The study protocols and all amendments were approved by the institutional review board or ethics committee of each participating center. All patients provided written informed consent. The trial was performed per the Declaration of Helsinki, the Good Clinical Practice guidelines of the International Conference on Harmonization, and relevant laws and directives in China. This trial is registered with CLINICALTRIALS. GOV (NCT05262335).

This phase II multicenter, multicohort, single-arm, investigator-initiated exploratory study (ALTER-G-001) enrolled adult patients (aged 18 to 75 years) who had histologically or cytologically confirmed GI cancer with unresectable liver metastasis (with or without extrahepatic metastases). Cohort A consisted of patients with stage IV CRC, while Cohort C consisted of patients with GI cancers other than CRC and ESCC (such as gastric cancer, pancreatic cancer, and biliary tract cancer). Included patients also had at least one measurable metastatic liver lesion per the Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1. Other key eligibility criteria were an Eastern Cooperative Oncology Group (ECOG) performance status of 0–1, adequate hematologic and organ function, and a life expectancy of ≥12 weeks. Patients should have received no prior systemic therapy, including chemotherapy, targeted therapy, or immune therapy. Patients with recurrent disease, those who received prior neoadjuvant chemoradiotherapy plus curative surgical resection, and those treated with prior adjuvant chemoradiotherapy or curative concurrent chemoradiotherapy were eligible if treatment was completed more than six months before enrollment. The key exclusion criteria were a history of primary malignant cancer, with the exception of cured nonmelanoma skin cancer or cured lentigo maligna melanoma, and cured carcinoma in situ; active bleeding in the primary and/or metastatic lesions within two months of enrollment; arterial and/or venous thromboembolic events within six months of enrollment; current thrombolytic or anticoagulation treatment; and GI diseases with bleeding diathesis at risk of bleeding, as determined by the investigators, perforation, obstruction, or fistulation. Patients with brain and leptomeningeal metastases were ineligible. Patients who had received radiation therapy or surgery within 30 days of enrollment were also excluded. The full eligibility criteria are provided in the trial protocol.

For induction therapy, patients in Cohort A received 12 mg oral anlotinib (Chia Tai Tianqing Pharmaceutical Co., Ltd.) once daily on days 1 to 14 of each cycle31 plus the CAPEOX regimen (850 mg/m2 capecitabine given orally twice a day on days 1 to 14 of each cycle and 130 mg/m2 oxaliplatin given intravenously on day 1 of each cycle). Patients in Cohort C received 12 mg oral anlotinib once daily from days 1 to 14 of each cycle plus an SOC chemotherapeutic regimen at the discretion of the investigators. Each cycle lasted three weeks. The resectability of the liver metastases was assessed via multidisciplinary consultation after every two cycles during the six cycles of induction therapy. Patients who were eligible for surgery discontinued treatment four weeks prior to surgery or at the discretion of the investigators. Local treatment of liver metastases was also allowed. Patients who were ineligible for surgery and who achieved disease control (CR, PR or stable disease [SD]) per RECIST, version 1.1, received maintenance therapy consisting of 12 mg of anlotinib once daily for two weeks with one week off plus metronomic capecitabine chemotherapy (500 mg, twice daily, per os) (Cohort A), or anlotinib plus chemotherapy, in which the chemotherapy was preferably capecitabine metronomic chemotherapy (500 mg, twice daily, per os) (Cohort C). Maintenance therapy was continued until disease progression, death, or unacceptable toxicity.

Anlotinib was reduced by up to two doses (12 mg/d to 10 mg/d and 10 mg/d to 8 mg/d). The protocol-defined criteria for dose modification of anlotinib, capecitabine or other chemotherapeutic agents are available in the trial protocol.

Responses were evaluated radiologically with contrast computed tomography (CT) or magnetic resonance imaging (MRI) by investigators every two cycles during induction, every three cycles thereafter and at the end of the study per the RECIST (v1.1) criteria. CR and PR were confirmed 4 weeks apart, and SD should persist for at least 8 weeks. Patients were followed-up every six months to assess survival until the death of the patient or until the data cutoff date, whichever came first.

The occurrence, frequency, and severity of AEs were assessed throughout the treatment period and up to 30 days after the final treatment dose was given according to the National Cancer Institute Common Terminology Criteria of Adverse Events (NCI CTCAE) version 5.0. The incidence of perioperative TEAEs was assessed from the end of systemic treatment to 30 days after surgery. Serious AEs were observed in this study.

The primary end point of the study was the investigator-confirmed ORR, which is defined as the proportion of patients who achieved CR and PR as the best overall response per RECIST, version 1.1. The secondary endpoints included PFS, defined as the duration from treatment initiation to disease progression in patients who did not undergo surgery, or death, whichever occurred first; OS, defined as the duration from the date of treatment initiation to the date of death from any cause; DCR, defined as the proportion of patients who achieved CR, PR and SD as the best overall response; DoR, calculated from the first documented CR or PR to the first documented disease progression or death, whichever occurred first; and the conversion rate of liver metastasis, defined as the proportion of patients with initially unresectable liver metastases who become eligible for surgical resection after undergoing conversion therapy.

Based on an ORR of 47% with first-line CAPEOX plus bevacizumab in NO16966,32,33 and assuming a dropout rate of 20%, a sample size of 45 patients in Cohort A was required to achieve an ORR of 70% with first-line CAPEOX plus anlotinib and maintenance therapy. Furthermore, based on an ORR of 19.4–47.8% with first-line SOC chemotherapy for GI cancers,34,35,36,37,38 a sample size of 40 patients in Cohort C was anticipated to achieve an ORR of 47% with first-line SOC chemotherapy plus anlotinib and maintenance therapy, which resulted in 20% attrition.

This study followed the ITT principle, and the full analysis set included all patients who had received at least one dose of the study medications. Efficacy analysis was based on the ITT population. The ORR, DCR and conversion rate of liver metastases and their 95% CIs were estimated via the Clopper‒Pearson method. The follow-up duration was estimated using the reverse Kaplan‒Meier method. The median PFS and OS were estimated with the Kaplan‒Meier method, and the corresponding 95% CIs were calculated using the Brookmeyer‒Crowley method via log transformation. Patients who underwent surgery were censored for PFS at the final radiological evaluation before surgery.

The safety set included all patients who had received at least one dose of the study medications and had available safety records. Safety was analyzed mainly by descriptive statistics.

All P values were 2-sided, with P < 0.05 considered statistically significant. Statistical analysis was performed via SAS version 9.4.