Preclinical safety profile of RC88-ADC：a novel mesothelin-targeted antibody conjugated with Monomethyl auristatin E

Abstract Mesothelin (MSLN) is an attractive therapeutic target for antibody drug conjugates (ADC) because of significant differences in expression pattern between diseased and normal tissues. RC88-ADC is a novel ADC, targeting MSLN, and inhibits tumor growth significantly in mice xenograft models. We performed an 11-week repeated dose toxicity study of RC88-ADC via intravenous injection in Cynomolgus Monkeys with an 8-Week recovery period according to International Conference on Harmonization (ICH) S9 and S6(R1). RC88-ADC was administered to groups of 5 male and 5 female monkeys at dose levels of 2.5, 5, and 10 mg/kg/2 weeks, meanwhile vehicle, naked antibody, small molecule groups were set up as the control. 4 animals died in 10 mg/kg group of RC88-ADC. The clinical symptoms mainly included ocular toxicity, weight loss and food intake decrease in the middle and high dose groups of RC88-ADC. RC88-ADC caused dose-related reversible myelosuppression, manifested as hematologic toxicity, which was consistent with the small molecule toxicity profile of its coupling. The highest non-severely toxic dose of RC88-ADC was 5 mg/kg in monkeys after repeated dosing. Nonetheless, the integrated analysis showed that RC88-ADC demonstrated an acceptable safety profile and provided an improved treatment window. These results pave the way for further investigation of RC88-ADC in humans.


Introduction
Mesothelin (MSLN) is a 40 kDa membrane glycoprotein mainly distributed in the monolayers of mesothelial cells constituting the pleura, peritoneum, pericardium, ovarian epidermis, sheath, testicular reticulum, and fallopian tube epidermis (Lv et al. 2019). The biological function of MSLN has not been fully elucidated, mice with MSLN gene knockout showed normal growth and development state (Tapan and Nera 2000). MSLN is overexpressed in a variety of tumors including epidermal mesothelioma (85-90%), pancreatic cancer (80-85%), ovarian cancer (60-65%), lung cancer (60-65%), and its expression in gastric cancer, bidirectional synovial sarcoma and uterine cancer is variable (Morello et al. 2016). The limited expression in normal tissues and overexpression in some tumors makes MSLN an attractive anti-tumor therapeutic target.
At present, chimeric antigen receptor-T cells (CAR-T), monoclonal antibodies and antibody drug conjugates (ADC) are the main therapeutic options against MSLN overexpressing tumors in clinical trials. CAR-T showed significant antitumor efficacy in the treatment of hematological tumors, but minimal efficacy in solid tumors, while monoclonal antibody drugs targeting MSLN were generally well-tolerated in patients but the clinical efficacy was not satisfying. ADC drugs consist of antigen targeting antibodies and cytotoxic payloads, combined through cleavable or uncleavable linkers. The general action mechanism of ADC can be described as follows (Tsuchikama and An 2018), after the ADC is administrated into the blood stream, antibody component of ADC binds to cell-surface antigen receptor to form an ADC-antigen complex, then the complex is internalized, the internalized complex undergoes lysosomal processing and the cytotoxic payload is released inside the cell. The released payload binds to its target, leading to cell death. In comparison with traditional small molecule and monoclonal antibody drugs, ADC drugs have the advantage of high efficacy and low toxicity, for now 12 ADCs have been approved by FDA for tumor treatment, including those targeting CD22, CD30 and HER2 (Yaghoubi et al. 2020). Currently, there are no approved MSLN targeting drugs, considering the unmet clinical need and advantages of ADC drugs, we developed a MSLN targeting ADC drug, RC88-ADC, in which cytotoxic monomethyl auridine E (MMAE) were bond to MSLN targeting antibody through cleavable linkers. In pharmacological studies, the mechanism of action of RC88-ADC has been clearly described, and RC88-ADC showed significant antitumor effects on xenograft models with different levels of MSLN expression (The data have not yet been published). In monkey. The study protocols were approved by the Animal Care and Use Committee of the National Shanghai Center for new drug safety evaluation and research (approval No. IACUC-2017-M-012).

Study design
The tolerated dose of MMAE based ADCs was 12 mg/kg after single administrated or 6 to 9 mg/kg after repeated administrated fSaber, 2015 #165gfJiang, 2020 #253g. So we performed a dose range finding study to determine suitable dose levels for subsequent toxicology study. 6 Monkeys(1/ sex/group) were assigned to 3 groups, and administered RC88-ADC via intravenous infusion at doses of 8 mg/kg, 10 mg/kg, and 12 mg/kg, once every 2 weeks for 3 times totally. All animals in 12 mg/kg group died, and all animals survived in other groups. Accordingly, 10 mg/kg dose was selected as the highest dose to be tested for the subsequent toxicity study. Moreover, considering that the effective dose of RC88-ADC in mouse xenograft model was 1 mg/kg, descending doses using a 2-fold interval factor according to the guideline OECD 409 recommendations was selected. Thus, the low dose was 2.5 mg/kg, higher than the effective dose.
According to International Conference on Harmonization (ICH) S9 and S6(R1), monkeys as relevant animal species to evaluate the safety of RC88-ADC are appropriate. 70 naïve monkeys were randomized by sex to RC88-ADC low, medium and high dose groups at 2.5 mg/kg, 5 mg/kg, and 10 mg/kg, RC88-ADC solvent vehicle control group, 5 and 10 mg/kg RC88 naked antibody groups, 0.1 mg/kg MMAE small molecule group, a total of 7 groups. Animals were dosed 5 mL/kg by intravenous infusion biweekly for a total of 6 doses (Table 1). A recovery period of 8 weeks was set up to observe the recovery of toxicity. All studies were conducted in accordance with the principles of GLP.

Clinical observations
All animals were observed twice daily for morbidity, mortality, and clinical signs, including evaluations of the dose site, skin, fur, eyes, ears, nose, oral cavity, thorax, abdomen, external genitalia, limbs and feet, respiration, movements, urination, defection and behavior changes. Ophthalmic examinations were performed on all animals before first administration, the end of treatment and end of recovery period.
2.5. Body weight, food consumption and body temperature Body weight and food consumption were measured weekly. Body temperature was tested seven times, twice at acclimation period, 4 hours and 24 hours post the first and last dosing, and end of recovery period.

Cardiovascular and respiratory
Electrocardiograms (ECG) (Lead II) examination was conducted once in quarantine period, 4 times in administration period and 1 time in recovery period by ECG-6951E (Japan). ECG measurements including heart rate (HR), RR interval, QT interval, and corrected QT (QTc) were evaluated quantitatively using the Fridericia method. Blood pressure(BP), and respiratory rate were measured at the same frequency as ECG. BP measurements included systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP).

Urinalysis and feces test
Urine and feces test were carried out once in quarantine period, administration period and recovery period respectively. Specific gravity (SG), Glucose, protein bilirubin, ketones nitrite, occult blood clarity (by visual inspection), urobilinogen Color (by visual inspection), white blood cells pH, vitamin C, urine sediment (microscopic) were generated by the Uritest-300 Urine Analyzer or observed by microscopic or naked eyes.

Necropsy and organ weight
Gross necropsy was performed in all animals including unscheduled death/moribund animals. The animals were fasted over night before sacrifice and anesthetized by entobarbital sodium (approximately 30-45 mg/kg, adjusted according to the anesthetic effect), intravenous injection via peripheral veins. After exsanguination, all monkeys were given a complete pathological necropsy. The following organs were weighed from all animals at scheduled necropsy: brain, heart, liver, kidney, adrenal, thymus, spleen, testis, ovary, epididymis, uterine (including cervix and oviduct), thyroid/parathyroid gland. Main organs/tissues (listed in the Table S1) and observable lesions were collected from all animals at scheduled necropsy. Eyeball, optic nerve, testis and epididymis were preserved in the modified Davidson's fixative. Other tissues were preserved in 10% neutral buffered formalin. All collected organs/tissues were embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H-E) for histopathological examination conventionally.

Toxicokinetic analysis
Blood samples were collected per protocol (Table 1). Briefly, samples were collected from animals in the RC88-ADC group and RC88 naked antibody group prior to 1 to 5 dose and at 10 time points (the end of infusion, postdo-se1h,4h,24h,48h,96h,8d,10d) after the first and fifth doses, with a total of 12 samples per animal. Samples were collected from animals in the MMAE group prior to the first and fifth doses and at 8 time points (the end of infusion, postdose 1 h,4h,8h,24h,48h,72h,168h) after dose, with a total of 9 samples per animal. Then samples were centrifuged at 1000 g for 10 minutes at 4 C and stored below À70 C until analysis. Validated test methods were employed to analyze the absolute content in the samples. ADC and total antibodies were measured by enzyme-linked immunosorbent assay (ELISA) and MMAE was measured by liquid chromatographytandem mass spectrometry (LC-MS/MS) fJiang, 2020 #253g. For quantitation of RC88-ADC or total antibodies in serum, mice anti-MMAE antibody (AbMax Biotechnology Co., Ltd) or Human Mesothelin protein (Sino Biological) were coated in 96-well microplates, RC88-ADC or total antibodies in serum bond to the coated substance. Then it was detected using Human IgG-heavy and light chain monkey-adsorbed Antibody Conjugate HRP. TMB was used in staining procedures as substrate for ELISA detection, the reaction was stopped by H 2 SO 4 solution. The absorbance at 450 nm was read by Multimode Reader. The LLOQ was 12.5 ng/mL. The LC-MS/MS system consisted of a SHIMADZU UPLC system and SHIMADZU tandem mass spectrometer (LCMS-8060) were employed to detect MMAE. The analytical column C18 (Kinetex C18, 2.6 lm, 50 mm Â 2.1 mm) was used. The mobile phase was purified water (0.1% Formic acid and 1 mM Ammonium acetate) and Acetonitrile (0.1% Formic acid) for gradient elution. The column temperature was 40 C, flow rate was 0.6 mL/min; autosampler temperature was 8.0 C. The injection volume was 5 lL; Run time was 2.3 min. The LLOQ was 20 pg/mL. Area under the concentration-time curve (AUC) and maximum plasma concentration (C max) were calculated using WinNonlin 8.0 software.

Statistical analysis
Statistical analysis was performed using two-sided tests and the significance level was set at 5% or P 0.05. Data were expressed as mean ± standard deviation (SD) and were processed using Social Science Software Version 23 (IBM Inc. USA) or SAS 9.2 software (SAS Institute, Cary, NC, USA). One-way analysis of variance (ANOVA) or Kruskal-Wallis tests (nonparametric method) were performed according to the significance of Levene's test. An independent sample t-test was used to compare the differences in toxicokinetic parameters between males and females.

Survival and clinical observations
Mortality occurred in high dose of RC88-ADC group and in small molecule toxin group, as detailed below: ADC high dose group: 3 females were found dead or moribund since D10 and 1 male animal died on D37. Small molecule group: 1 female animal was moribund on D10. Three animals in the high dose group of RC88-ADC died before the second dose, so we adjusted the dosage from 10 mg/kg to 7.5 mg/kg, starting from the second dose. During the study, the clinical symptoms associated with the test article were found in the middle, high dose of RC88-ADC groups and small molecular toxin group. The common symptoms included decreased body temperature, decreased activity, abscess of lower abdomen and tachypnea. Specific symptoms of RC88-ADC group were lacrimation excessive, eye discharge, eyelid swollen, conjunctival swollen, swelling of nasal wing.

Body weight, food consumption, and body temperature
During the dosing and the recovery periods: No significant differences in body weights were observed in vehicle control group, naked antibody group, low dose RC88-ADC group and small molecule group (Tables 2 and 3).
In middle dose ADC group: Although no statistical significance was noted in body weights of males and females, compared to the vehicle group; Significant body weight decrease in one female animal was observed from D14, compared with its own baseline value; Largest decrease of body weighs was observed on D63 (D63 body weight vs baseline value II: 3.2 kg vs 4.2 kg, #24%). The animal recovered during the recovery period. There was no significant difference of body weight in male animals of this group.
Decreased food consumption was found in two female animals from middle dose RC88-ADC group and nine animals (4 females and 5 males) from high dose RC88-ADC group during the dosing period. All changes of food consumption recovered at the end of recovery period.
During the dosing and recovery periods, no test articlerelated change of body temperature was noted in males and females of all treated groups.

Hematology and blood chemistry
Dose-dependent hematological toxicity was found in animals received RC88-ADC or MMAE administration. Hematology results showed that RBC, HGB and HCT decreased, and Retic decreased or increased in both ADC and MMAE groups. WBC (mainly in NEUT and LYM) decreased in the ADC mediumand high-dose groups and MMAE groups after dosing, then recovery or compensatory increase were found. In general, the decrease was significant 7 days after dosing but bounced back to a higher level before the next dose. The degree of change in high dose group of RC88-ADC and MMAE group was similar, significantly higher than that in middle dose group of RC88-ADC, which contained the same amount of MMAE as MMAE group. In addition, platelets (PLT) significantly increased in males at the mid dose of ADC and in both sexes at the high dose of ADC during the dosing period. At the end of the recovery period, these changes recovered (Tables 4 and 5, Figures 1 and 2).
During the study, the changes in serum biochemical parameters that were considered to be related with RC88-ADC or small molecule toxins included decreases of P, ALB and ALB/GLO, increased GLO (Table 6). All above changes recovered during the recovery period.
No other test article related change, including markers of cardiotoxicity (Trop I), was observed in the study.

Urinalysis and feces test
During the experiment, no obvious abnormality was found in the urinalysis and feces test of all animals.

Organ weight
Compared with the vehicle control group, the absolute weight and relative weight of thymus in males of low, middle 4  Unit: kg; D: Dosing Period; n: the number of animals in each group; & : Beginning with the second dose, ADC High dose was adjusted from 10 mg/kg to 7.5 mg/ kg. Data were expressed as x± SD; P > 0.05 as compared with Vehicle Control. Statistical analysis was not performed on the data of the recovery period due to n < 3.  A change of percentage compared with the vehicle control group at the same time; -represents a decrease; bold font has statistical difference compared with the vehicle control group; group 1: vehicle control group; group 2: low-dose ADC group; group 3: middle-dose ADC group; group 4: high-dose ADC group; group 5: middle-dose naked antibody group; Group 6: high dose naked antibody-group; Group 7: small molecule dose group.
and high dose groups of RC88-ADC showed a decreased tendency at the end of the dosing period, which was considered to be related to the test article. Statistical differences were observed in individual tissues and organs without correlated histopathological changes, which was not considered to be test article related (Tables S8-S11). Mean RBC levels over time from the repeat-dose monkey study. Dose-dependent decreases and feedback to higher level in RBC were noted in the RC88-ADC treatment groups. Similar pattern was found in MMAE group. The changes fully recovered after 8 weeks of recovery. 'B', baseline; 'R', recovery; Ã p < 0.05, ÃÃ p < 0.01, ÃÃÃ p < 0.001, ÃÃÃÃ p < 0.0001. Figure 1. Mean WBC levels over time from the repeat-dose monkey study. Dose-dependent decreases and feedback to higher level in WBC were noted in the RC88-ADC treatment groups. Similar pattern was found in MMAE group. The changes fully recovered after 8 weeks of recovery. 'B', baseline; 'R', recovery; Ã p < 0.05, ÃÃ p < 0.01, ÃÃÃÃ p < 0.0001. Compared with the vehicle control group at the same time, the percentage of change; -represents a decrease; the bold font has statistical difference compared with the vehicle control group; Group 1: vehicle control group; Group 3: ADC middle dose group; Group 4: ADC high dose group; Group 7: small molecular dose group.

Necropsy and pathology
To unscheduled death, thymus remnant; dark red spot in the lung were found. To scheduled sacrifice, the gross abnormalities associated with the test article were found in the right lung, pericardium and pleura adhesions in one animal of high dose group of RC88-ADC. At the end of the dosing period, tracheal mucosal epithelial atrophy, focal erosion/ulceration, epicardial inflammation, focal myocardial degeneration/necrosis with inflammatory infiltration, pericardial inflammation, mesodermal cell hypertrophy, thymic atrophy, and decreased hematopoietic cells in the bone marrow (sternum), and necrosis and increased mitosis of single epidermal cells of the skin (mammary gland sites and administration sites), degeneration/necrosis of corneal epithelial cells of the eyeballs and eyelids, increased mitoses, and conjunctival inflammation were observed at ADC doses !5 mg/kg. The number of sinusoidal parietal cells and nuclear division of sinusoidal parietal cells/hepatocytes increased in the high dose group of ADC. At the end of the recovery period, epicarditis, pericarditis and blepharoconjunctivitis were observed when ADC was administered !5 mg/kg; Atrophy of tracheal epithelium, pulmonary pleurisy, pleural hemorrhage, reactive hyperplasia of eyelid and cornea, inflammation of corneal stroma and lens degeneration were observed in ADC high dose group (Table 7).
Small molecule group: at the end of the treatment period, thymus atrophy, decreased hematopoietic cells, increased mitotic figures in bone marrow (sternum), increased epidermal single cell necrosis, and increased mitotic figures in skin (dosing site) were noted. At the end of the recovery period, all the above lesions recovered.

Toxicokinetic analysis
To RC88-ADC group, after the first dose, C max of ADC and total antibody increased dose-proportionally from 2.5 to 10 mg/kg, and AUC increased more than dose proportionality. C max and AUC of free MMAE increased more than dose increase. After five doses, the exposures were greatly affected by anti-drug antibodies, and there may be some deviation in the determination of dose proportionality. C max and AUC of ADC and total antibody increased more than the dose increase; C max of free MMAE did not increase dose-proportionally and AUC increased more than dose proportionality.
After multiple doses of low, medium and high doses of RC88-ADC in cynomolgus monkeys, no obvious drug accumulation in ADC and total antibody was observed in each group, but MMAE exposure was significantly higher than that in the first dose group.
After multiple doses of 5 and 10 mg/kg of naked antibody in cynomolgus monkeys. There were no significant gender differences in naked antibody exposure after the first and fifth doses. In the dose range of 5-10 mg/kg, the increase of C max and AUC of naked antibody was less than the increase of dose after the first dose. After the fifth dose, C max of naked Liver, increased sinusoid cells, increased mitotic figure.

Discussion
Due to the progress of antibody engineering and the improvement of the stability of the linker, ADC has attracted the attention of scholars (Carter and Senter 2008, Burris 2011, Foyil and Bartlett 2011, Dan et al. 2018. For now, 12 ADCs have been approved by FDA for treatment of various cancers. Patients have benefited greatly from these agents regardless of prolonged overall survival or improved quality of life (Ducry and Stump 2010, Teicher and Chari 2011, Lambert and Chari 2014. RC88-ADC is a novel MSLN-targeting ADC drug that has shown significant therapeutic efficacy in animal models. In accordance with ICH S9 and S6(R1), we conducted safety studies mainly in cynomolgus monkeys, pharmacologically relevant species. In preclinical safety assessments, RC88-ADC demonstrated no effect on cardiovascular and respiratory systems, but showed a hematologic toxicity phenotype consistent with the effects of small molecule MMAE-microtubule inhibitors. The common toxicities analyzed in the RC88-ADCtreated and MMAE-treated groups were thymic atrophy, decreased hematopoietic cells in the bone marrow, increased mitotic figures, and single-cell necrosis of the skin epidermis.
Based on the literature (Saber andLeighton 2015, p. 165, Jiang et al. 2019) and dose range finding study, we preliminarily inferred that the toxicity of the ADC with same small molecule was similar, therefore, the dose tested for the repeat dose toxicity study was selected on the basis of the single dose toxicity study. The results showed unexpected differences in toxicity, despite similar changes in hematologic toxicity, target toxicity was more pronounced, particularly in the heart and lung, which was consistent with the targets distribution profile of RC88-ADC.
In the repeat-dose toxicity study, 4/10 animals in the high dose group of RC88-ADC died after dosing. Ocular toxicity Table 9. Summary of TK parameters (TAb) in cynomolgus monkeys (Mean ± SD).
Parameter Unit RC88-ADC 2.5 mg/kg (n ¼ 10) RC88-ADC 5 mg/kg (n ¼ 10) RC88-ADC 10 mg/kg (n ¼ 10) ᭝ RC88-0 5 mg/kg (n ¼ 10)  was more pronounced in the middle and high dose groups, besides dose-related histopathological changes such as epicarditis and pneumonia. Ocular toxicity may be related to the rapid proliferation of corneal epithelium, which makes it more susceptible to MMAE.
One animal in the small molecule group died, and all animals survived in the ADC medium dose group with the same small molecule content. Although targeted toxicity was found in this group, hematologic changes associated with small molecules were mitigated and systemic toxicity was low. MMAE exposure (C max and AUC) in the ADC medium dose group were 416-fold and 3-fold lower than that in the MMAE group, respectively (C max 0.25 ng/mL, 103.96 ng/mL, AUC 39 ng/mL Ã hr, 124.9 ng/mL Ã hr, respectively). This proves the design idea of ADC, supporting the hypothesis that ADC improves the efficacy of cytotoxic agents, by improving the therapeutic index. The highest non-severely toxic dose of RC88-ADC was 5 mg/kg in monkeys. Meanwhile, 2.5 mg/kg was defined as the no observed adverse effect level, because no test article related changes were found in animals in this dose group.
RC88-ADC target-mediated toxicities found in animal studies were similar to those in patients, including serositis and pleural effusion reported in the phase I clinical trial of DMOT4039A, which share the same linker and small molecule as RC88-ADC (Weekes et al. 2016). It can be speculated that non clinical results can be well transformed into clinical.
Based on the results of nonclinical studies, the National Medical Products Administration (China) has approved RC88-ADC to conduct a clinical study on 19 November 2018 for treatment of patients with MSLN-positive tumors. A phase I clinical trial is being performed to evaluate the safety, pharmacokinetics and effect of RC88-ADC in subjects with advanced malignant solid tumors(NCT04175847).
In conclusion, nonclinical studies of RC88-ADC have shown significantly improved tolerability over MMAE. We have fully characterized RC88-ADC-related toxicity and found that toxicity is predictable and acceptable, consistent with the pharmacological activity mechanism of MMAE and antigen distribution. These results paved the way for further investigation of RC88-ADC in humans.