February 20, 2020 - 9:48am EST by
2020 2021
Price: 9.32 EPS 0 0
Shares Out. (in M): 34 P/E 0 0
Market Cap (in $M): 316 P/FCF 0 0
Net Debt (in $M): 150 EBIT 0 0
TEV ($): 466 TEV/EBIT 0 0
Borrow Cost: General Collateral

Sign up for free guest access to view investment idea with a 45 days delay.



AMAG short. 

AMAG Pharmaceuticals, Inc (“AMAG”) is a levered specialty pharma company that has suffered setbacks in core assets, and we believe there is material downside from current levels. A year ago, AMAG had two strongly performing, cash generating products—Makena, a preterm birth prevention drug, and Feraheme, one of only two high-dose IV iron products on the U.S. market. Makena was battling generics, but AMAG had developed an autoinjector formulation to preserve the franchise. Feraheme was viewed by the market as a compounding growth story. Feraheme sales were $168mm in 2019 and the Makena autoinjector was at a ~$160mm annual run-rate as of Q3 ’19—we believe both products have ~85-90% gross margins. Additionally, AMAG was about to launch a product it was extremely bullish on, Vyleesi, a self-administered injection for low sexual desire in women. 

Since then, Makena and Vyleesi have collapsed to the point where AMAG is discontinuing its women’s health salesforce, putting Makena into run-off and abandoning Vyleesi. This was necessitated by Makena failing to show that it actually delayed pre-term births in the long-term confirmatory study that was required for its original approval, which led an FDA Advisory Committee (Adcom) to recommend its withdrawal from the market. The FDA has yet to act, but, as we await their decision, generics continue to erode branded Makena sales, and we expect that to accelerate as AMAG discontinues promotion.  In addition to Makena’s problems, the Vyleesi launch has been a failure, one that would seem predictable given the characteristics of the drug: an injection required to be self-administered ahead of an expected sexual encounter that causes nausea as its main side effect and has ho-hum efficacy. AMAG has implicitly acknowledged the launch failure and has announced plans to sell Vyleesi from a position of weakness. AMAG also announced plans to sell Intrarosa, a languishing vaginal insert for which AMAG once targeted hundreds of millions in sales.

Feraheme, the supposed crown jewel of the AMAG portfolio, faces several competitive threats that, in our opinion, AMAG prefers not to talk about. Per a settlement, Sandoz can launch a generic version of Feraheme in July 2021. Even if it can’t get its own product approved, in July 2022, Sandoz can launch an AMAG-supplied authorized generic for a year, interrupting AMAG’s exclusivity. The last patent for Feraheme expires in 2023, inviting additional generics. A more imminent and ominous competitive threat has surfaced with the January 2020 FDA approval of Monoferric, a third—and clearly superior—high dose IV iron. This product has a long ex.-U.S. track record and is the only high dose IV iron that requires a single dose (both Feraheme and market-leader Injectafer require two doses days apart). We believe Monoferric will have a disruptive impact on its competitors and severely diminish Feraheme's value.

The company's pipeline has also faced setbacks and we believe the assets under development carry significant clinical risk. Further, there's a $320mm convert due in 2022 (far out of the money) and recent developments cannot be good for refinancing efforts. We believe that activist Caligan Capital’s successful campaign for Board representation has created some enthusiasm for the stock, but see their actions to date, such as firing CEO William Heiden and pushing successfully for the wind down of women’s health, as less about value creation and more about averting future value destruction. We believe $5/share is conservative fair value. 

Company Overview

Below is an overview of each component of the story.


Hydroxyprogesterone caproate (HPC), a synthetic hormone, has been available from compounding pharmacies for decades. The primary use case is prevention of preterm birth (PTB), despite no established mechanism of action. Studies over the years had conflicting findings regarding HPC’s efficacy, prompting a 2003 clinical trial called Meis. Meis showed a treatment effect but several factors limited its generalizability and it wasn’t designed to support a drug approval. 

In a process involving several sponsors and multiple review cycles, a 2006 NDA (relying on the Meis data) to make HPC an FDA-approved drug was granted accelerated approved in 2011. Accelerated approval came with 7 years of market exclusivity (only obstacle to generics); however, it also required successful completion of a confirmatory trial, PROLONG. PROLONG was deemed necessary due to the shortcomings of Meis. Makena was owned by KV Health at the time of approval—AMAG acquired the product in 2014.

Prior to Makena, HPC was available from compounding pharmacies for $15 per injection. It was initially marketed at 100x this price. The FDA bars compounded versions of approved drugs; however, they turned a blind eye to compounding of HPC due to the price gouging. By the time AMAG acquired the drug, Makena was at a $180mm annual sales run rate. Concerns over the safety of compounded drugs emerged in 2012, allowing Makena to gradually take share. By 2017, prior to generic entry, Makena hit a peak of ~$390mm of revenues.

2018 was the beginning of the end for Makena. The first HPC generics arrived, starting to dent sales. AMAG responded by introducing an autoinjector form of the drug (smaller needle, convenient administration). Through a patient access / HCP support program, AMAG clung to a meaningful slice of the market and was optimistic about the autoinjector’s durability.

In March 2019, topline results were announced from PROLONG. The results showed a clear lack of efficacy, including in all subgroups. The company’s narrative is that the Eastern European population dominating PROLONG had an inherently low risk for PTB compared to the largely (60%) African American population studied in Meis. Despite no treatment effect in the Black / African American subgroup in PROLONG, AMAG contends that withdrawal of Makena will disadvantage such high-risk populations. 

At an FDA Adcom meeting in October 2019, the panel voted 9-7 to withdraw Makena. The drug’s fate lies in the hands of the FDA commissioner. The Adcom vote is already taking a toll—Q4 2019 witnessed a sharp decline in Makena sales. Even if Makena remains on the market, there are repercussions short of withdrawal that should continue eroding sales—payers could stop covering it and doctors could stop prescribing it as they digest the compelling evidence that it has no effect. Management is “proactively managing product to be cash-flow positive” in 2020 and has a rather muted sales outlook.  


AMAG’s first approved products were medical imaging agents employing iron nanoparticles. Ferumoxytol, the compound now known as Feraheme, was developed as both an IV iron and an MRI contrast agent. Feraheme was approved in 2009 for the treatment of iron deficiency anemia (IDA) in patients with chronic kidney disease (CKD). Feraheme’s advantage over existing IV irons was the high dose, allowing for full iron replenishment in two quick visits. Lower-dose forms required up to 10 longer visits over the course of a month.

In 2013, a second high-dose IV iron, Injectafer, was approved in the U.S. Unlike Feraheme, Injectafer was labeled for use in all patients with IDA (not only CKD). AMAG was pursuing this label expansion; however, there were increasing concerns over the product’s safety. Feraheme was initially approved as a quick, undiluted injection. This led to many serious allergic reactions and deaths. In 2015, the FDA issued a Boxed Warning requiring a slower IV infusion followed by 30 minutes of monitoring. A combination of Injectafer competition, narrower label and safety profile limited Feraheme’s growth—it did ~$60mm in sales during 2010, its first full year, and was at ~$105mm in 2017.

In early 2018, Feraheme finally got the label expansion, doubling its addressable market. 2018 and 2019 saw strong sales growth, partly volume driven but augmented by steady price increases. Feraheme and Injectafer have also benefited from the Medicare Part B model rewarding use of higher priced drugs (doctors reimbursed based on ASP + % spread).

2019 Feraheme sales were ~$168mm, an impressive 2-year expansion. We believe this is the product’s peak. AMAG thinks the product is too hard to copy; however, most analysts assume generic entry and erosion of the franchise post-2023. We think this disruption could come sooner if Sandoz gets approval and at a minimum, the Sandoz settlement guarantees an exclusivity interruption in the form of an authorized generic. Feraheme’s perceived value relies on robust performance ahead of this cliff.

Unfortunately for AMAG, generics no longer present the only threat. Last month, the FDA approved a third high-dose IV iron product, Monoferric. Monoferric has been sold in other countries for years and is the first product approved in the U.S. to replenish iron levels in one infusion. As validation of the competitive threat, Injectafer’s U.S. owner has sued the maker of Monoferric for patent infringement. We are monitoring this litigation; however, we believe it will be difficult to keep this superior treatment from patients. Whenever it is finally marketed, Monoferric’s dosing advantage should help it steal substantial share from Injectafer and Feraheme. 

Women’s Health – Assets to Be Divested

Vyleesi is a treatment for HSDD, or low sexual desire in premenopausal women. It is meant to be injected prior to an encounter and over 40% of users develop nausea upon use. Coupled with weak data and the only comp, Addyi, failing spectacularly, we’ve always been skeptical of Vyleesi. AMAG was in the full throes of a launch; however, we believe their sales and marketing efforts failed to drive enough early traction to justify the spend.

Intrarosa is a vaginal insert intended to treat painful intercourse accompanying menopause. Despite some analysts projecting peak sales as high as $300-400mm, Intrarosa did $16mm in 2018 and $21mm in 2019. Prior to announcing the divestitures, AMAG had already deemphasized Intrarosa, making Vyleesi the primary focus of the shared salesforce.


Overall, we believe AMAG has a weak pipeline with limited upside. 

Ciraparantag is a reversal agent for new blood thinners like Xarelto and Eliquis. The use case is in patients on these drugs that develop bleeding complications or need emergency surgery. AMAG bought the asset from Perosphere in 2018 for ~$70mm upfront, $140mm of regulatory milestones and $225mm of sales-based milestones. Until recently, Daiichi Sankyo was a collaborator and was to fund part of the Phase III program. Daiichi has since bailed on the agreement, paying AMAG a $16mm termination fee. AMAG is yet to advance Ciraparantag beyond the Phase IIb completed by Perosphere, a healthy volunteer study that used a manual, potentially subjective blood clotting assay as its primary endpoint. A more typical reagent-based assay could not be used, since Ciraparantag binds to and interferes with the reagents. FDA requested an automated assay and Perosphere developed a “coagulometer” device. For the clinical program to move forward, this device needs to get an IDE and Phase IIb needs to be repeated. We think there may be issues with the coagulometer. Failure to obtain an IDE has already delayed the program by over a year—a reliable assay is a gating item.

AMAG-423 is AMAG’s candidate for severe preeclampsia, a pregnancy complication causing hypertension and swelling with maternal and neonatal implications. The molecule is digoxin immune fab (DIF) and the last trial, a Phase II proof-of-concept study, was completed in 2010. The data from this trial showed an improvement in sodium pump activity (proposed mechanism of action); however, no significant difference was observed in neonatal outcomes. 

In June 2015, AMAG entered into an option agreement to acquire the drug from Velo Bio upon the completion of a (still ongoing) 200 patient Phase IIb/IIIa study. Enrollment is extremely challenging—you need an expectant mother to present with severe preeclampsia at a study site. The first patient was enrolled by Velo in June 2017. In September 2018, AMAG re-cut and exercised its option, bringing AMAG-423 in-house and taking over responsibility for the trials (only 26 patients were enrolled at this point). It paid $12.5mm upfront, with a $35mm approval milestone and several sales milestones. When the latest trial started, guidance was for NDA approval by 2020. AMAG is not providing enrollment updates but recently suggested at least two more years of substantial program-related spend. We believe the drug won’t be approved until 2023, if at all. All patents expire in November 2022, so AMAG will only have the 7 years of exclusivity accompanying the orphan designation.     


Our valuation uses a sum-of-the parts approach. For Feraheme, we model a flat 2020 as Monoferric launches, and a 20% decline in 2021 as it ramps. Starting in 2022, we model more meaningful declines as generics enter the mix and patents expire. On a DCF basis, we estimate Feraheme is worth ~$275 million with downside possible on a stronger Monoferric launch.

For Makena, AMAG’s guidance implies $35-85mm of revenues in 2020, assuming Consensus for Feraheme (guidance issued prior to Monoferric’s approval). In a midpoint “keep (Makena)” scenario, we model $60mm in 2020 revenues, declining to $30mm by 2024 on decreased usage of HPC, lower pricing and lower AMAG share. The DCF value of Makena with this trajectory is ~$30mm, given the need for ongoing spend to drive autoinjector prescriptions. This is our base case—sales / cash flows could be better in some “keep” scenarios; however, product could also be withdrawn outright.

Anyone taking over Vyleesi and Intrarosa will need to spend heavily on marketing and won’t pay much upfront. Further, AMAG is a licensee of both products, owing royalties and sales milestones to the respective licensors. A new licensee will inherit these arrangements and it’s unclear how much tail value AMAG can capture. To date, AMAG has paid ~$225mm for these products ($60mm upfront for Vyleesi + $80mm in regulatory milestones and $70mm + ~$13.5mm in stock for Intrarosa). Given how far both products are from initial estimates of commercial potential, we think it’s unlikely AMAG recoups more than 1/3 of what it paid for Vyleesi, with Intrarosa coming with it as a cheap option—upfront value of ~$50mm. Net of an estimated ~$25mm of ongoing expense borne by AMAG prior to divestiture, we ascribe $25mm of value to Vyleesi and Intrarosa. 

As for the pipeline, we estimate the NPV of R&D and non-Feraheme/Makena overhead to be negative $300mm through 2023, our estimate of when Ciraparantag and AMAG-423 could be approved. For Ciraparantag, $100mm is owed to Perosphere upon FDA approval and for AMAG-423, $35mm is owed to Velo Bio. This implies ~$400mm of investment (on NPV basis) to get both products approved. This is prior to any marketing spend on their launches. Given the uncertain future of each program, we think PoS-adjusted commercial cash flows are unlikely to exceed the required investment. Our base case assumes the pipeline is breakeven on an NPV-basis. 

In summary:




  • Makena not pulled AND sales stabilize

  • Someone sees value in Vyleesi or Intrarosa AND pays upfront

  • Monoferric does not launch

  • Pipeline assets succeed

I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise hold a material investment in the issuer's securities.




  • Final FDA decision to withdraw Makena from the market—could come any day

  • Increased investor awareness of Monoferric threat and a successful launch of the product

  • 2020 guidance revised to account for Monoferric impact

  • Sandoz ANDA approval for generic Feraheme

  • Poor divestiture terms announced for Vyleesi and Intrarosa—little to no upfront consideration

  • Further pipeline timeline slippage or bad readouts

  • Equity raise to address debt

-1       show   sort by    
      Back to top