doi: /j.ijrobp CLINICAL INVESTIGATION

Transcription

1 doi: /j.ijrobp Int. J. Radiation Oncology Biol. Phys., Vol. 76, No. 1, pp , 2010 Copyright Ó 2010 Elsevier Inc. Printed in the USA. All rights reserved /10/$ see front matter CLINICAL INVESTIGATION Lung ADDING IPSILATERAL V 20 AND V 30 TO CONVENTIONAL DOSIMETRIC CONSTRAINTS PREDICTS RADIATION PNEUMONITIS IN STAGE IIIA B NSCLC TREATED WITH COMBINED-MODALITY THERAPY SARA RAMELLA, M.D.,* LUCIO TRODELLA, M.D.,* TOMMASO CLAUDIO MINEO, y EUGENIO POMPEO, y GERARDINA STIMATO, PH.D.,* DIEGO GAUDINO, PH.D.,* VINCENZO VALENTINI, M.D., z FRANCESCO CELLINI, M.D.,* MARZIA CIRESA, M.D.,* MICHELE FIORE, M.D.,* ANGELO PIERMATTEI, PH.D., x PATRIZIA RUSSO, M.D., k ALFREDO CESARIO, M.D., {# AND ROLANDO M. D ANGELILLO, M.D.* *Radiotherapy Unit, Campus Bio-Medico University, Rome, Italy; y Department of Thoracic Surgery, Tor Vergata University, Rome, Italy; z Department of Radiation Therapy, Catholic University, Rome, Italy; x Department of Physics, Catholic University, Rome, Italy; jj Lung Cancer Unit, National Cancer Research Institute, Genoa, Italy; { Clinical CdC San Raffaele, Velletri, Italy; and # Department of Thoracic Surgery, Catholic University, Rome, Italy Purpose: To determine lung dosimetric constraints that correlate with radiation pneumonitis in non small-cell lung cancer patients treated with three-dimensional radiation therapy and concurrent chemotherapy. Methods and Materials: Between June 2002 and December 2006, 97 patients with locally advanced non smallcell lung cancer were treated with concomitant radiochemotherapy. All patients underwent complete three-dimensional treatment planning (including dose-volume histograms), and patients were treated only if the percentage of total lung volume exceeding 20 Gy (V 20 ) and 30 Gy (V 30 ), and mean lung dose (MLD) had not exceeded the constraints of 31%, 18%, and 20 Gy, respectively. The total and ipsilateral lung dose volume histogram parameters, planning target volume, and total dose delivered were analyzed and correlated with pneumonitis incidence. Results: If dose constraints to the total lung were respected, the most statistically significant factors predicting pneumonitis were the percentage of ipsilateral lung volume exceeding 20 Gy (V 20 ipsi), percentage of ipsilateral lung volume exceeding 30 Gy (V 30 ipsi), and planning target volume. These parameters divided the patients into low- and high-risk groups: if V 20 ipsi was 52% or lower, the risk of pneumonitis was 9%, and if V 20 ipsi was greater than 52%, the risk of pneumonitis was 46%; if V 30 ipsi was 39% or lower, the risk of pneumonitis was 8%, and if V 30 ipsi was greater than 39%, the risk of pneumonitis was 38%. Actuarial curves of the development of pneumonitis of Grade 2 or higher stratified by V 20 ipsi and V 30 ipsi were created. Conclusions: The correlation between pneumonitis and dosimetric constraints has been validated. Adding V 20 ipsi and V 30 ipsi to the classical total lung constraints could reduce pulmonary toxicity in concurrent chemoradiation treatment. V 20 ipsi and V 30 ipsi are important if the V 20 to the total lung, V 30 to the total lung, and mean lung dose have not exceeded the constraints of 31%, 18%, and 20 Gy, respectively. Ó 2010 Elsevier Inc. Lung cancer, Radiochemotherapy, Pneumonitis, Dosimetric constraints, Ipsilateral lung. INTRODUCTION In the overall treatment strategy for non small-cell lung cancer (NSCLC) patients, the pulmonary toxicity is an important limiting factor. Several authors have previously explored the role of dosimetric predictive factors, such as the percentage of total lung volume exceeding a defined dose (Vdose) and mean lung dose (MLD), to plan for an optimized strategy to reduce radiation pneumonitis (RP) and/or escalate radiation dose. Among these, Graham et al. (1) and Hernando et al. (2) have correlated the value of percentage of lung volume exceeding 20 Gy (V 20 ), percentage of lung volume exceeding 30 Gy (V 30 ), and MLD stratifying the risk of RP developing. Their data have been widely used by the community and have improved the way in which curative radiation doses are delivered. This applies, as well, in patients receiving concurrent chemoradiation (3). Nevertheless, the role of concurrent chemotherapy to radiation in treating locally advanced NSCLC has been gaining substantial value in the recent years, and this has encouraged Note An online CME test for this article can be taken at asro.astro.org under Continuing Education. Reprint requests to: Sara Ramella, M.D., Via E. Longoni 49, 00155, Rome, Italy. Tel: (+39) ; Fax: (+39) ; s.ramella@unicampus.it 110 Conflict of interest: none. Received Sept 9, 2008, and in revised form Jan 20, Accepted for publication Jan 21, 2009.

2 Ipsilateral lung constraints d S. RAMELLA et al. 111 some authors to review their data and report the RP rate in patients receiving a multimodality treatment (4). Paradoxically, the Vdose and MLD referring to both lungs as a single functional unit do not take into account the possible chances of an imbalance among the doses to the lung with the primary tumor and the contralateral lung; substantially, the final value of each dosimetric parameter is just a mean value for the total lung parenchyma. So, an extreme case could be characterized by no dose to the contralateral lung by concentrating the radiation beams on one lung. The focus of this trial is to assess the relevance of ipsilateral dosimetric parameters on RP. We report the results of our experience in patients treated with concurrent chemoradiotherapy for NSCLC obtained by analyzing the value of adding ipsilateral V 20 (V 20 ipsi), ipsilateral V 30 (V 30 ipsi), and ipsilateral MLD (MLDipsi) to total lung volume constraints. METHODS AND MATERIALS Between June 2002 and December 2006, after Institutional Review Board approval, 97 consecutive patients with locally advanced NSCLC underwent concomitant radiochemotherapy indicated and applied with radical or neoadjuvant (inductive) intent. The patients characteristics, tumor stage, and radiation dose delivered are shown in Table 1. Minimum follow-up was 6 months. In all patients a history was obtained and physical examination and staging studies were performed. These included chest, upper abdomen, and brain computed tomography (CT) and bone scans; positron emission tomography was included starting in June Pulmonary function was evaluated at the beginning and after the end of combined treatment according to standardized protocols following the American Thoracic Society guidelines to determine acceptability (5). The clinical stage was IIIA in 72 patients and IIIB in 25 patients. No patients in Stage IV were included. The primary tumor was located in the upper lobe in 34% of patients, in the middle lobe in 57% and in the lower lobe in the remaining 9%, 46% of tumors were in the right lung and 54% in the left lung. Chemotherapy included 80 mg/m 2 of cisplatin the first and last week of radiation therapy treatment and 350 mg/m 2 of gemcitabine every week during radiotherapy or gemcitabine alone. These approaches have been previously described in a Phase I and II trial (6, 7). Radiotherapy was delivered by use of a linear accelerator (Clinac C2100; Varian Medical Systems, Palo Alto, CA) with a 6- to 15-MV photon beam. All patients underwent three-dimensional treatment planning by use of Eclipse computer software (version 6.5; Varian Medical Systems). Treatment planning was based on CT scans with 5-mm section thickness and 5-mm intervals obtained in the treatment position; tissue heterogeneity correction was routinely performed. All patients were immobilized by customized devices. Radiotherapy was administered with an angled field technique (coplanar and noncoplanar fields) to include the entire target volume (planning target volume [PTV]) in the isodose 100% (range, 95% 107%) area. The median International Commission on Radiation Units & Measurements total referred dose was 50.4 Gy in the neoadjuvant setting and 59.4 Gy in the radical setting with standard fractionation (1.8 Gy/d). The gross tumor volume (GTV) was defined as tumor extension and metastatic lymph nodes. The clinical target volume was equal to GTV, and the PTV consisted of GTV plus a 1-cm margin. Elective nodal irradiation was never administered. The primary GTV was contoured on the CT parenchyma window, whereas the nodal GTV was contoured on the mediastinum only. Table 1. Patient characteristics and tumor stage Age (y) (median, 66) Gender Male 86 Female 11 ECOG Histologic type Squamous cell 50 Adenocarcinoma 26 Large cell 3 NSCLC NOS 18 Clinical disease stage IIIA 77 IIIB 20 Primary tumor location Upper lobe 33 Middle or lower lobe 64 Radiotherapy dose delivered 50.4 Gy Gy 23 Completed treatment early 3 Surgery Yes 46 No 51 Abbreviations: ECOG = Eastern Cooperative Oncology Group; NSCLC NOS = non small-cell lung cancer, not otherwise specified. The right lung and the left lung were contoured separately and then taken into consideration as a single structure called total lung, which was defined as follows, as in the study of Graham et al. (1): (Right lung + Left lung) PTV. Accordingly, the ipsilateral lung was defined as Right lung/left lung PTV. All treatment plans were approved if conventional dosimetric lung constraints were respected: V 20 of 31% or lower (1), V 30 of 18% or lower (2), and MLD of 20 Gy or lower (1, 2); moreover, the dose delivered to the spinal cord was limited to less than 40 Gy; the percentage of esophagus volume exceeding 55 Gy (V 55 ), 30% of the organ volume or lower (8); and finally, the percentage of heath volume exceeding 40 Gy (V 40 ), 50% or lower (9). V 20,V 30, and MLD were also evaluated on the ipsilateral lung (which is the lung on the side containing >50% of the PTV), and they were registered as V 20 ipsi, V 30 ipsi, and MLDipsi, respectively. Then, PTV and total dose delivered were correlated with the risk of pneumonitis and Vdose values. After treatment completion, patients were followed up every 4 weeks usually for up to 6 months and every 2 months if the general status was stable on an outpatient basis. RP was graded according to the National Cancer Institute Common Toxicity Criteria, version 3.0 (10). Statistics Univariate logistic regression analysis was used to test the association between the Vdose and MLD to RP. For the multivariate analysis, forward stepwise logistic regression analysis was adopted. Actuarial curves were analyzed by use of Kaplan-Meier survival analysis (11, 12). All statistical tests were 2-sided, and p # 0.05 was considered statistically significant (13).TheV 20 ipsi and V 30 ipsi values were selected according to the cumulative incidence of pneumonitis (1,2,4). RESULTS After a median follow-up period of 29 months (range, 6 45 months), 14 of the 97 patients had RP of Grade 2 or

3 112 I. J. Radiation Oncology d Biology d Physics Volume 76, Number 1, 2010 Table 2. Incidence and grade of pneumonitis Grade (Common Toxicity Criteria, version 3.0) Incidence (no. of patients) None (fatal) 2 higher, yielding a mean rate of 14.4% (Table 2). In the group of surgically resected patients, the incidence was 8.7% (4 of 46), whereas in the nonresected patients, it was 19.6% (10 of 51) (p = 0.15). For the entire cohort, the lung injury arose after a median period of 79 days from the beginning of radiotherapy; in the surgical and nonsurgical groups, the median time was 123 and 55 days, respectively. The two fatal toxicities arose in nonresected patients after 45 and 60 days from the start of treatment, respectively. The first developed in a patient with a contraindication to a chemotherapy alone based protocol (this particular patient was affected by an autoimmune disease, mixed connectivitis, which is known to be characterized by an abnormal fibrogenic reaction to radiation). The second was probably caused by an infection in a radiation-induced injury because a control pulmonary CT scan documented a bilateral and diffuse inflammatory reaction. Univariate analysis was performed on several constraints to analyze their role on the development of pneumonitis. Results are shown in Table 3. The most statistically significant factors predicting pneumonitis were V 20 ipsi and V 30 ipsi (percentage of ipsilateral lung volume exceeding 20 Gy and 30 Gy, respectively). If we analyze the no-rp group and the RP group, V 20 is always below the cutoff proposed by Graham et al. (1) (V 20 #31%); there are no differences in the mean values of the two groups (19.5 vs. 22.5, p = 0.121). There were also no differences for V 30 (13.1 vs. 15.5, p = 0.09), MLD (13.9 vs. 14.7, p = 0.687), or MLDipsi (22.5 vs. 26.7, p = 0.135). By contrast, the difference in ipsilateral dose between the RP group and the no-rp group is statistically significant not only for V 20 ipsi (38.8 vs. 48.4, p = 0.010) but also for V 30 ipsi (28.4 vs. 38.7, p = 0.001). Our analysis showed that these parameters divided patients into low- and highrisk groups; when V 20 ipsi was 52% or lower, the risk of pneumonitis was 9%, and when V 20 ipsi was greater than 52%, the risk of pneumonitis was 46% (Table 4). Actuarial curves of the development of pneumonitis of Grade 2 or higher stratified by V 20 ipsi and V 30 ipsi are shown in Figs. 1 and 2. PTV was investigated to understand whether the risk of pneumonitis is associated with large irradiated volumes. The mean PTV was 390 ml. The mean values in patients with RP and those with no RP were 547 ml and 367 ml, respectively (p = 0.007) (Table 5). The capability to reach lower values of V 20 ipsi is not limited to small PTV (p = 0.07). Multivariate analysis was performed for the development of pneumonitis of Grade 2 or higher, and the results are shown in Table 6. The only significant factors were V 20 ipsi, V 30 ipsi, and PTV. Total delivered dose was not correlated with pneumonitis (p = 0.802). Total dose was not correlated with total lung V 20 and MLD (p = 0.18 for V 20 and p = 0.29 for MLD), whereas V 30 was significantly lower in the group treated with a higher dose level. Neither ipsilateral constraints were influenced by total delivered dose (p = for V 20 ipsi, p = for V 30 ipsi, and p = 0.06 for MLD); this implies that higher doses are not necessarily correlated with higher dosimetric values. In our clinical practice, adding the ipsilateral constraints to the standard lung dosimetric factors determined a fall in the incidence of RP from 14.4% to 6.8%. DISCUSSION A review of published studies shows that the ideal dose volume histogram metric for the prediction of the risk of RP has not yet been identified (14). However, it has been established with certainty that adding chemotherapy to radiation therapy will cause a rise in toxicity rate (4). In the study of Graham et al. (1) half of the patients were treated with radiotherapy and half with radiochemotherapy. The incidence of pneumonitis was up to 20% at 24 months but just 7% if the V 20 value was 31% or lower. In the experience of Hernando et al. (2) the global rate of pulmonary toxicity was 19%, but it was only 6% when V 30 was 18% or lower. Instead, 57% of our pneumonitis patients had V 30 of 18% or lower. However, in that study only 6.5% of patients (13 of 201) were treated with concurrent radiochemotherapy. Table 3. Dosimetric constraints in RP and no-rp groups V 20 ipsi V 30 ipsi MLDipsi V 20 V 30 MLD No-RP group 38.8 (11.8) 28.4 (10) 22.5 (9.5) 19.5 (6.6) 13.1 (4.9) 13.9 (6.4) RP group 48.4 (17.5) 38.7 (13.6) 26.7 (9.4) 22.5 (7.6) 15.5 (5.3) 14.7 (3.6) p Value p = p = p = p = p = p = NOTE. Data are given as mean (SD). Abbreviations: RP = radiation pneumonitis; V 20 ipsi = percentage of ipsilateral lung volume exceeding 20 Gy; V 30 ipsi = percentage of ipsilateral lung volume exceeding 30 Gy; MLDipsi = ipsilateral mean lung dose; V 20 = percentage of lung volume exceeding 20 Gy; V 30 = percentage of lung volume exceeding 30 Gy; MLD = mean lung dose.

4 Ipsilateral lung constraints d S. RAMELLA et al. 113 Table 4. Risk groups of patients according to ipsilateral constraints Variable Cutoff Actuarial incidence of lung toxicity of Grade 2 or higher (Common Toxicity Criteria, version 3.0) p Value (Fisher exact test) V 20 ipsi #52% 9% p = >52% 46% V 30 ipsi #39% 8% p = >39% 38% MLDipsi #22 Gy 7% p = 0.04 >22 Gy 23% Abbreviations: V 20 ipsi = percentage of ipsilateral lung volume exceeding 20 Gy; V 30 ipsi = percentage of ipsilateral lung volume exceeding 30 Gy; MLDipsi = ipsilateral mean lung dose. In the trial of Tsujino et al. (4), all patients were treated by adding chemotherapy to radiotherapy in a concomitant set, and the incidence of pneumonitis rose up to 51% with the standard dosimetric constraints. Only in the case of very low values of V 20 did the pneumonitis risk become acceptable (8.7% for V 20 #20%). It follows that the use of constraints lower than V 20 becomes paramount to avoid pulmonary toxicity. In common clinical experience, these low constraints are difficult to achieve during the treatment planning phases. Consequently, this process becomes difficult and time-consuming for physicians and for physicists, especially when the volume of the irradiated field is large (15). This fosters additional applied research in investigating the hypotheses in radiation-induced pneumonitis. Fig. 2. Univariate analysis showing cumulative hazard of development of pneumonitis of Grade 2 or higher stratified by two levels of V 30 ipsi (p = 0.012). Our results are consistent with the hypothesis that V 20 ipsi is a new predictive constraint of lung injury in concurrent chemoradiation treatments. Pulmonary function is certainly the result of the right and left lung function; for this reason, in investigations related to toxicity, the lungs were usually considered as a unique organ unit. To date and to our best knowledge, only few authors have examined the effect of the radiation dose applied to the ipsilateral and contralateral (with respect to the tumor location) lungs separately. In the study of Oetzel et al. (16) the lungs were analyzed as separate organs. The toxicity rate was 0% if MLDipsi was 15 Gy or lower, 13% if MLDipsi was 17.5 to 20 Gy, 21% if MLDipsi was 22.5 to 25 Gy, and 43% if MLDipsi was 27.5 Gy or greater. Correlation between dose to the ipsilateral lung and pneumonitis has been reported in two studies: Seppenwoolde et al. (17) and Yorke et al. (18). The strongest correlations with pneumonitis in the last study were found for the percentage of lung volume exceeding 5 by (V 5 ) through V 30 in the ipsilateral lung, but the authors do not suggest a constraint. In our results V 20 ipsi was proven as an effective stratification criterion substantially dividing the patients into two Table 5. Mean value and SD of PTV in RP group and in no-rp group PTV (mean) (ml) SD Fig. 1. Univariate analysis showing cumulative hazard of development of pneumonitis of Grade 2 or higher stratified by two levels of V 20 ipsi (p = 0.001). No-RP group RP group Abbreviations: PTV = planning target volume; RP = radiation pneumonitis.

5 114 I. J. Radiation Oncology d Biology d Physics Volume 76, Number 1, 2010 Table 6. Multivariate analysis of factors related to development of pneumonitis of Grade 2 or higher p Value V 20 ipsi V 30 ipsi MLDipsi V V MLD PTV Dose Abbreviations: V 20 ipsi = percentage of ipsilateral lung volume exceeding 20 Gy; V 30 ipsi = percentage of ipsilateral lung volume exceeding 30 Gy; MLDipsi = ipsilateral mean lung dose; V 20 = percentage of lung volume exceeding 20 Gy; V 30 = percentage of lung volume exceeding 30 Gy; MLD = mean lung dose; PTV = planning target volume. groups (Fig. 1). The V 20 ipsi threshold value was 52% where toxicity rate rose from 9% (if V 20 ipsi was #52%) to 46% (if V 20 ipsi was >52%) (p < 0.05). The same was true for V 30 ipsi and MLDipsi (Table 4). V 20 ipsi and V 30 ipsi are important if the V 20 to the total lung, V 30 to the total lung, and MLD have not exceeded the constraints of 31%, 18%, and 20 Gy, respectively. Moreover, in relation to the development of pneumonitis, the volume (PTV) is a significant factor in confirming (intuitively) that large irradiated volumes are at risk of toxicity (15). What is interesting, however, is the possibility of obtaining low V 20 ipsi values even for large PTV values (in fact, there is no significant difference in V 20 ipsi for PTV according to dimension). Total dose was not correlated with pneumonitis. In addition, an interesting observation in our experience was that higher total doses were not related either to higher total lung constraints or to ipsilateral ones, so the possibility of delivering higher doses is not limited. Recently, some authors have stressed that a controversial area of predicting RP is the determination of whether to give a lot to a little or a little to a lot (19, 20). In the study of Willner et al. (19), conducted on radiochemotherapy-treated patients, the authors studied the percentage of lung volume exceeding 10 Gy (V 10 ), V 20,V 30, and V 40 and concluded that a large dose (e.g., 40 Gy) to a small volume of normal lung is worse than a small dose (e.g., 10 Gy) to a large volume. On the other hand, Gopal et al. (20) reached the opposite conclusion: a small dose (13 Gy) to a large volume of normal lung is worse than a large dose to a small volume for the development of lung injury. However, even if no clear conclusions can be deduced from the published literature, V 20 and V 30 are well-established predictors of pneumonitis and they represent an intermediate dose threshold. Surgery could be a confounding factor. Some historically reported data suggest that surgery itself may induce a pulmonary inflammatory reaction, clearly shown by the significant incidence of adult respiratory distress syndrome in the postoperative period. The exact correlation among surgery, inflammation, and the subsequent onset of RP remains unclear. It could be speculated, however, that surgery removes only high-dose PTV-containing lung regions. Regions treated with lower doses (20 and 30 Gy) are generally not removed and could play a role in the process of the development of pneumonitis. A demonstration could come from the trial of Albain et al. (21) in which Grade 3 or 4 lung toxicity was not so different between resected and nonresected patients (18% vs. 24%). In our experience, although the differences did not reach statistical significance, the sample size and the low incidence rate in both groups can contribute to the lack of power to detect a significant value. In fact, the data do not rule out the possibility that removing the high-dose radiotherapy region of the lung may reduce the risk of pneumonitis. Lung injury arose after a median period of 79 days from the beginning of radiotherapy; in the surgical and nonsurgical groups the median time was 123 and 55 days, respectively. This is only 1 to 2 weeks after radiation. In comparison to the common findings of 10 to 12 weeks after radiation for the peak incidence of RP, this is quite early. Data related to our previous experience with gemcitabinebased concurrent radiochemotherapy (6, 7) have been reanalyzed according to this suggestion. In all cases RP cases have been observed earlier in comparison to the common figure of 10 to 12 weeks after radiation. The peak incidence was recorded at 10 to 15 days after treatment. Given the outcome of this reanalysis, we agree that it could be speculated that gemcitabine can actively contribute to the early onset of RP. As a matter of fact, the potential of the drug as a radiosensitizer is very well known. According to Mehta s hypothesis (22), the safest approach is likely to be limiting both the volume of normal lung that is irradiated and the amount of radiation that it receives. Following the lines of simplification, we could summarize that the procedure of adding ipsilateral lung constraints to total lung constraints has a beneficial effect. In fact, as observed in our own experience and other authors trials (4), an increasing lung injury is recorded even if total lung constraints are respected. Differently from the trials of Graham et al. (1) and Hernando et al. (2) where reported RP rate was 6% and 7%, respectively, we [and Tsujino et al. (4)] have observed a 14.4% (and 51%) rate of lung toxicity in the same setting. We, therefore, recommend that treatment planning should be improved (by adding ipsilateral lung constraints) to fully exploit the potential of combined concurrent radiation and chemotherapy treatment strategies where, otherwise, enhanced toxicity could represent a significant limiting factor. In this setting V 20 ipsi and V 30 ipsi constraints would be useful and simple lung injury predictive factors. To respect these parameters, we suggest that all beam entrances should not be on the ipsilateral lung.

6 Ipsilateral lung constraints d S. RAMELLA et al. 115 REFERENCES 1. Graham MV, Purdy JA, Emami B, et al. Clinical dose volume histogram analysis for pneumonitis after 3D treatment for non small-cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys 1999;45: Hernando ML, Marks LB, Bentel GC, et al. Radiation-induced pulmonary toxicity: A dose volume histogram analysis in 201 patients with lung cancer. Int J Radiat Oncol Biol Phys 2001; 51: Socinski MA, Morris DE, Halle JS, et al. Induction and concurrent chemotherapy with high-dose thoracic conformal radiation therapy in unresectable stage IIIA and IIIB non-small-cell lung cancer: A dose-escalation phase I trial. J Clin Oncol 2004;22: Tsujino K, Hirota S, Endo M, et al. Predictive value of dose-volume histogram parameters for predicting radiation pneumonitis after concurrent chemoradiation for lung cancer. Int J Radiat Oncol Biol Phys 2003;55: Standardization of spirometry 1987 update. Statement of the American Thoracic Society. Am Rev Resp Dis 1987;136: Trodella L, Granone P, Valente S, et al. Phase I trial of weekly gemcitabine and concurrent radiotherapy in patients with inoperable non-small-cell lung cancer. J Clin Oncol 2002;20: Galetta D, Cesario A, Margaritora S, et al. Multimodality treatment of unresectable stage III non-small cell lung cancer: Interim analysis of a phase II trial with preoperative gemcitabine and concurrent radiotherapy. J Thorac Cardiovasc Surg 2006; 131: Belderbos J, Heemsbergen W, Hoogeman M, et al. Acute esophageal toxicity in non-small cell lung cancer patients after high dose conformal radiotherapy. Radiother Oncol 2005;75: Giraud P, Cosset JM. Radiation toxicity to the heart: Physiopathology and clinical data [In French]. Bull Cancer 2004;91: Trotti A, Colevas AD, Setser A, et al. CTCAE v3.0: Development of a comprehensive grading system for the adverse effects of cancer treatment. Semin Radiat Oncol 2003;13: Kaplan EM, Meier P. Non-parametric estimation for incomplete observations. J Am Stat Assoc 1958;53: StartXact user manual. Cambridge, MA: Cytel Software; p Cox DR. Regression models and life-table. J R Stat Soc 1972; 334: Rodrigues G, Lock M, D Souza D, et al. Prediction of radiation pneumonitis by dose-volume histogram parameters in lung cancer A systematic review. Radiother Oncol 2004;71: Armstrong JC. Target volume definition in 3D conformal radiation therapy of lung cancer. Br J Radiol 1998;71: Oetzel D, Schraube P, Hensley F, et al. Estimation of pneumonitis risk in three-dimensional treatment planning using dosevolume histogram analysis. Int J Radiat Oncol Biol Phys 1995;33: Seppenwoolde Y, De Jaeger K, Boersma LJ, et al. Regional differences in lung radiosensitivity after radiotherapy for non small-cell lung cancer. Int J Radiat Oncol Biol Phys 2004;60: Yorke ED, Jackson A, Rosenzweig KE, et al. Correlation of dosimetric factors and radiation pneumonitis for non small-cell lung cancer patients in a recently completed dose escalation study. Int J Radiat Oncol Biol Phys 2005;63: Willner J, Jost A, Baier K, et al. A little to a lot or a lot to a little? An analysis of pneumonitis risk from dose-volume histogram parameters of the lung in patients with lung cancer treated with 3-D conformal radiotherapy. Strahlenther Onkol 2003; 179: Gopal R, Tucker SL, Komaki R, et al. The relationship between local dose and loss of function for irradiated lung. Int J Radiat Oncol Biol Phys 2003;56: Albain KS, Scott CB, Rusch VR, et al. Phase III comparison of concurrent chemotherapy plus radiotherapy (CT/radiotherapy) and CT/radiotherapy followed by surgical resection for stage IIIA(pN2) non-small cell lung cancer (NSCLC): Initial results from Intergroup trial 0139 (RTOG 93-09) [Abstract]. Proc Am Soc Clin Oncol 2003;22: Mehta V. Radiation pneumonitis and pulmonary fibrosis in non small-cell lung cancer: Pulmonary function, prediction and prevention. Int J Radiat Oncol Biol Phys 2005;63:5 24.