Five-Year Results With Patisiran for Hereditary Transthyretin Amyloidosis With Polyneuropathy

Authors: David Adams, Jonas Wixner, Michael Polydefkis, John L. Berk, Isabel M. Conceição, Angela Dispenzieri, Amanda Peltier, Mitsuharu Ueda, Shaun Bender, Kelley Capocelli, Patrick Y. Jay, Elena Yureneva, Laura Obici

Abstract

Question

What is the long-term effect of patisiran, an RNA interference therapeutic, on disease progression and survival in hereditary transthyretin amyloidosis with polyneuropathy (hATTR-PN)?

Findings

In this 5-year open-label extension of a randomized clinical trial including 211 patients with hATTR-PN receiving patisiran, measures of disability, polyneuropathy, and quality of life showed modest changes. Patients who received patisiran in their parent study had better outcomes and survival than those receiving placebo.

Meaning

Patisiran treatment for hATTR-PN should be initiated as quickly as possible after diagnosis.

Introduction

Hereditary transthyretin amyloidosis (hATTR; or variant transthyretin amyloidosis) is an underdiagnosed, progressive, debilitating, and fatal disease. Pathogenic variants of the TTR gene cause the circulating protein to become unstable, misfold, and deposit as amyloid in the nerves, heart, gastrointestinal tract, and musculoskeletal tissues, resulting in heterogeneous clinical presentations. Polyneuropathy and cardiomyopathy are the most serious manifestations.^1,2,3^

Patients with hATTR steadily worsen over time.^4,5,6,7,8^ In late-onset Val30Met cases, patients with polyneuropathy develop difficulty walking within 3 years after symptom onset, require a walking aid within 2 to 3.8 years, and become wheelchair bound within 4 to 7 years.^7,8^ Median survival following the diagnosis of polyneuropathy or cardiomyopathy is 4.7 years and 3.4 years, respectively.^9,10,11,12^

TTR gene silencers degrade TTR messenger RNA via endogenous cellular mechanisms—patisiran and vutrisiran by RNA interference (RNAi) and inotersen and eplontersen by RNase H1—thereby reducing hepatic production of the amyloidogenic protein.^13^ In the 18-month Phase 3 Multicenter, Multinational, Randomized, Double-blind, Placebo-controlled Study to Evaluate the Efficacy and Safety of Patisiran (ALN-TTR02) in Transthyretin (TTR)-Mediated Polyneuropathy (Familial Amyloidotic Polyneuropathy-FAP) (APOLLO) study,^14,15^ patients with hATTR with polyneuropathy (hATTR-PN) receiving patisiran had meaningful improvements in polyneuropathy and quality of life (QOL) compared with placebo and stability or improvement of polyneuropathy and cardiomyopathy manifestations compared with baseline. In the subsequent global open-label extension (OLE) study, patients from APOLLO and a prior phase 2 OLE^16^ study all received patisiran. Previously published interim 12-month data showed sustained clinical stability for patients who received patisiran in the parent studies, and patients who received placebo in the APOLLO study exhibited stability or improvement of disease manifestations after switching to patisiran. However, the disease severity of these patients remained worse than that of patients who had received patisiran in the APOLLO study, illustrating the importance of early treatment intervention.

The phase 2 OLE, APOLLO, and global OLE studies of patisiran comprise the longest consecutively running trials to date of an RNAi therapeutic in ATTR amyloidosis. Here, we present the final results from the global OLE study, the longest experience to date of a TTR gene silencer in adults with hATTR-PN, in which some patients received patisiran for up to 7 years.

Methods

Study Design and Patients

The study followed the guidelines of the International Conference on Harmonisation, the Declaration of Helsinki, and the Health Insurance Portability and Accountability Act of 1996. Written informed consent was obtained from all patients before undergoing any protocol-specific tests or procedures that were not part of routine care. The study protocol and all amendments (Supplement 1) were approved by the institutional review boards and ethics committees at each site.

The full methodology of this multicenter, 5-year, OLE, single-arm study has been published.^17^ Patients were eligible to enter the global OLE if they had completed the phase 3 APOLLO (18 months) or phase 2 OLE (24 months) parent studies. Patients eligible for APOLLO were aged 18 to 85 years with a documented pathogenic TTR variant, a diagnosis of hATTR-PN, a Neuropathy Impairment Score (NIS) of 5 to 130, and a polyneuropathy disability (PND) score of IIIB or lower (ie, able to walk with 2 sticks, 1, or none). Patients in the phase 2 OLE had biopsy-confirmed hATTR with evidence of mild to moderate neuropathy and a Karnofsky Performance Status score of 60% or greater. Patients were excluded from the study if they had received a liver transplant or were above New York Heart Association (NYHA) Class II.

Patients were grouped based on parent study treatment in APOLLO or the phase 2 OLE, referred to as APOLLO placebo, APOLLO patisiran, and phase 2 OLE patisiran. All patients, regardless of parent study treatment, received patisiran during the global OLE.

Procedures and Outcomes

During the global OLE, all patients received patisiran, 0.3 mg/kg, intravenously every 3 weeks for up to 5 years. In the previously published APOLLO parent study, the primary end point was modified NIS+7. Secondary end points included, in hierarchical order, Norfolk Quality of Life-Diabetic Neuropathy (QOL-DN), NIS-weakness, Rasch-Built Overall Disability Scale (R-ODS), 10-m walk test, modified body mass index [mBMI], and Composite Autonomic Symptom Score 31.^14^ In the previously published phase 2 OLE study, the primary objective was to assess safety and tolerability of patisiran. Secondary end points included mNIS+7, European Quality of Life (EuroQoL) 5-dimension questionnaire, R-ODS, 10-m walk test, dynamometric grip strength, and mBMI.^17^ Details of exploratory end points have been previously published.^14,17^ During the global OLE, efficacy outcomes were evaluated at baseline and annually through year 5. These included assessments of ambulatory status (PND score), severity of polyneuropathy (NIS), nutritional status (mBMI), QOL (Norfolk QOL-DN score), and disability (R-ODS score). The end points were not evaluated in a statistical hierarchy.^16^ Norfolk QOL-DN was not collected during the phase 2 OLE.

Concerning safety, adverse events (AEs) were assessed throughout the 5 years of the global OLE and coded according to the Medical Dictionary for Regulatory Activities version 18.0 and higher. Probability of survival and mortality were also assessed; risk factors associated with mortality at parent study baseline and long-term survival are included as a post hoc analysis. The procedures for assessing serum TTR concentration and antidrug antibodies are found in the eMethods in Supplement 2.

Statistical Analyses

Efficacy outcomes are presented descriptively as observed mean (SD) change from global OLE baseline or parent study baseline without imputation for missing values. No formal significance tests were conducted for the clinical assessments, as this OLE study did not include a comparator group. The patisiran safety population consisted of all patients who received 1 or more doses of patisiran up to the data cutoff (December 7, 2022).

Post hoc analyses of survival were based on AE data. The vital status of every patient was known at the time of withdrawal from or completion of the global OLE. Survival was assessed using Kaplan-Meier analysis, with patients monitored from first date of treatment (placebo or patisiran) to last date of follow-up (the earliest of last dose of patisiran plus 90 days, database lock, or date of last contact). Exposure-adjusted mortality rates were calculated as the total number of deaths per 100 patient-years, with only deaths following the first dose of patisiran included.

Univariate Cox regression models assessed the association between characteristics at parent study baseline and survival (baseline characteristics analyzed are shown in eTable 1 in Supplement 2); covariates that showed statistical significance were carried forward into multivariate Cox regression analyses. Multivariate models included a full model that included each variable, a forward selection model, a backward selection model, and a stepwise selection model. Entry or exit criteria for selection approaches were a P value of .15. All statistical analyses were performed using SAS version 9.4 (SAS Institute).

Results

Patients

The global OLE study enrolled 211 of 212 patients from the parent APOLLO or phase 2 OLE studies (49 in the APOLLO placebo group, 137 in the APOLLO patisiran group, and 25 in the phase 2 OLE patisiran group). At the global OLE baseline, the mean (SD) age was 61.3 (12.3) years, 156 patients (73.9%) were male and 55 (26.1%) were female. Overall, 138 (65.4%) completed the global OLE, 117 of whom received patisiran during the parent studies (eFigure 1 in Supplement 2). Of the 73 patients (34.6%) who withdrew from the study, the most common reasons for withdrawal were death (38 [52.1%]), AEs (13 [17.8%]), and patient decision (16 [21.9%]).

At global OLE baseline, APOLLO placebo patients had more severe polyneuropathy than patients in the APOLLO patisiran and phase 2 OLE patisiran groups by multiple metrics, including NIS, Norfolk QOL-DN, PND, and R-ODS (eTable 1 in Supplement 2). The APOLLO placebo group also had higher levels of the heart failure marker, N-terminal prohormone of brain-type natriuretic peptide (NT-proBNP). The phase 2 OLE patisiran group had a higher proportion of patients with the Val30Met genotype and less severe disease at global OLE baseline than patients in either APOLLO group.

Functional Assessments of Polyneuropathy

During the APOLLO parent study, 108 patients treated with patisiran (73.0%) showed preserved (96/148 [64.9%]) or improved (12/148 [8.1%]) ambulatory ability by PND score. Twenty-three of 77 placebo-treated patients (29.9%) were stable, and none improved. In the global OLE, the PND score in the overall population was sustained (76/137 [55.5%]) or improved (13/137 [9.5%]) compared with the global OLE baseline (Figure 1). Of the 49 patients with a PND score IIIA or IIIB (requiring 1 or 2 sticks or crutches to ambulate) at global OLE baseline, 3 (6.1%) improved, 29 (59.2%) remained stable, and 17 (34.7%) worsened over 5 years. Nine (6.6%) patients were wheelchair bound (PND score IV) at year 5, 4 of whom were wheelchair bound at global OLE baseline. At year 5, there were 63 patients (46.0%) walking unaided (PND score I-II) and 65 (47.4%) walking with an aid (PND score IIIA-IIIB).

Similarly, the overall population showed a mean (SD) change from global OLE baseline of 10.9 (14.7) in NIS at year 5 on measures of muscle weakness, reflexes, and sensation. For comparison, the NIS of the APOLLO placebo group had worsened by 25.4 (20.8) points during the 18-month parent study, meaning there was an approximate 87% reduction in rate of worsening when these patients switched to patisiran treatment in the global OLE. In the global OLE, all 3 groups demonstrated sustained treatment benefits throughout the OLE period (Figure 2A).

Profound malnutrition, as quantified by mBMI, is a hallmark of progressive hATTR-PN due to autonomic dysfunction. During the APOLLO parent study, the mBMI (calculated as weight in kilograms divided by height in meters squared times serum albumin in grams per liter) in the placebo group decreased by a mean (SD) of 122.1 (96.7) over 18 months. Following initiation of patisiran in the global OLE, the APOLLO placebo group demonstrated marked improvement, recovering more than half of the mBMI lost during APOLLO (eFigure 2 in Supplement 2). Both parent study patisiran groups showed long-term mBMI stability during the global OLE and mean (SD) mBMI increased in the overall population by 46.4 (120.7) over 5 years.

Patient-Reported Assessments of Polyneuropathy

During the parent study, the APOLLO placebo group experienced worsening QOL with a mean (SD) 19.8 (20.1) increase in Norfolk QOL-DN score over 18 months. Throughout the global OLE, all treatment groups demonstrated a mean (SD) change of 4.1 (16.7), an approximately 94% slower rate of worsening compared with the placebo group during the parent study (Figure 2B).

Other functional measures revealed similar outcomes. At the end of APOLLO, the placebo group had greater difficulty performing activities of daily living, as demonstrated by a mean (SD) change of –9.9 (7.5) in R-ODS score. During the global OLE, R-ODS scores showed a mean (SD) change of –3.7 (6.2) in the overall population over 5 years, an approximate 89% reduction in the rate of worsening compared with the placebo group during APOLLO (Figure 2C).

Mortality

Overall, 41 patients (19.4%) died during the global OLE and all deaths were considered unrelated to patisiran by the investigators (eTable 2 in Supplement 2 lists causes of death). Survival rates were higher among patients who received patisiran in the parent studies compared with those who received placebo in the parent study and switched to patisiran on entering the global OLE study (Figure 3). The exposure-adjusted mortality rate per 100 patient-years was 12.8 (95% CI, 8.2-19.8) in the APOLLO placebo group, 3.3 (95% CI, 2.3-4.9) in the APOLLO patisiran group, 1.7 (95% CI, 0.5-5.2) in the phase 2 OLE patisiran group, and 4.4 (95% CI, 3.3-5.8) in the overall global OLE population.

Risk Factors Associated With Survival

Risk factors for mortality at the parent study baseline were evaluated (eTable 3 in Supplement 2). Univariate analyses identified associations between mortality and NT-proBNP level (>3000 ng/L vs ≤3000 ng/L), familial amyloid polyneuropathy stage (II/III vs I), genotype (late-onset V30M vs non-V30M vs early-onset V30M), left ventricular wall thickness (≥1.5 cm vs <1.5 cm), left ventricular mass (≥243.67 g vs <243.67 g), NYHA classification (II vs I), cardiac subpopulation (yes vs no), parent study treatment (placebo vs patisiran), National Amyloidosis Centre ATTR amyloidosis stage (2/3 vs 1), NIS (≥56.68 vs <56.68), and PND score (III/IV vs I/II) (eTable 4 in Supplement 2). In both full and selection-based multivariate analyses, only familial amyloid polyneuropathy stage (full hazard ratio [HR], 4.89; 95% CI, 1.69-14.12; P = .003; selection-based HR, 5.74; 95% CI, 2.76-11.92; P < .001) and parent study treatment (full HR, 5.52; 95% CI, 2.91-10.48; P < .001; selection-based HR, 5.19; 95% CI, 2.79-9.63; P < .001) were associated with mortality; all variable selection approaches resulted in the same variables being selected (Table 1).

Safety

Safety outcomes in patients during the placebo-controlled APOLLO trial have been previously published.^15^ During the global OLE period, AEs leading to study withdrawal occurred in 47 patients (22.3%). The rates of AEs and serious AEs were higher in the APOLLO placebo than the APOLLO patisiran and phase 2 OLE patisiran groups. Infusion-related reactions were the most common treatment-related AE (n = 34 [16.1%]). The types of AEs reported were similar to those observed in APOLLO, and the exposure-adjusted event rates were similar or lower than in APOLLO (Table 2). No safety concerns were identified during the global OLE, including concerns that could be related to the functions of TTR in carrying vitamin A or thyroid hormone in the circulation. Results relating to reduction of serum TTR and presence of antidrug antibodies can be found in the eResults in Supplement 2.

Discussion

The present analyses of the global OLE study support the benefits of patisiran for the long-term treatment of hATTR amyloidosis with polyneuropathy. The results build on the placebo-controlled APOLLO study, which first demonstrated the efficacy and safety of patisiran. The phase 2 OLE and APOLLO patisiran groups, which received patisiran for up to 7 years, demonstrated the greatest benefit and sustained clinical stability. Across multiple measures, the APOLLO placebo group experienced substantial disease progression over 18 months while receiving placebo, which was attenuated over 5 years of patisiran treatment, consistent with the impact of patisiran on serum TTR (eFigure 3 in Supplement 2). Treatment benefits were observed across the spectrum of disease severity, including patients with severe ambulatory deficits. Long-term mortality in the overall population was lower than expected from natural history and much lower among patients receiving early treatment in the parent studies, underscoring the adverse impact of delaying treatment.

Impact on Disability, Polyneuropathy, and QOL

hATTR-PN is characterized by rapid deterioration in health, resulting in loss of mobility, autonomic and peripheral nerve function, and QOL.^5,6,7,8,9,10^ As established by previous natural history of the disease, patients receiving placebo in APOLLO worsened by 25 NIS points—roughly 1 PND stage—over 18 months.^1,14,18^ Hence, untreated patients with PND stage IIIA or IIIB who require 1 or 2 walking aids will likely become wheelchair bound (PND stage IV) in less than 2 years. Over the 5-year global OLE, most patients showed stable or improved ambulation, and nearly all patients retained their ability to walk, findings that were supported by the analysis of PND scores.

Multiple quantitative functional assessments corroborate the sustained benefit of patisiran measures of ambulation. During long-term treatment in the global OLE, patients showed only modest quantitative changes in motor and sensory function as measured by NIS. Furthermore, improvements in mBMI reflect relative reversal of autonomic neuropathy, which causes constipation, diarrhea, and vomiting leading to malnutrition and life-threatening wasting.^19^ Improvements in mBMI may also result from reduced peripheral neuropathy, which is associated with muscle wasting in other diseases.^20^

Patients with hATTR-PN often experience rapid deterioration of QOL. An 8.8-point worsening in Norfolk QOL-DN score is meaningful and perceptible by patients.^21^ During APOLLO, patients receiving placebo worsened by twice that score over 18 months. In contrast, the modest quantitative changes in Norfolk QOL-DN scores of patients treated with patisiran during APOLLO and of all treatment groups during the global OLE highlight the impact of treatment on QOL.

Impact on Mortality

Mortality rates for the phase 2 OLE and APOLLO patisiran patients were less than half those reported in natural history studies, which range from 6.8 to 29 deaths per 100 patient-years.^11,18,22,23,24^ The APOLLO placebo mortality rate was at the lower end of this range. In the APOLLO placebo group, most deaths occurred during APOLLO and the first year of the global OLE.

The parent studies enrolled patients who had polyneuropathy as the predominant disease manifestation, but cardiovascular causes of death were common (eTable 2 in Supplement 2). Nearly all patients who initially present with polyneuropathy will develop a mixed phenotype with cardiomyopathy within 8 years.^25^ In APOLLO, 56% of patients showed evidence of cardiac amyloid involvement at baseline, predefined as a left ventricular wall thickness of 13 mm or greater, in the absence of aortic valve disease or hypertension.^15^ Only polyneuropathy impairment (familial amyloid polyneuropathy stage) and parent study treatment assignment (patisiran vs placebo) were associated with mortality risk in multivariate analyses, despite a univariate analysis suggesting the role of advanced cardiomyopathy-related parameters (high NT-proBNP and NYHA class). The results extend prior observations from APOLLO that patisiran appeared to halt or reverse the progression of cardiac manifestations of hATTR,^15^ suggesting that patisiran has a broad impact on multisystemic disease progression.

Safety

The safety profile of patisiran during this 5-year study remained consistent with prior analyses, with no new safety signals.^14,16,17^ Although TTR is known to have a role in transporting vitamin A and thyroid hormone, there was no evidence that decreased circulating TTR via RNAi resulted in any AEs that could be attributed to vitamin A or thyroid hormone deficiency. Vitamin A supplementation is recommended for patients who receive patisiran and other TTR gene silencing therapies. To our knowledge, there are no off-target effects of patisiran, consistent with its specificity for TTR mRNA. This long-term follow-up supports the safety of the therapeutic reduction of TTR levels.

Limitations

The present results should be considered in the context of the global OLE study design and duration. The results are descriptive due to the open-label treatment, lack of randomization, and lack of a placebo comparator that prohibit statistical inferences. As the parent and global OLE studies ran over 7 years, the results are susceptible to survivorship bias. Selection bias may also have influenced the study population, as participants were self-selected from the parent patisiran studies, although the retention rate in the transition from parent study to global OLE was high. Causes of death during the global OLE were not independently adjudicated, precluding a precise estimate of the percentage of deaths attributable to cardiovascular or other causes. Additionally, only 65% of patients who entered the global OLE completed the study. This is mitigated by the fact that, of the 170 patients who did not die during the study, 81% completed the study through the full follow-up.

Conclusions

In a disease characterized by unrelenting decline, the results depict the long-term clinical stability afforded by TTR reduction and the long-term safety of patisiran treatment. The extended follow-up provides compelling evidence for the capacity of patisiran treatment to improve survival. However, patients tend not to recover function that is lost before starting treatment, which, along with a potential survival benefit, bolsters the importance of diagnosis and early treatment initiation. Safe and effective rapid TTR reduction allows patients with hATTR to preserve functional capacity and maintain a better QOL.