Atossa Genetics — Update 30 November 2016

Atossa completed a $2.875m equity raise (1.15m shares sold at $2.50 per share) in early September 2016, increasing its outstanding shares to 3.78m shares. We expect the proceeds to be allocated to Atossa’s two active pipeline programs: its 30-pt Phase II study on its proprietary IDMC-delivering fulvestrant to breast ducts in patients scheduled for mastectomy or lumpectomy, and its oral endoxifen candidate, presently in pre-IND (investigational new drug) development stage as a potential treatment for breast cancer patients refractory to tamoxifen. We expect Atossa to report initial data from the IDMC-fulvestrant study by Q117, and the firm believes it could start human endoxifen studies in 2017.

Atossa’s IDMC intends to deliver therapeutics for breast cancer and/or precancerous conditions, with potentially higher local exposure and lower systemic exposure vs established therapies or delivery approaches. Atossa’s internal analysis suggests that the drug levels in breast tissue might be more than 20,000 times more concentrated with IDMC versus systemic administration.

The current IDMC clinical program is designed to irrigate and deliver fulvestrant (marketed as Faslodex by AstraZeneca) to each of the five to seven breast ducts. Fulvestrant is FDA-approved for estrogen receptor-positive (ER+) metastatic breast cancer (with $704m in global 2015 sales) and is normally administered by intramuscular (IM) injection (to the buttocks), usually consisting of a monthly dose of two injections (costing $10,000-14,000 a month in the US).

The ongoing open-label Phase II study, taking place at Columbia University, is comparing the safety, tolerability and pharmacokinetics following the IDMC instillation of fulvestrant, compared to IM administration, in a neoadjuvant setting1 in women scheduled for mastectomy or lumpectomy (with either breast cancer or ductal carcinoma in situ, or DCIS). Six patients will receive 500mg IM fulvestrant and 24 patients will receive up to 500mg of IDMC-fulvestrant. The study will also measure changes in the expression of Ki-672 as well as estrogen and progesterone receptors, between biopsies taken prior to fulvestrant therapy, and post-treatment surgical specimen. Breast imaging before and after drug administration in both groups will also examine any potential treatment effects on the lesions or on breast density.

Management expects to present the study at the CTRC-AARC San Antonio Breast Cancer Symposium to be held on 6-10 December 2016. It is unclear whether interim study data will be provided at this symposium but, if not, we anticipate interim clinical data will be provided by the end of Q117. We expect a larger pivotal study would be required for approval under the FDA 505(b)2 process. Under 505(b)(2), Atossa may be able to rely on much of the existing safety, efficacy and preclinical data already established on fulvestrant. If the 505(b)2 pathway is accepted by the FDA (as we forecast in our model), Atossa may only be required to complete pharmacokinetic equivalency studies (the current trial) and possibly only a relatively small Phase III study to gain regulatory approval (ie smaller in size and scope than conventional 505(b)1 new drug application pivotal trials). Atossa believes that IDMC administration could provide more effective targeting, potentially resulting in lower doses to achieve therapeutic efficacy and/or a less frequent dosing schedule, leading to improved convenience and potential cost savings for patients and/or insurers.

We assume that after the current IDMC-fulvestrant trial, a larger (200- to 400-patient) pivotal study will be needed before approval. We expect the company to partner the IDMC-fulvestrant program with an oncology-experienced medical devices and/or pharma firm before or in parallel to starting this pivotal study, with Atossa entitled to 20% royalties on net IDMC sales.

We expect the start of pivotal studies in late 2017 or early 2018, leading to potential product launch in 2020. This differs somewhat from management’s guidance, at it anticipates that the FDA may require fewer clinical data to approve IDMC-fulvestrant, and that the product could be launched as early as 2018. If Atossa is successful in obtaining IDMC-fulvestrant approval earlier than our estimates, there could be upside to our forecasts.

We assume that IDMC-fulvestrant will be used in the neoadjuvant setting in the treatment of DCIS and ER+ breast cancers. The American Cancer Society estimates that about 246,660 new cases of invasive breast cancer will be diagnosed in women per year and about 61,000 new cases of carcinoma in situ (CIS) will be diagnosed (CIS is non-invasive and is the earliest form of breast cancer). Approximately 75-80% of breast cancers are ER+3 (ie they grow in response to estrogen).

Our model anticipates that IDMC-fulvestrant will be used in a peak case of 25% of neoadjuvant treatment scenarios (reflecting only up to 40% of diagnosed ER+ breast cancers, primarily those at Stage II and III). We continue to estimate that the IDMC single-use device will be sold at launch at $3,500 per monthly application, with expected peak IDMC-fulvestrant product sales (consisting of the IDMC device and separate from the cost of the fulvestrant drug) of $182.8m in 2025 (leading to royalties to Atossa of $36.6m). We expect commercialization through 2030, when the IDMC technology’s core patents expire. While the IDMC device could be extended to other breast cancer drugs, we are currently only modeling the fulvestrant application.

In June 2016, Atossa announced that it was developing endoxifen as a treatment for breast cancer patients refractory to tamoxifen. Following surgical treatment for atypical hyperplasia (AH) or non-invasive ER+ breast cancers (such as DCIS), additional treatment with a selective estrogen receptor modulation (SERM) drug such as tamoxifen or raloxifene (Evista) is often recommended. A large-scale randomized study (IBIS-I) found that tamoxifen reduced breast cancer incidence in high-risk women by 30-50% over five years of treatment, for ER+ cancer and DCIS. Recent studies have shown that extended use of tamoxifen (10 years versus five years) further reduces recurrence risk and mortality, so US clinical practice guidelines now recommend consideration of adjuvant tamoxifen therapy for 10 years.4 5 6

Tamoxifen is effective in both pre-menopausal and post-menopausal cancers, and raloxifene, a newer SERM drug with a more advantageous AE profile vs tamoxifen,7 is only approved for use in post-menopausal women. Aromatase inhibitors, discussed in our initiation report, are another chemopreventative therapeutic option for post-menopausal women. However, given their well-documented long history of use and relatively inexpensive cost (ZenRx estimates that the wholesale cost of a daily 20mg tablet of generic oral tamoxifen can be as little as $0.15), tamoxifen remains the mainstay preventative or post-surgical treatment for ER+ breast cancers (or related conditions such as AH), particularly in pre-menopausal women.

Orally dosed tamoxifen is metabolized in the liver by enzymes (including cytochrome P450 isoforms) into multiple metabolites, yet only a few of these metabolites have an active ER antagonist effect (blocking estrogen from binding to its receptors). The most significant of these (in terms of ER antagonism contribution and plasma concentration in patients with normal tamoxifen metabolism) are endoxifen (4-hydroxy-N-desmethyltamoxifen) and, to a lesser extent, 4-hydroxytamoxifen;8 9 both have at least 100-fold higher ER binding and cancer cell proliferation suppression than the parent compound.10 11

Several researchers found that patients with a deficiency (due to genetic or other factors) in cytochrome P450 enzyme CYP2D6 have an impaired ability to metabolize tamoxifen into endoxifen, and that up to 15-20% of Europeans carry genetic CYP2D6 variants associated with an impairment in forming anti-estrogenic tamoxifen metabolites.9 Variations in other enzymes, including CYP3A, and several other factors (including which other systemic medications are taken) can also influence endoxifen levels12. Fox et al. found that that over 50% of tamoxifen-dosed patients with low endoxifen have no obvious cause.13

Fox et al. found that in 122 patients taking 20mg/day of tamoxifen (the standard dose), 24% had blood endoxifen levels of below 15 nmol/L, and suggests that 15nmol/L may be the critical level needed or anticancer effect. Several groups found that those with impaired tamoxifen metabolism (or low endoxifen) may have higher risk of cancer recurrence compared to those with normal metabolism; Madlensky et al. found that in 1,370 ER+ breast cancer patients taking tamoxifen, those in the lowest quintile for endoxifen concentration had a 35% increased recurrence rate than remaining patients.14 Saladores et al. found in a study of 587 pre-menopausal women taking tamoxifen that low plasma endoxifen was associated with shorter distant relapse-free survival and a 1.6-1.9 increased hazards ratio for cancer recurrence.15

These studies form a basis for dosing endoxifen directly in such patients. Atossa has filed composition of matter and methods of treatment patent applications (with patent lives potentially into 2036) for its endoxifen formulation, secured manufacturing with a Taiwan-based manufacturer (KriSan Biotech) and is working to file an IND to start human (Phase II) studies in 2017.

The planned Phase II study will be in pre-menopausal women with ER+ breast cancer in patients already taking tamoxifen (20mg/day). A companion diagnostic test will be used to measure plasma endoxifen levels in the blood. Patients below a pre-specified threshold of endoxifen will take the company’s endoxifen formulation (1-2mg/day) and those above that threshold will continue on tamoxifen (20mg/day). The study will compare blood endoxifen levels between both groups, as well as assess pharmacokinetics (PK) and safety.

As tamoxifen has a long-established history of systemic use, and as endoxifen is a metabolite of this drug (and with a similar “active moiety”16), Atossa believes it could be eligible for the 505(b)2 registration pathway. FDA draft guidance for 505(b)2 class applications indicates that product candidates which are active metabolites of existing approved drugs can in some cases be covered under this registration pathway.17 Under this scenario, a clinical trial demonstrating endoxifen’s efficacy may not be necessary for approval, and regulators may only require safety and PK studies.

While no details have been provided as to the study length and recruitment size of the currently planned endoxifen Phase II trial, we currently assume that the FDA will allow the program to be assessed under 505(b)2 criteria, which we believe will require an additional trial for approval. We estimate that the currently planned study will finish in H118, with the registration-enabling study to start in mid-2018 and lead to approval in 2020. Rather than commercialize endoxifen itself, we believe Atossa will seek to partner its endoxifen formulation with a pharma company in 2018 (prior to starting the pivotal trial), and will be entitled to 20% in net royalties.

Under 505(b)2, we do not believe that efficacy data showing improved survival rates (or lower recurrence rates) vs tamoxifen would be needed for approval. However, such data could strengthen the clinical case for physicians to select endoxifen vs tamoxifen in clinical settings, so we anticipate that Atossa’s commercial partner will eventually conduct a head-to-head, post-marketing trial vs tamoxifen.

A team of investigators at Mayo Clinic (Goetz M, Ames M, and collaborators) and the National Cancer Institute (NCI) reported in late 2013 that in a 22-patient Phase I study in women with ER+ breast cancer, a formulation of Z-endoxifen hydrochloride was safe (at doses up to 160mg per day) and showed early evidence of anti-tumor activity. The same group is currently conducting a 94-patient randomized Phase II trial comparing its endoxifen formulation with tamoxifen in treating patients with ER+ breast cancer (but negative for HER receptors) that has spread to nearby tissue or lymph nodes or other parts of the body (data, including progression-free survival, expected in mid-2017). The group is also conducting an ongoing 62-pt, open-label, dose-ranging Phase I study (seeking to find a maximum tolerated dose, with data expected in H117). The NCI is also conducting another dose-ranging Phase I study on this formulation (n=72; data expected in Q416).

While Atossa is filing patents for its own endoxifen formulation and methods of treatment, there is a material risk that competing studies from the Mayo/NCI investigators, should they lead to registration or commercialization-stage end products, could lead to intellectual property (IP) related competition challenges to Atossa’s eventual endoxifen product.

Approximately 300,000 US women take tamoxifen18 (for either cancer prevention or an adjunctive treatment to prevent recurrence) and approximately one million women take the drug worldwide. Madlensky et al.12 estimate that 20% of women currently taking tamoxifen do not achieve sufficient endoxifen concentrations, which is comparable with Fox et al.’s findings shown above. We assume that this (20%) reflects the potential target market for oral endoxifen (thus 60,000 persons in the US), and that peak market share for Atossa’s product would be 50% of this group, which would be attained within five years of launch (2025). Using the current price of branded oral SERM drug raloxifene (also used in the breast cancer prevention market) as a guide, we model a starting net price of $200/month for the drug on launch (in 2020). Given these assumptions, we anticipate peak sales in 2025 of $91m and $161m in the US and worldwide, respectively which, at our 20% assumed royalty rate, leads to net royalties of $32.1m to Atossa in 2025.

We highlight that there is the potential for some variability in our market size estimates. A study19 on 279 Polish women taking tamoxifen found that nearly 60% of these (including over 30% of patients with fully functional CYP2D6 status) had endoxifen concentrations below the predefined threshold of therapeutic efficacy. If convincing clinical data can be developed for physicians, patients and stakeholders on the potential benefits of oral endoxifen (vs tamoxifen) to a wider range of the current tamoxifen treatment population than we currently assume (20%), then there can be upside to our peak sales estimates.

Atossa reported Q316 results on 14 November 2016, with net income of $0.2m ($0.07 per share given 2.8m average shares outstanding in Q316); the results were buoyed by the one-time $1.76m payment (recorded as other income) from Besins Healthcare. This payment resulted from the company’s settlement agreement, under which Atossa relinquished its rights to the Afimoxifene Topical Gel (AfTG) program. Atossa acquired exclusive rights for AfTG from Besins Healthcare in 2015 to develop and commercialize it for breast hyperplasia but, in February 2016, entered into a legal dispute with Besins given Besins’ own AfTG development plans. Both parties agreed in August 2016 to terminate the original licence agreement and settle all claims and counterclaims, leading to the one-time $1.76m payment to Atossa. We had previously valued (in our 27 May 2016 update note) the discounted AfTG revenue opportunity to Atossa at $15.5m.

Excluding this one-time payment, Atossa reported an adjusted $1.56m loss ($0.56 per share). Q316 R&D costs were $0.085m and 9M16 R&D costs were $0.4m. We expect R&D costs to rise in coming quarters as recruitment for the ongoing fulvestrant study progresses and as the endoxifen program starts in-human studies.

As stated above, in addition to the $2.875m equity raise, Atossa received $1.76m from Besins during Q316. Atossa had a 9M16 cash burn rate (operating cash flow including net interest, minus net capex) of $4.0m. It had $4.4m net cash on 30 September 2016, and we assume the burn rate will increase slightly in coming quarters following increasing R&D costs. We estimate that Atossa’s burn rate in Q416 would be $2.1m. We believe the company will need to raise funds again in H117 to fund its operations. The company entered a $10m equity line of credit with Aspire Capital in May 2016, which could help satisfy the funding requirement. Our model assumes a 2017 burn rate of $13.3m, and that Atossa will raise $20m in funding in 2017. As per our usual policy, for modelling purposes, we assign these financings to long-term debt.

Our rNPV valuation now includes the prospects of the company’s oral endoxifen program, in addition to the IDMC program (previously also included in our model) and the AfTG opportunity has been removed (as AfTG rights were transferred back to Besins). We assume that Atossa will spend $4.6m in R&D on the IDMC-fulvestrant program in 2017 before partnering it, and that it will spend $6.5m on oral endoxifen R&D across 2017 and H118 (prior to entering a licence agreement on this program). We assign a 12.5% probability of success to the oral endoxifen program, which takes into account product development risk, as well as the IP risk and competitive risk from the Mayo/NCI investigations and trials for their own endoxifen formulation. We continue to apply a 25% probability for the IDMC-fulvestrant program. Given the addition of the endoxifen program, the removal of AfTG and rolling forward our forecasts, we now obtain an rNPV of $6.2m (up from $2.3m previously). The inclusion of the endoxifen program and rolling forward our forecasts leads to an increase in our rNPV. After including Q316 net cash of $4.4m, we obtain an equity valuation of $10.7m, or $2.82 per share.