Portal de Periódicos da CAPES

High-dose radiation improved local tumor control and overall survival in patients with inoperable/unresectable non–small-cell lung cancer: Long-term results of a radiation dose escalation study

2005; Elsevier BV; Volume: 63; Issue: 2 Linguagem: Inglês

10.1016/j.ijrobp.2005.02.010

1879-355X

Feng‐Ming Kong, Randall K. Ten Haken, Matthew J. Schipper, Molly A. Sullivan, Ming Chen, Carlos López, Gregory P. Kalemkerian, James A. Hayman,

Effects of Radiation Exposure

Purpose: To determine whether high-dose radiation leads to improved outcomes in patients with non–small-cell lung cancer (NSCLC). Methods and Materials: This analysis included 106 patients with newly diagnosed or recurrent Stages I–III NSCLC, treated with 63–103 Gy in 2.1-Gy fractions, using three-dimensional conformal radiation therapy (3D-CRT) per a dose escalation trial. Targets included the primary tumor and any lymph nodes ≥1 cm, without intentionally including negative nodal regions. Nineteen percent of patients (20/106) received neoadjuvant chemotherapy. Patient, tumor, and treatment factors were evaluated for association with outcomes. Estimated median follow-up was 8.5 years. Results: Median survival was 19 months, and 5-year overall survival (OS) was 13%. Multivariate analysis revealed weight loss (p = 0.011) and radiation dose (p = 0.0006) were significant predictors for OS. The 5-year OS was 4%, 22%, and 28% for patients receiving 63–69, 74–84, and 92–103 Gy, respectively. Although presence of nodal disease was negatively associated with locoregional control under univariate analysis, radiation dose was the only significant predictor when multiple variables were included (p = 0.015). The 5-year control rate was 12%, 35%, and 49% for 63–69, 74–84, and 92–103 Gy, respectively. Conclusions: Higher dose radiation is associated with improved outcomes in patients with NSCLC treated in the range of 63–103 Gy. Purpose: To determine whether high-dose radiation leads to improved outcomes in patients with non–small-cell lung cancer (NSCLC). Methods and Materials: This analysis included 106 patients with newly diagnosed or recurrent Stages I–III NSCLC, treated with 63–103 Gy in 2.1-Gy fractions, using three-dimensional conformal radiation therapy (3D-CRT) per a dose escalation trial. Targets included the primary tumor and any lymph nodes ≥1 cm, without intentionally including negative nodal regions. Nineteen percent of patients (20/106) received neoadjuvant chemotherapy. Patient, tumor, and treatment factors were evaluated for association with outcomes. Estimated median follow-up was 8.5 years. Results: Median survival was 19 months, and 5-year overall survival (OS) was 13%. Multivariate analysis revealed weight loss (p = 0.011) and radiation dose (p = 0.0006) were significant predictors for OS. The 5-year OS was 4%, 22%, and 28% for patients receiving 63–69, 74–84, and 92–103 Gy, respectively. Although presence of nodal disease was negatively associated with locoregional control under univariate analysis, radiation dose was the only significant predictor when multiple variables were included (p = 0.015). The 5-year control rate was 12%, 35%, and 49% for 63–69, 74–84, and 92–103 Gy, respectively. Conclusions: Higher dose radiation is associated with improved outcomes in patients with NSCLC treated in the range of 63–103 Gy. IntroductionTreatment outcomes of patients with non–small-cell lung cancer (NSCLC) have changed very little over the last 3 decades, with the 5-year overall survival rate remaining in the range of 10–15%. Although surgical resection is the treatment of choice for patients with early-stage disease, the majority of patients cannot undergo surgery because of either medical comorbidities or the extent of their disease. Based on Surveillance Epidemiology and End Results data in the United States between 1995 and 2001, only 70% of patients with localized lung cancer had surgery (1Surveillance Epidemiology and End Results (SEER) Available at: http://seer.cancer.gov/csr/1975_2001/results_merged/sect_15_lung_bronchus.pdf. Accessed October 2004.Google Scholar), whereas the majority of patients with locally advanced NSCLC cannot have surgery. Radiation, the mainstay of local nonsurgical treatment, is associated with poor outcomes. In several studies of patients with medically inoperable early-stage diseases, the 5-year overall survival rates ranged from 10–34% (2Graham M.V. Purdy J.A. Emami B. et al.Preliminary results of a prospective trial using three-dimensional radiotherapy for lung cancer.Int J Radiat Oncol Biol Phys. 1995; 33: 993-1000Abstract Full Text PDF PubMed Scopus (123) Google Scholar, 3Sibley G.S. Jamieson T.A. Marks L.B. et al.Radiotherapy alone for medically inoperable Stage I non–small-cell lung cancer The Duke experience.Int J Radiat Oncol Biol Phys. 1998; 40: 149-154Abstract Full Text Full Text PDF PubMed Scopus (318) Google Scholar, 4Cox J.D. Azarnia N. Byhardt R.W. et al.A randomized phase I/II trial of hyperfractionated radiation therapy with total doses of 60.0 Gy to 79.2 Gy: Possible survival benefit with > or equal to 69.6 Gy in favorable patients with Radiation Therapy Oncology Group stage III non-small-cell lung carcinoma: Report of Radiation Therapy Oncology Group 83-11.J Clin Oncol. 1990; 8: 1543-1555Crossref PubMed Scopus (460) Google Scholar), far inferior to the surgical rates of 50–70% (5Mountain C.F. Revisions in the international system for staging lung cancer.Chest. 1997; 111: 1710-1717Crossref PubMed Scopus (4522) Google Scholar). For patients with locally advanced unresectable disease treated with radiotherapy alone, the 5-year survival rate is 0.20–0.25, >0.25–0.31, >0.31–0.40, and >0.40. After 1997, model parameters (n = 1.00, m = 0.33, and TD50 = 33.00) were updated based on the trial data and both lungs were combined to estimate the Veff. The boundaries of the five bins were adjusted. The Veffs for the revised bins were as follows: 0–0.12, >0.12–0.18, >0.18–0.24, >0.24–0.31, and >0.31–0.40. Table 1 shows the dose assignment within individual bins. Radiation doses ranged from 63–103 Gy based on the Veff bins and the timing of enrollment. Tumors of similar size might have received different doses because of tumor location geometry, the presence of mediastinal nodal involvement, small total lung volume, or the level of dose escalation.Table 1Dose escalation schemaNTCP (%)Lung Veff bins (Gy)1 (0–0.12)2 (>0.12–0.18)3 (>0.18–0.24)4 (>0.24–0.31)5 (>0.31–0.40)184.069.3292.475.63102.984.065.1592.469.363.07102.975.665.11084.069.363.01492.475.665.12084.069.32892.475.6Abbreviations: Veff = effective volume; NTCP = normal tissue complication probability. Open table in a new tab Radiation therapy techniqueAll patients underwent planning and treatment while lying supine in a low-density cradle and breathing freely. At the time of simulation, fluoroscopy was performed from both the anterior and lateral projections to measure the maximum excursion of the primary tumor in the superior-inferior and anterior-posterior directions. Treatment-planning CT scans were performed with patients in the treatment position, included the entire thorax, used a minimum of 5-mm cuts through the target volume, and were performed using i.v. contrast. To minimize the volume of normal lung irradiated, we restricted the gross tumor volume (GTV) to include only the primary tumor, any hilar or mediastinal lymph nodes with a short-axis diameter of at least 1 cm on CT, and any abnormal findings detected on bronchoscopy or mediastinoscopy. The primary GTV was drawn using a lung CT window and level, whereas abnormal lymph node volumes were drawn using a mediastinal window and level. In cases of extensive atelectasis or pneumonia, the primary GTV was left to physician discretion. The clinical target volume (CTV) was routinely created by expanding the GTV by 0.5 cm. Clinically uninvolved hilar, mediastinal, and supraclavicular nodal regions were not purposely included in the CTV. The planning target volume (PTV) was created by expanding the CTV by a minimum of 0.5 cm for setup error. An additional margin (typically 0.5–1.0 cm) was added when necessary to account for respiratory motion. The goal of treatment planning in this study was to develop a plan that minimized both the dose to the surrounding normal tissue and the volume of normal lung irradiated (i.e., Veff) while providing coverage of the PTV by at least the 95% isodose surface. The 100% isodose line was defined at the isocenter, and dose (corrected for tissue electron density heterogeneity) was prescribed to this point.Typically, a single treatment plan was used that consisted of a median of three noncoplanar static fields (range, 2–7). All patients received daily treatment using 2.1-Gy fractions with beams ranging in energy from 6 MV to 25 MV. Tissue heterogeneity was corrected using the equivalent path length algorithm. The fraction size of 2.1 Gy was selected based on a retrospective assessment of the treatment plans of previously treated patients to be on average roughly equivalent to an uncorrected 2-Gy fraction when using an average lung density correction factor, and also accounting for subsequent changes in tumor and lung DVH shapes due to redistribution of computed dose.All patients who received chemotherapy were required to undergo treatment-planning CT scans both before receiving their first dose of chemotherapy and several days before starting treatment with radiation therapy. In patients who received chemotherapy, the GTV consisted of the postchemotherapy primary tumor mass and any nodes that had been larger than 1 cm in short-axis diameter on the prechemotherapy CT scan. If the primary tumor or enlarged lymph nodes previously seen on the prechemotherapy CT scan were no longer visible on the postchemotherapy scan (i.e., the patient experienced a complete response at that site), its epicenter was identified using the prechemotherapy CT scan, expanded by 1.5 cm, and included in the CTV.Follow-upPatients who completed protocol therapy were seen in follow-up to monitor and record toxicity, response, and recurrence at 1 month from completion of therapy, then every 3 months for 2 years, then every 4 months for 1 year, then every 6 months for 2 years, and yearly thereafter. Patients underwent a chest X-ray at every visit and a CT scan of the chest at least at 6 months. For patients with bronchoscopically visible disease at diagnosis, a repeat bronchoscopy was to have been performed at 6 months. Pulmonary function tests were repeated every 6 months until stable over two assessments.Endpoints and data analysisMedian survival (MS), actuarial survival rates, and time to locoregional disease progression were the endpoints of this analysis, and were dated from the initiation of radiation. Four different types of survival were analyzed; locoregional progression free (equivalent to locoregional tumor control), distant progression free, any progression free, and overall survival (OS). Locoregional progression was determined by a radiologist if there was a significant increase in the volume of tumor within the thorax when compared with the previous CT scan. Estimates of each survival type were calculated using the Kaplan-Meier method (19Kaplan E.L. Meier P. Nonparametric estimation from incomplete observations.J Am Stat Assoc. 1958; 53: 457-481Crossref Scopus (47667) Google Scholar). Categorical predictors of survival were tested for significance using log–rank and Wilcoxon tests. Continuous covariates were tested for significance with a likelihood ratio test within a Cox proportional hazards model. The proportional hazard assumption for each covariate was checked by testing the significance of the interaction of time and the covariate. Models comprising multiple variables were obtained for each survival type using stepwise regression in the proportional hazard model. Age, sex, Zubrod performance status, stage (overall grouping, T stage, and N stage), GTV, tumor dose received, lung Veff, radiation duration, and chemotherapy were evaluated for their association with each survival endpoint in both univariate and multivariate analysis. Various interactions of interest were also considered in the Cox models. Results are presented as mean and 95% confidence interval (CI) unless otherwise specified.ResultsPatient and treatment characteristicsBetween 1992 and 2002, 122 patients with NSCLC were consented at the University of Michigan Hospital and its affiliates. Excluding patients with major protocol violations (e.g., changes of fraction size), those failing to meet normal tissue dose constraints, and those who received <63 Gy, a total of 106 patients were included in this analysis. The estimated median follow-up duration is 102.7 months (= 8.5 years). All patients except for 1 (who was lost to follow-up after 4.5 years) were followed up for at least 5 years or until the date of death.The characteristics of the patients and treatments are listed in Table 2. The patients ranged in age from 40–84 years, with a median age of 67 years. Fifty-nine percent had a performance status of 0, whereas 37% and 4% had a performance status of 1 and 2, respectively. Sixty-three percent had no weight loss, 13% had 5% weight loss. Thirty-five percent of patients had Stages I and II disease, 33% had Stage IIIA, 24% had Stage IIIB, and 8% had local recurrences after surgical resection for early-stage NSCLC. Among the 8 recurrent cases, 6 had T2N0, 1 had T3N0, and 1 had T2N2 disease at the time of recurrence. Thirty-six patients (34% of total, 44% of male patients) were from the Veterans Affairs Medical Center (VAMC) in Ann Arbor. Nineteen percent of patients (20/106) received neoadjuvant chemotherapy of cisplatin and vinorelbine. Forty-seven percent (50/106) received a radiation dose >74 Gy.Table 2Patient, tumor, and treatment characteristicsFactorNo.PercentageAge (years)Median67Range40–84SexMale8176%Female2524%Zubrod/ECOG performance status06359%13937%244%Weight lossNone6763%≤5%1413%>5%2524%Clinical stage (AJCC 1997 and 2003)I2927%II98%IIIA3533%IIIB2524%Recurrent88%Neoadjuvant chemotherapyYes2018%No8682%Radiation dose (Gy)63.098.5%63.71⁎These patients refused to complete prescribed dose of radiation.0.94%65.11716.0%67.21⁎These patients refused to complete prescribed dose of radiation.0.94%69.32826.4%73.51⁎These patients refused to complete prescribed dose of radiation.0.94%75.61514.2%84.01615.1%92.41110.4%102.976.6%Abbreviations: ECOG = European Cooperative Oncology Group; AJCC = American Joint Commission on Cancer. These patients refused to complete prescribed dose of radiation. Open table in a new tab Survival, disease control, and pattern of failureFigure 1 shows OS curves of all patients. The median OS was 19 months (95% CI, 13–21). The OS rates at 1, 2, 3, and 5 years were 61%, 37%, 23%, and 13%, respectively. The causes of death were 85% from lung cancer, 9% from other diseases, 3% from a second primary malignancy, and 3% from massive hemoptysis. Overall progression-free survival rates at 1, 2, 3, and 5 years were 39%, 23%, 19%, and 16%, respectively. Locoregional progression-free survivals at 1, 2, 3, and 5 years were 58%, 40%, 31%, and 29%, respectively. Median locoregional progression-free survival was 14 months (95% CI, 10–24). Median distant metastasis-free survival was 16 months (95% CI, 9–25). Distant metastasis-free survival rates at 1, 2, 3, and 5 years were 57%, 42%, 36%, and 32%, respectively. Among 89 patients who progressed from NSCLC, progression was noted initially at local site alone in 37%, at nodal region and local site (1 patient also at distant site) in 6%, at local and distant sites in 11%, and at distant sites only in 46% (Fig. 2). Ultimately, 46% of patients had both local and distant progression, 6% regional and local-distant progressions, 21% local progression only, and 27% distant progression only. Therefore, 73% of patients ultimately had a component of locoregional failure.Fig. 2Pattern of first failure or initial tumor progression.View Large Image Figure ViewerDownload (PPT)Factors associated with overall survivalUnivariate analysisTable 3 lists p values and hazard ratios (if p < 0.1) for factors evaluated. Age, sex, and performance status between 0 and 2 were not significantly correlated with OS rates. There was also no difference in OS between men from the VAMC and the non-VA patients (log–rank p value = 0.022). Overall stage grouping, T stage, and the use of neoadjuvant chemotherapy were not significantly associated with OS. For patients with Stage III disease, the median survival (5-year OS) was 12 months (11%) and 16 months (10%) for patients with and without chemotherapy (p = 0.893), respectively. The chemotherapy effect remained insignificant after taking the duration of chemotherapy into the survival calculation. There was also no significant difference in OS between patients with newly diagnosed vs. recurrent disease. There was a trend toward decreased survival with increasing GTV. The hazard ratio for an increase of 100 cc in GTV was 1.12 (p = 0.808). Patients who experienced weight loss had a significantly higher risk of death than those without weight loss (p = 0.0177, hazard ratio = 1.66 for any weight loss vs. none). Although there was no significant difference in OS for patients with N1, N2, or N3 disease, the presence of any nodal metastasis was a significant negative prognostic factor for OS (p = 0.0489, hazard ratio = 1.49 for N1–3 vs. N0). Tumor dose in Gy was tested as a continuous variable and found to be a significant (p = 0.0008) predictor of OS; the hazard from death decreased by a factor of 3% with each increase of 1 Gy.Table 3Factors evaluated for overall survival (univariate analysis)Factorsp ValueHazard ratioAge0.8782NSSex (female vs. male)0.9192NSWeight loss (any vs. none)0.01771.662Zubrod-EORTC performance score0.7758NSStage group (IA, IB, IIA, IIB, IIIA, IIIB, and recurrent)0.3583NST stage (T1, 2, 3, 4)0.4477NSN stage (1–3 vs. 0)⁎These factors remain significant in multivariate analysis.0.04891.489GTV (per 100 cc)0.08081.121Lung Veff0.00471.036Total dose (per Gy)⁎These factors remain significant in multivariate analysis.0.00080.965Chemotherapy (no vs. yes)0.7970NSAbbreviations: EORTC = European Organization for Research and Treatment of Cancer; GTV = gross tumor volume; NS = not significant; Veff = effective dose. These factors remain significant in multivariate analysis. Open table in a new tab Multivariate analysisCox multivariate analysis reveals that weight loss and radiation dose were significant independent factors for OS. There was no significant interaction between these two predictors (p = 0.099). The final predictive model contains total dose (p = 0.0006, hazard ratio = 0.97 per Gy) and weight loss (p = 0.0113, hazard ratio = 1.72 for any weight loss vs. none). The GTV was tested in the model of multiple variables with total dose and was found to be not significant (p = 0.95). Furthermore, there was no significant interaction between GTV and total dose (p = 0.68). The median survival (5-year survival) was 10 months (11%) for patients with weight loss and 21 months (14%) for patients without weight loss.Radiation dose and overall survivalFigure 3 shows the relationship between OS and radiation dose delivered when dose was evaluated as a categorical variable. The median survivals (5-year OS) were 12 months (4%), 27 months (22%), and 22 months (28%) for dose levels of 63–69 Gy (mean = 67 Gy), 74–84 Gy (mean = 80 Gy), and 92–103 Gy (mean = 97 Gy), respectively (p = 0.0002; Fig. 3A, Table 4). There were significantly more deaths from causes other than lung cancers in the high-dose groups (p < 0.0001; Table 5).Fig. 3Radiation dose, weight loss, and actuarial overall survival. Cox proportional multivariate analysis revealed that radiation dose and weight loss are independent significant factors for overall survival. (A) Survival curves for the three dose groups: Patients who received 74–84 Gy and 92–104 Gy had significantly better survival than those who received 63–69 Gy. (B) Survival curves of patients with or without any weight loss: The risk of death is 1.66 times higher for patients who had weight loss compared with patients who did not have weight loss.View Large Image Figure ViewerDownload (PPT)Table 4Association of radiation dose and overall sur