Picking up on the informative postings by SpicyIP and Generic Pharmaceuticals and IP (Generic IP), there appears to be some confusion as to what protection Roche obtained in the U.S for the drug Valganciclovir.
As Generic IP has rightly pointed out, in addition to US Patent No. 6083953 (‘953) reported in the Livemint article, Roche was also granted an additional US Patent No.5840891 (‘891). Both patents derive from US continuation application nos. 08/453,223 and 08/281,893 (the latter forming the priority claim for the Indian application 959/MAS/1995).
It is clear that the main claim in ‘953 covers the crystalline form of Ganciclovir mono (L-valinate) hydrochloride. According to our analysis, with the help of our expert pharmaceutical chemist Prof J. Fortunak, it appears that claim 1 of ‘891 does not cover the mono L-valinate ester of Ganciclovir. We believe that it claims broad coverage for the L-valinate ester of Ganciclovir analogs only. Claim 2 to 6 further define the main claim. In essence, it seems that ‘953 claims protection for intermediates to make L-Valganciclovir, but not the actual drug itself.
Indeed, this is made clearer when one reviews the file wrappers for the various continuation applications made by Roche to claim the L-Valinate ester and pharmaceutically acceptable salts for Ganciclovir – which is believed to be the main claim for 959/MAS/1995. On each occasion, the examiner rejected Roche’s attempts and eventually Roche persuaded the USPTO that the prior art of Beauchamp (US Patent No. 5043339) and another publication by Beauchamp in 1992 did not anticipate the crystalline form as eventually patented in ‘953.
It is worth noting that Roche made a more recent attempt through another continuation application (US2004/0102462) to patent the L-valinate ester hydrochloride by amending the claim to read ‘a compound that is non-crystalline form of L-Valinate hydrochloride’. Luckily the application went to the same examiner as all the other applications and the claims were rejected. The application has now been abandoned.
Surprisingly, the EPO granted the main claim for the L-valinate ester and salts (EP0694547) despite knowing about the prior art of Beauchamp. Does the Indian Patent Office follow the USPTO, the EPO, or go its own way?
Which brings us to the other burning issue in this case, the application, or lack of, of s8 by the Chennai Patent Office. SpicyIP makes the relevant point that unless the patent office requests the further information on corresponding applications in other countries, technically the applicant does not have to volunteer the information.
There is another angle to s8 that has not been picked up. Section 8 was amended by the 2005 Act. Prior to the Amendment, under s8(1)(b), the applicant was only required to provide an undertaking to keep the Controller informed of corresponding applications for the same or substantially the same invention as applied for in India ‘up to the date of acceptance of the complete specification in India’. The critical question is whether the amended wording of s8(1)(b) to keep the Controller informed ‘up to the date of grant’ and the addition of s8(2) (and R12(3)) can apply retrospectively. It would be unfair on applicants if it was, especially if they have complied with the requirements under s8(1)(b) as they stood prior to 2005. If so, the relevant question to ask if using s8 as a ground for opposition against applications filed prior to 2005 is, when was the complete specification in India accepted.
Taking the above into account, if the Indian patent application was to be held invalid for some of the claims, but not those covered by ‘891, Roche could still potentially block generic companies from making Valgancicovir if they adopt the same intermediates route. It will be interesting to see whether the Indian companies can work around this.