Department at Stockholm Söder Hospital (1994–1998) for adjuvant post-operative RT were prospectively followed for pulmonary complications 1, 4 and 7 months after the completion of RT. Prescribed doses were 46 Gy or 50 Gy in 2 Gy fractions. The frequencies of RP requiring corticosteriods at Södersjukhuset were found to be 10 % among patients treated with loco-regional RT including the IMN.
The results also showed that irradiated lung volume was associated with RP and that the estimated risk increased if more than 30% of the ipsilateral lung volume received more than 20 Gy75. A subset of patients who were diagnosed with severe pulmonary complications needing cortisone treatment had a significant functional loss comparable to 15 years of normal ageing or the loss of 3/4 of a lung lobe compared to the patients who were asymptomatic when measured at 6 months after RT38.
In paper I radiological changes in the lung tissue after RT for BC were quantified with CT in 121 of the patients included above. The association of radiological changes with spatial dose distribution and symptomatic RP was documented. CT scans were performed at two levels through the lungs before RT and 4 months afterwards, one at a level of about 3 cm above the mamillary plane and one at an apical CT slice at the superior aspect of the clavicular head as shown in Figure 9.
The change in mean density (Hounsfield number) was measured in three different regions (entire slice, anterior half and anterior third) for each of the two levels. In Figure 10 typical density change after RT for loco-regional RT is shown.
Figure 9.
slice. In t. (A–C) Eval
the central slluation of rad
lice, three rediologic chan
egions were anges perform
analyzed. med in the ceentral and appical
Figure low de schem Of the RT an partia The p
The s illustr lung D large variat calcul system correc with 1 and el
e 10 Diagnos ensity (black) matically show
e 121 patien nd 97 for lo al mastectom patients were I. Loc brea wee II. Loc mam foss Gy/
III. Loc mam 50 G IV. Loc
mam with dos pan in relat rated for a s DVH for the
span in dos tions within lated using m uses a p ct for inhom 1 cm spacin lectron ener
stic CT scan ) to high dens wn.
nts included oco-regional my (28).
e treated wit cal RT after ast to a tota ek.
co-regional mmary lym sa supraclav /fraction, 5 co-regional mmary lym Gy at 2 Gy/
co-regional mmary, axil h photon be se of 46 Gy tive lung vo subgroup in e four differ ses to the lu each group the Therapl pencil beam mogeneities.
ng over the c rgies from 6
of a lung rea sity (white) fo
d in this stud al therapy fo th four diffe r partial mas al dose of 50 RT after pa mph nodes (I vicular, axil fractions a RT after pa mph nodes: a /fraction, 5 RT after m llary, and su eams and th
at 2 Gy per olumes irrad n Figure 11
rent RT tech ung for the p (not illustr lan system, m-like algori
The dose chest. Photo 6 MeV up to
action at arro for treatment
dy, 24 were ollowing eit erent RT tec stectomy: ta 0 Gy at 2 Gy artial mastec IMN): photo llar and IMN
week.
artial mastec as group II b fractions a w modified radi
upraclavicu he chest wal r fraction, 5 diated with t and shows hniques. Th different te rated here) w
version 5B ithm (the B planning w on energies o 16 MeV.
ow showing d technique IV
referred for ther radical chniques:
angential ph y per fractio ctomy inclu ons coverin N. Totaldos ctomy exclu but excludin
week.
ical mastect ular lymph n l with enfac fractions a the four trea
the mean cu he figure illu
echniques. H were also si B for both p Batho-powe was based on
ranged from
density chang V. Dose distri
r local-post mastectom
hotons cove on, 5 fractio uding intern ng the breast se 50 Gy at uding intern ng IMN. To tomy: the in nodes were ce electrons
week.
atment tech cumulative i ustrates the r
However, in ignificant. D papers I and er law algor n about 20 C m 4 MV up
2 ges, from ibution
operative y (69) or
ering the ons a nal
t and the 2 nal otaldose
nternal treated s. Total
hniques is ipsilateral relatively ndividual DVH was d II. This rithm) to CT slices to 8 MV
27
Figu differ Data on assessme clinically 0.
1.
2.
The defi patients centre to 28 patien diagnose had been The res symptom
• R
• T
• I No assoc confound
• C
• C
ure 11. Cu rent subgro potentially ent for RP y defined ac No RP: no r independen Mild RP: re fever) judg corticostero Moderate c
with or wit requiring c inition of m were asses o reduce the
nts (23%) de ed, and in 5 n given, the sult of the matic RP we RT techniqu Total doses Increased ag ciation was
ding factors Chemothera Concurrent
mulative D ups in study confoundin was done ccording to t
registered re nt of RT, as espiratory sy ged by the cl
oids.
omplication thout fever) corticosteroi
ild and mod ssed by a fe effect of di eveloped sy patients (4%
therapy was study sho ere associate
ues as low as 1 ge
found betw s were show
apy tamoxifen i
DVH of the y I.
ng factors w 1, 4, and 7 the followin espiratory sy
judged by t ymptoms (c linician to b ns: respirato judged by t d treatment derate reacti ew clinician
fferent obse ymptomatic
%) moderate s finished 3 owed that ed with:
6–30 Gy een lung de wn:
intake
e ipsilateral
were also col 7 months af ng criteria:
ymptoms or the clinician ough and/or be RT induc ory symptom
the clinician .
ions was sim ns working erver judgem
RP. In 23 p e RP was di to 8 weeks short-term
ensity chang
l lung volu
llected prior fter RT com r respiratory n.
r dyspnea w ed but not tr ms (cough an n to be RT in milar to the
closely tog ments in clin patients (19%
iagnosed. If before the s lung dens
ges and the f
ume showin
r to RT. A c mpletion. R y problems with or witho
reated with nd/or dyspn nduced and NCI-CTC32 gether in a nical finding
%), mild RP chemothera start of RT.
sity change
following
ng the
clinical RP was
out
nea
2. The single gs.
P was apy es and
0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70
Cumulative ipsilateral lungvolume (%)
Dose (Gy)
group I group II
group III group IV
29 In paper II, a cohort of 87 of the above 121 patients were further analyzed.
Correlations between the incidence of short-term pulmonary complications and lung-dose/volume and fractionation parameters were documented. Data from dose–volume histograms of the complete ipsilateral lung and incidence data were fitted with four different NTCP models. Each bin of the dose-volume histograms was corrected for fractionation to 2 Gy per fraction using the LQ model with a / ratio of 3 Gy. The end-points considered in this study were: symptomatic (clinical) RP, X-ray assessed lung density changes and CT assessed lung density changes.
Clinical pneumonitis was defined according to modified CTC-NCIC grades 0–2.
Patients with of grade-1 or -2 pneumonitis were grouped together in the analysis and compared with asymptomatic patients. Density changes on chest X-ray were scored according to a system originally suggested by Arriagada76 which divides the lung field into three regions: the apical-lateral, central-parahilar and basal-lateral.
The highest density grades in each region were added together to form scores ranging from 0 to 9. Total scores of 1–3 were considered to represent slight radiological pneumonitis and scores of 4–9 moderate to severe.
Density changes on chest CT were evaluated using two scan sets of the thorax: the first one was taken prior to RT for treatment planning purposes and the second 4 months after the completion of RT. Each lung was divided into three regions and density changes were scored and grouped together in the same way as described above for the evaluation of chest X-ray changes. Also in this analysis no/slight RP was compared to moderate/ severe events.
The four different NTCP models used were: the Lyman model with DVH reduced to the equivalent uniform dose (LEUD), the Logit model with DVH reduced to EUD (LOGEUD), the Mean Lung Dose (MLD) model and the Relative Seriality (RS) model. The data-fitting procedure was done using the maximum likelihood analysis, based on individual DVHs and binary incidence data (0 = no complication, 1 = complication).
Figure 12. Results for modelling with the LOGIT+EUD model. NTCP versus EUD is shown. The end-points were clinical RP (upper panel), incidence diagnosed by CT (middle panel) and X-ray assessed density changes (lower panel).
Figure 12 shows the results of modelling for the LOGIT+EUD model. The incidence data in the figure (solid symbols) have been pooled and sorted into bins of 2 or 4 Gy. The solid line shows the results of modelling and the dotted lines the two-dimensional 68% confidence interval for the NTCP curve. All the studied NTCP models fit quite accurately the endpoints considered. EUD or MLD were
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
0 5 10 15 20
NTCP
Dose (Gy)
"LOGIT+EUD MODEL"
clinical RP
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
0 5 10 15 20
NTCP
Dose (Gy) LOGIT+EUD MODEL CT incidence
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
0 5 10 15 20
NTCP
Dose (Gy) LOGIT+EUD MODEL X-ray assessed density change
31 The present study resulted in a D50 value significantly lower than many other studies, which might be explained by differences in the definition of the end-point.
In this study only early pulmonary complications were evaluated.
In a continued work to reduce RP in BC patients treated with loco-regional RT at Södersjukhuset, the use of dose-volume constraints significantly reduced post-RT radiological changes on chest X-ray77. Symptomatic pneumonitis was very rare in this study. Only one patient of 88 developed a moderate reaction. Mild reactions were detected in 6 patients. Furthermore, no relationship was found between symptomatic RP and radiological RP on chest X-ray or CT, but only one patient was diagnosed with RP, so no statistically relevant conclusion can be made.