Faculty Bibliography

PURPOSE: We report a comparison of the dosimetry and toxicity of three-dimensional conformal radiotherapy (3D-CRT) vs. intensity-modulated radiotherapy (IMRT) among patients treated in the prone position with the same fractionation and target of the hypofractionation arm of the Canadian/Whelan trial. METHODS AND MATERIALS: An institutional review board-approved protocol identified a consecutive series of early-stage breast cancer patients treated according to the Canadian hypofractionation regimen but in the prone position. Patients underwent IMRT treatment planning and treatment if the insurance carrier approved reimbursement for IMRT; in case of refusal, a 3D-CRT plan was used. A comparison of the dosimetric and toxicity outcomes during the acute, subacute, and long-term follow-up of the two treatment groups is reported. RESULTS: We included 97 consecutive patients with 100 treatment plans in this study (3 patients with bilateral breast cancer); 40 patients were treated with 3D-CRT and 57 with IMRT. IMRT significantly reduced the maximum dose (Dmax median, 109.96% for 3D-CRT vs. 107.28% for IMRT; p < 0.0001, Wilcoxon test) and improved median dose homogeneity (median, 1.15 for 3D-CRT vs. 1.05 for IMRT; p < 0.0001, Wilcoxon test) when compared with 3D-CRT. Acute toxicity consisted primarily of Grade 1 to 2 dermatitis and occurred in 92% of patients. Grade 2 dermatitis occurred in 13% of patients in the 3D-CRT group and 2% in the IMRT group. IMRT moderately decreased rates of acute pruritus (p = 0.03, chi-square test) and Grade 2 to 3 subacute hyperpigmentation (p = 0.01, Fisher exact test). With a minimum of 6 months' follow-up, the treatment was similarly well tolerated in either group, including among women with large breast volumes. CONCLUSION: Hypofractionated breast radiotherapy is well tolerated when treating patients in the prone position, even among those with large breast volumes. Breast IMRT significantly improves dosimetry but yields only a modest but confirmed benefit in terms of toxicities. If a concurrent boost to the tumor bed is not required, a conformal 3D-CRT approach can adequately deliver prone whole-breast hypofractionation radiotherapy

Purpose: Limited information is available comparing toxicity of accelerated radiotherapy (RT) to that of standard fractionation RT for early stage breast cancer. We report early and late toxicities of two prone regimens of accelerated intensity-modulated radiation therapy (IMRT) with a concomitant boost (CB) to the tumor bed delivered over 3 or 5 weeks as compared to standard 6 week RT with a sequential electron boost. Methods: From 2/2003 to 12/2007, 169 consecutive patients with Stage I-II breast cancer were offered the choice to undergo prone RT with either: a 6-week standard RT regimen of 46 Gy/23 fractions (fx) to the whole breast (WB), followed by a14 Gy sequential boost (SB) to the tumor bed (6wSB), a 5-week regimen of 50 Gy to WB with an IMRT CB of 6.25 Gy in 25 fx (5wCB); or a 3-week protocol of 40.5 Gy to WB with an IMRT CB of 7.5 Gy in 15 fx (3wCB). These regimens were estimated as biologically equivalent, based on alpha/beta = 4 for tumor control. Toxicities were reported using RTOG and LENT/SOMA scoring. Results: 51/169 patients chose standard 6wSB, 28 selected 5wCB, and 90 enrolled in 3wCB protocol. Maximum acute toxicity was Grade 3 dermatitis in 4% of the patients in the 6wSB compared 1% in 3wCB. In general, acute complications (breast pain, fatigue, and dermatitis) were significantly less in the 3wCB than in the other schedules (P < 0.05). With a median follow-up of 61 months, the only Grade 3 late toxicity was telangiectasia in two patients: one in 3wCB and one in 5wCB group. Notably, fibrosis was comparable among the three groups (P = NS). Conclusion: These preliminary data suggest that accelerated regimens of breast RT over 3 or 5 weeks in the prone position, with an IMRT tumor bed CB, result in comparable late toxicity to standard fractionation with a sequential tumor boost delivered over 6 weeks. As predicted by radiobiological modeling the shorter regimen was associated with less acute effects.

PURPOSE: To report setup variations during prone accelerated partial breast irradiation (APBI). METHODS: New York University (NYU) 07-582 is an institutional review board-approved protocol of cone-beam computed tomography (CBCT) to deliver image-guided ABPI in the prone position. Eligible are postmenopausal women with pT1 breast cancer excised with negative margins and no nodal involvement. A total dose of 30 Gy in five daily fractions of 6 Gy are delivered to the planning target volume (the tumor cavity with 1.5-cm margin) by image-guided radiotherapy. Patients are set up prone, on a dedicated mattress, used for both simulation and treatment. After positioning with skin marks and lasers, CBCTs are performed and the images are registered to the planning CT. The resulting shifts (setup corrections) are recorded in the three principal directions and applied. Portal images are taken for verification. If they differ from the planning digital reconstructed radiographs, the patient is reset, and a new CBCT is taken. RESULTS: 70 consecutive patients have undergone a total of 343 CBCTs: 7 patients had four of five planned CBCTs performed. Seven CBCTs (2%) required to be repeated because of misalignment in the comparison between portal and digital reconstructed radiograph image after the first CBCT. The mean shifts and standard deviations in the anterior-posterior (AP), superior-inferior (SI), and medial-lateral (ML) directions were -0.19 (0.54), -0.02 (0.33), and -0.02 (0.43) cm, respectively. The average root mean squares of the daily shifts were 0.50 (0.28), 0.29 (0.17), and 0.38 (0.20). A conservative margin formula resulted in a recommended margin of 1.26, 0.73, 0.96 cm in the AP, SI, and ML directions. CONCLUSION: CBCTs confirmed that the NYU prone APBI setup and treatment technique are reproducible, with interfraction variation comparable to those reported for supine setup. The currently applied margin (1.5 cm) adequately compensates for the setup variation detected

PURPOSE: The purpose of this study was to investigate how incorporation of magnetic resonance spectroscopy imaging (MRSI) into radiotherapy planning would increase the target volume for patients with recurrent glioma. METHODS: After prior standard radiotherapy, 25 patients with recurrent glioma were treated with bevacizumab and concurrent hypofractionated stereotactic radiotherapy (HFSRT), delivering 30 Gy in five fractions. MRSI were acquired for 12 patients. Areas with markedly higher choline levels relative to the levels of total creatine and N-acetylaspartate were identified and referred to as MRSI voxels with elevated metabolite ratios (EMR). Gross tumor volume (GTV) consisted of contrast-enhancing tumor on T1-weighted magnetic resonance images (MRI) and computed tomography. Clinical target volume (CTV) was GTV + 5 mm margin and MRSI voxels with EMR. Overall survival (OS) and 6-month progression free survival (PFS) for these patients were reported in a prior publication [Gutin et al., Int. J. Radiat. Oncol., Biol., Phys. 75(1), 156-163 (2009)], and the outcome was correlated with the GTV and the volume of MRSI voxels with EMR in this study. RESULTS: Seven of the 12 patients had MRSI voxels with EMR. If none of the MRSI voxels with EMR were included, the CTV would range from 9.2 to 73.0 cm3 with a median of 31.0 cm3, whereas if all voxels were included, the CTV would range from 27.4 to 74.4 cm3 with a median of 35.0 cm3. For three of the seven patients, including the voxels with EMR, would have increased the CTV by 14%-23%. For one patient, where the MRSI voxels with EMR did not overlap the GTV, including these voxels would increase the CTV by 198%. No correlation could be found between the OS and PFS and the GTV or the volume of MRSI voxels with EMR. CONCLUSIONS: Seven of 12 patients with recurrent glioma had MRSI voxels with EMR. For four of these seven patients, including the MRSI voxels with EMR, significantly increased the CTV. This study does not have statistical power to conclude on the importance of including areas with MRSI-suspect disease into the radiation target volume

We have previously demonstrated high pathologic response rates after neoadjuvant concurrent chemoradiation in patients with locally advanced breast cancer (LABC). We now report disease-free survival (DFS) and overall survival (OS) in the context of pathologic response. 105 LABC patients (White 46%, Non-White 54%) were treated with paclitaxel (30 mg/m(2) intravenously twice a week) for 10-12 weeks. Daily radiotherapy was delivered to breast, axillary, and supraclavicular lymph nodes during weeks 2-7 of paclitaxel treatment, at 1.8 Gy per fraction to a total dose of 45 Gy with a tumor boost of 14 Gy at 2 Gy/fraction. Pathological complete response (pCR) was defined as the absence of invasive cancer in breast and lymph nodes and pathological partial response (pPR) as the persistence of <10 microscopic foci of invasive carcinoma in breast or lymph nodes. Pathologic response (pCR and pPR) after neoadjuvant chemoradiation was achieved in 36/105 patients (34%) and was associated with significantly better DFS and OS. Pathological responders had a lower risk of recurrence or death (HR = 0.35, P = 0.01) and a longer OS (HR = 4.27, P = 0.01) compared with non-responders. Median DFS and OS were 57 and 84 months for non-responders, respectively, and have not yet been reached for responders. Importantly, pathologic response was achieved in 54% of patients with HR negative tumors (26/48). In conclusion, pathologic response to concurrent paclitaxel-radiation translated into superior DFS and OS. Half of the patients with HR negative tumors achieved a pathologic response

PURPOSE: NYU 05-181 protocol compared the CT simulation in both supine and prone positions for 400 patients with breast cancer (200 left-breast and 200 right-breast) to identify which setup is better at sparing heart and lung involvement in the treatment process. The results demonstrated that all right-breast patients benefited from the prone treatment position, while for left-breast patients, 85% were better treated prone and 15% were better treated supine. Using the clinical data collected from this protocol, the authors aimed at developing an automated tool capable of identifying which of the left-breast cancer patients are better treated supine without obtaining a second CT scan in the supine position. METHODS: Prone CT scans from 198 of the 200 left-breast cancer patients enrolled in NYU 05-181 protocol were deidentified and exported to a dedicated research planning workstation. Three-dimensional geometric features of the organs at risk and tumor bed were extracted. A two-stage classifier was used to classify patients into the prone class or the supine class. In the first stage, the authors use simple thresholding to divide the patients into two groups based on their in-field heart volume. For patients with in-field heart volume < or = 0.1 cc, the prone position was chosen as the preferred treatment position. Patients with in-field heart volume > 0.1 cc will be further classified in the second stage by a weighted support vector machine (SVM). The weight parameters of the SVM were adjusted to maximize the specificity [true-supine/(true-supine+false-prone)] at the cost of lowering but still maintaining reasonable sensitivity [true-prone/(true-prone+false-supine)]. The authors used K-fold cross validations to test the performance of the SVM classifier. A feature selection algorithm was also used to identify features that give the best classification performance. RESULTS: After the first stage, 49 of the 198 left-breast cancer patients were found to have > 0.1 cc of in-field heart volume. The three geometric features of heart orientation, distance between heart and tumor, and in-field lung were selected by the feature selection algorithm in the second stage of the two-stage classifier to give the best predefined weighted accuracy. The overall sensitivity and specificity of the proposed method were found to be 90.4% and 99.3%, respectively. Using two-stage classification, the authors reduced the proportion of prone-treated patients that need a second supine CT scan down to 16.3/170 or 9.6%, as compared to 21/170 or 12.4% when the authors use only the first stage (thresholding) for classification. CONCLUSIONS: The authors' study showed that a feature-based classifier is feasible for predicting the preferred treatment position, based on features extracted from prone CT scans. The two-stage classifier achieved very high specificity at an acceptable expense of sensitivity