Stacey A. Cohen

PURPOSE: Chemotherapy-induced thrombocytopenia (CIT) is a challenging and common complication of cytotoxic chemotherapy in patients with GI cancers. There is no widely available approved treatment. Avatrombopag is a potent, second-generation oral thrombopoietin receptor agonist promising for CIT management. METHODS: We conducted a multicenter, randomized, double-blind, investigator-initiated US trial of avatrombopag or placebo (1:1) in GI cancer patients with persistent CIT (platelets ≤85 × 10 ⁹ /L on day 1 of a chemotherapy cycle despite adequate time to recover from the prior cycle). The primary end point was successful correction of CIT and prevention of recurrence (defined as recovery of platelets to ≥100 × 10 ⁹ /L, no chemotherapy dose modification/delay, and prevention of CIT recurrence [platelets ≥100 × 109/L] at end of the on-study cycle). RESULTS: The trial was closed to enrollment by the data and safety monitoring board at the planned interim analysis after meeting prespecified stopping criteria for efficacy. Sixteen of 23 patients (70% [95% CI, 47 to 87]) receiving avatrombopag achieved the primary end point versus 4/24 patients (17% [95% CI, 5 to 37]) receiving placebo ( Z = 3.67, P < .001). The median (IQR) platelet count at the end of the on-cycle treatment period was 157 (136-202) × 10 ⁹ /L with avatrombopag versus 72 (68-134) × 10 ⁹ /L with placebo. Adverse events (AEs) and serious AEs occurred in 78% and 17% of patients receiving avatrombopag and 46% and 0% of patients receiving placebo, respectively. No serious AEs were study drug-related. There were no treatment-related AEs leading to death or discontinuation of study drug. CONCLUSION: In this randomized, placebo-controlled trial, avatrombopag was efficacious in the management of persistent CIT in patients with GI cancers. These findings are promising for a common, challenging oncologic complication that prevents delivery of full-dose, on-time, cancer-directed therapy.

Background: Circulating tumor (ct)DNA is a prognostic biomarker in gastrointestinal malignancies. In rectal cancer, its utility to inform perioperative management and predict recurrence, particularly in patients undergoing non-operative management (NOM), remains unclear. Studies are needed to clarify how post-neoadjuvant therapy (NAT) and post-surgical ctDNA status correlate with clinical outcomes in localized rectal cancer. Methods: We retrospectively analyzed ctDNA data from 220 patients with rectal cancer using a personalized tumor-informed assay (Signatera™, Natera, Inc., Austin, TX, USA). Of these, 148 (67.3%) underwent NAT followed by surgery, and 72 (32.7%) underwent NAT followed by NOM. We assessed associations between post-NAT ctDNA status and survival outcomes. In the surgical cohort, we examined associations between post-operative ctDNA status and clinical response, pathological response, survival outcomes, and NAR scores. Results: In the surgical cohort, ctDNA positivity at the post-operative MRD timepoint was a strong predictor of recurrence, with an 88.3% relapse rate compared to 11.5% in ctDNA-negative patients (p < 0.001). Among the 64 NOM patients with post-NAT ctDNA, 21.9% (14/64) were ctDNA-positive, of whom 100% (14/14) relapsed (92.9% local-only), 13 relapsed by the time of data cut-off, and one relapsed 8 months after the cut-off. Only 10% (5/50) of the ctDNA-negative NOM patients experienced local recurrence (p < 0.0001). ctDNA positivity post-NAT was associated with inferior DFS (p = 0.003). Conclusion: ctDNA was a strong predictor of recurrence in rectal cancer, including in NOM settings. In NOM patients, ctDNA detected local recurrences, highlighting its potential to guide post-NAT surveillance and treatment.

Supplemental Table 2: Summary patient characteristics for Comparison Set 1

Supplemental Figure 3: EVL methylation levels of normal colon for patients with no history of adenoma or cancer after initial colonoscopy (left group) and for levels patients with an adenoma or colorectal cancer detected after initial (middle group). Patients with no follow-up colonoscopies (right group) were not included in the analyses for Comparison Set 2. %mEVL=percentage of DNA that is methylated at the EVL locus assayed.

Supplemental Figure 1. Flowchart of patient selection. * Subjects with hereditary conditions or inflammatory bowel disease (IBD) were excluded. A flowchart showing the selection process for the subjects included in the studies is shown. The primary exclusion criteria that affected subject selection were lack of serial colonoscopy data and lack of tissue from the normal colon

Supplemental Figure 2: Methylation levels of normal colon mucosa in patients with no history of adenoma or cancer ever (group on left) and levels for patients with a history of adenoma(s) or colorectal cancer at initial or a follow-up colonoscopy (group on right) in Comparison Set 1. %mEVL=percentage of DNA that is methylated at the EVL locus assayed.

Supplemental Table 7: Summary statistics for methylated EVL measurements in Comparison Set 3

Supplemental Table 5: Summary statistics for methylated EVL measurements in Comparison Set 1

Supplemental Table 4: Summary patient characteristics for Comparison Set 3