FRAIL and FiND for Chemotherapy Toxicity in GI Cancer

Background and Rationale

Older adults receiving chemotherapy for gastrointestinal malignancies represent a clinically heterogeneous population. Chronological age alone does not adequately reflect differences in physiological reserve, functional capacity, comorbidity burden, nutritional status, or vulnerability to treatment-related adverse events. Brief frailty screening instruments may complement conventional oncology assessments by identifying clinically relevant vulnerabilities before systemic treatment.

The FRAIL scale and the Frail Non-Disabled (FiND) questionnaire assess overlapping but distinct aspects of vulnerability. FRAIL provides a five-item measure of frailty, whereas FiND distinguishes frailty-related characteristics from self-reported mobility disability. Although both instruments are practical for routine clinical use, their comparative associations with chemotherapy toxicity and their incremental predictive contributions beyond standard clinical factors require further evaluation.

Study Objectives

The primary objective is to evaluate the association between baseline FRAIL and FiND assessments and the occurrence of grade ≥3 chemotherapy-related toxicity within 180 days after initiation of adjuvant chemotherapy in older adults with nonmetastatic gastrointestinal cancer.

Additional objectives are to compare the discriminatory performance of FRAIL and FiND, examine their incremental predictive contributions beyond selected clinical risk factors, describe the relationship between the two frailty classifications, and characterize treatment modifications and unplanned hospitalizations during follow-up.

Study Design and Population

FRAIL-GI is a single-center observational cohort study conducted in the Medical Oncology Department of Ankara Etlik City Hospital, Türkiye.

The study population consists of adults aged 65 years or older with nonmetastatic gastrointestinal malignancies for whom adjuvant systemic chemotherapy has been selected as part of routine clinical management.

Chemotherapy regimen selection, dose modifications, supportive care, and decisions regarding treatment continuation are made by the treating oncology team independently of study participation. No experimental anticancer treatment is assigned.

Baseline Assessments

Baseline assessments include demographic characteristics, primary tumor site, chemotherapy regimen and treatment intensity, Eastern Cooperative Oncology Group performance status, selected comorbidities, anthropometric measurements, and routine laboratory parameters.

Frailty is evaluated using the FRAIL scale and the FiND questionnaire before chemotherapy initiation.

The FRAIL scale comprises five dichotomous components: fatigue, resistance, ambulation, illnesses, and loss of weight. The illnesses criterion is positive when at least five of 11 specified chronic conditions are present. The weight-loss component evaluates unintentional body-weight loss exceeding 5% during the preceding 12 months. Total FRAIL scores range from 0 to 5 and are categorized as robust (0), pre-frail (1-2), or frail (3-5).

The FiND questionnaire comprises two mobility-disability items and three frailty-related items. Participants are classified hierarchically as robust, frail without mobility disability, or mobility-disabled. A positive mobility-disability assessment takes precedence over frailty-related item responses. FiND is analyzed using these categories rather than an unvalidated additive total score.

Follow-up and Outcome Assessment

Study assessments are organized around baseline and approximately days 90 and 180 after chemotherapy initiation.

Chemotherapy-related adverse events are evaluated throughout the 180-day observation period using the Common Terminology Criteria for Adverse Events, version 5.0. The primary outcome is the occurrence of at least one chemotherapy-related adverse event of grade ≥3 within 180 days of treatment initiation.

The date of the first severe adverse event and the maximum recorded toxicity grade are documented when available. Treatment-related outcomes include chemotherapy dose reductions, treatment delays, early permanent treatment discontinuation, and unplanned hospitalization.

The availability of scheduled follow-up assessments is evaluated separately from ascertainment of clinical toxicity outcomes, because an incomplete scheduled assessment does not necessarily indicate loss of clinical outcome follow-up.

Statistical Analysis

The associations of baseline FRAIL scores and FiND classifications with severe chemotherapy toxicity are evaluated using logistic regression.

Clinical reference models incorporate age, baseline performance status, and chemotherapy intensity. The FRAIL-extended model additionally incorporates the continuous FRAIL score, whereas the FiND-extended model incorporates the categorical FiND classification.

Model discrimination is evaluated using the area under the receiver operating characteristic curve. Overall prediction error and calibration are assessed using the Brier score and calibration measures. Bootstrap resampling is used for internal validation and estimation of optimism-corrected model performance. Comparative analyses examine whether either frailty assessment provides additional predictive information beyond the clinical reference model.

Exploratory analyses may examine the joint contribution of FRAIL and FiND and the robustness of findings across alternative model specifications or clinical subgroups.

The study is intended to clarify the potential value and limitations of brief baseline frailty assessments for identifying older adults at increased risk of severe chemotherapy-related toxicity. Findings are interpreted in the context of the observational study design and the need for independent validation before clinical prediction tools can be recommended for routine decision-making.