a992jeongetal_molecularr

Figure 3. RECIST Responders Showed a Higher Molecular Response Rate P = 0.012 Odds ratio (95% CI) = 0.26 (0.090-0.75) MR+ (N) MR- (N) RECIST Responder 20 5 RECIST Non-responder 50 48 PPV (95% CI) 28% (19-40%) NPV (95% CI) 91% (80-96%) 20/25 (80% MRR) cOR 40/71 (56% MRR) SD 9/26 (35% MRR) 80% of responders per RECIST had a molecular response 91% of patients without molecular response were non-responders per RECIST PD 29% 57% 13% 1% 9% 58% 32% mPD mSD mPR mCR cOR Best Overall Response (RECIST) SD PD NA C1D1 C2D1 C3D1 C1D1 C2D1 C3D1 C1D1 C2D1 C3D1 −100 −50 0 50 100 Visit Pe rc en t c ha ng e in T P5 3 V A F fr om b as el in e (% ) 0 25 50 75 100 Fr ac ti on o f p ati en ts (% ) p = 0.0094 MR+ (N=70) MR− (N=53) A B A. MRR is highest in CR/PR RECIST response and lowest in PD. Dashed lines indicate the staged MR cutoffs. The color of lines indicate molecular response stages defined in Figure 2. B. Clinical response was enriched in MR+ cohort. Statistical significance was calculated using Chi‑squared test and Fisher’s exact test (two‑sided), respectively. Figure 4. Molecular Responders Showed Greater Tumor Size Reduction MR stage mCR mPR mSD mPD Be st p er ce nt c ha ng e fr om b as el in e in s um o f d ia m et er (% ) −100 −30 0 20 >100 Complete molecular responders showed a median target lesion size reduction of 31% based on RECIST criteria MR stage mCR mPR mSD mPD N 26 42 32 20 P-value 5.1e-5 Median tumor size change -30.6% -18.2% -7.3% 5.7% Best percent change in the sum of the diameters of target lesions is shown with color indication of molecular response stage. The inset table shows the median change across MR stages. Statistical significance from Kruskal‑Wallis test. Dashed lines indicate 20% and ‑30% RECIST criteria cutoffs. Figure 5. Molecular Responders Showed a Longer Time‑To‑Progression 70 55 39 20 8 1 0 0 0 53 31 15 7 3 1 1 1 1MR− MR+ Patients at risk 38 35 25 14 6 1 0 0 0 31 27 13 6 3 1 1 1 1MR− MR+ Patients at risk 0 0.25 0.5 0.75 1.0 0 0.25 0.5 0.75 1.0 0 2 4 6 8 10 12 14 16 Months 0 2 4 6 8 10 12 14 16 Months Entire cohort (N=123) Patients with stable disease at the first assessment (SD1A; N=69) Su rv iv al p ro ba bi lit y Su rv iv al p ro ba bi lit y MR+ MR− MR+ MR− MR+ MR- mTTP (months) (95% CI) 5.49 (4.14-6.70) P-valueHR (95% CI) 0.43 (0.29-0.65) 3.8e-5 2.69 (2.56-3.84) MR+ MR- mTTP (months) (95% CI) P-valueHR (95% CI) 6.14 (4.73-8.54) 0.44 (0.25-0.78) 3.8e-3 3.56 (2.73-4.37) TTP is compared between MR+ and MR‑ groups (left) in the entire evaluable cohort (N=123) and (right) in the patients whose first overall response was stable disease (N=69). mTTP and HR with 95% CI are shown in inset tables. P‑values are from log‑ranked test. Figure 6. Complete and Confirmed Molecular Responses Further Stratified Time‑to‑Progression mPR mCR mSD mPD Patients at risk 28 23 19 10 5 0 0 0 0 42 32 20 10 3 1 0 0 0 33 21 11 5 2 1 1 1 1 20 10 4 2 1 0 0 0 0 uMR+ cMR+ cMR− Patients at risk 38 37 29 15 6 1 0 0 0 10 8 3 2 0 0 0 0 0 17 15 6 5 3 1 1 1 1 Entire cohort (N=123) Patients with 2 on-treatment timepoints (N=65) Patients with 2 on-treatment timepoints (N=65) 0 0.25 0.5 0.75 1.0 0 0.25 0.5 0.75 1.0 0 2 4 6 8 10 12 14 16 Months Su rv iv al p ro ba bi lit y 0 2 4 6 8 10 12 14 16 Months Su rv iv al p ro ba bi lit y mPDmSDmPRmCR cMR−uMR+cMR+ mCR mPR mTTP (months) (95% CI) 6.47 (5.42-NA) 4.14 (3.25-6.14) mSD mPD 2.73 (2.56-4.14) 2.69 (1.41-4.11) P-value mCR vs mSD/mPD 2.2e-5 HR (95% CI) mCR vs mSD/mPD 0.30 (0.17-0.54) mCR mPR Confirmed MR+ 16 22 mSD mPD 0 0 3 4 3 0 0 0 10 7 cMR+ uMR+ mTTP (months) (95% CI) 6.14 (5.39-8.48) 3.84 (3.25-NA) cMR- 2.76 (2.66-7.72) P-value cMR+ vs cMR- 0.014 HR (95% CI) cMR+ vs cMR- 0.46 (0.24-0.87) Unconfirmed MR+ Confirmed MR- Median TTP was 6.14 months for confirmed MR compared with 2.8 months for cMR- Complete molecular response (mCR) and confirmed molecular response appeared independent Median TTP for mCR was 6.5 months, compared with 2.7 months in mPD TTP is illustrated (left) across staged MR groups in the entire evaluable cohort (N=123) and (right) across confirmed MR groups in the patients whose C2D1 and C3D1 samples are both available (N=65). Confirmed MR+ (cMR+) is defined as MR+ at both C2D1 and C3D1 timepoints, unconfirmed MR+ (uMR+) as MR+ at either timepoint, and confirmed MR‑ (cMR‑) as MR‑ at both timepoints. mTTP and HR with 95% CI are shown in inset tables. P‑values are from log‑ranked test. Figure 7. Molecular Response Preceded RECIST Outcome in Patients with Stable Disease at the First Assessment 0 6 12 18 24 30 36 42 48 54 60 66 72 Weeks from C1D1 Patients with stable disease at the first assessment (SD1A; N=69) Group 1 Group 2 Group 3 Group 4 Group 5 RECIST response Molecular response Response Stable disease Not evaluable Progression mCR/mPR mSD mPD Group Number of patients MR BOR MR-BOR correlation Lead time to BOR (weeks) Group 1 5 MR+ cOR Yes 14 (12-19) Group 2 2 MR- cOR No - Group 3 19 MR+ SD No - Group 4 26 MR- SD Yes 8 (2-31) Group 5 17 MR+/MR- NAa - - • MR+ identified 14 weeks prior to response (Group 1; n=5) • MR- identified 8 weeks prior to progression (Group 4; n=26) Timepoints of MR and RECIST overall response calls are shown for the SD1A cohort (N=69) (Figure on left). Patients were classified into 5 groups based on MR and BOR (Table on Right). For the groups in which MR predicted clinical response, median time benefit was calculated. aNot available due to treatment ongoing or discontinued without progression. RESULTSBACKGROUND • Azenosertib is a novel selective WEE1 kinase inhibitor showing promising activity in the treatment of gynecologic tumors, including HGSOC1,2 • The clinical development of azenosertib in HGSOC is conducted via several efficacy studies using objective response rate and progression‑free survival as primary and secondary endpoints • Molecular response (MR), as measured by longitudinal changes of tumor DNA fraction in cell‑free DNA (cfDNA), can represent a cost‑effective, and early efficacy endpoint, which needs further evaluation in HGSOC clinical studies METHODS Figure 1. (A) Clinical Setting for MR Analysis in Azenosertib‑Treated HGSOC. (B) Description of Azenosertib‑Treated HGSOC Cohort for MR Analysis Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 ••• 0Time (week) HGSOC 3 6 Plasma cfDNA (TP53 VAF) Azenosertib treatment Radiological tumor assessment (RECIST v1.1) 9 12 HGSOC patients treated with ≥300 mg of azenosertib (QD) on 5:2 intermittent dosing (N=220) Excluded (N=64) • Plasma not collected (N=55) • cfDNA NGS assay pending (N=9) Excluded (N=17) • cfDNA NGS assay QC failed Excluded (N=16) • No baseline TP53 variant detected Plasma cfDNA NGS assay processed (N=156) Eligible for MR analysis (N=139) Evaluable for MR analysis (N=123) A B A. Retrospective MR study was designed with HGSOC patients enrolled in azenosertib studies, ZN‑c3‑001 (NCT041583363), ZN‑c3‑005/DENALI (NCT051288254) and ZN‑c3‑006/MAMMOTH (NCT051988045). Patients treated with monotherapy at a total daily dose of 300 mg or higher of azenosertib (QD) on a 5:2 intermittent dosing schedule were selected. Longitudinal cfDNA collection was implemented in ZN‑c3‑001 after a protocol amendment in 2023 so only includes a subset of patients enrolled in this trial. Plasma samples were collected every 3 weeks and analyzed by Tempus xF+ Liquid Biopsy Assay to calculate TP53 variant allele fraction in cfDNA. Radiological tumor assessment (RECIST v1.1) was conducted every 6 weeks. Efficacy was extracted from an unlocked clinical database (data cutoff: 1/27/2025). B. All patients were treated with azenosertib monotherapy at a total daily dose of 300 mg or higher (QD) on a 5:2 intermittent dosing schedule. Among 139 patients whose plasma samples were properly collected and successfully analyzed, 123 patients showed TP53 variants detectable at baseline and were considered the evaluable cohort. On‑treatment plasma samples were collected at C2D1 and/or C3D1 to evaluate molecular response on treatment. Clinical response endpoints including best overall response, best tumor size reduction and time‑to‑progression (TTP) were used to validate molecular response. acOR defined as confirmed response per RECIST 1.1 criteria in subjects with at least one post‑baseline assessment. Figure 2. TP53 VAF Change Distribution and MR Definition MR stage Definition Molecular complete response (mCR) mVAF change = -100% Molecular partial response (mPR) -100% < mVAF change < -50% Molecular stable disease (mSD) -50% ≤ mVAF change ≤ 0% Molecular progressive disease (mPD) 0% < mVAF change MR stage mCR mPR mSD mPD 0 10 20 30 40 −100 −50 0 50 >100 Percent change in TP53 VAF from baseline (%)a N um be r of p ati en ts mCR mPR MR+ MR- mSD mPD Distribution of percent change in TP53 variant allele fraction at the earliest available timepoint (either C2D1 or C3D1) is shown with the ranges of different MR stages (left). The definition of different MR stages are described in the table (right). MR positivity includes mCR and mPR, whereas MR negativity includes mSD and mPD. aMedian VAF used in multiple TP53 variants detected. CONCLUSIONS • The study presents the largest evaluation of cfDNA‑based MR in HGSOC treated with an investigational drug (N=123 patients) • Clinical validity: cfDNA TP53‑based MR showed significant associations with BOR, RECIST tumor size change, and TTP – In particular, complete reduction or clearance of TP53 VAF identified patients with the greatest reduction in tumor size and longest TTP • Clinical utility: MR may predict outcome in HGSOC patients with stable disease at first RECIST assessment – Molecular responders had longer TTP – MR established within the first or second cycle of treatment provided an early indication of RECIST response (median 14 weeks) or progression (median 8 weeks) • Response characterization: Complete MR and confirmed MR appeared independent References 1. Pasic A, et al. Presented at the AACR 2022 Annual Meeting. Abstract number CT148. 2. Liu J, et al. Presented at the ASCO 2023 Annual Meeting. Abstract number GOG‑3072. 3. https://clinicaltrials.gov/study/NCT04158336. 4. https://clinicaltrials.gov/study/NCT05128825. 5. https://clinicaltrials.gov/study/NCT05198804. Acknowledgments This study is sponsored by Zentalis Pharmaceuticals, Inc. Editorial support for this poster was provided by Second City Science, LLC. Additional Information For more information on this study, visit www . zentalis . com or contact jjeong @ zentalis . com. Abbreviations 5:2, 5 days on, 2 days off; BOR, best overall response; CD, cycle day; cfDNA, cell‑free DNA; CI, confidence interval; cMR, confirmed molecular response; cMR+, confirmed molecular response positivity; cMR‑, confirmed molecular response negativity; cOR, confirmed objective response; HGSOC, high‑grade serous ovarian cancer; HR, hazard ratio; mCR, molecular compete response; mPD, molecular progressive disease; mPR, molecular partial response; MR, molecular response; mSD, molecular stable disease; mTTP, median time‑to‑progression; mVAF, median variant allele fraction; NA, not assessable; NE, not evaluable; NGS, next generation sequencing; NPV, negative predictive value; ORR, objective response rate; PD, progressive disease; PPV, positive predictive value; QC, quality control; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease; SD1A, stable disease at first assessment; TTP, time‑to‑progression; uMR+, unconfirmed molecular response positivity; VAF, variant allele fraction.PRESENTED AT: American Association for Cancer Research (AACR), April 25‑30, 2025, Chicago, Illinois, USA Poster #3254 Cell‑Free DNA Molecular Response Predicts Clinical Efficacy in HGSOC Patients Treated With Azenosertib Jinkil Jeong1, Jianhui Ma1, Monah Abed1, Heekyung Chung1, Doris Kim1, Nandini Molden1, Divya Rajendran2, Chang Shim1, Danielle D. Jandial1, Mark R. Lackner1, Fiona Simpkins3, Funda Meric‑Bernstam4, Leslie M. Randall5, Olivier Harismendy1 1Zentalis Pharmaceuticals, Inc., San Diego, CA, USA; 2Lyda Hill Department of Bioinformatics, UT Southwestern Medical Center, Dallas, TX, USA; 3Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Penn Ovarian Cancer Research Center, University of Pennsylvania, Philadelphia, PA, USA; 4Department of Investigational Cancer Therapeutics, University of Texas MD Anderson Cancer Center, Houston, TX, USA; 5Massey Comprehensive Cancer Center, VCUHealth, Richmond, VA, USA. HGSOC evaluable for MR analysis (N=123) Dose at 5:2 intermittent (mg) N (%) 300 24 (19) 350 3 (2) 400 96 (78) On‑treatment timepoint N (%) C2D1 47 (38) C3D1 11 (9) C2D1 and C3D1 65 (53) Best overall response (BOR) per RECIST 1.1 N (%) Confirmed Objective Response (cOR)a 25 (20) Stable disease (SD) 71 (58) Progressive disease (PD) 26 (21) NE 1 (1) Median follow‑up, months: 8.61 Exhibit 99.2