ABBV (and ENTA): The Corvette(s) to GILD’s Ferrari

Print Friendly, PDF & Email

Richard Evans / Scott Hinds / Ryan Baum


203.901.1631 /.1632 / .1627 richard@ / hinds@ /


September 2, 2014

ABBV (and ENTA): The Corvette(s) to GILD’s Ferrari

  • We believe consensus incorrectly views GILD’s pending hepatitis C (HCV) regimen as having consistently greater efficacy than ABBV’s pending regimen; in this note we compare pivotal trial results for the two regimens side by side, and show that the ABBV alternative offers comparable efficacy
  • The GILD regimen is of 8 (rather than 12) week duration in some but not all type 1 patients; the ABBV regimen is of 12 week duration in all patients. Patients are required to use ribavirin (RBV) more often on ABBV than on GILD regimens. Both regimens are highly tolerable; there were no withdrawals for adverse events (AE’s) in GILD’s three main trials; across ABBV’s five main trials, two saw zero AE related withdrawals, one trial saw a 0.5% rate, one a 1.0% rate, and one a 1.9% rate
  • The GILD regimen is one pill once daily; the ABBV regimen is no more than three pills twice daily
  • On net, the ABBV regimen offers the same efficacy, at the cost of greater treatment duration and/or the addition of RBV-associated tolerability issues. Despite this narrow relative disadvantage vs. GILD, in an absolute sense the ABBV regimen is both sufficiently brief and highly tolerable – making it a viable alternative, especially for cost-sensitive payors
  • In this note we detail the efficacy and tolerability gap between PegIntron and Pegasys, the two predominant pegylated interferons (IFNs) that have been the mainstay of HCV therapy for more than a decade. We show that despite being substantially less effective and less tolerable than Pegasys, that PegIntron has captured 34% of global pegylated IFN sales to date
  • The ABBV HCV regimens are far more closely matched to the GILD HCV regimens than PegIntron was to Pegasys; accordingly barring the effect of additional competitive entrants, we would expect the ABBV HCV regimens to capture global share on par with that captured by PegIntron (i.e. more than 30%)
  • Allowing for the effect of competitive entrants (the first with at least comparable efficacy to ABBV comes from BMY in 2016, the second with potentially comparable efficacy from MRK in 2017), we’re confident ABBV can capture at least 20% of global sales. Current consensus appears to credit ABBV with less than 10% share of global sales
  • One component (ABT-450) of the ABBV regimen is licensed from ENTA; we believe cumulative royalties to ENTA should range between $900M and $3.5B. ENTA’s current enterprise value is roughly $665M

ABBV’s pending (12/22/2014 PDUFA[1]) triple-regimen for hepatitis C (HCV) infection appears to be a worthy alternative to GILD’s pending (10/10/2014 PDUFA) combination of Sovaldi (sofosbuvir) and ledipasvir, particularly for cost-sensitive payors

Comparative Efficacy

In the following two exhibits we compare the reported efficacy[2] of ABBV’s ‘3D’ regimen[3] to that of GILD’s Sovaldi/ledipasvir combination, in genotype 1 treatment-naïve (Exhibit 1) and treatment-experienced (Exhibit 2) patients

In genotype 1a treatment-naïve patients, the 12 week ABBV regimen without ribavirin (RBV) compares favorably (90.2% SVR12, 95% c.i. 86.2% to 94.3%) to the 8 week GILD regimen (93.0% SVR12, 95% c.i. 88.1% to 96.3%). In this case both of the regimens trade convenience (ABBV without RBV and GILD for 8 weeks instead of 12) for lower efficacy levels than can be achieved with RBV (ABBV) or a greater duration of therapy (GILD). Adding RBV to the ABBV regimen raises efficacy in treatment-naïve 1a patients to 97.0% (95% c.i. 93.7% to 100%); extending the GILD regimen to 12 weeks raises efficacy to 99.3% (95% c.i. 94.6% to 100%)

In treatment-naïve 1b patients the ABBV 12 week regimen without RBV is 99.0% effective (95% c.i. 97.7% to 100%), as compared to a 97.7% SVR12 for the 8 week GILD regimen (95% c.i. 87.7% to 99.9%)

The companies’ reporting of efficacy rates in treatment-naïve type 1 patients with cirrhosis[4] is structured in such a way as to preclude a truly direct comparison (ABBV discloses the specific sub-genotype of type 1 cirrhotic patients but GILD does not). Nevertheless efficacy of the ABBV 12 week regimen with RBV appears to be on par with that of the 12 week GILD regimen

Moving to treatment-experienced patients (Exhibit 2); in type 1a the 12 week ABBV regimen with RBV is on par (96.0% SVR12, 95% c.i. 93.0% to 98.9%) with the 12 week GILD regimen (95.3% SVR12, 95% c.i. 88.5% to 98.7%). In treatment-experienced 1b, the 12 week ABBV regiment without RBV is on par (100% SVR12, no c.i. data available) with the 12 week GILD plus RBV regimen (100% SVR12, 95% c.i. 85.2% to 100%), and superior to the 12 week GILD without RBV regimen (87.0% SVR12, 95% c.i. 66.4% to 97.2%)

Again, the companies’ basis of reporting results for cirrhotic patients precludes direct comparison, but the ABBV 12 week with RBV regimen appears to be on par with the GILD 12 week without RBV regimen in treatment-experienced / cirrhotic type 1 patients

Comparative Tolerability / Convenience

As is clear from the preceding, the ABBV and GILD regimens offer comparable efficacy, however the ABBV regimen often requires adding RBV and/or an additional 4 weeks of therapy as compared to the GILD regimens. And, it should be noted that the ABBV regimen requires multiple pills[5] twice daily, where the GILD regimen is a single pill taken once daily, except when used with RBV (RBV is always twice daily, regardless of whether it’s used with the ABBV or GILD regimen)

Adverse-event related discontinuation rates in pivotal clinical trials were very low for the ABBV regimens (0% to 1.9%), and zero for the GILD regimens (Exhibit 3, top row). Rates of individual adverse events are comparable between the GILD and ABBV regimens when RBV is not included; as expected, when RBV is included in the ABBV regimens adverse events are more frequent than comparable GILD regimens (which do not include RBV)

The obvious question is whether this degree of difference in adverse events is sufficient to preclude, or even significantly limit, adoption of the ABBV regimen – we believe it is not, for two main reasons. First, the overall rate of adverse-event related discontinuations on the ABBV regimens is extremely low – much lower than would be expected if the ABBV regimens were truly difficult to tolerate. Second, in this same (HCV) market we’ve already seen a substantially less tolerable regimen (PegIntron) compete effectively against a more tolerable (and more effective) alternative (Pegasys). Exhibit 4 summarizes rates of adverse events from a large head-to-head trial sponsored by the maker of PegIntron; bolded line items indicate adverse events occurring at least 1.2x more frequently with PegIntron than with Pegasys

PEG v. PEG, and what this implies about ABBV v. GILD

As PegIntron and Pegasys were heading into late clinical development, the two drugs were viewed as highly similar, with a perceived efficacy advantage going to Pegasys because of its larger volume of distribution[6] and higher trough concentration[7] levels. After late stage trials were completed Pegasys’ reputation of potentially greater efficacy was joined by a general perception that Pegasys also was more tolerable than PegIntron; at that time PegIntron’s greater incidence of significant injection site reactions was a particular area of concern. As both drugs moved into the market, the sense that PegIntron was less tolerable grew, presumably from practical clinical experience, which resulted in Pegasys becoming the preferred agent. Pegasys’ efficacy advantage remained largely theoretical until Schering-Plough (the maker of PegIntron) sponsored the head-to-head IDEAL study, whose 2001 results showed substantially higher early and end-of-treatment response rates for Pegasys, though longer-term sustained response rates (SVR12) were similar between Pegasys and PegIntron. Adverse events in the IDEAL trial were generally higher for PegIntron (Exhibit 4, again). Our sense is that IDEAL amplified the impression that PegIntron was less tolerable, and was seen as supportive but not conclusive evidence that PegIntron was also less effective

In 2007 a review[8] of 5,944 Veterans’ Administration patient records found that patients receiving Pegasys were significantly more likely to achieve a sustained response than patients on PegIntron. This report seems to have had a major effect on relative global sales of the two agents; PegIntron sales began a slow decline in 2007 while Pegasys sales continued to grow (Exhibit 5). In 2009, a meta-analysis[9] of 8 clinical trials (4293 patients) found a higher response rate for Pegasys than for PegIntron (relative rate of 1.1:1). Finally in 2010, two independent investigators[10] compared Pegasys and PegIntron head-to-head, and both found higher sustained response rates for Pegasys (66% v. 54% in Rumi et al.; 68.8% v. 54.4% in Ascione et al). After the 2007 VA findings Pegasys was firmly established as the preferred agent, and the relative sales trajectories of PegIntron and Pegasys appear not to have reflected any added impact of the 2009 and 2010 findings (Exhibit 5, again)

The point of this comparative history is to show that despite coming out of the gate with a perceived efficacy disadvantage and a very real tolerability disadvantage, PegIntron maintained a 40% share of global sales until the publication of the VA findings in 2007, after which PegIntron’s share has gradually declined to its current level of roughly 28%. Across the lifetime thus far of the two pegylated interferons, PegIntron has achieved a 34% share (Exhibit 6)

We believe the efficacy and tolerability of the ABBV 3D regimen is closer to the efficacy and tolerability of the GILD Sovaldi/ledipasvir regimen that PegIntron was (or is) to Pegasys. Accordingly, we believe the ABBV 3D regimen can achieve a share of global sales at least on par with that achieved by PegIntron, provided it is not displaced by other new entrants

Other new entrants

BMY has multiple HCV regimens pending approval, many with PDUFA dates of 12/7/2014[11]. Of these near term approvals, we do not believe any are competitive with the ABBV 3D regimens

BMY’s asunepravir / daclatasvir regimen has lower (<= 90% in type 1b) efficacy rates and longer duration (24wks) of therapy than either the ABBV or GILD regimens. BMY’s daclatasvir / Sovaldi regimen has competitive efficacy (approaching 100%) in treatment-naïve type 1 patients at 12 weeks of therapy, but is likely to be price ineffective[12] in type 1 patients. Instead, we expect daclatasvir / Sovaldi to be niched to non-type 1 patients, where it appears to offer an efficacy advantage

Longer-term, BMY has a triple-regimen (daclatasvir / asunepravir / BMS-791325) on track for a 1Q15 filing (and presumably an early 2016 launch). Based on phase II results this regimen may offer comparable (but not superior) efficacy to the non-RBV ABBV regimen (+/- 92% SVR) at 12 weeks of therapy in type 1 treatment-naïve patients. Like the ABBV 3D regimens, the BMY triple-regimen requires twice-daily dosing

It is unclear whether MRK’s MK5172/MK8742/ +/- RBV regimens are competitive in type 1. Trials for the non-RBV regimen have targeted completion dates in 3Q15, which implies a 2016 filing and 2017 launch. Interim results make it fairly clear that the duration of treatment for this regimen will be 12 weeks, as an 8 week trial arm show poor relative efficacy

Consensus sales expectations

Consensus calls for cumulative global sales of approximately $80B (through 2018) for current generation HCV treatments; as we’ve argued in a prior research note[13] we believe these global sales expectations are realistic. Where we believe consensus errs is in assigning 79% of global sales to GILD, and only 11% to ABBV (Exhibit 7). Of the 2 ABBV analysts with product level forecasts (at least on Bloomberg), only one (Jeffries) has a forecast for the ABBV HCV products – thus our statement that consensus allocates 11% of global HCV sales to ABBV may in fact overstate the HCV expectations that are priced into ABBV. We doubt that the ABBV analysts who are not offering product level detail have higher HCV expectations, as the Jeffries total-company sales forecast for ABBV exceeds the consensus total-company forecast (Exhibit 8)

If we could completely rule out any impact from a competitive BMY entry in 2016 (daclatasvir / asunepravir / BMS-791325), or from a competitive MRK entry in 2017 (MK5172 / MK8742) – which of course we cannot – we would expect the ABBV regimens to do at least as well vs. GILD as PegIntron did against Pegasys. As we’ve argued, despite PegIntron having greater disadvantages relative to Pegasys than the ABBV regimens have relative to GILD, PegIntron has claimed 34% of the global pegylated-IFN market to date, and still holds a 28% share of current global sales. Thus absent any impact from competitive entrants we believe ABBV would capture more than 30% of global sales

Our argument that ABBV should capture ‘at least 20 percent’ of global sales is intended simply to allow room for competitive entries. As yet there is no evidence the 2016 BMY or 2017 MRK regimens will be superior to the ABBV regimens in terms of efficacy, but we cannot rule out that either or both of these regimens may offer convenience / tolerability advantages; and, we accept that expanding the number of viable HCV treatment alternatives would inevitably result in lower net pricing

ENTA as a higher-beta vehicle for the ABBV HCV bull case

Abbott (now ABBV) licensed the ABT-450 component of the 3D regimen from ENTA in November of 2006; the redacted agreement between ABBV and ENTA can be found in the latter’s form S-1 filed November 6, 2012

Per the agreement ABBV will owe ENTA ‘tiered royalties ranging on a blended basis from the low double digits up to the high teens’

The agreement recognizes the possibility of ABT-450 being used in combination with other products, but is non-specific as to how ABT-450’s share of the combination’s net sales would be determined, instead calling for ‘a good faith determination of the relative value of each pharmaceutically active ingredient, based on the estimated fair market value of each such therapeutically or biologically active ingredient …’. At the very least, we believe ABT-450 would be allocated one-quarter of net sales (on the simple premise that there are four[14] active agents in the regimen). More likely, because ritonavir is in the combination for the purposes of improving the pharmacokinetics of other agents, rather than for the purpose of direct anti-HCV activity; and, because ritonavir (as Norvir) is currently sold at prices well below one-quarter of the likely cost of the 3D regimen, we believe ritonavir will be credited with less than one-quarter of the 3D regimen’s sales[15]. This argues for ABT-450 being allocated as much as 32.9% (see footnote 15) of the 3D regimen’s net sales

In a recent note[16] we estimated cumulative global sales of between $154B and $235B for the entire lifecycle of the current generation HCV products. We further argued, as we’ve re-iterated in more detail here, that ABBV’s 3D regimen could realistically be expected to capture 20% or more of these global sales

In Exhibit 9, we provide a pro forma estimate of how ABBV’s share of global HCV sales might flow through to ENTA. As a low case, we take our low case global sales estimate of $154B, assume the ABBV regimens capture just 15% of this market, then further assume that ABT-450 is credited with 30% of 3D regimen sales, and that ENTA earns a 12.5% royalty on the allocated ABT-450 net sales. These assumptions translate into a low case estimate of royalties owed to ENTA of $900M (across the +/- 10 year lifecycle of the current generation of HCV agents). As a high case, we take our high global sales estimate of $235B, estimate ABT-450 is credited with 32.9% of net sales, and further estimate that ENTA’s royalties equal 18% of ABT-450 net sales. These assumptions translate into a high case estimate of $3.5B in royalties owed to ENTA, again across the +/- 10 year lifecycle of current generation HCV treatments. ENTA’s current market capitalization is roughly $800M, including roughly $135M in net cash and marketable securities. Thus ENTA’s enterprise value of roughly $665M is substantially less than the $2.2B (not time or risk-adjusted) midpoint of our low- and high-case estimates of lifetime ABT-450 royalties

  1. ‘Prescription drug user fee action’ date, i.e. the date on which the Food and Drug Administration (FDA) is expected to act on the company’s filing
  2. Efficacy is reported as the rate of sustained virological response at 12 weeks following the end of treatment (SVR12). Sustained virological response is defined as a lack of detectable HCV RNA (virus-specific ribonucleic acid, which indicates active viral replication)
  3. ABT-450 with ritonavir (ABT-450r), plus ombitasvir, plus dasabuvir; with or without ribavirin (RBV)
  4. Cirrhosis is defined as the presence of liver degradation as evidenced by adverse histologic (microscopically viewed tissue sample) changes
  5. ABT-450r and ombitasvir are co-formulated in a single pill taken once daily, dasabuvir is given twice daily
  6. Meaning Pegasys distributes more evenly and completely within the body than PegIntron
  7. Meaning the lowest concentration of Pegasys reached between doses is higher than the lowest concentration of PegIntron reached between doses
  8. Backus LI, et al., “Predictors of response of US veterans to treatment for the hepatitis C virus” Hepatology 2007;46:37-47
  9. Awad T, et al., “Peginterferon alpha-2a may achieve a higher sustained virological response than peginterferon alpha-2b in chronic hepatitis C: a Cochrane systematic review of randomized clinical trials” Hepatology 2009;50(suppl):707-708a
  10. Rumi M, et al., “Randomized study of peginterferon alfa-2a plus ribavirin versus peginterferon alfa-2b plus ribavirin in chronic hepatitis C” Gastroenterology 2010;138:108-115; also

    Ascione A et al., “Peginterferon alpha-2a plus ribavirin is more effective than peginterferon alpha-2b plus ribavirin for treating chronic hepatitis C virus infection” Gastroenterology 2010;138:116-122

  11. Daclatasvir was approved in the EU on August 26th, 2014
  12. The ABBV regimens are unlikely to price above the list price of GILD’s Sovaldi/ledipasvir combo, and in turn GILD is unlikely to price Sovaldi/ledipasvir much if any higher than the current price of a Sovaldi regimen. This means a Sovaldi combination regimen – such as BMY’s daclatasvir / Sovaldi – leaves no room for the other (in this case daclatasvir) agent to have any price at all
  13. “GILD v. ABBV …”, SSR Health LLC, August 7, 2014
  14. ABT-450, ritonavir, ombitasvir, and dasabuvir
  15. Norvir’s current price (WAC, or wholesale acquisition cost) is $257.17 for 30 100mg capsules. The 3D regimen includes 100mg of ritonavir daily for 12 weeks, thus if the value of ritonavir in the 3D regimen is calculated in terms of its known ‘street’ price, the value would be 3 * $257.17 or $771.51. If we further assume the ABBV 3D regimen sells for 25% less than the current cost ($84,000) of the GILD regimen, we can estimate this leaves (0.75 * $84,000) – $771.51 = $62,228.49 in net sales to be allocated across ABT-450, ombitasvir, and dasabuvir. If we further assume each of these direct-acting agents gets a third of this post-ritonavir remainder, we can assume ABT-450 would be credited with approximately 0.33 * $62,228.49 = $20,722.09, which represents 32.9% of the total estimated regimen price
  16. Ibid 13
Print Friendly, PDF & Email