[Ip-health] Some initial observations on the Satwant Reddy Report
chan park
chansoobak@yahoo.com
Wed Jun 6 13:42:14 2007
--
[ Picked text/plain from multipart/alternative ]
The "Report on Steps to be taken by Government of India in the context of =
Data Protection Provisions of Article 39.3 of the TRIPS Agreement" was sub=
mitted on 31 May 2007, on the last working day of the outgoing Secretary o=
f the Department of Chemicals and Petrochemicals - Satwant Reddy - who was=
one of its two primary authors (the other was Gurdial Singh Sandhu, Joint=
Secretary in the same department).
It is available publicly at http://chemicals.nic.in/DPBooklet.pdf.
At close to 60 pages, the report attempts to balance several conflicting=
positions. Perhaps that explains why - the day after submission - newspap=
ers reported such widely differing interpretations.
On June 1, the Economic Times said: "Multinational drug makers like Pfiz=
er, Glaxo and Novartis can look forward to delaying the entry of me-too ve=
rsions of their new drugs in India for five years.....It recommends a five=
-year data exclusivity period for pharmaceuticals and three-year exclusivi=
ty term for agrochemicals with a series of safeguards to prevent its abuse=
."
(http://economictimes.indiatimes.com/News/News_By_Industry/Healthcare__Bi=
otech/Pharmaceuticals/Pfizer_Glaxo__Novartis_will_have_to_wait_to_launch_ge=
neric/articleshow/2093016.cms)
The same day, Mint wrote: "Domestic pharmaceutical companies have won a =
significant victory, with a key committee of the government declining to c=
hange drug laws that currently allow them to make copycat versions of drug=
s by effectively using data filed by the patent-holder. The committee has =
refrained from granting drug makers exclusive use of chemical, test and oth=
er data they file with regulators while applying for drug approvals. Inste=
ad, it has recommended that data submitted by a drug maker be treated as a=
trade secret and get the same legal protection accorded to such confident=
ial data."
(http://www.livemint.com/2007/06/01234358/Govt-backs-local-drug-cos-on-d.=
html)
Subsequently, the Hindu and the Business Standard offered their appraisa=
ls of the report.
(http://www.hindu.com/2007/06/04/stories/2007060401361300.htm)
(http://www.business-standard.com/common/storypage.php?leftnm=3Dlmnu2&sub=
Left=3D1&autono=3D286458&tab=3Dr)
Part of the reason there is confusion over what the report actually reco=
mmends is the arcane nature of the issue (data protection) and the arcane =
nature of the recommendations themselves, which are couched in so many alt=
ernate (and differing) scenarios that it is difficult to tell exactly how =
the report will be implemented into law in the long run.
So what does this report actually say?
There are three sections to the recommendations - for pharmaceuticals, f=
or agro-chemicals and for traditional medicines. For the latter two catego=
ries, the report recommends immediate data exclusivity.
With respect to pharmaceuticals, the recommendations are split into two =
stages. For now, the report recommends that data be protected as a trade s=
ecret, with explicit changes made to the Drugs & Cosmetics Act to allow th=
is, as well as changes to the procedures for submission, storage and handl=
ing of such data. In a nutshell, this is the "data protection" model which =
is somewhat similar to Brazil, and close to the effects of similar prov=
isions in Argentina.
The report then recommends a "transition period" - which is unspecified =
- to move to a 5-year "data exclusivity" system. This data exclusivity sys=
tem comes with safeguards, which at first glance appear to be generally mo=
re comprehensive than those implemented in models elsewhere in the world.
The report says that both the transition period (if it should be, and fo=
r how long) as well as the subsequent data exclusivity are "subject to fur=
ther discussions." One reading of this is that the report's only actionabl=
e recommendations for pharmaceuticals pertain to a data protection regime,=
to be implemented immediately, and nothing further. If - and when - gover=
nment decides to tackle the speculative sections of the report, it may be =
assumed that a subsequent committee will decide on whether, and how, the t=
ransitional period and data exclusivity will fall into place.
Another reading is that, as such, the recommendations are final. What is=
to be delibarated is merely the length of the transition, as well as the =
fine tuning of the safeguards.
Either way, there is some confusion as to where this could lead. But why=
a transition period at all? The report makes it unambiguously clear that =
a data-protection system fulfils the minimum requirement of TRIPs. That is=
to say, the first step of its recommendation satisfies the criteria it wa=
s originally tasked to deliver on: as to what Article 39.3 mandates vis-a-=
vis data protection.
It further goes on to provide reasons why this would work well for India=
:
6.2.4: "Positive Impact: (A) It will allow greater competition to take p=
lace by not giving monopoly rights to the innovator drug companies on the =
non-patented drugs. This will help keep a check on the prices of these dru=
gs. (B) It will help greater availability of generic drugs as the domestic=
companies will continue to make generics of the pioneer drugs (not under =
patent) since the Drug Regulator will continue to rely on the data provide=
d by the first applicant while approving the subsequent applicants for sim=
ilar drugs."
Then it goes on to describe the "Negative Impact"
"a) It may have some adverse impact on FDI in pharmaceuticals sector in In=
dia since the perception with most international corporates may be that In=
dian does not provide adequate IPR protection to new products (b) This may=
slow down the early launch of some of the new drugs in India for lack of =
adequate protection."
If data exclusivity can harm the Indian public and the Indian economy no=
w, as the report makes explicit, it is not clear at all as to why it won't=
in - say - 10 years' time. Nowhere is this explanation provided except fo=
r a weak analysis by the authors=92 to conclude that data exclusivity will=
lead to increased pharma FDI to India, and the =93early launch=94 of inter=
national drugs.
What we are told =96 quite clearly =96 is that this committee would have=
liked to recommend data exclusivity today, if not for the fact that MNC p=
harma public relations just hasn=92t done its job: =93this [transitional] =
period can be utilized to educate the public and industry so as to allay t=
heir apprehensions on the issue.=94
Apparently a few years plus a lot of money will provide us the education=
we need and the foresight we require to =93allay apprehensions.=94 But ju=
st for the record, and before MNC spin-doctors get to telling us what we s=
hould think=85.
***
The logic that data protection only =96 the immediate recommendation =96=
is =93unfair=94 is spurious. For one thing, the proportional investment m=
ade for clinical trials for India (specifically) is often so low that it i=
s recouped several times over by premium pricing and early entry by the o=
riginator. That is to say, there is absolutely no need for any further inc=
entive other than the market itself.
Consider Gleevec, a drug that has been much in the news. Gleevec earned =
Novartis $2.5 billion in 2006. Novartis reports that it had all of 45 payi=
ng patients from India during the year. At a cost of $2600 per patient pe=
r month, Indian sales represented 0.05% of the world market. If we accept =
the overblown $150 million cost of clinical trials, it means that the frac=
tion of this cost attributable to India is roughly $82,500. As such, the =
payment by the first 3 patients for the drug would cover the cost of clini=
cal trials =96 assuming a 90% profit in comparison to the generic price.
Or we could consider Plavix =96 a blockbuster heart medication marketed =
by Bristol-Myers Squibb and Sanofi-Aventis that earned them a whopping $6.=
3 billion in 2006. Retail sales data of Plavix tells us that the first 25 =
patients in India would adequately compensate the cost of clinical trials=
to bring this drug to market in India.
An odd twist is that India is rapidly becoming a very important locatio=
n for clinical trials outsourcing. Because of lower costs pharmaceutical c=
ompanies are looking to use India to reduce clinical development timelines =
so as to be able to launch their products sooner in the west.
Pharmaceutical FDI is generally not dependent on data exclusivity provis=
ions. The experience of several countries around the world suggests this. =
An in depth study of Jordan's data exclusivity implementation (Oxfam 2007)=
as part of its bilateral treaty with the United States found that there w=
as 'no FDI by foreign drug companies into Jordan since 2001 to synthesize =
or manufacture medicines in partnership with local generics companies', an=
d the only FDI was setting up sales offices. For the 13 major trading part=
ners of the United States for which pharmaceutical FDI is available easi=
ly there is no clear pattern as to the impact of implementing data exclusi=
vity laws. For 6 of the 13 countries, real foreign direct investment fell =
in the three years including and following the implementation of data excl=
usivity compared with the three years preceding it.
India, as a growing consumer market for pharmaceuticals is already - and=
will continue to be - a destination for MNC pharma.
***
The most serious problem with this report is that the =93safeguards=94 i=
t envisions for the data exclusivity to come, are, in fact, largely meanin=
gless. What=92s worse is that they are couched in language that might lead =
people to believe otherwise.
Two minor points first.
(1) To exclude =93drugs for life threatening diseases such as HIV/AIDS=94 =
from data exclusivity, as the report suggests, is an interesting safeguard=
. It would be even more interesting to understand exactly what prescriptio=
n drugs this committee finds non-life-threatening: that list would undoubt=
edly be very small. To make this safeguard work, the committee would have =
to exclude virtually every existing drug in the Indian market (well, excep=
t perhaps drugs for erectile dysfunction, baldness and insomnia).
(2) Data exclusivity with safeguards doesn=92t seem to stay that way. On=
ce data exclusivity is in, it usually gets worse. If India is considering =
the idea of data exclusivity to stave off pressure from the USTR, in the f=
orm of the Special 301 Report that the US releases annually, it might want=
to reconsider given Canada=92s experience. A mainstay on USTR=92s specia=
l 301 report since 1995, Canada (despite allowing 5 year data exclusivity)=
drew much censure for allowing generics to be pre-approved before they co=
uld be marketed, among other things. The USTR only let off on Canada =96 o=
n this score =96 after that country pushed data exclusivity upwards to 8 y=
ears and removed some of the safeguards.
The major point next.
The safeguards that the Reddy Report recommends appear to be, on their f=
ace, comprehensive =96 so comprehensive, in fact, that some may conclude a=
t first blush that all of the pre-existing =93fears=94 surrounding data ex=
clusivity have been adequately addressed by these safeguards. Upon closer=
analysis, however, it becomes clear that the safeguards are both unworkab=
le and incomplete, and have the potential of undermining the real public h=
ealth safeguards that exist in our patent law.
The first safeguard, for instance, states that data exclusivity will onl=
y apply to those new chemical entities discovered on or after 1 Jan 1995. =
Presumably, this will prevent those drugs that were patented in other co=
untries prior to India=92s entry into the WTO in 1995 from obtaining a mar=
ket monopoly. Sounds good in theory, but rarely, if ever, is the picture =
so straightforward.
Take, for instance, the antifungal drug lubiprostone, which was approved=
by the US FDA in 2006. A query in the FDA=92s Orange Book shows that the=
re are not one, but five patents associated with this particular drug. Lo=
ok up each of these patents in the US Patent and Trademark Office website,=
and you discover that they were all filed on different dates, and that tw=
o of the five were actually filed (and thus =93discovered=94) pre-1995, wh=
ile the remaining three were =93discovered=94 in 1999, 2000, and 2002. Th=
en consider that of the 5 patents listed in the Orange Book for lubiprosto=
ne, 4 relate to a =93new method of treatment=94 and =93new form=94. All of=
these patents relate to the drug that was approved as an NCE.
So is lubiprostone an NCE? And was lubiprostone =93discovered=94 prior t=
o 1995? Or in 1999? Or in 2000? Or in 2002?
If the answer here is that the Drugs Controller must apply data exclusivi=
ty as soon as s/he sees even one patent that was filed after 1995, then th=
e safeguard is really no safeguard at all. In a forthcoming paper (that =
we are currently completing), we looked into the US FDA=92s drug approvals=
for =93new molecular entities=94 from January 2005 to present, and the pa=
tents associated with each drug in the Orange Book. Of the 36 NME approval=
s during this time, 22 of them (61%) had patents associated with them that=
were =93invented=94 prior to 1995. However, all but four of these drugs=
also had other follow on patents associated with it in addition to the or=
iginal pre-1995 patent.
That the vast majority of new drugs have not one, but several patents as=
sociated with it is hardly surprising, and is merely a confirmation of the=
well-known problem of =93evergreening=94 that Parliament intended to disc=
ourage with the enactment of section 3(d) and other safeguards in the Indi=
an patent law. However, this particular =93safeguard=94 of making data ex=
clusivity only applicable to drugs =93discovered=94 after 1995 would perve=
rsely create the incentive for drug companies to file for and obtain frivo=
lous follow-on patents on older drugs, precisely so that it could also qua=
lify for data exclusivity.
The alternative - and equally unpalatable - solution would be for the Dr=
ugs Controller to examine the numerous patents associated with any given d=
rug, and to make a substantive determination as to whether a given patent =
should be =93counted=94 for data exclusivity purposes.
But then the question arises: How, exactly, is the DCGI qualified to ma=
ke substantive decisions regarding patents? With all due respect to the DC=
GI=92s office, it simply lacks the capacity, resources and bandwidth to ma=
ke judgments on such a difficult topic as patent law. Indeed, even the U=
S FDA has acknowledged that it is simply not competent to make any judgmen=
ts on the validity of patents, and that forcing it to do so would only cau=
se further delay and confusion: =93The [FDA] believes that, even if it had=
the authority and expertise (which it does not), such a review would not =
speed the availability of generic drugs. Rather, it would instead add a lay=
er of complexity and delay, leading to litigation between FDA and the gene=
ric or innovator, in addition to any litigation between the generic and i=
nnovator.=94 [Testimony of Daniel Troy, Chief Counsel, US FDA, before the=
Senate Committee on the Judiciary, 1.8.2003; available at
http://www.hhs.gov/asl/testify/t030801.html]
Thus, this seemingly simple and straightforward safeguard, on closer ins=
pection, turns out to be either a perverse incentive to engage in evergree=
ning or an unworkable and unbearable burden placed upon the DCGI that is l=
ikely to result in even further delays.
The Reddy Report also recommends as a further safeguard, that =93in the =
case of data protection for patented drugs, the period of protection shoul=
d in no case go beyond the 20-year period of patent protection in India.=
=94 However, the phenomenon of a data exclusivity period extending beyond=
the 20-year patent term is a relatively rare occurrence. Among the 36 FD=
A approved drugs that we examined, we found only one instance in which the=
exclusivity period would extend beyond the term of the last-to-expire pat=
ent. [Paliperidone, a treatment for schizophrenia, was approved in 2006 a=
nd granted data exclusivity until 2011. However, the lone patent listed =
for this drug will expire in 2009. Thus, data exclusivity would give two =
additional years of market exclusivity after the patent expires.]
Thus, this =93safeguard=94 is largely a red herring to distract our atte=
ntion away from the far larger problem: Data exclusivity will also be conf=
erred on those drugs that have no patent protection at all. The Reddy Rep=
ort quietly sneaks this in on p. 5, stating that =93Data Protection [read:=
data exclusivity] is required both for patented as well as for non-patent=
ed drugs..,=94 while conveniently neglecting to mention this key fact in t=
he recommendations. However, the impact of making data exclusivity availa=
ble for unpatented (or unpatentable) drugs is hard to overstate, particula=
rly in India, where Parliament has enacted several safeguards to ensure th=
at frivolous patents are not granted.
A large percentage of =93new=94 drugs coming in to the market would not =
qualify for patent protection in India. This means that if data exclusivi=
ty were in place, Indian patients (and the section of the generic drug ind=
ustry that caters to the domestic market) would be denied access to use or=
production of these drugs.
For instance, Section 3(c) of the Patents Act prevents the patenting of =
=93any living thing or non-living thing occurring in nature.=94 In 2005,=
the FDA approved a chemical called exenatide for the treatment of type-2 =
diabetes. Exenatide, however, was not created in a laboratory. In fact, i=
t was discovered in the venomous saliva of the gila monster, a poisonous l=
izard found in the wild deserts of the western United States and Mexico.=
Nevertheless, the United States Patent and Trademark Office duly granted=
it a patent. Luckily for the 62 million Indians who live with type-2 dia=
betes, lizard spit would not gain patent protection in India. However, Ex=
entaide also has data exclusivity in the US until 2010.
Section 3(d) of the Patents Act, as has been made famous by the ongoing =
Novartis litigation, prevents drug companies from getting a new patent by =
creating a new =93form=94 of an already existing drug, unless this new for=
m makes the drug more effective. However, a lesser known but equally impor=
tant clause in section 3(d) states that a new use of an already known comp=
ound cannot be patented. In other countries, drug companies routinely obt=
ain patents on new uses of known substances. For instance, the US FDA appr=
oved a drug called deferasirox in 2005 as a treatment for iron overload. =
For those who require frequent blood transfusions, such as people with sev=
ere anaemia, an excess of iron can build up in the body, often leading to =
severe organ damage and even death. Deferasirox is a derivative of a chem=
ical that was known and used for some time as an herbicide.
Novartis discovered that this chemical could also be used for treating i=
ron overload, and obtained two separate patents in the United States for =
this new use. However, due to the ban on the patenting of new uses of kno=
wn compounds, Novartis would again be unable to gain patent protection on =
this product in India. However, Novartis also owns data exclusivity on def=
erasirox =96 which doesn=92t expire in the US until 2010.
An invaluable weapon in the evergreener=92s arsenal is the ability, once=
the active ingredient has been discovered and patented, to obtain a new p=
atent, some years later, on the =93composition=94 of the final drug produc=
t. Patent offices are inundated with patent applications seeking a 20-yea=
r monopoly for combining an already known active ingredient with a variety=
of commonly used inactive substances to obtain the final product =96wheth=
er in tablet form, injectable solution, transdermal patch or other dosage =
form. These =93innovations,=94 of course, often do nothing to increase th=
e efficacy of the active ingredient, and serve very little purpose other t=
han to extend the patent life of the original drug patent. However, Indian=
law severely restricts the ability of drug companies from engaging in suc=
h practices with section 3(e), which excludes from patentability =93mere a=
dmixtures=94 that result in a product that is nothing more than the sum of=
its parts. An example of
such practice can be seen in the recent US FDA approval of anidulafungin,=
an antifungal of potential significance in treating the most commonly enc=
ountered AIDS-related opportunistic infections. With the original patent=
s on this drug set to expire in 2012, a subsequent patent application was =
filed in 2002 that claimed a formulation combining the active ingredient w=
ith =93an aqueous solvent.=94 The =93aqueous solvent,=94 the patent goes =
on to explain, can be either =93water or saline.=94
Luckily for us, the original patents on this drug are too old to be pate=
nted in India, and section 3(e) would prevent the patenting of the later =
=93invention=94 of combining a drug with salt water. American consumers =
are not so lucky. They will have to wait until the expiration of the seco=
nd patent in 2022 before they can access a cheaper generic version of this=
drug. Indian consumers would be lucky as well to access generics immediat=
ely, barring the data exclusivity on this product, which lasts till 2011 i=
n the US.
MNC pharmaceutical companies routinely disguise their claims of a new us=
e of a known compound with clever language stating that they have discover=
ed new =93methods of treating=94 human beings by administering the already=
known product. This has been an extremely successful tactic for drug com=
panies to recycle old compounds for newly discovered uses. However, secti=
on 3(i) of the Indian Patents Act states that =93any process for the medic=
inal=85or other treatment of human beings=94 cannot be patented. This is=
welcome news for those suffering from multiple myeloma, a rare form of bl=
ood cancer that is almost always fatal. The US FDA, in 2005, approved len=
alidomide as a treatment for multiple myeloma. The active compound that fo=
rms lenalidomide, however, was known as far back as 1972. Despite this, t=
here are eleven patents covering this drug in the U.S., all but two of wh=
ich are =93method of treatment=94 patents that would be invalidated by sect=
ion 3(i) of the Indian
law. (The other two, by the way, are =93composition=94 patents and would =
be invalidated by 3(e)).
Thus, instead of waiting until 2023 for an affordably priced generic to =
enter the market, those suffering from multiple myeloma can entertain the =
hope that an Indian generic will develop a cheaper version soon. But lenal=
idomide comes with data exclusivity that only expires in 2010 in the US.
Indeed, when we examined the patents associated with the 36 NME approval=
s from 2005-2007, we found that up to 80% of these drugs would come with p=
atents that would not be valid in India. Of the 110 patents that are atta=
ched to these 36 NMEs, we found that as many as 96 of these patents would =
fail under one or more of the safeguards that have been implemented in our=
patent act. Of these, only 18 of them would be invalid simply because th=
ey were filed prior to 1995. The remainder of them would be invalidated o=
n one or more of the other safeguards that exist in this act.
***
As these numbers show, the safeguards in the patent act, if rigorously a=
nd correctly applied, have the potential to invalidate many bad patents th=
at create wholly unjustified monopolies.
However, with the introduction of data exclusivity =96 even with all of=
the alleged =93safeguards=94 that are detailed in the Reddy Report =96 th=
ese stringent patent provisions would be rendered meaningless.
In conclusion (this is merely a preliminary note prompted by the release=
of the Satwant Reddy Committee Report) we strongly suggest that governmen=
t take no further action on either the transitional period, or the eventua=
l data exclusivity suggested, until the empirical evidence is assembled, a=
nalysed and thoroughly examined.
Chan Park
Achal Prabhala
Arjun Jayadev
---------------------------------
The fish are biting.
Get more visitors on your site using Yahoo! Search Marketing.