The Nonclinical Safety Profile of GalNAc-conjugated RNAi Therapeutics in Subacute Studies

Abstract

Short interfering RNAs (siRNAs) and antisense oligonucleotides (ASOs) are the most clinically advanced oligonucleotide-based platforms. A number of N-acetylgalactosamine (GalNAc)-conjugated siRNAs (GalNAc-siRNAs), also referred to as RNA interference (RNAi) therapeutics, are currently in various stages of development, though none is yet approved. While the safety of ASOs has been the subject of extensive review, the nonclinical safety profiles of GalNAc-siRNAs have not been reported. With the exception of sequence differences that confer target RNA specificity, GalNAc-siRNAs are largely chemically uniform, containing limited number of phosphorothioate linkages, and 2'-O-methyl and 2'-deoxy-2'-fluoro ribose modifications. Here, we present the outcomes of short-term (3-5 week) rat and monkey weekly repeat-dose toxicology studies of six Enhanced Stabilization Chemistry GalNAc-siRNAs currently in clinical development. In nonclinical studies at supratherapeutic doses, these molecules share similar safety signals, with histologic findings in the organ of pharmacodynamic effect (liver), the organ of elimination (kidney), and the reticuloendothelial system (lymph nodes). The majority of these changes are nonadverse, partially to completely reversible, correlate well with pharmacokinetic parameters and tissue distribution, and often reflect drug accumulation. Furthermore, all GalNAc-siRNAs tested to date have been negative in genotoxicity and safety pharmacology studies.

Keywords: cell(ular) pathology; drug development; liver; monkey pathology; preclinical research and development; preclinical safety assessment/risk management; rat pathology.

Figure 1.

GalNAc-siRNA targeted delivery mechanism. (A) Following GalNAc-mediated delivery to hepatocytes via ASGPR and endolysosomal escape, siRNAs are loaded into a multi-subunit RNAi-induced silencing complex (RISC). After siRNA duplex unwinding, the antisense strand (red) remains bound to RISC and directs site-specific cleavage of the complementary target RNA sequence (green), resulting in RNA degradation and reduced expression of the target protein. (B) siRNA chemical modifications enhance stability and conjugation with a trivalent GalNAc ligand at the 3’ end of the sense strand allows for targeting to the ASGPR on hepatocytes. Alnylam’s Enhanced Stabilization Chemistry (ESC) GalNAc-siRNAs are fully modified at the 2’ position of the ribose with 2′-O-methyl (2′-OMe) and 2′-deoxy-2′-fluoro (2′-F) and contain limited number of strategically placed phosphorothioate (PS) linkages (Nair et al. 2014; Foster et al. 2018). ASGPR = asialoglycoprotein receptor; GalNAc = N-acetylgalactosamine; siRNA = short interfering RNA.

Figure 2.

Hepatocellular vacuolation in rats administered GalNAc-siRNAs once weekly in 3-week repeat-dose toxicity studies at doses up to 300 mg/kg. (A, B) Hematoxylin and eosin stained liver sections in 0.9% sodium chloride–treated control rats (A) and GalNAc-siRNA-treated rats (B). (C, D) Oil Red O staining of the liver shows neutral lipids in control rats (C) and increased amounts in GalNAc-siRNA-treated rats (D). (E, F) Transmission electron micrographs of the liver of control rats (E) and GalNAc-siRNA-treated rats (F). Arrows indicate vacuoles containing lipid droplets. GalNAc = N-acetylgalactosamine; siRNA = short interfering RNA.

Figure 3.

Receiver operating characteristic (ROC) curve for hepatocellular vacuolation. Diagnostic performance of clinical chemistry parameters was evaluated for hepatocellular vacuolation by ROC analysis (performed using GraphPad Prism version 6.07 for Windows, GraphPad Software, La Jolla, CA, www.graphpad.com). The method of Hanley (Hanley and McNeil 1982) was used to calculate the area under the curve (AUC) values with a 95% confidence interval. A diagnostic test with an AUC of 0.8–1.0 is considered to have good predictive ability, whereas a test with an AUC of 0.5–0.7 indicates a random or nondiagnostic test result with poor predictive ability (Ennulat et al. 2010). Clinical and anatomical pathology data from 933 animals over 28 studies were compiled. Chemistry parameters were normalized to the respective control mean and expressed as the fold change of each individual compared to the control mean. ROC analysis was performed on five liver parameters: Alanine Aminotransferase (ALT), Aspartate Aminotransferase (AST), Alkaline Phosphatase (ALP), Total Bilirubin (TBIL), and Glutamate Dehydrogenase (GLDH). The observation of hepatocellular vacuolation was separated into two bins, no observation and grades 1 to 4. AUC for all five parameters were below 0.7, indicating they were not predictive as biomarkers for hepatocellular vacuolation.

Figure 4.

Increased hepatocellular SCN and mitosis in rats administered GalNAc-siRNAs once weekly in 3-week repeat-dose toxicity studies at doses up to 300 mg/kg. (A) Instances of hepatocellular SCN are indicated by arrows. (B) Mitotic events (arrows) are often observed with SCN, consistent with regenerative activity. Hematoxylin and eosin stained liver sections are shown. GalNAc = N-acetylgalactosamine; SCN = single-cell necrosis; siRNA = short interfering RNA.

Figure 5.

Drug accumulation in rats administered GalNAc-siRNAs once weekly in 3-week repeat-dose toxicity studies at doses up to 300 mg/kg. (A, B) Proximal renal tubular cells of rats administered GalNAc-siRNAs. H&E staining (A) shows cytoplasmic basophilic granules and IHC (B) using an antibody recognizing 2’F-containing oligonucleotides confirms these granules to be GalNAc-siRNA drug. (C, D) The subcutaneous injection site of rats administered GalNAc-siRNAs. H&E staining (C) shows vacuolated mononuclear cells (arrows) in the superficial dermis, and IHC (D) confirms the presence of GalNAc-siRNA drug in these cells. GalNAc = N-acetylgalactosamine; H&E = hematoxylin and eosin; IHC = immunohistochemistry; siRNA = short interfering RNA.

Figure 6.

Drug accumulation in monkeys administered GalNAc-siRNAs once weekly in 5-week repeat-dose toxicity studies at doses up to 300 mg/kg. (A, B) Hepatic Kupffer cells of monkeys administered GalNAc-siRNAs. H&E staining (A) reveals basophilic granules (arrows), and IHC (B) using an antibody recognizing 2’F-containing oligonucleotides confirms these granules to be GalNAc-siRNA drug (arrows). Positive IHC labeling is also observable in the cytoplasm of hepatocytes. (C, D) Macrophages in the medulla of systemic lymph nodes of monkeys administered GalNAc-siRNAs. H&E staining (C) demonstrates vacuolated macrophages in the lymph node medulla, and IHC (D) confirms the presence of GalNAc-siRNA drug in these cells. The mesenteric lymph node often contains the most prominent granules versus the draining lymph nodes from the subcutaneous injection site or other lymph nodes. GalNAc = N-acetylgalactosamine; H&E = hematoxylin and eosin; IHC = immunohistochemistry; siRNA = short interfering RNA.