The following are conclusions from significant trials that we covered from this year’s American College of Cardiology meeting, held from March 24 to 27. Data from a number of major trials was presented this year, though for the most highly-awaited studies, ARISE and MERLIN, limited top-line data had been released previously.

Conclusions: AGI-1067 (AGIX), ISIS 301012 (ISIS), CSL-111 (CSL Ltd), LY518674 (LGND, LLY), Crestor (AZN), Torcetrapib (PFE), Ranexa (CVTX), Cleviprex & Angiomax (MDCO), SCH 530348 (SGP), Tolvaptan (Otsuka)

DYSLIPIDEMIA AND ATHEROSCLEROSIS

AGI-1067 – ARISE
As previously announced, AGI-1067 did not meet its primary endpoint. Today it was reported that AGI-1067 did meet a prespecified composite endpoint of cardiovascular death, resuscitated cardiac arrest, myocardial infarction (heart attack) and stroke, but given that AGI-1067 and placebo had very similar outcomes for the primary endpoint, this must mean that the need for revascularization/unstable angina endpoints had unfavorable imbalances. Given its unfavorable effect on lipids (increasing LDL-cholesterol and reducing HDL-cholesterol), this makes us nervous, though hard clinical outcomes are more important than these lab value changes. There were other positives and negatives as well. The drug apparently reduced the onset of diabetes, though the benefit in A1C (a measure of glucose control) was modest. There was a trend towards more heart failure hospitalizations. While the difference in liver toxicity was small, it is also worrisome, as the combination of increases in liver enzymes and bilirubin can be a harbinger of serious liver problems.

Thus the results are mixed. As we noted previously, another large trial would in all likelihood be needed for approval of the drug to confirm any benefits, and engaging in the trial would be quite risky. Atherogenics has said they have the funds to pursue it on their own--it is still unclear whether AstraZeneca will continue in the partnership. All in all, we are still negative on the drug's chances, and are leaving our likelihood of approval below average.

ISIS 301012 - Monotherapy
While the LDL lowering effects of the high dose (400 mg) group was impressive, most of the patients developed elevated liver function tests over 3 times the upper limit of normal. While it is not certain that patients would go on to develop serious liver toxicity, it is still worrisome, and likely the reason the press release noted that the anticipated doses of the drug are up to 300 mg/week. For more discussion on liver function tests with the lower doses, please see our past event. Data for the high dose group in combination with statins is expected tomorrow (March 26). Given that there was no new data on the lower doses, and the 400 mg dose is unlikely to be pursued, today's data does not really impact our projections for the drug.

ISIS 301012 – Statin Add-on Trial
The current release presents new data for the 400 mg/week cohort in a study where ISIS 301012 was added to statin treatment. Most of the data for the other groups is similar to what has been presented before. Unfortunately, while the 400 mg dose did not lead to dramatically increased liver function test abnormalities over the other dose groups (as in another study released at the ACC conference for monotherapy), there was also no improvement in the percent change in lipid parameters versus the 300 mg/week dose. Hence the current release on the 400 mg dose does not really change our assessment of the drug, though it does add confidence to the strong results seen for the 300 mg dose.

ISIS 301012 - Homozygous Familial Hypercholesterolemia
Though the data are only for a few patients in an open-label study, they are good, with ISIS 301012 lowering LDL-cholesterol on top of heavy standard treatment in patients with a serious form of a hypercholesterolemia genetic disorder. On a percentage basis, results were not greater than those seen for the 300 mg/week dose in the Phase II statin study of a more general population, and two of the patients in the current study continued to have very high LDL-cholesterol levels. Still, patients with this condition do not have many treatment options, and the reductions in LDL-cholesterol were substantial. ISIS 301012 appeared safe, which is particularly important given the high doses and multiple drugs patients were on. More data on ISIS 301012 as an add on to statins in the more general hypercholesterolemia population should be available later today. Given the encouraging, albeit preliminary, results in the current study and the likely advancement of ISIS 301012 for Phase III testing in this genetic condition, we are raising our likelihood of approval by another 5%.

CSL-111 (Reconstituted HDL) - ERASE
CSL-111 is reconstituted HDL consisting of apolipoprotein A-I from human plasma combined with material from soybeans. It has chemical and biological properties that are similar to human HDL.

In the current study of patients with a recent acute coronary syndrome episode, short-term infusion of the drug did not lead to significant reductions in a measure of atherosclerosis (percentage change in atheroma volume). Nevertheless, the numerical difference seen with placebo was suggestive, and we think further study is warranted.

Of interest, the magnitude of the change from baseline with CSL-111 was on the order of that seen in a small trial of a synthetic variant of HDL, ETC-216, though it is disappointing that the placebo subtracted results for ETC-216 appear better. Still these are small, short term studies, and results are quite preliminary.

The increase in liver function tests with the higher dose that had to be discontinued is somewhat troubling and could be an issue, though it is unclear whether this indeed represents liver toxicity.

Overall, given that we think further study is warranted, we are raising our likelihood of approval by 3%. CSL Limited will be discussing the results with experts to decide next steps.

LY518674 – Phase II
Two studies with LY518674 were reported, and results were generally disappointing. One study involved patients with high cholesterol, where some patients received concomitant Lipitor, and the other patients with the atherogenic dyslipidemia (high triglycerides and low HDL cholesterol).

LY518674 is 10,000 times more potent than fenofibrate (Tricor). However, in patients with dyslipidemia, while LY518674 decreased triglycerides and increased HDL-C (the "good" cholesterol), results were not dramatically different from fenofibrate. Moreover, LY518674 increased LDL-cholesterol to a greater extent than fenofibrate, which is very undesirable. At higher doses, LY518674 effect on HDL-cholesterol also started to decrease. On the safety side, it was troubling that LY518674, as well as fenofibrate, led to increases in serum creatinine, a measure of kidney function--though the mechanism for this and whether it is a serious problem is unclear.

In the hypercholesterolemia study, LY518674 did lead to greater reductions in triglycerides and greater elevations in HDL-cholesterol than did statin Lipitor, but its effect on LDL-cholesterol was much lower. And when added to Lipitor, there was no additional lowering of LDL-cholesterol, though there were still benefits on HDL-cholesterol and triglycerides. There was also some evidence of an effect on creatinine.

The increase in LDL-cholesterol in the dyslipidemia study makes it unlikely LY518674 would be a desirable drug in that situation. That effect on LDL-cholesterol along with the creatinine side effect, also raises questions for further development in conjunction with statins for patients with hypercholesterolemia (the study noted that the increase in LDL-cholesterol appeared to be correlated with elevated triglycerides--precisely the situation where such a drug would be most useful). Lilly has not announced its intentions, but given the discouraging results, we are lowering our likelihood of approval by 25%.

Crestor - METEOR
This was a study in low risk individuals, who per current guidelines would either not be recommended for drug treatment or for whom drug treatment is optional (please see our Hypercholesterolemia Indication Report for more). The high dose of Crestor was used. While the study did not demonstrate that Crestor significantly led to regression, there was significantly less advance in atherosclerosis compared to placebo. The difference was not necessarily larger than that seen in other statin trials, but was notable because patients in the current trial were at lower risk than in prior trials. Given the low risk of the groups, there were too few ischemic events to compare clinical outcomes, though oddly, all of the ischemic events occurred in the Crestor group.

The implications of the study, however, are unclear, particularly given the large number of patients who could be affected. We do not think that of themselves, these results would change guidelines. While the trial could spur more treatment of "optional" patients, we do not think physicians will want to necessarily use the high dose of Crestor for all of these patients. Further study on clinical outcomes and which patients are most appropriate for treatment are needed. Of note, AstraZeneca is conducting a very large clinical outcomes study, JUPITER, in patients with low to normal levels of LDL-cholesterol, but elevated C-Reactive Protein, an inflammatory marker. Final results are not due out, though, for another few years.

Torcetrapib imaging studies - ILLUSTRATE and RADIANCE 1
Results from the three Torcetrapib imaging studies were released today. The trials compared Torcetrapib plus Lipitor to Lipitor alone, and used imaging studies to measure atherosclerosis:
-ILLUSTRATE (the trial above) examined patients with coronary artery disease with intravascular ultrasounds (IVUS) of coronary arteries,
-RADIANCE 1 studied patients with heterozygous familial hypercholesterolemia (a genetic condition leading to elevated LDL-cholesterol levels) with ultrasound measures of the carotid artery intima-media thickness (CIMT, the carotid artery is in the neck, but is used as a surrogate for atherosclerosis in the heart), and
-RADIANCE 2 looked at patients with mixed hyperlipidemia (elevated triglycerides/low HDL-cholesterol) measuring CIMT.

In all of the trials, HDL-cholesterol was increased and LDL-cholesterol decreased with Torcetrapib/Lipitor relative to Lipitor alone, but all found no difference between the groups on their primary endpoint measurements of atherosclerosis. ILLUSTRATE found a small benefit for Torcetrapib on one of the secondary atherosclerotic measures (change in normalized atheroma volume), but RADIANCE 1 found worsening in one of its secondary endpoints (annualized change in mean carotid intima–media thickness for the common carotid artery). Fewer details are available for RADIANCE 2.

The RADIANCE studies had roughly a two-fold increase in serious cardiovascular events, whereas in ILLUSTRATE, the rate of such events was fairly close for the two groups (numerically favoring Lipitor alone).

All of the studies showed evidence of Torcetrapib raising blood pressure to a significant degree. However, based on calculations of expected results for the LDL-cholesterol lowering seen with the Torcetrapib combo against expected changes for the average blood pressure change, the RADIANCE 1 investigators estimated that there should be a net benefit for the Torcetrapib/Lipitor group. They noted that this left no room to impute any benefit for the increase in HDL-cholesterol. On the other hand, based on information provided by the investigators for ILLUSTRATE, it appears the blood pressure effect could more than offset the expected change in percent atheroma volume from the LDL-cholesterol lowering effect of Torcetrapib/Lipitor, and the actual change seen in percent atheroma volume may have come out slightly better than one would expect based on these estimates. It should be noted, though, that such calculations and estimates are quite imprecise.

The investigators have noted that the studies do not give a definitive picture of why Torcetrapib led to the worse cardiovascular outcomes that stopped its large clinical outcomes trial (ILLUMINATE). The possibilities include dysfunctional HDL, blood pressure effects, and other toxic effects (that may also be involved in the blood pressure increase). More analysis may help point to the answer, such as adjusting the data for blood pressure effects. While these effects most likely did offset any potential benefits somewhat, we tend to think there is a good chance that the large HDL molecules produced by CETP inhibition are also dysfunctional, and so are pessimistic about the chances for other drugs in this class.

ANGINA

Ranexa - MERLIN TIMI-36
Top line data had been released a few weeks ago (please see that event for a note on our revenue projections). As we noted, while disappointing that the main efficacy endpoints were not met, it is reassuring that the drug demonstrated safety in terms of arrhythmias, given that the drug prolongs the QT interval on EKG. From today's release, it appears that Ranexa did instead reduce arrhythmias (though the type of arrhythmias was not specified in the release), which had been suggested by preclinical studies, but the effect was not large enough to show up as a reduction in symptomatic arrhythmias. On the safety side, we would like more detail on the episodes of syncope, though this has been reported before as a side effect.

It was also reassuring that, as one would expect, Ranexa reduced several angina/ischemia endpoints, though we need to see more detail on the magnitude of this, as well as differences in myocardial infarctions.

While the negative primary endpoint means Ranexa will not be used after ACS for prevention of major cardiac events, we think the safety results seen so far do support broader usage in angina. Ranexa is likely to still be most used in patients refractory to other agents, since these are well established in the market. The drug now needs to prove itself in the market, and it remains to be seen at what rate revenues pick up. We also think that Ranexa would likely perform better with a major commercialization partner.

HYPERTENSION – INTRAVENOUS TREATMENT

Cleviprex – ECLIPSE
These results show that Cleviprex is generally able to control blood pressure better than major drugs currently used for this. The trial provided interesting new detailed information on the relationship between blood pressure control and mortality risk.

However, from the information released, there was only very limited evidence that this translated into a difference between drugs in terms of reducing major events: a lower mortality rate for Cleviprex than sodium nitroprusside. Still, most of the numerical differences favored Cleviprex, and the studies do demonstrate Cleviprex's safety (though we would like more details on the renal function outcomes). It was reassuring to see that rates of atrial fibrillation were similar for Cleviprex, since this had been a safety concern previously. One caveat, however, is that these studies were open-label, which weaken conclusions somewhat.

The information released does support approval of Cleviprex, and we are raising our likelihood of approval by another 5%. More data on Cleviprex's use in the acute treatment of hypertension in other settings (from the VELOCITY trial) are expected the first half of this year. ACUTE CORONARY SYSTEM

SCH 530348 – Phase II
SCH 530348 is the first of a new class of agents that blocks the platelet PAR-1 receptor, a receptor that binds thrombin. Thrombin is the most potent platelet activator, and also has a major role in the formation of strands of fibrin, the other component besides platelets involved in clot formation. Because SCH 530348 blocks the thrombin receptor on platelets, it does not interfere with thrombin’s effect on fibrin generation, which may theoretically help to reduce the risk of bleeding.

In the current study, SCH 530348 was added to standard therapy (including other anti-platelet agents and anti-coagulants, including heparin and Angiomax) prior to angiography (a test looking at what is blocking the heart arteries, to decide what type of treatment should be done). In those patients receiving percutaneous coronary interventions (PCI, such as angioplasty and stent placement) it was continued for 60 days at a maintenance dose. In those PCI patients, rates of TIMI major and minor bleeding were generally similar to the placebo group, though they were numerically higher (though not statistically significant) for the lesser category of non-TIMI bleeding. There also appeared to be some increase in bleeding in patients undergoing CABG (coronary artery bypass graft surgery). More details on outcomes for this latter group of patients is needed, though the group was relatively small.

Impressively, SCH 530348 resulted in a numerical reduction in cardiac events in the PCI patients, with evidence of a dose-response (though this was not statistically significant, given the size of the groups).

More detail is needed on whether or how much of the potential benefits of SCH 530348 is due to its action early on around the procedure versus from longer term use. In addition, there are other agents being tested for acute and chronic use in place of Plavix, which could change the treatment paradigm (please see our ACS Indication Report for more). It is still unclear whether SCH 530348 would ultimately be used in addition to both aspirin and Plavix (or another similar platelet inhibitor), the standard anti-platelet drugs now used after PCI, or could replace Plavix use, at some point in long-term treatment.

On balance, today’s results are encouraging results, and with the move into Phase III pending, we are raising our likelihood of approval by 25%.

Angiomax – ACUITY One Year Results
Today's release, showing similar 1-year mortality and ischemic events for the different arms of the ACUITY study are reassuring, though not entirely unexpected. The data do relieve concerns that using Angiomax alone (with provisional glycoprotein IIb/IIIa inhibitor (GPI)) could lead to worse ischemia and outcomes later. More detail is needed, however, on the subgroup of patients who were troponin positive (an enzyme measured in the blood signifying a worse condition), since this subgroup had a numerically higher rate of ischemic events (1% difference) at 30 days. Detail is also needed on patients who were not pretreated with clopidogrel, since these patients had a statistically significant worse outcome with Angiomax alone at 30 days.

Pending further details, the information does not change our expectations for Angiomax. There may also be some impact on usage of the drug, though, from other data released at the ACC showing that in general, drug treatment in patients with stable angina leads to equivalent major outcomes as percutaneous coronary interventions (PCI) (the COURAGE study). This could lead to moderation of PCI usage in lower risk patients, though a third of the drug treated patients in the study did ultimately need to have a PCI to control symptoms. On the other hand, there is a strong financial incentive for PCI and it did improve anginal symptoms in the study. Our PCI models for stable angina are under review as we gather more information.

Addendum
For the subgroup with elevated troponin (the more serious patients), the composite ischemia endpoint was seen in 17.7% of patients treated with Angiomax alone versus 16.4% in the heparin/GPI group, but mortality was 4.8% versus 5%. The difference in composite ischemia was close to what it was at 30 days (9.4% versus 8.4%). Also, for the general population, the difference in heart attacks widened slightly from 5.4 Angiomax versus 4.9 to 7.1 versus 6.2. While the results may give some physicians pause about using the drug in these more serious patients, the Medicine’s company is trying to make the point that increased bleeding is a more serious prognostic factor. Given that we are already projecting minimal use in the highest risk (STEMI) patients, we are leaving our revenue models unchanged.

Regarding implications from COURAGE, there is a lot of controversy about how the trial should affect practice. Proponents of PCI note that over 90% of the patients assessed for inclusion were screened out; that PCI did show greater benefits for angina control, and this has always been the main reason supported in the literature to do the procedure in this group of patients, and that it may be difficult to achieve optimal medical control in “real world” practice. Other authorities note that patients undergoing the procedure expect better outcomes, and the new information is likely to delay the procedure until it is more needed. We are slightly moderating our numbers of PCIs in patients with stable angina.

ACUTE DECOMPENSATED HEART FAILURE

Tolvaptan - EVEREST
The EVEREST study involved two short term studies and a long-term study on major clinical outcomes. Patients had acute decompensated heart failure, for which they were hospitalized. To date, there are no agents that have been shown to improve long-term clinical outcomes in these patients, and there are safety questions regarding a number of the drugs used.

In the short-term portion of the study, oral tolvaptan in addition to standard therapy led to modest improvements in outcomes such as body weight (a reflection of how much excess fluid a patient had) and shortness of breath, but not in global clinical status (at day 7 or discharge). While tolvaptan increased minor side effects, it was generally safe. This was reinforced in the longer term study, where tolvaptan was shown to be non-inferior to standard treatment in terms of mortality. It was disappointing though, that tolvaptan did not show any improvement in long-term outcomes, which had been suggested in past study. Tolvaptan did lead to lower weight after hospitalization, but results were no longer statistically significant by the end of the study. Tolvaptan also tended to improve low levels of sodium, but there were only a small percentage of patients with this problem at baseline, so a major benefit was not demonstrated.

Overall, the results suggest that while tolvaptan may have some benefit for short term use and appears safe, it does not improve major outcomes and there is no justification for long term usage. Viewing these modest benefits in light of the need for treatment in this indication, we are raising our likelihood of approval by 3%.

ACUTE CORONARY SYSTEM

SCH 530348 – Phase II
SCH 530348 is the first of a new class of agents that blocks the platelet PAR-1 receptor, a receptor that binds thrombin. Thrombin is the most potent platelet activator, and also has a major role in the formation of strands of fibrin, the other component besides platelets involved in clot formation. Because SCH 530348 blocks the thrombin receptor on platelets, it does not interfere with thrombin’s effect on fibrin generation, which may theoretically help to reduce the risk of bleeding.

In the current study, SCH 530348 was added to standard therapy (including other anti-platelet agents and anti-coagulants, including heparin and Angiomax) prior to angiography (a test looking at what is blocking the heart arteries, to decide what type of treatment should be done). In those patients receiving percutaneous coronary interventions (PCI, such as angioplasty and stent placement) it was continued for 60 days at a maintenance dose. In those PCI patients, rates of TIMI major and minor bleeding were generally similar to the placebo group, though they were numerically higher (though not statistically significant) for the lesser category of non-TIMI bleeding. There also appeared to be some increase in bleeding in patients undergoing CABG (coronary artery bypass graft surgery). More details on outcomes for this latter group of patients is needed, though the group was relatively small.

Impressively, SCH 530348 resulted in a numerical reduction in cardiac events in the PCI patients, with evidence of a dose-response (though this was not statistically significant, given the size of the groups).

More detail is needed on whether or how much of the potential benefits of SCH 530348 is due to its action early on around the procedure versus from longer term use. In addition, there are other agents being tested for acute and chronic use in place of Plavix, which could change the treatment paradigm (please see our ACS Indication Report for more). It is still unclear whether SCH 530348 would ultimately be used in addition to both aspirin and Plavix (or another similar platelet inhibitor), the standard anti-platelet drugs now used after PCI, or could replace Plavix use, at some point in long-term treatment.

On balance, today’s results are encouraging results, and with the move into Phase III pending, we are raising our likelihood of approval by 25%.

Angiomax – ACUITY One Year Results
Today's release, showing similar 1-year mortality and ischemic events for the different arms of the ACUITY study are reassuring, though not entirely unexpected. The data do relieve concerns that using Angiomax alone (with provisional glycoprotein IIb/IIIa inhibitor (GPI)) could lead to worse ischemia and outcomes later. More detail is needed, however, on the subgroup of patients who were troponin positive (an enzyme measured in the blood signifying a worse condition), since this subgroup had a numerically higher rate of ischemic events (1% difference) at 30 days. Detail is also needed on patients who were not pretreated with clopidogrel, since these patients had a statistically significant worse outcome with Angiomax alone at 30 days.

Pending further details, the information does not change our expectations for Angiomax. There may also be some impact on usage of the drug, though, from other data released at the ACC showing that in general, drug treatment in patients with stable angina leads to equivalent major outcomes as percutaneous coronary interventions (PCI) (the COURAGE study). This could lead to moderation of PCI usage in lower risk patients, though a third of the drug treated patients in the study did ultimately need to have a PCI to control symptoms. On the other hand, there is a strong financial incentive for PCI and it did improve anginal symptoms in the study. Our PCI models for stable angina are under review as we gather more information.

Addendum
For the subgroup with elevated troponin (the more serious patients), the composite ischemia endpoint was seen in 17.7% of patients treated with Angiomax alone versus 16.4% in the heparin/GPI group, but mortality was 4.8% versus 5%. The difference in composite ischemia was close to what it was at 30 days (9.4% versus 8.4%). Also, for the general population, the difference in heart attacks widened slightly from 5.4 Angiomax versus 4.9 to 7.1 versus 6.2. While the results may give some physicians pause about using the drug in these more serious patients, the Medicine’s company is trying to make the point that increased bleeding is a more serious prognostic factor. Given that we are already projecting minimal use in the highest risk (STEMI) patients, we are leaving our revenue models unchanged.

Regarding implications from COURAGE, there is a lot of controversy about how the trial should affect practice. Proponents of PCI note that over 90% of the patients assessed for inclusion were screened out; that PCI did show greater benefits for angina control, and this has always been the main reason supported in the literature to do the procedure in this group of patients, and that it may be difficult to achieve optimal medical control in “real world” practice. Other authorities note that patients undergoing the procedure expect better outcomes, and the new information is likely to delay the procedure until it is more needed. We are slightly moderating our numbers of PCIs in patients with stable angina.

ACUTE DECOMPENSATED HEART FAILURE

Tolvaptan - EVEREST
The EVEREST study involved two short term studies and a long-term study on major clinical outcomes. Patients had acute decompensated heart failure, for which they were hospitalized. To date, there are no agents that have been shown to improve long-term clinical outcomes in these patients, and there are safety questions regarding a number of the drugs used.

In the short-term portion of the study, oral tolvaptan in addition to standard therapy led to modest improvements in outcomes such as body weight (a reflection of how much excess fluid a patient had) and shortness of breath, but not in global clinical status (at day 7 or discharge). While tolvaptan increased minor side effects, it was generally safe. This was reinforced in the longer term study, where tolvaptan was shown to be non-inferior to standard treatment in terms of mortality. It was disappointing though, that tolvaptan did not show any improvement in long-term outcomes, which had been suggested in past study. Tolvaptan did lead to lower weight after hospitalization, but results were no longer statistically significant by the end of the study. Tolvaptan also tended to improve low levels of sodium, but there were only a small percentage of patients with this problem at baseline, so a major benefit was not demonstrated.

Overall, the results suggest that while tolvaptan may have some benefit for short term use and appears safe, it does not improve major outcomes and there is no justification for long term usage. Viewing these modest benefits in light of the need for treatment in this indication, we are raising our likelihood of approval by 3%.