This is a prospective observational cohort study conducted at Fundación CTIC in Bogotá, Colombia. It characterizes the oral, fecal and intratumoral microbiome of Colombian adults with advanced solid tumors (gastric, colorectal, breast, cervical and head-and-neck cancer) who receive first-line immunotherapy as standard of care, and compares them with healthy volunteers. Using multi-omics (HiFi metagenomics, 16S, tumor RNA-Seq and untargeted metabolomics), the study aims to identify microbial signatures associated with treatment response and survival, building the initial Colombian cohort of a Cancer Microbiome Atlas with Latin American projection.

Single-center, prospective, observational cohort study. 120 patients with advanced (stage III unresectable or IV) solid tumors candidates for first-line immunotherapy and 30 healthy controls are enrolled (total N=150). Patients provide saliva, stool and tongue-scraping samples plus archival FFPE tumor tissue at baseline (V0) and at 6 (V1) and 12 (V2) months; controls provide saliva and stool at V0 only. Microbiomes are profiled by PacBio HiFi shotgun metagenomics and full-length 16S; tumor transcriptome by RNA-Seq; and saliva/stool metabolome by GC-MS in a subcohort (n=60). Tumor response is assessed by RECIST 1.1. Microbial, transcriptomic and metabolomic features are integrated with clinical outcomes (objective response, PFS, OS) using multivariable and machine-learning models.

Inclusion Criteria:

• Age 18 years or older.
• Histologically confirmed advanced (stage III unresectable or IV) gastric, colorectal, breast, cervical or head-and-neck cancer.
• Candidate for first-line immunotherapy per current clinical guidelines.
• Available FFPE tumor block in institutional or reference pathology archive.
• Able to provide saliva and stool samples at V0 and follow-up.
• ECOG performance status within protocol limits; life expectancy over 3 months.
• Cognitive capacity to give informed consent; signed EVA-BIOBANCO and Atlas consents.

Inclusion Criteria (healthy controls)

• No prior cancer diagnosis.
• No active autoimmune disease; no inflammatory bowel disease.
• No antibiotics, invasive dental treatment, immunosuppressants or corticosteroids in the prior 3 months; no severe active periodontal disease.

Exclusion Criteria:

• Systemic antibiotics within 3 months (except short course under 5 days for uncomplicated infection).
• Probiotics or prebiotics within 30 days.
• Chronic proton-pump inhibitors for more than 3 continuous months.
• Inflammatory bowel disease (Crohn, ulcerative colitis) or extensive bowel resection.
• Pregnancy or lactation.
• BMI under 18.5 or over 40 kg/m².
• Uncontrolled diabetes (HbA1c at or above 9% in prior 3 months).
• Invasive dental treatment within 3 months.
• Patients: simultaneous randomized trial of a non-INVIMA-approved experimental immunotherapy; or no evaluable FFPE block meeting quality thresholds.
• Inability to ensure 12-month clinical follow-up.

Colombian adults with advanced solid tumors (gastric, colorectal, breast, cervical, head-and-neck) candidates for first-line immunotherapy, plus healthy adult volunteers. Single center: Fundación CTIC, Bogotá.

De-identified genomic sequences (oral, fecal and intratumoral microbiome) will be deposited in public repositories (NCBI SRA / EBI ENA) with minimal metadata (tumor type, age by decade, sex). Analysis pipelines and code will be versioned and shared via GitHub and Zenodo. Raw data and code will be made public at study close.

Available at study close (estimated 2029); retained at least 15 years per ICH-GCP archiving policy.

Open access via public repositories (SRA, ENA, Zenodo) for de-identified sequence data and analysis code.