A modular, first time in human, open label, multiple dose, accelerated escalation with cohort expansion study of the safety and pharmacokinetics of intravenous infusion of CP-506, a tumor agnostic Hypoxia Activated Prodrug in patients with HRD/FAD solid tumors or tumor types with high incidence of HRD/FAD in monotherapy or in combination with carboplatin or patients with solid tumor and oligoprogressive disease receiving immune checkpoint inhibitors (ICI): a phase I-IIa

Hypoxia Activated Produgs (HAP): Given its pivotal role in tumour progression
and resistance to conventional therapies, several strategies have been
developed to overcome tumour hypoxia, including the use of hypoxia-activated
prodrugs (HAPs). Over time, the characteristics of an optimal HAP necessary to
achieve maximal therapeutic potential have been defined and refined. Firstly, a
HAP should be reductively metabolized at oxygen levels sufficiently low to
prevent activation under physiological oxygen conditions, which range from 1-9%
O2 depending on the tissue. Secondly, its active metabolites should be able to
kill non-proliferating cells typically present in hypoxic tumour regions.
Finally, it should possess a sufficiently large bystander effect, i.e. the
ability of the effector molecule to diffuse from the hypoxic activation zone to
neighboring tumour cells with intermediate oxygen levels.

Several HAPs have already been evaluated in both the preclinical and clinical
setting. Evofosfamide (TH-302), a 2-nitroimidazole-based phosphoramidate
mustard HAP, demonstrated antitumour activity as a single agent, which
correlated well with the degree of tumour hypoxia. When combined with
conventional anticancer therapies, its therapeutic efficacy was significantly
increased, both preclinically and clinically. However, in two large Phase 3
trials in pancreatic cancer (MAESTRO; NCT01746979) and soft tissue carcinoma
(TH-CR-406/SARC021; NCT01440088), addition of evofosfamide to conventional
therapies failed to improve overall survival. Potential reasons explaining this
failure include a lack of patient stratification based on tumour oxygenation
status, unexpected lowering of pharmacokinetic exposures due to reformulation
of the prodrug during transition from Phase II to Phase III trials, and a
limited bystander effect of the active metabolite of evofosfamide. PR-104, a
phosphate pre-prodrug of the dinitrobenzamide nitrogen mustard HAP PR-104A,
demonstrated selective activation in severe hypoxia in a wide range of cancer
cells in vitro (PR-104A) and human tumour xenografts in vivo (PR-104).
Furthermore, combination of PR-104 with radio- or chemotherapy enhanced its
antitumour effects. Based on these highly promising results, clinical trials
were initiated. However, safety and tolerability of PR-104 assessed in Phase I
trials demonstrated dose-limiting thrombocytopenia and neutropenia. Subsequent
preclinical studies elucidated that PR-104A can be activated independent of
tissue oxygenation by human two-electron aldo-keto reductase 1C3 (AKR1C3),
highly expressed on myeloid progenitor cells, explaining the observed
dose-limiting toxicity. Further clinical evaluation in hematological cancers
revealed that myelotoxicity was the only serious adverse event preventing
dose-escalation of PR-104 administration.

Based on this rationale, PR-104 was redesigned at the University of Auckland to
give rise to CP-506 , a next-generation HAP with more favorable properties.
First, CP-506 is designed to be resistant to AKR1C3 metabolism. Second, CP-506
is a water-soluble piperazine salt, avoiding the need for a phosphate
solubilization strategy as utilized for PR-104, and is thus resistant to
glucuronidation of the alcohol present in PR-104A, potentially leading to an
improved pharmacokinetic profile due to slower clearance. Glucuronidation of
PR-104A has been shown to be a major clearance pathway in humans hence
compromising plasma half-life. Third, unlike PR-104, CP-506 has the potential
to be orally bioavailable. Fourth, the di-nitro substituents adopted for
PR-104, where the nitro group ortho to the mustard contributes to facile
metabolic loss via self-alkylation, is avoided as CP-506 is a mono-nitro HAP.
Finally, the physiochemical properties of CP-506 and its metabolites readily
permit a bystander effect, unlike evofosfamide where the released
phosphoramidate mustard cytotoxic metabolite carries a negative charge and is
essentially cell entrapped.

This study has been transitioned to CTIS with ID 2023-506818-45-00 check the CTIS register for the current data.

The study*s primary objective is to assess the safety and tolerability of
CP-506 alone or in combination therapy. The 3 different modules explore the
administration of CP-506: alone (Module 1) or in combination with Carboplatin
(Module 2) or in combination with ICI (Module 3). Each module consists of a
dose escalation Part A which will have a maximum of 24 patients and determine
safety, tolerability with the determination of the Maximum Tolerated Dose (MTD)
and Recommended Phase 2 Dose (RP2D), pharmacokinetics, pharmacodynamics and
preliminary efficacy with the determination of the minimal biological effective
dose. Each module has an associated optional part B which consists of a dose
expansion cohort of 10 (Module 1B) or 22 (Module 2B and 3B) patients. These
expansion cohorts have as primary objective further safety assessment of CP-506
and as secondary objective efficacy and pharmacodynamics.

This study has been designed as a First in Human, early phase (I/IIa), 3
modules, 2 parts (A- dose escalation and B- cohort expansion), open-label,
uncontrolled, multi-center, multiple dose, accelerated 3+3 dose escalation
study.
The study will consist of 3 study modules. Module 1 is monotherapy with
accelerated dose escalation. Modules 2 and 3 are also accelerated dose
escalation modules, but with a combination of CP-506 with either Carboplatin or
ICI. Initiation of modules 2 and 3 are dependent on the decision by the
Independent Data Monitoring Committee (IDMC) based on the safety, tolerability
and feasibility information from the study as a whole.
Each study modules will consist of a Part A (dose finding) and an optional Part
B (cohort expansion). The option to start Part B will be decided by the IDMC
based on safety, tolerability and feasibility information from the study as a
whole. The expansion cohorts will explore preliminary efficacy.
Modules 2 and 3 will start after the end of module 1. The dose to start modules
2 and 3 is the *minimum biological effective dose* (MBED) as the dose that
results in 1) a decrease of 20% of hypoxia radiomics score on sequential CT of
at least one lesion OR 2) a decrease of 20% of the initial tumour volume of at
least one lesion.
MBED will be assessed by CT scan at visit 12, 17 and 18. After observation of
an MBED in three patients, the data supporting the MBED assessment will be
reviewed by the IDMC for confirmation. Based on this evaluation together with
the collected safety data, the IDMC will propose the start of modules 2 and 3
after the conclusion of module 1.

A maximum of 126 patients with advanced or metastatic cancer will be recruited
in this study.
• Module 1 - Monotherapy: up to 24 patients for Part A and 10 patients for Part
B for whom no standard of care or known effective treatment options are
available and with cancers that show a considerable incidence of Homologous
Recombination or Fanconi Anaemia DNA repair defects (HRD/FAD)
• Module 2 - Combination with Carboplatin: up to 24 patients for Part A and 22
patients for Part B with cancer types for which carboplatin is the standard of
care.
• Module 3 - Combination with ICI: up to 24 patients for Part A and 22 patients
for Part B with advanced or metastatic cancer for whom standard of care immune
checkpoint inhibitor (ICI) are currently administered for at least 6 months,
would still be administered according to the treating clinician, outside any
clinical trial but resulted in oligoprogression.

Each patient will receive their assigned dose of CP-506 every 21 days for 3
cycles.

20 STRUCTURED RISK ANALYSIS
20.1 Potential issues of concern
20.1.1 Level of knowledge about mechanism of action
CP-506 belongs to the category of Hypoxia Activated Prodrugs such as
Tirapazamine, AQ4N, PR-104, EO9 (apaziquone), TH-302 (evofosfamide) and SN30000
(89) but differs from these in terms of its chemical nature being a
water-soluble mesylate salt. Hypoxia Activated Prodrugs are designed to be only
active in hypoxic conditions such as inside solid tumors. The requirement of
hypoxia for CP-506 activation has been supported in vitro by an HCR (hypoxic
cytotoxic ratio) ranging from 4 to 157 in a panel of 51 tumor cell lines tested
in a cell-based potency assay. In animals, CP-506 showed a consistent
anti-tumor activity in 13 tumors models except in two tumor models whose tumor
did not present any regions of hypoxia, confirming the requirement of hypoxia
for CP-506 activity in vivo. CP-506 is activated in hypoxic regions and leads
to DNA damage of tumor cells with a bystander effect due to local metabolite
redistribution (IMPD section 3.2.4.2). CP-506 is not activated by endogenous
reductases, unlike its predecessor PR-104 (IMPD section 3.2.4.1.3). CP-506
pharmacokinetics has been assessed in animals.
Animal pharmacokinetics summary:
• Single-dose infusion of CP-506 revealed a fast elimination in rats and dogs,
suggesting lack of accumulation.
• The volume of distribution was intermediate in rats and dogs, consistent with
extravascular/tissue distribution.
• Preliminary results in rats and dogs were consistent with urinary excretion
of the parent compound.
• Protein binding in human plasma was about 50 % and slightly higher in rat and
dog plasma.
• Studies on hepatic metabolism revealed a fast metabolism of CP-506 in human,
rat, dog, monkey and minipig liver microsomes, with a number of primary and
secondary metabolites; rat and dog microsomes showed the closest similarity to
human material.
• Like liver microsomes, human hepatocytes showed a fast metabolism of CP-506.
Of the microsomal metabolites, only the most prominent product (M1, formed by
dealkylation of the bromide arm), was detected in hepatocytes. As also this
compound showed a fast decay, it may be considered that primary CP-506
metabolites undergo rapid conversion to secondary (most probably inactive)
products. For the most part, very similar results were obtained with rat and
dog hepatocytes.
• Most metabolites are half-mustards and hence, unlike the activated compounds
(CP-506H and CP-506M), would be unable to cross-link DNA.
*
20.1.2 Previous exposure of human beings with the test product(s) and/or
products with a similar biological mechanism
This is the first trial in humans. The mechanism for CP-506 to clear tumor
cells in humans is expected to be similar to the mechanism seen in animal
models (IMPD section 3.2.4).

20.1.3 Can the primary or secondary mechanism be induced in animals and/or in
ex-vivo human cell material?
The primary and secondary mechanisms of action have been investigated in human
tumor cell lines, mice, rats, dogs and xenograft human tumor models (IMPD
section 3.2.4).

20.1.4 Selectivity of the mechanism to target tissue in animals
The mechanism of action is based on activation upon hypoxic conditions. CP-506
exerts its cytotoxicity by the induction of DNA double-strand breaks and DNA
interstrand crosslinks. CP-506 selectively kills hypoxic tumor cells in
addition to inhibiting growth and clonogenic survival of hypoxic cells (IMPD
section 3.2.4.1.5 and 3.2.4.1.6). CP-506 effect is dose dependent. CP-506 shows
resistance to off-target activation via AKR1C3 metabolism, resulting in highly
specific hypoxia-dependent metabolism and activation (IMPD section 3.2.4.1.3).

20.1.5 Analysis of potential effects
CP-506 has demonstrated anti-tumor activity in a wide range of pre-clinical
tumor models (sensitive to alkylating agents as well as insensitive models).
CP-506 can profoundly impact tumor growth and induce regression in models
sensitive to cisplatin, confirming its alkylating agent properties. Anti-tumor
efficacy is dose-dependent and highly effective at doses of 600 mg/kg or
higher. The mode of action of CP-506 has been demonstrated and it requires the
presence of hypoxia for the activation of CP-506. CP-506 induces DNA damage in
hypoxic areas of tumor and it has a profound anti-tumor effect in pre-clinical
models sensitive to alkylating agents such as cisplatin. Repeated dosing dense
regimens proved to induce tumor regression and the control of tumor growth
overtime and CP-506 sensitizes tumors to chemotherapy and to immunotherapeutic
approaches.
Based on data from toxicology studies the following potential targets have been
identified: 1) the cardiac function, 2) the haematopoietic system, 3) the
liver, 4) the kidney and 5) the gastro-intestinal tract (IB section 4.3.3).
Safety pharmacology studies indicate that CP-506 affect cardiac repolarization
linked to selective inhibition of IKr potassium channels. This effect
translates into prolongation of cardiac action potential in vitro and a trend
to QT prolongation in dogs.
In toxicity studies, adverse systemic effects occurred, mainly in
haematological system. They consisted in dose-dependent and reversible effects
on white blood cells (particularly lymphocytes and neutrophils), red blood
cells, hemoglobin, hematocrit, reticulocytes and platelets. These findings
correlated with hypoplasia in hematopoietic organs. The gastrointestinal tract,
kidney and liver were identified as possible additional toxicological targets
of CP-506. Overall, the adverse effects of CP-506 are similar to those
generally observed after treatment with nitrogen mustards. In the case of
CP-506 they are consistent with a probable activation by hypoxic conditions in
the target organs.
The NOAEL dose was considered to be 200 mg/kg b.w./day in rats but could not be
established in dogs. The MTD of CP-506 in rats and dogs, was considered to be
400 mg/kg b.w./day and 150 mg/kg b.w./day respectively. The equivalent doses
expressed by body surface were calculated according to FDA recommendation and
correspond to 1200 mg/m2 and 2400 mg/m2 as respectively the NOAEL and MTD in
rats and 3000 mg/m2 as the MTD in dogs. Mortality occurred in rats at 600 mg/kg
which correspond to 3600 mg/m2.
Based on the FDA guideline
S9-Nonclinical-Evaluation-for-Anticancer-Pharmaceuticals, the starting dose
should be calculated as 1/10th of STD10 in rodent or 1/6th of HNSTD in
non-rodent, whichever is lower. In this case we would have 360 mg/m² or 500
mg/m² respectively. Therefore, the starting dose in clinical trial is currently
fixed as 360 mg/m².
Based on the non-clinical studies summarized herein, predictions of potential
adverse effects in the clinic include mainly effects on cardiac rhythm and
haematology. Effects on liver, kidney and gastro-intestinal tract are also
possible. ECG should be regularly measured for evaluation of possible QT
prolongation and/or arrhythmic events, necessitating rapid intervention. Blood
should be regularly sampled for evaluating effects on blood cells population
and markers of liver and/or kidney adverse effects. Possible effects on the
gastro-intestinal function need to be carefully monitored. In particular
presence of blood in the faeces should be evaluated.
No local adverse effects are expected if CP-506 is administered by strict
intravenous route. Nevertheless, the injection point should always be carefully
examined during treatment and on the following days.
Overall, therefore premonitory endpoints for adverse effects of CP-506 are
available for clinicians to assess arising toxicity in human patients

20.1.6 Pharmacokinetic considerations
Pharmacokinetics of CP-506 have been analysed in mouse, rat and dog plasma (IB
section 4.2). Single-dose infusion of CP-506 revealed a fast elimination in