Im. J Rndrorron Oncology Bml Phys, Vol. 18. pp. 173-180 Printed m the U.S.A. All n&s reserved.
Copyright
0360.3016190 $3.00 + .oO 0 1990 Pergamon Press plc
??Brief Communication
INFLUENCE OF RADIATION THERAPY ON THE LUNG-TISSUE IN BREAST CANCER PATIENTS: CT-ASSESSED DENSITY CHANGES AND ASSOCIATED SYMPTOMS SAMUEL ROTSTEIN, M.D.,*
INGMAR LAX, PH.D.+ AND GUNILLA SVANE, M.D.*
Karolinska Hospital,S-104 0 1 Stockholm,Sweden The relative electron density of lung tissue was measured from computer tomography (CT) slices in 33 breast cancer patients treated by various techniques of adjuvant radiotherapy. The measurements were made before radiotherapy, 3 months and 9 months after completion of radiation therapy. The changes in lung densities at 3 months and 9 months were compared to radiation induced radiological (CT) findings. In addition, subjective symptoms such as cough and dyspnoea were assessed before and after radiotherapy. It was observed that the mean of the relative electron density of lung tissue varied from 0.25 when the whole lung was considered to 0.17 when only the anterior lateral quarter of the lung was taken into account. In patients with positive radiological (CT) findings the mean lung density of the anterior lateral quarter increased 2.1 times 3 months after radiotherapy and was still increased 1.6 times 6 months later. For those patients without findings, in the CI pictures the corresponding values were 1.2 and 1.1, respectively. The standard deviation of the pixel values within the anterior lateral quarter of the lung increased 3.8 times and 3.2 times at 3 months and 9 months, respectively, in the former group, as opposed to 1.2 and 1.1 in the latter group. Thirteen patients had an increase in either cough or dyspnoea as observed 3 months after completion of radiotherapy. In eleven patients these symptoms persisted 6 months later. No significant correlation was found between radiological findings and subjective symptoms. However, when three different treatment techniques were compared among 29 patients the highest rate of radiological findings was observed in patients in which the largest lung volumes received the target dose. A tendency towards an increased rate of subjective symptoms was also found in this group. Radiotherapy,
Breast cancer, Lung complication,
Lung density.
INTRODUCTION
exist both within the lung as well as between lungs in different individuals (9, 23). As the use of adjuvant radiotherapy in breast cancer patients is gradually increasing, due to an increased acceptance of breast conserving treatments, it was considered of importance to further study the effect of radiotherapy on lung tissue in breast cancer patients. For this purpose 33 breast cancer patients, treated with various techniques for adjuvant radiotherapy, were chosen as a study group. All patients were examined by CT before radiotherapy to obtain the lung density in various parts of the lung, as this investigation was especially intended to determine the lung densities relevant for dose planning of breast cancer. CT examinations were also made at 3 months and 9 months after completion of radiotherapy. At the latter two examinations radiation induced radiological findings (CT) were also observed and compared to the changes in the density values. In addition,
induced lung complications in the treatment of breast cancer can be a major clinical problem (16-l 8, 20). In these situations the clinical manifestations are often accompanied by histological changes of the lung parenchyma (2, 12, 17). Traditionally, conventional x-ray films have been used to illustrate damages to the lung tissue caused by radiotherapy in breast cancer patients (3, 4, 8, 13, 19). However, the higher sensitivity of computer tomography (CT) to detect such lesions has led to several investigations of radiation therapy induced lung damages (12, 22). The CT scanner is the most accurate clinical instrument to map the density of lung tissue, which is necessary for an accurate calculation of the dose to lung tissue in radiotherapy of breast cancer. It has been shown, from CT studies of lung density, that considerable density variations Radiation
* Radiumhemmet, Department of Oncology. + Department of Hospital Physics. * Department of Diagnostic Radiology. Reprint requests to: Samuel Rotstein.
This study has been supported by a grant from the Cancer Society of Stockholm. Accepted for publication 19 July 1989.
173
I. .I. Radiation Oncology
174
0 Biology 0 Physics
respiratory tract symptoms (cough and dyspnoes) before and after radiotherapy were recorded by a questionnaire. METHODS
AND
January 1990. Volume
Table 1b. Percentages of irradiated breast cancer patients with positive radiological findings and increased subjective symptoms in patients equal to or more than 40 vears
MATERIALS
Patients The patient material in this study consisted of 33 previously healthy pre- and postmenopausal breast cancer patients operated either with a partial mastectomy or a modified radical mastectomy. All the patients were treated within the framework of the ongoing management program for breast cancer patients in the Stockholm-Gotland region. Ten patients were treated with a partial mastectomy and postoperative irradiation to the affected breast only (group A). Nineteen patients were treated with a modified radical mastectomy and received postoperative loco-regional radiotherapy to the chest wall and ipsilateral lymph nodes in the axilla, supraclavicular fossa, and parastemal regions (group B and group C). The 29 patients included in the three groups A, B and C were given radiotherapy with three different techniques. Mean ages for the three groups are given in Table 1a. Four patients were not eligible for any of the ongoing therapeutic trials and were thus given individualized treatment. Only one patient received chemotherapy. The treatment consisted preoperatively of 4 i.v. courses of FAC (5-fluorouracil. adriamycin, cyclofosfamide) every third week and postoperatively six courses of i.v. CMF (cyclofosfamide, metotrexate, 5-fluorouracil) were given every third week.
RadiotherapJj Patients in group A were given local radiation therapy postoperatively to the affected breast only (Fig. 1). A total dose of 50 Gy was delivered to the target volume by opposing tangential photon beams. The target volume for all patients in group B and C included the chest wall and lymph nodes in the axilla, the supraclavicular and internal mammary nodes.
I
18, Number
No. of patients Mean age (years) S.D. age (years) Positiv radiol. findings (%) 2nd, 3rd exam Increase in subjective symptoms (%) 2nd, 3rd exam
A
B
C
7 53.6 13.2 29, 29
6 58.3 10.1 50, 17
8 59.0 11.3 88. 75
50. 25 (4 pts.)
50, 25 (4 pts.)
50, 50 (8 pts.)
Group A: Partial mastectomy; photons to breast only (50 Gy). Group B: Modified radical mastectomy: electrons to chest wall. photons to regional lymph nodes (46 Gy). Group C: Modified radical mastectomy; photons to chest wall and lymph node regions (46 Gy).
All ten patients in group B were treated with an oblique electron-beam to the chest wall and an anterior %o beam to cover the parasternal nodes (Fig. 2). In nine cases (group C) the chest wall and internal mammary nodes were included in the same target volume by tangential photon beams (Fig. 3). In both B and C groups the axilla was treated by anterior and posterior 6oCo fields while the fossa supraclavicular region was treated with an anterior beam only. The total dose to the target volume was 46 Gy in groups B and C. The tangential beams used in groups A and C were given with “Co gamma rays or 6 MV photons. The electron energies given with a linear accelerator in group B ranged from 6 MeV to 10 MeV. The fractionation for all three groups was 2.0 Gy/fraction. 5 fractions a week. The dose to the chest wall was specified as the mean dose within the target area as defined by the oncologist. In group B the dose to the parasternal lymph nodes was specified at 3 cm depth. The dose in groups B and C was specified
Table la. Percentages of irradiated breast cancer patients with positive radiological findings and increased subjective symptoms
No. of patients Mean age (years) SD. age (years) Positiv radiol. findings (%) 2nd, 3rd exam Increase in subjective symptoms (%) 2nd, 3rd exam
A
B
C
10 46. 3 16, 5 30 30
10 49, 1 14, 2 40 IO
9 56, 7 12, 6 78 67
40 20 (5 pts.)
38 25 (8 pts.)
56 56 (9 pts.)
Group A: Partial mastectomy; photons to the affected breast only (50 Gy). Group B: Modified radical mastectomy; electrons to the chest wall, photons to lymph node regions (46 Gy). Group C: Modified radical mastectomy: photons to chest wall and lymph node regions (46 Gy).
Fig. 1. Typical dose plane for the tangential photon beams technique used for Stage 1 patients, group A. The broken line shows the target area.
Influence
of RT on lung-tissue
0 S. ROTSTEIN et al.
175
Lung density determination
Fig. 2. Typical dose plane for Stage II patients with thin chest wall, treated with an oblique electron beam, group B. The broken line shows the target area.
at the anterior-posterior midline to the axilla and at 3 cm depth to the fossa supraclavicular region. Two patients, operated with a modified radical mastectomy, were treated with an additional electron boost to the chest wall and received doses of 6 1 Gy. One patient was operated with a partial mastectomy followed by radiotherapy to the breast and lymph node regions in the axilla and supraclavicular fossa regions. In the latter case, 50 Gy was given to the affected breast and 48 Gy to the lymph node regions. In one case, 47 Gy was delivered to the chest wall and parasternal nodes only. The photon beam dose planning was made with a commercial system. In- homogeneity corrections were made for the lung tissue. The relative electron density was assumed for all patients to be 0.25. The outline of the lung was determined from three radiographs: anterior, lateral and oblique. This was chosen, instead of using the CT data, as the investigation was intended for the routine methods used at Radiumhemmet. The electron beam dose planning was manual at the time of this investigation.
A CT scanner was used in this investigation. The examinations were made at 120 kV and the examination time was 3 sec. All patients were examined during quiet breathing. It was shown by van Dyk et al. that for normal breathing negligible differences in the result occur between a scan time of 3 set and 26 set (23). For this scanner, a CT number of -5 12 was related to the relative attenuation coefficient for air and 0 for water. The relative electron density of the lung tissue was accordingly obtained from the relation = 1.O + 0.00 195NCT. Throughout this paper lung density is defined as the electron density relative to that of water, which is the quantity of main interest for dose planning. Numerically this is equal to the specific gravity (g cme3) within 2% for low atomic number materials (1). It has previously been shown that there exists a linear relationship between CT number and electron density for low atomic number materials with densities between 0 and 1 (7). From CT scans of a lung equivalent phantom material, the maximum error in lung density determination was estimated to be within kO.02. Each patient was examined with six CT-scans, f60 mm, +30 mm, 0 mm, -30 mm, -60 mm, and -80 mm. The scans were performed in transverse sections. The thickness of the slices were 10 mm. The zero level was located at the nipple of the remaining breast, and the plus and minus signs represent the cranial and caudal direction, respectively. A height of 14 cm was chosen to cover the height of the radiotherapy field to the chest wall. During the CT examinations the patients were positioned in the same way as for the radiotherapy, that is, on a sloping back (the patients upper torso is elevated by 15”). The arm on the tumor side was, however, kept in a more superior position during CT. With this position the chest wall was almost horizontal. Three patients could not be examined with the arms folded above the head because of pain in the scar of the axilla or limited shoulder motion. The two lowest slices (-60, -80 mm) were located below or in the level of the diaphragma in all patients. The slice at -30 mm was also below diaphragma in 22 patients. The lung densities were determined from the mean of the pixel values (CT number) within three different areas in the CT slices. The standard deviation of the pixel values was also determined. This is of interest for the detection of small localized high density regions in the lung tissue. The areas were: 1. The whole lung 2. The anterior half of the lung 3. The anterior-lateral quarter of the lung
Fig. 3. Typical dose plane for Stage II patients with thick chest wall, treated with tangential photon beams, group C. The broken line shows the target area.
The areas were manually determined by using the cursor on the viewing consol. The anterior half of the lung was determined as the lung region anterior of a transversal line corresponding to the middle of the lung hilus. The anterior-lateral region of the lung had the same lateral limit but correspond medially to the contralateral edge of the sternum. In all cases a margin of 0.5- 1.O cm was ex-
176
I. J. Radiation Oncology 0 Biology 0 Physics
chided towards the pleura. The big blood vessels in the hilus region were also excluded. In each region the mean of the CT numbers and the standard deviation was noted. RadiologicalJindings A second and a third CT examination were made at 3 months and 9 months, respectively, after the radiotherapy was completed. At these examinations the lung density was determined from measurements within the anteriorlateral region of the lung, that is, that part of the lung which was mainly included in the radiotherapy beams. The levels of the CT slices at the second and third examinations were the same as in the first examination, except that scans below the diaphragma were excluded. The mean of the CT numbers as well as the standard deviation were registered within the region. The CT pictures were examined for radiotherapy induced findings in the lung by an experienced diagnostic radiologist. Questionnaire Subjective symptoms were assessed using a questionnaire which was retrospectively mailed to all 33 participating patients 9 months after the radiotherapy was completed. The questionnaire was designed to record respiratory symptoms (cough and dyspnoea) before radiotherapy, 3 months and 9 months after completion of radiotherapy. All patients were asked to mark each symptom on a four level scale (none, mild, moderate or severe). Symptoms before radiation therapy were compared to symptoms 3 and 9 months after radiation. RESULTS Lung density The result of the lung density investigation from the pre radiotherapy scan is shown in Table 2. The mean values of the lung density are shown separately for the three most cranial slice locations. The standard deviation was calculated for the 33 patients in each group, shown in Table 2. Rudiologicalfindings and subjective symptoms The results of the radiological findings and variations in symptoms are individually shown in Table 3a for the 29 patients in the three radiotherapy groups A, B and C described in Methods and Materials. In Tables 1a and 1b Table 2. Mean values of lung densities a of the lung ant-lateral part Mean density Slice + 60 mm Slice + 30 mm Slice 0
0, 17 0, 17 0, 15
January 1990, Volume 18, Number 1 the results are summarized
for the three groups. In Table 1a all 29 patients are included and in Table 1b only those patients with an age equal to or more than 40 years. Both Table 3a and 1a indicate that positive radiological findings are more frequent in group C than in the other two groups. However, the mean age is also higher. To find out whether the positive findings could be explained by the ages of patients, all patients less than 40 years of age were excluded (Table 1b). The positive radiological findings, however, remained much more frequent in group C. The four patients given individualized treatment are presented in Table 3b. The location in the lungs of the radiological findings for the two groups B and C are schematically shown in Figure 4. The size of these findings, seen in the transverse slice, was in no case less than 2 cm. In general the findings were seen in more than one slice. For each patient the mean lung density and the standard deviation of the pixel values was determined within the anterior-lateral quarter of the lung, before radiotherapy and 3 months and 9 months after radiotherapy. Figure 5 shows the relative mean value of the mean density for three different patient groups before radiotherapy, 3 months and 9 months after radiotherapy. The relative density is given separately for the total number of patients, those with radiological findings and those patients without radiological findings (normal). The number of slices in each group is given in the Figure. In patients with positive radiological (CT) findings the mean lung density of the anterior lateral quarter increased 2.1 * 0.14 (Mean + SE) times 3 months after radiotherapy and was till increased 1.6 + 0.16 times, 6 months later. For those patients with normal lungs, in the CT pictures the corresponding values were 1.2 f 0.04 and 1.1 ? 0.05, respectively. The differences between the mean lung density values among those with radiological findings and those without were statistically significant both at 3 and 9 months after radiotherapy (p < 0.001). Figure 6 shows the mean values of the standard deviation for the same three patient groups as in Figure 5. The standard deviation of the pixel values within the anterior lateral quarter of the lung increased 3.8 k 0.42 times and 3.2 -t 0.28 times at 3 months and 9 months, respectively, in the former group, as opposed to 1.2 f 0.05 and 1.1 + 0.06 in the latter group. The differences between patients with and without radiological findings were highly significant @ < 0.001).
and standard
deviations
for 33 patients
Whole lung
4 of the lung ant part SD 0, 04 0, 04 0, 04
Mean density 0, 19 0, 19 0, 20
Note: Slice level zero is at the height of the nipple of the remaining zero.
SD
Mean density
SD
0, 04 0, 04 0, 06
0, 23 0, 25 0, 27
0, 04 0, 04 0, 06
breast. The other two slices are located cranial of slice level
Influence
of RT on lung-tissue
Table 3a. Radiological B.
Radiological findings
Age
2nd ex
3rd ex
41 49 43 32 19 37 69 75 47 51
N N Y N N Y Y N N N
Y Y N N N Y N N N N
findings and changes in subjective Breast 46 Gy, electron
Subjective symptoms* 2nd ex
3rd ex
t
t
+ +
+ _
t
t
t _ _ t _ t
t _ _ t _ t
symptoms C. Breast Parasternal
beam
Parasternal Fossa suprac. 46 Gy Axilla I
A. Breast 50 Gy, tangential photon beams
177
0 S. ROTSTEINetal.
46 Gy, tang. beams I
Fx;;,aasuprac’
46 Gy I
Radiological findings
Subjective symptoms*
Age
2nd ex
3rd ex
2nd ex
3rd ex
59 32 36 71 63 39 34 44 49 64
Y Y N N Y N N N N Y
N N N N Y N N N N N
_
_ _ _
+ t
t
+ t
+ _ t _
+
Radiological findings
Subjective symptoms*
Age
2nd ex
3rd ex
2nd ex
63 38 43 67 61 67 40 62 69
Y N Y N Y Y Y Y Y
Y N Y Y N Y N Y Y
3rd ex
_
_
+ _ _
+ _ _
+ + + _
+ + + _
+
+
+
* + increase, - no change or decrease. t Patient refused to answer the questionnaire. Y = Yes: N = No.
Table 3b. Radiological
Age 76 75 43
findings and changes in subjective
Radiotherapy
* + increase,
in patients
Radiological
findings
given individualized
radiation
Subjective
therapy
symptoms*
2nd ex.
3rd ex.
2nd ex.
3rd ex.
Yes
Yes
+
+
Yes
Yes
+
+
No
No
_
_
No
+
+
Breast electrons, 46 Gy Boost, breast em, 15 Gy Breast photons, 46 Gy Boost, breast em, 15 Gy Breast photons, 50 Gy Axill, fossa supracl, 48 Gy Breast electrons, 47 Gy Parasternal photons, 47 Gy
f**\ ( II) 65
symptoms
Yes
- no change or decrease.
*0
*81 \
‘\ .
Fig. 4. Schematic illustration of the location of the radiological findings in the lung for groups B and C. For illustration purposes, all findings are drawn in the left lung. Circles denote findings in group D, (compare to Fig. 2) and stars denote findings in group C, (compare to Fig. 3).
A total of 33 questionnaires were dispatched to the participating patients to record changes in symptoms 3 and 9 months after start of radiotherapy. At completion of the study 82% (27) of the questionnaires had been returned. As can be seen from Tables 3a and 3b, 13 patients had increased respiratory tract symptoms 3 months after completion of radiotherapy. Six months later I 1 patients still experienced an increased symptomatology compared to before therapy. In three patients, however, the symptoms were milder than at 3 months.
DISCUSSION
Lung density It is now well recognized that the electron density of lung tissue relative to that of water is a crucial factor for
178
I. J. Radiation Oncology 0 Biology 0 Physics
2.5
i ‘!
0.19 and 0.17 (Table 2). This is in good agreement with van Dyk ef al. who reported 0.18 for the anterior 5 cm of the lung (2 1) and Rosenblum et al. (15) who reported a density difference of 0.12 between the anterior and posterior parts. It is thus important to observe that a lung density of 0.17 should be used for most dose planning of mammary carcinoma and not the mean value for the whole lung volume of 0.25.
Before radlotherapy 3-m
@%I 9mc4lthsafter
January 1990, Volume 18, Number 1
T
Radiologicalfindings
Total
Pos. Findings
Normal
Fig. 5. Mean of the relative lung density within the anteriorlateral quarter of the lung. Standard error is shown on top of each pile. Values are given for the total number of slices, those with radiological findings and those with normal lungs. The
number of slices examined is given in the figure.
accurate dose planning in the radiotherapy for breast cancer. A thorough investigation of the lung density was made by van Dyk et uf. (23). They reported a mean lung density of 0.25 for a cross section at the height of the greatest AP lung diameter. This was obtained during normal breathing in patients with a mean age of 55 years. This is the same mean age as in the present patient material. From Table 2 it can be seen that the mean density for the whole lung volume including all three slice locations is 0.25. The difference in lung density for the three slice locations is not significant (15, 23). For the anterior half and the anterior-lateral quarter of the lung area the mean values of the lung densities are
Total
Pos.
Findings
Normal
Fig. 6. Standard deviation of the pixel values within the anteriorlateral quarter of the lung. Standard error is shown on top of each pile. Values are given for the total number of slices, those with radiological findings and those with normal lungs. The number of slices examined are shown in the figure.
The investigation of radiological findings by CT was not designed to cover the entire lung volume treated by radiotherapy. The distance between the slices was 30 mm and the slice thickness was 10 mm. This method was thus not excepted to find all radiological changes, but a representative sample. However, as mentioned above no finding was less than 2 cm in the transversal slice and in general the findings were seen in more than one slice. As can be seen from Table 1b the difference in percentages of radiological findings between groups B and C is striking especially at 9 months in spite of similar age distributions. target volumes, and target doses. The distribution between right and leftside treatments were approximately the same in the two treatment groups as were the radiological findings. In explaining the difference between groups B and C the dose distributions and absolute dose levels for the two treatment techniques should be considered, as it is today generally accepted that the relative biological effectiveness for high energy photons and electrons is the same (5). The accuracy in the specified dose is different for the tangential photon beams technique (groups A and C) and the electron beam technique (group B). Regarding the former groups, it has been shown by Kniiijs et al. (6) that the photon beam dose planning algorithm used overestimates the dose to the target by about 5% when tangential beams are used. This is mainly due to the reduced volume of scattering material. The same could also be expected for the lung tissue. However, the dose planning in the present investigation was made with a smaller beam height than the one actually used in the treatments in order to balance the mentioned overestimation of the dose. In this study a standard lung density of 0.25 was used instead of the more realistic 0.17, which underestimated the lung dose by about 2%. The increased range of Compton and pair production electrons in the lung tissue compared to the reference medium, water, is not accounted for in the photon beam dose planning. This effect will result in a less steep dose gradient in the penumbra region than what is shown in the dose plane (8). In the manual electron beam dose planning, used for the group B patients, the multiple scattering is disregarded. The penetration of the electrons into the lung tissue is thus in many cases overestimated (10). The general effect of this is that the area of high dose levels is overestimated
Influence of RT on lung-tissue 0 S. ROTSTEINd al.
and areas of low dose levels are underestimated. The dose distribution is more “smeared” out than what is shown in the dose plane. The magnitude of this effect is difficult to quantify as it depends on the anatomy of the individual patient. It can thus be concluded that the dose to the lung in the high dose region is overestimated in the dose planning for both electron beam technique and the tangential photon beams technique. According to Mah et al. (12) the incidence of acute radiation induced pulmonary damage to fractionated irradiation can be predicted fairly well by the estimated single dose (ED) (24). The biological endpoint used by Mah et al. was the same as used in this investigation (i.e. increase in lung density as observed by a diagnostic radiologist on CT). Using this model, it is evident that only the high dose region within the lung is relevant for the incidence of pulmonary damage. For the present fractionation scheme (groups B and C) a 50% incidence will be at the 78% relative dose level and the 80% incidence will be at the 97% relative dose level. From Figures 2 and 3, which show typical dose planes for the two radiotherapy groups B and C, it is evident that the lung volume in group B (Fig. 2), with an incidence probability of more than 50% (78% isodose), is small. It involves essentially the junction region between the electron beam and 6oCo beam. In group C (Fig. 3) the lung volume, with an incidence probability of more than 50%, covers essentially a quarter of the lung. Also, the lung volume with an incidence probability of more than 75% (95% isodose) is large. The photon beam technique used was chosen to treat the target volume adequately. Other photon beam techniques presently in use to treat the chest wall and the homolateral parasternal nodes mostly include small volumes of underdosage within the target volume. These techniques, however, will probably result in less extensive radiological changes ( 14). A comparison with the results of Mah et al. can be made with the present results. In Table 4 an approximate average dose in the high dose region of the lung is given for the three techniques A, B and C. Corresponding ED values have been calculated from the number of fractionations and the average doses. Given ED values, the incidence of radiological findings can be extracted from Mah rt al. The corresponding figures from the present study are also given in Table 4. Taking into account the small
Table 4. Technique
A
B
C
Average lung dose ED Radio]. findings from Mahela/.(lI) Radio]. findings this study
50 Gy 12.1
39 Gy 9.8
46 Gy 11.5
88% 50%
48% 40%
83% 90%
179
number of patients in the three groups and the corresponding statistical uncertainty, the results seem to be in agreement for groups B and C. For group A the difference is, however, higher. The location of the radiological findings in groups B and C, schematically shown in Figure 4, supports the relation given by Mah et al., that is, the findings are located in the high dose region ( 12). In all cases, except two, these changes were confined to the anterior-lateral quarter of the lung (Fig. 4). In two cases the findings extended into the hilus region. These two patients received an electron beam boost to the tumor bed. Note that only one finding is seen in group B (10 patients) at the junction region between the electron beam and the 6oCo beam. In this area a local overdosage, sometimes up to 30% in mean dose within 2 cm2, is always seen in the dose plan; this has been confirmed by phantom measurements. It thus appears that a volume effect exists for small volumes such that a small volume, given a high dose, will give less biological effects compared to a larger volume. For larger volumes, however, Mah et al. did not observe any volume effect for the endpoint studied here. As can be seen from Figure 5 the mean value of the lung density within the anterior-lateral quarter of the lung shows a significant increase for the group of patients with positive findings compared to the normal group. This difference between the two groups is pronounced in Figure 6, which shows the standard deviation of the pixel values for the same quarter of the lung. From the information given by Figures 5 and 6 the conclusion can be made that the findings are localized and have a higher density than normal lung tissue. For the normal group, the CT picture of lung tissue after radiotherapy is thus very smooth. No conclusion can be drawn on the number of localized findings in each examined lung. However, the results of this study show that the increased pixel values correspond to a finding confined to a limited area. During the evaluation of the CT scans, the standard deviation of pixel values was found to be a valuable tool to detect lung abnormalities induced by radiotherapy. In fact, some of the findings were primarily detected from an increase in the standard deviations. Mah and van Dyke (11) have reported a relative density in the lung tissue for patients with radiological findings of 1.20 & 0.10 after radiotherapy. This value is much lower than the results of this study (Fig. 5). There are several factors that might explain this difference: First, a difference in the lung volume irradiated. In their study only 12 of 54 patients were treated for breast cancer, and almost half of the patients were treated for a primary lung cancer. The irradiated lung volume could thus be expected to be larger. A lower relative density will follow ifthe ratio between the area (in the CT-slice) of the radiological finding to the irradiated lung area decreases. A second factor might be that the average lung dose for all the patients in the high dose region was less than in the present study.
180
1. J. Radiation Oncology
0 Biology 0 Physics
A third factor might also be the anatomical site of the primary tumor and thus the difference in the part of the lung that was irradiated.
Subjective symptoms Thirteen out of 27 patients experienced increased respiratory tract symptoms 3 months after completion of radiotherapy. Most of these complains were of minor importance (mild) but in six patients symptoms increased from none or mild before start of radiotherapy to moderate or severe. Three patients had persisting symptoms of moderate degree 6 months later. Altogether, 11 patients complained of some sort of respiratory inconvenience 9 months after radiotherapy. As can be seen from Tables 3a and 3b there was no clear relation between radiological findings and subjective symptoms; this is in agreement with earlier reports (12). There is, however. a tendency for more symptoms to occur among those patients who had the largest lung volumes irradiated, but in only one case, which had previously been treated with pre- and
January
1990, Volume
18, Number
postoperative chemotherapy, some form of treatment.
1
did these symptoms
require
CONCLUSIONS The relative electron density of the whole lung volume is found to be 0.25, and for the anterior-lateral quarter, 0.17. This latter lung density is recommended for radiotherapy dose planning of breast cancer if a standard density is used. Radiation induced radiological findings are shown to appear in lung volume given a high dose. The number of patients in this study is too small to be conclusive regarding the best choice of radiotherapy technique to minimize lung damage. However, it appears that radiotherapy of the chest wall by electron beams compared to tangential photon beams is preferable with regard to radiological findings using the aforementioned techniques. It also appears that the percentage of increase in subjective respiratory symptoms is not worse for the electron beam treated group. compared to the tangential beam group.
REFERENCES 1. Brooks, R.; Michell, L.; O’Connor, C.; Di Chiro. G. On the relationship between computed tomography numbers and specific gravity. Phys. Med. Biol. 26: 14 1- 147; I98 1. 2. Gross, N. Pulmonary effects of radiation therapy. Ann. Int. Med. 86:8 l-92; 1977. 3. Hagen, S.; Kolbenstvedt. A. Radiologic pulmonary changes following cobalt 60 treatment of mammary carcinoma. Acta Radiol. Ther. 11:386-94; 1972. 4. Hietala, S. 0.; Hahn, P. Pulmonary radiation reaction in the treatment of carcinoma of the breast. Radiography 42: 225-30; 1976. Commission on Radiation Units and 5. ICRU (International Measurements). Radiation dosimetry: electron beams with energies between 1 and 50 MeV. Report 35. Bethesda, Maryland: ICRU; 1984. 6. KnoBs, T.; Ahlgren, L.; Nilsson, M. Comparison of measured and calculated absorbed doses from tangential irradiation of the breast. Radiother. Oncol. 7:8 I-88: 1986. 7. Knobs, T.; Nilsson, M.; Ah&en, L. A method for conversion of Hounsfield number to electron density and predication of macroscopic pair production crossections. Radiother. Oncol. 5:337-345: 1986. 8. Kornelsen, R.; Young, M. Changes in the dose-profile of a 10 MV X-ray beam within and beyond low density material. Med. Phys. 9:114-l 16; 1982. 9. Lagrange, J. L.; Brassard, N.; Costa, A.; Auband, D.; H&y, M.; Bruneton, J. L.; Lalanne, C. M. CT measurement of lung density: The role of patient positioning and value for total body irradiation. Int. J. Radiat. Oncol. Biol. Phys. 13: 941-944; 1987. 10. Lax, I. Development of a generalized Gaussian model for absorbed dose calculation and dose planning in therapeutic electron beams. Thesis, University of Stockholm: 1986. 11. Mah, K.; Van Dyk, J. Quantitative measurement of changes in human lung density following irradiation. Radiother. Oncol. 11:169-179; 1988. 12. Mah, K.; Van Dyk, J.; Keeane, T.; Poon, P. Acute radiationinduced pulmonary damage: a clinical study on the response to fractionated radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 13:179-188; 1987.
13. Oppeldal, T.; Kolbenstvedt, A. Pulmonary contraction following 6OCo irradiation of mammary carcinoma. Acta Radiol. Ther. 15:329-36; 1976. 14. Roberson, P. L.; Lichter, A. S.; Bodner. A.; Fredrickson. H. A.; Padikal, T. N.; Kelly, B. A.; Van de Geijin, J. Dose to lung in primary breast irradiation. Int. J. Radiat. Oncol. Biol. Phys. 9:97-102; 1982. L.: Mauceri, R.: Wellenstein, D.; Bassano, D.; 15. Rosenblum, Cohen, W.; Heitzman, E. Computed tomography of the lung. Radiol. 129:521-524; 1978. 16. Roswit, P.; Whise, D. C. Severe radiation injures of the lung. Am. J. Roentgenol. 129:127-136; 1977. 17. Rothwell, R. 1.: Kelly, S. A.: Joslin, C. A. F. Radiation pneumonitis in patients treated for breast cancer. Radiother. Oncol. 4:9- 14: 1985. 18. Rubin, P.; Casarett, G. Respiratory system. In: Clinical radiation pathology, Chapter 12, Philadelphia: W. B. Saunders; 1968:423-470. 19. Scheier, M.; Rhomberg, W.: Schelling, F. Verlnderungen ins Thoraxrontgenbild nach postoperative Bestrahlvag von Mamma Karzinomen. Strahlenter. Oncol. 163: 103- 108; 1987. Amer. Rev. Res. Dis. 87: 20. Smith, C. Radiation pneumonitis. 647-654; 1963. 21. Van Dyk, J.; Battista, J.; Rider, W. Half body radiotherapy: the use of computed tomography to determine the dose to lung. Int. J. Radiat. Oncol. Biol. Phys. 6:463-470: 1980. 22. Van Dyk, J.; Keane, T. J.; Kan, S.; Rider, W. D.; Fryer, C. J. H. Radiation pneumonitis following large single dose irradiation: a reevaluation based on absolute dose to lung. Int. J. Radiat. Oncol. Biol. Phys. 7:461-467; 1981. 23. Van Dyk, J.; Keane, T. J.; Rider, W. D. Lung density as measured by computerized tomography: implications for radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 8: 1363- 1372; 1982. 24. Wara, W.; Phillips, T.; Margolis, L.; Smith, V. Radiation pneumonitis: a new approach to the derivation oftime-dose factors. Cancer 32:547-553; 1973.
Copyright
0360.3016190 $3.00 + .oO 0 1990 Pergamon Press plc
??Brief Communication
INFLUENCE OF RADIATION THERAPY ON THE LUNG-TISSUE IN BREAST CANCER PATIENTS: CT-ASSESSED DENSITY CHANGES AND ASSOCIATED SYMPTOMS SAMUEL ROTSTEIN, M.D.,*
INGMAR LAX, PH.D.+ AND GUNILLA SVANE, M.D.*
Karolinska Hospital,S-104 0 1 Stockholm,Sweden The relative electron density of lung tissue was measured from computer tomography (CT) slices in 33 breast cancer patients treated by various techniques of adjuvant radiotherapy. The measurements were made before radiotherapy, 3 months and 9 months after completion of radiation therapy. The changes in lung densities at 3 months and 9 months were compared to radiation induced radiological (CT) findings. In addition, subjective symptoms such as cough and dyspnoea were assessed before and after radiotherapy. It was observed that the mean of the relative electron density of lung tissue varied from 0.25 when the whole lung was considered to 0.17 when only the anterior lateral quarter of the lung was taken into account. In patients with positive radiological (CT) findings the mean lung density of the anterior lateral quarter increased 2.1 times 3 months after radiotherapy and was still increased 1.6 times 6 months later. For those patients without findings, in the CI pictures the corresponding values were 1.2 and 1.1, respectively. The standard deviation of the pixel values within the anterior lateral quarter of the lung increased 3.8 times and 3.2 times at 3 months and 9 months, respectively, in the former group, as opposed to 1.2 and 1.1 in the latter group. Thirteen patients had an increase in either cough or dyspnoea as observed 3 months after completion of radiotherapy. In eleven patients these symptoms persisted 6 months later. No significant correlation was found between radiological findings and subjective symptoms. However, when three different treatment techniques were compared among 29 patients the highest rate of radiological findings was observed in patients in which the largest lung volumes received the target dose. A tendency towards an increased rate of subjective symptoms was also found in this group. Radiotherapy,
Breast cancer, Lung complication,
Lung density.
INTRODUCTION
exist both within the lung as well as between lungs in different individuals (9, 23). As the use of adjuvant radiotherapy in breast cancer patients is gradually increasing, due to an increased acceptance of breast conserving treatments, it was considered of importance to further study the effect of radiotherapy on lung tissue in breast cancer patients. For this purpose 33 breast cancer patients, treated with various techniques for adjuvant radiotherapy, were chosen as a study group. All patients were examined by CT before radiotherapy to obtain the lung density in various parts of the lung, as this investigation was especially intended to determine the lung densities relevant for dose planning of breast cancer. CT examinations were also made at 3 months and 9 months after completion of radiotherapy. At the latter two examinations radiation induced radiological findings (CT) were also observed and compared to the changes in the density values. In addition,
induced lung complications in the treatment of breast cancer can be a major clinical problem (16-l 8, 20). In these situations the clinical manifestations are often accompanied by histological changes of the lung parenchyma (2, 12, 17). Traditionally, conventional x-ray films have been used to illustrate damages to the lung tissue caused by radiotherapy in breast cancer patients (3, 4, 8, 13, 19). However, the higher sensitivity of computer tomography (CT) to detect such lesions has led to several investigations of radiation therapy induced lung damages (12, 22). The CT scanner is the most accurate clinical instrument to map the density of lung tissue, which is necessary for an accurate calculation of the dose to lung tissue in radiotherapy of breast cancer. It has been shown, from CT studies of lung density, that considerable density variations Radiation
* Radiumhemmet, Department of Oncology. + Department of Hospital Physics. * Department of Diagnostic Radiology. Reprint requests to: Samuel Rotstein.
This study has been supported by a grant from the Cancer Society of Stockholm. Accepted for publication 19 July 1989.
173
I. .I. Radiation Oncology
174
0 Biology 0 Physics
respiratory tract symptoms (cough and dyspnoes) before and after radiotherapy were recorded by a questionnaire. METHODS
AND
January 1990. Volume
Table 1b. Percentages of irradiated breast cancer patients with positive radiological findings and increased subjective symptoms in patients equal to or more than 40 vears
MATERIALS
Patients The patient material in this study consisted of 33 previously healthy pre- and postmenopausal breast cancer patients operated either with a partial mastectomy or a modified radical mastectomy. All the patients were treated within the framework of the ongoing management program for breast cancer patients in the Stockholm-Gotland region. Ten patients were treated with a partial mastectomy and postoperative irradiation to the affected breast only (group A). Nineteen patients were treated with a modified radical mastectomy and received postoperative loco-regional radiotherapy to the chest wall and ipsilateral lymph nodes in the axilla, supraclavicular fossa, and parastemal regions (group B and group C). The 29 patients included in the three groups A, B and C were given radiotherapy with three different techniques. Mean ages for the three groups are given in Table 1a. Four patients were not eligible for any of the ongoing therapeutic trials and were thus given individualized treatment. Only one patient received chemotherapy. The treatment consisted preoperatively of 4 i.v. courses of FAC (5-fluorouracil. adriamycin, cyclofosfamide) every third week and postoperatively six courses of i.v. CMF (cyclofosfamide, metotrexate, 5-fluorouracil) were given every third week.
RadiotherapJj Patients in group A were given local radiation therapy postoperatively to the affected breast only (Fig. 1). A total dose of 50 Gy was delivered to the target volume by opposing tangential photon beams. The target volume for all patients in group B and C included the chest wall and lymph nodes in the axilla, the supraclavicular and internal mammary nodes.
I
18, Number
No. of patients Mean age (years) S.D. age (years) Positiv radiol. findings (%) 2nd, 3rd exam Increase in subjective symptoms (%) 2nd, 3rd exam
A
B
C
7 53.6 13.2 29, 29
6 58.3 10.1 50, 17
8 59.0 11.3 88. 75
50. 25 (4 pts.)
50, 25 (4 pts.)
50, 50 (8 pts.)
Group A: Partial mastectomy; photons to breast only (50 Gy). Group B: Modified radical mastectomy: electrons to chest wall. photons to regional lymph nodes (46 Gy). Group C: Modified radical mastectomy; photons to chest wall and lymph node regions (46 Gy).
All ten patients in group B were treated with an oblique electron-beam to the chest wall and an anterior %o beam to cover the parasternal nodes (Fig. 2). In nine cases (group C) the chest wall and internal mammary nodes were included in the same target volume by tangential photon beams (Fig. 3). In both B and C groups the axilla was treated by anterior and posterior 6oCo fields while the fossa supraclavicular region was treated with an anterior beam only. The total dose to the target volume was 46 Gy in groups B and C. The tangential beams used in groups A and C were given with “Co gamma rays or 6 MV photons. The electron energies given with a linear accelerator in group B ranged from 6 MeV to 10 MeV. The fractionation for all three groups was 2.0 Gy/fraction. 5 fractions a week. The dose to the chest wall was specified as the mean dose within the target area as defined by the oncologist. In group B the dose to the parasternal lymph nodes was specified at 3 cm depth. The dose in groups B and C was specified
Table la. Percentages of irradiated breast cancer patients with positive radiological findings and increased subjective symptoms
No. of patients Mean age (years) SD. age (years) Positiv radiol. findings (%) 2nd, 3rd exam Increase in subjective symptoms (%) 2nd, 3rd exam
A
B
C
10 46. 3 16, 5 30 30
10 49, 1 14, 2 40 IO
9 56, 7 12, 6 78 67
40 20 (5 pts.)
38 25 (8 pts.)
56 56 (9 pts.)
Group A: Partial mastectomy; photons to the affected breast only (50 Gy). Group B: Modified radical mastectomy; electrons to the chest wall, photons to lymph node regions (46 Gy). Group C: Modified radical mastectomy: photons to chest wall and lymph node regions (46 Gy).
Fig. 1. Typical dose plane for the tangential photon beams technique used for Stage 1 patients, group A. The broken line shows the target area.
Influence
of RT on lung-tissue
0 S. ROTSTEIN et al.
175
Lung density determination
Fig. 2. Typical dose plane for Stage II patients with thin chest wall, treated with an oblique electron beam, group B. The broken line shows the target area.
at the anterior-posterior midline to the axilla and at 3 cm depth to the fossa supraclavicular region. Two patients, operated with a modified radical mastectomy, were treated with an additional electron boost to the chest wall and received doses of 6 1 Gy. One patient was operated with a partial mastectomy followed by radiotherapy to the breast and lymph node regions in the axilla and supraclavicular fossa regions. In the latter case, 50 Gy was given to the affected breast and 48 Gy to the lymph node regions. In one case, 47 Gy was delivered to the chest wall and parasternal nodes only. The photon beam dose planning was made with a commercial system. In- homogeneity corrections were made for the lung tissue. The relative electron density was assumed for all patients to be 0.25. The outline of the lung was determined from three radiographs: anterior, lateral and oblique. This was chosen, instead of using the CT data, as the investigation was intended for the routine methods used at Radiumhemmet. The electron beam dose planning was manual at the time of this investigation.
A CT scanner was used in this investigation. The examinations were made at 120 kV and the examination time was 3 sec. All patients were examined during quiet breathing. It was shown by van Dyk et al. that for normal breathing negligible differences in the result occur between a scan time of 3 set and 26 set (23). For this scanner, a CT number of -5 12 was related to the relative attenuation coefficient for air and 0 for water. The relative electron density of the lung tissue was accordingly obtained from the relation = 1.O + 0.00 195NCT. Throughout this paper lung density is defined as the electron density relative to that of water, which is the quantity of main interest for dose planning. Numerically this is equal to the specific gravity (g cme3) within 2% for low atomic number materials (1). It has previously been shown that there exists a linear relationship between CT number and electron density for low atomic number materials with densities between 0 and 1 (7). From CT scans of a lung equivalent phantom material, the maximum error in lung density determination was estimated to be within kO.02. Each patient was examined with six CT-scans, f60 mm, +30 mm, 0 mm, -30 mm, -60 mm, and -80 mm. The scans were performed in transverse sections. The thickness of the slices were 10 mm. The zero level was located at the nipple of the remaining breast, and the plus and minus signs represent the cranial and caudal direction, respectively. A height of 14 cm was chosen to cover the height of the radiotherapy field to the chest wall. During the CT examinations the patients were positioned in the same way as for the radiotherapy, that is, on a sloping back (the patients upper torso is elevated by 15”). The arm on the tumor side was, however, kept in a more superior position during CT. With this position the chest wall was almost horizontal. Three patients could not be examined with the arms folded above the head because of pain in the scar of the axilla or limited shoulder motion. The two lowest slices (-60, -80 mm) were located below or in the level of the diaphragma in all patients. The slice at -30 mm was also below diaphragma in 22 patients. The lung densities were determined from the mean of the pixel values (CT number) within three different areas in the CT slices. The standard deviation of the pixel values was also determined. This is of interest for the detection of small localized high density regions in the lung tissue. The areas were: 1. The whole lung 2. The anterior half of the lung 3. The anterior-lateral quarter of the lung
Fig. 3. Typical dose plane for Stage II patients with thick chest wall, treated with tangential photon beams, group C. The broken line shows the target area.
The areas were manually determined by using the cursor on the viewing consol. The anterior half of the lung was determined as the lung region anterior of a transversal line corresponding to the middle of the lung hilus. The anterior-lateral region of the lung had the same lateral limit but correspond medially to the contralateral edge of the sternum. In all cases a margin of 0.5- 1.O cm was ex-
176
I. J. Radiation Oncology 0 Biology 0 Physics
chided towards the pleura. The big blood vessels in the hilus region were also excluded. In each region the mean of the CT numbers and the standard deviation was noted. RadiologicalJindings A second and a third CT examination were made at 3 months and 9 months, respectively, after the radiotherapy was completed. At these examinations the lung density was determined from measurements within the anteriorlateral region of the lung, that is, that part of the lung which was mainly included in the radiotherapy beams. The levels of the CT slices at the second and third examinations were the same as in the first examination, except that scans below the diaphragma were excluded. The mean of the CT numbers as well as the standard deviation were registered within the region. The CT pictures were examined for radiotherapy induced findings in the lung by an experienced diagnostic radiologist. Questionnaire Subjective symptoms were assessed using a questionnaire which was retrospectively mailed to all 33 participating patients 9 months after the radiotherapy was completed. The questionnaire was designed to record respiratory symptoms (cough and dyspnoea) before radiotherapy, 3 months and 9 months after completion of radiotherapy. All patients were asked to mark each symptom on a four level scale (none, mild, moderate or severe). Symptoms before radiation therapy were compared to symptoms 3 and 9 months after radiation. RESULTS Lung density The result of the lung density investigation from the pre radiotherapy scan is shown in Table 2. The mean values of the lung density are shown separately for the three most cranial slice locations. The standard deviation was calculated for the 33 patients in each group, shown in Table 2. Rudiologicalfindings and subjective symptoms The results of the radiological findings and variations in symptoms are individually shown in Table 3a for the 29 patients in the three radiotherapy groups A, B and C described in Methods and Materials. In Tables 1a and 1b Table 2. Mean values of lung densities a of the lung ant-lateral part Mean density Slice + 60 mm Slice + 30 mm Slice 0
0, 17 0, 17 0, 15
January 1990, Volume 18, Number 1 the results are summarized
for the three groups. In Table 1a all 29 patients are included and in Table 1b only those patients with an age equal to or more than 40 years. Both Table 3a and 1a indicate that positive radiological findings are more frequent in group C than in the other two groups. However, the mean age is also higher. To find out whether the positive findings could be explained by the ages of patients, all patients less than 40 years of age were excluded (Table 1b). The positive radiological findings, however, remained much more frequent in group C. The four patients given individualized treatment are presented in Table 3b. The location in the lungs of the radiological findings for the two groups B and C are schematically shown in Figure 4. The size of these findings, seen in the transverse slice, was in no case less than 2 cm. In general the findings were seen in more than one slice. For each patient the mean lung density and the standard deviation of the pixel values was determined within the anterior-lateral quarter of the lung, before radiotherapy and 3 months and 9 months after radiotherapy. Figure 5 shows the relative mean value of the mean density for three different patient groups before radiotherapy, 3 months and 9 months after radiotherapy. The relative density is given separately for the total number of patients, those with radiological findings and those patients without radiological findings (normal). The number of slices in each group is given in the Figure. In patients with positive radiological (CT) findings the mean lung density of the anterior lateral quarter increased 2.1 * 0.14 (Mean + SE) times 3 months after radiotherapy and was till increased 1.6 + 0.16 times, 6 months later. For those patients with normal lungs, in the CT pictures the corresponding values were 1.2 f 0.04 and 1.1 ? 0.05, respectively. The differences between the mean lung density values among those with radiological findings and those without were statistically significant both at 3 and 9 months after radiotherapy (p < 0.001). Figure 6 shows the mean values of the standard deviation for the same three patient groups as in Figure 5. The standard deviation of the pixel values within the anterior lateral quarter of the lung increased 3.8 k 0.42 times and 3.2 -t 0.28 times at 3 months and 9 months, respectively, in the former group, as opposed to 1.2 f 0.05 and 1.1 + 0.06 in the latter group. The differences between patients with and without radiological findings were highly significant @ < 0.001).
and standard
deviations
for 33 patients
Whole lung
4 of the lung ant part SD 0, 04 0, 04 0, 04
Mean density 0, 19 0, 19 0, 20
Note: Slice level zero is at the height of the nipple of the remaining zero.
SD
Mean density
SD
0, 04 0, 04 0, 06
0, 23 0, 25 0, 27
0, 04 0, 04 0, 06
breast. The other two slices are located cranial of slice level
Influence
of RT on lung-tissue
Table 3a. Radiological B.
Radiological findings
Age
2nd ex
3rd ex
41 49 43 32 19 37 69 75 47 51
N N Y N N Y Y N N N
Y Y N N N Y N N N N
findings and changes in subjective Breast 46 Gy, electron
Subjective symptoms* 2nd ex
3rd ex
t
t
+ +
+ _
t
t
t _ _ t _ t
t _ _ t _ t
symptoms C. Breast Parasternal
beam
Parasternal Fossa suprac. 46 Gy Axilla I
A. Breast 50 Gy, tangential photon beams
177
0 S. ROTSTEINetal.
46 Gy, tang. beams I
Fx;;,aasuprac’
46 Gy I
Radiological findings
Subjective symptoms*
Age
2nd ex
3rd ex
2nd ex
3rd ex
59 32 36 71 63 39 34 44 49 64
Y Y N N Y N N N N Y
N N N N Y N N N N N
_
_ _ _
+ t
t
+ t
+ _ t _
+
Radiological findings
Subjective symptoms*
Age
2nd ex
3rd ex
2nd ex
63 38 43 67 61 67 40 62 69
Y N Y N Y Y Y Y Y
Y N Y Y N Y N Y Y
3rd ex
_
_
+ _ _
+ _ _
+ + + _
+ + + _
+
+
+
* + increase, - no change or decrease. t Patient refused to answer the questionnaire. Y = Yes: N = No.
Table 3b. Radiological
Age 76 75 43
findings and changes in subjective
Radiotherapy
* + increase,
in patients
Radiological
findings
given individualized
radiation
Subjective
therapy
symptoms*
2nd ex.
3rd ex.
2nd ex.
3rd ex.
Yes
Yes
+
+
Yes
Yes
+
+
No
No
_
_
No
+
+
Breast electrons, 46 Gy Boost, breast em, 15 Gy Breast photons, 46 Gy Boost, breast em, 15 Gy Breast photons, 50 Gy Axill, fossa supracl, 48 Gy Breast electrons, 47 Gy Parasternal photons, 47 Gy
f**\ ( II) 65
symptoms
Yes
- no change or decrease.
*0
*81 \
‘\ .
Fig. 4. Schematic illustration of the location of the radiological findings in the lung for groups B and C. For illustration purposes, all findings are drawn in the left lung. Circles denote findings in group D, (compare to Fig. 2) and stars denote findings in group C, (compare to Fig. 3).
A total of 33 questionnaires were dispatched to the participating patients to record changes in symptoms 3 and 9 months after start of radiotherapy. At completion of the study 82% (27) of the questionnaires had been returned. As can be seen from Tables 3a and 3b, 13 patients had increased respiratory tract symptoms 3 months after completion of radiotherapy. Six months later I 1 patients still experienced an increased symptomatology compared to before therapy. In three patients, however, the symptoms were milder than at 3 months.
DISCUSSION
Lung density It is now well recognized that the electron density of lung tissue relative to that of water is a crucial factor for
178
I. J. Radiation Oncology 0 Biology 0 Physics
2.5
i ‘!
0.19 and 0.17 (Table 2). This is in good agreement with van Dyk ef al. who reported 0.18 for the anterior 5 cm of the lung (2 1) and Rosenblum et al. (15) who reported a density difference of 0.12 between the anterior and posterior parts. It is thus important to observe that a lung density of 0.17 should be used for most dose planning of mammary carcinoma and not the mean value for the whole lung volume of 0.25.
Before radlotherapy 3-m
@%I 9mc4lthsafter
January 1990, Volume 18, Number 1
T
Radiologicalfindings
Total
Pos. Findings
Normal
Fig. 5. Mean of the relative lung density within the anteriorlateral quarter of the lung. Standard error is shown on top of each pile. Values are given for the total number of slices, those with radiological findings and those with normal lungs. The
number of slices examined is given in the figure.
accurate dose planning in the radiotherapy for breast cancer. A thorough investigation of the lung density was made by van Dyk et uf. (23). They reported a mean lung density of 0.25 for a cross section at the height of the greatest AP lung diameter. This was obtained during normal breathing in patients with a mean age of 55 years. This is the same mean age as in the present patient material. From Table 2 it can be seen that the mean density for the whole lung volume including all three slice locations is 0.25. The difference in lung density for the three slice locations is not significant (15, 23). For the anterior half and the anterior-lateral quarter of the lung area the mean values of the lung densities are
Total
Pos.
Findings
Normal
Fig. 6. Standard deviation of the pixel values within the anteriorlateral quarter of the lung. Standard error is shown on top of each pile. Values are given for the total number of slices, those with radiological findings and those with normal lungs. The number of slices examined are shown in the figure.
The investigation of radiological findings by CT was not designed to cover the entire lung volume treated by radiotherapy. The distance between the slices was 30 mm and the slice thickness was 10 mm. This method was thus not excepted to find all radiological changes, but a representative sample. However, as mentioned above no finding was less than 2 cm in the transversal slice and in general the findings were seen in more than one slice. As can be seen from Table 1b the difference in percentages of radiological findings between groups B and C is striking especially at 9 months in spite of similar age distributions. target volumes, and target doses. The distribution between right and leftside treatments were approximately the same in the two treatment groups as were the radiological findings. In explaining the difference between groups B and C the dose distributions and absolute dose levels for the two treatment techniques should be considered, as it is today generally accepted that the relative biological effectiveness for high energy photons and electrons is the same (5). The accuracy in the specified dose is different for the tangential photon beams technique (groups A and C) and the electron beam technique (group B). Regarding the former groups, it has been shown by Kniiijs et al. (6) that the photon beam dose planning algorithm used overestimates the dose to the target by about 5% when tangential beams are used. This is mainly due to the reduced volume of scattering material. The same could also be expected for the lung tissue. However, the dose planning in the present investigation was made with a smaller beam height than the one actually used in the treatments in order to balance the mentioned overestimation of the dose. In this study a standard lung density of 0.25 was used instead of the more realistic 0.17, which underestimated the lung dose by about 2%. The increased range of Compton and pair production electrons in the lung tissue compared to the reference medium, water, is not accounted for in the photon beam dose planning. This effect will result in a less steep dose gradient in the penumbra region than what is shown in the dose plane (8). In the manual electron beam dose planning, used for the group B patients, the multiple scattering is disregarded. The penetration of the electrons into the lung tissue is thus in many cases overestimated (10). The general effect of this is that the area of high dose levels is overestimated
Influence of RT on lung-tissue 0 S. ROTSTEINd al.
and areas of low dose levels are underestimated. The dose distribution is more “smeared” out than what is shown in the dose plane. The magnitude of this effect is difficult to quantify as it depends on the anatomy of the individual patient. It can thus be concluded that the dose to the lung in the high dose region is overestimated in the dose planning for both electron beam technique and the tangential photon beams technique. According to Mah et al. (12) the incidence of acute radiation induced pulmonary damage to fractionated irradiation can be predicted fairly well by the estimated single dose (ED) (24). The biological endpoint used by Mah et al. was the same as used in this investigation (i.e. increase in lung density as observed by a diagnostic radiologist on CT). Using this model, it is evident that only the high dose region within the lung is relevant for the incidence of pulmonary damage. For the present fractionation scheme (groups B and C) a 50% incidence will be at the 78% relative dose level and the 80% incidence will be at the 97% relative dose level. From Figures 2 and 3, which show typical dose planes for the two radiotherapy groups B and C, it is evident that the lung volume in group B (Fig. 2), with an incidence probability of more than 50% (78% isodose), is small. It involves essentially the junction region between the electron beam and 6oCo beam. In group C (Fig. 3) the lung volume, with an incidence probability of more than 50%, covers essentially a quarter of the lung. Also, the lung volume with an incidence probability of more than 75% (95% isodose) is large. The photon beam technique used was chosen to treat the target volume adequately. Other photon beam techniques presently in use to treat the chest wall and the homolateral parasternal nodes mostly include small volumes of underdosage within the target volume. These techniques, however, will probably result in less extensive radiological changes ( 14). A comparison with the results of Mah et al. can be made with the present results. In Table 4 an approximate average dose in the high dose region of the lung is given for the three techniques A, B and C. Corresponding ED values have been calculated from the number of fractionations and the average doses. Given ED values, the incidence of radiological findings can be extracted from Mah rt al. The corresponding figures from the present study are also given in Table 4. Taking into account the small
Table 4. Technique
A
B
C
Average lung dose ED Radio]. findings from Mahela/.(lI) Radio]. findings this study
50 Gy 12.1
39 Gy 9.8
46 Gy 11.5
88% 50%
48% 40%
83% 90%
179
number of patients in the three groups and the corresponding statistical uncertainty, the results seem to be in agreement for groups B and C. For group A the difference is, however, higher. The location of the radiological findings in groups B and C, schematically shown in Figure 4, supports the relation given by Mah et al., that is, the findings are located in the high dose region ( 12). In all cases, except two, these changes were confined to the anterior-lateral quarter of the lung (Fig. 4). In two cases the findings extended into the hilus region. These two patients received an electron beam boost to the tumor bed. Note that only one finding is seen in group B (10 patients) at the junction region between the electron beam and the 6oCo beam. In this area a local overdosage, sometimes up to 30% in mean dose within 2 cm2, is always seen in the dose plan; this has been confirmed by phantom measurements. It thus appears that a volume effect exists for small volumes such that a small volume, given a high dose, will give less biological effects compared to a larger volume. For larger volumes, however, Mah et al. did not observe any volume effect for the endpoint studied here. As can be seen from Figure 5 the mean value of the lung density within the anterior-lateral quarter of the lung shows a significant increase for the group of patients with positive findings compared to the normal group. This difference between the two groups is pronounced in Figure 6, which shows the standard deviation of the pixel values for the same quarter of the lung. From the information given by Figures 5 and 6 the conclusion can be made that the findings are localized and have a higher density than normal lung tissue. For the normal group, the CT picture of lung tissue after radiotherapy is thus very smooth. No conclusion can be drawn on the number of localized findings in each examined lung. However, the results of this study show that the increased pixel values correspond to a finding confined to a limited area. During the evaluation of the CT scans, the standard deviation of pixel values was found to be a valuable tool to detect lung abnormalities induced by radiotherapy. In fact, some of the findings were primarily detected from an increase in the standard deviations. Mah and van Dyke (11) have reported a relative density in the lung tissue for patients with radiological findings of 1.20 & 0.10 after radiotherapy. This value is much lower than the results of this study (Fig. 5). There are several factors that might explain this difference: First, a difference in the lung volume irradiated. In their study only 12 of 54 patients were treated for breast cancer, and almost half of the patients were treated for a primary lung cancer. The irradiated lung volume could thus be expected to be larger. A lower relative density will follow ifthe ratio between the area (in the CT-slice) of the radiological finding to the irradiated lung area decreases. A second factor might be that the average lung dose for all the patients in the high dose region was less than in the present study.
180
1. J. Radiation Oncology
0 Biology 0 Physics
A third factor might also be the anatomical site of the primary tumor and thus the difference in the part of the lung that was irradiated.
Subjective symptoms Thirteen out of 27 patients experienced increased respiratory tract symptoms 3 months after completion of radiotherapy. Most of these complains were of minor importance (mild) but in six patients symptoms increased from none or mild before start of radiotherapy to moderate or severe. Three patients had persisting symptoms of moderate degree 6 months later. Altogether, 11 patients complained of some sort of respiratory inconvenience 9 months after radiotherapy. As can be seen from Tables 3a and 3b there was no clear relation between radiological findings and subjective symptoms; this is in agreement with earlier reports (12). There is, however. a tendency for more symptoms to occur among those patients who had the largest lung volumes irradiated, but in only one case, which had previously been treated with pre- and
January
1990, Volume
18, Number
postoperative chemotherapy, some form of treatment.
1
did these symptoms
require
CONCLUSIONS The relative electron density of the whole lung volume is found to be 0.25, and for the anterior-lateral quarter, 0.17. This latter lung density is recommended for radiotherapy dose planning of breast cancer if a standard density is used. Radiation induced radiological findings are shown to appear in lung volume given a high dose. The number of patients in this study is too small to be conclusive regarding the best choice of radiotherapy technique to minimize lung damage. However, it appears that radiotherapy of the chest wall by electron beams compared to tangential photon beams is preferable with regard to radiological findings using the aforementioned techniques. It also appears that the percentage of increase in subjective respiratory symptoms is not worse for the electron beam treated group. compared to the tangential beam group.
REFERENCES 1. Brooks, R.; Michell, L.; O’Connor, C.; Di Chiro. G. On the relationship between computed tomography numbers and specific gravity. Phys. Med. Biol. 26: 14 1- 147; I98 1. 2. Gross, N. Pulmonary effects of radiation therapy. Ann. Int. Med. 86:8 l-92; 1977. 3. Hagen, S.; Kolbenstvedt. A. Radiologic pulmonary changes following cobalt 60 treatment of mammary carcinoma. Acta Radiol. Ther. 11:386-94; 1972. 4. Hietala, S. 0.; Hahn, P. Pulmonary radiation reaction in the treatment of carcinoma of the breast. Radiography 42: 225-30; 1976. Commission on Radiation Units and 5. ICRU (International Measurements). Radiation dosimetry: electron beams with energies between 1 and 50 MeV. Report 35. Bethesda, Maryland: ICRU; 1984. 6. KnoBs, T.; Ahlgren, L.; Nilsson, M. Comparison of measured and calculated absorbed doses from tangential irradiation of the breast. Radiother. Oncol. 7:8 I-88: 1986. 7. Knobs, T.; Nilsson, M.; Ah&en, L. A method for conversion of Hounsfield number to electron density and predication of macroscopic pair production crossections. Radiother. Oncol. 5:337-345: 1986. 8. Kornelsen, R.; Young, M. Changes in the dose-profile of a 10 MV X-ray beam within and beyond low density material. Med. Phys. 9:114-l 16; 1982. 9. Lagrange, J. L.; Brassard, N.; Costa, A.; Auband, D.; H&y, M.; Bruneton, J. L.; Lalanne, C. M. CT measurement of lung density: The role of patient positioning and value for total body irradiation. Int. J. Radiat. Oncol. Biol. Phys. 13: 941-944; 1987. 10. Lax, I. Development of a generalized Gaussian model for absorbed dose calculation and dose planning in therapeutic electron beams. Thesis, University of Stockholm: 1986. 11. Mah, K.; Van Dyk, J. Quantitative measurement of changes in human lung density following irradiation. Radiother. Oncol. 11:169-179; 1988. 12. Mah, K.; Van Dyk, J.; Keeane, T.; Poon, P. Acute radiationinduced pulmonary damage: a clinical study on the response to fractionated radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 13:179-188; 1987.
13. Oppeldal, T.; Kolbenstvedt, A. Pulmonary contraction following 6OCo irradiation of mammary carcinoma. Acta Radiol. Ther. 15:329-36; 1976. 14. Roberson, P. L.; Lichter, A. S.; Bodner. A.; Fredrickson. H. A.; Padikal, T. N.; Kelly, B. A.; Van de Geijin, J. Dose to lung in primary breast irradiation. Int. J. Radiat. Oncol. Biol. Phys. 9:97-102; 1982. L.: Mauceri, R.: Wellenstein, D.; Bassano, D.; 15. Rosenblum, Cohen, W.; Heitzman, E. Computed tomography of the lung. Radiol. 129:521-524; 1978. 16. Roswit, P.; Whise, D. C. Severe radiation injures of the lung. Am. J. Roentgenol. 129:127-136; 1977. 17. Rothwell, R. 1.: Kelly, S. A.: Joslin, C. A. F. Radiation pneumonitis in patients treated for breast cancer. Radiother. Oncol. 4:9- 14: 1985. 18. Rubin, P.; Casarett, G. Respiratory system. In: Clinical radiation pathology, Chapter 12, Philadelphia: W. B. Saunders; 1968:423-470. 19. Scheier, M.; Rhomberg, W.: Schelling, F. Verlnderungen ins Thoraxrontgenbild nach postoperative Bestrahlvag von Mamma Karzinomen. Strahlenter. Oncol. 163: 103- 108; 1987. Amer. Rev. Res. Dis. 87: 20. Smith, C. Radiation pneumonitis. 647-654; 1963. 21. Van Dyk, J.; Battista, J.; Rider, W. Half body radiotherapy: the use of computed tomography to determine the dose to lung. Int. J. Radiat. Oncol. Biol. Phys. 6:463-470: 1980. 22. Van Dyk, J.; Keane, T. J.; Kan, S.; Rider, W. D.; Fryer, C. J. H. Radiation pneumonitis following large single dose irradiation: a reevaluation based on absolute dose to lung. Int. J. Radiat. Oncol. Biol. Phys. 7:461-467; 1981. 23. Van Dyk, J.; Keane, T. J.; Rider, W. D. Lung density as measured by computerized tomography: implications for radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 8: 1363- 1372; 1982. 24. Wara, W.; Phillips, T.; Margolis, L.; Smith, V. Radiation pneumonitis: a new approach to the derivation oftime-dose factors. Cancer 32:547-553; 1973.
