| Japanese Journal of Clinical Oncology | Pages |
Risk Factors for Severe Radiation Pneumonitis in Lung Cancer
Introduction
Patients and Methods
Results
Discussion
Acknowledgment
References
Risk Factors for Severe Radiation Pneumonitis in Lung Cancer
Methods: 111 cases of primary lung cancer, treated with radiotherapy or chemoradiotherapy, were retrospectively analyzed.
Results: Severe radiation pneumonitis occurred in 17 cases (15.3%). The ratio of interstitial change in lungs before radiotherapy and radiotherapy to the contralateral mediastinum with >40 Gy in the radiation pneumonitis group (RP group) was significantly higher than in patients without radiation pneumonitis (control group) (47.1% vs 5.3%; P < 0.001 and 58.8% vs 27.7%; P = 0.037, respectively). Using logistic regression analysis, interstitial changes before radiotherapy and radiotherapy to the contralateral mediastinum of >40 Gy were significant risk factors associated with severe radiation pneumonitis.
Conclusions: These data suggest that pre-existing interstitial changes detected by chest radiography or computed tomography and radiotherapy to the contralateral mediastinum (>40 Gy) may predict the development of severe radiation pneumonitis.
INTRODUCTION
Radiotherapy plays an important role in the treatment of localized primary lung cancer (1-3). However, therapy is limited by pulmonary complications. A well defined area of pulmonary fibrosis confined to the field of radiation often appears on the chest radiograph 6-12 months after radiotherapy (4-6). In 5-15% of patients, an early acute pneumonitis, which is characterized by non-productive cough, fever and dyspnea on exertion, develops 2-12 weeks after radiotherapy (4,7-9). It is thought that the early onset of symptoms is indicative of a more serious and protracted clinical course (10). In some cases, fibrosis spreads outside the radiation field, even to the contralateral lung, progressing to acute, often fatal, respiratory failure (11-14). Therefore, it would be very important if one were able to anticipate the occurrence of severe radiation pneumonitis (RP) spreading beyond the radiation field. The present study was undertaken to identify risk factors associated with RP following radiotherapy or chemoradiotherapy for primary lung cancer.
PATIENTS AND METHODS
The subjects were 111 patients with primary lung cancer who received radiotherapy or combined modality therapy (chemotherapy and radiotherapy) between July 1994 and August 1997. All patients underwent a history and physical examination, radiographic evaluation for staging purposes [chest radiograph, bone scan and computed tomography (CT) of the chest, upper abdomen and whole brain] according to the Japanese Lung Cancer Society (15) and had histological confirmation of primary lung cancer. Performance status (PS) was rated on the Eastern Cooperative Oncology Group (ECOG) scale.
Radiation was administered with 10 MV X-rays from a linear accelerator. The treatment schedule was primarily once daily with standard fractionation and the fraction size was 2 Gy in 104 patients. Seven patients had hyperfractionated radiotherapy, consisting of 1.5 Gy/fraction, 2 fractions/day, 5 days/week. The initial radiation volume included the primary tumor, ipsilateral hilar nodes and ipsilateral or bilateral mediastinal nodes by antero-posterior parallel opposed fields with a 1-1.5 cm margin around the primary tumor and grossly involved nodes.
The diagnosis of radiation pneumonitis was established retrospectively by clinical symptoms of non-productive cough, fever, dyspnea and characteristic chest radiograph and CT findings: an infiltrate that spread beyond the radiation field. If possible, sputum cultures were obtained to rule out secondary infections.
We assessed the prognostic significance of age, gender, PS, histology, clinical stage, Brinkman Index (BI), oxygenation (PaO2), lactate dehydrogenase (LDH), C reactive protein (CRP), underlying lung disease, primary site, pulmonary function tests and total radiation dose. In this study, `interstitial change' means radiographic changes, detected by plain chest films and/or CT scans, regardless of clinical symptoms. We also noted the radiation dose at the primary site, ipsilateral mediastinum (ips. med.) and contralateral mediastinum (cont. med.), as well as the radiation field area, chemoradiation schedule (concurrent vs sequential or alternating combined modality therapy), chemotherapy regimen, corticosteroid therapy and clinical symptoms and signs. Comparisons of LDH, CRP and PaO2 prior to radiotherapy and at the onset of RP were performed using the Wilcoxon signed-rank test. Differences in rates of various factors were tested by Fisher's exact probability test. Differences in total radiation dose, radiation doses at specific sites, radiation field area, laboratory data before radiotherapy, BI and pulmonary function were examined by the Mann-Whitney U-test. Multivariate analysis was based on multiple stepwise regression and included BI, laboratory data before radiotherapy, baseline interstitial changes of the lung, DLco/VA (permeability coefficient), radiation field area, exposure of the contralateral mediastinum to radiation >40 Gy and chemotherapy. Significance was accepted if P < 0.05.
RESULTS
Seventeen patients developed radiation pneumonitis (RP) which had spread beyond the radiation field; 94 patients without RP were the control group. No patients in the RP group had infections preceding this diagnosis. In this study, no patient had been previously irradiated. At the time of their presentation with pneumonitis, there were no immunosuppressive conditions, such as leukopenia or neutropenia, that would implicate opportunistic infections. Patient demographics included in this study are listed in Table 1. No obvious relationship was found between chemotherapy and radiation pneumonitis. By contrast, the frequency of underlying lung disease before radiotherapy was much higher in the RP group (47.1%) than controls (5.3%; P < 0.001). In six of eight patients in the RP group, baseline interstitial change of the lung was detected by both plain chest films and CT scans and in the other two patients, it was detected only by CT scans. The mean permeability coefficient (DLco/VA) in the RP group tended to be smaller than the control group (P = 0.071). CRP in the control group was significantly higher than the RP group (P = 0.049).
Table 1.
| Parameter | Pneumonitis | P value | |
| + | - | ||
| No. of patients | 17 | 94 | |
| Male/female | 16/1 | 83/11 | |
| Mean age (range) | 70.2 (57-85) | 69.0 (44-86) | |
| PS (0/1/2/3) | 1/12/4/0 | 10/56/19/9 | |
| Stage | 0.211 | ||
| I | 1 | 8 | |
| II | 1 | 14 | |
| IIIA | 10 | 37 | |
| IIIB | 5 | 35 | |
| Histology | 0.638 | ||
| Sq. | 9 | 59 | |
| Sm. | 5 | 17 | |
| Ad. | 2 | 14 | |
| La. | 1 | 4 | |
| Chemotherapy | 12 | 48 | 0.222 |
| Underlying lung disease | <0.001 | ||
| Interstitial change | 8 | 5 | |
| Emphysema | 6 | 31 | |
| Other | 0 | 7 | |
| None | 3 | 51 | |
| BI | 1160 ± 603 | 913 ± 569 | 0.112 |
| Laboratory data | |||
| PaO2 (Torr) | 71.4 ± 8.4 | 70.2 ± 8.2 | 0.559 |
| A-aDO2 (Torr) | 22.0 ± 14.6 | 19.0 ± 8.5 | 0.917 |
| LDH (IU/l) | 314.4 ± 65.8 | 315.7 ± 70.8 | 0.878 |
| CRP (mg/dl) | 0.9 ± 1.5 | 1.91 ± 3.0 | 0.049 |
| DLco/VA (ml/min.mmHg.l) | 4.6 ± 0.9 | 6.8 ± 5.1 | 0.071 |
Table 2.
| Parameter | Pneumonitis | P value | |
| + | - | ||
| Dose of radiation (Gy) | 53.8 ± 11.2 | 60.3 ± 6.7 | 0.006 |
| Dose according to radiation site (Gy) | |||
| Primary site and hilum | 53.8 ± 11.2 | 59.2 ± 6.4 | 0.98 |
| Ipsilateral meda | 53.8 ± 11.2 | 51.2 ± 17.8 | 0.89 |
| Contralateral med | 43.8 ± 8.7 | 34.6 ± 19.7 | 0.051 |
| No. of patientsb | 10 | 26 | 0.037 |
| Radiation field (cm2) | 135.6 ± 48.9 | 114.5 ± 43.6 | 0.137 |
| Timing of thoracic irradiation | 0.880 | ||
| Sequential | 10 | 39 | |
| Alternative | 0 | 1 | |
| Concurrentc | 2 | 8 | |
| Primary site | 0.388 | ||
| Upper lobe/sup. seg.d | 12 | 78 | |
| Middle/lower lobee | 5 | 16 | |
Table 3.
| Case No. | Primary focusa | Underlying lung diseaseb | RT field area (cm2) | RT dose (Gy)c Prim. | Ips. med. | Cont. med. | Chemotherapy regimend | Chemotherapy course No. | Timing of RTe |
| 1 | r-LL | None | 120 | 58 | 58 | 58 | P, E | 4 | Seq |
| 2 | r-UL | Emp | 102 | 22 | 22 | 22 | None | ||
| 3 | r-LL | Interst | 120 | 60 | 40 | 40 | P, V | 2 | Seq |
| 4 | r-LL | Emp | 116 | 44 | 44 | 44 | P, V | 2 | Seq |
| 5 | l-UL | Emp | 123.5 | 66 | 66 | 42 | None | ||
| 6 | l-UL | Interst | 93.5 | 68 | 68 | 40 | None | ||
| 7 | l-LL | Emp | 156 | 62 | 62 | 44 | CB, E | 4 | Seq |
| 8 | r-UL | Interst | 114 | 40 | 40 | 40 | P, CB, E | 3 | Seq |
| 9 | r-UL | Interst | 120 | 36 | 36 | 36 | P, V | 2 | Seq |
| 10 | r-LL | Interst | 288 | 54 | 54 | 54 | CAV/PE | 6 | Seq |
| 11 | r-LL | None | 182 | 62 | 62 | 40 | CPT-11 | 6 | Conc |
| 12 | l-LL | Emp | 75 | 66 | 66 | 46 | None | ||
| 13 | l-LL | Interst | 186 | 54 | 54 | 54 | CB, E | 4 | Seq |
| 14 | l-UL | Emp | 120 | 54 | 54 | 54 | P, CPT-11 | 1 | Seq |
| 15 | r-UL | intest | 144 | 60 | 60 | 42 | CPT-11 | 6 | Conc |
| 16 | r-UL | None | 120 | 56 | 56 | 42 | CB, E | 2 | Seq |
| 17 | r-UL | Interst | 162 | 50 | 50 | 10 | None |
Table 4.
| Beforea | Onsetb | P value | |
| PaO2 | 71.4 ± 8.4 | 60.0 ± 9.9 | 0.0052 |
| CRP | 0.86 ± 1.48 | 6.60 ± 3.34 | 0.0004 |
| LDH | 314.4 ± 65.8 | 419.7 ± 123.3 | 0.0010 |
Table 2 compares radiation-related factors in the two groups. Pneumonitis occured in seven patients before completion of radiotherapy and radiotherapy was therefore discontinued, which is why the radiation dosage of the RP group was smaller than in the control group.
The radiation dose to the contralateral mediastinum in the RP group tended to be larger than in controls (P = 0.051). The number of patients with exposure of the contralateral mediastinum to >40 Gy in the RP group was larger than in controls (P = 0.037).
Table 3 lists individual factors in the RP group.
CRP, LDH and PaO2 in the RP group were compared before radiotherapy and at the onset of radiation pneumonitis (Table 4). LDH and CRP were significantly increased (P = 0.001 and 0.0004, respectively) and the level of PaO2 was significantly decreased (P = 0.0052).
Table 5 lists the results of the multivariate analysis. Interstitial change at the baseline stands out as the dominant prognostic factor (P = 0.013). Radiotherapy to the contralateral mediastinum of >40 Gy was a secondary factor (P = 0.028).
Table 5.
| Independent | Coefficient | Standard error | P value |
| Interstitial change | 2.417 | 0.970 | 0.013 |
| Radiotherapy to cont. med.a | 1.867 | 0.849 | 0.028 |
In the 10 RP patients who completed their radiation course, the mean to onset of radiation pneumonitis was 14.9 days following completion. Fourteen patients in the RP group died, 11 due to respiratory insufficiency and cor pulmonale caused by radiation pneumonitis; the mean time from onset of radiation pneumonitis to death was 64.0 days (range 26-145 days). Three RP patients died of cancer progression. Three RP patients are alive with cancer progression.
Fever and gradual elevation of the CRP without evidence of infection were present in 13 RP patients 1-2 weeks before the onset of pneumonitis. When radiation pneumonitis became evident, fever, non-productive cough and dyspnea were noted in 15, 8 and 15 patients, respectively. Small amounts of sputum were present in three patients, but cultures were negative.
All patients with RP received corticosteroids and antibiotics as soon as pneumonitis presented (prednisone 10-60 mg/day or methylprednisolone 24-40 mg/day). Fifteen patients initially responded; two, however, progressed to fatal acute respiratory distress despite corticosteroid administration. Recurrence of pneumonitis was recognized in 12 patients despite corticosteroids being gradually tapered (prednisone 10-30 mg/day or methylprednisolone 16-28 mg/day). Pulse therapy (methylprednisolone 1000 mg/day over 3 days) was used in 13 patients; the effect was transient in all cases but one.
Autopsy of two RP patients revealed typical changes of radiation pneumonitis, including proliferation of atypical type II pneumocytes, thickening of the basement membrane and marked septal fibrosis (16).
 : A
![]() |
 : B
![]() |
Figure 1. Chest radiograph (A) on admission and (B) at onset of radiation pneumonitis, showing infiltrates over the radiation field.
The course of one patient is described here. Patient No.16 in Table 3 was a 70-year-old man. Chest radiograph and CT showed a 4.3 × 3.8 cm mass in the anterior segment of the right upper lobe (Figs 1 and 2). Transbronchial biopsy specimens showed small-cell lung cancer. Radiotherapy consisting of 56 Gy in 28 fractions over 6 weeks was given to the right upper chest and mediastinum after two courses of systemic chemotherapy. Two weeks after completion of radiation, the patient noted a non-productive cough and fever. Chest radiograph and CT showed diffuse bilateral interstitial infiltrates (Figs 1 and 2). Prednisone (40 mg/day) and antibiotics were given. The patient's oxygenation gradually worsened as the pulmonary infiltrates progressed. He expired from cardiorespiratory arrest. At autopsy, both lungs were poorly expanded and consolidated, with diffuse interstitial fibrosis. Areas outside the radiation field showed interstitial fibrosis and proliferation of type II pneumocytes (Fig. 3).
A
![]() |
B
![]() |
Figure 2. Chest CT (A) before and (B) 2 weeks after completion of radiotherapy, showing diffuse bilateral infiltrates.
Figure 3. A section of lung from autopsy exhibits thickening of the alveolar walls, hyaline membranes deposited in alveolar spaces and proliferation of type II alveolar cells. (hematoxylin and eosin staining; original magnification, ×200). Radiation pneumonitis is a poorly understood phenomenon with many variations and complications. It is very important to distinguish between acute and late onset, because the mechanisms of action are probably different. There also is evidence that radiation pneumonitis is not necessarily confined to the treated lung volume. Using bronchoalveolar lavage lymphocyte subset analysis, Roberts et al. (17) have shown that a lymphocytic alveolitis developed in both lung fields after strictly unilateral thoracic irradiation in patients treated for breast cancer. More recently, Morgan et al. (18) proposed two distinct forms of radiation-induced lung injury. One was termed classical radiation pneumonitis, which ultimately leads to pulmonary fibrosis and is confined to the radiation field area. The other was termed sporadic radiation pneumonitis, which is an immunologically mediated process resulting in bilateral lymphocytic alveolitis and an `out-of-field' response to localized pulmonary irradiation. The latter is similar to hypersensitivity pneumonitis in many respects. Similarly, it was suggested that irradiation may induce accumulation of activated T cells (HLADR and ICAM-1-positive) in the lungs and such accumulation linked closely to radiation-induced lung injury (19). CT and gallium-67 citrate scans have produced evidence that radiation pneumonitis can occur at points distant from the site of irradiation (20,21). Furthermore, adult respiratory distress syndrome (ARDS) after thoracic radiotherapy has been reported to occur within days following completion of irradiation; this has a fulminant course, similar to some of our cases (12,14). The recognition that radiation treatment can produce pneumonitis or fibrosis remote from the treated area may obviate the need to investigate invasively the cause of infiltrates in selected cases. Chemotherapy can enhance the incidence of pneumonitis (22-25). Pneumonitis and fibrosis have occured during treatment with actinomycin D (22), cyclophosphamide (22), vincristine (22), bleomycin (23), methotrexate (24), mitomycin (25) and doxorubicin (26,27). However, our data did not show an obvious relationship between chemotherapy and radiation pneumonitis since chemotherapy in the present study was not controlled. Underlying pulmonary disease has been shown to modify the radiation effect. Baseline impairment of pulmonary function will add to radiation damage and cause symptoms with smaller radiation volumes than in patients with normal function (28,29). It has been shown that if radiation passes through an area in which there was tumor, fibrosis, fluid, atelectasis or pneumonia, the exit dose is correspondingly decreased (30). Therefore, it is probable that the presence of interstitial change increases the likelihood of severe radiation pneumonitis. The inclusion of the mediastinum or hilum in the treament volume in continuity with adjoining lung carries the highest risk of radiation injury (31,32). Smith (31) has suggested that damage to the lymphatic outflow tract in the lungs and mediastinum, with subsequent sensitization to radiation of that region of the lung, was the most important element in the pathogenesis of radiation pneumonitis. Our results suggest a significant correlation between the incidence of severe radiation pneumonitis and dosages of radiation to the contralateral mediastinum of >40 Gy. We offer two possible explanations for this observation: first, the radiation field is often larger, including the contralateral mediastinum; second, contralateral lymphatic outflow may be damaged. It has been reported previously that markers such as LDH (33), KL-6 (34), type III procollagen N-terminal peptides (35) and TGF-[beta] (36,37) may help to establish the diagnosis of radiation pneumonitis and evaluate the activity of this disease. In this study, changes in LDH, CRP and PaO2 were significant with the development of acute radiation pneumonitis and therefore may help to detect radiation pneumonitis at an early stage. Symptoms of radiation pneumonitis develop insidiously and become clinically evident 2-3 months after the completion of radiotherapy. An early onset implies a more serious and protracted clinical course (4). In our study, most cases showed early onset and their prognosis was poor. Therefore, it is critical to detect early signs and symptoms of radiation pneumonitis and to establish factors related to prognosis. Fever and a gradual elevation of CRP without evidence of infection may indicate the early likelihood of severe radiation pneumonitis. Some authors have reported improvement of radiation pneumonitis with corticosteroids and others have noted exacerbation of pneumonitis with abrupt steroid withdrawal (38,39). Ward et al. (40) have reported that corticosteroids suppress alveolitis after lung irradiation in the rat. Prednisone is used most frequently at doses of 60-100 mg/day (4,10,28). We suggest that when corticosteroids are employed for radiation pneumonitis, the initial dosage should be continued until fever, dyspnea and cyanosis are completely eliminated and the chest radiograph returns to normal. To avoid recurrence, we further recommend that corticosteroids should be gradually tapered at the rate of 5-10% of the initial dose every 2-4 weeks, depending on symptoms. Azathioprine may be a well-tolerated option in patients with radiation pneumonitis and concomitant corticosteroid toxicity (41). It has also been suggested that interferon-[gamma] inhibits radiation-induced neutrophil and protein influx into alveoli (42). Further investigations are required validate these therapies. In conclusion, we suggest that baseline interstitial changes and radiotherapy to the contralateral mediastinum of >40 Gy are important risk factors for radiation pneumonitis. Fever and changes in LDH, CRP and PaO2 may herald the onset of early radiation pneumonitis. New agents in the treatment of radiation pneumonitis may be prospectively evaluated with these data. The author is very grateful to Mr Kei-ichi Honda, Biometric Analysis Department, Shionogi, for statistical advice.
DISCUSSION
Acknowledgment
References
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification: 13 Apr 1999
Copyright© 1999 Foundation for Promotion of Cancer Research.
![]()
CiteULike
Connotea
Del.icio.us What's this?
This Article ![]()
![]()
Abstract
![]()
Alert me when this article is cited
![]()
Alert me if a correction is posted
![]()
Services ![]()
![]()
Email this article to a friend
![]()
Similar articles in this journal
![]()
Similar articles in ISI Web of Science
![]()
Similar articles in PubMed
![]()
Alert me to new issues of the journal
![]()
Add to My Personal Archive
![]()
Download to citation manager
![]()
Search for citing articles in:
ISI Web of Science (12)
![]()
Request Permissions
![]()
Google Scholar ![]()
![]()
Articles by Makimoto, T
![]()
Articles by Mori, M
![]()
Search for Related Content
![]()
PubMed ![]()
![]()
PubMed Citation
![]()
Articles by Makimoto, T
![]()
Articles by Mori, M
![]()
Social Bookmarking ![]()
![]()
What's this?



