Interpreting the Stereochemistry of Phosphorus and Proving Commercial Success: How Gilead's Nucleoside Phosphoramidate Prodrug Patent "Captured" Sofosbuvir and Survived Invalidation

In pharmaceutical patent disputes, claim interpretation can be decisive because it defines the claimed subject matter against which novelty and inventive step are assessed. A single stereochemical notation in a structural formula may determine whether a patented claim covers a commercially successful drug worth billions of dollars.

The dispute in the Gilead Sciences Inc. (hereinafter “Gilead Sciences”) patent (“Nucleoside Phosphoramidate Prodrug patent”) centers on whether the solid line at the chiral phosphorus atom denotes a specific stereochemical configuration or left the stereochemistry undefined (Figure 1). With the marketed product being the single Sₚ diastereomer, the interpretation of that solid line determines whether the product is covered by the granted claim..

Figure 1. Claim 1 of the patent shows a compound with several chiral centers. Most of the bonds to those chiral centers are drawn as hashed/wedged bonds. However, the bond to the chiral phosphorus of the phosphoramidate moiety is drawn as a solid line.

The interpretation of the stereochemistry at the phosphorus had important consequences, made more complex due to the prosecution history: if Claim 1 was interpreted to encompass Sofosbuvir, did Gilead Sciences nevertheless surrender protection for Sofosbuvir by deleting the phrase “or its stereoisomers” during examination? These issues were considered by the China National Intellectual Property Administration (“CNIPA”) in a recent invalidation decision involving Gilead Sciences’ Nucleoside Phosphoramidate Prodrugs patent, a case selected as one of CNIPA's Top Ten Reexamination and Invalidation Cases.¹ The decision provides important guidance on both the interpretation of stereochemical notation and the circumstances in which claim amendments may amount to a surrender of subject matter.

Background

Sofosbuvir is an NS5B polymerase inhibitor approved by the U.S. FDA in 2013 and is widely regarded as a breakthrough therapy that transformed the treatment of hepatitis C, enabling cure rates exceeding 90% in many patient populations.

Gilead Sciences is the owner of Sofosbuvir as well as the patent at issue (“Nucleoside Phosphoramidate Prodrugs” Chinese patent no. ZL200880018024.2, granted on 2 May 2023). Beijing Kawin Technology Co., Ltd (hereinafter “Kawin Technology”) filed an invalidation request following patent disputes relating to Sofosbuvir.

Kawin Technology challenged the patent on both novelty and inventive step grounds. The case eventually centered on these issues:

1. How to interpret the stereochemical configuration at the phosphorus centre in Claim 1?
   
2. Did the prosecution history result in Gilead Sciences’ surrender for S-configuration Sofosbuvir?
   
3. Can post-filing publications and third-party interpretations be relied upon in invalidation proceedings?
   
4. Can the commercial success of Sofosbuvir support the inventiveness of Claim 1?
   

The granted patent contained five claims² directed to a nucleoside phosphoramidate compound, pharmaceutical compositions and uses thereof for treating HCV infection, and a process for preparing the compound. Kawin Technology alleged that claim 1 lacked novelty and claims 1-5 lacked inventive step, seeking invalidation of the patent in its entirety.

The Core Dispute: What Does the Solid Line at the Phosphorus Atom Mean?

Claim 1 was directed to a specific nucleoside phosphoramidate compound represented by the structural formula shown in Figure 1. The dispute centered on one feature: a solid line (“—”) that bonds the phosphorus atom to the rest of the molecule.

Kawin Technology argued that:

• Claim 1 should be interpreted as a specific compound with a defined three-dimensional structure, namely, the compound of Example 25, which was allegedly a racemic mixture at the phosphorus center. Therefore, claim 1 does not include Sofosbuvir, which has an S-configuration at the phosphorus atom.
  

Gilead Sciences disagreed, arguing that:

• Claim 1 does not define the stereochemical configuration at the phosphorus atom. It should therefore be interpreted as encompassing the compound in different stereoisomeric forms, as well as mixtures thereof, and not as being limited to a racemate. Therefore, claim 1 encompasses Sofosbuvir.
  

Why CNIPA Viewed the Stereochemistry at Phosphorus as Unspecified

CNIPA first considered how a skilled person in the field would interpret the stereochemical notation used in the claim. First, because the specification does not define the meaning of the solid line ("—"), its interpretation must be based on the ordinary conventions in the field of chemistry. Although Claim 1 contains multiple chiral centers, only the phosphorus center is depicted with a solid line bond rather than a wedged or hashed bond. Since a hashed/wedged bond denotes a defined stereochemical configuration whereas a solid line ordinarily leaves stereochemistry unspecified, the CNIPA concluded that Claim 1 encompasses all possible stereochemical configurations at the phosphorus center rather than being a racemate. CNIPA further noted that racemates are typically identified by “(±)” or “rac”, not by a solid line alone. This interpretation was consistent with the specification, which repeatedly acknowledges the chirality of the phosphorus atom and its existence of multiple stereochemical configurations.³⁴ This indicates that at the time of drafting, Gilead Sciences was aware that the compounds at issue could exist in different stereochemical forms.

Examples in the Specification Do Not Define the Claim Scope

Kawin argued that Claim 1 should be limited to the alleged racemic mixture of Example 25 because both share the same structural formula.⁵ CNIPA rejected this argument, holding that Example 25 disclosed only a mixture of R- and S-configurations at phosphorus, not necessarily a racemate. Most importantly, claims represent generalized technical solutions, whereas examples illustrate specific embodiments within the claim scope. Accordingly, in the absence of a clear indication in the specification, the technical solution generalized by Claim 1 cannot be equated with the product obtained through specific reaction steps in the examples (e.g. Example 25). Therefore, claim 1 was not limited to a racemic mixture of the phosphorus stereoisomers.

Prosecution History Showed No Surrender for S-configuration Sofosbuvir

Kawin relied on the doctrine of estoppel, arguing that amendments made by Gilead Sciences during substantive examination had surrendered protection for S-configured Sofosbuvir and therefore Gilead Sciences could not later reintroduce the subject matter into the scope of protection. The relevant prosecution history is summarized below:

In this matter, CNIPA found that there was no indication that Gilead Sciences had intended to abandon protection for the S-configured compound. Although Gilead Sciences amended the compound name in Claim 1 to its structural formula and deleted the phrase “or its stereoisomers,” it simultaneously retained and added claims specifically directed to the phosphorus stereoisomers. Even if those claims were later deleted, Gilead Sciences stated that the stereoisomers were already encompassed by the general formula of Claim 1. In addition, statements made during prosecution referring to Example 25 were viewed as arguments supporting inventive step based on a representative embodiment, rather than disclaimers excluding other isomers or mixtures falling within the scope of Claim 1. As a result, the prosecution history did not support a finding that protection for S-configured Sofosbuvir had been surrendered.

The decision suggests that CNIPA will not infer surrender merely from the deletion of claim language. Instead, what matters is whether the prosecution history, viewed as a whole, demonstrates a clear and consistent intention to relinquish particular subject matter. Here, every significant prosecution step indicated that Gilead Sciences regarded the phosphorus stereoisomers as remaining within its invention, making it difficult to conclude that protection for Sofosbuvir had been intentionally abandoned.

The Impact of Post-filing Publications and Third-party Interpretations in Invalidation Proceedings

Kawin alleged that Gilead Sciences' post-filing publications and a STN search of the CAS database records identified the exact structural formula of Claim 1 as the specific compound PSI-7851 (i.e. a 1:1 mixture of the S_p and R_p stereoisomers of the nucleoside phosphoramidate compound shown in Claim 1). Therefore, Kawin argued that the structural formula in Claim 1 represents the nucleoside phosphoramidate compound as a racemic mixture of the R_p and S_p stereoisomers at the phosphorus center. Additionally, Kawin submitted multiple opinions from different institutions to support its interpretation of the claims.

CNIPA nevertheless rejected these arguments. CNIPA emphasized that, in invalidation proceedings, claim interpretation is ultimately a legal determination to be made by the panel from the perspective of a person skilled in the art, relying on internal evidence such as the claims, specification, and prosecution history, as well as external evidence such as textbooks and reference work. Accordingly, in principle, statements made by the inventors in their post-filing publications and characterizations appearing in databases could not be relied upon to determine the scope of the claims, while opinions issued by other institutions merely reflected their own views and were not binding on the invalidation proceedings.

As a result, CNIPA concluded that a skilled person would interpret Claim 1 as covering a compound or a mixture thereof that encompass all stereochemical configurations reasonably possible at the phosphorus atom.

About Commercial Success in Inventive Step Assessment

CNIPA first concluded that Claim 1 was inventive because the prior art provided no clear teaching that the specific phosphoramidate ProTide group could be successfully applied to the prior arts’ nucleoside. Another key issue of this case was whether commercial success could be recognized as a secondary consideration in the assessment of inventive step. According to the Guidelines for Patent Examination, commercial success may support a finding of inventiveness⁶ where such success is directly attributable to the technical features of the claimed invention. Thus, the key question was whether the Gilead Sciences’ product’s commercial success was directly attributable to the claimed technical features.

In this case, Gilead Sciences submitted extensive evidence showing that the compound of Claim 1 achieved remarkable commercial success as a treatment for hepatitis C. Kawin argued that the commercialized product was the S_p stereoisomer (i.e., Sofosbuvir), rather than the compound of Example 25 (1:1 mixture of the S_p and R_p stereoisomers of the nucleoside phosphoramidate compound shown in Claim 1). Kawin further argued that the phosphorus R_p stereoisomer isomer had never been marketed and therefore the commercial success of Sofosbuvir could not be extrapolated to all embodiments encompassed by Claim 1.

For pharmaceutical products, commercial success may be influenced by factors such as marketing and sales, but the product's therapeutic benefits remain the primary reason why physicians and patients choose it. In this case, CNIPA accepted that Sofosbuvir, which falls within the scope of Claim 1, achieved exceptional clinical and commercial success due to its superior efficacy in treating hepatitis C. CNIPA further held that, given the unique nature of pharmaceutical development and commercialization, a patentee is not required to demonstrate commercial success for every embodiment encompassed by a claim. Accordingly, the commercial success of Sofosbuvir alone was sufficient to support the inventiveness of Claim 1.

In view of the foregoing, CNIPA maintained the patent in full.

In practice, patentees often invoke commercial success, but relatively few cases succeed because proving that the product's commercial success is directly attributable to the claimed technical features is not straightforward. This case demonstrates that commercial success can be a meaningful secondary consideration when the causation is clear. Nevertheless, the patent had already survived the novelty and inventive-step analysis on conventional grounds. The commercial-success evidence therefore reinforced, rather than independently cause, the finding of inventiveness.

Eagle IP Thoughts

This decision provides several important lessons for pharmaceutical patentees.

Commercial Success in Supporting Inventive Step

This case provides an example of commercial success being considered in support of inventiveness in China, where its success is directly attributable to the technical features of the claimed invention rather than external factors such as marketing or sales efforts. However, in practice, commercial success remains difficult to establish as a support of inventiveness. Given that inventive step had already been established on conventional grounds, the commercial success of Sofosbuvir served to reinforce that finding and should not understood as commercial success alone has become easier to rely on in support of inventive step.

Preserving Claim Flexibility

This decision provides some guidance as to when estoppel applies in China. Although Gilead deleted the phrase "or its stereoisomers" and later removed claims directed to individual phosphorus stereoisomers, the prosecution record consistently reflected its position that those stereoisomers remained encompassed by Claim 1. The decision suggests that estoppel depends on whether the applicant relinquished a particular subject matter during prosecution to secure patent grant. Here, Gilead’s prosecution statements made clear that the stereoisomers remained within the scope of Claim 1, despite the amendments made during examination.

Muted Impact of Later Publications/Disclosures to a Patent

The decision has also addressed the relevance of post-filing statements in claim interpretation. Despite relying on post-filing scientific publications, database records, and opinions from multiple institutions, Kawin was unable to persuade CNIPA to adopt its interpretation of Claim 1. Instead, CNIPA held that claim construction must be based primarily on the claims, specification, prosecution history, and the understanding of a person skilled in the art at the time of filing, rather than on how the invention was subsequently described by inventors or third parties.

Contrast this with the US, where inventor or applicant statements made after the patent filing date have been used as damaging evidence in patent litigation⁷. In China, it appears that post-filing statements characterizing an invention in a particular way will not necessarily be treated as evidence that the patentee intended to relinquish/add certain subject matter. Rather, the patent itself remains the primary reference point for determining claim scope.

Drafting Strategies

In chemical drafting, patent applicants should clearly explain the stereochemical notation of structural formulas in the specification, particularly when chiral centers are involved. If a claim is intended to cover multiple stereochemical configurations, the specification should explicitly and unambiguously disclose all the relevant configurations. A general statement that all configurations are contemplated may not be sufficient where a particular configuration is neither disclosed in nor directly and unambiguously derivable from the specification even when combined with knowledge available in the art.

Finally, claim interpretation should be grounded in the specification, read in light of the knowledge available to a person skilled in the art at the filing date. As seen in this case, post-filing publications, database entries, expert opinions, and foreign decisions carried far less weight in determining claim scope. Applicants should therefore ensure that the specification, understood in that context, supports the intended claim scope from the outset.

If you would like to have more information on this matter or would like to have our advice, please feel free to contact us at [email protected].

This article is for general informational purposes only and should not be considered legal advice or a legal opinion on a specific set of facts.

1. 国家知识产权局 复审无效十大案件 案例三：“核苷氨基磷酸酯前药”发明专利权无效宣告请求案
   ↩︎
2. Claim 1, a compound represented by a structural formula; Claim 2, a composition comprising the compound of Claim 1 and a pharmaceutically acceptable medium; Claim 3, a composition for treating HCV infection; Claim 4, use of the compound in the manufacture of a medicament for treating an HCV-infected subject; and Claim 5, a process for preparing the compound by reacting compound 4" with nucleoside analogue 5', where X' is a leaving group.
   ↩︎
3. Paragraph [0818] of the specification (CN101918425B) states that “ Furthermore, the inventors recognised that the phosphorus atom in the phosphoramidate moiety represents an additional source of chirality. Although the following structures do not specifically depict the chirality of phosphorus, the inventors recognised that stereochemical configurations are possible. Thus, in the staggered (or saw-horse) linear representation, the oxygen substituent may project toward the viewer while the OR¹ substituent projects away from the viewer, or vice versa. Accordingly, the Cahn-Ingold-Prelog stereochemical designation at phosphorus may be either R or S. The following structures therefore encompass all stereochemical configurations reasonably possible at the phosphorus atom.”
   ↩︎
4. Paragraph [0072] of the specification (CN101918425B) states that “The term “P” indicates that the phosphorus atom is chiral and has the corresponding Cahn-Ingold-Prelog designation of “R” or “S”, each having its ordinary and well-understood meaning. Compounds of Formula I are expected to be racemic due to the chirality of the phosphorus atom. The applicant contemplates the use of racemates and/or resolved enantiomers. In some instances, an asterisk is not shown adjacent to the phosphorus atom of the phosphoramidate group. In such cases, it should nevertheless be understood that the phosphorus atom is chiral, and a person skilled in the art would likewise understand it to be chiral, unless the substituents bonded to phosphorus preclude the possibility of chirality, such as in P(O)Cl₃.”
   ↩︎
5. The compound of Example 25 is prepared by specific reaction steps and the specification records only its NMR and MS data; on that information it can be determined only that the compound is a mixture of R- and S-configurations at phosphorus, and the conclusion that it is a "racemate" is difficult to reach. That conclusion, moreover, is not derived from the structural formula but from a composite assessment drawing on other information such as the isomer resolution described in Example 81. Example 81 records that, because of the chirality at phosphorus, certain exemplified compounds were obtained as mixtures of diastereomers, separated by SFC on a Chiralpak-AS-H column, and that the absolute stereochemistry of the P-chiral centre of the diastereomers was not determined.
   ↩︎

6. Section 5.4 of Chapter 4, Part II of the Guidelines for Patent Examination, “When a product of an invention achieves commercial success, if such success is directly attributable to the technical features of the invention, this not only reflects that the invention has beneficial effects but also indicates that the invention is not obvious; therefore, such an invention possesses outstanding substantive characteristics and significant progress, and thus meets the requirements for inventiveness.”

   ↩︎
7. For example, in the patent interference between the University of California and the Broad Institute, the inventor Jennifer Doudna’s statements made post-filing (“[w]e weren't sure if CRISPR/Cas9 would work in eukaryotes”) was used as evidence to show that a later filed application Broad patent claiming use of CRISPR in eukaryotic cells was inventive.
   ↩︎