- Email: [email protected]
Radiation pneumonitis following treatment of non–small-cell lung cancer with continuous hyperfractionated accelerated radiotherapy (CHART)
Int. J. Radiation Oncology Biol. Phys., Vol. 56, No. 2, pp. 360 –366, 2003 Copyright © 2003 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/03/$–see front matter
doi:10.1016/S0360-3016(02)04491-7
CLINICAL INVESTIGATION
Lung
RADIATION PNEUMONITIS FOLLOWING TREATMENT OF NON–SMALLCELL LUNG CANCER WITH CONTINUOUS HYPERFRACTIONATED ACCELERATED RADIOTHERAPY (CHART) PETER JENKINS, PH.D., F.R.C.R., KAREN D’AMICO, B.SC., KIM BENSTEAD, M.D., F.R.C.R., AND SEAN ELYAN, M.D., F.R.C.R. Gloucestershire Oncology Centre, Cheltenham General Hospital, Cheltenham, UK Purpose: To determine whether partial volume lung irradiation influences the risk of developing acute radiation pneumonitis after the treatment of non–small-cell lung cancer with continuous hyperfractionated accelerated radiotherapy (CHART). Methods and Materials: We conducted an analysis of 32 patients treated with CHART at the Gloucestershire Oncology Center. Twelve patients were treated using conventional two-dimensional treatment techniques and received elective nodal irradiation (ENI). Their treatment plans were subsequently recapitulated using a three-dimensional treatment planning system. Twenty patients were planned using this system from the outset. For these patients, elective nodal irradiation was omitted. Dose–volume histograms (DVH) were constructed and several parameters analyzed for their ability to predict for the development of pneumonitis. Results: Univariate analysis revealed that the percentage lung volume receiving more than 20 Gy (V20) and the mean lung dose are of predictive value for the development of pneumonitis after CHART. There is a strong correlation between these two parameters. Importantly, partial volume lung irradiation using CHART appears to be better tolerated than conventionally fractionated radiotherapy. The omission of ENI considerably reduces V20. Using a commonly employed 3-beam technique it was also noted that the shape of the planning target volume (PTV) in the transverse plane (expressed as an elliptical index) affects the conformity of the V20 isodose to the PTV. This influences the scope for dose escalation with irregularly shaped tumors. Conclusions: In relation to acute radiation pneumonitis, CHART appears to have a superior therapeutic index than conventionally fractionated radiotherapy. V20 and mean lung dose are useful factors for predicting the risk of this complication. The use of these parameters will aid the selection of optimal treatment plans and provides a basis for future dose escalation studies. © 2003 Elsevier Inc. CHART, Pneumonitis, Lung cancer, Three-dimensional treatment planning.
Radiation therapy used either alone or in combination with other modalities plays a major role in the treatment of non–small-cell lung cancer (NSCLC). However, conventional techniques and fractionation regimens are of only limited effectiveness in achieving local disease control (1). The continuous hyperfractionated accelerated radiotherapy (CHART) protocol (54 Gy delivered in 36 fractions over 12 days) was devised to counteract tumor cell proliferation, thought to occur during protracted treatment schedules. A large multicenter randomized control trial demonstrated that treatment with CHART improves local control and survival when compared with a higher dose of conventionally fractionated radiotherapy (2, 3). Overall, there was a 22% reduction in the relative risk of death equivalent to an absolute improvement in 2-year survival of 9% (from 20%
to 29%). The success of this study has led to the development of similar accelerated protocols that are currently the subject of clinical trials in the United States and Europe. Although the late treatment toxicity experienced by patients undergoing CHART is not significantly different from that seen with conventional fractionated radiation therapy, acute effects are more marked (4). For example, esophagitis developing in the immediate posttreatment period occurs earlier and is more pronounced with CHART. Radiation pneumonitis and fibrosis were assessed both radiologically and clinically in the CHART trial (2). The frequency of acute radiation pneumonitis as determined by chest radiograph was slightly greater in the CHART group, 65% vs. 56%. When assessed by clinical criteria, 19% of the conventionally treated group had symptoms of radiation pneumonitis requiring treatment vs. 10% in CHART. Thus,
Reprint requests to: Peter Jenkins, Ph.D., F.R.C.R., Gloucestershire Oncology Centre, Cheltenham General Hospital, Cheltenham, GL53 7AN, UK. Tel: (44) 1242 274019; Fax: (44) 1242 273506. E-mail: [email protected]
Presented at the annual meeting of the American Society of Clinical Oncology, Orlando, Florida, May 18 –21, 2002. Received Apr 19, 2002, and in revised form Oct 11, 2002. Accepted for publication Nov 18, 2002.
INTRODUCTION
360
Pneumonitis following CHART
despite the fact that treatment is completed in only 12 days, CHART appeared to be better tolerated in terms of the prevalence of posttreatment pneumonitis. The advent of three-dimensional conformal radiation therapy (3D-CRT) has enabled the delivery of radiation to a defined target to be optimized, both improving target volume coverage and reducing normal tissue irradiation. Preclinical and Phase I/II studies have demonstrated that the application of 3D-CRT to the treatment of lung cancer can improve the therapeutic ratio (5, 6). We are particularly interested in combining the advantages of this technology with CHART in an effort to optimize local tumor control. As a prelude to the formal introduction of 3D-CRT into the CHART program at our center, a reanalysis of lung toxicity has been conducted. Our aim was to determine whether partial lung volume irradiation with CHART is better tolerated than conventional radiotherapy as well as to identify parameters that might enable us to estimate the risk of radiation pneumonitis. A more accurate knowledge of partial volume normal organ radiation tolerances for CHART would facilitate decisions regarding the relative merits of candidate treatment plans. It would also allow patients who do not fulfil the CHART selection criteria by virtue of tumor size, to be treated using this protocol. Finally, obtaining this data is a necessary prerequisite for future dose escalation studies. This paper represents the first publication concerning partial lung radiation tolerances for patients treated solely with nonconventional fractionation schemes. The data provided should be of use to other workers modeling the risks of radiation pneumonitis. METHODS AND MATERIALS Thirty-two patients treated with CHART at the Gloucestershire Oncology Center for medically or surgically inoperable NSCLC form the basis of this report. All patients had minimal weight loss at presentation, baseline forced expiratory volume in 1 s (FEV1) ⬎ 1.5 L, no prior surgery, and World Health Organization performance status 0 –1. Two patients received induction chemotherapy as part of a separate study (3 cycles of mitomycin C, vinblastine, and carboplatin). Adjuvant chemotherapy after CHART was not used. Patient demographics and tumor characteristics are shown in Tables 1 and 2. Twelve patients presenting between November 1999 and November 2000 were treated with the traditional two-phase technique as stipulated in the CHART protocol (2). Briefly, patients were planned using two-dimensional techniques based on a central computed tomographic (CT) slice taken at the level of the tumor. Phase I included the mediastinum (ipsilateral hilar and paratracheal nodes but not contralateral hilar nodes) and primary tumor with a 1-cm margin and received a dose of 37.5 Gy in 25 fractions given three times daily. Supraclavicular fossa nodes were not irradiated electively. The maximum anterior field size allowable was 240 cm2. A second phase followed immediately thereafter, delivering a dose of 16.5 Gy in 11 fractions to a smaller volume including only
● P. JENKINS et al.
361
Table 1. Patient and tumor characteristics Age Median Range Race (Caucasian:Afrocaribbean)
71 (70)* 47–83 (59–80) 31:1 (2:0)
Gender Male Female
23 (1) 9 (1)
Pathology Squamous Adenocarcinoma Not specified No histologic confirmation
20 (1) 4 6 2 (1)
Site Right/Light Central/Peripheral
13/19 (1/1) 20/12 (2/0)
* The numbers in brackets represent the characteristics of the 2 patients who went on to develop pneumonitis.
the demonstrable disease and a 1-cm margin. For this phase of treatment, the maximum anterior field size was not to exceed 140 cm2. In total, a dose of 54 Gy was delivered in 36 fractions over 12 days. For these patients the treatment plan was recapitulated on a three-dimensional planning system (FOCUS, CMS Associates, St. Louis, MO). This was performed using the entire planning CT scan to contour normal patient anatomy and the planning target volume (PTV). The original two-dimensional treatment plan (field sizes and wedge factors) was then reconstructed on the three-dimensional volume. The given dose applied to each of the treatment fields was entered and a composite Phase I/II plan produced corrected for lung heterogeneities. The 20 patients treated after November 2000 have been planned entirely on the three-dimensional planning system. For these patients, elective nodal irradiation (ENI) has been omitted and the demonstrable tumor–nodal complex (defined by CT scan) plus a 1–2-cm margin treated in a single phase. The planning scan protocol for all patients reported was contiguous 1-cm slices encompassing the entire volume of the lungs. We commonly employ a 3-beam arrangement designed to minimize the dose to the contralateral lung. The size of the largest and anterior field for all patients was well within that stipulated in the CHART protocol. Conformal beam shaping was used only for the 2 patients treated with induction chemotherapy. Table 2. TNM staging for the 32 patients treated
N0 N1 N2 N3 Total
T1
T2
T3
T4
Total
2 0 1 0 3
10* 0 3* 0 13
3 1 3 1 8
5 0 3 0 8
20 1 10 1 32
* Represents a patient who developed pneumonitis.
362
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 2, 2003
Dose–volume histograms (DVH) were calculated for both lungs as a single organ and for the whole patient. The PTV and large airways were excluded from the lung contours. Normal tissue complication probabilities (NTCP) were derived using the method of Kutcher and Burman (7). The elliptical index of the PTV was defined as the longest major axis/longest minor axis. Patients were usually followed up at weekly intervals until the acute radiation reaction subsided and then every 3 months. A diagnosis of acute radiation pneumonitis was made on the basis of clinical symptoms and radiologic findings developing in the first 3 months after treatment in the absence of any other likely cause. This was graded using the Southwest Oncology Group (SWOG) acute lung morbidity scoring. Due to the difficulty in accurately scoring lower grades of pneumonitis in patients with lung tumors, only Grade 2 and above are reported. Statistical analysis was performed using the general linear model (GLM) univariate and multivariate procedure (SPSS 10.0 software). RESULTS A typical DVH for both lungs using our standard 3-beam arrangement is shown in Fig. 1. As our treatment techniques are designed to minimize the dose to the contralateral lung, the ipsilateral lung receives a much higher mean dose. However, in common with other authors, we have chosen to consider both lungs as a single organ for the basis of all calculations. Three-dimensional treatment plans were used to calculate a number of biophysical parameters that have previously been shown to predict for the development of pneumonitis after conventional fractionated radiotherapy. Two of our patients developed Grade 2 pneumonitis within the first 3 months after the completion of CHART. This corresponds to a crude risk of 6%. Both these patients had centrally located tumors, and both received ENI. The symptoms resolved after treatment with steroids. Univariate analysis has shown that the percentage volume of both lungs receiving greater than 20 Gy (V20) is better correlated with the development of this complication than any of the other factors (Table 3A). None of the factors tested was significant on multivariate analysis. A strong correlation was seen between V20 and mean lung dose (Fig. 2). This most likely reflects the 3-field beam arrangement (anterior, anterior/ posterior oblique, and lateralized oblique) that is commonly employed at our institution. No case of clinically significant (Grade 2) pneumonitis was observed with V20 values ⬍40% or a mean lung dose ⬍20 Gy. A wide range of lung doses was recorded, reflecting the varied anatomic relationship of the tumor–node complex to normal tissue in each patient. As expected, the 20 patients in whom ENI was omitted clustered toward the bottom of this distribution (Fig. 3). The median V20 for this group of patients was 22% compared with 34% for the patients in whom the nodes were electively irradiated. In the course of this analysis, it became apparent that in addition to the inclusion of ENI, the shape of the PTV can
Fig. 1. (A) Typical DVH for both lungs using our standard 3-beam technique. (B) Integral DVH for the ipsilateral lung plotted for the same patient as in A. Two maxima were frequently seen at approximately 40 Gy and 54 Gy corresponding to the intersection of the treatment fields outside the PTV. (C) Diagram illustrating the origin of these maxima for a symmetrically shaped PTV. Dark shading: intersection of three treatment fields. Light shading: intersection of two treatment fields. As the target volume becomes more elliptical these areas of overlap increase.
Pneumonitis following CHART
Table 3A. Univariate analysis of possible predictive factors for the development of pneumonitis* Parameter
p Value
Maximum surface area (of field) Mean lung dose V20 V25 V30 NTCP (n ⫽ 12)
0.41 0.041 0.024 0.026 0.033 0.171
* V20, V25, and V30 are the percentage volumes of both lungs irradiated with more than 20 Gy, 25 Gy, and 30 Gy, respectively. NTCP were calculated for 12 patients all treated with a two-phase technique.
Table 3B. Comparison of the absolute values (median and range) for patients developing pneumonitis with those that did not
All patients
Patients developing pneumonitis
Patients without pneumonitis
Mean lung dose (Gy) 13.2 (5.9–24.9) 21.8 (21.3–22.4) 13.1 (5.9–24.9) V20 (%) 24 (8–46) 42 (40–44) 23.5 (8–46) V25 (%) 22 (7–44) 38 (36–40) 21 (7–44) V30 (%) 17.7 (6–39) 31 (31–31) 17.2 (6–39)
● P. JENKINS et al.
363
also influence V20. In particular, the more ellipsoid the PTV in the transverse plane the lower the conformity index of V20 to PTV (Fig. 4). This principally results from the overlap of the entry and exit beams outside the PTV, which is also responsible for the peaks seen in the integral DVH for the ipsilateral lung (Figs. 1B and 1C). As the PTV becomes less elliptical (more circular), these peaks become smaller and the conformity of the V20 isodose to the PTV improves.
DISCUSSION Acute radiation pneumonitis is a common complication of curative radiotherapy schedules for NSCLC (8). Pathophysiologically, this process involves all components of the lung parenchyma and comprises an inflammatory infiltrate of the alveolar wall, destruction of the type I and II pneumocytes, endothelial injury, and small vessel thrombosis (9). The syndrome is usually subacute, developing 6 –12 weeks after completion of radiotherapy, but more severe grades develop earlier and are potentially life-threatening. Although pneumonitis is not the only dose-limiting toxicity following thoracic radiotherapy, with accelerated fractionation schemes like CHART such acute effects are a major concern. The incidence of radiation pneumonitis appears to be dependent on total dose, fractionation schedule, and irradi-
Fig. 2. Scatterplot of the mean lung dose vs. the percentage of the lung volume irradiated with more than 20 Gy (V20). Solid squares represent patients developing Grade 2 pneumonitis. Correlation coefficient 0.98.
364
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 2, 2003
Fig. 3. Boxplot showing the percentage lung volume receiving greater than 20 Gy (V20) in patients treated with (n ⫽ 12) and without (n ⫽ 20) ENI. The 2 patients who developed pneumonitis both received ENI. Median (solid line), interquartile range (shaded box), and extreme values are shown.
ated volume, although the precise nature of this relationship remains unclear (10). Thus, when the entire lung is irradiated with a single large fraction as in total body irradiation, there is a steep dose–response relationship starting at 8.2 Gy (11). Conversely, treatment with multiple daily fractions to a whole lung dose of 17.5 Gy appears to be well tolerated (12). It is worthy of note that, irrespective of the threshold value, the dose–response relationship for every fractionated regimen studied is steep. Clinical experience also indicates that partial lung irradiation is better tolerated than whole lung exposure. Several groups have attempted to define tolerance doses for partial lung irradiation because the ability to predict which patients might be at increased risk of developing lung damage would be of considerable clinical use. The influential study of Graham et al. correlated the risk of pneumonitis with V20 (13). They found that with V20 values between 22% and 31%, the actuarial incidence of Grade 2 or higher pneumonitis is 7%. For V20 values of 32% to 40% and ⬎40%, the respective figures are 13% and 36%. These data were based on 1.8 –2 Gy fraction sizes to typical doses of 70 Gy in 35 fractions. Similar findings using a threshold of 25 Gy and 30 Gy have been reported by others (14, 15). Mean lung dose and mean biologic lung dose have also been shown to be predictive for the development of pneumonitis (16 –18). However, it would be hazardous to simply extrapolate the results from these reports to altered fraction schemes that are likely to have
biologically different effects on lung tissue. On the basis of the smaller fraction size in CHART (1.5 Gy), one might predict that this schedule would be less likely to produce pneumonitis. However, this effect may be offset by the shorter overall treatment time and reduced opportunity for sublethal damage repair (19, 20). We therefore sought to analyze a group of patients who had been treated with CHART to determine whether the physical dose distribution influences the risk of pneumonitis. We have confirmed that simple parameters derived from the DVH can be used to predict pneumonitis risk with CHART. As with conventional treatment schedules V20 and mean lung dose appear to correlate strongly with the development of this complication (Table 3B and Fig. 2). This is despite the fact that the V20 for any given treatment plan with CHART is smaller than for conventional radical doses. For example, for a standard plan using a one-phase technique, the V20 volume for conventional radiation therapy to 70 Gy is defined by the 28% isodose surface delivered in fraction sizes of 0.57 Gy. For CHART to 54 Gy, the similar V20 volume would be encompassed by the smaller 37% isodose surface and delivered in a fraction sizes of 0.55 Gy. Field size, as specified in the CHART study, is not predictive of pneumonitis risk, and we would recommend that this no longer be used to select patients for treatment with 3D-CRT. The poor correlation with NTCP is perhaps not surprising given the fact that the parameterization for this
Pneumonitis following CHART
● P. JENKINS et al.
365
Fig. 4. Scatterplot of the elliptical index (longest major axis/longest minor axis) vs. the V20 conformity index for 10 patients. (The V20 conformity index was defined as the volume of the patient irradiated to more than 20 Gy/PTV.) Correlation coefficient 0.63 (significant at the 0.05 level, two-tailed).
model is derived from conventionally fractionated radiation therapy. Two patients in our series (6%) developed Grade 2 pneumonitis in keeping with the crude prevalence rates reported in the CHART trial (2). It is noteworthy that we did not see any cases of pneumonitis below a V20 of 40% or a mean lung dose of 20 Gy. When compared with the data of Graham et al. it appears that partial volume lung irradiation with CHART is less likely to cause radiation pneumonitis than conventionally fractionated treatment (13). However, in addition to radiobiologic factors, treatment technique and the use of induction chemotherapy may contribute to this finding. In common with other reports, our treatment plans show considerable heterogeneity in the lung doses received (Fig. 3). However, when ENI is omitted there is a significant reduction in V20, suggesting that there is potential for dose escalation with CHART. It was also evident that the shape of the PTV in the transverse plane influences the V20 value. As the PTV becomes less circular and more elliptical, the conformity of V20 to the PTV declines. This may be of limited importance if the 20 Gy isodose substantially overlaps the mediastinum, but for target volumes within the lung it becomes more significant. It should be noted that, as a first approximation, the PTV becomes more ellipsoid when the hilar nodes are included in the treatment plan. While better conformity can be achieved by using additional beams,
conformal shaping, and the omission of ENI, this observation suggests that symmetrically shaped target volumes are the most suitable for dose escalation. The superiority of the CHART regimen over conventionally fractionated radiotherapy confirmed that cellular repopulation during protracted treatment courses is a major factor in local failure. The success of this regimen has engendered interest in dose escalation using CHART or variants thereof such as CHARTWEL and HART. Preliminary experience from one such schedule has shown an increase in esophageal and lung toxicity (21). On the basis of our data, we would recommend that future dose escalation protocols take account of the partial volume of lung irradiated. Finally, it must be appreciated that the development of acute radiation pneumonitis is multifactorial. Our analysis takes no account of atelectasis and hypoperfusion related to the tumor or to comorbid lung conditions which can influence lung function. Factors such as interleukin-6 and transforming growth factor-beta have also recently been shown to be important in the development of pneumonitis (22, 23). This may account for the observed heterogeneity in the risk of pneumonitis seen at a single dose level. Ultimately, it is likely that no one variable will adequately predict pneumonitis risk and that multivariate models will be needed (15, 24). In summary, this analysis has characterized the volumet-
366
I. J. Radiation Oncology
● Biology ● Physics
ric relationship between the dose of radiation and the risk of radiation pneumonitis after CHART. It will be of use in determining guidelines for the selection and safe treatment of
Volume 56, Number 2, 2003
patients with this protocol. The prospective evaluation of these parameters will permit further refinement of these data and aid the development of this accelerated fractionation protocol.
REFERENCES 1. Le Chevalier T, Arriagada R, Quoix E, et al. Radiotherapy alone versus combined chemotherapy and radiotherapy in nonresectable non-small-cell lung cancer: First analysis of a randomized trial in 353 patients. J Natl Cancer Inst 1991;83: 417–423. 2. Saunders M, Dische S, Barrett A, et al. Continuous hyperfractionated accelerated radiotherapy (CHART) versus conventional radiotherapy in non-small-cell lung cancer: A randomised multicentre trial. CHART Steering Committee. Lancet 1997;350:161–165. 3. Saunders M, Dische S, Barrett A, et al. Continuous, hyperfractionated, accelerated radiotherapy (CHART) versus conventional radiotherapy in non-small cell lung cancer: Mature data from the randomised multicentre trial. CHART Steering Committee. Radiother Oncol 1999;52:137–148. 4. Dische S, Saunders MI. The CHART regimen and morbidity. Acta Oncol 1999;38:147–152. 5. Armstrong JG, Burman C, Leibel S, et al. Three-dimensional conformal radiation therapy may improve the therapeutic ratio of high dose radiation therapy for lung cancer. Int J Radiat Oncol Biol Phys 1993;26:685–689. 6. Armstrong JG, Zelefsky MJ, Leibel SA, et al. Strategy for dose escalation using 3-dimensional conformal radiation therapy for lung cancer. Ann Oncol 1995;6:693–697. 7. Burman C, Kutcher GJ, Emami B, et al. Fitting of normal tissue tolerance data to an analytic function. Int J Radiat Oncol Biol Phys 1991;21:123–135. 8. Marks LB. The pulmonary effects of thoracic irradiation. Oncology (Huntingt) 1994;8:89–106. 9. McDonald S, Rubin P, Phillips TL, et al. Injury to the lung from cancer therapy: Clinical syndromes, measurable endpoints, and potential scoring systems. Int J Radiat Oncol Biol Phys 1995;31:1187–1203. 10. Seppenwoolde Y, Lebesque JV. Partial irradiation of the lung. Semin Radiat Oncol 2001;11:247–258. 11. Van Dyk J, Keane TJ, Kan S, et al. Radiation pneumonitis following large single dose irradiation: A re-evaluation based on absolute dose to lung. Int J Radiat Oncol Biol Phys 1981; 7:461–467. 12. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21:109–122. 13. Graham MV, Purdy JA, Emami B, et al. Clinical dose–volume
14. 15.
16.
17.
18.
19. 20. 21.
22. 23. 24.
histogram analysis for pneumonitis after 3D treatment for non–small cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys 1999;45:323–329. Armstrong J, Raben A, Zelefsky M, et al. Promising survival with three-dimensional conformal radiation therapy for nonsmall cell lung cancer. Radiother Oncol 1997;44:17–22. Marks LB, Munley MT, Bentel GC, et al. Physical and biological predictors of changes in whole-lung function following thoracic irradiation. Int J Radiat Oncol Biol Phys 1997;39: 563–570. Martel MK, Ten Haken RK, Hazuka MB, et al. Dose–volume histogram and 3-D treatment planning evaluation of patients with pneumonitis. Int J Radiat Oncol Biol Phys 1994;28:575– 581. Oetzel D, Schraube P, Hensley F, et al. Estimation of pneumonitis risk in three-dimensional treatment planning using dose–volume histogram analysis. Int J Radiat Oncol Biol Phys 1995;33:455–460. Kwa SL, Lebesque JV, Theuws JC, et al. Radiation pneumonitis as a function of mean lung dose: An analysis of pooled data of 540 patients. Int J Radiat Oncol Biol Phys 1998;42: 1–9. Van Dyk J, Mah K, Keane TJ. Radiation-induced lung damage: Dose-time-fractionation considerations. Radiother Oncol 1989;14:55–69. Bentzen SM, Skoczylas JZ, Bernier J. Quantitative clinical radiobiology of early and late lung reactions. Int J Radiat Biol 2000;76:453–462. Saunders MI, Rojas A, Lyn BE, et al. Experience with dose escalation using CHARTWEL (continuous hyperfractionated accelerated radiotherapy weekend less) in non-small-cell lung cancer. Br J Cancer 1998;78:1323–1328. Chen Y, Rubin P, Williams J, et al. Circulating IL-6 as a predictor of radiation pneumonitis. Int J Radiat Oncol Biol Phys 2001;49:641–648. Anscher MS, Kong FM, Andrews K, et al. Plasma transforming growth factor beta1 as a predictor of radiation pneumonitis. Int J Radiat Oncol Biol Phys 1998;41:1029–1035. Fu XL, Huang H, Bentel G, et al. Predicting the risk of symptomatic radiation-induced lung injury using both the physical and biologic parameters V(30) and transforming growth factor beta. Int J Radiat Oncol Biol Phys 2001;50:899– 908.
doi:10.1016/S0360-3016(02)04491-7
CLINICAL INVESTIGATION
Lung
RADIATION PNEUMONITIS FOLLOWING TREATMENT OF NON–SMALLCELL LUNG CANCER WITH CONTINUOUS HYPERFRACTIONATED ACCELERATED RADIOTHERAPY (CHART) PETER JENKINS, PH.D., F.R.C.R., KAREN D’AMICO, B.SC., KIM BENSTEAD, M.D., F.R.C.R., AND SEAN ELYAN, M.D., F.R.C.R. Gloucestershire Oncology Centre, Cheltenham General Hospital, Cheltenham, UK Purpose: To determine whether partial volume lung irradiation influences the risk of developing acute radiation pneumonitis after the treatment of non–small-cell lung cancer with continuous hyperfractionated accelerated radiotherapy (CHART). Methods and Materials: We conducted an analysis of 32 patients treated with CHART at the Gloucestershire Oncology Center. Twelve patients were treated using conventional two-dimensional treatment techniques and received elective nodal irradiation (ENI). Their treatment plans were subsequently recapitulated using a three-dimensional treatment planning system. Twenty patients were planned using this system from the outset. For these patients, elective nodal irradiation was omitted. Dose–volume histograms (DVH) were constructed and several parameters analyzed for their ability to predict for the development of pneumonitis. Results: Univariate analysis revealed that the percentage lung volume receiving more than 20 Gy (V20) and the mean lung dose are of predictive value for the development of pneumonitis after CHART. There is a strong correlation between these two parameters. Importantly, partial volume lung irradiation using CHART appears to be better tolerated than conventionally fractionated radiotherapy. The omission of ENI considerably reduces V20. Using a commonly employed 3-beam technique it was also noted that the shape of the planning target volume (PTV) in the transverse plane (expressed as an elliptical index) affects the conformity of the V20 isodose to the PTV. This influences the scope for dose escalation with irregularly shaped tumors. Conclusions: In relation to acute radiation pneumonitis, CHART appears to have a superior therapeutic index than conventionally fractionated radiotherapy. V20 and mean lung dose are useful factors for predicting the risk of this complication. The use of these parameters will aid the selection of optimal treatment plans and provides a basis for future dose escalation studies. © 2003 Elsevier Inc. CHART, Pneumonitis, Lung cancer, Three-dimensional treatment planning.
Radiation therapy used either alone or in combination with other modalities plays a major role in the treatment of non–small-cell lung cancer (NSCLC). However, conventional techniques and fractionation regimens are of only limited effectiveness in achieving local disease control (1). The continuous hyperfractionated accelerated radiotherapy (CHART) protocol (54 Gy delivered in 36 fractions over 12 days) was devised to counteract tumor cell proliferation, thought to occur during protracted treatment schedules. A large multicenter randomized control trial demonstrated that treatment with CHART improves local control and survival when compared with a higher dose of conventionally fractionated radiotherapy (2, 3). Overall, there was a 22% reduction in the relative risk of death equivalent to an absolute improvement in 2-year survival of 9% (from 20%
to 29%). The success of this study has led to the development of similar accelerated protocols that are currently the subject of clinical trials in the United States and Europe. Although the late treatment toxicity experienced by patients undergoing CHART is not significantly different from that seen with conventional fractionated radiation therapy, acute effects are more marked (4). For example, esophagitis developing in the immediate posttreatment period occurs earlier and is more pronounced with CHART. Radiation pneumonitis and fibrosis were assessed both radiologically and clinically in the CHART trial (2). The frequency of acute radiation pneumonitis as determined by chest radiograph was slightly greater in the CHART group, 65% vs. 56%. When assessed by clinical criteria, 19% of the conventionally treated group had symptoms of radiation pneumonitis requiring treatment vs. 10% in CHART. Thus,
Reprint requests to: Peter Jenkins, Ph.D., F.R.C.R., Gloucestershire Oncology Centre, Cheltenham General Hospital, Cheltenham, GL53 7AN, UK. Tel: (44) 1242 274019; Fax: (44) 1242 273506. E-mail: [email protected]
Presented at the annual meeting of the American Society of Clinical Oncology, Orlando, Florida, May 18 –21, 2002. Received Apr 19, 2002, and in revised form Oct 11, 2002. Accepted for publication Nov 18, 2002.
INTRODUCTION
360
Pneumonitis following CHART
despite the fact that treatment is completed in only 12 days, CHART appeared to be better tolerated in terms of the prevalence of posttreatment pneumonitis. The advent of three-dimensional conformal radiation therapy (3D-CRT) has enabled the delivery of radiation to a defined target to be optimized, both improving target volume coverage and reducing normal tissue irradiation. Preclinical and Phase I/II studies have demonstrated that the application of 3D-CRT to the treatment of lung cancer can improve the therapeutic ratio (5, 6). We are particularly interested in combining the advantages of this technology with CHART in an effort to optimize local tumor control. As a prelude to the formal introduction of 3D-CRT into the CHART program at our center, a reanalysis of lung toxicity has been conducted. Our aim was to determine whether partial lung volume irradiation with CHART is better tolerated than conventional radiotherapy as well as to identify parameters that might enable us to estimate the risk of radiation pneumonitis. A more accurate knowledge of partial volume normal organ radiation tolerances for CHART would facilitate decisions regarding the relative merits of candidate treatment plans. It would also allow patients who do not fulfil the CHART selection criteria by virtue of tumor size, to be treated using this protocol. Finally, obtaining this data is a necessary prerequisite for future dose escalation studies. This paper represents the first publication concerning partial lung radiation tolerances for patients treated solely with nonconventional fractionation schemes. The data provided should be of use to other workers modeling the risks of radiation pneumonitis. METHODS AND MATERIALS Thirty-two patients treated with CHART at the Gloucestershire Oncology Center for medically or surgically inoperable NSCLC form the basis of this report. All patients had minimal weight loss at presentation, baseline forced expiratory volume in 1 s (FEV1) ⬎ 1.5 L, no prior surgery, and World Health Organization performance status 0 –1. Two patients received induction chemotherapy as part of a separate study (3 cycles of mitomycin C, vinblastine, and carboplatin). Adjuvant chemotherapy after CHART was not used. Patient demographics and tumor characteristics are shown in Tables 1 and 2. Twelve patients presenting between November 1999 and November 2000 were treated with the traditional two-phase technique as stipulated in the CHART protocol (2). Briefly, patients were planned using two-dimensional techniques based on a central computed tomographic (CT) slice taken at the level of the tumor. Phase I included the mediastinum (ipsilateral hilar and paratracheal nodes but not contralateral hilar nodes) and primary tumor with a 1-cm margin and received a dose of 37.5 Gy in 25 fractions given three times daily. Supraclavicular fossa nodes were not irradiated electively. The maximum anterior field size allowable was 240 cm2. A second phase followed immediately thereafter, delivering a dose of 16.5 Gy in 11 fractions to a smaller volume including only
● P. JENKINS et al.
361
Table 1. Patient and tumor characteristics Age Median Range Race (Caucasian:Afrocaribbean)
71 (70)* 47–83 (59–80) 31:1 (2:0)
Gender Male Female
23 (1) 9 (1)
Pathology Squamous Adenocarcinoma Not specified No histologic confirmation
20 (1) 4 6 2 (1)
Site Right/Light Central/Peripheral
13/19 (1/1) 20/12 (2/0)
* The numbers in brackets represent the characteristics of the 2 patients who went on to develop pneumonitis.
the demonstrable disease and a 1-cm margin. For this phase of treatment, the maximum anterior field size was not to exceed 140 cm2. In total, a dose of 54 Gy was delivered in 36 fractions over 12 days. For these patients the treatment plan was recapitulated on a three-dimensional planning system (FOCUS, CMS Associates, St. Louis, MO). This was performed using the entire planning CT scan to contour normal patient anatomy and the planning target volume (PTV). The original two-dimensional treatment plan (field sizes and wedge factors) was then reconstructed on the three-dimensional volume. The given dose applied to each of the treatment fields was entered and a composite Phase I/II plan produced corrected for lung heterogeneities. The 20 patients treated after November 2000 have been planned entirely on the three-dimensional planning system. For these patients, elective nodal irradiation (ENI) has been omitted and the demonstrable tumor–nodal complex (defined by CT scan) plus a 1–2-cm margin treated in a single phase. The planning scan protocol for all patients reported was contiguous 1-cm slices encompassing the entire volume of the lungs. We commonly employ a 3-beam arrangement designed to minimize the dose to the contralateral lung. The size of the largest and anterior field for all patients was well within that stipulated in the CHART protocol. Conformal beam shaping was used only for the 2 patients treated with induction chemotherapy. Table 2. TNM staging for the 32 patients treated
N0 N1 N2 N3 Total
T1
T2
T3
T4
Total
2 0 1 0 3
10* 0 3* 0 13
3 1 3 1 8
5 0 3 0 8
20 1 10 1 32
* Represents a patient who developed pneumonitis.
362
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 2, 2003
Dose–volume histograms (DVH) were calculated for both lungs as a single organ and for the whole patient. The PTV and large airways were excluded from the lung contours. Normal tissue complication probabilities (NTCP) were derived using the method of Kutcher and Burman (7). The elliptical index of the PTV was defined as the longest major axis/longest minor axis. Patients were usually followed up at weekly intervals until the acute radiation reaction subsided and then every 3 months. A diagnosis of acute radiation pneumonitis was made on the basis of clinical symptoms and radiologic findings developing in the first 3 months after treatment in the absence of any other likely cause. This was graded using the Southwest Oncology Group (SWOG) acute lung morbidity scoring. Due to the difficulty in accurately scoring lower grades of pneumonitis in patients with lung tumors, only Grade 2 and above are reported. Statistical analysis was performed using the general linear model (GLM) univariate and multivariate procedure (SPSS 10.0 software). RESULTS A typical DVH for both lungs using our standard 3-beam arrangement is shown in Fig. 1. As our treatment techniques are designed to minimize the dose to the contralateral lung, the ipsilateral lung receives a much higher mean dose. However, in common with other authors, we have chosen to consider both lungs as a single organ for the basis of all calculations. Three-dimensional treatment plans were used to calculate a number of biophysical parameters that have previously been shown to predict for the development of pneumonitis after conventional fractionated radiotherapy. Two of our patients developed Grade 2 pneumonitis within the first 3 months after the completion of CHART. This corresponds to a crude risk of 6%. Both these patients had centrally located tumors, and both received ENI. The symptoms resolved after treatment with steroids. Univariate analysis has shown that the percentage volume of both lungs receiving greater than 20 Gy (V20) is better correlated with the development of this complication than any of the other factors (Table 3A). None of the factors tested was significant on multivariate analysis. A strong correlation was seen between V20 and mean lung dose (Fig. 2). This most likely reflects the 3-field beam arrangement (anterior, anterior/ posterior oblique, and lateralized oblique) that is commonly employed at our institution. No case of clinically significant (Grade 2) pneumonitis was observed with V20 values ⬍40% or a mean lung dose ⬍20 Gy. A wide range of lung doses was recorded, reflecting the varied anatomic relationship of the tumor–node complex to normal tissue in each patient. As expected, the 20 patients in whom ENI was omitted clustered toward the bottom of this distribution (Fig. 3). The median V20 for this group of patients was 22% compared with 34% for the patients in whom the nodes were electively irradiated. In the course of this analysis, it became apparent that in addition to the inclusion of ENI, the shape of the PTV can
Fig. 1. (A) Typical DVH for both lungs using our standard 3-beam technique. (B) Integral DVH for the ipsilateral lung plotted for the same patient as in A. Two maxima were frequently seen at approximately 40 Gy and 54 Gy corresponding to the intersection of the treatment fields outside the PTV. (C) Diagram illustrating the origin of these maxima for a symmetrically shaped PTV. Dark shading: intersection of three treatment fields. Light shading: intersection of two treatment fields. As the target volume becomes more elliptical these areas of overlap increase.
Pneumonitis following CHART
Table 3A. Univariate analysis of possible predictive factors for the development of pneumonitis* Parameter
p Value
Maximum surface area (of field) Mean lung dose V20 V25 V30 NTCP (n ⫽ 12)
0.41 0.041 0.024 0.026 0.033 0.171
* V20, V25, and V30 are the percentage volumes of both lungs irradiated with more than 20 Gy, 25 Gy, and 30 Gy, respectively. NTCP were calculated for 12 patients all treated with a two-phase technique.
Table 3B. Comparison of the absolute values (median and range) for patients developing pneumonitis with those that did not
All patients
Patients developing pneumonitis
Patients without pneumonitis
Mean lung dose (Gy) 13.2 (5.9–24.9) 21.8 (21.3–22.4) 13.1 (5.9–24.9) V20 (%) 24 (8–46) 42 (40–44) 23.5 (8–46) V25 (%) 22 (7–44) 38 (36–40) 21 (7–44) V30 (%) 17.7 (6–39) 31 (31–31) 17.2 (6–39)
● P. JENKINS et al.
363
also influence V20. In particular, the more ellipsoid the PTV in the transverse plane the lower the conformity index of V20 to PTV (Fig. 4). This principally results from the overlap of the entry and exit beams outside the PTV, which is also responsible for the peaks seen in the integral DVH for the ipsilateral lung (Figs. 1B and 1C). As the PTV becomes less elliptical (more circular), these peaks become smaller and the conformity of the V20 isodose to the PTV improves.
DISCUSSION Acute radiation pneumonitis is a common complication of curative radiotherapy schedules for NSCLC (8). Pathophysiologically, this process involves all components of the lung parenchyma and comprises an inflammatory infiltrate of the alveolar wall, destruction of the type I and II pneumocytes, endothelial injury, and small vessel thrombosis (9). The syndrome is usually subacute, developing 6 –12 weeks after completion of radiotherapy, but more severe grades develop earlier and are potentially life-threatening. Although pneumonitis is not the only dose-limiting toxicity following thoracic radiotherapy, with accelerated fractionation schemes like CHART such acute effects are a major concern. The incidence of radiation pneumonitis appears to be dependent on total dose, fractionation schedule, and irradi-
Fig. 2. Scatterplot of the mean lung dose vs. the percentage of the lung volume irradiated with more than 20 Gy (V20). Solid squares represent patients developing Grade 2 pneumonitis. Correlation coefficient 0.98.
364
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 2, 2003
Fig. 3. Boxplot showing the percentage lung volume receiving greater than 20 Gy (V20) in patients treated with (n ⫽ 12) and without (n ⫽ 20) ENI. The 2 patients who developed pneumonitis both received ENI. Median (solid line), interquartile range (shaded box), and extreme values are shown.
ated volume, although the precise nature of this relationship remains unclear (10). Thus, when the entire lung is irradiated with a single large fraction as in total body irradiation, there is a steep dose–response relationship starting at 8.2 Gy (11). Conversely, treatment with multiple daily fractions to a whole lung dose of 17.5 Gy appears to be well tolerated (12). It is worthy of note that, irrespective of the threshold value, the dose–response relationship for every fractionated regimen studied is steep. Clinical experience also indicates that partial lung irradiation is better tolerated than whole lung exposure. Several groups have attempted to define tolerance doses for partial lung irradiation because the ability to predict which patients might be at increased risk of developing lung damage would be of considerable clinical use. The influential study of Graham et al. correlated the risk of pneumonitis with V20 (13). They found that with V20 values between 22% and 31%, the actuarial incidence of Grade 2 or higher pneumonitis is 7%. For V20 values of 32% to 40% and ⬎40%, the respective figures are 13% and 36%. These data were based on 1.8 –2 Gy fraction sizes to typical doses of 70 Gy in 35 fractions. Similar findings using a threshold of 25 Gy and 30 Gy have been reported by others (14, 15). Mean lung dose and mean biologic lung dose have also been shown to be predictive for the development of pneumonitis (16 –18). However, it would be hazardous to simply extrapolate the results from these reports to altered fraction schemes that are likely to have
biologically different effects on lung tissue. On the basis of the smaller fraction size in CHART (1.5 Gy), one might predict that this schedule would be less likely to produce pneumonitis. However, this effect may be offset by the shorter overall treatment time and reduced opportunity for sublethal damage repair (19, 20). We therefore sought to analyze a group of patients who had been treated with CHART to determine whether the physical dose distribution influences the risk of pneumonitis. We have confirmed that simple parameters derived from the DVH can be used to predict pneumonitis risk with CHART. As with conventional treatment schedules V20 and mean lung dose appear to correlate strongly with the development of this complication (Table 3B and Fig. 2). This is despite the fact that the V20 for any given treatment plan with CHART is smaller than for conventional radical doses. For example, for a standard plan using a one-phase technique, the V20 volume for conventional radiation therapy to 70 Gy is defined by the 28% isodose surface delivered in fraction sizes of 0.57 Gy. For CHART to 54 Gy, the similar V20 volume would be encompassed by the smaller 37% isodose surface and delivered in a fraction sizes of 0.55 Gy. Field size, as specified in the CHART study, is not predictive of pneumonitis risk, and we would recommend that this no longer be used to select patients for treatment with 3D-CRT. The poor correlation with NTCP is perhaps not surprising given the fact that the parameterization for this
Pneumonitis following CHART
● P. JENKINS et al.
365
Fig. 4. Scatterplot of the elliptical index (longest major axis/longest minor axis) vs. the V20 conformity index for 10 patients. (The V20 conformity index was defined as the volume of the patient irradiated to more than 20 Gy/PTV.) Correlation coefficient 0.63 (significant at the 0.05 level, two-tailed).
model is derived from conventionally fractionated radiation therapy. Two patients in our series (6%) developed Grade 2 pneumonitis in keeping with the crude prevalence rates reported in the CHART trial (2). It is noteworthy that we did not see any cases of pneumonitis below a V20 of 40% or a mean lung dose of 20 Gy. When compared with the data of Graham et al. it appears that partial volume lung irradiation with CHART is less likely to cause radiation pneumonitis than conventionally fractionated treatment (13). However, in addition to radiobiologic factors, treatment technique and the use of induction chemotherapy may contribute to this finding. In common with other reports, our treatment plans show considerable heterogeneity in the lung doses received (Fig. 3). However, when ENI is omitted there is a significant reduction in V20, suggesting that there is potential for dose escalation with CHART. It was also evident that the shape of the PTV in the transverse plane influences the V20 value. As the PTV becomes less circular and more elliptical, the conformity of V20 to the PTV declines. This may be of limited importance if the 20 Gy isodose substantially overlaps the mediastinum, but for target volumes within the lung it becomes more significant. It should be noted that, as a first approximation, the PTV becomes more ellipsoid when the hilar nodes are included in the treatment plan. While better conformity can be achieved by using additional beams,
conformal shaping, and the omission of ENI, this observation suggests that symmetrically shaped target volumes are the most suitable for dose escalation. The superiority of the CHART regimen over conventionally fractionated radiotherapy confirmed that cellular repopulation during protracted treatment courses is a major factor in local failure. The success of this regimen has engendered interest in dose escalation using CHART or variants thereof such as CHARTWEL and HART. Preliminary experience from one such schedule has shown an increase in esophageal and lung toxicity (21). On the basis of our data, we would recommend that future dose escalation protocols take account of the partial volume of lung irradiated. Finally, it must be appreciated that the development of acute radiation pneumonitis is multifactorial. Our analysis takes no account of atelectasis and hypoperfusion related to the tumor or to comorbid lung conditions which can influence lung function. Factors such as interleukin-6 and transforming growth factor-beta have also recently been shown to be important in the development of pneumonitis (22, 23). This may account for the observed heterogeneity in the risk of pneumonitis seen at a single dose level. Ultimately, it is likely that no one variable will adequately predict pneumonitis risk and that multivariate models will be needed (15, 24). In summary, this analysis has characterized the volumet-
366
I. J. Radiation Oncology
● Biology ● Physics
ric relationship between the dose of radiation and the risk of radiation pneumonitis after CHART. It will be of use in determining guidelines for the selection and safe treatment of
Volume 56, Number 2, 2003
patients with this protocol. The prospective evaluation of these parameters will permit further refinement of these data and aid the development of this accelerated fractionation protocol.
REFERENCES 1. Le Chevalier T, Arriagada R, Quoix E, et al. Radiotherapy alone versus combined chemotherapy and radiotherapy in nonresectable non-small-cell lung cancer: First analysis of a randomized trial in 353 patients. J Natl Cancer Inst 1991;83: 417–423. 2. Saunders M, Dische S, Barrett A, et al. Continuous hyperfractionated accelerated radiotherapy (CHART) versus conventional radiotherapy in non-small-cell lung cancer: A randomised multicentre trial. CHART Steering Committee. Lancet 1997;350:161–165. 3. Saunders M, Dische S, Barrett A, et al. Continuous, hyperfractionated, accelerated radiotherapy (CHART) versus conventional radiotherapy in non-small cell lung cancer: Mature data from the randomised multicentre trial. CHART Steering Committee. Radiother Oncol 1999;52:137–148. 4. Dische S, Saunders MI. The CHART regimen and morbidity. Acta Oncol 1999;38:147–152. 5. Armstrong JG, Burman C, Leibel S, et al. Three-dimensional conformal radiation therapy may improve the therapeutic ratio of high dose radiation therapy for lung cancer. Int J Radiat Oncol Biol Phys 1993;26:685–689. 6. Armstrong JG, Zelefsky MJ, Leibel SA, et al. Strategy for dose escalation using 3-dimensional conformal radiation therapy for lung cancer. Ann Oncol 1995;6:693–697. 7. Burman C, Kutcher GJ, Emami B, et al. Fitting of normal tissue tolerance data to an analytic function. Int J Radiat Oncol Biol Phys 1991;21:123–135. 8. Marks LB. The pulmonary effects of thoracic irradiation. Oncology (Huntingt) 1994;8:89–106. 9. McDonald S, Rubin P, Phillips TL, et al. Injury to the lung from cancer therapy: Clinical syndromes, measurable endpoints, and potential scoring systems. Int J Radiat Oncol Biol Phys 1995;31:1187–1203. 10. Seppenwoolde Y, Lebesque JV. Partial irradiation of the lung. Semin Radiat Oncol 2001;11:247–258. 11. Van Dyk J, Keane TJ, Kan S, et al. Radiation pneumonitis following large single dose irradiation: A re-evaluation based on absolute dose to lung. Int J Radiat Oncol Biol Phys 1981; 7:461–467. 12. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21:109–122. 13. Graham MV, Purdy JA, Emami B, et al. Clinical dose–volume
14. 15.
16.
17.
18.
19. 20. 21.
22. 23. 24.
histogram analysis for pneumonitis after 3D treatment for non–small cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys 1999;45:323–329. Armstrong J, Raben A, Zelefsky M, et al. Promising survival with three-dimensional conformal radiation therapy for nonsmall cell lung cancer. Radiother Oncol 1997;44:17–22. Marks LB, Munley MT, Bentel GC, et al. Physical and biological predictors of changes in whole-lung function following thoracic irradiation. Int J Radiat Oncol Biol Phys 1997;39: 563–570. Martel MK, Ten Haken RK, Hazuka MB, et al. Dose–volume histogram and 3-D treatment planning evaluation of patients with pneumonitis. Int J Radiat Oncol Biol Phys 1994;28:575– 581. Oetzel D, Schraube P, Hensley F, et al. Estimation of pneumonitis risk in three-dimensional treatment planning using dose–volume histogram analysis. Int J Radiat Oncol Biol Phys 1995;33:455–460. Kwa SL, Lebesque JV, Theuws JC, et al. Radiation pneumonitis as a function of mean lung dose: An analysis of pooled data of 540 patients. Int J Radiat Oncol Biol Phys 1998;42: 1–9. Van Dyk J, Mah K, Keane TJ. Radiation-induced lung damage: Dose-time-fractionation considerations. Radiother Oncol 1989;14:55–69. Bentzen SM, Skoczylas JZ, Bernier J. Quantitative clinical radiobiology of early and late lung reactions. Int J Radiat Biol 2000;76:453–462. Saunders MI, Rojas A, Lyn BE, et al. Experience with dose escalation using CHARTWEL (continuous hyperfractionated accelerated radiotherapy weekend less) in non-small-cell lung cancer. Br J Cancer 1998;78:1323–1328. Chen Y, Rubin P, Williams J, et al. Circulating IL-6 as a predictor of radiation pneumonitis. Int J Radiat Oncol Biol Phys 2001;49:641–648. Anscher MS, Kong FM, Andrews K, et al. Plasma transforming growth factor beta1 as a predictor of radiation pneumonitis. Int J Radiat Oncol Biol Phys 1998;41:1029–1035. Fu XL, Huang H, Bentel G, et al. Predicting the risk of symptomatic radiation-induced lung injury using both the physical and biologic parameters V(30) and transforming growth factor beta. Int J Radiat Oncol Biol Phys 2001;50:899– 908.