On Tuesday, 13th January 2015, the Patent Office rejected Gilead’s attempt to patent Hepatitis C drug, sofosbuvir (brand name: Sovaldi). The application had earlier been opposed through 2 pre-grant oppositions – by generic pharmaceutical company Natco, and as well as by non-profit group I-MAK. While this is certainly good news for the 18 million Hepatitis C patients in India, it does throw up some very interesting questions vis-a-vis the Indian pharmaceutical landscape. I’ll first look at what this means for the price of sofosbuvir in India, and then take a look at some curiosities surrounding the order.
1. Pricing strategies
In the US, Gilead’s Sovaldi, which was being hailed a wonder drug, with a 90% cure rate and a short therapy duration, was priced at about $1000 per day, for 84 days. Following the controversies that popped up around the world over its extremely high price, Gilead soon announced a lowered price of $300 / bottle ($900 or about Rs 54,000 for 84 days) for India as well as other lower income countries, as well as entered in a license agreement with 7 Indian generic drugs to sell the drug in 91 countries, with a 7% royalty rate. However, now that Gilead does not have a patent on the drug, it would appear that whatever price these 7 generics sell it at, it’ll be 7% higher than it could’ve been. It’s worth noting that Natco, who opposed the patent, was not one of these 7 licensees – so there is a chance that Natco, and others like it (that have not entered into license agreements with Gilead), can still manufacture sofosbuvir at at least slightly, if not significantly cheaper rates. From thereon, one can hope that competition within the generics will bring the price down to something more affordable. Would it be too optimistic to hope for the 100$/12 week course that Andrew Hill at Liverpool University predicted possible?
Nonetheless, the fact that 7 large Indian generic companies (Cipla, Cadila Healthcare, Hetero, Strides Arcolab, Ranbaxy, Sequent Scientific and Mylan) were willing to enter into license agreements when there certainly were strong questions of the patent application being rejected is something worth pondering over.
2. The order
The order dated 13th January, 2015, by Hardev Karar, (Dy Controller General, Patent Office, Delhi) rejected Gilead’s patent application for sofosbuvir on the basis of it failing to clear Section 3(d). While the order deals with some of the argument very well, it does raise rather than answer some questions towards the end.
The order first starts with the controller laying out the arguments from the pre-grant oppositions as well as the arguments put forth by the applicants. Karar agrees that the compounds are novel and inventive and moves on to the applicability of Section 3(d). The applicants argue that Section 3(d) only applies to new forms of ‘known substances’, and since the compounds were novel and inventive, they could not be considered ‘known substances’.
The order quickly turns this down, stating: “This kind of variation of orientation of the groups can make the compound novel and Inventive however, in the eyes of section 3(d) this novel and inventive substance is “considered to be the same substance, unless they differ significantly in properties with regard to efficacy”.” … “In other words we can say that a molecule with minor changes in addition to the novelty must show significantly enhanced therapeutic efficacy as compared to the nearest prior art molecule which is structurally and functionally close.” [Emphasis my own]. This seems a little off, as 3(d)’s ‘known substance’ should not (and does not) include any prior art simply because it is the ‘nearest’ prior art molecule that is structurally and functionally close.
In any case, as I understand it (and I’m open to correction), in this particular order, this deviation/error does not matter, as the order goes on to say that the compound in question is a stereoisomer of a known substance, and thus would, in fact, qualify to be considered a ‘known substance’ as per the explanation of Section 3(d).
Therapeutic Efficacy
However, the bigger questions come up after this when the order goes on to deal with therapeutic efficacy.
“The applicants showed the cytotoxcity data to prove the difference in properties which is insufficient to proove significant increase in the therapeutic efficacy. The data does not show any clinical trials to prove the improvement in the terapeutic efficacy.” (sic)
Firstly, having established that the compound needs to show enhanced therapeutic efficacy over previously known compounds, the order mentions that clinical trial data may be used to show this! Asking for clinical trial data at the patent stage is asking for an excessively high level of proof, and certainly is unreasonable to ask. By asking for such data, the order opens itself up for criticism.
And secondly, was the question around the issue of toxicity vis-a-vis therapeutic efficacy. The applicants had submitted data showing that seemingly minor changes in substituents at the 2′ position of the nucleoside result in large changes (unexpectedly better) in HCV activity and cytotoxicity. [Relevant part from submission pasted in table alongside this text]. However, the Controller says this data ‘cannot be considered sufficient and appropriate to show the enhancement of the therapeutic efficacy.’
Reading the order though, I’m still left confused as to what exactly was meant. Whether:
a) Cytotoxicity cannot be considered as a factor in accounting for therapeutic efficacy? Or
b) That Gilead showed significant improved and unexpected results over comparator compounds but did not make the link of showing how these result in enhanced therapeutic efficacy.
As the Controller said that the data in Table 1 [the table posted above] is not ‘appropriate’ to show the enhancement of therapeutic efficacy, I would lean towards the first explanation – that toxicity cannot be considered a factor in Section 3(d). If so, I would consider this problematic as significantly lower toxicity can play a large role in whether a patient can even take a drug! That being said, if I recall correctly, pharmacology considers slightly different meanings for efficacy and toxicity. Though the Supreme Court did not delve into these issues in the Novartis case, Shamnad had brought up a similar point in his amicus brief. In any case, I would’ve liked to see a little more discussion on this point rather than vague-ish seeming text which has seems to have gone straight ahead and interpreted it.
In any case, possible infirmities with the order aside, the result is good news for patients!
Edit: Order is available here.
Excuse my ignorance but I have been led to believe this is a breakthrough drug of potentially immense benefit to patients compared to existing treatments for HepC. I am also fully aware of the controversies about its pricing. But that is nothing to do with its patentability. I am also fully in support of 3d) in the Indian patent law. So I fail to understand how it was rejected (even if I read the order I would not feel competent to offer a reasoned analysis about it). So can someone explain in ways that ordinary people can understand? Or if there is no coherent explanation does it not raise questions about how the Patent Office interprets the legislation?
Thanks Charles,
I’m still mulling over parts of the order myself. We’ll try to come out with a more coherent analysis of how exactly 3(d) was applied here in the next few days.
Swaraj,
Just a few quick comments:
1. 3(d)’s ‘known substance’ should not (and does not) include any prior art simply because it is the ‘nearest’ prior art molecule that is structurally and functionally close.
It is important to note that s3d does state that “other derivatives” of known substance shall be considered to be the same substance unless they differ significantly in properties with regard to efficacy.
Given that the compounds in Gilead Pharmasset’s app derive from the earlier prior art cited, technically the prior art can be used as a basis for determining a known form or what is a derivative of a known form.
2. The data provided in the table as submitted with the application only relates to the cytidine analogs, not the uridine analog which forms the base of sofosbuvir in its prodrug form (subject of the application 3658/KOLNP/2009). Pharmasset itself has said in earlier publications that (see Clark 2005) that the uridine base demonstrated no activity (which could translated to efficacy) or cytoxicity. This is the case also against the known forms which Gilead Pharmasset compound and the uridine analog derives from. It’s for this very reason Gilead Pharmasset cannot show how the uridine analog is more efficacious than the known forms.
Thanks for adding to this discussion.
1. I agree with this comment. Hence, I went on (after that quoted line) to state that the loose wording doesn’t matter in this particular case, since the prior art here “does” fall within the s3d explanation of known substance (i.e, salts… other derivatives).
The issue I was highlighting with that quoted line was simply that the Controllers language was not very clear and could lead to a wrong interpretation wherein any ‘near’ prior art, and not just ‘salts… other derivatives’ could be considered.
2. Thanks for this bringing this up. I wasn’t aware of it and will look up more on this.