By Jake King
Intercept Pharmaceuticals' (ICPT) shares immediately took off from its late October IPO price of $15 per share, soaring to a closing high of $41.70 on January 29th. However, in recent weeks the sellers have come out, with ICPT declining from $38 on March 25th to the current low-$30 range. Why the fallout? For one thing, shareholders that acquired stock in the aftermarket and participated in the rally are taking profits, but more interestingly, the "6-month lock up period" for investors in the IPO expired on April 8th, which explains why the stock has declined nearly every day since the lock-up ended. This low-liquidity stock presents a significant challenge for investors seeking to exit, but more importantly there's a looming risk that has not made its way into analyst reports, the company's presentation materials, or its financial statements. This risk could make the selling less orderly, in our view, as more research is conducted on the subject.
The key here is that Intercept's only clinical-stage compound, obeticholic acid (OCA), faces a major competitive threat in its lead indication, Primary Biliary Cirrhosis (PBC); it's a risk that has yet to be factored into the company's rich valuation. Generic fibrates (traditionally used to lower bad cholesterol and triglycerides) continue to be studied in multiple clinical trials for the treatment of PBC, including one Phase III study that is ongoing. Importantly, adding cheap fibrates to the standard of care for PBC may obviate the need for Intercept's drug. PBC is an orphan indication, so there are already few patients out there, and evidence that we've unturned suggests that the company may be overestimating the size of the market. Additionally, data from clinical trials indicate that the addition of fibrate drugs, which are already generic and widely available, to standard-of-care therapy (ursodeoxycholic acid, aka UDCA) may, in fact, produce similar, and perhaps superior effects, when compared to the company's OCA candidate, in patients that don't fully respond to first-line UDCA treatment. These studies also offer insight that Intercept's proposed Phase III clinical trial endpoint may not receive the FDA's blessing, as analysts expect. Given the run that ICPT has had thus far, the end of the IPO lock-up period, the risk that generic fibrates could represent a significant threat to the commercial opportunity for OCA, and the risk of acceptability around the Phase III clinical trial endpoint, shares of ICPT, we believe, are unlikely to hold their current valuation and could fall back to the initial IPO levels.
Generic fibrates as a treatment for PBC may negatively impact Intercept estimates. As we noticed the pressure on ICPT and dug into the story, we found the surprising risk that generic fibrate drugs have shown success in multiple clinical studies in refractory PBC patients, the same indication that the company's lead drug candidate, OCA targets. Analysts covering Intercept have failed to identify this risk, and therefore, have yet to factor it into their OCA revenue estimates, hence, numbers could be taken down. Furthermore, the company doesn't appear to mention it in its presentation materials or financial statements, despite that some of the same clinical trial sites running the OCA study are carrying out studies utilizing fibrates in PBC. The expectation from analysts and investors is that OCA will command exorbitant pricing (perhaps as much as $100k per treatment course) given that no drugs are approved for PBC patients refractory to UDCA. But should cheap, widely available fibrates demonstrate similar, or perhaps superior effects relative to OCA, pricing expectations and/or market penetration estimates are likely to be lowered significantly.
The need for refractory PBC treatments is there, but the market may be small and generic fibrates offer a cheap solution. While Intercept is testing OCA in a number of indications, the story is primarily based on OCA as a 2nd-line treatment for PBC. PBC is a relatively rare chronic liver disease characterized by progressive destruction of the interlobular bile ducts, which can eventually lead to liver cirrhosis and death. The only approved medical treatment is ursodeoxycholic acid (UDCA), which is sold by several generic drug manufacturers. While UDCA works in the majority of patients with PBC, a percentage of the population is "refractory" (biomarkers remain elevated and liver damage progresses), hence the need for additional treatment options. We see the risk of generic fibrates as multifold:
- They may prove at least as good as OCA in treating refractory PBC patients.
- The number of patients that will go on to OCA treatment is much smaller than the company predicts given optimistic market estimates and the unanticipated fibrate competition.
- Pricing for OCA, if successful in Phase III trials, will need to be much lower than analysts anticipate.
- Generic fibrates are already on the market, and their visibility in treating this disease is increasing. Some providers that we spoke with already use fibrates in 2nd-line patients who are without options.
- The significant number of clinical trials evaluating fibrates in refractory PBC offer information that suggests the company's Phase III clinical endpoint is not what independent clinicians may consider acceptable.
Below we address these points in further detail.
Efficacy of the combinations: UDCA plus fibrates vs. UDCA plus OCA. Intercept's 217-patient Phase III POISE trial (NCT01473524) evaluates refractory PBC patients treated with a combination of UDCA plus OCA (10mg dose cohort and 5mg-10mg dose titration cohort) vs. UDCA alone (standard of care). Importantly, Intercept's Phase II data for the combination (OCA plus UDCA) showed a mean reduction in alkaline phosphatase (ALP) levels of about 24% in refractory patients taking OCA 10mg for 12 weeks (baseline ALP values of about 300 U/L declined to about 230 U/L) at 12 weeks.
For comparison, a pilot 20-patient study by Levy C. et al. showed that UDCA + fenofibrate resulted in a mean ALP reduction of about 49.5% (baseline ALP values of 351 U/L declined to 177 U/L), a strong result that demonstrates fenofibrates' potential superiority to the Phase II data for OCA. A separate study by Hazzan R. et al. demonstrated that UDCA + bezafibrate in refractory PBC patients also had about a 50% reduction in ALP levels. Notably, there are numerous small studies demonstrating the benefit of UDCA and fibrate combination therapies (most commonly bezafibrate) in refractory PBC patients, with normalization of key clinical markers such as ALP, immunoglobulin M (IGM), gamma-glutamyl transferase (GGT), and aspartase aminotransferase (AST) (see here). What's really impressive is that the rate of response for UDCA plus a fibrate reached into the 80% range based on a 2011 Japanese trial, compared to Intercept's Phase II response rate of 40% (Intercept's Company Presentation - Phase II data applied to the Phase III POISE trial criteria -- ALP reduction to <1.67 times the upper limit of normal in patients treated with OCA 10mg and ALP reduction of just 15%).
As a result, a higher response rate and potentially better efficacy in reducing ALP levels for UDCA + a fibrate (plus normalization of other key clinical markers) is likely to present stiff competition for the UDCA + OCA combination regimen that Intercept proposes. The lower cost of fibrate therapy may be another nail in the coffin, and OCA's dose-limiting pruritus (severe itching) side effect may also lessen the ability to gain significant market share, if OCA is approved (more below).
Estimate of patient numbers may be too high, and potential fibrate competition worsens the picture. Intercept states that up to 50% of PBC patients are non-responsive or intolerant to standard-of-care UDCA, yet existing evidence suggests that this number is overstated. In addition, Phase III trial results for bezafibrate due around the same time OCA is expected to make its market debut render this picture even less compelling. Digging through the literature, we put together an assembly of non-responder criteria from previous trials.
- Most recently, a major study published in March in Gastroenterology, which used a cross-section of the United Kingdom-PBC patient cohort, demonstrated that 79% (inversely, 21% are non-responders) of 2353 patients in the cohort "met the Paris I criteria for adequate treatment response" with standard UDCA therapy (P < .05). This study, in fact, is one of the same that Intercept cites as support for ALP as an appropriate surrogate endpoint.
- Depending on criterion used for evaluation, Kuiper et al. reported UDCA non-response rates after one year of treatment of: 24.1% (75 of 311) according to Rotterdam criteria; 33.7% (105 of 311) according to Corpechot criteria; or 38.2% (119 of 311) according to Pares criteria. Either way you slice it, the refractory population appears to be realistically in the 20-40% range of PBC patients, not 50%.
- Chapman and Halliday in their editorial calling for "No more [fibrate] pilot studies," claim that 30-40% of patients are non-responders.
- Pares et al. demonstrated a responder rate of 61% (inversely, a non-responder rate of 39%) in a 2006 study of 192 PBC patients titled "Excellent long-term survival in patients with primary biliary cirrhosis and biochemical response to ursodeoxycholic acid."
Of course, asymptomatic vs. symptomatic patients (disease progression), early- vs. late-stage diagnosis, and other patient-specific characteristics may play a role in these estimates, but the bottom line is that the 50% "refractory" estimate being used by many analysts is likely too optimistic for OCA's addressable patient population. And if, as researchers anticipate, fibrates continue to demonstrate the ability to improve the hallmark biomarkers of PBC, as well as generate a clinical benefit, OCA's addressable market will only continue to shrink as these generic drugs may have significant advantages, including very low cost (outlined below). What initially looked like a Home Run for Intercept -- a clear, unmet medical need -- is beginning to look more like a Single at best.
OCA unlikely to command typical Orphan Drug pricing if generic fibrates become a widely-used alternative in refractory PBC. Intercept outlines the market opportunity for OCA thus: "There are approximately 300,000 people with PBC in developed countries… approximately 60,000 have been diagnosed and are on ursodiol therapy. Based on this estimate, we believe there are up to 30,000 PBC patients who may currently be eligible for treatment with OCA." Note that this estimate is not just for the U.S. but all developed countries, and that competing against a generic alternative outside the U.S. is likely a non-starter. Again, the company's outlook and resulting analyst estimates seem overly optimistic about OCA's application in PBC, but it may not even matter if OCA lacks pricing power. Most analysts are anticipating Orphan Drug pricing for OCA in the tens of thousands of dollars annually, some upwards of $100k annually given the pricing power that accompanies a drug for a tiny patient population with no alternatives. But breaking down the costs of a potential UDCA/fibrate combination for the treatment of PBC, which research points to as a potential go-to therapy, paints a picture in which OCA will have little chance to price with an obnoxious premium.
For example, a 70kg PBC patient (middle-aged woman) requires ~1,050kg of UDCA per day (based on the AASLD's recommended dose of 13-15 mg/kg/day). A 100-tablet bottle of 300mg UDCA can be purchased for $96.99. This bottle should last her about a month, thus UDCA costs roughly $1300.00 annually for an average, middle-aged female patient (most PBC patients are female).
Bezafibrate, the frontrunner in the testing of fibrates, can be purchased for $1.30 per 400mg capsule. Patients may require 1 per day (based on the ongoing Phase III bezafibrate study), thus, bezafibrate therapy should cost around $475.00 annually.
Overall, the total annual cost of a bezafibrate/UDCA combination will be less than $2000.00 per patient.
Bear in mind, this does not factor in the effects of health insurance reimbursement, which lowers costs to patients much more. It's difficult to imagine a scenario in which prescribers would utilize OCA, which analysts are expecting will cost in the tens or hundreds of thousands annually, when an effective treatment option is available for less than $2000 a year. Clearly, for OCA to be competitive, orphan drug pricing may just not be an option in the 2nd-line indication, if fibrates pan out as a compelling treatment option; analyst expectations may need to be lowered significantly.
The timeline for OCA to get to market is similar to late-stage fibrate results. Intercept expects to have top-line data from the 12-month double-blind portion of the Phase III POISE study (NCT01473524) in 2Q14. This would peg a potential approval for OCA about a year after these results are released (by mid-2015). Interestingly, results from the Phase III bezafibrate study ongoing in France (NCT01654731) are also expected in mid-2015, according to the clinicaltrials.gov listing.
The results of the late-stage bezafibrate study, if positive, could lead to widespread adoption of the UDCA/bezafibrate combination. Providers are seeking another treatment option for PBC patients, and fibrates are already approved in other indications, making off-label use in PBC simple. The final gating factor, physicians have indicated to us, is for fibrates to demonstrate efficacy and safety in a Phase III trial. Some practitioners that we spoke with said that they're already using fibrates off-label simply because their patients have no treatment alternatives. If UDCA-refractory patients continue without treatment of any kind, the final option in PBC progression is a liver transplant. Physicians, patients, and payers avoid these expensive and prohibitive procedures at all costs, and the results of the Phase III bezafibrate study should confirm what many already suspect -- that fibrates make an excellent 2nd-line PBC treatment in combination with UDCA. We also believe that as word gets out about new treatment options fueled by the visibility of the Intercept story, off-label bezafibrate use may start to increase ahead of the Phase III results. It stands to reason that at the same time OCA comes to market, physicians will have just seen the results of the Phase III bezafibrate study and may choose to use the fibrate combination first, and reserve OCA for last-line drug use. If OCA's pricing is as high as analysts anticipate, the drug could enter a virtual vacuum of demand in 2015.
FDA may not accept ICPT's proposed Phase III "clinical marker" endpoint. The primary endpoint for the company's Phase III POISE trial is the achievement of both an ALP level of < 1.67 times the upper limit of normal, with at least a 15% reduction from baseline, and a normal bilirubin level, compared to placebo. There is question that this primary endpoint will be acceptable, as it's based on the observation of clinical markers rather than clinical outcome (Intercept plans to file under an accelerated approval using this "surrogate" endpoint). While we do think that these clinical markers are predictive of outcomes, the ALP threshold chosen by the company may not be rigorous enough. Interestingly, looking at the available literature for studies evaluating UDCA + fibrates, we notice that independent investigator-sponsored studies utilize >1.5 times the upper limit of normal ALP as a threshold to determine whether a patient is refractory to UDCA therapy (see here). This suggests that Intercept's < 1.67x upper limit of normal endpoint could still be in the "refractory" range according to independent clinicians. In other words, treated patients that achieve ALP levels of between 1.5 and 1.67 times the upper limit of normal would be a success in the POISE trial, but in practice, would be rendered a failure in the eyes of some independent investigators. Additionally, the endpoint criteria of a 15% reduction in ALP levels from baseline is a surprisingly low bar set by the company, particularly, when we see UDCA + fibrates reducing ALP levels by as much as 50% in clinical studies. While the company seeks to demonstrate to the FDA that treating to this endpoint is clinically meaningful, we believe that typical clinicians seeing 1.67x the upper limit of normal as "refractory" weighs against convincing the FDA to accept this endpoint. In fact, the 1.67x ULN threshold is curious, and seems to be an arbitrary parameter, or simply the company "fitting" its clinical trial to meet the capabilities of OCA 10mg. We wonder what the response rate would drop to if the endpoint were changed from <1.67x upper limit of normal to <1.5x upper limit of normal. As a result, this is a key risk to the story, in our view.
POISE trial significantly managed for drug tolerability, and patient exclusion criteria raises questions on overall market size. There is significant evidence that OCA's tolerability profile is dose-limiting, and Intercept has managed the Phase III trial design accordingly. OCA, in fact, worsens a common PBC symptom: severe pruritus (itching). Patients in a study of OCA as a monotherapy demonstrated increased rates of pruritus: 94% in patients treated with 50mg and 70% in patients on 10mg, vs. just 30% in patients on placebo. Essentially, OCA significantly worsens this key symptom. Importantly, all drop-outs due to adverse events were attributable to pruritus, suggesting that the symptoms were severe. Standard of care treatment, UDCA, meanwhile, has demonstrated an ability to reduce pruritus in PBC patients, and with that in mind, Intercept likely decided to combine the two therapies for a second Phase II trial and ultimately, focus on combination therapy in the POISE trial. Nevertheless, the Phase II combination of UDCA + OCA still generated pruritus of 80% in the 50mg arm, 85% in the 25mg arm, and 47% in the 10mg arm, versus 50% in the placebo arm. So, it appears that the 10mg dose in the POISE trial was clearly chosen to avoid this side effect, AND, the combination with UDCA may in be part, to temper the pruritus side effect seen with OCA monotherapy. We don't like when therapies need to be dose-managed for side effects, let alone combined with other treatments for similar reasons. A clinical researcher that PropThink contacted stated:
"I do believe that fibrates will play a significant role in the treatment of PBC in the specific case of suboptimal response to UDCA, even after OCA takes place in the market. Fibrates are not expensive, are usually well tolerated, just as UDCA is, and do not generate potential disabling pruritus as OCA does."
Additionally, we point out that Intercept's POISE trial has excluded some of the sickest and most refractory PBC patients, those with portal hypertension and cirrhosis, narrowing the patient population in the study and leaving the question of just how many patients are really candidates for drug treatment. Intercept excludes patients with "portal hypertension and complications…cirrhosis with complications, including history or presence of: spontaneous bacterial peritonitis, hepatocellular carcinoma, bilirubin > 2x ULN." Portal hypertension is not uncommon in PBC. Pierre Michel et al. suggested that while previous studies had shown a rate of portal hypertension in PBC patients of between 20% and 80%, the group's 135-patient study demonstrated a 35% portal hypertension rate. Perhaps estimates need to reflect that the number of potential treatment candidates is again, smaller than expected, when adjusted for patients that truly may not be available for treatment. It's interesting that neither the ongoing Phase III trial in France (NCT01654731) nor the trial conducted in Florida (NCT00575042) exclude these sicker UDCA-refractory patients.
Finally, at least two analysts covering ICPT attribute 85% and 73% in value of their near-$50 and mid-$50 price targets to OCA in PBC, respectively, suggesting that expectations for the company are currently built almost entirely on success of the PBC indication. Given the risks that we've outlined, we believe that ICPT will have a hard time holding its current valuation and could trade down to significantly lower levels, perhaps back to its original IPO price.