ROVI Laboratorios Farmaceuticos — DORIA low risk, high reward

ROVI, a profitable, Spanish speciality pharmaceutical company, is engaged in the research and development, as well as the manufacturing and marketing of a broad range of small molecule and speciality biologic products primarily in its domestic market (Spain accounted for 71% of net revenues in 2017). Internally developed biosimilar enoxaparin is the first to launch in the European market (Germany and the UK) and is a key driver of near-term revenue and profit. Sustainable long-term growth is dependent on successful R&D investments. ROVI’s efforts here lie in investing in its ISM technology assets: Risperidone ISM (DORIA), a long-acting, injectable, atypical antipsychotic therapy for the treatment of schizophrenia, and Letrozole ISM, a long-acting aromatase inhibitor for hormone-dependent breast cancer. This note focuses on DORIA’s opportunity in schizophrenia. DORIA is a high gross margin asset (85-95%), and will be the critical long-term driver of operating margins for the business. For details of ROVI’s marketed products, see our initiation note, Ace of Spain.

We value ROVI at €1.16bn or €23.3 per share, on a sum-of-the-parts basis, based on a three-stage DCF including our forecasts to 2025 (10% discount rate, long-term tax rate of 15%, 2.0% terminal growth rate) for the core business (excluding DORIA), and risk-adjusted NPV for DORIA US and the EU opportunity. Our DCF model consists of ROVI’s base business (product sales and toll manufacturing revenues), including our forecasts for biosimilar enoxaparin. We now separately value Phase III asset, Risperidone ISM (DORIA), which we include on a risk-adjusted basis. Compared to ROVI’s current portfolio of drugs and footprint, the US opportunity for DORIA is large and a key value driver (accounting for 16% of our valuation; EU DORIA accounts for 9%).

ROVI is subject to various sensitivities common to speciality pharmaceutical companies, including commercialisation (pricing, reimbursement, uptake and competition), manufacturing and financing risks. The key sensitivities for ROVI relate to successful European commercialisation of both Hibor and biosimilar enoxaparin (two years after launch [2019] enoxaparin represents 13% of our total revenue forecasts), while crystallising value from its R&D pipeline will prove critical in the longer term. Ongoing in-licensing deals are required to renew the portfolio offering in Spain and replace some of the mature products, which are facing patent expiry over the next few years. We do not value the technology platform, early-stage ISM (in situ micro particle implants) R&D pipeline and any future collaborations, all of which could represent upside. ROVI is a majority family-owned business; the principal shareholder (Norbel Inversiones) holds 69.64% of the business, thus the limited free float has an impact on liquidity.

In 2017 ROVI reported total revenues of €275.6m (+4% y-o-y) and EBITDA of €30.5m (-22%), leading to a lower EBITDA margin of 11.1% vs 14.8% in 2016. 2017 saw a significant rise in R&D expenses (62%) to €28.3m to support investment in portfolio products, Risperidone ISM and Letrozole ISM. In the medium term we expect operational leverage from the fully vertically integrated LMWH manufacturing and distribution business to aid margin expansion. We forecast an absolute 1.30% improvement in the operating margin from 10.7% in 2016 to 12.0% in 2020. Beyond this period, the enoxaparin contribution will continue to contribute to margin expansion. However, the game changer will be DORIA; by 2024 we believe this high-margin asset could support margins of 29%.

ROVI’s proprietary, patented ISM technology is based on the formation of in situ microparticle implants (ISM) for sustained drug release (long-acting intramuscular, administered by injection). The advantages for the patient are that it reduces plasma drug concentration variability, reduces the number of doses necessary for treatment, removes the need to remember to take a daily oral medication, and thus improves patient adherence and clinical outcomes. ROVI’s focus is on targeting long-term illnesses (schizophrenia and breast cancer), but we expect its ISM technology to be extended to other chronic indications. The ISM technology combines the advantages of technologies such as preformed microparticles with implants. Risperidone ISM (DORIA), the first clinical asset developed using its proprietary drug development technology (US and EU approval likely in 2021) would not only serve to validate this technology but, importantly, introduce a high-margin, high sales potential asset to ROVI’s P&L.

ROVI’s sustained-release injectable technology (ISM) is based on two separated syringes; one contains the active pharmaceutical ingredient and a polymer (solid state), and the other contains the solvent (liquid state). The ISM technology is patent protected until 2033. ROVI has been successful in the aseptic filling of the dry powder active pharmaceutical ingredient (API) into a syringe, which has eluded other formulation developers. There is no need for refrigeration of the two separate components of the injection. Patients are administered the injection once a month by a healthcare professional. At time of injection, the two components are mixed together so the active ingredient plus polymer mixes with the liquid and the suspension is then injected intramuscularly into the deltoid or gluteal muscle. The suspension precipitates into a biodegradable solid or semisolid implant at the injection site. The implant contains the API and, as it erodes, it releases the active substance over time in a sustained and controlled manner (Exhibits 1 and 2 demonstrate DORIA’s pharmacokinetic profile). ROVI’s proprietary technology is intended to overcome some of the disadvantages of prolonged-release oral or injectable formulations. Management believes that the advantages include an injection that is simpler to administer; high encapsulation efficacy; greater stability of active substance; and greater control of the initial release of the drug. Taking the last two points in context, we note that DORIA achieves therapeutic blood levels within hours of injection, whereas none of the currently available LAIs is (either concomitant oral drug therapy overlap is required for the first few weeks, or a booster injection after one week before regular maintenance dosing is possible, in the case of Johnson & Johnson’s [J&J] Invega Sustenna).

ROVI’s long-acting injectable (LAI) formulation of risperidone (Risperidone ISM or DORIA) is an improved version of the long-acting antipsychotic drug, risperidone, currently available on the market (J&J’s Risperdal Consta). While multiple long-acting formulations of different antipsychotic drugs are approved, none has the perfect profile. DORIA will provide another product in the LAI schizophrenia treatment armament that will benefit physicians and patients as it addresses some of the issues with other LAIs on the market and in late-stage development. While multiple LAI antipsychotic drug formulations should be available by the time we assume DORIA launches (US and EU in 2021), and DORIA’s key advantages aside, our analysis shows that the overall LAI market in the US is underutilised. In the US 8% of patients are treated with LAI vs 40% in Europe in 2014. A multitude of scientific literature and a change in World Federation of Societies of Biological Psychiatry (WFSBP) guidelines (2013) point to increased use of LAIs.

The Phase II PRISMA-2 clinical trial evaluated the safety and tolerability of DORIA in 67 patients with schizophrenia. This open-label study looked at a 75mg dose, injected intramuscularly (gluteal or deltoid) every 28 days for four months. Importantly, oral supplementation with risperidone was not required. Recently published data reported that the mean Cmax of the active moiety was achieved 24-48 hours (Tmax) after each administration and ranged over four consecutive doses from 39.6-53.2ng/mL and 54.1-61ng/mL when given in the gluteal or deltoid, respectively (see Exhibits 1 and 2). Importantly, all patients achieved therapeutic levels (>10ng/mL for the active moiety) between two and eight hours after drug administration and sustained release of drug was available throughout the 28-day dosing period. Side effects were in line with risperidone medications (hyperprolactinaemia 53.7%, injection site pain 32.8%). There were no significant changes across the study either on the PANSS (see later) or extrapyramidal symptoms scale.

The ongoing Phase III PRISMA-3 (ClinicalTrials.gov Identifier NCT03160521) clinical trial is evaluating the efficacy, safety and tolerability of DORIA (75mg and 100mg administered intramuscularly into the gluteal or deltoid muscle once a month for three successive months) in patients with acute schizophrenia in a multi-centre, randomised, double blind, placebo-controlled trial. Patients who have never taken risperidone before are required to undertake a brief trial of oral risperidone to ensure there is no clinically significant adverse hypersensitivity reaction before receiving the first DORIA dose. The primary outcome measure is the PANSS score mean change from baseline to endpoint over a 12-week period (plus one year follow-up). The Positive and Negative Syndrome Scale (PANSS) is a standardised clinical interview assessing and scoring 30 standardised symptoms (seven positive symptoms, seven negative symptoms and 16 general psychopathology symptoms). PRISMA-3 is assessing whether an improvement in PANSS from day 4 is achievable and statistically significant. If so and if included in its label, DORIA would have a unique selling point versus its competitor LAIs. Secondary outcome measures are the Clinical Global Impression-severity (CGI-S) score mean change from baseline to endpoint, CGI-I score mean at endpoint, overall response rate and PANSS response rate at endpoint. The trial is also looking at health economic-related outcomes (Exhibit 3), which could be relevant in pricing and reimbursement negotiations. PRISMA-3 is expected to read out in Q219, leading to NDA and marketing authorisation applications (MAAs) filing in late 2019. We forecast DORIA launch in 2021.

Exhibit 7 provides a visual guide to DORIA’s advantages versus competing LAIs on the market and about to launch. While we highlight these advantages we note that the complexity of schizophrenia means one size does not fit all and DORIA provides an additional formulation for physicians to consider on an individual patient basis. Later in this note we highlight the scientific evidence from medical literature on why LAI antipsychotics on the whole are increasingly considered to be better than oral medications in disease control. While we would not expect LAI to replace oral medications completely, the reality is that LAIs are underutilised, particularly in the US. DORIA’s advantages are as follows:

In the US ROVI will pursue regulatory NDA filing under section 505(b)(2) of the FD&C Act (as agreed at the special protocol assessment [SPA] meeting with the FDA). Under this section provision (2) of the 505 act, the FDA is permitted to rely on data not developed by the applicant for approval of an NDA. ROVI will use the published literature on oral risperidone as part of the application for the treatment of schizophrenia, in addition to the PRISMA-1, PRISMA-2 and PRISMA-3 efficacy outcomes and safety data. Under the amendment to the act, the FDA aimed to have a slighter faster approval time frame to NME (by months), but in practice this has not panned out.

In Europe ROVI will file a hybrid application with the European Medicines Agency (EMA) for the treatment of acute exacerbation of schizophrenia. The EMA defines a hybrid medicine as “a medicine that is similar to an authorised medicine containing the same active substance, but where there are certain differences between the two medicines such as in their strength, indication or pharmaceutical form”. We forecast a normal approval time frame for DORIA rather than an expedited one.

According to the US National Institutes of Health, the prevalence of schizophrenia in the US adult population is 1.1%. We apply the same prevalence rate to the European population to derive total US/EU schizophrenia patient numbers of 5.6 million. According to the National Institute of Mental Health’s Clinical Antipsychotic Trials of Intervention Effectiveness (the CATIE programme), 75% of patients are incomplete responders (including drug discontinuation due to side effects). In the US 8% of patients are treated with LAI vs 40% in Europe. Given the low LAI penetration rates in the US and an increase in the number of differing LAIs on the market or to be launched within the next few years, we believe the improved formulations should see a rise in overall LAI penetration rates. This derives our eligible LAI patient population in 2021. Given that DORIA’s clinical development programme (PRISMA-3) is focused on acutely exacerbated schizophrenic patients, we apply a conservative 5% share of the total LAI market in both Europe and the US. In the US the average LAI is priced at ~$10,700 pa – we note that Invega Sustenna monthly maintenance (based on 0.75ml of 117mg) is priced at $1,200 per month (retail price). In Europe the average LAI is priced at $3,500 pa. For DORIA we have assumed US pricing of $10,500 pa and EU pricing of $4,000 pa. Depending on the HEOR outcomes data from PRISMA-3 (see above), DORIA could command higher pricing versus our expectations.

ROVI’s commercialisation strategy for DORIA will vary between the US and EU. In the US ROVI is filing for ‘treatment of schizophrenia’ as the labelled indication vs ‘treatment of acute exacerbation of schizophrenia’ in Europe. The small but important nuance here is that the US label implies DORIA can be used at any time during the course of the disease. In practice, US physicians are more likely to consider an LAI in the acute setting (either onset or relapse) as this is the most pertinent time to consider changing a medication or switching to a long-acting formulation, given that, in many cases, lack of patient adherence to medication may have precipitated the relapse. ROVI has developed DORIA specifically as a once-a-month administration (vs weekly, fortnightly or every three months) to coincide with monthly outpatient visits. LAIs facilitate regular contact between patients and psychiatrists, which is essential for monitoring patient welfare and clinical progress.

We anticipate that in Europe ROVI will launch via its subsidiaries in key European markets: Germany, UK, Spain, Italy, France and Portugal. ROVI will need to hire specialist medical reps to support launch and marketing to physicians. In terms of sales and marketing expenses, it will need to extend its sales rep footprint to target the psychiatrists working in the acute medical setting – these physician numbers are relatively small and we believe ROVI can service the major European countries through a salesforce the size of 100 reps. In Europe LAI pricing is $3,650 pa on average and use of LAIs is estimated at 40% penetration. As such, we believe the US opportunity is larger (due to higher pricing – $10,700 pa and relative underutilisation of LAIs versus oral meds). In the US ROVI will need to build a salesforce of a similar size (80 reps) to cover the ~4,000 psychiatrists working in the acute clinical setting. However, ROVI could elect to partner the product in return for royalties and milestones on sales. Competition in the US is entrenched, with J&J, BMS, Lilly and AstraZeneca dominating the market. Nonetheless, given DORIA’s profile and assuming ROVI elects to market, we believe a 5% share of the relevant market is achievable. Furthermore, we expect DORIA to piggyback on the LAI segment versus the oral market from the increasing detailing by competitors.

In the US market we believe DORIA operating margins can reach 40% by 2024. DORIA fits well into ROVI’s fully vertically integrated pharmaceuticals business strategy and we believe the company’s manufacturing efficiency could enable high gross margins (95% in the higher-priced US market and 85% in Europe). In the US we assume a salesforce of 80 reps and additional overhead expenses totalling €49m by 2024 to place SG&A expenses at 56% of sales. This seems reasonable for the early launch years. However, beyond 2024 we would anticipate further operating leverage. In Europe we forecast DORIA operating margins will reach 14% in 2024 and 23% in 2025. We model 100 reps for Europe, but lower overhead costs given ROVI’s existing presence in the major European countries through its subsidiaries. The main drag on margins in comparison to the US is the lower pricing opportunity in Europe and thus higher gross margins (we model 85%).

Schizophrenia is a chronic and severe mental disorder that affects how a person thinks, feels and behaves. Prevention of relapse is a key challenge in schizophrenia as poor adherence to antipsychotic medications leads to high rates of hospitalisation rates and related burden of care, particularly at the onset or during acute episodes/relapses. Genetic predisposition (multiple genes have been implicated) combined with environmental exposures (eg recreational drugs) and/or stresses during pregnancy and childhood are thought to contribute or trigger risk of disorder. The condition affects 1.1% of the world’s population over the age of 18 (source: National Institute of Mental Health); actual rates vary slightly from country to country (about 0.5-1% of the population). Patients are typically diagnosed with the disorder in early adulthood.

The disorder is characterised by a breakdown in the relation between thought, emotion and behaviour, leading to faulty perception and the inability to function normally. Signs and symptoms of schizophrenia can vary by severity, duration and frequency, although the incidence of severe psychotic symptoms often decreases during a patient’s lifetime. Non-adherence to taking medications as prescribed, use of alcohol or recreational drugs, and stressful situations tend to increase symptoms. Generally, symptoms are classified into several categories:

While the condition cannot be cured, the success rate of treatment with antipsychotic medications and psychosocial therapies can be high but is dependent on a variety of factors (eg side effects, patient compliance and adherence, patient insight and insurance coverage). Drug therapy usually requires long-term maintenance treatment using one or more antipsychotic drugs (the two main classes of drugs are the ‘typical’ and the ‘atypical’ antipsychotic drug classes). Antipsychotic medications can be taken orally on a daily basis or less frequently (fortnightly/monthly) by intramuscular or subcutaneous injection of LAI formulations of antipsychotic drugs.

There is no cure for schizophrenia and drug treatments focus on eliminating symptoms; the current mainstay of which are the atypical and typical antipsychotic drugs. Both oral and long-acting injectable antipsychotics have shown efficacy, tolerability and safety as treatments for patients with schizophrenia.

First-generation or so-called typical antipsychotics were developed in the 1950s. Examples include chlorpromazine (1950s), fluphenazine (introduced in the 1960s) and haloperidol. These drugs are dopamine (D2) antagonists so side effects include sedation, hypotension and in some cases non-reversible tardive dyskinesia (drug-related involuntary movements).

Clozapine was the first second-generation drug developed; these drugs work as serotonin/dopamine antagonists, D2 partial agonists or serotonin partial agonists at 5-HT. Clozapine was not available in the US (risk of agranulocytosis) but is available in some countries in Europe. Subsequently, other second-generation or atypical antipsychotics were developed in the 1990s. Examples include risperidone (Janssen’s Risperdal), aripiprazole (BMS’s Abilify), olanzapine (Lilly’s Zyprexa), and quetiapine (AstraZeneca’s Seroquel). While this newer generation of atypical antipsychotics improved on the toxicity profile of the typical antipsychotics, drug-related side effects remain troublesome and include sedation, metabolic changes (weight gain, diabetes, hyperlipidemia) and endocrine changes (sexual side effects). Nonetheless, second-generation oral antipsychotics have become the first-line treatment in schizophrenia. In terms of unit volume, risperidone is the most used active drug principle.

Antipsychotic drugs with LAI formulations were developed to reduce the problem of non-adherence, which is estimated to be as high as 40%. The first LAI antipsychotic to be developed was fluphenazine enanthate (ER Squibb & Sons) in 1966, which displayed numerous potential advantages (including better correlation between dose and blood plasma concentrations, better monitoring of compliance, ease of administration and less risk of overdose). Disadvantages included slow titration to dose and a longer duration to therapeutic dosing. During the past 20 years improved versions of second-generation LAIs have been available on the market: Risperdal Consta in 2004 and its follow-up, Invega Sustenna, in 2009.

Long-term drug treatment of schizophrenia has major limitations; the US National Center for Biotechnology Information (NCBI) estimates that 25-33% of patients are treatment resistant and relapse rates remain high (relapse rates over two years in medication-treated chronic schizophrenia patients are approaching 41% (source: Crow et al 1986); the cumulative relapse rate for first-episode patients with good adherence over a three-year period was 36%, whereas the rate for poorly adherent patients was 57% (Chen et all Schizophr Res. 2005). Non-adherence to antipsychotic medication is common among patients with schizophrenia, and is the greatest challenge for recovery and prevention of relapse with greater risk of hospitalisation. This is particularly so during the early days of diagnosis when patients are coming to terms with their disorder and may not appreciate the necessity of adhering to the medication regimen. Discontinuation of antipsychotic medication in patients achieving remission leads to a relapse in more than 52% of patients in 6.5 months (source: Gilbert et all, 1995). Leucht et all points to a higher risk of relapse if antipsychotic drug medication is discontinued in patients despite them being stable on treatment for five years; this is reflected in treatment guidelines such as those of the World Federation of Societies of Biological Psychiatry, which recommend continuation of antipsychotic drugs for the treatment of first-episode psychosis for at least two years after first remission, and with a minimum of five years relapse-free stability (maybe even throughout life) before considering slow drug withdrawal (over six to 24 months). There is an unmet need in schizophrenia for novel mechanism and new formulations of antipsychotic drugs that can improve patient adherence, reduce persistent psychotic symptoms in antipsychotic-treated patients with fewer side effects and improve on negative and cognitive symptoms.

LAI antipsychotics are often used only when oral medications have failed rather than first-line therapy. Evidence increasingly supports their use as a first choice (WFSBP 2013 updated guidelines). Multiple effectiveness studies show the superiority of LAI antipsychotics, particularly in the case of risperidone.

According to IMS Midas, the US was the largest market for antipsychotics, with 2015 sales of $14.2bn, and sales in the top five largest European markets were $2.6bn. In the US oral formulations represented 87% of the value ($12.4bn) and LAIs accounted for the remaining 13% ($1.8bn) of the value. In the top five European markets, oral represented 69% and injectable 31% in terms of value in 2015. However, as antipsychotic drugs are used to treat a myriad of conditions (including schizophrenia but also depression, bipolar disorder, schizoaffective disorders and are also used off-label to treat acute psychotic episodes in dementia), a better understanding of market opportunity is required to assess the historic schizophrenia injectable market size.

According to IMS Midas, the US is the largest schizophrenia injectable market with MAT Q317 sales of €2.8bn and 2.0m units. Exhibit 5 shows the 25% CAGR (2011-15) in the injectables segment in value. Average pricing for LAIs in the US is ~$10,700 per annum. The market leader by volume (in units) in 2015 was risperidone (32% market share) followed by paliperidone (25% market share). In terms of value in 2015, paliperidone had the highest market share (60%), followed by risperidone (25%). The difference between unit and value relate to pricing; risperidone generics have been available in the US since 2008.

IMS Midas reports top five European injectable market sales of €793m and 3.2m units. Exhibit 6 shows the 7% CAGR (2011-15) in the injectables segment by value – average LAI pricing per annum is €3,650. IMS Midas forecasts monthly/fortnightly formulations to have 55% market share in 2021 (every two months or every three months formulations). The market leader by volume (in units) is risperidone (59% market share) followed by paliperidone (33% market share). In terms of value in 2015 paliperidone had the highest market share (55%) followed by risperidone (39%).

Exhibit 7 highlights the main competitors in the LAI antipsychotic market and DORIA’s competitive positioning on key clinical and practical metrics. An important general comment is that the choice of LAI will firstly depend on the antipsychotic drug itself; some patients will fare better with risperidone rather than aripiprazole or paliperidone or vice versa.

The key competing LAI antipsychotic products are:

Advances in the treatment of breast cancer over the last 20 years have, in certain cases (hormone-dependent or HER-2 status), led to improved survival and many patients becoming cancer free. Novartis’s Femara (letrozole), an aromatase inhibitor, is used to treat post-menopausal women with oestrogen receptor positive primary breast cancer. Letrozole is prescribed as an adjuvant to surgery. The drug is usually continued for five years, as aromatase inhibitors decrease the risk of developing a new breast cancer in the same or opposite breast (studies are underway to evaluate longer-term use, eg 10 years). However, once a patient is breast cancer free, adherence to an oral daily ‘cancer medication’ can be an issue as some patients elect to or forget to take the product. The opportunity for a once every three months, slow-release, intramuscular depot formulation of the drug is multi-fold and include improved patient quality of life (lower dose frequency and reminder of illness), reduced healthcare costs and possible improved clinical outcomes. ROVI has initiated Phase I (LISA-1) clinical trials to evaluate the pharmacokinetic and pharmacodynamic profile of Letrozole ISM at differing doses in healthy post-menopausal women. We do not currently reflect Letrozole explicitly in our ROVI valuation.

ROVI is subject to various sensitivities common to speciality pharmaceutical companies, including commercialisation (pricing, reimbursement, uptake and competition), manufacturing and financing risks. The key sensitivities for ROVI relate to successful European commercialisation of both Hibor and its biosimilar enoxaparin, and crystallising value from the ISM pipeline. Enoxaparin represents 13% of our total revenue forecasts in 2019 (its second year of commercial availability). The largest near-term driver of our sales and net profit expectations is enoxaparin. Compared to ROVI’s current portfolio of drugs and footprint, the US opportunity for DORIA is large and a key valuation driver (accounting for 16% of our valuation; EU DORIA accounts for 9%).

Our revised valuation of €1.16bn or €23.3/share reflects a change to our sum-of-the-parts methodology (Exhibit 8). Our previous three-stage, DCF-based valuation of ROVI at €0.96bn or €19.1/share used our sales and P&L model, and included peak sales of €76.2m for DORIA in these cash flows. The major source of uplift reflects the change in our assumptions for US and European DORIA sales and operating cost assumptions. We now value DORIA US and EU through standalone NPV calculation (Exhibit 10) and derive value for the rest of the business by using a DCF of our sales and P&L model excluding DORIA (Exhibit 9). We have not changed our sales assumptions for the rest of the business. Compared to ROVI’s current portfolio of drugs and footprint, the US opportunity for DORIA is large and a key valuation driver (accounting for 16% of our valuation; EU DORIA accounts for 9%).

Our sum-of-the-parts valuation consists of:

ROVI reported 4% growth in operating revenue to €275.6m in 2017 (vs 8% to €265.2m in 2016), aided by strong growth in the toll manufacturing business. Total revenues grew 12% to €76.0m in Q118 driven by strong growth in the speciality pharmaceutical business (+23% to €63.8m). Biosimilar enoxaparin reported fledgling Q118 sales of €4.1m despite only being available in Germany since September 2017 and in the UK since March 2018. For more details on Q118 results and FY17 results see our notes Stunning growth in speciality pharmaceuticals and Steady as she grows.

We have made no changes to our financial forecasts. We expect operating margins to further decrease in 2018 (from 10.7% in 2016 and 6.9% in 2017) to 4.6% in 2018, mainly due to higher R&D expenses but also to the small increase in SG&A to support new product launches. We anticipate margin growth from 2019 to 7.6%, mainly due to operational leverage as we still anticipate c €32m in R&D costs in 2019. We forecast an absolute 1.3% improvement in operating margin from 10.7% in 2016 to 12.0% in 2020. Longer-term margins should continue to ramp up beyond this period as the operational leverage from enoxaparin sales starts to flow through to the P&L, and the impact of high-margin asset, DORIA, becomes evident. In the US market we forecast that DORIA operating margins could reach 40% by 2024 and in Europe we believe DORIA operating margins could reach 23% in 2025.

At 31 March 2018 ROVI had €40.4m in bank borrowings and debt, and cash and cash equivalents of €39.0m.