Biologic–Biologic and Biologic–JAK Inhibitor Combination Therapy in Refractory Systemic Autoinflammatory Diseases

Abstract

Objectives:

Systemic autoinflammatory diseases (SAIDs) arise from genetic defects in innate immunity, leading to dysregulated activation of inflammatory pathways, including interleukin (IL)-1, IL-6, TNF, and JAK/STAT. Clinical manifestations range from recurrent fever to severe complications such as encephalitis and AA amyloidosis. Management aims to control inflammation using immunosuppressive agents and targeted monotherapies (biologics or JAK inhibitors). Advanced combination therapy (ACT), defined as the use of biologics and/or JAK inhibitors in combination, has emerged as a strategy for refractory disease.

Methods:

In this observational retrospective longitudinal cohort study, patients with SAIDs treated with ACT were included. Demographic, clinical, treatment, and safety data were collected. Treatment response was assessed using a composite outcome incorporating corticosteroid dose, C-reactive protein (CRP), and clinical improvement and categorized as non-response, partial response, or complete response.

Results:

Thirty-eight patients (median age 30 years [range 4–76]) were included. The most common indications for ACT were pyogenic arthritis, pyoderma gangrenosum and acne (PAPA), mevalonate kinase deficiency (MKD), and undifferentiated SAIDs. Most patients had disease-related complications and were dependent on glucocorticoids and/or opioids to control inflammation and pain, respectively. Following multiple ACT trials, complete response was observed in 21 patients (55.3%), partial response in 12 (31.6%), and no response in 5 (13.1%). Overall, 65 ACT regimens were administered, most commonly combining IL-1 and TNF inhibitors. Thirty-nine regimens were discontinued because of lack of efficacy, secondary loss of response, or adverse events. At the final follow-up, 26 patients (68%) remained on ACT, with a median treatment duration of 60 months (range, 11–186).

Conclusions:

ACT offers significant clinical benefits for patients with difficult-to-treat SAIDs, though challenges such as secondary loss of efficacy and infection risks remain.

Graphical Abstract

Introduction

Systemic autoinflammatory diseases (SAIDs) are a group of conditions characterized by systemic inflammation resulting from genetic defects involved in innate immunity. The key mechanisms driving SAIDs include gain-of-function (GOF) mutations in positive regulators, loss-of-function (LOF) mutations in negative regulators of signaling pathways, specific enzymatic defects, and the accumulation of misfolded proteins. The main inflammatory mechanisms implicated in SAIDs include interleukin (IL)-1, IL-6, IL-12, IL-18, tumor necrosis factor-alpha (TNF), interferons (IFN), nuclear factor kappa B (NFKB), and JAK/STAT pathways [1]. Immune hyperactivation in SAIDs can be either continuous or intermittent, depending on the underlying mechanism and environmental triggers [2]. This hyperactivation may involve one or multiple inflammatory pathways simultaneously. In some cases, autoinflammation can overlap with other major immune diseases, such as autoimmunity, immune deficiency, atopy, or lymphoproliferation [3]. Moreover, while genetic alterations are typically confined to immune cells, in some SAIDs, they can also affect the biology of non-immune cells [1].

SAIDs present a wide range of symptoms, including but not limited to fever, eye and ear inflammation, serositis, arthritis, and inflammatory skin conditions. Moreover, life- and organ-threatening manifestations are not uncommon, such as retinitis, encephalitis, pyoderma gangrenosum (PG), enterocolitis, macrophage activation syndrome, and AA amyloidosis. The main goals of treatment are controlling symptoms to improve patients’ quality of life and productivity, while also minimizing inflammatory activity to prevent long-term damage [4]. Glucocorticoids, known for their broad-spectrum immunosuppressive effects, can rapidly suppress inflammation but contribute to accrual damage. Advanced or targeted therapies—specific cytokine inhibitors (biologics) and oral small-molecule inhibitors (i.e. Janus Kinase Inhibitors, JAKi)—have become the cornerstone of SAIDs management due to their efficacy and safety [4]. In SAIDs, despite the activation of multiple inflammatory pathways, a single advanced therapy is typically sufficient to control disease activity in most patients. However, certain conditions, like PG, may remain refractory, requiring multi-pathway targeting for optimal disease control.

Advanced combination therapy (ACT) involves the simultaneous use of two or more advanced therapies to achieve optimal disease control in patients with inflammatory diseases. Early studies in the 2000s on the combined use of biological agents in rheumatoid arthritis (RA) showed harm without additional benefit, leading to the box warning against the use of biological combinations for many years [5]. However, with the ample use of biologic agents and DMARD combinations in inflammatory diseases, a significant number of patients —including those with RA, spondyloarthritis (SpA), psoriatic arthritis (PsA), and inflammatory bowel disease (IBD)— have been identified as refractory or difficult to treat, showing nonresponse to multiple lines of advanced treatments. In this context, ACT has gained renewed attention as a viable approach [6]. JAK inhibitors inherently have multi-target activity, and bispecific targeted inhibitors are currently in the drug pipeline to better treat various immune-mediated inflammatory diseases [7]. Moreover, recent findings in IBD clinical trials suggest that induction treatment with ACT is more effective than either agent alone, further shifting the focus toward ACT [8]. Efforts for finding optimal ACT are underway in many inflammatory diseases including IBD, SpA, PsA, and lupus [9].

Given the substantial morbidity and potential mortality associated with SAIDs, effective glucocorticoid-sparing strategies remain a major unmet clinical need. Although ACT represents a promising therapeutic approach, evidence regarding its use in this setting is scarce [10, 11]. To address this gap, we present our experience with ACT in a large cohort of patients with SAIDs, focusing on indications for its use, efficacy in terms of glucocorticoid-sparing effects, opioid requirements, hospitalization frequency, and adverse events requiring careful monitoring.

Methods

Patients and data collection

This observational, longitudinal retrospective cohort study was conducted at the National Human Genome Research Institute (NHGRI) Autoinflammatory Disease Clinic. We enrolled patients with genetically confirmed SAIDs as well as undifferentiated systemic autoinflammatory disease (USAID) patients who had received two or more advanced treatments simultaneously between January 2002 and December 2024. To reduce selection bias, all consecutive patients receiving advanced combination therapy during the study period were included. Coexisting inflammatory diseases were diagnosed according to established classification or diagnostic criteria where available (SpA [12] and IBD [13]) or based on compatible clinical features supported by imaging (SpA, myocarditis, osteitis, retinitis, and central nervous system inflammation) and/or histopathology or surgical findings where appropriate (Hidradenitis suppurativa (HS), nodulocystic acne, PG, IBD, and AA amyloidosis). These conditions were treated in accordance with current standards of care where available (FMF, MKD, TNF receptor-associated periodic syndrome (TRAPS), SpA, IBD, and HS) or, in the absence of formal recommendations, according to the best available evidence and contemporary clinical practice at the time of treatment [14–16]. Socio-demographic, clinical, laboratory, treatment efficacy, and safety data as well as disease characteristics, prior treatments, medication start and end dates, treatment indications, and drug combination outcomes were extracted using standardized electronic medical records and predefined outcome measures. The incidence and severity of treatment-related and serious adverse events were also documented. A serious adverse event (SAE) was defined according to the National Institute on Aging (NIA) guidelines as any event at any dose that results in death, is life-threatening, requires inpatient hospitalization or prolongation of an existing hospitalization, results in persistent or significant disability/incapacity, or is a congenital anomaly/birth defect [17]. All patients provided written informed consent, and the study was approved by the Institutional Review Board of the National Institutes of Health (Protocol No. 94-HG-0105) and registered at ClinicalTrials.gov (NCT00001373). Patients were fully informed about the off-label use of ACT, and the potential risks and uncertainties were discussed as part of the shared decision-making process.

Assessment of treatment response

The response to each combination therapy was evaluated with a composite score based on three key parameters. The first was the steroid-sparing effect of ACT, defined as the ability to reduce the glucocorticoid dose to ≤7.5 mg/day of prednisone or its equivalent. The second parameter was the improvement in C-reactive protein (CRP) levels, specifically the suppression of CRP to below 10 mg/L and values were recorded at the time of glucocorticoids are at their lowest administered dose. The third parameter was the change in clinical symptoms and organ damage. Given the lack of disease-specific activity tools for each condition, we employed a generic clinical response measure: the Clinical Global Impression Improvement (CGI-I) scale [18]. For CGI-I scoring, during each follow-up visit after initiating therapy, the clinician compared the patient’s overall clinical status to the week prior to treatment initiation and rated the change on a seven-point scale: 1 = very much improved; 2 = much improved; 3 = minimally improved; 4 = no change; 5 = minimally worse; 6 = much worse; 7 = very much worse (Table 1).[18] The CGI-I is widely used in clinical trials as a simple, standardized tool to quantify treatment response [19]. The final treatment response was classified into one of three categories: non-response (NR), partial response (PR), or complete response (CR), according to the predefined criteria (Table 2).

Table 1.

Clinical Global Impression Improvement (CGI-I) scale

CGI-I score	Global change	Definition
1	Very much improved	nearly all better; good level of functioning; minimal symptoms; represents a very substantial change
2	Much improved	notably better with a significant reduction of symptoms and an increase in the level of functioning, but some symptoms remain
3	Minimally improved	slightly better with little or no clinically meaningful reduction of symptoms. Represents very little change in basic clinical status, level of care, or functional capacity
4	No change	symptoms remain essentially unchanged
5	Minimally worse	slightly worse but may not be clinically meaningful; may represent very little change in basic clinical status or functional capacity
6	Much worse	clinically significant increase in symptoms and diminished functioning
7	Very much worse	severe exacerbation of symptoms and loss of functioning

Adapted from Busner J & Targum SD, 2007. The clinical global impressions scale: applying a research tool in clinical practice. Psychiatry (Edgmont). 2007 Jul;4(7):28–37.

Table 2.

Composite response criteria for assessing response based on daily prednisone requirement at last assessment, C-reactive protein, and Global Clinical Impression Improvement (CGI-I).

Clinical Global Impression Improvement (GCI-I) scores: 1 = very much improved with the initiation of combination treatment; 2 = much improved; 3 = minimally improved; 4 = no change; 5 = minimally worse; 6 = much worse; 7 = very much worse since the initiation of treatment. Colors denote: □ non-response; □ partial response; and □ complete response.

Combination Strategy

All patients had exhausted available therapeutic options labelled for the treatment of their underlying conditions (e.g., methotrexate, cyclosporin A, mycophenolate, retinoic acid, IL-1 inhibitors, TNF inhibitors, etc.). They were refractory to multiple conventional drugs and sequentially used advanced drug monotherapies. Corticosteroids were frequently used to control inflammation caused by coexisting inflammatory diseases in patients with IL-1-mediated diseases, including FMF. NSAIDs were utilized for their anti-inflammatory and analgesic effects, while opioid analgesics were restricted to selected cases with severe, debilitating pain due to concerns regarding their prolonged use in inflammatory conditions. The selection of ACT combinations was guided by several key considerations, including disease phenotype, comorbidities, prior treatment failures, drug pharmacodynamics, and clinical experience gained from former phenotype-similar patients. All employed medications were approved for immun mediated inflammatory diseases, and the dosing adhered to FDA-approved guidelines. The ACT approach involved combining the two most effective and safest advanced treatment options. Where possible, short half-life drugs were chosen and then if the safety of this strategy was established and long-term treatment was required, medications with longer durations of action in the same class were substituted. In patients who had achieved good inflammation control (evidenced by normal acute phase reactants in between flares) but experienced occasional flares, the second agent was prescribed on a flare-directed (on demand, pro re nata, PRN) basis.

Statistical Analyses

To reduce selection bias, all consecutive patients receiving advanced combination therapy during the study period were included. Data were extracted using standardized electronic medical records and predefined outcome measures. Data were analyzed using SPSS version 22 (IBM, Armonk, NY, USA). Descriptive statistics were generated for the study cohort, including frequencies for categorical variables and means or medians for continuous variables, as appropriate based on data distribution. The significance of pre- and post-treatment nominal, dichotomous, and ordinal-level variables was tested using McNemar and Wilcoxon tests. The incidence rate of serious adverse events was calculated as the number of incident events divided by the total patient-years of follow-up and expressed per 100 patient-years. No imputation for missing data was performed. Statistical significance was set at p < 0.05 for all tests.

Results

A total of 38 patients who received ACT were identified. Among the cohort, 20 patients (52.6%) were female, with a median age at enrollment of 30 years (range: 4–76 years). The median age at symptom onset was 5 years (range: 0–58 years) with only five patients having adult-onset symptoms. The median disease duration was 19 years (range: 4–57 years). The median age at the initiation of the first ACT regimen was 22 years (range: 0–70 years), with 13 patients classified as being in the pediatric age group. The details of clinical characteristics, co-morbidities, complications and outcomes of patients are summarized in supplementary Table 1.

Pyogenic arthritis, pyoderma gangrenosum and acne (PAPA), mevalonate kinase deficiency (MKD), and USAIDs were the most prevalent SAIDs necessitating ACT in our cohort. The most affected organs/systems included the gastrointestinal tract, musculoskeletal system, and skin (see Supplementary Table 1). Concomitant immunodeficiency was evident in six patients. Regarding complications, all SAIDs patients, except for two with MKD, exhibited some form of permanent disease- or treatment-induced damage at the time of ACT initiation. In addition to glucocorticoid toxicity, the most frequently observed complications included joint restriction, skin disfigurement, anemia, and growth retardation. At the time of ACT initiation, 26 patients were dependent on glucocorticoids to control severe inflammation, 17 were dependent on opioid analgesics for severe pain control, and most required frequent hospitalizations because of disease flares or infections. Prior to ACT, all patients had received multiple lines of advanced treatment monotherapies, along with conventional DMARDs and immunomodulators. The median number of biologics or small molecules used before starting ACT regimens was 3 (range: 1–9), and the median number of conventional oral medications was 4 (range: 0–8).

Table 3 outlines the types and indications for various ACT regimens. The primary indications for initiating ACT included active systemic disease with frequent flares (n=14), PG (n=11) and/or nodulocystic acne (NCA, n=9), co-morbid inflammatory bowel disease (n=4), deforming arthritis, spondylitis and osteitis (n = 7), HS (n=3), central nervous system (CNS) inflammation (n=2), uveitis (n=1), myocarditis (n = 1), and AA amyloidosis with uncontrollable inflammation (n=1). In total, 65 ACT regimens were trialed, 57 of which involved the regular use of two or more advanced treatments whereas eight ACT regimens were administered as flare-directed, involving anakinra in combination with an anti-TNF agent during disease flares (MKD, PAPA, and USAIDs). Three ACT regimens consisted of triple-drug combinations (anti-IL-1 plus anti-TNF plus JAKi, all used in PAPA patients), while the remainder consisted of dual biologic agent combinations. The most common ACT regimens included a combination of an IL-1 inhibitor (anti-IL-1) and a TNF inhibitor (anti-TNF), or an anti-IL-1 combined with a JAKi.

Table 3.

Systemic autoinflammatory diseases, main indications for advanced combination treatments and the ACT responses

^*Total number of trialed combinations=65. Each number represents an ACT regimen. Color codes represent response to treatments; green: complete response, yellow: partial response, red: non-response. ABA: abatacept, ADA: adalimumab, ANA: anakinra, anti-IL-1: anti-interleukin-1 agent, anti-TNF: anti-tumor necrosis factor alpha agent, CAN: canakinumab, CNS: Central nervous system, DUP: dupilumab, FMF: familial Mediterranean fever, HA20: Haploinsufficiency A20, HS: hidradenitis suppurativa, IBD: inflammatory bowel disease, JAKi: Janus kinase inhibitor, MKD: mevalonate kinase deficiency, MWS: Muckle-Wells syndrome, NCA: nodulocystic acne, PG: pyoderma gangrenosum, PASH: pyoderma gangrenosum, acne, and hidradenitis suppurativa syndrome, ROSAH: retinal dystrophy, optic nerve edema, splenomegaly, anhidrosis and headache syndrome, SAPHO: Synovitis, Acne, Pustulosis, Hyperostosis, and Osteitis, SPA: spondyloarthritis, TCZ: tocilizumab, TRAPS: tumor necrosis factor receptor-associated periodic syndrome, USAID: undifferentiated systemic autoinflammatory disease, UST: ustekinumab, VEXAS: vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic syndrome.

After several ACT trials, complete response was achieved in 21 patients (55.3%), partial response in 12 (31.6%), and no response in 5 (13.1%). At the time of the last assessment, 23 patients were on a regular ACT regimen, three were on a flare-directed ACT regimen, 11 patients were receiving an advanced treatment monotherapy, and one patient (HA20) had undergone hematopoietic stem cell transplantation (HSCT). At the final assessment, ACT regimens were ongoing for a median duration of 60 months [range: 11–186 months]. ACT regimens led to significant improvements in clinical outcomes, including improvements in acute phase reactants, anemia, glucocorticoid and opioid requirements, and hospitalization needs in treatment responsive SAID patients (Table 4).

Table 4.

The outcome of patients who had sustained successful ACT combination (n=26)

	Before ACT	During ACT	p
ESR, mm/h	19 [7.5–40]	8 [4–16]	0.002
CRP, mg/L	9.9 [2.5–29.8]	1.5 [0.4–4.2]	<0.001
Hemoglobin, g/dL	12.8 [10.6–13.6]	13.5 [12.4–14.4]	0.009
Platelets, x103/mm3	243 [203–319]	289 [248–322]	0.29
WBC, x103/mm3	7.7 [5.8–11.2]	6.8 [5.0–8.6]	0.065
Albumin, mg/L	4.3 [3.6–4.6]	4.4 [4.1–4.6]	0.013
Glucocorticoid dose, mg/day	20 [0–30]	0 [0]	<0.001
Diabetes, n (%)	3 (11.5)	1 (3.8)	0.5
Hypertension, n (%)	9 (34.6)	10 (38.5)	1.0
Opioid use, n (%)	12 (46.2)	2 (7.7)	0.002
NSAID use, n (%)	19 (73.1)	12 (46.1)	0.13
Number of hospitalizations*	1 [0–3]	0 [0]	0.001

All values are presented as median [IQR 25–75] or frequency (%).

^*Sixteen patients required hospitalization prior to ACT, which decreased to three after ACT. The numbers in the table represent yearly occurrences, comparing the year before and after ACT (adjusted for treatment duration).

Regarding underlying conditions and ACT indications, the combination of anti-IL-1 and anti-TNF therapy appeared effective for co-morbid IBD, SpA, osteitis, peripheral arthritis, HS, MKD, and TRAPS flares. Multisystemic involvement in HA20 and VEXAS syndromes was refractory to the ACTs used. In three cases of PAPA syndrome where the combination of anti-IL-1 and anti-TNF therapies was effective but insufficient, the addition of a JAK inhibitor provided further improvement. Ustekinumab monotherapy was successful in a patient with severe ulcerative mucosal disease along with other systemic manifestations after a failed anti-IL-1 and anti-TNF combination. On the final assessment, 39 of trialed ACT regimens were discontinued. Reasons for discontinuation included primary non-response (n= 11), loss of efficacy over time (n= 13), pregnancy (n=1), adverse events (infections in six, exacerbation of underlying disease in three), and patients’ own will (n = 5).

Secondary loss of efficacy (secondary non-response) was observed in 6 patients with PAPA syndrome, 3 patients with USAIDs, and one patient each with ROSAH and FMF, after a median duration of 32 months [range: 4–84 months]. Of these secondary non-responses, 11 involved the anti-IL-1 plus anti-TNF combination, with 8 of these patients receiving concomitant methotrexate (10–15 mg/week) alongside the ACT regimen. However, switching to a drug within the same class (e.g., infliximab to golimumab or anakinra to canakinumab) appeared effective in these patients.

The only SAEs were the infections requiring hospitalization as per our NIA-SAE definition. A total of 37 infectious episodes were observed in 20 patients, 16 episodes of which required hospitalization. The majority of infections were similar to those encountered prior to initiating ACT, including sinusitis, recurrent urinary tract infections, and complicated PG, HS and/or NCA. ACT was administered to six patients with underlying immunodeficiency; two of them developed infections, including exacerbation of pre-existing condyloma acuminatum lesions, and staphylococcal skin abscess in an FMF patient with neutropenia, and IgA deficiency, which was managed with oral antibiotics. Sixteen SAEs (all infections) were observed over 151.9 patient-years of follow-up corresponding to an incidence rate of 10.5 event in 100-patient-years. However, the annualized rate of serious infections did not differ between the year preceding the first ACT regimen, during which patients were receiving advanced treatment monotherapy, and the period following ACT initiation, with median (range) values of 0 (0–5) and 0 (0–3), respectively. No patient deaths occurred during the follow-up.

Discussion

In this study, we report the use of 65 ACT regimens in 38 patients with SAIDs, of whom 26 (68%) maintained clinical benefit at the final assessment, including 23 receiving fixed-schedule continuous therapy and three using ACT on an on-demand basis. The ACT regimens resulted in significant clinical and laboratory improvements, as well as reductions in glucocorticoid and opioid requirements, outcomes that had not been achieved with prior advanced monotherapy in these multi-refractory patients with poor prognoses. Combination regimens appeared to be relatively safe, with infections being the leading cause of treatment discontinuation; however, there were no cases of unexpected or opportunistic infections. On the other hand, certain SAIDs, particularly those with severe involvement, such as HA20 and VEXAS syndromes, appeared refractory to ACT.

The exact pathogenesis of many SAIDs remains unclear, though GOF or LOF genetic alterations result in the simultaneous activation of multiple inflammatory pathways by affecting key pro-inflammatory or regulatory mechanisms. Despite significant advancements in monoclonal antibodies and small molecule targeted treatments of inflammatory diseases, few have been specifically developed for SAIDs due to their rarity, complexity, and heterogeneous presentations. As a result, advanced treatments designed for more common inflammatory diseases are often repurposed for SAIDs [4, 20]. Glucocorticoids are frequently used alongside these advanced therapies, despite the well-known complications associated with long-term use. HSCT offers an alternative for refractory patients; however, it comes with significant short- and long-term morbidity and mortality risks. Moreover, HSCT may not be suitable or effective for patients with USAIDs, particularly those with unknown disease mechanisms, genetic variants involving non-hematopoietic cell lines, multimorbidity, or poor performance status. Therefore, ACT may represent a viable treatment option for selected SAIDs patients who have a history of failure with multiple drugs or who exhibit refractory symptoms.

FMF, TRAPS, MKD and cryopyrinopathies are the most common genetically confirmed hereditary SAIDs in autoinflammatory clinics and IL-1 antagonists made a breakthrough in these patients with controlling debilitating inflammatory flares in most patients [21, 22]. However inflammatory co-morbid conditions that are refractory to IL-1 targeting such as IBD, SpA, PG, HS and vasculitides remain a treatment challenge in these patients. Anti-TNFs and JAKis are among the treatment options in IBD and SpA in routine clinical practice. Our findings suggest that combination therapy with anti-TNF and anti-IL-1 agents may be an effective treatment option for patients with these coexisting inflammatory conditions, with an acceptable safety profile in our cohort. In some FMF and MKD patients, anti-TNF monotherapy can suppress autoinflammatory disease flares when used to manage co-morbid inflammatory conditions [23, 24]. Thus, in these diseases, adding anti-IL-1 therapy to anti-TNF treatment may be reserved for patients who continue to experience flares despite anti-TNF therapy for their co-morbid conditions. Alternatively, anti-IL-1 agents could be used on a flare-directed basis during attacks if inflammatory activity is well controlled with anti-TNF agents alone [25–27]. Other potential approaches to warrant investigation include integrin receptor antagonists and IL-12/23 inhibitors for co-morbid IBD, as well as IL-17 inhibitors for associated SpA.

Inflammatory skin diseases were the most common organ-specific involvement requiring ACT in our cohort. PG, NCA and HS are notoriously difficult to treat, often requiring multiple lines of biologic and non-biologic immunosuppressants, and frequently necessitating major surgical interventions—grafting and amputation. Adalimumab, secukinumab and bimekizumab are approved treatments for HS [28], and isotretinoin and doxycycline for NCA [29]. However, no approved treatments exist for PG, and paradoxically, some commonly used biologics may even trigger PG [30]. Infections, including cellulitis, abscesses, fistula formation, bacteremia, and sepsis, further complicate the management of HS, NCA, and PG, adding significant challenges to treatment planning. In a recent study, we demonstrated that the IL-1, IL-18, IL-15, and interferon-γ axes create a self-amplifying inflammatory loop in PAPA syndrome, which can be suppressed with combination of advanced treatments [11]. In our study, anti-IL-1 and anti-TNF dual combination was effective while three patients still required triple combination with anti-IL-1, anti-TNF and JAKi to control disease flares. Although less explored in real-world settings, ACT regimens involving IL-17 and IL-12/23 inhibitors, along with small molecule targeted therapies might be effective as well [30]. Moreover, as biologics are generally less effective when used as second- or later-line therapies than as first-line treatment, as shown in RA and SpA, early ACT may be more effective in above mentioned notoriously monotherapy-refractory diseases before epigenetic and metabolic reprogramming occurs.

Data on the use of ACT has raised concerns about an increased risk of adverse or unknown effects [31]. However, our study found no new safety concerns associated with ACT regimens. Infections were the only SAEs observed, with an incidence rate of 10.5 events per 100 patient-years, notably higher than reported rates for biologic monotherapies in SAIDs [32]. On the other hand, this figure is significantly lower than the reported SAE incidence rate in etanercept-anakinra combination trial for RA [33]. We propose that severe inflammatory activity in SAIDs can be effectively managed with ACT, likely with less impairment of immune defense. Supporting this, a recent breakthrough deep immunophenotyping study suggested that FMF and IBD are characterized by high levels of inflammation, while RA, SpA and lupus are associated with lower levels of inflammation [34]. Consequently, the dual inhibition of potent cytokines in RA, a low-level inflammatory disease, may contribute to an increased risk of infections observed in combination [31]. Additionally, compared to RA patients, those with SAIDs are typically younger and have fewer comorbidities, making them less susceptible to serious infections.

Our study has several limitations. The retrospective and uncontrolled design, limited sample size, heterogeneity of SAID phenotypes, and varying follow-up periods preclude drawing robust conclusions. Patients receiving ACT experienced clinical and laboratory improvements. However, given the uncontrolled retrospective design of the study, these findings should be interpreted cautiously, as they may also reflect selection bias, regression to the mean, and other inherent limitations of observational studies. Our findings suggest that ACT may provide meaningful clinical benefit in carefully selected patients with refractory SAIDs; however, these observations require confirmation in prospective controlled studies. Additionally, we did not use quality of life (QoL) instruments, and we were unable to determine the effect of ACT on QoL. While our follow-up duration was relatively long, with a median of 60 months, this may still be inadequate for a lifelong disease. Notably, we observed a dramatic response to ustekinumab monotherapy in a patient with mucosal PG who had previously failed ACT regimens, suggesting that novel biologics with different modes of action may still be effective as monotherapy. Our experience with these novel biological agents is limited, underscoring the need for further studies in this area. Finally, because this study was conducted at a tertiary referral center with expertise in rare autoinflammatory diseases, the findings may not be fully generalizable to all patients with SAIDs.

In conclusion, ACT provides significant clinical benefits for selected patients with refractory SAIDs, despite challenges such as secondary loss of efficacy and infection risks. However, important questions remain regarding the associated costs, the optimal timing of ACT initiation, its optimal duration, and the most effective treatment combinations, underscoring the need for further research in these areas.

Table 5.

Infectious episodes observed in SAIDs patients with ACT regimen.

Pt #	SAID	ACT regimen	ACT indication	Infection	Hospital admission	Pre-ACT infections
1	FMF	CAN+UPA	IBD & SPA	Staphylococcal skin abscess	Yes	Skin abscesses
2	FMF	CAN+ADA	IBD & NCA	Skin abscess	Yes	Polymicrobial skin abscesses
3	FMF	CAN+ADA	IBD & NCA	Sinusitis and paronychia	No	No
4	PAPA	ADA+CAN+RUX	PG	Zona zoster	No	No
5	PAPA	GOL+RUX	NCA	Staphylococcal skin abscess and cellulitis	Yes	Infected NCA and skin abscesses
7	PAPA	TCZ+ANA	Arthritis	Recurrent sinusitis	No	Recurrent URTI
9	PAPA	GOL+ANA & GOL+CAN	NCA & PG	Episodes of skin abscesses with Strep Anginosus bacteriemia, dental abscess, osteomyelitis	Yes	Polymicrobial skin abscesses
10	PAPA	CAN+GOL+RUX	PG	Otitis externa, sinusitis, paronychia	No	Recurrent cellulitis & infected PG, otitis externa, sinusitis
11	PAPA	INF+ CAN	PG	Skin abscess with P. aeruginosa and bacteriemia, Pneumonia, Candida stomatitis	Yes	Polymicrobial skin abscesses and infected PG lesions
12	PASH	GOL+CAN	PG & HS	Cellulitis secondary to PG	No	Polymicrobial Hidradenitis
17	VEXAS	ETA+ANA	Myocarditis	Pneumonia	Yes	No
		TOF+ANA	Myocarditis	Ramsey Hunt	Yes	No
19	ROSAH	ADA+ANA	Retinitis	Tooth abscesses	No	Tooth & gum abscesses due to dry mouth
20	TRAPS	ETA+ANA	Amyloidosis	Catheter infection	Yes	No
23	MKD	ETA+ANA (PRN)	Flares	Recurrent UTI and bronchitis	No	Recurrent UTI
25	MKD	ADA+CAN	Flares	Recurrent sinusitis	No	Recurrent URTI
28	USAID	ADA+ANA	PG	Candida stomatitis	No	No
29	USAID	INF+ANA	PG & osteitis	Cellulitis	Yes	Complicated PG
30	USAID	GOL+CAN	PG	Recurrent sinusitis	No	Sinusitis
34	USAID	TCZ+ANA	Flares	Recurrent UTI	No	Recurrent UTI
35	USAID	GOL&ANA	Flares	Parotitis	Yes	No
37	USAID	ADA+ANA	Flares	Condyloma acuminatum exacerbation	No	Condyloma acuminatum

ACT: advanced combination treatment, ADA: adalimumab, ANA: anakinra, CAN: canakinumab, ETA: etanercept, FMF: familial Mediterranean fever, GOL: golimumab, HS: hidradenitis suppurativa, IBD: inflammatory bowel disease, IFX: infliximab, MKD: mevalonate kinase deficiency, NCA: nodulocystic acne, PG: pyoderma gangrenosum, PASH: pyoderma gangrenosum, acne, and hidradenitis suppurativa syndrome, PG: pyoderma gangrenosum, ROSAH: retinal dystrophy, optic nerve edema, splenomegaly, anhidrosis and headache syndrome, RUX: ruxolitinib, SAPHO: Synovitis, Acne, Pustulosis, Hyperostosis, and Osteitis, SPA: spondyloarthritis, TCZ: tocilizumab, TOF: tofacitinib, TRAPS: tumor necrosis factor receptor-associated periodic syndrome, UPA: upadacitinib, URTI: upper respiratory tract infection, USAID: undifferentiated systemic autoinflammatory disease, UST: ustekinumab, UTI: urinary tract infection, VEXAS: vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic syndrome.

Data availability:

Due to legal constraints related to the underlying data, individual-level data from this study is not publicly available.