Kazia Therapeutics — Ambiguity in GBM but pipeline still busy

In July 2022, Kazia Therapeutics announced that the study sponsors (GCAR) of the Phase III GBM AGILE study had advised the company that the paxalisib arm of the trial had completed recruitment but did not meet pre-defined criteria for continuing to the second stage. As a reminder, the GBM AGILE study (NCT03970447) is an adaptive platform trial split into two stages. Stage 1 is a randomised screening stage consisting of c 150 patients in which participating drugs are evaluated against a common control. If the therapy displays satisfactory safety, recruits the full patient sample size and hits a pre-defined efficacy threshold, then it will be graduated to Stage 2. This would be an expansion cohort consisting of c 50 additional patients, evaluating the primary endpoint of overall survival (OS) in active and control patients.

As paxalisib has been shown in prior studies to possess a good safety profile, we expect that the drug failed to reach the pre-defined efficacy threshold for graduation to Stage 2. However, patients already enrolled, which we expect to be c 150 patients, will continue in Stage 1 until the trial ends, with final survival and response data expected in H2CY23. We note that as the trial is still fully blinded and the degree to which paxalisib has missed the efficacy threshold will not be known until completion of the final analysis, we are unable to draw definitive results currently.

The graduation decision from Stage 1 to Stage 2 is centred on a complex algorithm using probabilities of OS superiority over the control arm based on a variety of different biomarker signatures, which we expect sets a high bar for graduation. Once the paxalisib data is fully analysed, Kazia will have randomised data from c 300 patients (c 150 within paxalisib arm, c 150 control) including c 12 month follow up data, which should add considerable statistical power to the results. If response and survival data are still clinically meaningful, it is possible that the company may have the data necessary to gain approval from the FDA, given the significant unmet medical need in GBM. Given the recent news, we expect that if approved, the breadth of paxalisib’s approval may be affected or may be granted on conditional basis, with further Phase III randomised data needed for full approval.

While it is Kazia’s most clinical advanced trial, the Phase III GBM AGILE study represents only a part of management’s development strategy for paxalisib (Exhibit 1). By leveraging its network of expert investigators, the company is investigating paxalisib’s use in various brain cancer indications in six active clinical trials (including GBM AGILE). Separate to the Phase III GBM AGILE study, Kazia is conducting a Phase II trial (NCT03522298) evaluating the safety, pharmacokinetics and efficacy of paxalisib in newly diagnosed GBM. The study is now in the follow-up stage, after demonstrating a progression free survival (PFS) of 8.6 months (vs temozolomide 5.3 months) and an overall survival (OS) of 15.9 months (vs temozolomide 12.7). We note that this study was not a double-blinded randomised trial and the comparison data for temozolomide is from different studies, limiting its comparability. Despite this, given the significant unmet medical need in GBM, these results could support paxalisib’s approval, in our view, if the GBM AGILE findings are seen as positive. The company is also conducting a Phase II study in primary CNS lymphomas (NCT04906096), led by investigations at the Dana-Farber Cancer Institute, which began enrolment in late 2021.

More recently, Kazia has focused on the use of paxalisib in rare pediatric cancers. As such, two studies in diffuse intrinsic pontine glioma (DIPG) are underway. The first, a Phase I study (NCT03696355) led by investigators at the St Jude Children’s Research Hospital, is evaluating paxalisib’s safety and efficacy in newly diagnosed DIPG patients after radiotherapy. We expect final data from this by end-CY23. The second, a Phase II trial (NCT05009992) led by investigators at the Pacific Pediatric Neuro-Oncology Consortium, is investigating paxalisib in combination with ONC201, again in newly diagnosed DIPG patients after radiotherapy. We expect interim data from this study in CY23. In addition, the company is considering a Phase I clinical trial in another rare childhood brain cancer, atypical teratoid/rhabdoid tumors (AT/RT), following encouraging preclinical data. Kazia has received a rare pediatric disease designation and orphan drug designation (ODD) for paxalisib in AT/RT and in DIPG, offering considerable potential benefits if paxalisib is approved in these indications. For more detail on Kazia’s childhood brain cancer program, see our recent Deep dive into childhood brain cancer.

Kazia Therapeutics is also investigating paxalisib’s utility in the treatment of BMs. At the Society for Neuro-oncology – American Society of Clinical Oncology (SNO-ASCO) brain metastases conference in August 2022, investigators from the MSK will report interim data from a Phase I trial (NCT04192981) of paxalisib in combination with whole-brain radiotherapy (WBRT), for the treatment of patients with solid tumor BMs or leptomeningeal metastases harboring PI3K pathway mutations. The abstract for the oral presentation shows that, at the time of analysis (January 2022), the MTD of paxalisib had been determined at 45mg/day (the lowest planned dose), completing part A of the study. Encouragingly, robust responses were observed in all evaluable patients (n=9, out of 12 enrolled), resulting in a 100% overall response rate within three months of the protocol therapy. We believe this is a positive result for Kazia, however we add the caveat that the reported data are based on a small sample set and therefore care must be taken when interpreting the results.

The investigator-led study was a single-arm, open-label trial, aiming to identify the MTD of paxalisib (doses at 45mg, 60mg and 75mg daily), using a 3+3 dose escalation design. The trial calls for patients (estimated total enrolment n=24) to receive paxalisib in combination with WBRT (30Gy in 10 fractions) over a two-week period. Now that the MTD has been determined at 45mg/day, an additional 12 patients will be enrolled in the part B expansion cohort. We expect top-line data from the study to be communicated in early CY23. Another investigator-led Phase II study of paxalisib in BM is currently underway.

As part of the company’s development program, two separate Phase II trials, investigating the use of paxalisib in the treatment of BM are also underway. One (NCT03994796), led by investigators from the Alliance for Clinical Trial in Oncology is investigating the use of genetic testing to guide treatment for patients with BM (PI3K-mutant participants treated with paxalisib monotherapy) and will measure objective intracranial response rates. Participants may have already failed prior radiotherapy. The second (NCT03765983), led by investigators at the Dana-Farber Cancer Institute, is investigating paxalisib in combination with trastuzumab (HER2 antibody) in the treatment of patients with HER2-positive breast cancer brain metastases.

BM occurs when cancerous cells from a primary tumor enter a patient’s blood stream and settle in the brain tissue, forming a new secondary tumor. It is estimated by the clinical community that c 20% of patients with cancer will develop a BM, but this figure could be much higher as those not displaying neurological symptoms are unlikely to be screened for the condition. Indeed, autopsy studies indicate that the incidence may be as high as 40%. BM is commonly associated with highly prevalent types of cancer, including lung, breast, melanoma, renal and colorectal (Exhibit 1). Notably, BM is diagnosed at a high rate in individuals with lung or breast cancer, two indications with large global patient populations (global lung cancer cases in 2020 were c 2.2 million, breast cancer c 2.3 million).

Development of BM is known to be affected by various factors, including age, sex, race, cancer molecular subtype, tumor source and tumor stage. However, despite advances in identifying high-risk patients, the clinical value of these prognostic factors has not been verified. The prognosis for patients with BM is often poor. Evidence suggests that, if left untreated, median survival for BM patients is only one month and, even with treatment, survival beyond six months is unlikely.

BM begins when primary tumor cells undergo a process called epithelial-mesenchymal transition (EMT). Essentially, EMT allows cells to detach from basal membranes and become more mobile, entering the bloodstream and eventually settling in the brain. Many of the cellular processes that lead to EMT are controlled by the PI3K/mTOR pathway, paxalisib’s target. In addition, aberrations in the PI3K/mTOR pathway can lead to angiogenesis and immune modulation, the combination of which (together with EMT) creates a supportive environment for metastases to form and grow. Mutations affecting this pathway are frequently observed in BM, with one study reporting that 43% of enrolled BM patients had at least one PI3K pathway mutation. Importantly, mutations in the PI3K pathway have been observed in BM independent of primary tumor histology, indicating that distinct therapeutics may be needed to treat BM separately from the originating tumor. Considering this evidence, we believe paxalisib may be well suited to address BM in the subset of patients harboring PI3K/mTOR pathway aberrations.

Surgical resection is the first line of treatment for patients diagnosed with BM. However, its use is limited due to the risk of neurological defects, infection, bleeding and the inaccessible nature of many brain tumors. Despite recent advances in neurosurgery and targeted drug and immunotherapy development, radiotherapy continues to form the basis of many BM treatment regimens. WBRT is a radiological intervention that, as the name suggest, involves applying radiation to the whole brain to kill tumor cells. Another option is the use of stereotactic radiosurgery (SRS), a focused method capable of delivering high doses of radiation to tumor cells with precision. Treatment for patients with BM often includes a mix of surgery, WBRT, SRS and chemotherapy.

The recent Phase I trial is investigating paxalisib in combination with WBRT, which is the most commonly used radiotherapy in patients with BM as it can be initiated quickly, is widely available and can provide palliative relief for patients. It is estimated that c 200,000 patients receive WBRT in the United States every year for the treatment of BM. As an adjuvant post-SRS, multiple studies have shown that WBRT increases the control rate of tumors and decreases the risk of new metastases. However, this seemingly positive effect is not associated with increased overall survival compared to radiosurgery alone. An important consideration for clinicians prescribing WBRT is the procedure’s neurological toxicity. Cognitive decline is often associated with treatment regimens including WBRT, with particularly pronounced effects on episodic memory having been observed. Randomized clinical trials have demonstrated worsened neurological function and cognitive decline after WBRT. For example, a Phase III trial (NCT00377156) investigating the use of SRS with and without WBRT showed that 91.7% of patients who received WBRT experienced cognitive deterioration compared with SRS alone (63.5%). Clinical data also suggest that WBRT can have a significant effect on patient quality of life (QOL). Given the clinical data available, we believe that paxalisib could potentially improve response rates and/or increase overall survival compared to WBRT alone. The demonstration of any potential benefits in safety and patient QOL will be important, in our view.

Due to potential cognitive deterioration and neurotoxicity, US and European guidelines now recommend against the use of WBRT in patients with limited BMs who have already received SRS. Accordingly, SRS is growing in popularity as a method of treatment among clinicians and patients. Randomized clinical data support a less severe neurological side effect profile with SRS alone, although disease control is accepted as better with WBRT. Hence a decision on which radiotherapy regimen to pursue will be based on a multitude of factors, including the number of BM, patient QOL and cancer histology. We anticipate that Kazia may choose to expand paxalisib’s development to investigate combinations with SRS and surgery in patients with defined numbers of BM if clinical results continue to be positive. In our view, this would maximize paxalisib’s commercial potential in BM, if approved.

Conventional cytotoxic chemotherapies generally lack the ability to cross the blood-brain barrier (BBB) and therefore have limited utility in the treatment of BM. Historically, patients with BM were excluded from many clinical trials, which meant that data on the effectiveness of many drugs in this indication were scarce. However, modern advances in targeted therapies and immunotherapies have expanded the role of systemic therapy in the treatment of BM, as many of these drugs can cross the BBB. In particular, next-generation tyrosine kinase inhibitors (TKIs), for example erlotinib and Osimertinib, and immune checkpoint inhibitors (ICIs), for example a nivolumab/ipilimumab combination, have shown promise in the field. We note however, that many of these TKIs/ICIs are already approved for the treatment of the primary cancer and have not been studied in combination with WBRT/SRS, so data may not be directly comparable to the recent paxalisib results in BMs.

Despite the promising clinical data displayed in various TKIs and immunotherapies, guidelines warn against delaying radiotherapy for systemic treatment because of the poor prognosis of patients with BM. In specific cases, off-label use of TKIs and ICIs is occasionally recommended to treat BM. For instance, ASCO, SNO and American Society of Radiation Oncology (ASTRO) guidelines recommend that osimertinib, alectinib or pembrolizumab (anti-PD-L1 antibody) may be offered to EGFR-mutant, ALK-rearranged or PD-L1-high NSCLC patients, respectively, despite the fact that none of these drugs is approved to treat BM. Considering the poor prognosis for BM patients, the potentially debilitating side effects of mainstay radiotherapies and lack of specifically approved therapeutics, we see this as an area of significant unmet medical need.

Given the high rate of BM observed in large market indications (breast cancer, NSCLC, melanoma – see Exhibit 3), we believe there is considerable potential for paxalisib in the treatment of BM. For example, EvaluatePharma expects the market for NSCLC drugs to grow to US$51bn by 2028. Assuming 20% (lower end of estimates) of patients develop BM, this represents a potential market value of c US$10bn in NSCLC alone. Kazia’s development strategy continues to focus on the treatment of less prevalent cancers in adults and children. However, if future data readouts confirm paxalisib’s utility in BM, we expect that Kazia may reassess its strategic focus. In our view, considering the relative stages of development, a shift in strategy is unlikely to affect current clinical programs.

Considering the clear unmet medical need for systemic treatments, the considerable patient populations affected and the encouraging clinical data recently presented on paxalisib, we believe there is a significant potential opportunity for paxalisib’s use in BM. However, to maximize the potential medical and commercial impact of paxalisib in this setting, we believe the company should focus on the following:

We value Kazia Therapeutics at US$151.3m or US$10.86 per basic ADR, previously US$294m or US$22.28 per basic ADR. A breakdown of our valuation is presented in Exhibit 4. Considering the recent news from the Phase III GMB AGILE study, we have adjusted the probability of success for paxalisib in GMB to 20% from 35%. We believe 20% is justified as the results are still blinded and we expect the bar for success for graduation to Stage 2 was set high. We have also adjusted our peak market penetration estimate to 15% from 25% as we expect the GBM AGILE result will likely affect the extent of paxalisib’s approval. With these changes we now expect peak sales for paxalisib in GBM of US$270m, previously US$450, and a risk-adjusted NPV of US$85.6m. Our remaining valuation assumptions are unchanged (for detail, see our recent update).

Kazia reported net cash of US$11.0m (A$15.2m) at end December 2021, which we estimate will fund it into October 2022. In the company’s appendix 4C statement, management reported gross cash of A$7.36m (US$5.33m) at end-June 2022 and a Q422 operating cash burn rate of A$6.5m. We expect that Kazia will need to seek c US$51m in financing (including US$22m in FY23 and US$22m in FY24) to see it past the final clinical readout from the Phase III GBM AGILE study (expected in H2 CY23) and the commercialization of paxalisib. Kazia established a US$35m at-the-market program in April 2022, which may meet a portion of our projected funding needs. At end June 2022, Kazia issued 486,281 American depository shares (ADSs) using this program, at an average price of $6.08. Total gross proceeds amounted to US$2.96m (A$4.26m). Management has stated that fees for this financing facility are c 50% of traditional financing and shares are issued with no warrants or discounts.