Demographics and patient disposition
In total, 124 subjects were screened, 72 enrolled (Table 1), and 70 subjects completed Parts A and B. Two subjects on active treatment did not receive the third immunization, due to severe injection site reactions (ISR) after the second immunization. Fifty subjects received the booster immunization in study Part C, with 49 subjects completing the study (Fig. 1).
Both immunotherapeutics were safe and well tolerated, with no deaths, no treatment-related SAEs (Table 2, Supplementary Table 1) and no subjects withdrawn due to TEAEs (any AE occurring after the first treatment). Among 72 enrolled subjects, 67 (93.1%) experienced at least one systemic TEAE and 71 (98.6%) subjects experienced ISRs.
Related systemic TEAEs were experienced by 46% (AT04A), 75% (AT06A), and 58% (placebo) of subjects. However, the majority of systemic TEAEs were of mild or moderate intensity, with only one severe systemic TEAE classified as probably related to immunization (AT06A), comprising a transient episode of asthma rapidly controlled with inhalation of fenoterol/ipratropium bromide (Supplementary Tables 2 and 3). The most commonly experienced systemic TEAEs were headache, fatigue, and myalgia.
Injection site reactions (ISRs) accounted for 63% of the recorded AEs classified as related to study treatment, occurring more frequently in active treatment groups (Table 2). Erythema, induration, swelling, granuloma, and pain were reported most frequently, mostly mild or moderate in intensity. Severe ISRs were reported in 8 (16.7%) subjects, most were transient, three (6.3%) required short-term medication. The frequency of injection site reactions was constant over time during the three priming immunizations. However, the number of subjects with severe ISRs increased after the booster with AT04, but not AT06.
Baseline safety parameters were within physiological ranges and were similar across treatment groups. There were no clinically significant changes over time in vital signs, ECG, hematology, coagulation, clinical chemistry of immune parameters (including the absence of complement activation, or an increase in circulating immune complexes) in both treatment groups relative to placebo.
Regular analyses of PMBCs obtained at baseline, following the priming immunizations and until six weeks after the booster immunization, ruled out a systemic activation of cytotoxic T cells specific for PCSK9.
Both SAIT candidates induced a strong PCSK9-reactive antibody response
Both AT04A (Fig. 2a) and AT06A (Fig. 2b), induced a strong and long-lasting humoral immune response (isotype pattern containing IgG1, IgG3 and IgG4) against the immunizing peptides AT04 and AT06. The geometric mean of the half-max titer increased after two immunizations (i.e., within 6 weeks) from baseline (titer ≤ 1:10) to 1:159 and 1:101 in the AT04A and AT06A group, respectively, with values after the third injection (week 10) only slightly higher at 1:169 and 1:159, respectively. Titers declined over time with an elimination half-life of approximately 12 weeks, reaching baseline levels at week 60. Booster immunization reactivated a strong antibody response with rapid onset in both treatment groups (Fig. 2).
Anti-PCSK9 epitope geometric mean titers increased after three immunizations from baseline (≤ 1:10) to 1:134 and 1:125 in the AT04A and AT06A group respectively, and were marginally below the titers against the immunizing peptides, indicating strong cross-reactivity of treatment-induced antibodies to the target. Titers against the native PCSK9 target epitope displayed a time profile very similar to that observed with AT04 and AT06 (Fig. 2a, b). Seroconversion, defined as a fourfold increase of titers over baseline, was obtained in 21 (87.5%) and 23 (95.8%) subjects in the AT04A and AT06A groups, respectively. Placebo-immunized subjects exhibited no immune response against AT04, AT06 or the PCSK9 target epitope (Fig. 2). The concentration of IgG antibodies against the immunizing peptides was analyzed by CFCA using surface plasmon resonance. Group mean antibody levels reached serum concentrations around 1–2 µg/mL 2 weeks after the second immunization (week 6) and 2 weeks (week 62) after the booster (data not shown).
No differences in total and free PCSK9 concentrations between groups were detected during the study (Supplementary Tables 4 and 5).
Impact of the immunizations on lipid parameters
Analysis of the relative change of lipid parameters from baseline was performed post hoc, based on the 48 subjects who received the booster immunization and completed the study per protocol, with one subject excluded (prohibited concomitant medication, atorvastatin).
The immune response against the PCSK9 target epitope was equally high in both treatment groups, however, the effect on lipid metabolism was more pronounced in AT04A-immunized subjects, with a mean peak reduction in serum LDLc of 11.2% and 13.3% from baseline at weeks 20 and 70, respectively, compared to placebo. Also, a statistically significant reduction of LDLc over the whole study period of 90 weeks was observed in the AT04A group with a mean reduction of 7.2% (P < 0.0001). In contrast, no significant difference to placebo was observed for the AT06A group (Fig. 3b, Supplementary Table 6). The relative change in LDLc values from baseline in subjects of the AT04A (n = 14) and the placebo (n = 18) groups at week 70 (interim analysis) and week 90 (final analysis) are illustrated in Fig. 4. As expected, the individuals in the placebo group showing an increase or decrease of LDLc levels compared to baseline are equally distributed. In the AT04A treated group, however, 9 and 12 out of 14 subjects at week 70 and 90, respectively, showed LDLc lowering (Fig. 4a, b).
All subjects in the AT04A group demonstrating a strong immune response, with a PCSK9 target epitope titer at week 62 > 50 (expected to have PCSK9-specific serum antibody concentrations > 1 µg/ml), showed a decrease in LDLc values from baseline at week 70 and week 90 (Fig. 4c). In this group, the maximal individual LDLc decrease was 39% at week 90, suggesting that higher immunogenicity has a more favourable impact on lipid metabolism.
Although maximal antibody titers were observed at week 10 (study Part A) and week 62 (Part C; Fig. 2a), the maximal LDLc decrease was detected several weeks later in week 20 (Part A) and week 90 (Part C), respectively (Fig. 3a). In both active treatment groups, HDLc did not change over time (data not shown).
To understand the relationship between immunogenicity and effect on lipid parameters, a correlation analysis was performed. An inverse correlation between LDLc levels and immune response against AT04A was observed (especially after the booster) with an increasingly significant inverse correlation covering the period from week 66 (4 weeks after the peak titer at week 62) to week 82 with a Spearman’s correlation coefficient of up to r = − 0.51, supporting the theory that higher antibody concentrations lower LDLc.