How the antihypertensive losartan was discovered

How the antihypertensive
losartan was discovered
Gaurab Bhardwaj229 Tomasso Hall, Babson College, Babson Park, MA 02457, USA Based on interviews and publications, this case study is a history of how DuPont scientists discovered losartan, the first angiotensin II receptor antago- nist. Essential aspects of the story include: i) the discovery occurred at a young and inexperienced pharmaceutical business; ii) three bench scientists had recently graduated from PhD programmes and only the fourth had any industrial research experience; iii) pivotal to its success was the support and risk-taking of the highly experienced and recently hired head of pharmaceu-tical research; iv) a timely patent issued to Takeda Chemical Industries sug- gested a new line of research; v) a mistake made by an inexperienced pharmacologist yielded pivotal information; vi) the bench scientists were given the freedom to explore while being supported by research managers; vii) luck favoured the scientists in losartan’s subreceptor-binding and metabo- lite; and viii) the marketing group insisted that losartan not be developeduntil Merck expressed an interest in the drug candidate. Today, losartan is amultibillion dollar drug.
Keywords: angiotensin II receptor antagonist, discovery process, risk taking
Expert Opin. Drug Discov. (2006) 1(6):609-618
1. Introduction
A pioneering, multibillion dollar antihypertensive drug almost came not to be.
Discovered in March 1986 by scientists on their first assignments at a corporate
research lab after completing their PhDs, losartan was the first in a new class of
antihypertensive drugs, angiotensin II (Ang II) receptor antagonists. The com-
pany, DuPont, was also new to pharmaceuticals. Oil shocks of the 1970s and
high prices for petroleum and natural gas feedstocks for its traditional chemical
businesses, led senior executives to seek new businesses that would reduce
dependence on these inputs and generate higher profit margins [1]. Following its
long-practice of creating new businesses and growth from research in its labs,
DuPont had launched efforts by 1980 to create pharmaceuticals and other life
sciences businesses (Figure 1). The losartan work was among these early lab
endeavours. This is its history of discovery based on oral histories of scientists
and research managers, patents and publications.
Losartan’s discovery came from taking risks, scientists having the freedom to explore with the unstinting support of research managers, creativity, meticulous-ness, trying new approaches, making assumptions in the absence of data, mis-takes and luck. Inexperience and experience both proved essential; theycomplemented one another. In a pharmacologist’s mistake, inexperience provedto be a boon. However, it needed an experienced research manager to recognisepromise in the mistake’s results and persuade others that the new line of researchthat had been opened up, needed to be pursued. However, DuPont’s inexperi-ence later proved to be a liability when it almost decided not to develop losar-tan. This time, it needed an experienced company, Merck, to recognise thepotential of losartan, which persuaded DuPont that the drug candidate hadpromise and should be developed.
10.1517/17460441.1.6.609 2006 Informa UK Ltd ISSN 1746-0441 How the antihypertensive losartan was discovered
DuPont research
Corporate DuPont
FDA approvals
Figure 1. Timeline of key events in the discovery of losartan.
Expert Opin. Drug Discov. (2006) 1(6)
Bhardwaj
2. Research on the renin–angiotensin system
captopril from Squibb, had been approved by the FDA in in the early 1980s
April 1981. It was evidence that interfering with RAS couldlower blood pressure, increasing industry’s interest in Hypertension results from the high force of blood on vessel discovering drugs targeted at RAS, especially ACE inhibitors.
walls. The pumping of the heart sends rhythmic pulses of Lastly, a drug can be designed to bind with Ang II’s recep- blood surging through blood vessels. Under normal condi- tors and prevent it from doing so. No Ang II receptor antago- tions, vessels dilate and contract to accommodate the undu- nist drugs existed until losartan was discovered. Industry lating flow of blood. For many reasons, vessels may lose their attempts over many years had resulted in a few antagonists, flexibility or muscle tissues in vessels may contract. To move peptide analogues of Ang II, that were active in vitro, but the same amount of blood that is needed by the body, the could not be made into drugs. They lacked oral absorption heart must now pump with a greater force to overcome the and had short half-lifes – some of just a few minutes. Some higher resistance. Over time, this creates high blood pres- sure. Besides doing damage to the heart, hypertension can In the early 1980s, scientists were aiming their drug discovery be detrimental to other organs such as the liver, kidney and brain. Consistent high blood pressure can increase thechances of heart attack, stroke, kidney diseases and eye 3. Origins of the work
The causes of hypertension are many, complex, and In March 1982, DuPont hired RI Taber, a scientist with not entirely known, leaving room for many different nearly 20 years of research experience at the Schering Corpo- approaches to targeting the disease. By the 1970s, pharma- ration, to head pharmaceutical research at DuPont. By now, ceutical companies had begun drug discovery efforts aimed the company’s thrust into pharmaceuticals was well under way at the renin–angiotensin system (RAS), which was known and a number of research programmes were in progress. On to play a role in regulating blood pressure (Figure 2). How-
joining, Taber began reviewing them all [3].
ever, much was still unknown in 1982 when DuPont scien- Cardiovascular research at DuPont had been going on for tists turned their attention to RAS in their quest for an many years, but the effort was small compared with those at established pharmaceutical companies. A number of com- RAS regulates blood pressure by a sequence of steps, start- pounds had been synthesised and tested, but none had made ing with the production of the protein angiotensinogen in it as a drug. Taber found the scientists working on a variety of the liver [2]. Released into the bloodstream, it is not harmful cardiovascular compounds for a number of targets. He by itself. The kidneys release the enzyme renin into the noticed some of these were of the types that had been aban- bloodstream. Renin is specific to angiotensinogen, cleaving doned by others in the industry. Among vasodilators were it to form angiotensin I (Ang I), a decapeptide hormone.
some that were similar to minoxidil. Similar compounds had Subsequently, angiotensin-converting enzyme (ACE), which been shown to produce cardiac lesions in dogs and were no is nonspecific, cleaves Ang I to form Ang II, an octapeptide longer being pursued by others in the industry. Another dis- hormone. Ang II is active in causing high blood pressure and covery attempt combined different types of compounds, such other physiological effects. The Ang II molecule binds to as ACE inhibitors, β-blockers and vasodilators. These yielded receptors on the surface of muscle cells lining blood vessels.
large molecules that were not stable. Taber decided to quickly The binding results in the contraction of the vessels and put an end to these lines of cardiovascular research. He did rising of blood pressure. Thus, RAS presented at least three not think any were likely to result in a drug.
targets for drugs to interfere with its functioning and lower On completing his survey of all the pharmaceutical research efforts at DuPont, Taber concluded that the overall Drugs can be designed to inhibit the two enzymes – renin work was fragmented and spread thin. So he began by focus- and ACE. In blocking renin, a renin inhibitor drug would ing efforts in a few areas that included cardiovascular, central prevent the cleaving of angiotensinogen and the subsequent nervous system, anti-inflammatory and a small cancer pro- steps that lead to hypertension. Renin was an attractive target gramme. He suggested setting up three targets. Taber said, because in being specific to angiotensinogen its physiological ‘We will go off and look for compounds for these three targets and function was limited. A renin inhibitor drug would, thus, wherever we succeed, that will be our research programme’.
probably have few side effects. A few compounds that inhib- AL Johnson, who headed the medicinal chemistry group in ited renin had been discovered by the early 1980s, but they cardiovascular research, underlined the importance of such could not be made orally active to turn into drugs. Even by choices, ‘I think one of the major problems is identifying a good mid-2006, after decades of research, no drug in this class had area and going after it, and committing the resources to it to make sure it succeeds. The trouble with research is we can never Targeting ACE to prevent the formation of Ang II and sub- predict (which area will turn out to be good)’.
sequent steps leading to hypertension had been more Of the three targets, one that Taber suggested was the successful by the early 1980s. The first ACE inhibitor, Ang II receptor. Taber reasoned that instead of targeting the Expert Opin. Drug Discov. (2006) 1(6)
How the antihypertensive losartan was discovered
Renin–angiotensin system
Drug discovery efforts
Figure 2. Drug discovery efforts aimed at the renin–angiotensin system in the early 1980s.
Angiotensin II receptor sub-types were not known in the early 1980s.
ACE: Angiotensin-converting enzyme.
enzymes in RAS, the receptor might yield something novel this approach before. Although no peptide analogue had yet led and beneficial. No Ang II receptor antagonist drug existed.
to an Ang II receptor antagonist, the approach did not preclude The first in this new class of antihypertensives would have an the possibility. Besides, he did not have a better lead. ‘There was edge in the market. With captopril already being sold and not much out there scientifically’, noted Johnson of those days.
ACE inhibitors being pursued by many others, there was little Duncia felt that the peptide approach at least had some point in chasing those for a small research group such as advantages. There was some literature on peptide antagonist DuPont’s. Moreover, Taber felt that the history of pharma- compounds. Plus, the structure of captopril had originated ceutical research showed that a drug with a new mode of in peptides. Squibb scientists had taken a peptide obtained action often yielded important findings and led to unforeseea- from snake venom and designed analogues of it that led, in ble therapeutic benefits. Why not try an alternative route to 1975, to the discovery of the nonpeptide captopril. Peptides drug discovery? He argued, ‘An angiotensin receptor antagonist held the prospect of a more rational approach to drug would do everything that an ACE inhibitor would do and per- design. Duncia reasoned that if he could find an analogue haps would not be as toxic’. However, the DuPont scientists that antagonised Ang II, he would try fixing the undesirable characteristics. However, dealing with peptides was not easy.
They were floppy molecules, making drug design difficult.
4. Initial work with peptides
Nonpeptides were easier to work with because they wererigid and small. They were also easier to manufacture and JV Duncia, a chemist who had joined DuPont after obtaining more effective as drugs. However, he did not have any a PhD in organic chemistry from Princeton University in 1981, was assigned to the Ang II receptor work. He chose to Duncia began making small peptide analogues of different synthesise peptides, despite no one having had any success with fragments of Ang II to see if any would bind to the receptor Expert Opin. Drug Discov. (2006) 1(6)
Bhardwaj
and prevent Ang II from doing so [4]. The literature contained receptor-binding assay to first determine whether a particular some information, but it was contradictory. He pursued a compound would bind with Ang II receptors in isolated tis- claim that a central tetrapeptide had strong binding activity.
sues. Wong then evaluated the molecule in a functional assay His analogues of that fragment did not exhibit any strong to check whether in binding the molecule actually binding activity. Another report noted that a terminal antagonised. Another functional assay checked to see whether tetrapeptide had good binding activity. Duncia created an the molecule diminished the vasoconstrictor effect due only to analogue and found binding activity that was barely detecta- Ang II. Other chemicals in the body also cause vascular con- ble. He synthesised analogues of other fragments, many traction that is essential and does not result in high blood that were two or three amino acids long, but found noth- pressure. If the synthesised molecule reduced vasoconstriction ing. Eventually, after ∼ 1 year and many inactive molecules due to these chemicals as well, it would not be selective in its later, he concluded that contrary to the assumption with action – undesirable in a drug. Subsequently, the molecule which he had begun his work, analogues of small fragments would be tested in vivo to verify whether in vitro effects also of Ang II would not work. Most of Ang II was probably involved in binding, so a peptide with six to eight amino Chiu’s receptor-binding assay results of the Takeda acids would most likely be necessary to attain binding compounds synthesised by Carini caused disappointment.
activity. However, that was far too large a fragment to They were at odds with the patents’ claims. S-8307 and the analogue. Duncia’s peptide analogue approach had failed to other molecules did bind with the Ang II receptor, but extremely weakly [5,6]. They could have been dropped for lack The scientists also took another approach by setting up a of promise. They were at some other companies [7]. However, receptor-binding assay to screen compounds from DuPont’s instead of stopping with these results, Chiu and Wong contin- chemical library. Despite screening > 10,000 compounds, ued onto confirming the results in functional assays and then they found nothing useful. About four molecules were in vivo. They were being thorough, but follow-on tests in the ‘barely active’. Even the hint of promise had eluded their absence of anything worthwhile in the receptor-binding assay search for an Ang II receptor antagonist. ‘Finding it was like could be viewed as a waste of resources and time. However, looking for a needle in a haystack’, observed Johnson.
Chiu believed that in vitro tests had to be correlated and veri- ‘Back in 1982, we thought it was a formidable objective’, fied in vivo because many factors could cause results to differ Wong had also joined DuPont only recently, after com- 5. A mistake and pivotal findings
pleting his PhD in pharmacology in 1981 from the Univer-sity of Minnesota. The Ang II receptor antagonist After months of unyielding search, the scientists got their first programme was his first assignment. Accustomed to aca- break at the end of 1982. A routine patent search for Ang II demic research in pharmacology, Wong was not used to the receptor antagonists found two promising patents assigned to differences involved in industrial pharmacology. Responsi- Takeda Chemical Industries [101,102]. The patents, issued in ble for testing the Takeda compound in vivo, he injected a July and October 1982, described a family of compounds that large quantity (100 mg/kg) of it into a rat. In a human were novel imidazole derivatives which the company claimed weighing 70 kg, it would amount to an intravenous dose of had ‘excellent angiotensin II antagonistic activity and hypotensive 7000 mg. As a comparison, losartan taken orally by a activity and are useful as a hypotensive agent’ [102]. Excitingly, human only has 50 mg of the compound. Taber observed, these were that rarity – nonpeptides.
‘It was like adding the animal to the drug instead of adding Many other companies besides DuPont must have noticed the patents and Takeda had a head-start. Hurriedly, DJ Car- When Wong showed others his results, the immediate ini, a chemist who had just arrived at DuPont in 1982 after reaction was of dismay at the mistake. However, Taber, himself completing his PhD in organic chemistry from the Massa- a pharmacologist, quickly noticed, this time with amazement, chusetts Institute of Technology, was given his first assign- that although the molecule was extremely weak, it was selective.
ment to synthesise the Takeda compounds so that their It was evidence of the mode of action they were seeking. A drug effects could be verified. If found promising, they would, with such selectivity would have the desired effect with proba- finally, be a new line of attack. Carini synthesised com- bly few side effects. As the compound was extremely weak, it pounds that seemed most promising from the family took a vast quantity injected directly into the rat’s bloodstream described in the patents. One was named S-8307 (Figure 3).
to show selectivity [5,6]. The Takeda patents made no mention Meanwhile, despite his lack of success, Duncia persevered of selectivity, only potency that had, in any case, proved to be with the synthesis of peptides because nothing conclusive extremely weak. Wong also showed that S-8307 was a competi- had yet been shown about the Takeda compounds – there tive antagonist that indeed lowered blood pressure in rats, was no guarantee they would lead anywhere.
although weakly, with the very high dosage [5,6].
Two pharmacologists, AT Chiu and PC Wong, were Wong’s mentor in graduate school had impressed upon him responsible for testing the compounds. Chiu used a that selectivity was more important than potency. Focused on Expert Opin. Drug Discov. (2006) 1(6)
How the antihypertensive losartan was discovered
EXP-7711
EXP-6803
(losartan)
EXP-6155
Figure 3. Important molecules in the discovery of losartan.
evaluating the former, and knowing from in vitro results that discussion: ‘At that time it was felt that our chances of getting the compound was extremely weak, he did not think twice that kind of increase in activity were not good, and, admittedly, about the massive intravenous dose. He was unaware that they probably weren’t. The fact that it worked does not change the such an amount was unacceptably large for drug discovery.
fact that the odds were against us when we started’. However, Had he known that, he would have injected the ‘right’ (lower) Taber had little doubt that S-8307 was a viable lead that just amount. They would not have found selectivity and the had to be followed. Carini later described Taber inquiring Takeda compounds would have been dropped.
during the debate, ‘Well, do you have anything else? No? Well Through personal conversations, one of the DuPont scien- then, hell yes, it is a lead’. It was 1983.
tists later learned that Takeda had tried unsuccessfully toincrease the potency of their compounds. After a while, they 6. Nonpeptides and losartan
gave up and turned their attention elsewhere. Given the inter-est in Ang II receptor antagonists, the Takeda compounds Carini began synthesising a series of structural variants of were no doubt pursued at many other companies. In conduct- S-8307. Their activity, he found, was at best no better than ing their tests correctly, they would not have found any results that of S-8307. Most were inactive. Not long after, Duncia of promise. It took a mistake to discover selectivity.
joined him, abandoning his peptide work.
Thrilled by the results, Taber suggested directing discovery Duncia began by synthesising a couple of simple analogues efforts at the Takeda lead to design compounds that were and then moved to computer modelling to take a more similarly selective, but far more potent and which could be rational approach. He wanted to overlap and compare the taken orally. His view was not shared by everyone. The structures of Ang II and S-8307 to see how more potent ana- increase in potency needed was unrealistic, they said, and they logues could be made by better mimicking the binding were skeptical the molecule could be made orally available.
portion of Ang II. The chemists now believed S-8307 was They agreed that the Takeda compounds were the first indica- weak because it was too small compared with most of the tion of a nonpeptide Ang II receptor antagonist, but argued Ang II they believed was involved in binding. The question that this was also an untested approach. Carini recalled this still remained: which portion of Ang II bound and how? Expert Opin. Drug Discov. (2006) 1(6)
Bhardwaj
Duncia found a model of Ang II’s three-dimensional continued modifying and expanding it while retaining the structure from some published spectroscopic work [8]. This carboxylic acid group Duncia had earlier added [2,8]. An model hypothesised Ang II’s conformation in solution, not appendage on the molecule, an amine, was used to ‘hook on’ when it was binding to the receptor. There was no X-ray data other functional groups to enlarge the molecule. They made a available on the structure. However, starting with that con- large number of analogues. The effort eventually yielded formation, he realised he could use modelling to hypothesise EXP-6803, which showed another 10-fold boost in potency.
about Ang II’s contact points with the receptor. Then, the structure of S-8307 could be overlapped with that of Ang II Further synthesis now proceeded based on EXP-6803 [9,10].
to see what modifications were necessary in S-8307. The The chemists had been given a deadline 6 months earlier to computer modelling required making a number of assump- find orally active compounds; otherwise their line of research tions that later research showed were not entirely accurate.
would be terminated. It was now 1 month past the deadline, Nonetheless, in the absence of better information, assump- but no one had yet mentioned it. Someone surely would, tions had to be made. Fortunately, modelling led to ideas soon. Despite the big advances they had made so far, without that worked. The first assumption was taking the particular oral activity they had no drug. The molecule had to pass Ang II model as valid. There were also others in the literature through cell membranes, which are greasy, to get into the and the information was not consistent. Duncia chose his bloodstream. To pass through, the molecule too had to be model because it was based on Ang II’s conformation in greasy. Carini thought that a very different molecule, structur- water. He reasoned it was closer to reality than conforma- ally simpler, was required compared with the ones they had tions in other kinds of solution. Moreover, Ang II was been synthesising. At the same time, it had to retain those floppy. Its shape in solution could be very different from its characteristics they had found boosted potency. He hypothe- shape when binding with the receptor. ‘It was a big leap of sised that an amide bond linking two aromatic rings on their faith’, said Duncia. Next, assuming that S-8307 partially molecule was the culprit – it was not greasy enough. Further- mimicked Ang II in order to overlap their structures was more, often in turning a peptide into an orally active nonpep- another such leap. Unfortunately, there was no information tide the amide groups were replaced. Thus, he began available on the receptor’s structure.
synthesising analogues of the linkage to replace the amide. He The computer equipment needed for modelling in those substituted various groups. They did not work. Eventually, he days was huge, taking up an entire room. There was only one attached the hydroxymethyl group to the fifth position on the such machine at DuPont and a single individual, WC Ripka, imidazole ring. It made the molecule greasy. The simpler ran it. Ripka plugged-in the structural models of the two mol- structure would later shorten the synthesis work and contrib- ecules in the computer. The two structures were then over- ute in producing a metabolite with unexpected benefits. They lapped with Duncia’s instructions and moved around to find now had the molecule EXP-7711 that was orally active and it the right alignment. More assumptions went into this process.
was able to lower blood pressure for an extended period [9,10].
For example, both molecules had carboxylic acids. Duncia However, there was no increase in potency. assumed they were congruent. He also assumed that both The next series of molecules were modifications of molecules bound in the same way with the receptor. It need EXP-7711 [2,4,9]. With oral activity attained, the aim was to not have been so. Adding to the difficulty of computer model- continue increasing potency. EXP-7711 was still too weak to ling was the crude technology. ‘It was like a hand overlap. The be made into a drug. Much of the effort revolved around computer did not do anything. We overlapped it by hand, twist- replacing a carboxylic acid group with a variety of acidic func- ing it until finally I saw the groups line up the way I wanted tional groups. None were better than EXP-7711. After syn- them to line up’, said Duncia. Overlapping suggested that thesising a number of molecules, Duncia attached an unusual S-8307 may lack a second acidic functionality compared with acidic functional group called tetrazole that led to yet another the Ang II computer model. Adding that would make S-8307 10-fold increase in potency. They had losartan [103]. It was larger. Takeda had not patented a second functionality. If they could find an appropriate one, it would also make their mole- A tetrazole is a five-membered ring where four of the five cule different and patentable. Duncia added carboxylic acid as atoms are nitrogen. It is found in vinegar and some other the second functionality and created a new molecule, common chemicals. Tetrazoles had been used in medicinal EXP-6155 [8]. ‘The modelling was based on a lot of assumptions. chemistry, but never as an acidic isostere in an approved And the probability of something working after all those assump- drug. It was unusual enough that a number of people tions was very small’, he noted. Potency shot up 10-fold when inquired what it was. Why did Duncia use the tetrazole? Calling it a miracle, he explained, ‘I did not see it in the litera- The improvement in potency was encouraging, but not ture even though it was there, but I did not happen to come sufficient. EXP-6155 was not orally active in rats, even in across it. I just started drawing a heterocyclic analogue of a large doses. The working hypothesis remained that Ang II carboxylic acid. And I drew four nitrogens and I saw, yeah, the needed almost all of its amino acids for its activity, so valency is okay; you know, the number of bonds is okay; EXP-6155 had to be enlarged to raise potency. They everything is satisfied. I wonder if this thing really exists? And I Expert Opin. Drug Discov. (2006) 1(6)
How the antihypertensive losartan was discovered
had a chemistry book that I had bought for $2 at Barnes & receptor. Rather than being a ‘me too’ drug, it was a scien- Noble on 18th Street and 5th Avenue in Manhattan… in the tific breakthrough. It was the first orally active, potent, non- old days they used to have bins of books for a buck or two. As a peptide Ang II receptor antagonist. Although aimed at the graduate student I would take a bus ride (from Princeton, New same disease, it lowered blood pressure through an entirely Jersey) to go up there to see what books I could get for a couple of new mode of action. Pharmaceutical history was full of dollars. And I got this organic chemistry book written by a cou- instances where a new mode of action yielded unexpected ple of German authors [11]. Sure enough, in that book it gives benefits. It may not have the side effects associated with you a recipe for making tetrazole. I go, my goodness, I got to do ACE inhibitors. In any case, the concerns about losartan this. I asked Carini, Dave you have a precursor for making the could only be settled after some development work. As this tetrazole. Can I borrow some to make a tetrazole? He said, ‘Sure was a new class of drugs, there did not already exist a body John.’ It was very kind of him to do that. And I made it’.
of necessary scientific information. Data from at least Phase I Duncia continued, ‘Sometimes you have to take leaps of and early Phase II clinical trials were needed to draw valid faith. Sometimes you are not lucky, but sometimes you get that conclusions. Only then should economic and market analy- 10-fold kick in activity. And if you are very conservative and ses be performed. It was premature to reject losartan’s devel- are afraid to step out, you probably won’t get there’. Adding the opment and recommend its out-licensing. They were not tetrazole resulted in the final breakthrough – losartan or basing the decision on appropriate data. In both scientific DuP-753. The tetrazole improved oral absorption and and marketing terms, losartan would contribute to DuPont’s increased potency. Losartan was 1000-times more potent reputation in the pharmaceutical industry.
than S-8307. Between the two molecules lay hundreds P Timmermans (who had been hired to head cardio- others that had to be synthesised to reach the desired levels vascular research) noted that none of the major drugs repre- senting new classes in hypertension and congestive heartfailure could have been ‘predicted based on paperwork’. A 7. Selling losartan within DuPont
similar analysis would have resulted in none of them everhaving been developed. Taber, Timmermans, and others Despite the excitement losartan generated among the scien- spent great effort in convincing those outside drug discovery tists, it elicited little enthusiasm among the marketers in at DuPont of the promise of losartan. Some initial develop- DuPont’s pharmaceutical business. Taking losartan to subse- ment work was done, but it did not receive high priority. Its quent phases of development and commercialisation proved continued development remained clouded. Then, Merck to be a difficult sell. In late 1987, the company estimated stepped into the picture and changed it.
that development would require approximately a couple ofhundred million dollars and take ∼ 10 years. Comparable 8. Codevelopment with Merck
sums would then have to be spent on marketing, distribu-tion, and promotion. The marketers presented their report.
With a few promising molecules in various therapeutic They doubted losartan could be successful, based on the areas, DuPont recognised that it lacked the expertise to con- characteristics of the compound, the drugs already available, duct the sophisticated clinical trials needed. The company and the nature of the hypertension market. It was not worth approached Merck, for its experience and capabilities in developing and commercialising. The best move, they rec- drug development, to explore the possibility of codeveloping ommended, was to out-license losartan. It was a ‘me too’ the drug candidates emerging from DuPont’s young compound, in being a ‘modified ACE inhibitor’. With cap- pharmaceutical labs. Merck saw losartan’s potential. Given topril selling successfully, enalapril from Merck recently its well-established position in the industry, its reaction approved by the FDA in 1985, and other ACE inhibitors in carried weight. Doubts at DuPont turned to resolve. The various stages of development, there was little to differenti- struggling drug candidate suddenly gained credibility and ate losartan. Why would anyone buy it? Losartan would need a vastly superior safety and efficacy profile that it was The two companies signed an agreement to collaborate unlikely to have. In the 10 years that it would take to launch on the development and marketing of losartan, starting in it, new products would continue to enter the market and January 1990. Besides development capabilities, Merck also generics would exist. The market would be saturated by then.
had marketing and sales capabilities that were lacking at Most importantly, there were no existing unmet medical and DuPont and invaluable in launching and establishing a marketing needs in this therapeutic area. Not only was drug in a new class. For Merck, with its large presence in DuPont unlikely to earn a profit, it may not even recoup its the cardiovascular market, here was an opportunity to be expenditures in development and commercialisation.
part of a new class of drugs. The relationship between the The DuPont scientists and research managers disagreed two was further strengthened in January 1991 by the for- vehemently. It was scientifically incorrect of the marketers to mation of a joint venture, the DuPont Merck Pharma- label losartan a ‘modified ACE inhibitor’ and a ‘me too’ ceutical Company. The discovery of a new drug led to the drug. It did not inhibit an enzyme, but antagonised a Expert Opin. Drug Discov. (2006) 1(6)
Bhardwaj
9. Unexpected benefits
(losartan is the generic name of the drug; Merck sells losar-tan under the trade names Cozaar™ and Hyzaar™, which Pharmacological studies in rats revealed that losartan created an are registered trademarks of E. I. Du Pont de Nemours and active and potent metabolite [2,12]. Fortunately, it amplified the Company, Wilmington, DE; Hyzaar is a combination of antagonistic effect of losartan and extended its duration. Trou- losartan and a diuretic). The two companies shared the blingly though, animal tests showed that the metabolite’s creation revenue. Losartan’s new mode of action proved very effec- varied by species. It was created in rats, but not in dogs. Natu- tive and its selectivity probably resulted in its having few rally, there were worries about its creation and effects in humans.
side effects [14]. With its potency, long duration, and mild Clinical trials revealed that losartan created a major, active metab- side effects, losartan was generating > US$ 3 billion in olite in humans also. Luckily, it had beneficial effects. Working in annual sales by 2005 [201]. At the time of its launch, ana- tandem with losartan, the metabolite lowered blood pressure lysts had estimated annual sales of ∼ US$ 200 million [202].
much further than losartan could have on its own. Together, losa- The market for antihypertensives proved larger than had rtan and its metabolite were 10,000-times more potent than originally been estimated by DuPont’s pharmaceutical mar- S-8307. Whereas losartan had a half-life of 2 h, the metabolite keting group. By 2006, it was estimated that > 60 million had a half-life of 6 h. Losartan’s antagonistic effects, thus, lasted people in the US suffered from high blood pressure [15].
much longer and the drug could be taken just once daily.
Along with losartan, the market also sustained other drugs Subsequent research showed that the Ang II receptor actually that operated through various modes of action to lower consisted of two subtypes – AT1 and AT2 [12]. Although Ang II bound with both subreceptors, it was only the binding with AT1 Losartan’s structure served as prototype for others in their that caused vascular constriction, leading to high blood pressure.
design of new AT1 receptor antagonist compounds, such as Despite Ang II also binding with AT2, the latter played no role in candesartan, irbesartan, saprisartan, tasosartan, telmisartan, hypertension. The physiological function of AT2 would remain valsartan and zolasartan [2,16,17]. By 2002, the FDA had unclear for many years to follow [13]. Astonishingly, it emerged approved seven AT1 receptor antagonists and ten ACE that losartan binds only to AT1. The molecule was not consciously inhibitors (Figure 1). Among those was eprosartan that also
designed to do so. This binding contributed to its efficacy.
owed its origins to the Takeda patents, but was discovered As Taber remarked, drawing on > 20 years of research and independently by SmithKline Beecham using a different management experience in drug discovery, ‘You also have to be very lucky in the field of pharmaceutical research. There are many In 1997, the American Chemical Society awarded Dun- things that can happen that are totally outside your control’.
cia, Carini, Wong and two scientists from Merck its Awardfor Team Innovation. The joint venture between the two 10. Expert opinion and conclusion
companies was dissolved in 1998. DuPont continuedalone for a while until exiting the pharmaceutical business Approved by the US FDA in April 1995, losartan was in 2001. Merck continues to successfully sell Cozaar launched that month as the first Ang II receptor antagonist Bibliography
PC Wong - July 22 (1999); June 23 (2000).
Science and Corporate Strategy: Du Pont R&D, 1902-1980, Cambridge University et al.: The discovery of DuP 753, a potent, subtypes. Blood Press. (1996) 5:11-14.
DUNCIA JV, CHIU AT, CARINI DJ et al.: receptor antagonist. Med. Res. Rev. (1992) IRVIN JD et al.: Nonpeptide angiotensin II 12:149-191.
receptor antagonists: The next generation in WONG PC, CHIU AT, PRICE WA et al.: a new class of potent antihypertensives. antihypertensive therapy. J. Med. Chem. J. Med. Chem. (1990) 33:1312-1329.
(1996) 39:625-656.
characterization of 2-n-butyl-4-chloro-1- (2-chlorobenzyl)-imidazole-5-acetic acid, Adv. Med. Chem. (1993) 2:153-195.
DJ Carini - June 15 (1999), June 21 (2000); J. Pharmacol. Exp. Ther. (1988) 247:1-7.
et al.: Nonpeptide angiotensin II receptor et al.: Nonpeptide angiotensin II receptor antagonists: The discovery of a series of antagonists. II. Pharmacology of S-8308. N-(biphenylylmethyl)imidazoles as potent, Eur. J. Pharmacol. (1988) 157:13-21.
orally active antihypertensives.
J. Med. Chem. (1991) 34:2525-2547.
Expert Opin. Drug Discov. (2006) 1(6)
How the antihypertensive losartan was discovered
HILGETAG G, MARTINS A: Preparative Websites
Organic Chemistry, John Wiley & Sons, SURVEYS: Healthcare: Pharmaceuticals. et al.: Angiotensin II receptor antagonists receptor antagonists: an emerging new class Hypertens. Res. (1999) 22:147-153.
3:211-216.
Affiliation
new considerations in their mechanism of discovery of DuP-753. Curr. Opin. Assistant Professor of Strategy and Management, action. J. Clin. Hypertens. (2006) Thera. Patents. (1992) 2:1693-1718.
229 Tomasso Hall, Babson College, Babson Park, 8:381-385.
that act on renin-angiotensin system with emphasis in AT1 antagonists. Mini Rev. Med. Chem. (2001) 1:207-217.
Expert Opin. Drug Discov. (2006) 1(6)

Source: http://solvingbigproblems.com/s/Bhardwaj-EoDD-2006-idow.pdf