Synthon BV v Smithkline Beecham Plc

Factual and Procedural Background

The case concerns a dispute over the validity of a UK patent (No. 2,336,364) held by Company B ("SB") relating to paroxetine methanesulfonate and its crystalline forms. Company A ("Synthon") challenges the validity of SB's patent on the basis of a prior patent application (No. WO98/56787) filed by Synthon but not published before SB's priority date. Under section 2(3) of the Patents Act 1997 (corresponding to Article 54(3) EPC), Synthon's earlier application is deemed to have been published for the purpose of assessing novelty against SB's patent.

The dispute involves complex technical issues about polymorphism, crystallisation, and analytical techniques such as infrared (IR) spectroscopy and nuclear magnetic resonance (NMR). Both parties rely on expert evidence concerning the preparation and characterization of paroxetine methanesulfonate crystals.

The court is tasked with determining whether SB's patent claims are novel in light of Synthon's earlier application, focusing particularly on whether Synthon's disclosure enables the skilled person to prepare the crystalline form claimed by SB.

Legal Issues Presented

1. Whether SB's UK patent No. 2,336,364 is novel having regard to Synthon's earlier patent application No. WO98/56787, deemed prior published under s.2(3) of the Patents Act 1997.
2. Whether Synthon's earlier application is enabling such that it anticipates SB's patent claims by disclosing the same invention.
3. Whether the differences in IR spectra reported by the parties indicate distinct crystalline forms or measurement errors.
4. The application of the legal principles governing anticipation, including the distinction between enabling disclosure and inevitable result anticipation.

Arguments of the Parties

Company A (Synthon) Arguments

• Synthon contends that its earlier patent application invalidates SB's patent by virtue of anticipation under s.2(3) of the Patents Act 1997.
• Synthon argues their specification enables the skilled person to make paroxetine methanesulfonate crystals, despite some experimental difficulties and deficiencies in the "Experimental" section.
• They claim that the differences in IR spectra reported by SB result from measurement inaccuracies rather than different polymorphs.
• Synthon submits that the general teachings of their specification, combined with common knowledge and ordinary experimentation, would enable successful crystallisation.

Company B (SB) Arguments

• SB maintains that its patent claims a specific crystalline form of paroxetine methanesulfonate characterized by distinctive IR peaks, which is novel over Synthon's disclosure.
• SB challenges the repeatability of Synthon's "Experimental" example and contends that Synthon's disclosure is not enabling for the claimed crystalline form.
• SB argues that the differences in IR spectra reflect genuinely different crystalline forms, supporting the novelty of their invention.
• SB disputes that the skilled person could inevitably produce the SB crystalline form by following Synthon's teaching, emphasizing the absence of detailed crystallisation instructions.

Table of Precedents Cited

Court's Reasoning and Analysis

The court first examined the technical background of polymorphism and crystallisation, acknowledging the complexity and variability in producing specific crystalline forms. It recognized that both parties' patents broadly disclosed paroxetine methanesulfonate and its crystallisation, with neither patent indicating particular difficulties or critical solvents for crystallisation. Both patents taught a wide range of preparation and crystallisation methods, and both acknowledged the possibility of polymorphs and solvates without focusing on them as central to the invention.

The court then scrutinized the experimental evidence concerning the repeatability of Synthon's example for producing crystals. It found that the original "Experimental" example in Synthon's specification was not straightforwardly repeatable by a skilled team, and that additional unmentioned steps and considerable skill were required to eventually produce crystals. However, the court accepted submissions that the specification as a whole, combined with common general knowledge and ordinary experimentation, would enable the skilled person to make crystalline paroxetine mesylate.

Regarding the differences in IR spectra, the court found on the balance of probabilities that the differences resulted from measurement errors rather than the existence of distinct polymorphs. Both parties’ later work produced the same crystalline form characterized by the IR peaks reported in SB's patent. No other polymorph had been reliably demonstrated.

In applying legal principles, the court emphasized the strict nature of the novelty test under section 2(3) of the Patents Act 1997 and Article 54(3) EPC. It highlighted the distinction between enabling disclosure and inevitable result anticipation, rejecting the latter as the basis for Synthon's case because the prior disclosure did not inevitably produce the SB crystalline form.

The court adopted a rational, informational approach to novelty, focusing on whether Synthon's disclosure enabled the skilled person to make the claimed invention, not on formal claim comparisons or minor differences in characterization. It concluded that Synthon's earlier application disclosed and enabled the invention claimed by SB, including the crystalline form, and that SB's patent did not add any relevant new information.

Holding and Implications

The court held that SB's patent claims lack novelty and are anticipated by Synthon's earlier patent application under section 2(3) of the Patents Act 1997.

This decision means that Synthon's earlier application, deemed prior published, disclosed and enabled the invention claimed by SB, including the specific crystalline form of paroxetine methanesulfonate. The court found no valid basis to distinguish the SB patent claims as novel over Synthon's disclosure.

The ruling directly affects the parties by invalidating SB's patent claims, but it does not establish any new legal precedent beyond the application of established principles of anticipation and enabling disclosure in the context of chemical patents involving polymorphs.