Predictors And Risk Factors To Esophageal Or Pulmonary Injury Induced By Platinum-based Neoadjuvant Chemotherapy And Thoracic Radiotherapy

1. Impact of oral negative contrast on esophagus delineation and the predictability of physical parameters to acute radiation esophagitisPurpose: To present our observation on the impact of oral negative contrast on contouring the esophagus, on the variations in esophageal volumetric, dosimetric parameters and further its predictability to acute radiation esophagitis in patients treated with thoracic radiotherapy (TRT)Methods: Between January and December 2003, twenty-five patients (22 with lung cancer, 2 with malignant thymoma and 1 with mediastinum tumor) who received incidental thoracic irradiation were enrolled onto this prospective clinical study. There were 2 females and 23 males with a median age of 53 years (range 36-73 years). Delivered radiation dose to the isocenter was 30 - 66 Gy, median 60 Gy.Treatment planning images were taken at 0.5 cm or 0.8 cm intervals by a Philips ACQSEM? large bore CT simulator. All patients were scanned twice in exactly the same position and at the same intervals. Before the second scanning, thirty-milliliter oral negative contrast was given to each patient. CT images were then transferred to Cad-plan? (version 6.02) 3D treatment planning system to delineate target volumes and normal anatomical structures. The external surface of esophagus was contoured by a single physician on each non-contrasted (C-) or contrasted (C+) axial image from the level of the lung apices superiorly to the gastroesophageal junction inferiorly. The internal lumen and mucosa layer (defined as tissue encircling the internal lumen lmm in depth) were also delineated by a single physician uniformly. The number of segments of well-visualized internal lumen was noted for either C- or C+ images.The following volumetric and dosimetric parameters were extracted from the 3D dose data set for internal lumen, mucosa layer and the whole esophagus acquired from C- or C+ scans, respectively: 1) Volume; 2) Maximum (Dmax) and mean dose (Dmean); 3) V10 to V55 which is the percent of esophageal volume received at least 10 to55 Gy, respectively. The predictive ability of these (C- or C+) dosimetric parameters was accessed based on the area beneath receiver operating characteristic (ROC) curve.Results: After taking oral contrast, more CT segments presenting well-visualized esophageal internal lumen were revealed in 14 patients (56%). There were no changes in three patients, however, in deed fewer well-visualized images in the other eight patients at least partly due to swallowing. There were totally 388 well-visualized segments in C+ images, comparing 363 segments in C- images (363/622 vs. 388/621, p = 0.37). Esophageal volumes did changed significantly after taking oral contrast (p = 0.017-0.028). Dmean and V10-50 were highly correlated. There was no statistically significant changes between each pair of dosimetric predictor extracted from C- or C+ data set (p = 0.07-0.97).Sixteen Of 25 (64%) patients developed CTC 3.0 Grade >2 acute esophagitis. Mean dose of internal lumen or mucosa layer or the whole esophagus was a better single metrics (area under ROC curve = 0.72-0.79) comparing Dmax (area under ROC curve = 0.62-0.68) in predicting acute esophagitis.Conclusions: The application of oral negative contrast gave assistance in contouring the esophagus in approximately 60% of patients (14/25). However, there was no numeric improvement for the other 40% of patients (11/25) under this approach, which further offset the overall benefit and finally leads to no statistical significance in this group of patients. The differences between each pair of dosimetric parameters extracted from C- or C+ data set, or its predictability to acute radiation esophagitis were not notable. Therefore, we decided not to use oral contrast in our further study evaluating models for predicting acute radiation esophagitis.2. Serum pro-inflammatory cytokines, serum manganese superoxide dismutase (MnSOD) and dosimetric metrics as predictors for acute esophagitis induced by neoadjuvant platinum-based chemotherapy and thoracic radiotherapy-A prospective studyPurpose: To investigate prospectively changes in serum pro-inflammatory cytokines [interleukin (IL) la> IL 1/3^ IL 6^ tumor necrosis factor-a(TNF-a)], serum MnSOD (antioxidant enzyme) and dosimetric metrics as predictors for acute esophagitis induced by neoadjuvant platinum-based chemotherapy and thoracic radiotherapy (TRT)Methods: Between January and December 2004, eighty-two patients (71 with lung cancer, 6 with malignant thymoma and 5 with mediastinum tumor) who received incidental thoracic irradiation and neoadjuvant platinum-based chemotherapy were enrolled onto this prospective clinical study. There were 10 females and 72 males with a median age of 56 years (ranged 19-74 years). All patients received neoadjuvant platinum-based chemotherapy. Chemo-regimen consisted of NP, NIP, EP and CAP. The medium pre-RT cycle number was one (ranged 1-6 cycles). Radiation dose ranged from 30 - 66 Gy. Cad-plan? (version 6.08) and Pinnacle3? (version 7.0) treatment planning system were applied to generate dosimetric metrics as following: maximum dose (Dmax), mean dose (Dmean), VI0 to V55 of. internal lumen, mucosa layer or the whole esophagus. Serial serum concentration of IL lo\ IL l/3> IL 6% TNF-a and MnSOD were determined by enzyme-linked immunosorbent assay (ELISA) prior to and weekly during RT. Since great interpatient divergence was revealed for baseline serum cytokines and MnSOD levels, we calculated weekl/pre-RT ratios (wl/0 ratio, defined as serum concentration in the 1st week of RT/Pre-RT level) as surrogates. In univariate analysis, various dosimetric, serum cytokines, serum MnSOD, treatment (conventional fractionated RT vs non-conventional RT) and clinical (age, smoke) parameters were studied. Logistic step-wise regression analysis was then used to determine significant predictors to CTCAE 3.0 Grade >2 acute esophagitis. The predictive ability of these parameters was also evaluated by the area beneath receiver operating characteristic (ROC) curve.Results: Forty-five Of 82 (55%) patients developed CTCAE 3.0 Grade >2 acute esophagitis, 36 developed Grade 2 and 9 Grade 3. The most statistically significant single parameters for predicting acute esophagitis were mean dose of internal lumen and TNF-a wl/0 ratio, tis. The single best predictor was mean dose of internal lumenwith the area under ROC curve 0.67, comparing area under ROC curve 0.59 for TNF-a wl/0 ratio. The cut-off was 21.6 Gy with sensitivity of 76% (34/46) and specificity of 44% (16/36). On logistic regression analysis, mean dose of internal lumen and TNF-a wl/0 ratio remained significantly correlated with Grade >2 acute esophagitis, p = 0.0245 and 0.0494, respectively. Finally, a model considering mean dose of internal lumen, wl/0 ratio of TNF-a, IL 1/3 and MnSOD seemed to be better correlated with Grade >2 acute esophagitis (area under ROC curve = 0.732). Conclusions: The single best predictor for CTCAE 3.0 Grade >2 acute esophagitis was mean dose of internal lumen, with sensitivity of 76% and specificity of 44% at the cut-off value of 21.6 Gy. On logistic regression analysis, mean dose of internal lumen and TNF-a wl/0 ratio (serum concentration in the 1st week of RT/Pre-RT level) were significantly correlated with acute esophagitis. A multiparameter model considering mean dose of internal lumen, wl/0 ratio of pro-inflammatory cytokines TNF-a, IL 1/3 and antioxidant enzyme MnSOD seemed to better predict outcome (area under ROC curve = 0.732) than any single parameter in this group of patients. Further studie is underway to test this model.3. Relating radiation-induced regional lung injury to changes in pulmonary function tests.Purpose: To determine if the sum of radiation therapy (RT)-induced reductions in regional lung perfusion is quantitatively related to changes in global lung function as assessed by reductions in pulmonary function tests (PFTs).Methods: 207 patients (70% with lung cancer) that received incidental partial lung irradiation had PFTs (FEV1 -forced expiratory volume in 1 second, DLCO- Diffusion Capacity for Carbon Monoxide) prior to, and repeatedly following, RT as part of a prospective clinical study. Regional lung function was serially assessed pre- and post-RT by SPECT (single photon emission computed tomography) perfusion scans. Of these, 53 patients had 105 post-RT evaluations of changes in both regional perfusion and PFTs, at approximately the same time post-RT (within 0-4 weeks), and were without evidence of intra-thoracic disease recurrence that might influence regional perfusion and PFTs, and were not on steroids. The summation of the regional functional perfusion changes was correlated with changes in PFTs, using linear regression. Follow-up ranged from 3 to 86 months (median 19).Results: Overall, there is a significant relationship between the sum of changes in regional perfusion and changes in PFTs (p = 0.002-0.24, depending on particular PFT index). However, the correlation coefficients are low (r = 0.16-0.41). Conclusions: There is a statistically significant relationship between RT-induced changes in regional function (i.e. perfusion) and global function (i.e. PFTs). However, there is a lot of variability in the data, making it different to relate changes in perfusion to changes in PFTs. Thus, with our current techniques, prediction of changes in perfusion alone appears to be inadequate to accurately predict changes in PFTs. Therefore, we decided not to evaluate perfusion changes in further study.4. Factors relative to the incidence of CTCAE 3.0 Grade ^ pneumonitis or Grade ^ pulmonary fibrosis induced by neoadjuvant platinum-based chemotherapy and thoracic radiotherapy: A prospective studyPurpose: To investigate prospectively physical parameters and changes in serum interleukin (IL) la, IL 1/?, IL 6, IL 7, tumor necrosis factor-a (TNF-a), transforming growth factor-/31 (TGF-/31) and manganese superoxide dismutase (MnSOD) as indictors for CTCAE 3.0 Grade ^ pneumonitis Grade ^ pulmonary fibrosis induced by neoadjuvant platinum-based chemotherapy and thoracic radiotherapy (TRT) Methods: Between January and December 2004, eighty-six patients (66 with lung cancer, 10 with esophageal cancer, 5 with malignant thymoma and 5 with mediastinum tumor) who received incidental thoracic irradiation and neoadjuvant platinum-based chemotherapy were enrolled onto this prospective study. There were 9 females and 77 males with a median age of 57 years (ranged 19-74 years). All patients received neoadjuvant platinum-based chemotherapy. Chemo-regimen included NP, NIP, EP, IEP and CAP. The medium pre-RT cycle was one (ranged 1-6 cycles). Radiation dose ranged from 30 - 66 Gy. Cad-planTM (version 6.08) and Pinnacle3TM (version 7.0) treatment planning system were applied to generate MLD (mean lung dose) and V20 (percent of lung receiving ^0 Gy). Serial serum concentration of IL la, IL 1(8, IL 6, IL 7, TNF-a, TGF-/31and MnSOD were determined by enzyme-linked immunosorbent assay (ELISA) prior to and weekly during RT. Since great inter-patient divergence was revealed for baseline serum cytokines and MnSOD levels, we calculated weekl/pre-RT ratios (wl/0 ratio, defined as serum concentration in the 1st week of RT/Pre-RT level) as surrogates. On univariate analysis, MLD, V20, serum cytokines, serum MnSOD, smoking history (<20 pack-year vs. ^0 pack-year) and clinical parameters (conventional fractionated RT vs. non-conventional RT, age) were studied. Cox proportional hazard model was used to determine significant indictors to CTCAE 3.0 Grade >3 acute pneumonitis.Results: Sixteen of 86 (18.6%) patients developed CTCAE 3.0 Grade 2 pulmonary fibrosis. There was no grade 4-5 toxicity. On univariate analysis, baseline FEV1 and age were significantly relative to severer acute pneumonitis, p=0.015 and p=0.041, respectively. MnSOD wl/0 ratio approached statistical significance (p=0.087). On multivariate analysis, baseline FEV1 (p = 0.0042, HR 0.221), MnSOD wl/0 ratio (p = 0.0061, HR 0.038) and heavy smoker (>20 pack-year, p = 0.0353, HR 3.388)...