Objective To investigate the result and side effect of late course accelerated three-di-mensional conformal radiotherapy (3DCRT) for esophageal carcinoma. Methods From July 2003 to March 2006, 55 patients with esophageal carcinoma receiving 3DCRT were randomly divided into late course accel-erated radiation group (group A, 27 patients) and conventional fractionation group (group B, 28 patients). The prescribed dose in group B was 64 -66 Gy, 2 Gy per fraction, 1 fraction per day, 5 fractions per week for about 6.5 weeks. Patients in group A received conventional fractionation irradiation for the first 4 weeks. Then the dose was increased to 3 Gy per fraction to a total dose of 67 -70 Gy. The treatment course in group A was about 6 weeks. The treatment response, acute site effects, 1-, 3-and 5-year local control rates and o-verall survival rates of the two groups were observed. Results In group A, 23 patients (85%) achievedcomplete response (CR) and 4(15%) achieved partial response (PR). While in group B, 16 patients (57%) achieved CR and 12(43%) achieved PR. The CR rate was significant higher in group A (χ~2 = 5.24,P=0.022). The 1-, 3-, 5-year local control rates were 85%, 54%, 54% in group A, and 70%, 56%, 33 % in group B (χ~2 = 0.68, P = 0.409), respectively. The 1 -,3-,5-year overall survival rates of the two groups were 81%, 37%, 29% and 61%, 39%, 23% (χ~2 = 0.06, P = O. 804), respectively. Both lo-cal control and overall survival were similar between the two groups. The incidences of acute radiation esoph-agitis in the two groups were similar (85% vs. 89% ;χ~2 =0. 00,P=0. 959), and the incidence of radiation pneumonitis was slightly higher in group A than in group B (67% vs 43% ;χ~2 =3.14,P =0.076). By the last follow up, 19 patients in group A and 21 in group B died. Among them, 10 in group A and 15 in group B died of local failure, while 7 in group A and 5 in group B died of metastasis. Conclusions When com-pared with conventional fractionation 3DCRT, late course accelerated 3DCRT for esophageal carcinoma can achieve better results in clinical response, though not in long-term local control or survival. The incidence of acute radiation esophagitis and pneumonitis is clinically acceptable.

Objective To observe the incidence of RP in NSCLC and esophageal carcinoma treated with 3DCRT and investigate the relationship between acute RP and lung function and dosimetric parameters.Methods From October 2006 to August 2008, 3DCRT plus concurrent chemotherapy of NP or LFP were applied to 64 patients with locally advanced NSCLC or esophageal carcinoma. twenty-three patients suffered form NSCLC and 41 patients from esophageal carcinoma, the prescription doses were 60 Gy/30fx and 58 -64 Gy/29 -32fx, respectively. Results For patients with esophageal carcinoma, 34% developed RP(9 grade 1,3 grade 2 and 2 grade 3). For patients with NSCLC, 96% developed RP(9 grade 1, 8 grade 2 and 5 grade 3). There was significant difference between the two groups(t =5. 55,P=0. 000). The FEV1.0/FVC and DLCO of patients with NSCLC were significantly lower than those of esophageal carcinoma, the ratio were 75.6%:82.7%(t=2.75,P=0.008)and 71.7%:81.0%(t=2.50, P=0.015),respectively. For patients whose FEV1.0, FEV1.0/FVC%, DLCO ＜80% and ≥80% before irradiation,the incidence of ≥2grade ARP were 35% vs 25% ,31% vs 26% and 35% vs 19%, respectively(x2 = 1.81,0.15,2. 13,P =0.179,0.697,0.144). While for patients whose FEV1.0 ＜ 70% and ≥70%, the incidence of severe ARP were 67% and 22% ,respectively(x2 =5.64, P =0.018). Spearman correlated analysis indicated that all the dosimetric parameters had relation with ≥ 2 grade ARP . The V20 of lung and MLD were found independently associated with RP according to multivariate analysis(x2 = 4.61,6.97, P = 0.032,0.008).Conclusions Parameters of basic lung function can predict the incidence of ≥2 grade RP to some extent,especially when the value of FEV1.0, FEV1.0/FVC%, and DLCO was lower. However, the V20 of lung and MLD may be the most valuable predictors.

Objective To investigate the value of PET-CT in TNM staging and three-conformal radiotherapy (3DCRT) in esophageal carcinoma. Methods From September 2007 to November 2008, 20 patients with pathologically confirmed esophageal carcinoma were enrolled, including 2 treated with surgery and 18 with 3DCRT. All the patients received PET-CT simulation before the treatment. The length and maximum transverse diameter of Grit based on esophagoscope, esophagography, CT, PET-CT_(SUV2.5) and PET-CT_(40%SUVmax) were compared. The TNM stages were compared based on CT and PET-CT images. Three treatment plans were produced and analyzed based on images of CT, PET-CT_(SUV2.5) and PET-CT_(40%SUVmax),respectively. Results The length of lesion on esophagoscope, esophagography, CT, PET-CT_(SUV2.5) and PET-CT_(40%SUVmax) was 4. 93 cm, 5.06 cm, 6. 67 cm, 5. 89 cm and 4. 84 cm, respectively. The corresponding maximum transverse diameter on the last 3 images was 4. 05 cm, 3.38 cm and 2. 95 cm, respectively. In all, 31, 21 and 14 positive lymph nodes were identified according to CT images, PET-CT images and the both. Five patients with CT diagnosed stage M_0 were found to have distant metastasis by PET-CT images, and 1 patient with CT diagnosed stage M_1 was excluded by PET-CT. The volumes of GTV_(CT) were similar with GTV_(SUV2.5) in 2 patients, smaller in 5 patients, and larger for the remaining 13 patients. For these 13 patients, the radiation dose of normal tissues based on GTV_(SUV2.5) was relatively lower. Conclusions The length of lesion based on PET -CT_(SUV2.5) matches the pathological length best , followed by esophagography. With PET-CT_(40%SUVmax) the actual lesion length may be underestimated. TNM stage might be changed by PET-CT, and then the target volumes and radiation doses of normal tissues might be reduced.

Objective To observe the incidence of radiation pneumonitis (RP) in patients with non-small cell lung cancer (NSCLC) or esophageal carcinoma treated by three-dimensional conformal radiation therapy (3DCRT), and investigate the predictive value of low dose volume of the lung on RP.Methods From Mar 2005 to Aug 2008, 161 patients with locally advanced NSCLC or esophageal carcinoma received 3DCRT. Fifty-three patients with NSCLC received radiotherapy of 60 Gy/30 -34 fx and concurrent chemotherapy of navelbine plus cisplatin (NP). Among the 108 patients with esophageal carcinoma with prescription doses ranging from 58 Gy/29 fx to 70 Gy/35 fx, 46 and 62 received 3DCRT alone or 3DCRT with concurrent chemotherapy of calcium folinate, 5-Fu and cisplatin (LFP), respectively. Univariate and multivariate analysis and receiver operating characteristics (ROC) curves were performed to assess the correlated factors of RP. Results The follow-up rate was 100%. The rate of RP was 57. 8% (93/161) for all patients, 94% (50/53,including 1 with Grade 4 and 1 with Grade 5) for patients with NSCLC, and 39.8% (43/108, none with Grade 4 and 5) for patients with esophageal carcinoma. The correlative factors included the sex, volume of gross tumor volume (GTV), mean lung dose (MLD), V5, V10, V15, V20, V25 and V30 of normal lung according to Spearman correlative analysis. Univariate analysis showed that all the 9 factors could predict RP. Only V5 and the volume of GTV were found independently associated with Grade 2 or more RP in multivariate analysis. ROC analysis indicated that the cut-off point of the curve was 55% with the area under ROC curve of 0. 684 (P = 0. 000). For the patients with lung V5 ≥ 55% and < 55% ,43% (36/84) and 18% (14/77) developed RP of Grade 2 or more, respectively. Conclusions Dosimetric parameters including MLD, normal tissue complication probability, and V5-V30 could predict RP. V5 may be the most valuable predictor. When V5 exceeds 55%, the probability of RP of grade 2 or more may increase notably. Besides the limitation of MLD, V20 and V30, the volume of low dose region should be also limited to a lower range during treatment planning.