Distribution Of Subclinical Lesions Of Early Breast Cancer In Premenopausal Patients And The Feasibility Of Individualized Tu-mor Bed Delineation

Part I:Analysis of the distribution of subclinical lesions of early breast cancer in premenopausal patients who undergo breast-conserving surgeryObjective:To analyze the distribution and the characteristic of subclinical le-sions of early breast cancer by using pathological-radiological method, to find the biological parameters which affect the distribution of subclinical lesions and to provide specific histopathological information for the tailored tumor bed target volume delineation after breast-conserving surgery.Matirial and Methods:The enrolled specimens were sectioned into5-mm-thick consecutive slices.The orientation of the slices was perpendicular to the nip-ple-lateral direction and X-ray images were obtained. By comparing the X-ray images, the invasive tumor and surrounding clusters of microscopic lesions were delineated and measured. After reconstruction of the tissue sections in three di-mensions, the distance and the corresponding direction of the microscopic exten-sion relative to the edge of the invasive tumor were established. Factors that may influence the distribution of subclinical lesions were analyzed, which included: patient factors (age), tumor factors (pathology subtypes, diameter and the result of immunohistochemical), and surgical factors (whether preoperative biopsy in-filtrating lesions contains intraductal component).Results:From January2011to October2011,29patients were enrolled. Micro-scopic disease extension outside the primary invasive tumor was found in20/29(69%) cases. In58.6%,24.1%,3.4%of tumors, microscopic diseases were found outside the gross invasive lesions of1cm,2cm and3cm. The proportion of in-traductal component around the primary invasive breast cancer was the only sig-nificant factor influencing the distribution of subclinical lesions (r=0.602, p<0.01). The resection margins were not symmetrical. The mean resection mar-gin in6edges were:22±11mm(superior margin);26±15mm(inferior margin);24±11mm (medial margin);31±13mm(lateral margin);8.4±4.3mm(anterior margin);8.6±4.4(deep margin).Conclusions:In breast-conserving surgery, the resection margins are asymmet-rical. It is necessary to individualize the tumor bed target volume delineation. In 90%of cases, the distribution of subclinical lesions of invasive disease were within30mm from the edge of the primary invasive lesion. Thus, it is appropri-ate to use30mm as the reference of tumor bed target volume delineation of early breast cancer in premenopausal patients who undergo breast-conserving surgery. Part II:The feasibility of individualized tumor bed delineationPurpose:To compare clinical target volumes generated using a standard10mm and15mm margin (CTVstandard10and CTVstandard15) with those generated using three-dimensional surgical excision margins (CTVtailored) in patients with breast-conserving surgery for breast cancer. Then to compare the dosimetrical effect on target volume and normal tissue by using these different TB-CTVs respectively.Materials and Methods:Ten women out of the tewenty-nine patients of Part I were enrolled. Tumor bed delineation was according to the protocol of RTOG0413/NSABP0319. TB-CTVstandard10and TB-CTVstandard15was generated by adding a uniform three-dimensional10mm and15mm margin, and TB-CTVtailored was generated by adding30mm minus the excision margin in three-dimensions. Concordance between TB-CTVstandard and TB-CTVtailored was quantified using conformity (Col), geographical-miss (GMI), normal-tissue (NTI) in-dices and dose-volume histogram of critical organs. An external-beam irradiation plan was generated to cover95%of TB-CTV with the95%isodose. Dose-volume histo-gram of treatment plans with TB-CTVstandard10and TB-CTVstandard15was compared with that with TB-CTVtailored.Results:TB-CTVstandard10and TB-CTVstandard15were both greater in volume than that of the TB-CTVtailored(63cm3vs.53cm3, p=0.03;97cm3vs.53cm3, p=0.01). The mean Col, GMI and NTI of TB-CTVstandard10and TB-CTVtailored were0.50,0.17and0.33, respectively. The mean Col, GMI and NTI of TB-CTVstandard15and TB-CTVtailored were0.48,0.07and0.45, respectively. CTVtailored was inade-quately covered by the95%isodose of the TB-CTVstandard10in5/10cases, and CTVtailored was inadequately covered by the95%isodose of the TB-CTVstandard15in3/10cases. There was no difference with the dose to the heart、ipsilateral lung and contralateral breast between the plans generated by TB-CTVstandard10and TB-CTVtailored. Neither was the difference of the dose to the heart and contralateral breast between the plans generated by TB-CTVstandard10and TB-CTVtailored, but the mean ipsilateral lung dose of the irradiation plan generated by TB-CTVstandard15was higher than that of the irradiation plan generated by TB-CTVtailored (1171.59cGy vs.1124.17cGy, p=0.01)Conclusions:CTVs defined using a standard uniform10mm and15mm TB-CTV margin are greater in volume than those based on3D actual excisional margin data. Standard TB-CTV margin could result in a geographical miss. There was no obvious difference with the dose of the heart, ipsilateral lung and contralateral breast between the irradiation plan generated by TB-CTVstandard and TB-CTVtailored. TB-CTVstandard15results in higher mean dose to the ipsilateral lung than that with TB-CTVtailored.