In total, 253 patients were enrolled into the study and were included in the full analysis set (Fig. 2). Two out of the 50 participants who discontinued the study during Epoch 1 withdrew consent because of AEs. One patient experienced fatigue, which was of moderate severity, probably related to treatment and later resolved/recovered. The AEs experienced by the second patient included pyrexia, headache, and myalgia, all of which were mild, possibly related to treatment, and later resolved/recovered. One of these patients was negative for binding anti-rHuPH20 antibodies and the other patient did not undergo testing. No participants discontinued the study during Epoch 2 owing to AEs. None of the enrolled patients tested positive for binding anti-rHuPH20 antibodies at any time prior to enrollment. Among 14 participants who were positive for treatment-emergent anti-rHuPH20-binding antibodies during Epoch 1, 13 entered Epoch 2 (1 patient withdrew consent). Most participants in the study had a history of Ig treatment (n = 242/253, 95.7%) and many patients had ongoing fSCIG 10% treatment at enrollment (n = 141/169, 83.4%). Participant demographics and characteristics are summarized in Table 1.

Patient disposition. aTwo fatal AEs (possible stress-related complication of chronic lymphocytic leukemia; cardiogenic shock) were reported, neither of which was considered to be treatment-related. bOther reasons included: unable to keep timely appointments (n = 2); switched to home infusion and would not complete diary and/or questionnaire (n = 2); switched to home infusion (n = 1); switched to another treatment (n = 1); transferred care (n = 1); did not restart fSCIG 10% (n = 1). cParticipant eligible for entry into Epoch 2 (anti-rHuPH20 antibody titer of ≥ 1:160 in Epoch 1) but withdrew consent before entry. AE adverse event, fSCIG facilitated subcutaneous immunoglobulin, rHuPH20 recombinant human hyaluronidase

Participants received fSCIG 10% treatment for a median (interquartile range [IQR]) duration of 10.0 (3.5–11.8) months. Participants who discontinued fSCIG 10% could remain in the study for anti-rHuPH20 antibody and safety assessments. Among the 37 participants who discontinued fSCIG 10% permanently at an Epoch 1 visit, 14 participants remained in the study. During Epoch 2, two participants discontinued fSCIG 10% and 21 participants discontinued fSCIG 10% at a completion/termination visit. fSCIG 10% discontinuation due to AEs occurred in four, one, and six participant(s) at Epoch 1, Epoch 2, and the completion/termination visits, respectively.

Infusion parameters for participants with available treatment data during follow-up are summarized in Table 2. The majority of infusions were administered every 4 weeks (1197/2201, 54.4%, n = 225) and at home (1395/2230, 62.6%, n = 227), with a median (IQR) number of infusion sites of 2 (2.0–2.0, n = 203). Of participants with available data (n = 227), most (n = 144, 63.4%) used two sites, 30 (13.2%) used only a single site, and 29 (12.8%) used one or two sites. The most frequently used infusion site location was the abdomen, including the left upper abdomen (22.5%), right upper abdomen (22.3%), right lower abdomen (18.1%), and left lower abdomen (17.0%).

Overall, 98.5% of infusions were administered without rate reduction, interruption, or discontinuation due to AEs. Treatment-related, non-serious AEs (including infections) were experienced by 52 participants (20.6%, 284 events) (Table 3). Overall, the rate of all SAEs was 0.207 per person-year (95% confidence interval [CI]: 0.159–0.266). The incidences of serious, non-serious, and local and systemic AEs related to fSCIG 10% are shown in Table 4. Most treatment-related AEs were either mild or moderate in severity. Two participants (0.8%) each experienced one treatment-related SAE (aseptic meningitis and deep vein thrombosis). Two fatal AEs (possible stress-related complication of chronic lymphocytic leukemia, and cardiogenic shock; 0.8%) were reported, neither of which was considered to be treatment-related. The incidence of all non-serious AEs (number of events per person-year, including infections) was lower among patients previously treated with fSCIG 10% (3.068, 532 events, n = 98/141) than those naive to fSCIG 10% treatment at enrollment (5.790, 161 events, n = 21/28). There was also a marked difference in the number of events per person-year of systemic non-serious AEs between participants previously treated with fSCIG 10% (0.623, 108 events, n = 21/141) and those naive to fSCIG 10% treatment (1.690, 47 events, n = 5/28).

Of 196 participants enrolled, for whom at least one anti-rHuPH20 assessment was completed, 14 (7.1%) tested positive for treatment-emergent anti-rHuPH20-binding antibodies (maximum titer 1:10 240). Samples from four patients with anti-rHuPH20 antibody titers ≥ 1:10 000 did not cross-react with other human hyaluronidases (hyaluronidase-1, hyaluronidase-2, and hyaluronidase-4). The incidence of a positive anti-rHuPH20 antibody test according to the specific PID diagnosis was highest in participants with specific antibody deficiency (11.8%; 95% CI: 3.3–34.3) followed by hypogammaglobulinemia (9.1%; 95% CI: 3.1–23.6) and common variable immunodeficiency (6.6%; 95% CI: 3.5–12.0). Among 150 participants with an assessment at baseline, eight (5.3%) were positive for anti-rHuPH20 antibodies, six (75.0%) of whom were treated with fSCIG 10% prior to enrollment.

Overall, among participants with positive antibody titers, titers generally increased over Epoch 1, with varying trends over time in Epoch 2 (Fig. 3). However, the incidence of non-serious AEs related to fSCIG 10% treatment did not increase after the first positive anti-rHuPH20 titer. No participants experienced any treatment-related SAEs before or after the first positive antibody test and no neutralizing antibodies were detected.

Anti-rHuPH20 antibody titers over Epochs 1 and 2. rHuPH20, recombinant human hyaluronidase

Most respondents reported self-administration to be easy (22/56, 39.3%) or very easy (18/56, 32.1%) at Month 12 of Epoch 1 (Supplementary Table S1). During Epoch 1, among respondents with available treatment preference data (n = 91), almost all (98.9%) planned to continue fSCIG 10% treatment.

In general, HRQoL measurements remained stable over the study. The mean (SD) change in the EQ-5D-3L VAS score from baseline to Month 12 of Epoch 1 was 2.5 (15.88) (Supplementary Table S2). The EQ-5D-3L questionnaire results are summarized in Supplementary Table S3. The proportion of respondents who reported some anxiety/depression or problems with activities decreased over Epoch 1 (33.6% at baseline to 23.9% at Month 12 of Epoch 1) (Supplementary Table 31). By contrast, the proportion of respondents reporting some problems with pain/discomfort increased over Epoch 1 (54.1% at baseline to 62.7% at Month 12 of Epoch 1).

SF-36v2 scores were indicative of normal health and remained stable over Epoch 1 (mean [SD] changes from baseline to Month 12 of Epoch 1 were − 0.4 [5.49] and 0.5 [8.16] for physical and mental component summary scores, respectively) (Supplementary Table S4).

At the time of study completion/discontinuation, 56 hospitalization events (event rate [number of events per person-year]: 0.191) and 287 days in hospital (event rate: 0.980) had occurred. In total, 15 participants (event rate [number of events per person-year]: 0.051) had at least one hospitalization event owing to an infection (Supplementary Table S5). There were no abnormal total IgG levels within the 4 weeks prior to or during hospitalization.