High dose rate intraluminal radiotherapy for carcinoma of the bronchus: outcome of treatment of 406 patients

High dose rate intraluminal radiotherapy for carcinoma of the bronchus: outcome of treatment of 406 patients

ADIOTHEI:IAPY #'ONCOLOGY ELSEVIER Radiotherapy and Oncology 33 (1994) 31-40 High dose rate intraluminal radiotherapy for carcinoma of the bronchus: ...

919KB Sizes 10 Downloads 21 Views

ADIOTHEI:IAPY #'ONCOLOGY ELSEVIER

Radiotherapy and Oncology 33 (1994) 31-40

High dose rate intraluminal radiotherapy for carcinoma of the bronchus: outcome of treatment of 406 patients S.W. Gollins *a, P.A. Burt a, P.V. Barber b, R.

Stout a

aDepartment of Radiotherapy, The Christie Hospital, Withingtan, Manchester M20 9BX, UK bDepartment of Respiratory Medicine, WythenshaweHospital, Manchester M23 9LT, UK Received 19 August 1994; accepted 15 September 1994

Abstract

In April 1988 the Christie Hospital started using the microSelectron-HDR machine to deliver intraluminal radiotherapy (ILT) to inoperable bronchial carcinomas causing symptoms due to endobronchial disease. Results of treatment in the first 406 patients with primary non-small-cell carcinoma are presented. Three main categories of patient were defined. Category 1 consisted of 324 patients (79.8%) who were previously unirradiated and received a single fraction of ILT as their primary treatment, mostly to a dose of 1500 cGy (76%) or 2000 cGy (23%) at 1 cm from the centre of the iridium-192 treatment source. The percentage of these patients whose symptoms or signs were improved at 6 weeks following ILT were as follows: stridor 92%, haemoptysis 88%, cough 62%, dyspnoea, 60%, pain, 50% and pulmonary collapse, 46%. Approximately two-thirds of these patients (67.30) derived long lasting palliation and required no further treatment during their lifetime. The other third of patients needed subsequent treatment at some stage because of recurrence of their symptoms and in this situation external beam radiotherapy (EB) or a repeat ILT treatment was effectively utilised. Category 2 consisted of 65 patients (16%) who had previously received EB but required ILT when their turnout recurred. At 6 weeks post-ILT levels of symptom palliation were broadly similar to those obtained if ILT was used in previously unirradiated individuals, although the improvement was not so well sustained with time and only 7% showed improvement in pulmonary collapse at 6 weeks. Category 3 consisted of 17 patients (4.2%) in whom ILT was used concurrently with EB as a combined initial treatment. Similar levels of palliation were seen when compared with patients who received a single ILT treatment only. Overall, ILT was well tolerated in terms of early and late morbidity. In conclusion, the efficiency of a single ILT treatment in palliating symptoms due to endobronchial tumour in previously unirradiated individuals is comparable with that reported in series where treatment for advanced lung cancer combines a prolonged course of EB concurrently with several ILT treatments.

Keywords: Bronchus; Carcinoma; Intraluminal radiotherapy; Endobronchial brachytherapy; Single fraction; Palliation

1. Introduction Up to 75% of patients with carcinoma of the lung have inoperable disease at presentation and still more have unresectable disease at thoracotomy. Up to 50% of patients develop symptoms and signs due to endobronchial disease, the main ones being haemoptysis, cough, dyspnoea and pulmonary collapse. Recently intraluminal radiotherapy (ILT) using endobronchial high dose rate brachytherapy with, for example, the microSelectron-HDR machine, has received increasing attention. In terms of palliation, response rates using ILT [1,2,4,5, 8,9,11-13,15,17,18] can be comparable with (or greater than, in the case of pulmonary collapse) those obtained using external beam radiotherapy (EB) [6,7,10,14].

Reported ILT regimes often use multiple ILT treatments in addition to a concurrent course of EB. In April, 1988 the Christie Hospital in Manchester started using ILT to palliate symptoms from endobronchial disease in lung cancer using the microSelectronH D R machine. When ILT was used as a primary treatment in previously unirradiated patients, a single fraction was usually given without the concurrent use of EB. This paper is a retrospective audit of the results of using ILT at the Christie Hospital between April 1988 and the end of 1992. 2.

Methods

In order to assess the extent of disease, patients had posteroanterior, lateral and penetrated anteroposterior

* Corresponding author. 0167-8140/94/$07.00 © 1994 Elsevier Science Ireland Ltd. All rights reserved SSDI 0167-8140(94)01462-C

32

S.W. Gollins et al./Radiother. Oncol. 33 (1994) 31-40

chest radiographs. Patients who had been assessed with a view to possible surgery had mediastinal CT scans carried out but these were not routinely done as part of an intensive staging procedure. Further investigations such as abdominal ultrasound scan or isotope bone scan were usually done if clinical symptoms or serum profile aroused suspicion. Patients were treated with ILT on a day-case basis using the microSelectron-HDR machine. A fibreoptic bronchoscopy was used under local anaesthesia to place the 2-mm diameter treatment applicator alongside the tumour. It was then advanced past the tumour until resistance was reached more peripherally in the lung, thus anchoring the applicator. No additional endobronchial device was used to fix the applicator in position. The applicator was then connected to the remote afterloading machine for a period of approximately 15 min whilst the iridium-192 source was transferred to the applicator. The treatment dose was prescribed at a distance of I cm from the central axis of the source. Details of the method have been described elsewhere [17]. At the time of first ILT treatment the presence of stridor, dyspnoea, haemoptysis, cough, pain and pulmonary collapse was recorded. In addition, the Medical Research Council respiratory performance scale was used to assess pulmonary function at the time of treatment. In a large proportion of patients who had pain at presentation, this was pleuritic in nature and associated with a distal obstructive pneumonitis rather than direct invasion by the tumour of thoracic structures. When patients were seen at follow-up it was found that a simple and effective assessment was for a judgement to be made by the doctor and patient as to whether the presenting symptoms and signs were improved, unchanged or worse. For the purpose of this study attempts were not made to score the degree of improvement or deterioration in an individual or make further assessments according to the MRC respiratory performance scale. 3. Patients

Between April 1988 and the end of 1992, 448 patients with inoperable endobronchial or endotracheal tumours Table 1 The number and percentage of patients with the various types of histological diagnoses for the 322 patients in whom the pathological diagnosis of non-small-cell lung cancer was available Histological/cytological diagnosis

Number

Squamous cell carcinoma Large cell carcinoma Adenoearcinoma Non-small-cell carcinoma

280 8 13 21

Total

322

Percentage 87.0 2.5 4.0 6.5

were treated with ILT. They were deemed inoperable because of the advanced state of the primary tumour, the presence of metastases or the patients general condition, age or pulmonary function. The following subgroups (some overlapping) are not considered further (numbers of patients are shown in parentheses): endobronchial metastases from an extrathoracic primary tumour (17), small cell carcinomas (12), carcinomas solely confined to the trachea (9), adenoid cystic carcinomas (5) and carcinoid tumours (3). The remaining 406 patients with primary bronchial carcinomas, the majority of whom had biopsy proven non-small-cell carcinomas, form the basis of this paper. The male:female ratio was 3.46:1 with an average age of 68.3 years (male) and 70.0 years (female). Histological confirmation was available in 322 (79.3%) of the 406 patients, the majority (87%) being squamous cell carcinomas (Table 1). Of the remaining 84 patients (20.7%o), the clinical, bronchoscopic and radiological presentation suggested non-small-cell carcinoma, although this was not histologically confirmed due to bronchoseopic biopsies or brushings not containing viable tumour tissue. Initial MRC respiratory performance status at the time of first ILT treatment showed fairly even distribution amongst patients as demonstrated in Table 2. At presentation the majority of patients had UICC Stage II or III disease but there were 33 patients (8.1%) with known Stage IV (TNM M1) disease. 4. Treatment

4.1. Treatment categories Three main categories of treatment could be defined for the 406 patients (Table 3). Category 1 consisted of 324 patients (79.8%) who were previously unirradiated and who received a single fraction of ILT as their initial treatment. Of these 324 patients, 218 (Group A) received Table 2 Medical Research Council Respiratory Performance Scale used to assess patients at the time of their first ILT treatment Description

Score

Climbs hills and stairs without dyspnoea Walks any distance at normal pace on the flat without dyspnoea Walks more than 90 m (100 yards) at own speed without dyspnoea Dyspnoea on walking 90 m Dyspnoea on mild exertion

0

53

19.6

1

41

15.1

2

58

21.4

3 4

57 62

21.0 22.9

Total 100

Information was available on 271 patients.

Number

271

Percentage

100

33

S. W. Gollins et al. / Radiother. Oncol. 33 (1994) 31-40

Table 3 Numbers and percentages of patients in the three main treatment categories subdivided into separate mutually exclusive treatment combinations, Groups A - E Treatment

Group

Treatment regime

Number

Percentage

Mean gap (range)

A B

Single ILT treatment only Initial ILT followed by a subsequent treatment, either EB or second ILT or EB plus second ILT

218

53.7

91 l0 5

22.4 2.5 1.2

233 (33-1185) 355 (42-1085) 224 (22-609)

Initial EB prior to a single ILT treatment Initial EB prior to first ILT followed by subsequent second ILT

62

15.3

366 (63-1953)

3

0.7

EB and single ILT given as a combined initial treatment

17

4.2

category

1

2

C D

3

E

Total

406

100

Mean gaps between treatments in days (with range) also illustrated.

no further radiotherapy during their lifetime whilst the other 106 patients (Group B) subsequently needed an additional treatment with either EB (in the majority of cases) or repeat ILT if initial ILT failed to control symptoms, or on symptomatic recurrence of their tumour. Category 2 consisted of 65 patients (16%) who had received EB as their initial treatment and were treated with ILT for symptomatic disease recurrence or failure of EB to control symptoms (Groups C and D). Finally, Category 3 consisted of 17 patients (4.2%) who received treatment initially with both ILT and EB as a combined modality treatment (Group E). ILT was not used alone as a primary treatment in patients with bulky tumours whose symptoms were dysphagla or pain because of tumour mass or infiltration of thoracic structures and who did not have any symptoms due to endobronchial tumour. In such cases EB was the preferred radiotherapeutic method. 4.2. Treatment prior to first I L T treatment Ten patients (2.5%) had received some form of chemotherapy and 7 patients (1.7%) laser treatment in the area of subsequent ILT. The average gap between previous laser treatment and ILT was 76 clays (range, 5-330 days). Details of previous EB treatment are illustrated in Table 4. The majority of patients had received EB with palliative intent before the disease recurrence that required ILT. 4.3. Details of first I L T treatment All 406 patients received their first ILT as a single fraction. Table 5 illustrates the frequency with which various sites in the bronchial tree were treated, showing

that right and left main bronchi were the most commonly treated sites. Twenty-four patients (5.9%) received ILT treatment to two tumours in different sites, requiring two endobronchial applicators rather than the usual one. The median length of bronchus treated via the first catheter was 90 mm (range, 40-150 mm). If a second catheter was used in addition, the median length of bronchus treated via this was 70 mm (range, 40-120 ram). The dose used was prescribed at a distance of 1 cm perpendicular to the central axis of the source. Table 6 illustrates the frequency with which various doses were used at first ILT treatment in the first and second catheters. Overall, 79.6% of patients received a dose of Table 4 Details of EB treatment in the 65 patients who received initial EB and required subsequent ILT because of tumour recurrence (Category 2) Number of fractions

Mean treatment volume (cm 3)

Mean tumour dose (cGy)

Number of patients

Percentage

I

--

1090

6

2

--

1600

I

1.5

4

--

2025

I0

15.4

8

9.2

--

2870

41

63.1

10 Total

--

3000

2 60

3.1 92.3

15 16 25 Total

655 163 720

4860 5500 5000

3 1 i 5

4.6 1.5 1.5 7.7

The space separates treatments given with palliative intent (above the space) from treatments given with radical intent. In practice, treatments using 4 or 8 fractions were usually prescribed to a maximum subcutaneous dose of 2250 cGy and 3000 cGy, rvspectively and given as a parallel opposed pair to a surface field size of 90-150 crn 2.

34

S.W. Gollins et al. / Radiother. Oncol. 33 (1994) 31-40

Table 5 The number and percentage of patients whose first ILT treatment ineluded the various sites shown

tent treating a volume of 522 cm to a dose of 5000 cGy in 16 fractions using a wedge pair technique.

Site

4.4. Treatment subsequent to first I L T treatment

Percentage of patients

Trachea Carina Right main bronchus Right upper lobe Right intermediate bronchus Right middle lobe Right lower lobe Left main bronchus Left upper lobe Left lingula lobe Left lower lobe

5.9 9.1 29.9 18.0 18.8 2.7 10.4 24.6 13.3 2.7 8.7

Total

144.1

Some treatments included more than one contiguous site. All patients illustrated who had some involvement of the trachea also had involvement of the carina or main bronchi.

1500 cGy with 18.5% receiving 2000 cGy. In the 324 previously unirradiated patients who received a single ILT treatment as their initial treatment (Category 1), 74 patients (22.8%) received 2000 cGy, 246 patients (75.9%) received 1500 cGy and 4 patients (1.2%) received 1250 cGy or 1000 cGy at 1 cm. Seventeen patients (Category 3) received EB concurrently with a single fraction of ILT as part of a combined initial treatment, usually to tumours with a large extrabronchial component causing pain or dysphagia due to tumour bulk in addition to symptoms due to endobronchial tumour. The EB treatment for 16 of these patients was given with palliative intent via parallel opposed pairs treating surface field sizes of 90-150 cm, 15 patients receiving a maximum subcutaneous dose of 3000 cGy in 8 daily fractions and one patient receiving a maximum subcutaneous dose of 2250 cGy in 4 daily fractions. The other patient received EB with radical in-

Ninety-six patients in Group B who were initially treated with ILT, later needed a course of EB because of failure of the initial ILT treatment to control symptoms, or because of symptomatic recurrence of their tumour. The majority (90 patients, 94%) received a palliative course of EB using 8 fractions (39 patients), 4 fractions (27 patients), or one fraction (23 patients) to cord tolerance. One patient received only 2 of his intended 8 fractions as treatment had to be curtailed because he sustained a pathological fracture of his femur necessitating transfer to another hospital for internal fixation. Six patients (6%) received a radical course of EB after a single fraction of ILT had re-expanded a segment of collapsed lung, thus enabling a small, radically treatable target volume to be defined. An average treatment volume of 512 cm was used, in 4 patients treating to 4750 cGy in 15 fractions and in 2 patients treating to 5000 cGy in 16 fractions. Eighteen patients (15 in Group B and 3 in Group D) received a second ILT treatment in the same location as first ILT treatment because of failure of the initial ILT treatment to control symptoms or because of later tumour recurrence. The median length of bronchus treated under these circumstances was 80 mm (range, 40-120 mm). Overall, doses were lower than at first ILT treatment, with 11 patients (61%) receiving 1000 cGy and 7 patients (39%) 1500 cGy at 1 cm. 5. R ~

5.1. Palliation of symptoms The presence or absence of seven main symptoms/signs was recorded at the time of first ILT treat-

Table 6 The overall frequency with which various doses were used in the first ILT treatment First applicator

Second applicator

Dose at I cra (cGy)

Number of patients

1000 1250 1400 1500 1850 2000

3 3 1 323 l 75

Total

406

Percentage 0.7 0.7 0.25 79.6 0.25 18.5 100

Dose at 1 cm (cGy)

Number of patients

Percentage

1000 1250

1 1

4.2 4.2

1500

20

83.3

2000

2

8.3

24

100

Doses are expressed in cGy at 1 cra perpendicular to the central axis of the source. Twenty-four patients had two tumours in non-contiguous locations treated simultaneously, requiring a second applicator as shown.

S. W. Go~/ins et al. / Radiother. Oncol. 33 (1994) 31-40 100

80

~

toms at the time of first ILT treatment were similar to the overall frequencies in the 406 patients (not shown). Symptoms present at the time of ILT were reassessed at 1.5, 4 and 12 months after treatment and recorded as being improved, unchanged or worse. Of the patients who initially had particular symptoms at the time of ILT, an average of 73% of surviving patients had recorded in their notes on follow-up whether the symptoms were unchanged, improved or worse. Fig. 2 shows the results for the overall group of 406 patients. At 6 weeks following treatment a high percentage of patients had their symptoms improved. The most readily definable endpoint was haemoptysis and in 89% of patients this symptom resolved. The number of assessable patients surviving to 4 and then 12 months fell off sharply, as can be seen from Fig. 2, but even at 12 months a considerable percentage of the surviving patients maintained symptomatic palliation. To demonstrate the palliative efficiency of a single ILT treatment in previously unirradiated patients, the 324 patients in Groups A and B, whose initial treatment consisted of a single fraction of ILT, are shown in Fig. 3. Again, high percentages of patients having their symptoms improved are apparent at 6 weeks following ILT treatment. In decreasing order of frequency improvements were seen in the following percentages of patients: stridor 92%, haemoptysis 88%, cough 62%, dyspnoea 60%, pain 50% and pulmonary collapse 46%. Improvement in collapse in this instance indicates reexpansion of a previously collapsed segment, lobe or whole lung on serial chest radiographs. Relief of pain

7(; a

58.9

60 *4

¢

¢0 ro

40

20 65 1.5

!

2

4

,3

5

G

7

l:stridov 2 dVspTzo~q 3:h~Pmnply~i~ 4-cnlt~h 5 ~ p a ~ E:collapse 7=dyspl ag~a

Fig. i. The overall percentage of the 406 patients with the various symptoms shown present at the time of the first ILT treatment.

ment (Fig. 1). The presenting features in descending order of frequency were dyspnoea, cough, haemoptysis, a degree of pulmonary collapse, pain, stridor and dysphagia. Within Groups A - E the presenting syrup-

1.5 m o n t h s

88.9

after first

ILT

4 months

after first

12 m o n t h s

ILT

:? o

89.O

35

,oo

/I//

84.4

,////,

~, 60

so.z

6

6o.4

a f t e r f i r s t ILT

100

76.5

60 .HI, ////,

54.5

"/i///, 40

40

20

20

0 ~tr~vors

1

2

3

4

5

6

1

2

3

4

5

6

1

2

3

4,

5

6

18

244

191

2f7

41

140

10

I12

128

150

30

89

1

53

51

51

11

36

l=stridor

2=dyspnoea

3=haevr~optysis 4 = c o u g h 5 = p a i n

6=collapse

Fig. 2. The whole group of 406 patients. Assessments were made at i.5, 4 and 12 months after first ILT treatment and the presenting symptoms judged improved, unchanged or worse. The histograms represent the percentage of assessed surviving patients who initially had the symptom shown, in whom this symptom was judged to have improved. The number of survivors who were assessed to derive each percentage shown is represented underneath the histograms.

S.W. Gollins et al./ Radiother. OncoL 33 (1994) 31-40

36

1.5 m o n t h s

after first

4 months

ILT

1oo

100"

12 m o n t h s

after first ILT

100

~

Ioo

92.3 88.4

87.9

80

80`

80

60`

60

a f t e r f i r s t ILT

100

N

80.0

62. 5

6f.6

60 50,0 o

.................

4o

:::::::::::::::: 4 0 . 9

:~:~:~:+:~

40"

!iiii+iii 20

0 sure.ors

1

2

3

13

201

I47

40

:JiA

++::++

zo

4

5

6

1

172

28

183

7

l=stridor

2=dyspnoea

i

iN)+ii

z0`

?.::?:.?~ (>

2

3

4

5

6

144

99

183

22

78

3=haemoptysis

1

2

3

4

5

6

50

45

46

8

35

4=cough 5=pailz 6=collapse

Fig. 3. The 324 previously unirradiated patients in Groups A and B (79.8% of the total 406) who initially had a single ILT treatment. The histograms represent the percentage of assessed surviving patients who initially had the symptom shown, in whom this symptom was judged to have improved. The number of survivors who were assessed to derive each percentage shown is represented underneath the histograms.

which was pleuritic in nature nearly always correlated with relief of a distal pneumonitis. In patients who survived to 4 and then 12 months, the percentage with improvement in their symptoms was reasonably similar to that seen at 6 weeks, with no symptom showing more than a third fall with time in terms of the degree of palliation achieved. In patients who survived to 12 months improvement in presenting symptoms was maintained in the following percentages of patients: haemoptysis 80%, pain 63%, dyspnoea 50%, cough 39% and pulmonary collapse 37%. Of the 324 patients who had a single ILT treatment as their initial treatment (Category 1), 218 patients (67.3%, Group A) derived long lasting palliation and required no further radiotherapy during their lifetime. The other 106 patients (32.7%, Group B) however, needed subsequent radiotherapy with either EB (in the majority of cases) or repeat ILT because of failure to control symptoms satisfactorily with initial ILT or because of symptomatic disease recurrence. Fig. 4 shows the times at which these 106 patients underwent their subsequent treatment. By 6 weeks following ILT 9 patients (2.8% of the 324 patients in Groups A and B) had undergone further treatment. By 4 months 32 patients (9.9%) had undergone further treatment and by 12 months a total of 86 patients (26.5 0%) had undergone further treatment. Of the 106 patients in Group B who needed a subsequent treatment because of tumour recurrence, histology showed a similar distribution to the overall group of 406 patients. In 88% the diagnosis was squamous cell carcinoma, in 2% large cell carcinoma, in 3% adenocarcinoma and in 7% non-small cell carcinoma.

50

40

35.8

30 '.ha

2t.7

o

~ 20 f5.1

9.4

8.s

!0 3.8

3,7

u N +N+++N t.9

0 0-1.5

t5-4

4 8

8-12

12-16

16-20

20-24

24+

m o n t h s s i n c e I s t ILT Fig. 4. The 106 patients in Group B who needed a subsequent treatment due to failure of ILT to control symptoms or due to symptomatic disease recurrence following initial ILT. The figure shows, in various time intervals post-ILT treatment, the percentage of patients out of the 106 in Group B who had a subsequent treatment (total 100%).

37

S. W. Gollins et al. / Radiother. Oncol. 33 (1994) 31-40

4 months

1.5 rno~tths a f t e r f i r s t ILT 100

100 87.5

8o.

80

75.0

6(7

20

¢ ~rs~v~rs

4

i

a f t e r f i r s t ILT

N ///Z ,,//z

60

~. 40

12 m o n t h s

a f t e r f i r s t ILT

I00

56.3

(<('~ 40

....

4O"

37.5

3&3

~ 1 zs.4 20 t2.5

7.1

B

o2._

kq

////

o

~/.4

o

.=o o

2

3

4

5

6

1

2

3

4

5

6

1

2

3

4

5

6

31

32

31

5

14

1

17

16

16

2

8

0

3

4

3

f

I

1=stridor 2=dyspr~oea 3 = h a e m o p t y s i s

4 = c o u g h 5=pair~ 6=collapse

Fig. 5. The 65 patients in Groups C and D (16% of the total 406) who had had prior EB and received a single 1LT treatment on failure of EB to control symptoms or on symptomatic tumour recurrence. The histograms represent the percentage of assessed surviving patients who initially had the symptom shown, in whom this symptom was judged to have improved. The number of survivors who were assessed to derive each percentage shown is represented underneath the histograms.

There was no consistent difference in the degree of palliation achieved in patients treated with 2000 coy at 1 cm from the central axis of the source as opposed to 1500 cGy (not shown). Fig. 5 demonstrates the palliative efficiency of a single fraction of ILT in the 65 patients in Category 2 (Groups C and D) who received a single fraction of ILT on recurrence of tumour following previous EB. At 6 weeks following ILT the numbers of patients showing improvement in symptoms of haemoptysis (88%o), stridor (75%), pain (60%), cough (55%) and dyspnoea (52%) bear comparison with the previously unirradiated patients in Category 1 described above, although the numbers of patients assessable for stridor and pain are small. However, it is clear that the efficiency with which ILT can relieve collapse of lung tissue in this group (7% of patients improved) is not as good as in previously unirradiated patients. Moreover, although numbers of survivors are declining rapidly, it is evident that for the symptoms of haemoptysis, cough and dyspnoea at least, the palliative effect is not as well sustained with time at 4 and then 12 months, when compared with previously unirradiated patients (Category 1). Similar levels o f palliation to t h o s e in C a t e g o r y 1 were seen in the 17 patients in C a t e g o r y 3 in w h o m I L T a n d EB were c o m b i n e d as an initial treatment, a l t h o u g h n u m b e r s were small (not s h o w n ) .

5.2.

Survival

All surviving patients were followed up for a minimum of 16 months. At the close of study 383 patients (94.3%) had died from their cancer, 7 patients (1.9%) had died from an indeterminate cause, 4 patients

100

90 80 70

10I

II

)

0

Years Fig. 6. Kaplan-Meier actuarial survival curve for all 406 patients. Sur, vival dates from the time of first ILT treatment.

38

S. W. Gollins et al. / Radiother. Oncol. 33 (1994) 31-40 5.3. Morbidity I00 90 80 70 ~o 6 0

.3

5o tO

40

i~:.i

! tI

,i i,

30 20 fO II

O(

I

Years

Fig. 7. Kaplan-Meier survival curves for the three main treatment categories. These were the 324 previously unirradiated patients (Groups A and B), the 65 patients who had receivedpreviousEB and weretreatedwith ILT for recurrence(GroupsC and D) and the 17 patientswho receivedconcurrentILT and EB (Group E). Survivaldates from the time of first ILT treatment.- - , Groups A + B; . . . . , Groups C + D; . . . . , Group E.

(1%) had died from a cause other than cancer, 1 patient (0.2%0) was alive with disease recurrence and 11 patients (2.7%) were alive without disease recurrence. Fig. 6 shows the actuarial survival curve for the whole group of 406 patients. Median survival was 173 days, 2-year survival was 7.2% and 5-year survival was 1%. Fig. 7 represents the separate actuarial survival curves for the three main treatment categories. Median, 2-year and 5-year survival for the 324 previously unirradiated patients (Category 1) were 199 days, 8.8% and 1.3%, respectively. The median and 2-year survival for the recurrent group of 65 patients who had received previous EB (Category 2) were 130 days and 1.5%. The median and 2-year survival for the 17 patients who received combined ILT and EB (Category 3) were 199 days and 0%0.The survival of the patients who had undergone previous EB treatment (Category 2) is significantly less than the other two treatment categories (p < 0.0001 using log rank). This is almost certainly because these patients were at a later stage in the natural history of their disease at the time they received ILT. Thirty-three (8.1%) of the 406 patients initially presented with metastatic disease (TNM M1). A further 173 patients (42.6%) had documented locally recurrent disease, 40 patients (9.9%) had documented distant metastatic disease and 27 patients (6.7°/0) had both documented local recurrence and distant metastatic disease at the close of study.

The bronchoscopy and immediate ILT treatment were well tolerated with almost all patients being treated as a day-case procedure. The main symptomatic sideeffect of ILT was cough, with patients commonly reporting a mild transient exacerbation which settled without specific treatment within 2-3 weeks. This was accompanied by a transient exacerbation of dyspnoea in 5% of patients which resolved with a similar time course to cough. There was no radiation oesophagitis and there were no treatment related fistulae. One patient needed removal of necrotic debris by bronchoscopic debridemerit 28 months after a dose of 2000 cGy at 1 cm because of symptoms due to persistent radiation bronchitis. Massive haemoptysis death as a terminal event was recorded in 32 patients (7.9%) and in these patients the median time at which this occurred was 214 days. 6. Discussion The present study represents a series of 406 patients with inoperable lung cancer who received ILT to palliate symptoms due to endobronchial disease. Several previous studies have shown that ILT is as efficient at palliating the symptoms due to endobronchial disease in lung cancer as EB and can be more so in the case of pulmonary collapse [1,2,4,5,8,9,11-13,15,17,18]. In making comparisons between the present study and others previously reported, several features and differences in approach need to be taken into account. In the palliation of symptoms due to endobronchial disease in previously unirradiated advanced tumours, many centres favour the approach of using EB concurrently with ILT. The EB is typically 6000 cGy in 6 weeks, with brachytherapy given on alternate weeks at a dose of 500-750 cGy/fraction, to 3 or 4 fractions total. This can perhaps make it difficult to dissect out the relative contributions of the EB and the ILT treatments in assessing symptomatic response and morbidity. The use of ILT alone is commonly reserved for patients who have tumours that have recurred after an initial course of EB. In this situation again, the ILT is often fractionated. The present study uses a single fraction of ILT only, as both primary treatment in previously unirradiated patients and in patients with disease recurrence following previous EB treatment. Several reports have previously described a reasonably high frequency of laser use prior to ILT treatment, for example 24% of patients entered into one series [15]. The extent and alms of prior laser treatment are variable. Sometimes large deposits of energy are used to remove bulky tumours and sometimes the laser is simply used to facilitate placement of the endobronchial

S. IV. Gollins et al. / Radiother. Oncol. 33 (1994) 31-40

catheter. In the present study only 7 patients had had previous laser treatment varying from between 5 and 330 days before ILT was given. We have not found it necessary to deliver prior laser treatment to facilitate placement of the ILT treatment applicator. Because of the short median survival associated with carcinoma of the lung it has often been difficult to have sufficient numbers of longer term survivors to present data on the durability of the palliative response following ILT treatment. In the present study we were fortunate enough to have sufficient numbers of patients to be able to follow them up for a year post ILT treatment and have enough survivors to make conclusions about the proportion deriving long lasting palliation. In the present study we have attempted to examine the effect of ILT treatment on a homogeneous group of patients, namely non-small-cell lung cancers. In order to do this 42 patients out of the original 448 treated were excluded. These included patients with endobronchial metastases from extrathoracic primary tumours, patients with small cell carcinomas, patients with tumours confined to the trachea, patients with adenoid cystic carcinomas and patients with carcinoid tumours. The main conclusion from the present study is that a single ILT treatment in previously unirradiated patients, giving a dose of 1500-2000 cGy at I cm from the central axis of the source, is capable of improving the symptoms due to endobronchial disease in lung cancer to a degree comparable with that reported using schedules of multiple ILT treatments combined with EB. A single ILT treatment appears to be adequate therapy for approximately two-thirds of patients who derive long lasting palliation and require no further treatment during their lifetime. For the other third of patients who need a subsequent treatment at some stage either because of failure of the initial ILT to control symptoms or recurrence of their symptoms following an initial response, EB or a repeat ILT treatment can be effectively utilised. ILT treatment is also effective in palliating symptoms in patients who have recurrent disease following previous EB, providing a useful second line therapy, although this response is not so well sustained with time and pulmonary collapse is not well palliated in this situation. As far as side-effects are concerned, a single fraction of ILT treatment was found to be well tolerated in the present study. The percentage of patients dying from massive haemoptysis as a terminal event (7.9%) is similar to that observed in a recent publication [16] and does not seem to be excessively high for inoperable bronchial carcinoma. In the situation where it is accepted that a patient has an advanced tumour which is incurable and that the patient has a limited life span, the aim of treatment is to provide the maximum sustained improvement in symp-

39

toms with the minimum inconvenience for the patient and the minimum number of side-effects. A recent publication has demonstrated that a single fraction of EB can achieve acceptable levels of palliation in advanced lung cancer [3]. The present report shows that a single fraction of ILT treatment can produce levels of palliation for symptoms due to endobronchial disease which compare favourably with those published [3]. Whether the addition of EB as a combined treatment might increase the effectiveness of ILT in palliating symptoms, or even increase survival, has not yet been addressed in a prospective fashion but is currently the subject of an ongoing prospective randomised trial at the Christie Hospital.

Acknowledgement We thank Mr M. Dougal for help in the statistical analysis.

References [1] Aygun, C., Weiner, S., Scariato, A. Spearman, D. and Stark, L. Treatment of non-small cell lung cancer with external beam radiotherapy and high dose rate brachytherapy. Int. J. Radiat. Oncol., Biol. Phys. 23: 127-132, 1992. [2] Bedwineck, J., Petty, A., Bruton, C., Sofield, J. and Lee, L. The use of high dose rate endohronchial brachytherapy to palliate symptomatic endobronchial recurrence of previously irradiated bronchogenic carcinoma. Int. J. Radiat. Oncol., Biol. Phys. 22: 23-30, 1992. [3] Blvcben, N.M., Girling, D.J., Machin, D. and Stepbens, R.J. A Medical Research Council (MRC) randomised trial of palliative radiotherapy with two fractions or a single fraction in patients with inoperable non-small-ceU lung cancer (NSCLC) and poor performance status. Br. J. Cancer 65: 934-941, 1992. [4] Burt, P.A., O'Driscoll, B.R., Notley, H.M., Barber, P.V. and Stout, R. Intraluminal irradiation for the palliation of lung cancer with the high dose rate micro-Selectron. Thorax 45: 765-768, 1990. [5] Chang, L.-F.L., Horvath, J., Peyton, W. and Ling, S.-S. High dose rate afterloading brachytherapy in malignant airway obstruction of lung cancer. Int. J. Radiat. Oncol., Biol. Phys. 28: 589-596, 1994. [6] Cbetty, K.G., Moran, E.M., Sassoon, C.S.H., Viravathana, T. and Light, R.W. Effect of radiation therapy on bronchial obstruction due to bronchogenic carcinoma. Chest 95; 582-584, 1989. [7] Collins, T.M., Ash, D.V., Close, H.J. and Thorogood, J. An evaluation of the palliative role of radiotherapy in inoperable carcinoma of the bronchus. Clin. Radiol. 39: 284-286, 1988. [8] Cotter, G.W., Lariscy, C., Ellingwood, K.E. and Herbert, D. Inoperable ¢ndobronchial obstructing lung cancer treated with combined endobronchiai and external beam irradiation; a dosimetric analysis. Int J. Radiat. Oncol., Biol. Phys. 27: 531-535, 1993. [9] Macha, H.N., Koch, K., Stadler, M., Schumacher, W. and Krumhaar, D. New technique for treating occlusive and stenosing tumours of the trachea and main bronchi: endobronchial irradiation by high dose iridium-192 combined with laser canalisation. Thorax 42:511-515, 1987.

40

S. IV. Gollins et al. / Radiother. Oncol. 33 (1994) 31-40

[10] Majid, D.A., Lee, S.K., Kushalani, S. and Seydei, H. The response of atelectasis from lung cancer to radiation therapy. Int. J. Radiat. Oncol., Biol. Phys. 12: 231-232, 1986. [1 i] Mehta, M., Shahabi, S., Jarjour, N., Steinmetz, M. and Kuhsad, S. Effect ofendobronchial radiation therapy on malignant bronchial obstruction. Chest 97: 662-665, 1990. [12] Nori, D., Allison, R., Kaplan, B., Samala, E., Osian, A. and Karbowitz, S. High dose rate intraluminal irradiation in brow chogenic carcinoma. Chest 104: 1006-1011, 1993. [13] Seagren, S., Harrell, H.J. and Horn, R.A. High dose rate intraluminal irradiation in recurrent endobronchial carcinoma. Chest 88: 810-814, 1985. [14] Slawson, R.G. and Scott, R.M.. Radiation therapy in brow chogenic carcinoma. Radiology 132: 175-176, 1979.

[15] Speiser, B.L. and Spratling, L. Remote afterloading brachytherapy for the local control of endobronchial carcinoma. Int. J. Radiat. Oncol., Biol. Phys. 25: 579-587, 1993. [16] Speiser B.L. and Spratling L. Radiation bronchitis and stenosis secondary to high dose rate endobronchial irradiation. Int. J. Radiat. Oncol., Biol. Phys. 25:589-597, 1993. [17] Stout, R. Endobronchial brachytherapy. Lung Cancer 9: 295-300, 1993. [18] Sutedja, G., Baris, G., Schaake-Koning, C. and van Zandwijk, N. High dose rate brachytherapy in patients with local recurrences after radiotherapy of non-small cell lung cancer. Int. J. Radiat. Oncol., Biol. Phys. 24: 551-553, 1992