Study on Adaptive Radiotherapy and Multimodal Information of Cervical Cancer Assisted by Artificial Intelligence

Study on Adaptive Radiotherapy and Multimodal Information of Cervical Cancer Assisted by Artificial Intelligence

The standard treatment for non-operative cervical cancer is concurrent external radiation therapy and chemotherapy followed by brachytherapy. During the period of radiotherapy, organ movement and tumor shrinkage may lead to insufficient or excessive radiation dose for the tumor and organs at risk. Adaptive radiotherapy can use images information acquired during treatment as feedback to reduce errors.

Total 122 cases of cervical cancer with stage IB2-IVA will be randomly enrolled. Concurrent external volumetric rotational intensity modulated radiotherapy and chemotherapy followed by image-guided adaptive brachytherapy is the treatment strategies of control group patients.

Concurrent adaptive external volumetric rotational intensity modulated radiotherapy and chemotherapy followed by image-guided adaptive brachytherapy is the treatment strategies of experimental group patients. CT repositioning will be performed after 15fractions of external radiotherapy, then new target volume will be contoured and new radiotherapy plan will be formulated with the assistance of artificial intelligence program. New radiotherapy plan will be performed from the 17th fraction external radiotherapy. Information on side effects, survival, dosimetry, imaging, clinical features, and cost-effectiveness will be collected. The statistical analysis is as follows, First is the difference in grade 3 side effects between the two groups. Second is 2-year PFS and OS differences between the two groups. Third is relationship between dosimetric differences and prognosis. Fourth one is to analyze the prognostic and predictive factors of adaptive radiotherapy from the patient's clinical characteristics, Positron emission tomography-computed tomography(PET/CT), Magnetic Resonance Imaging(MRI) and other multimodal information. Fifth is cost-benefit analysis of Artificial Intelligence(AI).

1. Introduction and background The standard treatment for non-operative cervical cancer is concurrent external radiation therapy and chemotherapy followed by brachytherapy. During the period of radiotherapy, organ movement and tumor shrinkage may lead to insufficient or excessive radiation dose for the tumor and organs at risk. Adaptive radiotherapy can use images information acquired during treatment as feedback to reduce errors.

2. Hypothesis and purpose Main endpoint: adaptive radiotherapy can reduce level 3 side effects or not. Secondary endpoint: 1. The differences of 2-year progression-free survival and overall survival between two groups. 2. To analyze Physical dosimetry differences between two groups, and the correlation between physical dosimetry differences and prognosis also will be evaluated. 3. To analyze the prediction and prognostic factors of adaptive radiotherapy for cervical cancer, and to provide supporting data for the subsequent optimization of cervical cancer treatment. 4. To evaluate the effectiveness of AI and conduct cost-benefit analysis.

3. Trial methodology and design Total 122 cases of IB2-IVA cervical cancer will be randomly enrolled. Concurrent external volumetric rotational intensity modulated radiotherapy and chemotherapy followed by image-guided adaptive brachytherapy is the treatment strategies of control group patients.

   Concurrent adaptive external volumetric rotational intensity modulated radiotherapy and chemotherapy followed by image-guided adaptive brachytherapy is the treatment strategies of experimental group patients. CT repositioning will be performed after 15fractions of external radiotherapy, then new target volume will be contoured and new radiotherapy plan will be formulated with the assistance of artificial intelligence program. New radiotherapy plan will be performed from the 17th fraction external radiotherapy.

   Meanwhile, concurrent chemotherapy regimen is cisplatin 40mg/m2/week (the maximum weekly dose should less than or equal to 70mg and no more than 6cycles). Information on side effects, survival, dosimetry, imaging, clinical features, and cost-effectiveness will be collected. The statistical analysis is as follows, First is the difference in grade 3 side effects between the two groups. Second is 2-year PFS and OS differences between the two groups. Third is relationship between dosimetric differences and prognosis. Fourth one is to analyze the prognostic and predictive factors of adaptive radiotherapy from the patient's clinical characteristics, PET/CT, MRI and other multimodal information. Fifth is cost-benefit analysis of AI.

4. Anticipated result and potential impact Adaptive radiotherapy can reduce side effects and obtain prognosis and prognostic factors of adaptive radiotherapy.

Inclusion Criteria:

1. pathologically confirmed cervical squamous cell carcinoma or adenocarcinoma without treatment before;
2. Age: ≥18 years old;
3. The International Federation of Gynecology and Obstetrics(FIGO) stage: IB2 to IVA, or IVB with only para-aortic lymph node metastasis, refused or could not be treated by surgery;
4. Eastern Cooperative Oncology Group(ECOG)score ≤2;
5. good bone marrow, hematopoietic and liver and kidney function: absolute neutrophil count (ANC) ≥ 1.5 ☓ 109 / L, the platelet count ≥100 ☓ 109 / L, or hemoglobin > 90 g/L, serum bilirubin < 1.5 ☓ upper limit of normal reference value(ULN), aspartate aminotransferase(AST) and alanine aminotransferase(ALT)< 2.5 ☓ ULN, serum creatinine clearance ≥ 50 ml/min.
6. provide informed consent.

Exclusion Criteria:

1. women in pregnancy or nursing;
2. contraindications to chemoradiotherapy;
3. subjects participating in other clinical trials or participating in other clinical trials within 30 days;