Radiomics-based Malnutrition for Cervical Cancer.

Zhejiang Provincial People's Hospital

Status: Enrolling

Conditions: Malnutrition; Cervical Cancer

Treatments: Other: There are no interventions.

Study type: Observational

Funder types: Other

Identifiers

NCT05709769

ZJPPH-RT-01

Details and patient eligibility

About

Loss of skeletal muscle, is one of the most prevalent symptoms of malnutrition, and has been frequently reported as a negative factor in cancer patients at any disease stage. In this study, we are planning to firstly analyze the radiomics features of psoas extracted at the level of the third lumbar vertebra (L3) and then, develop a CT-based radiomics nomogram prediction model for predicting malnutrition based on their Patient-Generated Subjective Global Assessment (PG-SGA) scores in patients with International Federation of Gynecology and Obstetrics (FIGO, 2014 version) stage IB1-IIA2 cervical cancer (CC) who received postoperative radiotherapy/chemoradiotherapy (RT/CRT).

Full description

Cervical cancer is still a significant health problem worldwide. Based on the pathological findings after surgery, patients with intermediate or high risk factors for recurrence are recommended to receive adjuvant pelvic RT and/or platinum (cisplatin or carboplatin) based CRT to reduce the risk of tumor recurrence. However, around 30% of individuals with CC will still eventually develop tumor relapse, necessitating the investigation of better supportive care, like nutritional support, to improve therapeutic tolerance and reduce toxic reactions in these patients. In this respect, how to early identification of malnutrition by PG-SGA tool is crucial.

Meanwhile, CT-based radiomics approaches have been successfully applied to generate imaging biomarkers as decision support tools for clinical practice. In our recently accepted research (not yet publish on line, abstract available at https://www.frontiersin.org/articles/10.3389/fnut.2023.1113588/abstract), we firstly analyzed the radiomics features of psoas extracted at the level of L3 and then, developed a nomogram prediction model for patients with FIGO stage IB1-IIA2 CC who received postoperative RT/CRT. Our results demonstrated that this nomogram prediction model showed promising ability for detecting malnutrition based on their PG-SGA scores. The aim of the current study is designed to verify the prediction accuracy of the developed radiomics-based nomogram prospectively.

Enrollment: 150 estimated patients

Sex: Female

Ages: 18 to 80 years old

Volunteers: No Healthy Volunteers

Inclusion criteria

1. Patients received pelvic lymphadenectomy and radical hysterectomy, and pathological diagnosis of CC;
2. Patients had stage IB1-IIA2 CC based on the 2014 FIGO staging system for cervical cancer;
3. Patients received postoperative RT/CRT within one week after admission at the ZJPPH;
4. Patients must have Eastern Cooperative Oncology Group performance status 0-2;
5. No treatments prior to radical surgery;
6. Normal marrow function and the blood tests must be collected within 7 days from enrollment with a hemoglobin of ≥ 80g/L (can be transfused with red blood cells pre-study), an white blood cell (WBC) counts of ≥ 3.0×109/L，a neutrophil count of ≥ 2.0×109/L, , a platelet count of ≥100×109/L, a total bilirubin (TBil) of ≤ 1.0 upper normal limitation (UNL), a creatinine (Cr) of ≤ 1.0 UNL, alanine aminotransferase (ALAT) and aspartate aminotransferase (ASAT) of ≤ 2.5 UNL, Alkaline phosphatase (AKP) ≤5.0 UNL. and no major electrocardiogram abnormalities.
7. Patient does not have a known allergy to platinum (cisplatin or carboplatin) or compounds of similar biologic composition.
8. Patients must be with good compliance and agree to accept nutritional therapy;
9. Informed consent signed. -

Exclusion criteria

1. Poor image quality or visible artifacts around the L3 psoas;
2. Prior treatments of chemotherapy or irradiation;
3. Poor bone marrow, liver and kidney functions, which would make chemotherapy or radiotherapy intolerable;
4. Participating in other clinical trials;
5. Pregnancy, breast feeding, or not adopting birth control;
6. Clinically significant and uncontrolled major medical conditions including but not limited to: active uncontrolled infection, symptomatic congestive heart failure, Unstable angina pectoris or cardiac arrhythmia, psychiatric illness/ social situation that would limit compliance with study requirements; any medical condition, which in the opinion of the study investigator places the subject at an unacceptably high risk for toxicities;
7. The subject has had another active malignancy within the past five years;
8. Poor image quality or visible artifacts around the L3 psoas. -

Trial design

150 participants in 2 patient groups

Training Group

Description:

A primary cohort of eligible patients from the cancer center of Zhejiang Provincial People's Hospital is used for developing the radiomics-based nomogram prediction model. In the training cohort, a sample size of 88 was required to accept the hypothesis that the prediction accuracy of the radiomics-based nomogram model was greater than 45% with 90% power and to reject the hypothesis that the prediction accuracy rate was less than 30% with an α error of 5%. Initially, we planned to enroll 77 patients in the first stage. If 27 or more prediction accuracy rates were observed, we planned to continue to the second stage for a total of 88 patients for the analysis. Considering some deviant cases, the preplanned accrual number was set to 100 patients in the training cohort.

Treatment:

Other: There are no interventions.

Validation Group

Description:

An independent cohort of eligible patients is used for external validation. we are planning to enroll an additional 50 patients to further validate this radiomics-based nomogram prediction model.

Treatment:

Other: There are no interventions.