Prima BioMed — Update 27 July 2016

Prima BioMed is an ASX-listed immunotherapy company headquartered in Sydney, Australia, which is focused on developing products based on the LAG-3 pathway, following the acquisition of the private French immunotherapy company Immutep in December 2014. Its lead product IMP321 is in Phase II in breast cancer and in Phase I in combination with the checkpoint inhibitor Keytruda – Prima retains all the product rights except for China (partnered with Eddingpharm). IMP731 is in Phase I with GSK for autoimmune diseases, while IMP701 is in Phase I for solid tumours with Novartis. Prima previously conducted Phase II trials for the dendritic cell vaccine CVac in ovarian and pancreatic cancer, but has out-licensed the programme to Sydys Corporation. The company has facilities in Paris, France; Leipzig and Berlin, Germany; and its headquarters are in Sydney, Australia.

Our DCF valuation is A$282m or 14c/share (9c/share fully diluted). The fully diluted value per share is based on potential dilution from the 1.3bn options, warrants and convertible notes that would be in the money at the undiluted valuation of 14c/share, including an assumption that the A$13.75m Ridgeback Capital convertible note is converted to 688m shares at 2c/share. There would be further upside if the LAG-3 products progress and if studies indicate broader potential in new indications. GSK has indicated that it expects to progress the IMP731 programme into Phase II in 2016, which would provide upside to our current valuation. The next catalysts include results from the two dose-finding cohorts in the Phase II study of IMP321 in metastatic breast cancer (MBC) in Q416, which would followed by the initiation of the 196-patient randomisation phase. We also anticipate newsflow from the ongoing Phase I trial of IMP701 with Novartis.

Prima BioMed acquired French private immunotherapy company Immutep in December 2014 for a total consideration of A$26.3m. In H116 Prima raised A$13.8m (gross) through equity issues and raised a further A$13.8m through the issue of a convertible note to Ridgeback Capital (first announced in May 2015). The Ridgeback notes mature in 2025 and can be converted into 688m shares (ie 2c per share). Ridgeback was also issued 380m warrants, which are exercisable at 2.37-2.5c per share and expire in 2020-25. Prima accounted for the financing transaction, including the issue of the warrants, as a share based payment to a strategic investor (Ridgeback) valued at A$47.4m, in the H116 accounts. The fair value of the convertible note, estimated at A$4.7m, is recorded as a non-current financial liability, with the balance recorded in equity. Gross cash position at the end of June 2016 was A$20.9m. Operating cash burn in FY16 was A$11.6m; we forecast a 30% increase in FY17 to A$15.1m. Our forecasts indicate that Prima has a cash reach beyond the end of FY17, but that an extra A$8m will be needed to fund operations until end FY18.

Prima is exposed to clinical, regulatory and commercialisation risks typical of all biotech companies. The key sensitivity is clinical progress of its pipeline of LAG-3 candidates, primarily the internally funded IMP321. While Prima has funds to initiate the randomisation stage of the IMP321 Phase II study in MBC, it would require a partnership or alternative forms of funding to complete the study and advance IMP321 further. Existing partnerships with big pharma reduce the financial and execution risk for IMP701 and IMP731; in addition, if the Phase I study of IMP701 reveals evidence of efficacy, this could lead GSK to extend the study to additional indications including RA and MS, which could increase the potential peak sales and therefore the value of the product. Separately, progress with CVac depends on partner Sydys raising sufficient funds to progress development.

In December 2014 Prima BioMed completed the acquisition of French private immunotherapy company Immutep for a total consideration of A$26.3m. Immutep had a pipeline of three products, two of which are partnered, based on pathways in the Lymphocyte Activation Gene 3 (LAG-3) immune control mechanism. Subsequently, Prima took the strategic decision to prioritise the development of its lead LAG-3 product IMP321, a first-in-class fusion protein over its previous sole product CVac (a dendritic cell therapy) and has recently out-licensed CVac to Sydys Corporation.

The Immutep acquisition brought established relationships with Eddingpharm, GSK and Novartis, together with potential milestones of over US$100m plus royalties. Immutep’s founder, Professor Frédéric Triebel, a leading expert on LAG-3, joined Prima BioMed as CMO/CSO.

Cancer immunotherapy is used to enhance the ability of the immune system to attack and kill tumour cells. Immunotherapy harnesses the patient’s own immune system via a range of mechanisms, either activating it to detect and destroy tumour cells or blocking signalling mechanisms, which tumours exploit to suppress anti-tumour immune responses.

The potential of immunotherapy to improve patient outcomes has been highlighted by the success in melanoma, lung, renal and bladder cancers of the drugs Yervoy (ipilimumab), Keytruda (pembrolizumab), Opdivo (nivolumab) and Tecentriq (atezolizumab) that block the CTLA4 or PD1/PD-L1 immune checkpoint inhibitor (ICI) pathways.

Many tumours are able to escape immune attack by activating pathways in the immune system that act as checkpoints to dampen down immune responses when they are no longer needed, such as after a viral infection has been eliminated.

There are a number of these immune checkpoint pathways that can be activated in the tumour microenvironment to suppress immune responses in addition to PD1 and CTLA4, including LAG-3, OX40, TIM3 and GITR.

Responses to the approved ICI drugs Yervoy, Keytruda, Opdivo and Tecentriq are frequently long-lasting, but response rates to single agent ICI therapy are often low, typically in the range 10-30%. This has led to increased interest in other immunotherapies that can be used either on their own or in combination with PD1 or CTLA4 ICI drugs to increase the proportion of patients who respond to therapy and to increase the efficacy in other types of cancers.

LAG3 is of particular interest as an immunotherapy because in different circumstances it can either act as an ICI and suppress T-cell activity or stimulate the activity of antigen-presenting cells and help initiate an anticancer immune response (Exhibit 2). LAG-3 exists in both soluble and membrane-bound forms, which have different functions.

LAG-3 is expressed on activated T-cells, natural killer cells, B-cells and a small subset of dendritic cells known as plasmacytoid dendritic cells. LAG3 is not expressed by resting T-cells, but is upregulated several days after T-cell activation. It is significant to note that LAG3 is upregulated on so-called exhausted T-cells (which have impaired function following prolonged exposure to antigen) compared with effector or memory T-cells.

LAG3 binds strongly to major histocompatibility complex (MHC) class II molecules expressed on the surface of antigen-presenting cells (APCs), including dendritic cells.

In certain configurations (eg multiplexed LAG-3 on the membrane or cross-linked soluble LAG-3), this binding to surface MHC molecules activates the APCs to mature and transport tumour antigens to the lymph nodes for presentation to T-cells. In turn, this increased antigen presentation stimulates the reactivation and expansion of cytotoxic CD8+ T-cells, which can circulate to the tumour via the bloodstream and attack and kill tumour cells, thereby activating and amplifying the anti-tumour immune response.

In addition to the ability to activate APCs, membrane bound LAG-3 can act as a classical inhibitory immune checkpoint in T-cells. This inhibitory effect occurs when membrane-bound LAG-3 on the surface of an activated T-cell binds to MHC class II molecules on an APC at the same time as adjacent T-cell receptors (TCR) are bound to MHC class I molecules on the APC that is bearing an antigen fragment.

Prima BioMed’s lead product IMP321 is a LAG-3 Ig fusion protein that is based on the soluble form of LAG-3 and can activate APCs. These activated APCs process tumour antigens, including those released from cells killed by chemotherapy, transport the antigens to lymph nodes and present the tumour antigens to T lymphocytes, thus activating and amplifying the immune response, as summarised in Exhibit 3.

Previous studies conducted by Immutep before its acquisition by Prima focused on the use of IMP321 in combination with chemotherapy and as an adjuvant to cell therapy vaccines to stimulate APC activity.

In a previous Phase I/IIa study in MBC (n=30) conducted by Immutep, administration of IMP321 the day after chemotherapy induced a doubling of the tumour response rate to 50% vs 25% over six months (in comparison to a historical cohort of chemotherapy-only patients), as shown in Exhibit 4.

In December 2015 Prima initiated the AIPAC (Active Immunotherapy PAClitaxel) Phase II trial of IMP321 in patients with MBC who are HER2 negative and therefore ineligible for treatment with Herceptin. The trial will recruit approximately 211 patients at 30 sites in six European countries.

The first cohort of six patients who received 6mg of IMP321 in combination with paclitaxel has safely completed treatment. No drug-related serious adverse events were observed and data demonstrated activation of blood monocytes/dendritic cells and CD8 T-cells. Recruitment is underway in the second cohort of nine patients who will receive 30mg of IMP321, which will complete the open-label, safety run-in stage of the trial. The results for both cohorts will be presented and compared in Q416 to confirm the recommended Phase II dose of IMP321.

The trial will then move to a randomized, placebo-controlled, double-blind stage where 196 patients will receive six months of treatment with paclitaxel plus either IMP321 or placebo. After six months, paclitaxel treatment will cease and responders and patients with stable disease will continue to receive either IMP321 or placebo monotherapy for a further 12 months. The trial will recruit MBC patients who have not undergone prior chemotherapy for metastatic disease and are hormone receptor-positive and HER2 negative.

The primary endpoint from the randomized stage of the trial is improvement in progression-free survival (PFS), while overall survival (OS) is a secondary endpoint; tumour biomarkers will also be monitored. The European regulator (EMA) has indicated that this trial could be sufficient to support a marketing authorisation if it achieves certain (undisclosed) clinical endpoints.

The company expects the trial to take about three years to complete, so results should be available in early 2019. Positive results could potentially allow Prima to file for approval in Europe in 2019, with approval possible in 2020. US approval could potentially be achieved in 2023 following a Phase III trial.

In January 2016 Prima initiated a Phase I trial called TACTI-mel (Two ACTive Immunotherapies in melanoma). This trial is testing IMP321 in combination with the anti-PD1 immune checkpoint inhibitor Keytruda (pembrolizumab, Merck) in 24 patients with unresectable or metastatic melanoma. Patients will receive ascending subcutaneous doses of IMP321 up to 30mg per injection. The study will evaluate anti-tumour activity and the nature of the immune response at various doses, in addition to safety, pharmacokinetics and pharmacodynamics.

This study is of particular interest because it combines the APC activation properties of IMP321, which helps to initiate an immune response, with an ICI which ‘releases the brakes’ on the immune effector cells, enabling a stronger immune response. Prima has shown in preclinical studies that combining IMP321 with immune checkpoint inhibitors increased the strength of an anti-cancer immune response and the speed and level of tumour regression. Exhibit 5 shows that combining IMP321 with a PD-1 blocking antibody increased the activation of cytotoxic T-cells when human white blood cells (peripheral blood mononuclear cells) were stimulated with a peptide antigen in vitro. Similarly, in a mouse model of colon cancer, combination therapy with the mouse analogue of IMP321 (mLAG-3Ig) plus a PD-L1 blocking antibody inhibited tumour growth to a greater degree than either treatment on its own (Exhibit 6).

Combining two immunotherapy treatments with different mechanisms of action has already been proven to be an effective strategy. The FDA approved the combination of the ICIs Yervoy (ipilimumab) and Opdivo (nivolumab) in patients with advanced melanoma. Patients treated with the combination experienced a 60% overall response rate compared to 11% in those treated with Yervoy alone.

Prima announced earlier this month that it will participate in an investigator-sponsored collaborative study of intratumoural injection of IMP321. The new clinical trial, INSIGHT, will explore the potential for IMP321 to act as an activator of dendritic cells found within solid tumours.

The study, which will be led by Professor Doctor Salah-Eddin Al-Batran at the Northwest University Hospital in Frankfurt, Germany, will commence once it has received regulatory and ethics approval. As the trial is investigator-initiated it will not require significant resource commitment from Prima.

Prima has partnerships with Novartis and GSK for two separate programmes, which use monoclonal antibodies (mAbs) to target membrane-bound LAG-3 in different ways. Each of the pharma partners has commenced a clinical trial programme based on the in-licensed LAG-3 drug.

IMP701, which is partnered with Novartis, is an antagonist mAb, which blocks the LAG-3-mediated inhibitory signal given to tumour-infiltrating T-cells and thus activates T-cell proliferation. LAG525 is Novartis’s humanised version of IMP701, which it has taken into clinical development.

In June 2015 Novartis commenced a Phase I/II trial of LAG525 under its collaboration and licensing agreement with Prima. The Phase I trial (ClinicalTrials.gov Identifier: NCT02460224) is testing LAG525 as a single agent, as well as in combination with its in-development anti-PD1 immune checkpoint inhibitor PDR001, in patients with solid tumours.

In June 2016 Novartis added a third treatment group to the trial and increased the size of the trial to 416 patients, a substantial increase from the original target of 240 patients. The third treatment arm involves treating Japanese patients with LAG525 as a single agent.

Novartis has also tightened the focus of the trial to melanoma, lung and renal cancers, having previously also included bowel and nasopharyngeal cancers. Following the expansion in trial size, the estimated completion date has been pushed back to August 2018.

The fact that Novartis is trialling LAG525 in combination with PDR001 underscores the potential that it sees for synergistic combinations between drugs that target LAG3 and PD1/L1 checkpoint inhibitors. We take this expansion of the trial size as a positive sign, which may indicate that Novartis has seen encouraging initial results from the study. PDR001 is currently in Phase II development in nasopharyngeal carcinoma.

Novartis has not disclosed which cancers it intends to target for subsequent development of LAG525; in our forecasts, we assume that the initial indication will be non-small cell lung cancer (NSCLC), which has already been successfully targeted by the PD1 immune checkpoint inhibitors pembrolizumab (Keytruda, Merck) and nivolumab (Opdivo, Bristol-Myers Squibb).

Prima’s second out-licensed LAG-3 programme utilises LAG-3 in a different way to target autoimmune disease. It is based on IMP731, which is a cytotoxic mAb that will kill the few LAG-3+ activated T-cells that infiltrate autoimmune disease sites.

In animal studies IMP731 was shown to target and selectively deplete LAG-3 marked T-cells, leaving normal T-cells intact. This offers an interesting new approach compared with standard treatments such as corticosteroids, which inhibit T-cells indiscriminately, suppressing immunity. The approach is potentially applicable to the wider autoimmune disease population; diseases such as multiple sclerosis and rheumatoid arthritis are associated with chronically stimulated T-cells.

IMP731 is under development by partner GlaxoSmithKline (GSK), which in-licensed the programme from Prima (Immutep) in 2011. GSK has developed a humanised form of the IMP731 mAb, which it has termed GSK2831781. GSK dosed the first patient in a Phase I study of GSK2831781 in January 2015 in patients with plaque psoriasis. GSK2831781 aims to kill the few activated LAG-3 positive T-cells that are autoreactive in autoimmune disease. At its R&D day in November 2015 GSK indicated that it expects to progress its anti-LAG-3 programme into Phase II trials in immuno-inflammatory disease in 2016, with the potential for a regulatory filing in a 2021-25 time frame. The Phase II trial does not appear to have commenced yet.

In May 2016 Prima announced an agreement to out-licence CVac to Sydys Corporation (OTC: SYYC, www.sydyscorp.com), a publicly traded New York company that has been repurposed to develop the CVac assets. There was no upfront payment, but Prima received a 9.9% equity stake in Sydys at the time of closing. Should CVac be successfully commercialised, Prima could receive over A$400m (US$293m) in development, regulatory and commercial milestone payments for achieving set commercial sales targets, in addition to low single-digit royalties on sales. The deal removes any requirement to fund further development of CVac, while allowing Prima to participate in upside if the vaccine proves effective in future clinical trials.

In addition to the licensing arrangement, Prima has transferred its contracts and some inventory and equipment to Sydys. Prima CEO Marc Voigt will join the board of Sydys.

We have updated our risk-adjusted DCF valuation to account for the CVac licence deal and have rolled forward our risk-adjusted DCF model to FY17. Our valuation of Prima has increased slightly to A$282m (previously A$273m) or 14c per share (undiluted, previously 13c per share) with a lower valuation for the CVac programme more than offset by rolling forward our model to FY17. On a fully diluted basis our valuation is unchanged at 9c per share, after taking into account the options, warrants and convertible notes on issue. Other core valuation assumptions are unchanged. Exhibit 7 summarises the constituent parts of our valuation, which is based on a discount rate of 12.5%.

We have altered our forecasts for CVac following the out-licencing to Sydys. As the ongoing development is contingent on Sydys raising further capital to fund clinical trials of CVac, we lower our overall likelihood of success for the programme to 20% from 30%.

No details have been provided on the split of the US$293m of potential CVac milestone payments. In our model we assume US$13m on filing for approval in the US, US$100m for approval in first disease indication, US$40m on approval in a second indication plus sales-based milestones totalling US$150m. Milestones are risked at 25% for the earliest down to 10% for the last sales milestone. We assume a 4% royalty rate (previously 12%). Our overall valuation of CVac falls to A$28m from A$44m, with most of the value now represented by the potential milestone payments.

Gross cash balance at end FY16 was A$20.9m. For valuation purposes we deduct the A$13.75 face value of the Ridgeback Capital convertible note in calculating end-FY16 net cash of A$7.1m as shown in Exhibit 7. We note that this is different to the accounting treatment of the convertible note, which includes only the A$4.7m estimated fair value of the convertible note as a non-current liability with the remainder treated as equity, resulting in a balance sheet net cash figure of A$16.1m as shown in Exhibit 9.

Our peak sales estimates for IMP321 and IMP701/LAG525 are based on pricing per patient of US$60k and US$40k in the US and Europe, respectively. The marketed ICIs Keytruda, Nivolumab and Tecentriq are all priced at about US$12,500 per month (US$150k per year) in the US, which suggests that our pricing assumptions may be conservative depending on the approved indications, duration of treatment and total cost of combination therapies.

Exhibit 8 shows that in addition to the 2,062m Prima shares currently in issue, there are a further 1,379m potential shares that could be issued on the exercise of options, warrants and convertible notes, including 1,301m that would be in the money at our 14c per share undiluted valuation. Exhibit 8 shows that after taking into account these potential shares, our diluted valuation is 9c per share. Prima is likely to require additional funding to complete the IMP321 clinical trials; our diluted valuation of 9c per share does not take into account potential dilution from any future capital raising.

The breadth of the LAG-3 pipeline means there could be further upside if Prima or its partners launch additional products into the clinic or broaden the indications being studied.

We include risk-adjusted milestones payable by current partners GSK for IMP731, Novartis for IMP701 and Sydys for CVac, plus milestones from prospective deals for IMP321. Possible catalysts include results of the second AIPAC dose-finding cohort and recommended Phase II dose, the start of the randomised phase of AIPAC, progression of the licensed anti-LAG-3 antibody into Phase II by GSK or news on partnering, all of which could provide upside to our current valuation.

Prima acquired Immutep for A$26.3m in H115. In H116 Prima raised A$13.8m (gross) through equity issues and raised a further A$13.8m through the issue of a convertible note to Ridgeback Capital (first announced in May 2015). The 13,750,828 convertible notes issued to Ridgeback have a face value of $1.00 per note, mature on 4 August 2025 and accrue interest at a rate of 3% pa, which may also be converted into shares. Each note can be converted into 50 shares (ie equivalent to A$0.02/share) for a total of 688m shares. Ridgeback was also issued 380m warrants which are exercisable at 2.37-2.5c per share and expire in 2020-25. Prima accounted for the financing transaction, including the issue of the warrants, as a share-based payment to a strategic investor (Ridgeback) valued at A$47.4m, in the H116 accounts. The fair value of the convertible note, estimated at A$4.7m, is recorded as a non-current financial liability, with the balance recorded in equity. Gross cash position at the end of June 2016 was A$20.9m. Operating cash burn in FY16 was A$11.6m, a 50% increase over the previous year; we forecast a further 30% increase in FY17 to A$15.1m. Our forecasts indicate that Prima has a cash reach beyond the end of FY17, but that an extra A$8m will be needed to fund operations until end FY18.

Prima is exposed to the same clinical, regulatory and commercialisation risks as all biotech companies. The key sensitivity is clinical progress of its pipeline of LAG-3 candidates, primarily the internally funded IMP321. While Prima has funds to initiate the randomisation stage of the IMP321 Phase II study in MBC, it would require a partnership or alternative forms of funding to complete the study and advance IMP321 further. Existing partnerships with big pharma reduce the financial and execution risk for IMP701 and IMP731; in addition, if the Phase I study of IMP701 reveals evidence of efficacy, it could lead GSK to extend the study to additional indications including RA and MS, which could increase the potential peak sales and therefore the value of the product. Separately, progress with CVac depends on partner Sydys raising sufficient funds to progress development.