Oligonucleotide Drug Products — Antisense, siRNA, and Aptamer CMC Under FDA and EMA Review

Oligonucleotide therapeutics have achieved commercial validation — multiple phosphorothioate ASOs, several GalNAc-conjugated siRNAs, and multiple aptamers are FDA-approved, and the pipeline is expanding rapidly. The solid-phase synthesis chemistry for oligonucleotide drug substances is reproducible, the analytical toolkit is established, and FDA CDER has reviewed dozens of oligonucleotide NDA submissions. And yet the CMC deficiency pattern for oligonucleotide submissions does not reflect this maturity.

An oligonucleotide impurity specification reporting a single pooled “n-x total” figure has answered a question ICH Q3A never asked, while leaving each individual positional deletion sequence — with its own distinct sequence and potentially distinct biological activity — completely unidentified.

Sequence-Specific n-1 Deletion Impurity Specification — Positional Deletion Identification by IP-RP-HPLC/MS and ICH Q3A Individual Limits Beyond the Pooled n-x Approach

Automated solid-phase phosphoramidite synthesis produces positional deletion sequences whenever a coupling cycle fails for a fraction of the growing oligonucleotide chains, and for a 20-mer with nineteen coupling steps, even a coupling yield in the high 99% range compounds into a meaningful cumulative deletion burden before purification ever begins. The critical regulatory point is that each positional deletion is not interchangeable with any other: a chain missing its third nucleotide is a structurally and sequentially distinct molecule from a chain missing its seventeenth, and these different deletion products can carry meaningfully different biological activity — some retaining partial pharmacological function, others essentially inert — which is precisely why ICH Q3A(R2)’s individual identification and specification requirement applies to each positional deletion above the identification threshold rather than to a pooled category. Achieving that individual specification requires an ion-pair reversed-phase HPLC method, typically using a triethylammonium acetate mobile phase system, developed and optimized specifically to baseline-resolve adjacent positional deletions from the parent and from each other; where two positional deletions happen to co-elute despite gradient optimization, LC-MS deconvolution using the unique mass of each co-eluting species becomes the fallback for individual quantitation. A 3.2.S.4 specification stating “n-1 and shorter impurities not more than 2.0% total” has not applied ICH Q3A as the guideline requires — it has substituted a single convenient number for the individual identification and safety qualification of impurities that may differ meaningfully in their biological activity, which is exactly the deficiency FDA chemistry reviewers raise when they request positional-specific limits with MS-confirmed identity for every impurity above the reporting threshold.

Chemical Modification Pattern Verification — ESI-MS/MS CID Fragmentation, w-Ion and a-B Ion Series, and Why Intact Mass Alone Cannot Confirm the PS and 2′-Modification Pattern

An intact mass measurement by ESI-MS confirms that a synthesized oligonucleotide has the correct overall molecular formula, but it cannot distinguish between two molecules with identical formulas and different modification patterns — a phosphorothioate linkage misplaced at one position instead of another produces no change in molecular formula whatsoever, since both a phosphodiester and phosphorothioate version of the full sequence would only differ by the total count of sulfur-for-oxygen substitutions, not their positions. That is the specific gap collision-induced dissociation tandem mass spectrometry closes: fragmenting the intact oligonucleotide produces a family of sequence-specific fragment ions running from each end of the molecule, and because each phosphorothioate substitution produces a small, consistent mass shift in every fragment ion downstream of that position, a complete fragment ion series lets an analyst confirm not just that PS substitutions are present somewhere in the molecule but exactly where each one sits relative to the sequence. The same logic applies to confirming the position of 2′-MOE, 2′-F, or other ribose modifications that define a gapmer or fully modified oligonucleotide’s nuclease resistance and pharmacological behavior. A characterization package presenting only intact mass data, with a close match between measured and theoretical molecular weight, has demonstrated that the correct atoms are present somewhere in the molecule — not that they are in the correct positions — and FDA reviewers close this gap by requesting the complete MS/MS fragmentation dataset confirming the modification pattern residue by residue.

Counterion Specification, GalNAc Conjugate Characterization, and the Duplex T_m Attribute That Bridges Strand Purity to Pharmacological Activity

The degree of sodium salt formation in a synthetic oligonucleotide drug substance is not a cosmetic counterion detail — it directly affects hygroscopicity, and hygroscopicity in turn affects drug product dissolution and handling, which is why a specification stating simply “sodium salt” without a quantitative sodium content range measured by ion chromatography leaves an actual formulation-relevant attribute uncontrolled from lot to lot. For GalNAc-conjugated siRNA therapeutics specifically, the CMC picture expands beyond a single-strand specification into a duplex characterization problem: both the sense strand carrying the GalNAc targeting ligand and the antisense strand must be individually specified for purity, the conjugation efficiency attaching the GalNAc cluster to the sense strand needs its own quantitative limit, and the receptor-binding activity of that GalNAc cluster against the hepatocyte asialoglycoprotein receptor is a functional attribute that individual strand purity data cannot substitute for. The duplex itself carries one more identity attribute that bridges these individual-strand measurements to actual pharmacological function: melting temperature, measured by UV hyperchromicity, confirms that the two strands are properly annealed into a stable duplex, and a T_m falling outside the expected range for the given duplex length and overhang design signals either improper annealing or an inconsistency in the modification pattern that individual strand purity testing alone would never catch — a batch that passes every single-strand purity specification can still fail downstream potency testing if the duplex T_m is out of range, because RISC loading depends on proper duplex formation, not on either strand’s individual cleanliness.

The XGene Oligonucleotide Drug Substance CMC Architecture — Synthesis CPP Control, IP-RP-HPLC Purity, MS/MS Modification Confirmation, Counterion Specification, Stability Program, NDA vs. BLA Pathway

The XGene Oligonucleotide Drug Substance CMC Architecture is a structured NDA CMC development framework built around the fact that oligonucleotide impurities are sequence variants of the API itself, not structurally distinct byproducts a conventional impurity specification anticipates.

1. Synthesis Process Control — Monitor per-cycle coupling efficiency in-process and design the IP-RP-HPLC purity method specifically to baseline-resolve positional deletion sequences. 2. Individual Positional Impurity Specification — Identify and specify each positional n-1 deletion above the ICH Q3A threshold individually, with MS-confirmed sequence identity, rather than a pooled n-x limit. 3. MS/MS Modification Pattern Confirmation — Generate complete fragment ion series data confirming the position of every phosphorothioate and 2′-modification, not just intact mass. 4. Counterion and Conjugate Characterization — Specify sodium content quantitatively by ion chromatography, and for GalNAc conjugates, build the conjugation efficiency, receptor-binding, and duplex T_m specifications as a complete package. 5. Regulatory Pathway Determination — Confirm NDA classification under 21 CFR 314 for fully chemically synthesized oligonucleotides, including GalNAc conjugates, before the CMC strategy is finalized.

The output is the complete oligonucleotide drug substance CMC package that treats each sequence-related impurity and each modification position with the individual rigor ICH Q3A and its oligonucleotide-specific application require.

Spinraza (nusinersen, NDA 209531, approved December 2016), the first FDA-approved treatment for spinal muscular atrophy, established the foundational phosphorothioate ASO NDA CMC precedent for positional deletion impurity specification and MS-based modification pattern confirmation. Leqvio (inclisiran, NDA 214012, approved December 2021) confirmed that a GalNAc-conjugated siRNA produced by fully chemical synthesis is classified as an NDA drug substance rather than a biologic BLA, establishing the duplex characterization and conjugation efficiency specification precedent that subsequent GalNAc-siRNA programs reference. Vitravene (fomivirsen sodium, NDA 020961, approved August 1998), the first FDA-approved antisense oligonucleotide, established the original regulatory foundation for oligonucleotide drug substance specification and the IP-RP-HPLC purity methodology every subsequent ASO program has built upon.

For your oligonucleotide NDA CMC package, can you confirm today that your 3.2.S.4 specification includes individual limits for each positional n-1 deletion sequence identified above the ICH Q3A reporting threshold, and that your 3.2.S.3 characterization section includes ESI-MS/MS fragmentation data confirming the modification pattern at every position in the sequence?