GILD/MRK/ABBV: Why Everyone Wins if US Pricing Falls (In a Certain Way)

Richard
Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0


Richard Evans / Scott Hinds / Ryan Baum

SEE LAST PAGE OF THIS REPORT FOR IMPORTANT DISCLOSURES

203.901.1631 /.1632 / .1627

richard@ / hinds@ / baum@ssrllc.com

February 3, 2016

GILD/MRK/ABBV: Why Everyone Wins if US Pricing Falls (In a Certain Way)

  • On the eve of Sovaldi’s approval we estimate just over 1M US persons were HCV positive and diagnosed. Fully treated, this population would generate 3.3M prescriptions. Since Sovaldi’s approval roughly 900,000 Rx’s have been filled by current generation (Sovaldi, Harvoni, Viekira Pak) agents – and demand has begun to decline
  • If this trend continues US demand will reach about 1.7M Rx’s in aggregate, implying that about half of diagnosed US patients will have been treated, far fewer than the roughly two-thirds of diagnosed US patients treated in the last generation of HCV therapy (pegylated interferons, or ‘peg-IFNs’)
  • Current generation agents require 12 weeks of oral therapy, but produce few side effects and are very nearly a guaranteed cure. In contrast the peg-IFNs required a year of weekly injections, produced appalling side effects, and offered very low cure rates
  • If two-thirds of patients are willing to show up for peg-IFN therapy, then surely a far greater proportion are willing to show up for therapy with the current generation agents. In reality the patients have shown up, only to be turned away because of pricing. If prices go down, the patients will be treated
  • Current generation net pricing (circa $45,000) is about 50% higher than (2015 dollar) net prices for the peg-IFNs at the peak of their demand. If manufacturers offer prices closer to the historic peg-IFN level we believe payors would remove all or most restrictions on patient access, opening the door to an additional $14B to $19B in aggregate US demand
  • As interesting as the US HCV story may be, ultimately it’s a distraction – ex-US demand for HCV ultimately should equal 2.2 – 2.5x US demand. For practical reasons much of this demand should go to pan-genotypic regimens, where GILD holds a convincing lead
  • We believe the consensus expectation for global HCV sales over the next 5 years ($89B) overstates true demand by $20-$30B, even if US pricing can unlock US patient restrictions. The issue is timing; we see $84-$108B in global demand left to play for, but history shows that ex-US demand moves slowly
  • We believe consensus also errs in giving too much global share (14%) to ABBV and too little (7%) to MRK. We see MRK’s Zepatier displacing ABBV’s Viekira Pak outright; and, there is as yet no evidence that ABBV’s pan-genotypic regimen can compete with GILD’s

Where we’re BULLISH: Biopharma companies with undervalued pipelines (e.g. AMGN, BMY, GILD, SHPG, VRTX); Biopharma companies with pending major product approvals (e.g. ABBV, ACAD, ADMA, ALIOF, BIIB, CHMA, CLVS, CPRX, CTIC, GILD, ICPT, JAZZ, LLY, LPCN, MRK, NVO, OCUL, PTCT, SRPT, TEVA, ZSPH); SNY on sales potential for Praluent (alirocumab); CFN, BCR, CNMD and TFX on rising hospital patient volumes; XRAY and PDCO on rising dental patient volumes and rising average dollar values of dental products and services consumed per visit; CNC, MOH and WCG on bullish prospects for Medicaid HMOs; and, DVA and FMS for the likely gross margin effects of generic forms of Epogen

bull

Where we’re BEARISH: PBMs facing loss of generic dispensing margin as the AWP pricing benchmark is replaced (e.g. ESRX); Drug Retail as dispensing margins are pressured by narrowing retail networks and replacement of AWP (e.g. WBA, CVS); Research Tools & Services companies as growth expectations and valuations are too high in an environment of falling biopharma R&D spend (e.g. CRL, Q, ICLR); and, suppliers of capital equipment to hospitals on the likelihood hospitals over-invested in capital equipment before the roll-out of the Affordable Care Act (e.g. ISRG, EKTAY, HAE)

US Rx demand is stalling, consistent with the traditional HCV boom / bust pattern

Consistent with patterns seen with prior generations of HCV therapy, US Rx demand for the current generation agents[1] appears to have peaked (Exhibit 1, green line). In Exhibit 2 we overlay US Rx demand for the current and prior generations by month since launch, and extrapolate (dotted green) the demand trend for the current generation forward for approximately 3 years. Aggregate demand for the current generation agents as of November 2015 is 884,934 US Rx’s; our simple extrapolation points to eventual aggregate demand of roughly 1.7M US Rx’s by the start of 2019, at which point US demand would have become negligible

Low aggregate US uptake of new generation agents indicates prices are too high

NHANES[2] surveys the health and nutritional status of a statistically significant sample of US residents, and as part of the survey process collects blood samples from a subset of respondents. These are tested for, among other things, HCV antibodies (which indicates HCV exposure, whether or not the infection is still active); samples that are HCV antibody positive are further tested for HCV RNA (presence of which indicates an active infection)

These HCV testing data, along with respondents’ answers to HCV-specific questions, allow us to closely estimate the odds of a person being HCV infected (column ‘c’, Exhibit 3), diagnosed (col ‘d’), and treated (col ‘e’). Given estimates of treatment success and compliance rates with prior generations (pegylated interferons and earlier) of HCV therapy, we estimate that 30% of patients treated were cured (col ‘f’). The NHANES data tie to surveys completed before the launch of the current generation of agents[3]; as such our NHANES-based estimate of treatment candidates is more specifically a reasonable estimate of HCV treatment candidates in place before the current generation of products launched

In Exhibit 4, we estimate that as of the end of 2012, there were between 0.97M and 1.2M diagnosed HCV infected treatment candidates, who most commonly fall into the 1945 – 1965 birth cohort, have incomes below 138FPL, and/or are privately insured. And, we estimate 1.2M to 1.4M undiagnosed treatment candidates, who also most commonly fall into the 1945 – 1965 birth cohort and have incomes below 138FPL; however, unlike diagnosed candidates, the undiagnosed candidates are most commonly uninsured. Note that institutionalized persons, particularly prisoners of whom as many as 450,000 may be HCV positive, are excluded from these totals

The mid-point of our estimate of diagnosed treatment candidates is 1.09M persons; at 3 prescriptions per patient, this population could be expected to generate 3.3M US prescriptions, as compared to the roughly 900k current generation Rx’s filled to date, and the roughly 1.7M that we estimate will be filled by the time demand fades. Absent changes, the clear implication is that on slightly more than half of non-institutionalized, diagnosed US treatment candidates will gain access to the newer generation agents

This roughly one-half uptake rate is at least a bit lower than the 59% to 70% uptake rate we estimated from NHANES data for the prior generation agents (Exhibit 3 again, column ‘f’). However, bear in mind that the prior generation agents – essentially the pegylated interferons (IFNs) – required weekly injections for 48 weeks and produced appalling rates[4] of adverse effects. Patients tolerating the peg-IFN regimens could expect no better than an 80 percent likelihood of remission (for relatively rare genotypes 2,3, and 4); for more common genotype 1 patients the rate of remission ranged from 29% to only 51%. By way of comparison, current generation agents require only 12 weeks of daily oral therapy, have modest adverse effects[5], and boast efficacy rates of 90 percent or more

Where the peg-IFN demand curve was patient-constrained (because of low tolerability and low efficacy), the current generation demand curve is payor-constrained (because of high prices, payors are denying access to treatment)

At the peak of their US demand (circa 2004) the list price (in 2015 dollars) of a peg-IFN regimen was approximately $32,000 – $36,000[6], with net prices of roughly $27,000 – $31,000. List prices for Harvoni and Viekira Pak are in excess of $80,000, with net prices of roughly $45,000 – about a 50 percent premium to peg-IFN pricing. MRK’s Zepatier was recently approved with a US list price of $54,600; this is well below the GILD and ABBV list prices, but also well above the prevailing net prices – thus it remains unclear whether MRK’s entry will be offered at a true discount

Payors’ approach to the US HCV market has been to slow-walk patient demand, and offer exclusive formulary positions in exchange for lower prices. What’s needed is for payors to eliminate restrictions on patients’ access in exchange for lower prices. This occurred in at least one instance; ESRX agreed to eliminate restrictions (beyond being RNA positive) in exchange for ABBV giving what was at that point an aggressive discount on Viekira Pak. GILD cut Harvoni’s net price to other payors in response – but did not effectively tie these price concessions to the elimination of treatment restrictions. As a result, net pricing fell, but most payors’ treatment restrictions remained in place

Prices need to fall further – arguably toward the circa $30,000 level seen in the earlier (peg-IFN) generation — in order for the remaining US patients to get treatment. What’s crucial is that manufacturers rigidly tie their price concessions to an easing of eligibility restrictions

Pharmacoeconomics are one thing, budgets another …

GILD has defended Sovaldi / Harvoni prices on the basis of pharmacoeconomics – and very few health economists would argue the point that current generation prices are very much in line with the average economic values of the cures these agents bring about

That being said, many of the payors (particularly state Medicaid agencies and the Veterans Administration) providing coverage to large pools of potential treatment candidates simply lack the available budget, at current prices, to treat any willing patient. And, having covered therapy only for more severely affected patients during the last 2 years, the average acuity of the remaining patients almost certainly is lower than it was when these agents were first approved. As patients’ acuity levels fall, the economic value of treating them also falls, because the odds of patients progressing to liver failure fall with declining acuity, and the time needed to progress to failure from lower levels of acuity can be considerable

For both of these reasons – budget constraints, and the gradually deflating economics of treating less ill patients – we’re convinced that US net prices have to come down if the remaining half (or so) of diagnosed treatment candidates are to be given access to therapy

MRK’s entry is almost certain to displace Viekira Pak, and may catalyze a necessary recalibration of US prices

MRK’s HCV regimen (Zepatier) is broadly comparable to ABBV’s Viekira Pak. Rather than expanding the US HCV market to three viable competitors, we see the MRK entry displacing Viekira Pak, because of the latter drug’s relatively greater rate of adverse events as compared to Harvoni and Sovaldi (Exhibit 5)

We compare adverse event reports on a cases per Rx, rather than on an events per Rx basis; and, we compare products on a matched time-since-launch basis. This approach is reflected in the results shown in Exhibit 5. As is well known FDA has already added a labeled warning to Viekira Pak because of reported adverse events; however as we reported previously[7], if we ignore the cases FDA referenced in last year’s relabeling, we still find that Viekira Pak has a statistically significant excess of adverse event cases relative to either Harvoni or Sovaldi

We seriously doubt any formulary will commit to Viekira Pak as their sole HCV treatment given its excess rate of adverse events. This implies Viekira Pak will only be on formularies that are non-exclusive, which in turn implies that the brand will only capture volume if it can convince physicians – who are free to write other brands, particularly Harvoni – to write for Viekira Pak. We simply don’t think this is a viable position for the brand, and as such we see very modest US sales prospects for Viekira Pak in the event it remains available

We expect MRK’s regimen to essentially take Viekira Pak’s place as the lower-cost alternative to GILD’s Harvoni and Sovaldi. Like Viekira Pak, MRK’s Zepatier cannot compete head to head with Harvoni on an open formulary, so MRK has to offer lower (than Harvoni) net pricing in an attempt to win exclusive formulary positions

What’s crucial is that both MRK and GILD tie lower (than current) net pricing to an easing of restrictions on which patients are eligible for therapy. If this happens, we believe there are nearly as many diagnosed US HCV patients yet to be treated as have already been treated since Sovaldi first launched

Longer term, bigger picture – ROW sales, and the considerable advantages of pan-genotypic regimens

Across the roughly 13-year lifecycle of the last generation of HCV therapy – the peg-IFNs – nearly three-quarters of global revenue came from outside the US (Exhibit 6). Given the cost of the newer generation agents in the US, it’s doubtful the US:ROW price ratio seen with the pegylated interferons will carry over into the new generation, meaning ex-US sales arguably should be somewhat less than three-quarters of total demand for the new generation agents

To get a better estimate on US v. ROW sales for the current generation, we calculated other countries’ HCV sales potential on a relative (to US) basis (Exhibit 7). If we were to make no adjustment for the new generation agents’ leap in US prices as compared to earlier generation HCV products, the analysis would imply that ex-US sales would be roughly 3.5 to 4.1x US sales. Conversely, if we assume that the generational leap in US HCV prices isn’t replicated outside the US, the analysis implies ex-US sales of 2.2 to 2.5x US sales

Anyway we look at it, ex-US demand ultimately is far larger than US demand. However as one moves away from the US, and in particular away from the more developed countries, two important things happen: genotypes other than genotype 1 become more predominant; and, the practical hurdles to genotype testing become greater

Genotyping for HCV is a relatively complex procedure[8] that requires basic PCR apparatus (+/- $10,000) – which in less developed areas will only exist in major population centers. Samples can be sent from remote collection sites to central testing; however, the samples have to be held at -20 Celsius or lower from the point of collection to the point of testing, and in less developed areas this is impractical. Because of the challenges to reliably genotyping large numbers of patients in less developed areas, obviously it would be far easier to be able to treat patients without knowing the patients’ genotypes – and this is of course possible with the pan-genotypic regimens

As we look at HCV regimens in development (Exhibit 8), we don’t see any agents that are likely to reverse Harvoni / Sovaldi dominance in genotype 1. MRK’s follow-on 8-week regimen could prove to be preferable to Harvoni / Sovaldi because of its shorter duration; however, it’s not likely to enter the US (or developed world) market before 2020, by which time any willing US or western European patient with a genotype 1 infection arguably will have been treated with an earlier-approved agent

As such we view the genotype 1 game as having been won, and see the development competition shifting to pan-genotypic regimens. GILD’s Sovaldi / velpatasvir regimen should be the first approved (2016), followed as soon as 2017 by ABBV/ENTA’s ABT-493/ABT-530. Of note we have dramatically more efficacy data on the GILD than on the ABBV regimen; there are 12 week sustained viral response (SVR12) efficacy data available for GILD across all 6 genotypes, whereas we have only SVR4 data for one genotype (GT1) for the ABBV regimen. And, GILD has a second, shorter-duration (6wk) pan-genotypic regimen that could launch as early as 2019, though only limited efficacy data are available

A realistic estimate of remaining global HCV sales potential

As of January 2016 we estimate roughly 1.8M US persons who are infected with HCV, roughly 550,000 of whom are diagnosed. Assuming lower prices, we believe a majority (+/- 75%) of those diagnosed can be treated. And, we assume some portion of those currently undiagnosed (+/- 10%) will be diagnosed over time, which for the purposes of our estimate is roughly 5 years. At net prices ranging from $25,000 to $35,000 per course, this points to remaining US sales potential of roughly $14B – $19B (Exhibit 9)

 

As we pointed out in the prior section and in Exhibit 7, we believe ex-US sales potential is 2.2x to 2.5x US sales. This is hardly a heroic assumption – there are 27 times as many HCV infected persons in these countries as there are in the US, and these countries have average per-capita Rx spending that is 17% of the US level. Simplistically, this would point to ex-US demand of more than 4 times US demand; however, by calculating estimates on a country by country basis and lowering the relative spending assumption to account for relatively extreme US pricing, we come to the more conservative 2.2x to 2.5x figure

US sales of Sovaldi, Harvoni, and Viekira Pak reached a cumulative $21.4B as of 3Q15; this plus our $14B – $19B estimate of remaining US sales potential points to a US total of $35B – $40B for the current generation agents. Continuing with the assumption that ex-US demand is 2.2 – 2.5x US demand, this indicates total cumulative ex-US sales potential of $77B to $96B, gross of the $7B in ex-US sales registered through 3Q15

So on net, there’s roughly $84B – $108B in global sales demand left to play for, the vast majority of which is outside the US

Current consensus calls for about $88B in global sales from 2016 – 2020, distributed as shown in Exhibit 10. We see two probable errors with consensus – sales are unlikely to develop this quickly; and, the percentages of sales allocated to ABBV are too high, and to MRK too low

In the last major generation of global HCV therapy, US peg-IFN sales peaked within 2 years of launch, but ex-US sales peaked fully six years after launch. In aggregate, global sales developed as shown in Exhibit 11. Eighty-eight percent of sales came in the 3rd through 20th years of sales, and 54 percent came in the 3rd through 8th years of sales – corresponding to the 2016 – 2020 period in the case of the new generation agents. As such, if the global new generation sales curve unfolds at the pace of the peg-IFN curve, we’d expect to see roughly 61 percent (54% of 88%) of remaining global sales potential in the 2016 – 2020 period. Using our range of $84B to $108B for remaining global sales this implies 2016 – 2020 sales of $51B – $66B, well below the $89B called for by consensus

Consensus gives 14% of remaining sales potential to ABBV, which implies that Viekira Pak remains viable, and that ABBV’s pan-genotypic regimen is competitive. We have serious doubts that Viekira Pak is commercially viable because of its AERS profile (see Exhibit 5, again), and we’re not yet ready to give ABBV’s pan-genotypic regimen credit because it is likely to launch after GILD’s regimen, and has yet to publish comparable efficacy data to the GILD regimen (see Exhibit 8, again). We think it’s far more likely that MRK’s regimen takes Viekira Pak’s place in the US and Europe, and that GILD’s pan-genotypic regimens become dominant in the developing world. Accordingly, we expect MRK’s share of remaining global HCV sales to be in the teens, and ABBV’s to be in the lower single digits. In line with consensus, we believe GILD is likely to achieve a 75% or greater share of remaining HCV sales

  1. Sovaldi, Harvoni, Viekira Pak
  2. National Health and Nutrition Examination Survey
  3. The NHANES data behind these exhibits is from the period 1999-2012
  4. Using Pegasys as an example, 98 to 99 percent of clinical trial patients experienced adverse reactions; half to two-thirds experienced flu-like symptoms, one-quarter or more experienced nausea or vomiting, and one in five experienced adverse psychiatric effects. Adverse effect related quit rates varied from 10 to 19 percent
  5. No patients quit a Harvoni clinical trial because of adverse effects; roughly 1 percent of Viekira Pak patients quit a clinical trial because of adverse effects
  6. In 2004, peg-IFN wholesale acquisition cost (WAC) was $334.20 – $339.12 per week (x48 weeks); ribavirin (Copegus) WAC was $5.53 per 200mg tab, and was dosed at 1000mg to 1200mg daily for 48 weeks. Prices are inflated at 2.0% CPI for 11 years; the low end of the range represents Peg-Intron plus 1000mg daily of Copegus; the high end of the range represents Pegasys plus 1200mg daily of Copegus
  7. “ABBV: Viekira Pak Likely to See More Safety-Related Formulary Losses …”, SSR Health LLC, November 12, 2015
  8. http://www.cdc.gov/nchs/data/nhanes/nhanes_05_06/HEPC_D_met_LBXHCG.pdf

 

©2016, SSR, LLC, 1055 Washington Blvd, Stamford, CT 06901. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein. The views and other information provided are subject to change without notice. This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results. In the past 12 months, through a wholly-owned subsidiary SSR Health LLC has provided paid advisory services to Pfizer Inc (PFE) and to Merck (MKGAY) on both securities-related and non-securities-related topicsexh9_ed

Print Friendly