Symptomatic Radiation-induced Lung Injury In Non-Small Cell Lung Cancer:Clinical Characteristics,Management, Outcome, And Biological Predictive Factors

Part I. Characteristics, management, outcomes of symptomatic radiation-induced lung injury in patients with non-small cell lung cancer and tolerance to subsequent radiotherapy or chemotherapy in these casesObjective:To analyze the clinical characteristics, management and outcomes of symptomatic radiation-induced lung injury (SRILI) in patients with non-small cell lung cancer (NSCLC), and the tolerance to subsequent radiotherapy or chemotherapy in these patients after occurrence of SRILI.Methods:NSCLC patients treated with radiotherapy from January2000to May2007in cancer hospital of CAMS were retrospectively reviewed and those who developed SRILI during therapy or follow-up and received related treatment in our hospital were enrolled into analysis. The symptoms, clinical signs, imaging findings, and hematological changes of the patients were observed, the management, outcomes and tolerance to subsequent radiotherapy or chemotherapy after diagnosis of SRILI were analyzed. The diagnosis and classification of RILI was evaluated according to common terminology criteria for adverse events version3.0criteria by2radiation oncologists and1radiologist independently. In the part of analysis of management and outcome, patients should have detailed record of SRILI related management information, and a same patient might be assigned into more than one grade of SRILI classification for analysis as he or she experienced more than one phase of RILI during the disease course.Results:A total of81SRILI patients were included in the analysis of clinical characteristics, with35(43.2%) cases of grade2(G2),42of G3,0of G4, and4of G5of SRILI respectively. The median time from beginning of radiation to onset of symptoms was8.3weeks (ranging from1.3to32.3weeks). Among the SRILI symptoms, Cough and dyspnea were observed in95%and69%patients respectively, and fever was noted in48%patients with the median highest temperature of38.3℃. Signs were recorded in a low percentage relatively, and rough breath sound being noted in about50%patients was the most common sign. The imaging findings included patterns of consolidation with ventilation of the bronchial, patch haziness, and linear fibrosis, most of which were found in the radiation field. The hematological changes included slightly elevated proportion of neutrophils with median percentage of77.4%but total number of leukocyte was in the normal range.A total of80patients were included into the analysis of management and outcomes, including49with G2,46with G3,1with G4, and4with G5SRILI events respectively. Eighty-seven percent of patients with G3SRILI and three patients with G2SRILI accompanied with fever were treated with corticosteroid. Some of them received prophylactic administration of first-line antibiotics. The initial dose of steroid ranged from20to66.7mg prednisone equivalent daily; the median effective dose was33.3mg daily. The dose of steroid decreased by slow taper and the median duration of treatment was33days. Of the34patients who underwent steroid treatment and received evaluation of response to steroid treatment,65%were steroid-sensitive, and32%were steroid-dependent. Thirty one percent of G2SRILI progressed to G3and9%of G3SRILI upgraded to G5.In thirty-seven patients, SRILI occurred during the radiation with22patients of G2and15patients of G3SRILI, and all of them continued receiving radiotherapy to the total dose of50-64Gy. Eighteen patients of G2and19patients of G3SRILI received subsequent chemotherapy, all the patients that kept concurrent chemotherapy going on got SRILI symptoms aggravated (2patients with G2and3patients with G3SRILI), only2/16and3/16patients that received post-radiation chemotherapy have got SRILI recur or symptoms aggravated (p=0.016for G2, p=0.002for G3). All the patients who were cured with no SRILI symptoms when they received subsequent chemotherapy haven’t got SRILI recur (6patients with G2and11patients with G3SRILI),5/12and5/8patients got SRILI symptoms aggravated that they still had SRILI symptoms when they received the subsequent chemotherapy (p=0.026for G2, p=0.001for G3).Conclusions:The symptoms of SRILI occurred in8.3weeks after the beginning of radiation in median. The clinical characteristics of SRILI include cough, dyspnea, fever, coarse breath sound, imaging findings of lung consolidation, patch haziness and linear fibrosis, which are mostly found in the radiation field, and slightly elevated percentage of neutrophils. Corticosteroid is the primary therapy for G3SRILI and may also be considered for G2SRILI accompanied with fever. Steroid is effective in the treatment of SRILI, and SRILI related mortality is low with timely steroid treatment. Patients who developed SRILI during the radiation could tolerant continuing radiotherapy alone to the total dose of50Gy at least. If the SRILI were cured with no symptoms, subsequent adjuvant or palliative chemotherapy would not make the SRILI recur. Concurrent chemotherapy would aggravate the SRILI symptoms. Part Ⅱ. Elevation of plasma TGF-β1during radiotherapy increase risk of symptomatic radiation-induced lung injury in non-small cell lung cancer:a prospective validation studyObjective:To validate the conclusion that the elevation of plasma transforming growth factor-β1(TGF-β1) during radiotherapy (RT) increase risk of symptomatic radiation-induced lung injury (SRILI) in non-small cell lung cancer, which was reported by our group before.Methods:NSCLC patients who received conventional fractionated radiotherapy in our mono-institute were prospectively enrolled. Their platelet-poor plasma was collected pre-RT and at the about4weeks (3-5weeks) during RT. TGF-β1was measured using an enzyme-linked immunosorbent assay. The primary endpoint was SRILI (≥Grade2RILI) which was evaluated by National cancer institute commom toxicity criteria version3.0. The time to onset of SRILI was calculated from radiation start. Cox-regression were used for risk factor evaluation.Results:A total of67patients of stage I-IV NSCLC who received radiotherapy from May2007to March2012in cancer hospital of CAMS were included. Fourteen patients received post-operation radiotherapy, and56patients received chemotherapy, including35cases of concurrent chemoradiotherapy. The median radiation dose was60Gy (range from31to71.5Gy), and the median mean lung dose (MLD) was15.4Gy. Eighteen patients (26.9%) experienced SRILI. The incidence of SRILI was52.9%in patients with a TGF-β1ratio>1vs.18.0%in patients with a TGF-β1ratio<1(p=0.007). Univariate analysis showed pre-RT pulmonary function FEV1%, ipsilateral lung (IL) V15, IL V13, and plasma TGF-β1ratio were significantly correlated with the development of SRILI (p=0.015,0.014,0.014,0.007). Multivariate analysis showed plasma TGF-β1ratio>1and FEV1%<63.2%were significantly correlated with the development of SRILI (hazard ratio=3.77,3.30; p=0.017,0.036).Conclusions:The conclusion that the elevation of plasma TGF-β1during radiotherapy increase risk of SRILI in NSCLC was validated. The clinical application of plasma TGF-β1ratio>1as predictor of SRILI may be considered. Part III. Pretreatment concentration of plasma vascular endothelial growth factor with prognosis of locally advanced non-small cell lung cancerObjective:To investigate whether there is correlation between pretreatment concentration of plasma vascular endothelial growth factor (VEGF) and the prognosis of locally advanced non-small cell lung cancer (LD-NSCLC).Methods:Untreated locally advanced NSCLC patients who received radiotherapy (RT) or concurrent chemoradiotherapy with curative intent were included. Their platelet-poor plasma was collected pre-RT. VEGF was measured using an enzyme-linked immunosorbent assay (ELISA). After treatment, patients receive regular follow-up to death. The primary endpoint was overall survival (OS). Cox-regression was used for risk factor evaluation.Results:A total of49patients of LD-NSCLC who received radiotherapy from May2007to December2009in cancer hospital of CAMS were included. The median radiation dose was60Gy (range from36to70Gy). Forty-one patients received chemotherapy, including36cases of concurrent chemoradiotherapy. The median follow-up time for the patients alive was46.8months. The median survival time was23.5months. The1-y,2-y,3-y overall survival was73.5%,46.9%and38.8%. The ELISA assay was conducted in4batches, and inter-batch difference was significant in comparison between several two groups, and the overall inter-batch difference trend to be significant, x2=6.908, p=0.075. The detect concentration of VEGF may be affected by time of sample stored, r=-0.246, p=0.088. The median pretreatment concentration of plasma VEGF was55.7pg/ml (range from17.4-733.1pg/ml). The median survival time was26.6months for patients presenting with a VEGF plasma level>55.7pg/ml, and20.2months for patients that VEGF level<55.7pg/ml, but the difference was not significant, p=0.733. In univariate analysis, radiotherapy alone (compared with chemoradiotherapy), total radiation dose less than60Gy, N3stage, pretreatment LDH level>186U/L, pretreatment CEA level>5ng/ml were significantly correlated with the overall survival (p=0.001-0.038). Multivariate analysis showed radiotherapy alone (compared with chemoradiotherapy), total radiation dose less than60Gy, N3stage, and pretreatment LDH level>186U/L were independent prognostic factors of LD-NSCLC (hazard ratio=5.13,3.89,3.20,3.34; p=0.002,0.008,0.015,0.009). Conclusions:This study shows that there’s no significant correlation between pretreatment concentration of plasma VEGF and the prognosis of LD-NSCLC. Because of the limitations of the study, we may not conclude whether plasma VEGF concentration is prognostic factor of LD-NSCLC. Further research is required that based on a well-designed study that overcome the confounding factors such as long period of sample store and inter-batch variation.