Selective TYK2 inhibitors vs. pan-JAK inhibitors

The JAK-STAT Framework and the Pan-JAKi Safety Problem

The JAK-STAT pathway is the primary signalling conduit for numerous cytokines critical in inflammation and immunity — and the central reason why both selective TYK2 inhibitors and pan-JAK inhibitors converge on the same biological axis, yet produce fundamentally different safety outcomes. Cytokine receptors lack intrinsic kinase activity; instead, they associate with specific JAK family members (JAK1, JAK2, JAK3, TYK2). Upon cytokine binding, receptor-associated JAKs trans-phosphorylate each other, activating their kinase domains (JH1). Activated JAKs then phosphorylate tyrosine residues on the receptor cytoplasmic tails, creating docking sites for Signal Transducer and Activator of Transcription (STAT) proteins. STATs are phosphorylated by JAKs, dimerize, translocate to the nucleus, and drive gene transcription.

The problem with first-generation pan-JAK inhibitors (tofacitinib, upadacitinib) is precisely their breadth. By competitively blocking the catalytic JH1 domain of JAK1, JAK2, and JAK3 simultaneously, they disrupt signalling across all pathways that depend on these kinases — including those entirely unrelated to the immune-mediated diseases being treated. Hematopoietic cytokines — EPO (erythropoiesis), TPO (thrombopoiesis), G-CSF, and GM-CSF (myelopoiesis) — signal primarily through JAK2. Inhibiting JAK2 therefore directly causes dose-dependent anemia, thrombocytopenia, and neutropenia. Gamma chain cytokines (IL-2, IL-4, IL-7, IL-15, IL-21), essential for lymphocyte development and homeostasis, require JAK1 and JAK3 — their suppression drives lymphopenia and broad immunosuppression.

The thromboembolic risk associated with pan-JAK inhibitors is particularly well-documented. JAK2 mediates signalling for key haemostatic factors, and its inhibition may alter platelet function, endothelial cell responses, and coagulation factor levels. The ORAL Surveillance trial of tofacitinib versus TNF inhibitors in rheumatoid arthritis highlighted significantly increased MACE and thromboembolic event risk, particularly in high-risk patients — a finding that prompted black box warnings across the pan-JAK inhibitor class, as recognised by regulators including those tracked by the FDA.

How Allosteric JH2 Inhibition Achieves Cytokine Selectivity

Selective TYK2 inhibitors like deucravacitinib and zasocitinib achieve their remarkable specificity by targeting a fundamentally different domain: the regulatory pseudokinase domain (JH2) of TYK2, not the catalytic JH1 domain targeted by pan-JAK inhibitors. This distinction is the entire basis for their improved safety profile. TYK2’s JH2 domain is structurally conserved but enzymatically inactive due to key amino acid substitutions in its catalytic pocket. Its primary function is to autoinhibit the JH1 domain, maintaining TYK2 in a basal, inactive state — a regulatory role unique in its importance among JAK family members.

When deucravacitinib binds to the unique pocket within the TYK2 JH2 domain, it induces a conformational change that strengthens the interaction between the JH2 and JH1 domains — effectively “locking” TYK2 in its inactive state. This is an allosteric mechanism: the drug does not compete with ATP at the catalytic site, but instead stabilises a regulatory conformation. The selectivity arises from two structural features. First, the drug-binding pocket within the JH2 domain is highly unique to TYK2; the amino acid sequence and three-dimensional structure of this pocket differ significantly from the JH2 domains of JAK1, JAK2, and JAK3. Second, the functional consequence of JH2 stabilisation only prevents TYK2-dependent signalling — it does not directly block the catalytic activity of other JAKs.

The cytokine pathways that critically depend on TYK2 are precisely those implicated in psoriasis and SLE. The IL-12 receptor (IL-12Rβ1/IL-12Rβ2) requires TYK2 (bound to IL-12Rβ1) paired with JAK2. The IL-23 receptor (IL-23R/IL-12Rβ1) requires TYK2 (bound to IL-12Rβ1) paired with JAK2. The type I IFN receptor (IFNAR1/IFNAR2) requires TYK2 (bound to IFNAR1) paired with JAK1. IL-12 and IL-23 drive Th1/Th17 differentiation and IL-17 production, central to plaque psoriasis pathogenesis. Hyperactivation of the type I IFN pathway is a hallmark of SLE pathogenesis. By selectively stabilising the inactive conformation of TYK2, these inhibitors potently block all three of these disease-relevant pathways while leaving hematopoietic and lymphocyte homeostasis pathways intact — a selectivity profile that, according to Nature-published structural biology research, was made possible by rational exploitation of the JH2 regulatory architecture.

Clinical Evidence in Plaque Psoriasis: Efficacy Without Hematologic Cost

Deucravacitinib’s clinical differentiation from pan-JAK inhibitors in plaque psoriasis is most clearly demonstrated by the POETYK PSO-1 and PSO-2 trials, where it achieved high efficacy comparable to a biologic while producing a safety profile resembling apremilast rather than a JAK inhibitor. In both trials, deucravacitinib demonstrated superior efficacy to placebo and apremilast, and non-inferior efficacy to the IL-23p19 inhibitor ustekinumab at Week 16 on PASI 75 and sPGA 0/1 endpoints. Responses were durable through 52 or more weeks, positioning it as a high-efficacy oral agent.

The safety differentiation in the psoriasis programme is striking. There were no clinically significant decreases in mean haemoglobin, neutrophil, or platelet counts compared to placebo — no anemia, neutropenia, or lymphopenia warnings are included in the label, and laboratory monitoring for cytopenias is not required. There was no signal of increased DVT, PE, or MACE risk in the psoriasis clinical programme or pooled analyses. Rates of serious infections and herpes zoster were comparable to placebo and apremilast — significantly lower than typically observed with pan-JAK inhibitors. Lipid changes were mild and transient, similar to apremilast and less pronounced than some pan-JAK inhibitors, with no impact on atherogenic ratios. No increased malignancy signal was observed.

“Deucravacitinib’s label lacks the major warnings associated with pan-JAK inhibitors — no black box for thromboembolism, MACE, serious infections, or malignancy — despite achieving efficacy non-inferior to ustekinumab.”

In contrast, tofacitinib is approved for psoriasis in some regions but carries black box warnings for serious infections, mortality, malignancy, MACE, and thrombosis. Upadacitinib’s psoriasis approval includes similar warnings. Dose-dependent decreases in haemoglobin, neutrophils, and lymphocytes are common with both agents, requiring monitoring. These risks restrict their use to later-line positions — generally after biologics have been tried. Deucravacitinib, by contrast, is positioned as a first-line, high-efficacy oral option, a distinction that reflects the direct mechanistic consequence of sparing JAK2 and JAK1/JAK3. The PatSnap drug discovery intelligence platform tracks more than 2 billion data points across the global innovation landscape, including the expanding IP estate around TYK2 selectivity mechanisms.

Clinical Evidence in SLE: Targeting the IFN Signature Safely

In systemic lupus erythematosus, the case for selective TYK2 inhibition is mechanistically compelling and increasingly supported by clinical data — particularly given the prominent role of type I IFN hyperactivation in SLE pathogenesis and the specific thromboembolic risks faced by SLE patients. The Phase 2 PAISLEY trial of deucravacitinib in SLE demonstrated statistically significant and clinically meaningful improvements versus placebo across multiple primary and key secondary endpoints at both doses (3 mg QD and 6 mg QD) at Week 48, including SRI-4, LLDAS, and BICLA. Significant improvements in joint and skin activity, reduction in flares, and steroid reduction were observed, alongside robust reductions in type I IFN gene signatures.

The contrast with baricitinib — a JAK1/JAK2 inhibitor — in SLE is instructive. Baricitinib showed efficacy in Phase 3 trials (SLE-BRAVE-I and SLE-BRAVE-II) for joint and skin manifestations and met the SRI-4 primary endpoint in BRAVE-II, though not in BRAVE-I. However, a significant increase in thromboembolic events (DVT/PE) was observed in the baricitinib SLE programme compared to placebo, leading to a warning and restricted use in its SLE label, permitting use only in patients without high TE risk factors. This is a particularly important limitation in the SLE population, which often comprises young women with potential antiphospholipid syndrome overlap or other cardiovascular risk factors — a clinical reality acknowledged by bodies including the WHO in its autoimmune disease burden assessments. Tofacitinib failed in SLE Phase 2.

Deucravacitinib’s safety profile in SLE was consistent with its extensive psoriasis experience. No clinically meaningful changes in haemoglobin, neutrophils, or platelets were observed. Rates of serious infections and herpes zoster were low and comparable to placebo in PAISLEY — consistent with the clean infection safety profile seen in the psoriasis programme. No increased malignancy signal was observed. Phase 3 trials (NCT05617677, NCT05662568) are ongoing and will provide the confirmatory evidence needed to establish deucravacitinib as a first-in-class oral therapy for SLE with a mechanism directly targeting the core IFN pathway. The PatSnap innovation intelligence reports track the evolving IP landscape around TYK2 inhibition in autoimmune indications beyond psoriasis.

Safety Differentiation: A Head-to-Head Comparison

The mechanistic differences between allosteric TYK2 JH2 inhibition and pan-JAK JH1 inhibition produce a consistent and predictable pattern of safety differentiation across every major risk category. This pattern is not coincidental — each difference traces directly to which JAK family members are spared or inhibited, as documented across clinical programmes and increasingly recognised in regulatory guidance from agencies tracked by the EMA.

This safety differentiation translates into a fundamentally different positioning in treatment algorithms. Pan-JAK inhibitors are generally considered later-line options in psoriasis — often after biologics have been tried — due to their safety profile. Deucravacitinib, by contrast, is positioned as a first-line, high-efficacy oral option in psoriasis, with a label comparable to apremilast in terms of warnings despite achieving efficacy comparable to ustekinumab. In SLE, where the patient population faces particular thromboembolic vulnerability, the absence of a TE signal with deucravacitinib versus the TE warning on baricitinib’s SLE label represents a clinically meaningful differentiation that could determine which patients can safely receive oral JAK pathway therapy. The mechanistic elegance of allosteric TYK2 inhibition translates directly into a superior therapeutic index — and ongoing Phase 3 trials in SLE and studies in other IL-12/IL-23/IFN-driven diseases, including inflammatory bowel disease and psoriatic arthritis, will further define the full potential of this novel class.