Radiotherapy and Oncology, 19 (1990) 317-322 Elsevier

317

RADION00785

Lung cancer in patients with borderline lung functions - zonal lung perfusion scans at presentation and lung function after high dose irradiation R a y m o n d P. Abratt

1 Paul A.

Willcox 2 a n d J a m e s A. Smith 3

Departments of 1Radiotherapy, ~Respiratory Medicine and 3Nuclear Medicine, University of Cape Town and Groote Schuur Hospital, Cape Town, Republic of South Africa (Received 6 November 1989, revision received 14 June 1990, accepted 19 June 1990)

Key words." Lung cancer; Lung function test; Perfusion scan

Summary We prospectively studied patients with lung carcinoma and borderline lung functions (forced expired volume [ FEV 1] of less than 1500 ml or a dyspnoea score of 2 [ 7 ] at presentation), who were treated with high dose irradiation. Patients were divided into those with suprahilar and hilar tumours. Lung perfusion was assessed in upper, middle and lower zones for each lung at presentation. The ipsi-lateral upper and middle zone were regarded as at risk from irradiation in patients with suprahilar tumours and the whole ipsi-lateral lung in patients with hilar tumours. Lung function was measured at presentation (18 patients) at 4-6 month follow up (16 evaluable patients = group 1) and again at 10-12month follow up (10 evaluable patients = group 2). A worsening of the dyspnoea score (3 in group 1 and 2 in group 2) occurred only in patients with a > 10~o decrease in transfer factor irrespective of the change in FEV1. A statistically significant correlation was found between decreased transfer factor at follow up and the perfusion in the lung zones regarded as at risk from irradiation at presentation (Spearman's rank correlation). There was no correlation between perfusion and changes in the FEV1. Patients in whom lung perfusion was less than 3 5 ~ in the zones at risk tended not to have decreased transfer factor at follow up. These findings indicate that worsening in the patients' dyspnoea score after irradiation is dependent on decreased transfer factor rather than FEV1 and that patients with borderline lung functions may be treated with irradiation if the perfusion in the zones at risk from radiation is less than 35 ~o.

Address for correspondence: Dr. R. P. Abratt, Senior Specialist, Department of Radiotherapy, L Block, LE 34, Groote Schuur Hospital, Observatory 7925, Cape Town, Republic of South Africa. 0167-8140/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

318 Introduction

Patients with lung carcinoma frequently have compromised lung function because of co-existent chronic obstructive airways disease. Surgery or irradiation for lung carcinoma can only be undertaken if it is predicted that patients will tolerate the loss of normal lung tissue. Surgery results in the removal of the air passages and parenchymal lung tissue. Patients are not normally considered candidates for surgery if their predicted post-operative FEV1 is less than 800 to 1000 ml as this is associated with carbon dioxide retention and decreased exercise tolerance [3]. This post-operative FEV1 may be predicted with the help of regional lung perfusion scans using the formula: post-operative FEV1 = preoperative FEV1 x percent perfusion of remaining lung [1,9] as there is a statistically significant correlation between predicted and measured results. Analogous studies following radiation are more difficult as there is a delay in the onset of radiation fibrosis and the mechanism of lung damage is more complex, occurring mainly at the capillaryalveolar level. The changes in FEV1, the standard used, have not correlated with those predicted from regional perfusion studies in patients with

TABLE I Modified Medical Research Council Dyspnoea Scale [7]. Grade

Description Not troubled with breathlessness except with strenuous exercise Troubled by shortness of breath when hurrying on the level or walking up a slight hill Walks slower than people of the same age on the level because of breathlessness or has to stop for breath when walking at own pace on the level Stops for breath after walking about 100 yards or after a few minutes on the level Too breathless to leave the house or breathless when dressing or undressing

poor lung function. The changes in lung function tests have not been related to a functional index such as the dyspnoea score [7] (see Table I) and it would be therefore incorrect to assume that relatively little lung damage has occurred in patients in whom the FEV1 has remained stable. A prospective study was undertaken in patients with borderline lung functions who were being treated with irradiation to evaluate the relationship between a worsening in their dyspnoea score and changes in their lung functions (spirometry and gas transfer). The possible predictive value of pre-irradiation zonal lung perfusion scans on post-irradiation lung function was also studied.

Materials and methods

Patient selection and follow up Patients with lung cancer were evaluated for irradiation for local disease control at a multidisciplinary clinic if they had loco-regional disease and good performance status (ECOG 0 to 2). The dyspnoea score and FEV1 are measured on all patients as part of their work up. Routine staging investigations were limited to serum chemistry; bone scans and liver ultra sound were done to exclude metastases only in patients with abnormal serum enzymes or suggestive symptoms. Patients receiving irradiation were included in this study if they were regarded as having borderline lung functions. This comprised patients with an FEV1 of less than 1500 ml or a dyspnoea score of 2. The FEV1 of 1500 ml is an arbitrary value but has been used to indicate the need for further evaluation in patients prior to lobectomy [9]. It would be more accurate to express the value as a percentage of the predicted normal. The FEV1, however, is a practical screening test to select patients for more sophisticated evaluation and can be readily measured in radiotherapy outpatients. Patients with a dyspnoea score of 2 can ill afford a worsening of one point in their dyspnoea score and indicates that there may be functional impairment not revealed by the FEV 1.

319 The patients were divided into those with hilar and suprahilar tumours on the basis of their chest radiograph. Lung perfusion scans (see below) were performed. Lung function was measured in terms of total lung capacity (TLC), FEV1, forced vital capacity (FVC) and gas transfer (TLCO) using a computerized lung function system (Jeager master lab, Wursburg, F.R.G.). Lung functions were repeated in evaluable patients between 4 and 6 months after presentation (group 1) and again between 10 and 12 months (group 2). Evaluable patients were free of overt locally progressive carcinoma and were not anaemic. Perfusion scans

Lung perfusion scans were taken with the patient lying supine using 100 MBq Technetium 99m labelled Macro aggregated albumin given intravenously. Anterior and posterior views were recorded and displayed on a monitor. Rectangular regions of interest were placed around each image and divided into three equal zones between the lung apex and base. Total counts in each zone (upper, middle and lower) on both anterior and posterior images were recorded and the mean of anterior and posterior counts were regarded as perfusion to each zone. The junction between the middle and lower zones was shown to approximate to the lung hilum by the following procedure. During lung perfusion scanning a radioactive marker was used to locate the apex and base of the lung and the division into upper, middle and lower zones. These divisions were marked with pins. A supine chest radiograph was exposed to display the relationship between the pins, the lung hilum and the zones on the perfusion scan. In patients with hilar tumours, all the ipsilateral zones were regarded as at risk from irradiation. In patients with suprahilar tumours, the ipsilateral upper and middle zones were regarded as at risk.

Radiation

Patients received either 50-55 Gy in 2.5 Gy fractions or 48 Gy in 3 Gy fractions on a cobalt teletherapy unit with anterior and posterior portals to the tumour, with a minimum 2 cm margin, and the adjacent mediastinum, ipsi-lateral hilum and supraclavicular nodes. The mediastinum and nodes were not irradiated in the patients with stage I disease. The spinal cord was shielded posteriorly at cord tolerance. Four fractions per week were given with both regimens. None of the patients had chemotherapy.

Results

Patient characteristics

Eighteen patients were entered into this study between February 1987 and June 1988. The mean age was 58 years (range 41-72 years). There were 9male and 9female. Eleven (61~o) had squamous carcinoma; 5 (28~o) adenocarcinorna and 2 (11~o) large cell carcinoma. Two patients had stage 1 disease and 16 stage 3A or 3B disease. The patients with stage 1 disease were not considered candidates for surgery because of poor lung function and the primary treatment in the remaining patients was irradiation for inoperable disease. Ten patients had hilar and 8 had suprahilar tumours. Fifteen patients had an FEV1 of less than 1500 ml which consisted of 4 patients with a dyspnoea score of 1 and 11 patients with a dyspnoea score of 2. Three patients had an FEV 1 of more than 1500 ml and all had a dyspnoea score of 2. The mean radiation portal size was 124 cm 2 (range 49-156 cmZ). Seven patients were treated with 2.5 Gy fractions and the remainder with 3 Gy fractions. The one and 2 years survival rate was 55 and 27 ~o, respectively with a minimum follow up of 12months in surviving patients (life table method). Seven patients are alive at follow up; 4

320 with no evidence of active disease at a mean follow up of 19 months and 3 with active disease at a mean follow up of 16 months. Fourteen of the patients have developed progressive disease in the lung and 10 of them were within the radiation ports with a local recurrence rate of 55 ~o. Sixteen of the 18 patients were assessed at 4 - 6 m o n t h s follow up, which excluded two patients with progressive local disease (group 1). All 10 of the patients at 10-12 months follow up were evaluated (group 2).

group 1 patients and p = 0.13 in the group 2 patients-Fisher's exact test). The worsening in dyspnoea score was unrelated to changes in FEV1. The worsening in dyspnoea score was not an adverse prognostic factor. One of these patients died of cerebral metastases at 9 months, one died with nodal metastases at 19 months and one is alive with no evidence of active disease at 24 months.

Lung perfusion Lung functions The lung function tests at presentation and the changes at follow up are shown in Table II. An increase in mean FEV1 and a decrease in mean TLCO was noted in group 1 at 6 months and a decrease in both FEV1 and T L C O was noted in group2 at 12months. The changes in lung function tests at follow up were not statistically significant in the two groups (paired t-test). The number of patients in whom the FEV1 and TLCO showed a more than 10~o decrease, remained unchanged or a more than 10 ~ increase are shown in Table III. Three patients had a worsening of one point in dyspnoea score and all had a decreased TLCO. This was statistically significant in the group 1 patients (p < 0.03 in

The mean lung perfusion in the zones of at risk was 32~o for hilar tumours and 28~o for suprahilar tumours (see Table IV). The mean lung perfusion was 16~o (range 0-26~o) in the mid zone for hilar tumours and 8~o (range 3-14~o) in the upper zone for suprahilar tumours. A statistically significant correlation was found between perfusion in the zones at risk as described above and a decreased transfer factor at follow up using Spearman's rank correlation test. (This is a non-parametric test for data which is not normally distributed and in which the results are ranked. Patients in whom the variation of TLCO was less than 1 0 ~ were all given the same rank.) In the 16patients in group 1 evaluated at 6months, r s = 0 . 5 4 and p < 0 . 0 5 . In the

T A B L E II M e a n lung function tests at presentation and the changes at follow up. G r o u p 2 : 1 0 patients

G r o u p 1: 16 patients

P e r f o r m a n c e status D y s p n o e a score FEV1 (ml) FEV1 S.D. F V C (ml) T L C (ml) T L C O (ml/min per m m H g ) T L C O S.D.

Presentation

4 - 6 mths

Presentation

10-12 mths

1 2 1360 310 2100 3850 11.6 2.8

+0 +0 + 45 390 + 220 + 265 - 0.2 2.8

1 2 1290 355 2115 4005 12.1 2.7

+0 +0 - 55 250 + 100 - 80 - 0.9 2.0

The differences are not statistically significantly different (paired t-test). F V C = F o r c e d vital capacity; S.D. = s t a n d a r d deviation; T L C = total lung capacity; T L C O = single b r e a t h c a r b o n m o n o x i d e transfer factor.

321 TABLE III The number of patients with changed FEV1 and transfer factor of > 10% and the number of patients in these subgroups who have had a worsening in their dyspnoea score. Assessment at 4 - 6 months (group 1) TLCO

Total

Worse dyspnoea score

FEV1

Total

Worse dyspnoea score

Decreased Unchanged Improved

6 5 5

3 0 0

Decreased Unchanged Improved

2 8 6

0 2 1

Decreased Unchanged Improved

4 4 2

1 1 0

Assessment at 10-12 months (group 2) Decreased Unchanged Improved

4 4 2

2 0 0

The incidence of worsened dyspnoea score in group 1 is statistically significantly higher in patients with a decreased TLCO (p = 0.03 Fishers exact test).

10 patients in group 2 evaluated at 12 months, r s = 0.66 and p < 0.05. These changes did not correlate with the initial portal or the radiation fraction size. There was no correlation between perfusion and changes with spirometry.

A decrease of more than 10~o in gas transfer tended to occur if perfusion was above 35 ~o in the zones at risk and a decrease rarely occurred if perfusion was below 35 ~o (see Table V). This was statistically significant (p < 0.5-Fisher's exact test).

TABLE IV Mean perfusion in ipsi-lateral lung zones (as ~o) followed by the range in 18 patients. Tumours

MZ + UZ

M Z + U Z + LZ

Hilar Suprahilar

22 (0-32) 28a (19-36)

32 a (5-53) 46 (27-51)

a The zones regarded as at risk (see text) from irradiation. LZ = lower zone; M Z = mid zone; U Z = upper zone.

TABLE V The proportion of patients who manifest a decreased TLCO at follow up is statistically significantly higher if the lung perfusion was more than 3 5 ~ in the zones at risk from radiation at presentation.

Perfusion > 35~o Perfusion < 35~o p = 0.024 Fisher's exact test.

Yes

No

5 1

2 8

Discussion Radiation-induced lung damage is a dynamic and interactive process. The most important target cells appear to be the type II pneumocytes and the capillary endothelium [6]. The end result of lung injury is fibrosis and thickening of the alveolar septae and a reduction in fine vasculature [5]. It would be anticipated therefore that gas diffusion capacity would be a more sensitive test for lung injury than lung volumes. A worsening in dyspnoea score was only found in this study in patients with a > 10~o decrease in gas transfer. A clinical study of fractionated radiation in patients without lung carcinoma (treated for Hodgkins disease) showed that TLCO was the most sensitive marker [4]. The changes in FEVI after radiation relate to the balance between direct tissue damage and the traction on adjacent tissue following fibrosis [2].

322 They are not as clinically significant as after surgery. In a study with repeat lung functions at 8-10 months, there was a correlation between predicted and measured values in patients with an FEV1 of over 50~o or more of the predicted normal value, but if the FEV1 was less than 50~o predicted, it decreased to the value expected from the formula in only 10~o of cases [2]. In the only other reported study [8], patients had an initial mean FEV1 of 1900 ml, there was a decrease (mean = 1 2 ~ ) close to but not below the predicted value at the 6-12 month follow up interval, There are other factors which may affect lung function, after irradiation, such as co-existent lung disease and radiation factors (dose, dose per fraction and treatment volume). Functional compensation may occur in patients who do not have borderline lung functions. In patients with lung cancer, less lung damage (as measured by TLCO) can be expected if there is decreased perfusion in the zones at risk from irradiation, as less functioning lung is irradiated. The tumour, which may be responsible for the decreased perfusion, is also being treated. The percentage of perfused lung affected by radiation has been assumed to be the region within the irradiation portals in previous studies. This study uses a novel but simple method to document lung zones for perfusion studies. They can be determined prior to radiation planning which is an advantage if they are used to select patients for therapy. Lung perfusion in the zones at risk was found to correlate with decreased transfer factor at follow up in patients with borderline lung function. A threshold value of 3 5 ~ lung perfusion in the zones at risk from irradiation was found below which the transfer factor rarely decreased. These findings suggest that zonal lung perfusion and the transfer factor are useful predictive tests and helpful in the selection of patients with borderline lung functions for high dose irradiation. Patients with borderline lung functions and perfusion of less than 35~o in their zones at risk

from irradiation should not be denied radiation therapy on the basis of their lung functions alone. Patients with low FEV1 but relatively normal TLCO may also be candidates for irradiation.

Acknowledgements We thank the National Cancer Association of South Africa for financial support and Mrs. J. Wilmott for collecting the data. We also thank Miss Hazel Murray for typing the manuscript.

References I Bria,W. F., Kanarek, D. J. and Kazemi,H. Predictionof postoperative pulmonary function following thoracic operations. Value of ventilation-perfusion scanning. J. Thorac. Cardiovasc. Surg. 86: 186-192, 1983. 2 Choi, N.C. Prospective prediction of postradiotherapy pulmonary function with regional pulmonary function data: promis and pitfalls.Int. J. Radiat. Oncol. Biol.Phys. 15: 245-247, 1988. 3 Diener, C. F. and Burrows, B. Further observations on the course and prognosis of chronic obstructive lung disease. Am. Rev. Respir. Dis. 111: 719-724, 1975. 4 Do Pico, G.A., Wiley, A.L., Rao, P. and Dickie, H.A. Pulmonary reaction to upper mantle radiation therapy for Hodgkin's disease. Chest 75: 688-692, 1979. 5 Gross, N.J. Pulmonaryeffectsof radiation therapy. Ann. Intern. Med. 86: 81-92, 1977. 6 Law, M.P. and Ahier, R.G. Vascular and epithelial damage in the lung of the mouse after X-rays or neutrons. Radiat. Res. 117: 128-144, 1989. 7 Mahler, D.A. and Wells, C.K. Evaluation of clinical methods for rating dyspnea. Chest 93: 580-586, 1988. 8 Rubenstein, J. H., Richter, M.P., Moldofsky,P.J. and Solin, L.J. Prospective prediction of post-radiation therapy lung function using quantitative lung scans and pulmonary function testing. Int. J. Radiat. Oncol. Biol. Phys. 15: 83-87, 1988. 9 Wernly, J. A., DeMeester, T.R., Kirchner, P.T., Myerowitz, P.D., Oxford, D.E. and Golomb, H.M. Clinicalvalue of quantitative ventilation-perfusion lung scans in the surgical management of bronchogenic carcinoma. J. Thorac. Cardiovasc. Surg. 80: 535-543, 1980.

Lung cancer in patients with borderline lung functions--zonal lung perfusion scans at presentation and lung function after high dose irradiation.

We prospectively studied patients with lung carcinoma and borderline lung functions (forced expired volume [FEV1] of less than 1500 ml or a dyspnoea s...
438KB Sizes 0 Downloads 0 Views