Patient Disposition and Baseline Characteristics

The Russian and RoW cohorts comprised 317 (Q4W, n = 119; Q8W, n = 88; PBO → Q4W, n = 110) and 685 (Q4W, n = 216; Q8W, n = 246; PBO → Q4W, n = 223) patients, respectively. Of these, four patients in the Russian cohort (Q4W, n = 3; Q8W, n = 1) and 34 patients in the RoW cohort (Q4W, n = 24; Q8W, n = 10) discontinued between weeks 24 and 52, and a further nine (Q4W, n = 7; Q8W, n = 2) and 28 (Q4W, n = 19; Q8W, n = 9) patients from the Russian and RoW cohorts, respectively, discontinued between weeks 52 and 100 (Fig. 1).

Fig. 1

Patient disposition. At week 16, all patients with < 5% improvement in SJC and TJC were eligible for early escape. Patients who met early escape criteria were allowed to initiate or increase the dose of one of the permitted concomitant medications up to the maximum allowed dose at the discretion of the investigator. aIncludes both patients from the guselkumab Q4W group and those from the placebo group who crossed over to guselkumab Q4W at week 24. Q4W every 4 weeks, Q8W every 8 weeks, RoW rest of the world, SJC swollen joint count, TJC tender joint count

Although baseline patient demographic and disease characteristics were generally consistent between the geographic cohorts, several nominally significant differences were noted. The Russian cohort was somewhat younger (44.9 vs. 46.8 years), had higher CRP levels (1.4 vs. 0.8 mg/dL), and had more severe skin disease (mean PASI score, 12.4 vs. 8.1) at baseline than the RoW cohort. The Russian cohort was also characterized by lower rates of corticosteroid use (12.0% vs. 20.1%) and higher rates of nonsteroidal anti-inflammatory drug (NSAID) use (82.6% vs. 56.8%) at baseline (Table 1). Of the 685 patients in the RoW cohort, 82% were recruited from sites in Eastern Europe (Bulgaria, Czech Republic, Estonia, Hungary, Lithuania, Latvia, Poland, and Ukraine).

Table 1 Baseline patient demographics and disease characteristics of the Russian vs. RoW cohortsEfficacyACR20 and ACR50 Response

In the Russian cohort, higher proportions of patients receiving guselkumab achieved an ACR20 response at week 24 (Q4W, 60%; Q8W, 51%) vs. patients receiving PBO (30%). ACR20 response rates increased through week 52 (range across guselkumab-randomized patients, 69–78%); at week 100 (DISCOVER-2), ACR20 response rates ranged from 77% to 80% and were consistent between the crossover and guselkumab-randomized groups. Response patterns for the Russian cohort were consistent with the RoW cohort across all timepoints (Supplementary Table 1, Fig. 2).

Fig. 2

Proportion of biologic-naïve patients achieving ACR20 responses at a week 24, b week 52, and c week 100, and ACR50 responses at d week 24, e week 52, and f week 100, by treatment group. Data for weeks 24 and 52 are from the pooled populations of the DISCOVER-1 and -2 trials. Data for week 100 are only from patients from DISCOVER-2. Data are reported using non-responder imputation. ACR20 ≥ 20% improvement in the American College of Rheumatology criteria, ACR50 ≥ 50% improvement in the American College of Rheumatology criteria, Q4W every 4 weeks, Q8W every 8 weeks, RoW rest of the world

Additionally, trends in ACR50 response were similar to those observed for ACR20 for both cohorts across all timepoints (Supplementary Table 1, Fig. 2).

PASI90 Response

In the Russian cohort, rates of PASI90 response in patients with ≥ 3% BSA and IGA ≥ 2 were higher with guselkumab (Q4W, 69%; Q8W, 56%) vs. PBO (7%) at week 24 and rates increased through week 52 (range 72–75%); at week 100 (DISCOVER-2), the rate of PASI90 response was 62–74% across the treatment groups. Response rates in the crossover group were consistent with those in guselkumab-randomized patients. Response patterns for the Russian cohort were consistent with the RoW cohort across all timepoints (Supplementary Table 1).

Resolution of Enthesitis and Dactylitis (in Patients with Domain Involvement at Baseline)

In the Russian cohort, higher proportions of patients receiving guselkumab achieved complete resolution of enthesitis and dactylitis at week 24 vs. patients receiving PBO. Between weeks 24 and 52, the rates of enthesitis and dactylitis resolution increased across all treatment groups; at week 100 (DISCOVER-2), the proportion of guselkumab-randomized patients with enthesitis resolution was 59–67% and with dactylitis resolution was 74–91%. At week 100, rates of dactylitis resolution were higher in the Russian vs. the RoW cohort; resolution of enthesitis and dactylitis was generally consistent between the Russian and RoW cohorts at all other timepoints. In both cohorts, response rates in the crossover groups were consistent with those in guselkumab-randomized patients (Supplementary Table 1, Fig. 3).

Fig. 3

Proportion of biologic-naïve patients achieving enthesitis resolution at a week 24, b week 52, and c week 100; and dactylitis resolution at d week 24, e week 52, and f week 100 by treatment group in patients with enthesitis or dactylitis at baseline. Data for weeks 24 and 52 are from the pooled populations of the DISCOVER-1 and -2 trials. Data for week 100 are only from patients from DISCOVER-2. Data are reported using non-responder imputation. Q4W every 4 weeks, Q8W every 8 weeks, RoW rest of the world

Patient-Reported OutcomesHAQ-DI Scores

In the Russian cohort with baseline HAQ-DI ≥ 0.35, the proportion of patients who achieved a clinically meaningful improvement of ≥ 0.35 in HAQ-DI score was higher with guselkumab (Q4W, 51%; Q8W, 46%) vs. PBO (33%) at week 24. These proportions increased through week 52 (range 49–60%); at week 100 (DISCOVER-2), the proportion of guselkumab-randomized patients achieving an improvement of ≥ 0.35 in HAQ-DI score generally improved from week 52. Response rates in the crossover group were consistent with those in guselkumab-randomized patients. Response patterns in the Russian cohort were consistent with the RoW cohort at all timepoints (Supplementary Table 1, Fig. 4).

Fig. 4

Proportion of biologic-naïve patients achieving an improvement of ≥ 0.35 in HAQ-DI at a week 24, b week 52, and c week 100, by treatment group in patients with baseline HAQ-DI ≥ 0.35. Data for weeks 24 and 52 are from the pooled populations of the DISCOVER-1 and -2 trials. Data for week 100 data are only from patients from DISCOVER-2. Data are reported using non-responder imputation. HAQ-DI Health Assessment Questionnaire–Disability Index, Q4W every 4 weeks, Q8W every 8 weeks, RoW rest of the world

Patient-Reported Pain

In the Russian cohort, improvements in the least squares mean changes in patient pain VAS scores from baseline were larger for patients receiving guselkumab (Q4W, − 4.49; Q8W, − 6.28) vs. PBO (− 1.56) at week 4. Improvements continued through week 52 in guselkumab-randomized patients (range − 26.98 to − 27.69); at week 100 (DISCOVER-2), least squares mean changes in patient pain VAS scores from baseline ranged from − 34.53 to − 35.30. Response rates in the crossover group were consistent with those in guselkumab-randomized patients. Changes in patient-reported pain in the Russian cohort were consistent with those in the RoW cohort through week 100 (Supplementary Table 1).

Composite Measures of Disease ActivityDAPSA LDA/Remission

In the Russian cohort, higher proportions of patients receiving guselkumab achieved DAPSA LDA/remission at week 24 (Q4W, 31%; Q8W, 33%) vs. patients receiving PBO (19%). These proportions increased through week 52 (52% for both guselkumab-randomized groups); at week 100 (DISCOVER-2), the proportion of guselkumab-randomized patients achieving DAPSA LDA/remission was 60–65%. Response rates in the crossover group were consistent with those in guselkumab-randomized patients. Rates of DAPSA LDA/remission were generally lower in the Russian vs. the RoW cohort through week 52 and were consistent across the two cohorts at week 100 (Supplementary Table 1).

Least squares mean changes in all individual DAPSA component scores, including CRP levels, improved over time through week 100 (Supplementary Fig. S2).

Minimal Disease Activity

In the Russian cohort, higher proportions of guselkumab-treated patients achieved MDA at week 24 (Q4W, 18%; Q8W, 19%) vs. patients receiving PBO (6%). These proportions increased through week 52 (range 31–34%); at week 100 (DISCOVER-2), the rate of MDA was 39–44% across all treatment groups. Response rates in the crossover group were consistent with those in guselkumab-randomized patients. Response patterns in the Russian cohort were consistent with the RoW cohort at all timepoints (Supplementary Table 1, Fig. 5).

Fig. 5

Proportion of biologic-naïve patients achieving MDA responses at a week 24, b week 52, and c week 100, by treatment group. Patients were considered to have achieved an MDA response if they met five of the following seven criteria: TJC ≤ 1, SJC ≤ 1, PASI ≤ 1, patient pain VAS score ≤ 15, PtGA ≤ 20, HAQ-DI ≤ 0.5, and ≤ 1 tender entheseal point. Data for weeks 24 and 52 are from the pooled populations of the DISCOVER-1 and -2 trials. Data for week 100 are only from patients from DISCOVER-2. Data are reported using non-responder imputation. HAQ-DI Health Assessment Questionnaire–Disability Index, MDA minimal disease activity, PASI Psoriasis Area and Severity Index, PtGA Patient Global Assessment of Disease Activity, Q4W every 4 weeks, Q8W every 8 weeks, RoW rest of the world, SJC swollen joint count, TJC tender joint count, VAS visual analog scale

Symptoms of Axial Involvement

In the Russian cohort, improvement in BASDAI scores in patients with imaging-confirmed axial involvement at baseline was higher with guselkumab (Q4W, − 1.96; Q8W, − 1.98) vs. PBO (− 0.62) at week 24; at week 100 (DISCOVER-2), changes in the guselkumab-randomized patients ranged from − 2.32 to − 2.53. Response rates in the crossover group were consistent with those in guselkumab-randomized patients. Changes in BASDAI scores were smaller in the Russian cohort vs. the RoW cohort across all timepoints (Supplementary Table 1; Fig. 6), and similar trends were observed in mBASDAI and ASDAS scores through week 100 (Supplementary Fig. S3).

Fig. 6

Least squares mean change from baseline in BASDAI scores in biologic-naïve patients at a week 24, b week 52, and c week 100, by treatment group in patients with PsA with imaging-confirmed axial involvement. Data for weeks 24 and 52 are from the pooled populations of the DISCOVER-1 and -2 trials. Data for week 100 are only from patients from DISCOVER-2. Data are reported using non-responder imputation. BASDAI Bath Ankylosing Spondylitis Disease Activity Index, PsA psoriatic arthritis, Q4W every 4 weeks, Q8W every 8 weeks, RoW rest of the world

In the Russian cohort, patient-reported spinal pain also improved (measured by least squares mean changes in patient-reported spinal pain) through week 24 in patients with axial PsA (Q4W, − 1.65; Q8W, − 2.45; PBO, − 0.24). Improvements continued through week 52 (range − 1.88 to − 2.36); at week 100 (DISCOVER-2), changes in patient-reported spinal pain were − 2.42 to − 2.37 for guselkumab-randomized patients. Response rates in the crossover group were consistent with those in guselkumab-randomized patients. Changes in patient-reported spinal pain were smaller in the Russian cohort vs. the RoW cohort across all timepoints (Supplementary Table 1).

Safety

Rates of AEs with guselkumab were similar to those with PBO through week 24 and similar across all treatment groups in the Russian cohort (Table 2). The rate of AE occurrence at week 24 was 146.9 per 100 PYs (n = 83, 40.1%) in patients treated with guselkumab, vs. 162.4 per 100 PYs (n = 47, 42.7%) in patients receiving PBO. The rate of patients experiencing MACE was 1.0 per 100 PYs (n = 1, 0.5%) in patients treated with guselkumab. Among patients receiving guselkumab, the rate of patients experiencing serious AEs was 3.1 per 100 PYs (n = 3, 1.4%) and the rate of patients experiencing an AE leading to drug discontinuation was 3.1 per 100 PYs (n = 3, 1.4%). Amongst patients receiving PBO, these rates were 5.9 per 100 PYs (n = 3, 2.7%) and 3.9 per 100 PYs (n = 2, 1.8%), respectively. No increase in time-adjusted incidence rates was observed through week 52.

Table 2 Summary of adverse events in the biologic-naïve Russian cohort of the DISCOVER-1 and -2 trials

Through week 112 in DISCOVER-2, the rate of AE occurrence among patients treated with guselkumab was 89.7 per 100 PYs (n = 170, 62.7%). Among patients receiving guselkumab, the rate of patients experiencing serious AEs was 4.5 per 100 PYs (n = 21, 7.7%) and the rate of patients experiencing an AE leading to drug discontinuation was 0.9 per 100 PYs (n = 5, 1.8%).

No deaths or malignancies occurred, and no cases of inflammatory bowel disease, opportunistic infection, or active TB were reported.