Abstract
Purpose
To investigate the correlation among DVH (lung dose–volume histogram) parameters, clinical factors, and grade ≥ 2 radiation pneumonitis (RP) in patients with locally advanced non-small-cell lung cancer (NSCLC) treated with three-dimensional conformal radiotherapy (3D-CRT), and the differences between patients treated with 3D-CRT alone or that combined with chemotherapy on RP.
Patients and methods
As much as 93 patients of stage III NSCLC were treated with 3D-CRT, among which 36 were treated with chemotherapy after 3D-CRT, 57 received 3D-CRT treatment alone. The radiation dose was 62.5–65 Gy (BED = 68–72.7 Gy).
Results
The morbidity of grade ≥ 2 RP was 49.5%, of which grade 2 and grade 3 were 33.3 and 16.1%, respectively. The morbidity of RP in those patients treated with chemotherapy after radiotherapy was evidently higher than that of patients treated with radiotherapy alone (61.1 vs. 42.1%). According to the single factor analysis, V5–V50 and MLD of both the ipsilateral and the whole lung were all related to the occurrence of RP; comparing grade 3 with grade 2 within the same group, except V45, V50, TV20, TV30, and TMLD, other parameters also had their statistical significance (P < 0.01); comparing the non-chemotherapy-treated group with the chemotherapy-treated group, TV30 and TV35 had their statistical significance. According to logistic regression analysis; the occurrence of RP was evidently associated with the comprehensive value of DVH parameters, chemotherapy, and gender. Chemotherapy has increased the risk of RP 7.6 times. The increase of each score in the comprehensive value of DVH parameters would increase the risk of RP 22.7 times.
Conclusion
The comprehensive values of DVH parameters, chemotherapy, and gender have independent effects on the occurrence of RP. Most of DVH parameters were associated with the occurrence of RP. The curve shape composed of multiple points in DVH parameters was more important than any single DVH parameter.
Keywords: DVH (lung dose–volume histogram), Three-dimensional conformal radiotherapy (3D-CRT), Non-small-cell lung cancer (NSCLC), Radiation pneumonitis (RP), Chemotherapy
Introduction
Radiation pneumonitis (RP) is one of the most significant dose-limiting factors in the radiation treatment of non-small-cell lung cancer (NSCLC). As three-dimensional conformal radiotherapy (3D-CRT) is widely applied, researchers are paying more attention to the relationship between DVH parameters and RP. It is widely accepted that V20 and V30 (the percentage of lung volume when dosage is higher than 20 Gy and 30 Gy) can be used to predict the occurrence of RP (Blom-Goldman et al. 2007; Schallenkamp et al. 2007; Tsujino et al. 2006; Piotrowski et al. 2005). However, there are still different opinions when considering the function of other DVH parameters (such as V5 (Marks et al. 1997; Wang et al. 2006), V10 (Schallenkamp et al. 2007)) in the evaluation of RP, and whether dosage or volume is of more importance (Marks et al. 1997; Willner et al. 2003). Besides DVH parameters, many other clinical factors (such as chemotherapy and gender) also contribute to the occurrence of RP, even though their influences still require further investigation. Moreover, there are great variations among patients, and there is no common standard to guide individual treatment plan at present. Thus, we retrospectively analyzed 93 cases of NSCLC patients treated with 3D-CRT. We setup a strict restriction to categorize all the patients in the group over factors like the clinical stage, the history of surgery and chemotherapy, the ways of radiation treatment, and the dosage used. By analyzing the clinical plan and RP occurrence between the group receiving radiotherapy alone and the group receiving chemotherapy after radiation treatment, several DVH parameters and the clinical factors relating to the occurrence of RP were determined. This study aims to find individualized plan to guide the treatment plan for the NSCLC patients.
Patients and methods
Patients
From May 2004 to March 2007, 93 patients of locally advanced NSCLC were treated with 3D-CRT at the Radiotherapy Department of the First Affiliated Hospital of China Medical University in Shenyang. Among which, 72 are male and 21 are female. The median age was 63 years (range 84–19 years old). All patients had no history of chronic obstructive pulmonary disease (COPD), and had not received previous surgery or radiation treatment. The clinical stages were IIIa and IIIb (65 were at the IIIa stage and 28 were at the IIIb stage.) The Karnofsky Performance Scale score before radiation was higher than or equal to 80 (KPS ≥ 80). As much as 57 patients received only radiotherapy. As much as 36 patients were treated with chemotherapy after radiotherapy for 2 to 3 cycles during the follow-up, 31 of them received NP regimen (Vinorelbine 25–30 mg/m2 in the first and the fifth day, Cisplatin 20–25 mg/m (Schallenkamp et al. 2007) in the first 3 days), and 5 of them received DP regimen (taxanes and Cisplatin). Chemotherapy began 1–2 weeks after radiotherapy. The duration of follow-up was defined from the time after radiation to 6 months later, with a follow-up rate of 100%.
Methods
All patients were treated with linear accelerator, 6MV-X rays of 3D-CRT, five fields were usually used in the treatment plan. The direction of the beam depended on various cases of the patients (e.g. size and location of lesions, etc.). The median dose per fraction (PTV’s accepted dose surrounded by 95% of isodose curve) was 2.3 Gy (range 2.0–2.5 Gy) and six fractions in one week, with a total dosage of 62.5–65 Gy (BED = 68–72.7 Gy). CT scans with slices 5-mm in thickness were obtained from the mandible to the lower edge of the liver before radiotherapy and when the dosage had reached 35–40 Gy. Whether the treatment plan should be adjusted depended on the shrinkage of the lesions.
DVH parameters
All patients were assigned to two groups, chemotherapy-treated group and non-chemotherapy-treated group. Each group had two subgroups, i.e. diseased-lung group (see the diseased-lung as the organ for evaluation) and whole-lung group (see the whole-lung as the organ for evaluation). Based on the DVH of the two treatment plans (at the beginning of the radiation, and when radiation dosage had reached 35–40 Gy), the DVH parameters of the two groups were worked out by ADAC treatment planning system (TPS), including V5, V10, V15, V25, V30, V35, V40, V45, V50, and MLD (mean lung dose).
Evaluation of RP
Early RP and late lung fibrosis are different stages of radiated lung injury, which are usually separated by 3 or 6 months after radiation (Weibai 2008). For this analysis, we mainly estimated the early RP, so 6 months after treatment was used as a cutoff for diagnosing RP, while late lung fibrosis is still kept in following observations. The diagnosis of RP is usually established by a history of radiotherapy, radiographic evidence, and clinical presentation. The typical radiologic manifestation is areas of ground-glass opacity or consolidation in the irradiated lung that conforms to the shape and size of the treatment portals (Choi et al. 2004). The symptoms of RP are dry cough, low-grade fever, chest pain, and shortness of breath. RP can also present as ipsilateral pleural effusion and consolidation of the lung (Wang et al. 2006). All patients were restaged through chest CT each month after radiation treatment until the follow-up ended. The grade of RP was evaluated by their treating radiation oncologists based on the patients’ clinical presentation and CT images. As most of the patients had typical radiographic evidence and clinical presentation, the bronchoscopy or biopsy was not adopted. Occasionally, the patients who were hard to be diagnosed with RP was consulted by other physicians to get rid of possibility of other disease (such as inflammation or heart diseases, etc.). RP often occurs 2–3 months after radiation. If patients at the acute stage were not treated on time, the grade of RP would increase. Conversely, if patients received proper treatment, the RP in some of them would be degraded and some still gradually led to the stage of lung fibrosis. Thus, the grade of RP was developing dynamically. To setup a unified standard, we had finally graded RP when the patients had the worst clinical features during follow-up. RP was graded according to the NCICTC3.0 (National Cancer Institute-Common Toxicity Criteria) (Wang et al. 2006).
Evaluation of clinical outcome
The evaluation of clinical short-term outcome was carried out according to clinical presentation and the CT images taken every month after radiation. The evaluation had been referred to the evaluation criteria of WHO as reference, namely, CR (complete remission) refers to the complete remission of the tumor, PR (partial remission) refers to the remission of more than 50% of the tumor, NR (non-remission) refers to the remission of <50% of the tumor or enlargement of <25% of the tumor, and PD (progressive disease) refers to the enlargement of more than 25% of the tumor or development of new lesions.
Statistical analysis
SPSS 12.0 statistical software package was used to analyze the data. The DVH parameters in different grades of RP were analyzed by a two-way analysis of variance (ANOVA) followed by a protected least significant difference (LSD) test (two-tailed) (α = 0.05). The probit model of probability density function was mainly used to calculate the probability of RP. Principal component analysis was used to fit DVH parameters into aggregative indicator which was applied to the analysis of RP occurrence. Logistic regression model was mainly used to carry out the multivariate analysis of impact factors of RP. The significant test level was P < 0.05.
Results
Out of the 93 patients, 46 had grade ≥ 2 RP (49.5%), among which 31 were grade 2 RP (33.3%) and 15 were grade 3 RP (16.1%). The RP morbidity of chemotherapy-treated group was 61.1% (22/36), 15 patients suffered from grade 2 RP (41.7%), 7 suffered from grade 3 RP (19.4%). The RP morbidity of non-chemotherapy-treated group was 42.1% (24/57), 16 patients suffered from grade 2 RP (28.1%), 8 suffered from grade 3 RP (14.0%). Most of the RP frequently occurred 2–3 months after radiotherapy, with a median time of 2.7 months.
Evaluation of recent effect
There were 16 cases of CR patients (17.2%), 60 PR patients (64.5%), and 17 NR patients (18.3%). The total effective rate (CR + PR) was 81.7%.
Tables 1 and 2 show DVH parameters (V5–V50, MLD) of patients who suffered from different grade of RP in non-chemotherapy-treated group and chemotherapy-treated group (D = diseased-lung and T = whole-lung). A multiple comparison among grade 0 + 1, 2, and 3 within the same group and a comparison between the two group was carried out by analysis of a two-way analysis of variance (ANOVA) followed by a protected least significant difference (LSD) test. Comparing grade 2 and 3 with grade 1, respectively, within the non-chemotherapy-treated group and the chemotherapy-treated group, the parameters all had their statistical significance (P < 0.01); comparing grade 3 with grade 2, except V45, V50, TV20, TV30, and TMLD, other parameters also had their statistical significance (P < 0.01). Comparing the non-chemotherapy-treated group with the chemotherapy-treated group, only TV30 (P = 0.040) and TV35 (P = 0.037) subgroups had their statistical significance; comparing the same grade within the two subgroups further, all had their statistical significance, P < 0.05.
Table 1.
DVH parameters of different grade RP in non-chemotherapy-treated group and chemotherapy-treated group (diseased-lung subgroup)
| Group | Non-chemotherapy group | Chemotherapy group | F a | P | F b | P | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| 0 + 1 | 2 | 3 | 0 + 1 | 2 | 3 | |||||
| DV5 | 0.68 ± 0.11 | 0.78 ± 0.12* | 0.83 ± 0.13*, ** | 0.66 ± 0.09 | 0.74 ± 0.07* | 0.87 ± 0.07*, ** | 0.185 | 0.668 | 19.940 | <0.001 |
| DV10 | 0.62 ± 0.10 | 0.72 ± 0.13* | 0.79 ± 0.14*, ** | 0.60 ± 0.09 | 0.68 ± 0.07* | 0.82 ± 0.08*, ** | 0.356 | 0.552 | 21.016 | <0.001 |
| DV15 | 0.57 ± 0.11 | 0.68 ± 0.13* | 0.76 ± 0.14*, ** | 0.54 ± 0.09 | 0.63 ± 0.06* | 0.77 ± 0.07*, ** | 0.881 | 0.350 | 22.191 | <0.001 |
| DV20 | 0.52 ± 0.11 | 0.63 ± 0.14* | 0.71 ± 0.14*, ** | 0.48 ± 0.09 | 0.58 ± 0.07* | 0.71 ± 0.08*, ** | 1.272 | 0.263 | 22.744 | <0.001 |
| DV25 | 0.44 ± 0.12 | 0.56 ± 0.14* | 0.65 ± 0.14*, ** | 0.41 ± 0.08 | 0.52 ± 0.08* | 0.64 ± 0.10*, ** | 1.387 | 0.242 | 22.722 | <0.001 |
| DV30 | 0.37 ± 0.11 | 0.48 ± 0.14* | 0.57 ± 0.15*, ** | 0.33 ± 0.08 | 0.44 ± 0.10* | 0.53 ± 0.11*, ** | 2.214 | 0.140 | 18.713 | <0.001 |
| DV35 | 0.30 ± 0.10 | 0.39 ± 0.13* | 0.49 ± 0.14*, ** | 0.27 ± 0.07 | 0.37 ± 0.10* | 0.42 ± 0.11*, ** | 2.736 | 0.102 | 15.692 | <0.001 |
| DV40 | 0.24 ± 0.09 | 0.32 ± 0.11* | 0.39 ± 0.11*, ** | 0.21 ± 0.06 | 0.31 ± 0.09* | 0.34 ± 0.10*, ** | 1.972 | 0.164 | 13.885 | <0.001 |
| DV45 | 0.20 ± 0.08 | 0.26 ± 0.10* | 0.33 ± 0.10* | 0.17 ± 0.05 | 0.26 ± 0.07* | 0.27 ± 0.09* | 1.791 | 0.184 | 13.180 | <0.001 |
| DV50 | 0.16 ± 0.07 | 0.21 ± 0.09* | 0.26 ± 0.09* | 0.14 ± 0.05 | 0.21 ± 0.07* | 0.21 ± 0.10* | 1.906 | 0.171 | 9.318 | <0.001 |
| DMLD (cGy) | 1,981.40 ± 560.32 | 2,731.36 ± 810.88* | 3,267.92 ± 736.22*, ** | 2,063.53 ± 508.32 | 2,611.56 ± 478.67* | 3,289.80 ± 553.23*, ** | 0.001 | 0.977 | 14.456 | <0.001 |
D diseased-lung
* Grades 2 and 3 compared with grade 0 + 1 in the same group, P < 0.01
** Grade 3 compared with grade 2 in the same group, P < 0.01
aF compared with non-chemotherapy-treated group and chemotherapy-treated group
bF compared with grade 0 + 1, 2, and 3 of RP in the same group
Table 2.
DVH parameters of different grade RP in non-chemotherapy-treated group and chemotherapy-treated group (whole-lung subgroup)
| Group | Non-chemotherapy group | Chemotherapy group | F a | P | F b | P | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| 0 + 1 | 2 | 3 | 0 + 1 | 2 | 3 | |||||
| TV5 | 0.62 ± 0.12 | 0.75 ± 0.12* | 0.81 ± 0.12*, ** | 0.64 ± 0.11 | 0.71 ± 0.09* | 0.81 ± 0.05*, ** | 0.183 | 0.670 | 16.992 | <0.001 |
| TV10 | 0.49 ± 0.14 | 0.62 ± 0.15* | 0.69 ± 0.15*, ** | 0.47 ± 0.13 | 0.55 ± 0.10* | 0.65 ± 0.13*, ** | 1.714 | 0.194 | 11.695 | <0.001 |
| TV15 | 0.39 ± 0.13 | 0.51 ± 0.14* | 0.59 ± 0.16*, ** | 0.38 ± 0.11 | 0.46 ± 0.11* | 0.56 ± 0.13*, ** | 0.691 | 0.408 | 11.841 | <0.001 |
| TV20 | 0.33 ± 0.12 | 0.43 ± 0.13* | 0.50 ± 0.17* | 0.32 ± 0.09 | 0.39 ± 0.11* | 0.48 ± 0.14* | 0.773 | 0.382 | 10.624 | <0.001 |
| TV25 | 0.26 ± 0.10 | 0.36 ± 0.12* | 0.45 ± 0.17*, ** | 0.25 ± 0.06 | 0.32 ± 0.10* | 0.38 ± 0.12*, ** | 2.628 | 0.109 | 12.348 | <0.001 |
| TV30 | 0.21 ± 0.08 | 0.29 ± 0.10* | 0.37 ± 0.16* | 0.19 ± 0.04*** | 0.25 ± 0.09*, *** | 0.28 ± 0.11*, *** | 4.361 | 0.040 | 11.301 | <0.001 |
| TV35 | 0.16 ± 0.07 | 0.22 ± 0.08* | 0.30 ± 0.14*, ** | 0.15 ± 0.03*** | 0.20 ± 0.08*, *** | 0.22 ± 0.09*, **, *** | 4.501 | 0.037 | 9.895 | <0.001 |
| TV40 | 0.13 ± 0.06 | 0.17 ± 0.06* | 0.23 ± 0.09*, ** | 0.11 ± 0.03 | 0.15 ± 0.06* | 0.17 ± 0.07*,*** | 3.643 | 0.060 | 10.169 | <0.001 |
| TV45 | 0.10 ± 0.05 | 0.14 ± 0.05* | 0.18 ± 0.07* | 0.09 ± 0.02 | 0.17 ± 0.16* | 0.14 ± 0.07* | 0.123 | 0.726 | 5.340 | 0.006 |
| TV50 | 0.08 ± 0.04 | 0.11 ± 0.05* | 0.14 ± 0.06* | 0.07 ± 0.03 | 0.10 ± 0.04* | 0.11 ± 0.06* | 3.801 | 0.054 | 9.554 | <0.001 |
| TMLD (cGy) | 1326.13 ± 425.82 | 1908.55 ± 571.76* | 2340.46 ± 851.89* | 1464.96 ± 383.99 | 1854.26 ± 416.67* | 2143.58 ± 594.49* | 0.059 | 0.809 | 10.035 | <0.001 |
T whole lung
aF compared with non-chemotherapy-treated group and chemotherapy-treated group
bF compared with grade 0 + 1, 2, and 3 of RP in the same group
* Grades 2 and 3 compared with grade 0 + 1 in the same group, P < 0.01
** Grade 3 compared with grade 2 in the same group, P < 0.01
*** Compared with the same grade of RP in non-chemotherapy-treated group, P < 0.01
Tables 3 and 4 used the probit model of probability density function to calculate the DVH parameters (V5–V50, MLD) of diseased-lung and whole-lung in non-chemotherapy-treated group and chemotherapy-treated group when the probability of grade ≥ 2 RP was 5–50%, respectively. As each DVH parameter increased, the probability of RP increased progressively. The DVH parameters of each chemotherapy-treated subgroup were generally lower than those of the non-chemotherapy-treated group, especially in whole-lung subgroup.
Table 3.
The mean of DVH parameters in diseased-lung (D) group when probability of Grade ≥ 2 RP is 5–50%
| Group | Probability of RP (*) | DV5 | DV10 | DV15 | DV20 | DV25 | DV30 | DV35 | DV40 | DV45 | DV50 | DMLD |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Non-chemotherapy-treated group | 5 | 0.48 | 0.41 | 0.36 | 0.30 | 0.25 | 0.15 | 0.08 | 0.03 | 0.00 | 0.00 | 916 |
| 10 | 0.54 | 0.48 | 0.43 | 0.37 | 0.31 | 0.22 | 0.15 | 0.10 | 0.06 | 0.03 | 1,224 | |
| 20 | 0.62 | 0.56 | 0.51 | 0.46 | 0.39 | 0.31 | 0.24 | 0.17 | 0.13 | 0.09 | 1,597 | |
| 30 | 0.68 | 0.62 | 0.57 | 0.52 | 0.45 | 0.37 | 0.30 | 0.23 | 0.18 | 0.14 | 1,867 | |
| 40 | 0.73 | 0.67 | 0.62 | 0.57 | 0.50 | 0.43 | 0.35 | 0.28 | 0.23 | 0.18 | 2,097 | |
| 50 | 0.77 | 0.72 | 0.67 | 0.62 | 0.55 | 0.48 | 0.40 | 0.32 | 0.27 | 0.22 | 2,312 | |
| Chemotherapy-treated group | 5 | 0.51 | 0.44 | 0.40 | 0.37 | 0.28 | 0.18 | 0.10 | 0.06 | 0.05 | 0.00 | 1,104 |
| 10 | 0.55 | 0.48 | 0.44 | 0.40 | 0.32 | 0.22 | 0.14 | 0.10 | 0.08 | 0.00 | 1,340 | |
| 20 | 0.60 | 0.53 | 0.49 | 0.44 | 0.36 | 0.27 | 0.10 | 0.14 | 0.12 | 0.04 | 1,626 | |
| 30 | 0.63 | 0.57 | 0.52 | 0.47 | 0.39 | 0.30 | 0.22 | 0.17 | 0.14 | 0.07 | 1,832 | |
| 40 | 0.66 | 0.60 | 0.56 | 0.49 | 0.42 | 0.33 | 0.25 | 0.20 | 0.17 | 0.10 | 2,008 | |
| 50 | 0.68 | 0.62 | 0.57 | 0.52 | 0.44 | 0.36 | 0.28 | 0.22 | 0.19 | 0.13 | 2,173 |
Table 4.
The mean of DVH parameters in whole-lung (T) group when probability of Grade ≥ 2 RP is 5–50%
| Group | Probability of RP (%) | TV5 | TV10 | TV15 | TV20 | TV25 | TV30 | TV35 | TV40 | TV45 | TV50 | TMLD |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Non-chemotherapy-treated group | 5 | 0.46 | 0.20 | 0.11 | 0.05 | 0.05 | 0.02 | 0.00 | 0.00 | 0.00 | 0.00 | 482 |
| 10 | 0.52 | 0.29 | 0.20 | 0.14 | 0.11 | 0.08 | 0.05 | 0.03 | 0.02 | 0.01 | 727 | |
| 20 | 0.59 | 0.41 | 0.31 | 0.24 | 0.20 | 0.15 | 0.11 | 0.08 | 0.06 | 0.05 | 1,023 | |
| 30 | 0.65 | 0.49 | 0.38 | 0.32 | 0.26 | 0.20 | 0.15 | 0.12 | 0.09 | 0.07 | 1,237 | |
| 40 | 0.69 | 0.56 | 0.45 | 0.38 | 0.31 | 0.25 | 0.19 | 0.15 | 0.12 | 0.09 | 1,420 | |
| 50 | 0.73 | 0.62 | 0.51 | 0.44 | 0.36 | 0.29 | 0.23 | 0.18 | 0.15 | 0.11 | 1,590 | |
| Chemotherapy-treated group | 5 | 0.38 | 0.10 | 0.01 | 0.01 | 0.01 | 0.03 | 0.00 | 0.00 | 0.00 | 0.00 | 509 |
| 10 | 0.43 | 0.18 | 0.09 | 0.07 | 0.07 | 0.07 | 0.01 | 0.02 | 0.02 | 0.00 | 720 | |
| 20 | 0.50 | 0.28 | 0.18 | 0.15 | 0.13 | 0.11 | 0.06 | 0.05 | 0.04 | 0.01 | 975 | |
| 30 | 0.55 | 0.34 | 0.25 | 0.21 | 0.17 | 0.14 | 0.09 | 0.07 | 0.06 | 0.03 | 1,160 | |
| 40 | 0.60 | 0.40 | 0.31 | 0.25 | 0.20 | 0.16 | 0.11 | 0.09 | 0.08 | 0.05 | 1,317 | |
| 50 | 0.64 | 0.45 | 0.36 | 0.30 | 0.24 | 0.19 | 0.14 | 0.11 | 0.09 | 0.06 | 1,464 |
Due to the colinearity between variations like V5–V50 and MLD, we could not induct them into the regression model simultaneously when the multivariate analysis was carried out. Principal component analysis (PCA) is a statistical method to convert more related data to less unrelated data, less data conveys more information, which are the main functions of PCA. Therefore, we first use PCA to compress the data. Table 5 shows the contribution rate and the cumulated contribution rate of each principal component, three of the characteristic roots of principal components were higher than 1 (the characteristic roots were 14.981, 2.148, and 1.594, respectively). As the accumulation of total variation was 93.614%, we worked out three factors, and took the contribution rate of each factor as weight numbers to calculate the comprehensive value. Table 6 was the coefficient by which variables were multiplied to obtain the principal components as described in our results. The variables are correlated with the three principal components. Then we inducted variants like age, sex, chemotherapy, and comprehensive value of principal component into logistic regression model (the rate of being selected was 0.05, while the rate of being picked out was 0.10). Finally, sex, chemotherapy as well as comprehensive value were inducted into the model. The result has shown that the RP morbidity of female was lower (P = 0.033). Chemotherapy had increased the risk of RP 7.6 times, 95% CI (1.7–33.0). When the comprehensive value fit by the principal component had increased one score, the risk of RP would increase 22.7 times, 95% CI (5.4–95.3) (Table 7).
Table 5.
Total variance explain of principal component analysis (PCA)
| Component | Initial Eigen values | Extraction sums of squared loadings | ||||
|---|---|---|---|---|---|---|
| Total | Percentage of variance | Cumulative % | Total | Percentage of variance | Cumulative % | |
| 1st principle Component | 14.981 | 74.903 | 74.903 | 14.981 | 74.903 | 74.903 |
| 2nd principle Component | 2.148 | 10.740 | 85.643 | 2.148 | 10.740 | 85.643 |
| 3rd principle Component | 1.594 | 7.971 | 93.614 | 1.594 | 7.971 | 93.614 |
Contribution rate and cumulated contribution rate of each principal component, 1st principle component, 2nd principle component, and 3rd principle component were taken which enclose 93.614% of all the original data
Table 6.
Component score coefficient matrix used for principal component analysis (PCA)
| Variables | 1st Principle component | 2nd Principle component | 3rd Principle component |
|---|---|---|---|
| D5 | 0.054 | −0.178 | −0.228 |
| L5 | 0.058 | −0.194 | −0.072 |
| D10 | 0.056 | −0.165 | −0.227 |
| L10 | 0.057 | −0.200 | 0.121 |
| D15 | 0.060 | −0.118 | −0.205 |
| L15 | 0.057 | −0.191 | 0.163 |
| D20 | 0.062 | −0.071 | −0.185 |
| L20 | 0.058 | −0.150 | 0.223 |
| D25 | 0.063 | 0.019 | −0.160 |
| L25 | 0.060 | −0.073 | 0.246 |
| D30 | 0.062 | 0.092 | −0.144 |
| L30 | 0.061 | −0.003 | 0.231 |
| D35 | 0.061 | 0.149 | −0.106 |
| L35 | 0.061 | 0.051 | 0.223 |
| D40 | 0.059 | 0.191 | −0.113 |
| L40 | 0.060 | 0.099 | 0.200 |
| D45 | 0.057 | 0.210 | −0.124 |
| L45 | 0.034 | 0.163 | 0.150 |
| D50 | 0.052 | 0.254 | −0.109 |
| L50 | 0.057 | 0.175 | 0.166 |
The coefficient by which variables are multiplied to obtain the principal components as described in our results. The variables are correlated with the three principal components
Table 7.
Multivariate Logistic Regression Analysis of clinical parameters associated with RP
| Clinical parameters | RP | P value | OR (95% CI) | |
|---|---|---|---|---|
| + (n = 46) |
− (n = 47) |
|||
| Age (years) | 60.51 ± 9.30 | 61.87 ± 13.93 | 0.152 | 1.1 (1.0–1.4) |
| Sex (male) | 37 (80.4%) | 35 (74.5%) | 0.033 | 0.2 (0.0–0.9) |
| Chemotherapy | 23 (50%) | 13 (28.9%) | 0.007 | 7.6 (1.7–33.0) |
| Comprehensive value (score) | 0.507 ± 0.776 | −0.502 ± 0.641 | 0.000 | 22.7 (5.4–95.3) |
OR odds ratios were calculated using logistic regression analysis for clinical parametric data
Discussion
Radiotherapy is one of the major treatments for locally advanced NSCLC. However, the effects of conventional radiotherapy are not satisfactory in the past, with only 5–10% of 5-year survival rate; lower local control rate is one of the principal reasons. The clinical dosimetry studies indicate that an increase in dosage would raise local control rate and survival rate, but due to the limit of lung tissue tolerance dosage, it is impossible to use a dosage of more than 60 Gy in conventional radiotherapy. In recent years, the renovated techniques were widely used in radiotherapy, such as increasing the radiation dosage to the utmost and decreasing the period of radiation, the radiation injury could be controlled effectively. This gradually becomes true. As it was reported by Sim et al. (2001) 70 patients who suffered from locally advanced NSCLC had received only radiotherapy treatment, with a median dosage of 70.2 Gy, and the 2-year local control rate was 35.4%. Rengan et al. (2004) had observed 72 patients with GTV > 100 cm3 locally advanced NSCLC and found that the median survival time was 20 and 15 months (P < 0.05) with radiation doses of >64 Gy and <64 Gy. Zhao et al. (2007) and Zhu et al. (2008) held that the increase in radiation had the tendency of prolonging long-term survival of patients with locally advanced NSCLC. In this study, the patients at the III clinical stages and in good condition were chosen for analysis (KPS ≥ 80, without previous surgery and COPD). The radiation doses were 2.0–2.5 Gy each time and 6 times in one week, with a total dosage of 62.5–65 Gy (BED = 68–72.7 Gy). The purpose of this treatment was to shorten the treatment period and increase the tumor control rate. Although the morbidity of grade 2 and 3 RP was quite high (49.5%), it was probably due to higher per fraction in doses of radiation and total dosage. Until the end of follow-up (6 months after radiation), most of the patients were in stable condition (cough was alleviated, CT scans showed that the lung had focal fibrosis), and none of them died due to RP. The total effective rate (CR + PR) was 81.7%. Their long-term survival rate required further follow-up.
As to the effects of DVH parameters in RP prediction, it was generally acknowledged that V20 and V30 had played their roles in the past. When V20 < 25%, V30 < 20%, the treatment plan was generally accepted. In recent years, some researchers had proposed that parameters like V5 (Wang et al. 2006; Yamashita et al. 2007), V10 (Schallenkamp et al. 2007), V13 (Seppenwoolde et al. 2003) had effects in RP prediction. However, there were great individual differences among patients actually treated; the occurrence of RP was not determined by any of the single factors. Various basic lung function, previous surgery, previous chemotherapy, etc., had contributed to the patients’ tolerance to radiation. Especially in recent years, with the wide application of comprehensive therapy, an increasing attention had been paid to the influences of chemotherapy on RP. As it was found out by Onishi et al. (2003) in their study, that for patients of stage III NSCLC, while treated with radiotherapy and combined with taxotere in chemotherapy, the morbidity of grade ≥ 3 RP was 47%. In our results, NP regime was mainly used to patients in the chemotherapy-treated group. It was found out that the grade 2–3 RP in chemotherapy-treated group was distinctly higher than that of the non-chemotherapy-treated group in the study (61.1% vs. 42.1%). This raised the concern of how did chemotherapy affect RP. Biologically, after alveolar type cells and endothelial cells are radiated, the pro-inflammatory cytokines (IL-1, IL-6, TNFα, etc.) are released to induce macrophage release profibronic cytokines (TGFβ, PDGF, etc.) which further promote RP. But chemotherapeutics (such as Bleomycin, Paclitaxel, etc.) can also induce macrophage to release IL-1, TGFβ, etc. The collaboration of both could make the RP worse (Weibai 2008; Madani et al. 2007). Furthermore, Gopal (2005) considered the lung irradiation also leads to a reduction in lung diffusion capacity (DLCO), and the chemoradiation could result in a larger reduction in DLCO than radiation alone. In our results, we have compared grade 2 and 3 with grade 0 + 1 RP, respectively, within the non-chemotherapy-treated group and the chemotherapy-treated group, the DVH parameters all had their statistical significance (P < 0.01). The mean value of V5–V50 and MLD for grade 2 and 3 RP in chemotherapy-treated group was generally lower than that of the non-chemotherapy-treated group. The above results had provided evidence to support the notion that chemotherapy reduced the patients’ tolerance to radiation from the aspects of dose–volume, and the chemotherapy had an effect on the occurrence of RP independently. Our results provided the mean of V5–V50 and MLD for grade 2 and 3 RP in the two conditions of radiotherapy alone and that combined with chemotherapy, and were referentially important in making individualized treatment plans. Moreover, since grade 3 RP of patients was more vulnerable to the treatment, we have further compared grade 2 with grade 3 RP within the two group, respectively, except V45, V50, TV20, TV30, and TMLD, other parameters also had their statistical significance (P < 0.01). Thus, comparing the non-chemotherapy-treated group with the chemotherapy-treated group, only TV30 (P = 0.040) and TV35 (P = 0.037) had their statistical significance, but based on the result of multivariate logistic regression analysis, the group of patients combined with chemotherapy had increased the risk of RP occurrence 7.6 times, 95% CI (1.7–33.0).
Among DVH parameters, it was fiercely debated that which of them was most relevant to the occurrence of RP, a small volume of lung with a high dose of radiation or a large volume of lung with a low dose of radiation. Marks et al. (1997) held that V5 had the most relevance with the occurrence of grade 3 RP, and that irradiated volume of the normal lung tissues was more important than dose of radiation. Gopal (2003) and Yorke et al. (2002) held that a large volume of lung with a low dose of radiation would bring more harm to the lung functions than a small volume of lung with a high dose of radiation. Opposite to the above-mentioned opinions, Willner et al. (2003) held that a large volume of lung with a low dose of radiation (for instance, 10 Gy) would be more appropriate than a small volume of lung with a high dose of radiation (for instance, 40 Gy). As it was seen from the results of this paper, all the DVH parameters were evidently related to the occurrence of RP (P < 0.05). If any one of parameters was above a certain value, there might be risk of RP. In the process of multivariate analysis, some authors had directly inducted several DVH parameters into logistic regression model, and drew the conclusion that certain DVH parameters were the most relevant to the occurrence of RP. We had analyzed the colinearity among each DVH parameters, and found a single parameter had a limited effect on the RP prediction. Therefore, the analysis in this respect was terminated. Data compression was carried out through principal component analysis, and the contribution rate of the corresponding DVH factors was taken as the weight number to calculate the DVH comprehensive value. Then the DVH comprehensive value of principal component was inducted into logistic regression model. Our results showed that the comprehensive value of DVH parameters had the greatest contribution to the occurrence of RP. Each score added in the comprehensive value had increased the risk of RP 22.7 times. Lung is a parallel organ and so the functional subunits are connected in parallel. Although a large volume of lung with a low dose of radiation would harm the functions of several subunits, the much higher dose of radiation given to a small lung volume might enlarge the impairment progressively and finally lead to the impairment of the whole-lung function. Thus, we believe that the factors of dosage and volume are equally important. While in the process of RP evaluation, comparing the curve shape on the DVH with one or several points on the curve, the information provided by the former one would be more comprehensive. The occurrence of RP would not be determined by any fixed value of DVH parameter, different parameter values are corresponding to the different risks of RP. We had further observed the relationship between DVH parameters and the probability of RP occurrence in non-chemotherapy-treated group and the chemotherapy-treated group. Those data had provided quantitative information for the treatment plan. It helped us to predict the risk of RP occurrence accurately while making the treatment plan. It would help us in the selection and adjustment of treatment plan, or in the adoption of concrete prevention measures clinically as early as possible.
The effect of MLD in RP prognoses had been widely accepted (Oetzel et al. 1995; Hernando et al. 2001; Kim et al. 2005; Graham 1997). According to the study of Hernando et al. (2001), RP morbidity was 10% with an MLD < 10 Gy, and 27% with an MLD of 21–30 Gy. Graham et al. (1997) had found that RP morbidity was 8% with an MLD < 20 Gy, and 24% with an MLD > 20. In our results, the mean of TMLD for grade 2 RP in non-chemotherapy-treated group and chemotherapy-treated group was 19.1 ± 5.8 Gy and 18.5 ± 4.2 Gy, respectively. For grade 3 RP, it was 23.4 ± 8.5 Gy and 21.4 ± 5.9 Gy, respectively. The mean of TMLD for 20% of grade ≥ 2 RP in the two groups was 10.2 Gy and 9.8 Gy, respectively. The mean of MLD for grade 2 and 3 RP in the chemotherapy-treated group was generally lower than that in the non-chemotherapy-treated group. As to whether ipsilateral-lung MLD (DMLD) or whole-lung MLD (TMLD) was of more significance in RP evaluation, Oetzel et al. (1995) and Willner et al. (2003) held that DMLD was more important than TMLD. According to our results, DMLD and TMLD were both related to RP morbidity evidently, without showing their differences.
As to the correlation between gender and RP, Robnett et al. (2004) had analyzed 144 cases of patients with lung cancer, and found that the RP morbidity was higher in female than in male (15% vs. 4%, P = 0.01). Their results suggest that the lung volume of female was relatively smaller, and it would be more likely for RP to occur in the same radiation field. In addition, RP may be a kind of hypersensitivity, similar to autoimmune disease, which was more often in female. Our study has found that the risk of RP morbidity in female was evidently lower than that in male. The morbidity in female was only 19.5% of that in male. The interpretation of this discrepancy is difficult, the possible contributing factors, such as sample size for each gender, difference of living environment, personal smoking history, and genetic background could not be completely ruled out.
As it is shown in our study, chemotherapy and DVH parameters are important predictors of RP morbidity. V5–V50 and MLD are obviously related to RP morbidity. A large volume of lung with a low dose of radiation and a small volume of lung with a high dose of radiation both have the risk of RP morbidity. Dosage and lung volume are equally important to RP morbidity, which cannot be determined by a single DVH parameter. The curve shape composed of multiple points in DVH parameters can reflect the amount of radiation exposure of normal lung tissues in an all-round way, and can predict RP morbidity more accurately.
References
- Blom-Goldman U, Svane G, Wennberg B et al (2007) Quantitative assessment of lung density changes after 3-D radiotherapy for breast cancer. Acta Oncol 46(2):187–193 [DOI] [PubMed] [Google Scholar]
- Choi YW, Munden RF, Erasmus JJ et al (2004) Effects of radiation therapy on the lung: radiologic appearances and differential diagnosis. Radiographics 24:985–998 [DOI] [PubMed] [Google Scholar]
- Gopal R (2005) Pulmonary toxicity associated with the treatment of non-small cell lung cancer and the effects of cytoprotective strategies. Semin Oncol 32(2 Suppl 3):S55–S59 [DOI] [PubMed] [Google Scholar]
- Gopal R, Tucker SL, Komaki R et al (2003) The relationship between local dose and loss of function for irradiated lung. Int J Radiat Oncol Biol Phys 56:106–113 [DOI] [PubMed] [Google Scholar]
- Graham MV (1997) Predicting radiation response. Int J Radiat Oncol Biol Phys 39:561–562 [DOI] [PubMed] [Google Scholar]
- Hernando ML, Marks LB, Bentel GC et al (2001) Radiation-induced pulmonary toxicity: a dose-volume histogram analysis in 201 patients with lung cancer. Int J Radiat Oncol Biol Phys 51:650–659 [DOI] [PubMed] [Google Scholar]
- Kim TH, Cho KH, Pyo HR et al (2005) Dose-volumetric parameters for predicting severe radiation pneumonitis after three-dimensional conformal radiation therapy for lung cancer. Radiology 235:208–215 [DOI] [PubMed] [Google Scholar]
- Madani I, De Ruyck K, Goeminne H, De Neve W, Thierens H, Van Meerbeeck J (2007) Predicting risk of radiation-induced lung injury. J Thorac Oncol 2(9):864–874 [DOI] [PubMed] [Google Scholar]
- Marks LB, Munley MT, Bentel GC et al (1997) Physical and biological predictors of changes in whole-lung function following thoracic irradiation. Int J Radiat Oncol Biol Phys 39:563–570 [DOI] [PubMed] [Google Scholar]
- Oetzel D, Schraube P, Hensley F et al (1995) Estimation of pneumonitis risk in three-dimensional treatment planning using dose-volume histogram analysis. Int J Radiat Oncol Biol Phys 33:455–460 [DOI] [PubMed] [Google Scholar]
- Onishi H, Kuriyama K, Yamaguchi M et al (2003) Concurrent two-dimensional radiotherapy and weekly docetaxel in the treatment of stage Ш non-small cell lung cancer: a good local response but no good survival due to radiation pneumonitis. Lung Cancer 40:79–84 [DOI] [PubMed] [Google Scholar]
- Piotrowski T, Matecka-Nowak M, Milecki P (2005) Prediction of radiation pneumonitis: dose-volume histogram analysis in 62 patients with non-small cell lung cancer after three-dimensional conformal radiotherapy. Neoplasma 52(1):56–62 [PubMed] [Google Scholar]
- Rengan R, Rosenzweig KE, Venkatraman E et al (2004) Improved local control with higher doses of radiation in large-volume stage III non-small-cell lung cancer. Int J Radiol Oncol Biol Phys 60:741–747 [DOI] [PubMed] [Google Scholar]
- Robnett TJ, Machtay M, Vines EF et al (2004) Factors predicting severe radiation pneumonitis in patients receiving definitive chemoradiation for lung cancer. Int J Radiat Oncol Biol Phys 48:89–94 [DOI] [PubMed] [Google Scholar]
- Schallenkamp JM, Miller RC, Brinkmann DH et al (2007) Incidence of radiation pneumonitis after thoracic irradiation: dose-volume correlates. Int J Radiat Oncol Biol Phys 67(2):410–416 [DOI] [PubMed] [Google Scholar]
- Seppenwoolde Y, Lebesque JV, De Jaeger K et al (2003) Comparing different NTCP models that predict the incidence of radiation pneumonitis. Int J Radiat Oncol Biol Phys 56:106–113 [DOI] [PubMed] [Google Scholar]
- Sim S, Rosenzweig KE, Schindelheim R et al (2001) Induction chemotherapy plus three-dimensional conformal radiation therapy in the definitive treatment of advanced non-small cell lung cancer. Int J Radiat Oncol Biol Phys 51:660–665 [DOI] [PubMed] [Google Scholar]
- Tsujino K, Hirota S, Kotani Y et al (2006) Radiation pneumonitis following concurrent accelerated hyperfractionated radiotherapy and chemotherapy for limited-stage small-cell lung cancer: dose-volume histogram analysis and comparison with conventional chemoradiation. Int J Radiat Oncol Biol Phys 64(4):1100–1105 [DOI] [PubMed] [Google Scholar]
- Wang S, Liao Z, Wei X et al (2006) Analysis of clinical and dosimetric factors associated with treatment-related pneumonitis (TRP) in patients with non-small-cell lung cancer (NSCLC) treated with concurrent chemotherapy and three-dimensional conformal radiotherapy (3D-CRT). Int J Radiat Oncol Biol Phys 66(5):1399–1407 [DOI] [PubMed] [Google Scholar]
- Weibai Y (2008) Radiation oncology, vol 4. Peking Union Medical College Publishing Company, Beijing, pp 645–652. ISBN978-7-81072-948-2
- Willner J, Jost A, Baier K et al (2003) 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 179:548–556 [DOI] [PubMed] [Google Scholar]
- Yamashita H, Nakagawa K, Nakamura N (2007) Exceptionally high incidence of symptomatic grade 2–5 radiation pneumonitis after stereotactic radiation therapy for lung tumors. Radiat Oncol 2:21–32 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yorke ED, Jackson A, Rosenzweig KE et al (2002) Dose-volume factors contributing to the incidence of radiation pneumonitis in non-small-cell lung cancer patients treated with three-dimensional conformal radiation therapy. Int J Radiat Oncol Biol Phys 54:329–339 [DOI] [PubMed] [Google Scholar]
- Zhao LJ, West B, Hayman JA et al (2007) High radiation dose may reduce the negative effect of large gross tumor volume in patients with early stage non-small cell lung cancer]. Int J Radiat Oncol Biol Phys 68:103–110 [DOI] [PubMed] [Google Scholar]
- Zhu XZ, Wang LH, Zhao LJ et al (2008) Impact of gross tumor volume and radiation therapy dose on survival in stage III non-small-cell lung cancer treated with three-dimensional radiation therapy. Chin J Radiat Oncol 17(1):26–29 [Google Scholar]