baseline of prednisone daily dosage using a paired t-test. Values of p<0.05 in two-sided testing were considered statistically significant. Data were captured in RAVE (Medidata Solutions, New York, New York, USA) and analysed using SAS V.9.4 (SAS Institute, Cary, North Carolina, USA).

Role of the study sponsor

Pfizer provided study medication and funding for the study. The trial was designed by HB in collaboration with Pfizer. The sponsor was not involved in the overall management of the trial, data collection or data analysis, except for the performance of pharmacokinetics assays and analysis. The first manuscript draft was written by the authors (IR, HB) with subsequent drafts revised by all authors.

The authors have no competing interests related to this study.

Results

Subject disposition, demographics and baseline characteristics

Between August 2017 and June 2022, 11 patients were enrolled in the study and 4 subjects discontinued from the study before week 72, none for reasons of safety or tolerability (figure 1). As summarised in table 1 most subjects were female with a mean±SD age of 23.0±5.87 years and disease duration of 5.73±5.23 years at baseline. All subjects continued their baseline treatment with almost all (n=10, 90.0%) taking HCQ and six (54.5%) subjects taking daily prednisone. At baseline, as expected, all subjects had moderately active MC disease (mean CLASI-A=16.55) and none had mild or severe MC disease based on their CLASI-A Scores. There was, on average, moderate MC damage (CLASI-D=4.91), moderate overall disease activity, mostly from extra-MC activity, and low global damage, with five (45.5%) subjects having no damage (SDI=0) (table 1).

Figure 1.

Patient disposition. wk, week.

Table 1.

Demographic and clinical characteristics of the study population

	Mean±SD	N (% of 11)
Number of subjects; N	11	11
Age (years)	23.0±5.87
Female; n (% of N)		10 (91%)
Race (white); n (% of N)		7 (63.6%)
Ethnicity (Hispanic)		0%
Disease duration (years)	5.7±5.23
Weight (kg)	70.0±25.0
Mucocutaneous Disease Scores
CLASI Activity	16.55±8.03
BILAG MC domain	8.27±5.16
SLEDAI MC domain	2.18±0.60
CLASI Damage	4.91±1.87
SLE activity	5.82±6.98
SLEDAI total	7.1±1.87
SLEDAI extra-MC domain	4.91±1.87
Global damage (SDI)	0.64±0.67
Medications:	8.13±2.59
Prednisone (mg/day)
Prednisone; n (% of N)		6 (54.6%)
Mycophenolate mofetil		2 (18.2%)
Hydroxychloroquine		10 (90.9%)

BILAG, British Isles Lupus Assessment Group; CLASI, Cutaneous Lupus Erythematosus Disease Area and Severity Index; MC, mucocutaneous; SDI, Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SLEDAI, SLE Disease Activity Index.

Pharmacokinetics

Tofacitinib pharmacokinetics analysis was evaluated in seven subjects and showed expected profiles. The geometric mean area under the plasma concentration-time curve during the dosing interval of 12 hours (AUC_tau) was 176 ng*hour/mL (percentage coefficient of variation (CV) =23.5%). For two subjects, AUC_tau was computed based on k_el extrapolation, as the 12-hour concentration values were higher than the preceding point, precluding a k_el estimate. The geometric mean maximum plasma concentration (C_max) was 51.7 ng/mL (%CV=21.6%) with a median T_max (range=0.5–2) of 1 hour. The mean t_½, apparent clearance (CL/F) and Vz/F were 2.1 hours (SD =±0.321), 28.4 L/hour (%CV=23.5%) and 85.2 L (%CV=25.6%), respectively. The central tendencies of plasma tofacitinib concentrations after tofacitinib intake are similar in SLE compared with JIA (figure 2). Plotting pharmacokinetics results showed overlaying data from prior JIA studies. Additionally, background treatment did not influence tofacitinib distribution.

Figure 2.

Comparison of tofacitinib exposure in young adult patients with SLE-CL (red) versus in patients with JIA (blue and green). JIA, juvenile idiopathic arthritis.

Safety and tolerability

During the total observation period of 629-patient weeks of tofacitinib exposure, 10 (91%) subjects developed at least 1 AE, with a total of 73 AEs occurred (table 2; see online supplemental table 1 for additional detail), including 3 SAEs in 2 subjects, but no severe AEs. All reported SAEs that required hospitalisation (appendicitis, migraine with aura, pelvic inflammatory disease) responded promptly to usual treatment and resolved without complications.

Table 2.

Safety of tofacitinib in the study population

Adverse events (AEs)	Number of events	Per cent of total events, %
Total number of AEs	73	100
Mild	55	75
Moderate	18	25
Severe	0	0
Total number of serious AEs	3	4.1
Subjects with events	10	91
The overall duration of safety follow-up, weeks	629
Serious AEs (preferred term)
Appendicitis	1	1.37
Migraine with aura	1	1.37
Pelvic inflammatory disease	1	1.37
AEs (system organ class)
Infections and infestations	33	45.2
Gastrointestinal disorders	15	20.5
Musculoskeletal and connective tissue disorders	5	6.8
Nervous system disorders	5	6.8
Skin and subcutaneous tissue disorders	3	4.1
Injury, poisoning and procedural complications	2	2.7
Neoplasms benign, malignant and unspecified (including cysts and polyps)	2	2.7
Psychiatric disorders	2	2.7
Blood and lymphatic system disorders	1	1.4
Ear and labyrinth disorders	1	1.4
Eye disorders	1	1.4
Metabolism and nutrition disorders	1	1.4
Renal and urinary disorders	1	1.4
Reproductive system and breast disorders	1	1.4

Among non-serious AEs, 55 AEs were mild and 18 were moderate severities, mostly from infections and infestations (33/73=45.2%), gastrointestinal disorders (13/73=20.5%), nervous system disorders or skin and subcutaneous tissue disorders each (5/73=6.8%). All infections had uncomplicated courses and responded well to standard therapy. There were nine episodes of vomiting (12.3%), of which eight occurred in one subject. There were no events of special interest, such as cerebrovascular accident, myocardial infarction/acute coronary disease, pulmonary embolism, malignancy, lymphoproliferative disease, reactivation of chronic infection, serious organ dysfunction, serious hypersensitivity reaction, as well as episodes of diverticulitis, interstitial perforation, thromboembolic events or death. There were no withdrawals from the study or early termination due to AEs.

Effectiveness of tofacitinib for MC manifestations

As shown in figure 3, Panel A, the study met its effectiveness endpoint in that 73% of subjects achieved CLASI-PR by week 8, with the response maintained until the end of the study at week 72. We observed improvement in MC and extra-MC manifestation of SLE activity throughout the study. As summarised in table 3, CLASI-A Scores significantly improved as early as week 4 from baseline (p=0.003) and then further improved until the end of the study at week 72. Likewise, the SLEDAI-MC and BILAG-MC Scores significantly improved by week 8 compared with baseline, with improvement maintained throughout the study (table 3). Alopecia with active SLE improved throughout the study (mean±SD of CLASI-A alopecia item scores: baseline/week 4/week 8/week 24/week 72: 2.09±1.14/1.45±1.21/0.91±1.14/1.18±1.08/1.09±1.14) with improvement reaching statistical significance as early as week 8 of the study (p<0.05). Notably, CLASI-D Scores stabilised without significant changes through the study (p=0.94).

Figure 3.

(A) Partial and complete response as measured by the CLASI (ITT analysis). (B) Quality of Life as measured by the Skindex-29 during the study. CLASI, Cutaneous Lupus Erythematosus Disease Area and Severity Index; ITT, intention-to-treat.

Table 3.

Effectiveness of tofacitinib during the study (ITT analysis)*

	BL (1)	Week 4 (2)	Week 8 (3)	Week 24 (4)	Week 72 (5)	P value†
						(1) versus (2)	(1) versus (3)	(1) versus (4)	(1) versus (5)	(1) to (5)
MC disease activity and damage
CLASI-A	16.6±8.03	10.9±8.40	8.6±7.80	8.8±7.90	7.0±9.70	0.003	0.003	<0.001	0.02	<0.001
CLASI-D	5.82±6.98	5.55±6.96	6.27±7.17	5.73±7.11	4.6±7.06	0.25	0.08	0.88	0.73	0.94
BILAC-MC	11.8±6.97	5.7±4.22	2.5±3.24	4.3±5.59	5.6±5.82	0.02	<0.001	<0.001	0.003	<0.001
SLEDAI-MC	2.18±0.60	1.91±0.54	1.27±0.79	1.36±0.5	1.4±0.90	0.19	0.01	0.005	0.10	<0.001
Quality of life
Skindex-29 summary	37.6±24.16	26.1±23.58	24.1±22.63	24.3±28.67	25.5±30.71	0.02	0.01	0.03	0.14	0.009
Symptoms	42.6±19.29	31.0±20.52	25.8±19.21	26.8±23.46	30.7±28.64	0.11	0.04	0.06	0.14	0.02
Function limitation	24.5±20.88	18.5±19.44	20.4±21.03	20.8±29.34	21.7±29.19	0.15	0.28	0.19	0.74	0.36
Emotional distress	49.6±36.43	31.7±33.66	27.5±29.24	26.6±32.79	26.5±35.2	0.01	0.004	0.009	0.04	<0.001
Patient well-being	2.7±1.47	1.6±0.77	1.1±0.7	1.3±1.2	1.5±1.54	0.01	0.008	0.03	0.12	0.01
Pain	3.1±2.64	4.2±2.98	3.1±3.18	2.1±2.5	1.3±2.11	0.22	0.98	0.22	0.01	0.06

p<0.05 highlighted in bold indicates statistical significance.

*Mean±SD of outcome measure scores, unless stated otherwise from intention to treat analysis with the last observation carried forward.

†Paired t-test is used for the comparison of baseline (BL,1) versus week 4 (week 4;2), BL versus week 8 (3), BL versus week 24 (4) and BL versus week 72 (5); or a linear mixed-effects model is used to assess trends across the study denoted as (1) to (5).

BILAG, British Isles Lupus Assessment Group; CLASI, Cutaneous Lupus Erythematosus Disease Area and Severity Index; CLASI-A, CLASI-Activity; ITT, intention-to-treat; MC, mucocutaneous; SLEDAI, SLE Disease Activity Index.

Effectiveness of tofacitinib for extra-MC manifestations and tapering of background medications

SDI Scores remained stable during the study, while extra-MC manifestations trended towards improvement (mean±SD of SLEDAI extra-MC scores: baseline/week 4/week 8/week 24/week 72: 4.9±1.87/3.9±1.64/3.1±2.17/3.2±1.99/3.0±2.24). For musculoskeletal signs of diseases specifically, there were 2 (18%)/1 (9%) subjects at baseline with SLEDAI item scores for arthritis/myositis, which decreased over time (week 4: 2/1; week 8: 1/0; week 24: 0/0; week 72: 1/0).

Despite tofacitinib therapy, none of the six subjects on prednisone at baseline discontinued or even markedly decreased the dose of prednisone.

Patient reported outcomes

As depicted in figure 3, Panel B, patient-reported outcomes significantly improved during the study. On average, moderate impairment of QoL as measured by the Skindex-29 summary score improved to mild impairment as early as week 4 (mean±SD of change; baseline to week 4: −11.5+23.87, p=0.02). The most pronounced improvement occurred in the Skindex-29—emotional distress domain, followed by the Skindex-29—symptoms domain, and Skindex-29—function limitation, which improvement was clinical (table 3). Likewise, the subject-reported overall well-being significantly improved during the study, as did pain (table 3). The as-observed effectiveness of tofacitinib showed similar results (online supplemental table 2).

Discussion

We studied the pharmacokinetics, safety, tolerability and effectiveness of open-label tofacitinib for up to 72 weeks when used in young adults with SLE who experienced at least moderate MC involvement despite standard treatment, consisting of HCQ and prednisone in most subjects. The pharmacokinetics of tofacitinib in SLE was comparable to that when used in paediatric patients with JIA. Subjects tolerated tofacitinib well, and CLASI-PR was generally achieved. As demonstrated in the results, there was a statistically significant improvement of MC manifestations as early as week 4, with improvements maintained until the end of the study, suggesting that JAK inhibition interrupted key cytokines that initiate and/or perpetuate cutaneous SLE. Extra-MC manifestations also improved, as did subject well-being. Tofacitinib treatment was found to be associated with improvement of QoL related to MC manifestations as measured by the Skindex-29.

Both tofacitinib and HCQ are metabolised by cytochrome P450 enzymes, specifically CYP2C19 and/or CYP3A4.³⁰ Further, high levels of inflammation or vasculopathy are known to influence tofacitinib exposure.^{31 32} Nonetheless, tofacitinib exposure in the study population, which generally used HCQ background therapy, was comparable to JIA, which was used as a comparison cohort.^{16 33} Indeed, the apparent tofacitinib clearance (CL/F in L/hour) in the study cohort with SLE was 28.4 L/hour and has face validity as it is well within the range of that observed in healthy participants (34.9 L/hour),³⁴ and in patients with rheumatoid arthritis (18.4 L/hour),³⁴ psoriasis (26.7 L/hour),³⁴ JIA (26.1 L/hour),^{16 33} ulcerative colitis (26.3 L/hour),³⁵ psoriatic arthritis (20.4 L/h)³⁴ and ankylosing spondylitis (27.1 L/hour; data on file from Pfizer).

The benefits of JAKi use for the treatment of cutaneous manifestations with SLE have been studied in the past.^{8 36–38} Hasni et al ⁸ reported on 30 mostly Hispanic subjects with SLE treated with tofacitinib for 3 months. Like in our cohort, tofacitinib was associated with an acceptable safety profile, mild reduction in overall disease activity (SLEDAI, BILAG), but MC improvement was less robust, possibly due to the relatively short study period.

We observed robust improvement in MC manifestations, including improvement of alopecia, although rarely complete resolution of MC manifestations with SLE as has been reported for baricitinib in SLE.^{36 37} Importantly, a prior clinical trial of baricitinib in SLE showed that higher dosages of JAKi than those approved for other autoimmune diseases might lead to a more pronounced improvement of MC manifestations of SLE,^{36 37} suggesting that we might not induce maximal response since the approved daily dosage for moderate to severe rheumatoid arthritis was used in our study. Potential dose escalation of tofacitinib to optimise MC improvement must be carefully weighed against the increased risk of cardiovascular events with JAKi use, especially at high doses,^{39 40} possibly amplifying an already higher baseline risk of patients with SLE compared with rheumatoid arthritis and the general healthy population.³⁹

Whether other injectable medications that have been shown beneficial for the treatment of MC manifestations in the context of active SLE, such as anifrolumab, rituximab or belimumab, are more beneficial than JAKi will need a comparative, larger study.

Further, the effect of tofacitinib may be incomplete in cases of higher MC activity, and we would not expect pre-existing MC damage, especially scarring and long-standing dyspigmentation, to improve despite effective therapy. Recent dyspigmentation might be amenable to improvement, but we were unable to perform subanalysis to determine whether different types of SLE MC lesions differentially respond to tofacitinib.

The safety and tolerability of tofacitinib in this study were comparable to the reported safety profile of the medication. As expected, infections, mainly upper respiratory infections, were the most common AEs. The frequency and type of infections were not different from those reported in the literature. There were no cases of Herpes zoster infections or cardiovascular events, likely because of the limited size of the study population and the younger median age of the subjects in the study.

Although there were no reports of thrombotic events in this study, recent research suggests that there is a higher risk of thrombosis with SLE, antiphospholipid syndrome, and in the setting of antiphospholipid antibody positivity.⁴¹ This should be considered before JAKi use in patients with SLE, including that antiphospholipid antibody status can change with disease activity.

The effect of JAKi use on patient QoL and well-being as it relates to MC manifestations of SLE has not been well described previously. To the best of our knowledge, we are the first to show that tofacitinib resulted in early, rapid improvement of Skindex-29 Scores and well-being. Like others,⁴² we found that MC manifestations result in high levels of emotional distress, and to a lesser extent, a reduction in QoL from limitation of pain and functional impairment (Skindex-29). Arguably, considering the high frequency of MC involvement with SLE and its profound effect on QoL, drug development and effective treatment of MC manifestations deserve more attention. Likely because MC manifestations are rarely life-threatening or organ-threatening, most current drug development programmes focus on kidney or other manifestations of SLE.

We confirmed adherence to tofacitinib, and background medications remained largely stable during the study, and effectiveness measures were assessed by highly experienced/trained investigators to complete study indices, suggesting that observed treatment effects are due to tofacitinib. Nonetheless, the open-label character of the study might have led to an overestimation of the effects of tofacitinib, specifically on patient-reported outcomes, since all subjects were aware that they were receiving tofacitinib. Further, the overall systemic disease activity with SLE (SLEDAI) was relatively low, and we did not measure commonly used clinical trial outcomes, such as the SLE Responder Index 4 (SRI-4), to assess the overall response of SLE disease activity during the study. This limited the ability to assess the treatment effect of tofacitinib on the overall course of SLE. However, the principal objectives of the study were pharmacokinetics and to delineate the effects of tofacitinib on MC manifestations of SLE.

As recommended,^{43 44} we performed two different statistical analyses, intention-to-treat and as-observed analyses, to assess the effectiveness of tofacitinib on MC outcomes. Subjects who discontinued the study did so mostly because of scheduling or time constraints, rather than failure to respond to tofacitinib. Importantly, both analytical approaches yielded comparable results with respect to improvement of MC manifestations during the study and the improvement of patient-reported outcomes. The as-observed analysis was performed for QoL assessment due to better applicability for analysing data.

In summary, tofacitinib improved MC manifestations of SLE and associated reduction in QoL. Further, tofacitinib at dosages approved for rheumatoid arthritis resulted in comparable tofacitinib exposure despite background therapies that could impact its pharmacokinetics.

Data availability statement

Data are available in a public, open access repository.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).

Ethics approval

This study involves human participants and was approved by Cincinnati Children’s Hospital Medical Center (CCHMC), Cincinnati, USA Institutional Review Board (FWA #00002988) and MetroHealth System, Cleveland, USA MetroHealth Institutional Review Board (FWA #00003983). Participants gave informed consent to participate in the study before taking part.