Lung Cancer (2006) 52, 149—154

avai lab le at www.sc iencedi rec t .com

journa l homepage: www.e lsev ier .com/ locate / lungcan

Radiation therapy alone in elderly withearly stage non-small cell lung cancer

Sol San Jose ∗, Maria Dolores Arnaiz, Anna Lucas, Valentin Navarro,Gala Serrano, Mariel Zaderazjko, Branislav Jeremic1, Ferran Guedea

Department of Radiation Oncology, Catalan Institute of Oncology, Barcelona, Spain

Received 31 March 2005; received in revised form 28 November 2005; accepted 5 December 2005

KEYWORDSRadiation therapy;Non-small cell lungcancer;Elderly;Early stage

SummaryBackground: Radiation therapy (RT) alone is frequently used in elderly patients with medicallyinoperable early stage (I/II) non-small cell lung cancer (NSCLC). We retrospectively investigatedthe effectiveness of RT alone in this patient population treated in our institution.Material and methods: Between 1995 and 1999, a total of 33 patients were treated with RT alonein our institution, all being males. RT doses ranged 66—78 Gy (median, 70 Gy) using standardfractionation (2.0 Gy per fraction). The age range was 71—97 years (median, 75 years) with 11patients being ≥80 years old. Twenty-two (67%) patients had a squamous cell carcinoma. Therewere 24 (73%) stage I and nine (27%) stage II patients.Results: Radiographic objective response rate was observed in 23 (70%) patients. The mediansurvival time was 37.4 months and 3-year survival time was 50%, while the median cause-specific survival time was 48.1 months and a 3-year cause-specific survival rate was 55.3%. Themedian time to local recurrence was 36.8 months and a 3-year local recurrence-free survivalrate was 50.2%, while the median time to distant metastasis was not achieved yet, the 3-yeardistant metastasis-free survival rate being 71.4%. One (3%) patient died of RT-induced acutelung toxicity, while only two (6%) patients experienced late grade 3 lung toxicity. No otherhigh-grade toxicity was observed during this study.Conclusions: RT alone was effective and low toxic in elderly with early stage (I/II) NSCLC andcould be considered as treatment of choice in this patient population.© 2006 Elsevier Ireland Ltd. All rights reserved.

∗ Corresponding author. Tel.: +34 93 260 7722;fax: +34 93 260 7725.

E-mail address: ssanjose@iconcologia.net (S. San Jose).1 Per invitation.

1. Introduction

The standard treatment approach in patients having tech-nically operable, but medically inoperable early stage (I/II)non-small cell lung cancer is radiation therapy (RT) [1]. In anunselected patient population with stage I/II, RT alone canachieve a median survival time (MST) of up to >30 months

0169-5002/$ — see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.lungcan.2005.12.010

150 S. San Jose et al.

(>40 months in T1N0), with 5-year survival rates of 25—30%(40% in T1N0) [2—19]. This was achieved in spite of the factthat this patient population represents a negative selec-tion, due to pre-existing comorbidity precluding surgeryor advanced age, both factors indicating frequent cancer-unrelated deaths and clinical staging. Advanced age, inparticular, becomes progressively important factor becauselung cancer is a typical disease of the elderly patient, beingthe first among the cancer-killers in both sexes.

While treatment approaches with curative intention arefeasible in patients with localized disease, the evidence isbased on studies, which are usually performed with selectedpatients. The elderly patients are generally underrepre-sented in clinical trials, which in cases of RT in elderly, arelacking. Furthermore, a frequent observation in daily prac-tice is that elderly are less likely to be vigorously screenedand staged, and frequently less aggressively treated [20].When, however, evaluated the specific features they didnot seem to have different characteristics at presentation,particularly related to stage of disease, performance statusand histology, when compared to their non-elderly coun-terparts, although other characteristics such as type andnumber of comorbidities and organ function differ in thetwo groups of populations [21]. Indeed, a number of studieson RT in early NSCLC provided a subgroup analysis accord-ing to age, indicating no difference in the various treat-ment outcome endpoints according to the age [7,10,12,14].Some, however, observed detrimental effect of advancedattn[

unWebp

2

Ippafotncs

thrfic

p

istered using 6 and 18 MV photons from linear accelera-tors. As a general policy, the patients presenting with thetumors localized more centrally, anterior—posterior treat-ment fields were mostly used until 36—38 Gy to treat allvisible tumor, with 2—1.5 cm margin, and the half or thewhole mediastinum, with a 1—1.5 cm margin, together withthe ipsilateral supraclavicular fossa, in cases of upper tumorlocation. After this, a combination of 2—3 oblique treat-ment fields followed to treat visible tumour and a part ofmediastinum up to 50 Gy using the same margin. After thisthe final tumor-only boost was used, raising the total doseup to at least 66 Gy. For the patients with more peripheraltumors, a combination of oblique fields was mostly used totreat all visible tumors with a 2—1.5-cm margin without asystematic approach to electively cover the nodes at riskof subclinical involvement. Maximum tumor dose was 78 Gyand minimum tumor dose was 66 Gy (in one patient only).Total tumor doses 66—70 Gy were given to 25 patients (74%),while the rest (n = 9; 26%) received 72—78 Gy. The maxi-mum dose allowed for the spinal cord was 46 Gy. Standardfractionation (2.0 Gy) was used throughout the RT course.Prescription of dose was made to ICRU-reference point[26], and tissue inhomogeneity correction was used for dosecalculation.

Response to treatment was evaluated after the end of RTwith the tools outlined above. Response criteria were by theWorld Health Organization (WHO) [27]. The RT-related toxi-city was assessed by involving radiation oncologist accordingt

t�

rtgAp

3

BNcdaeoiacttemsermo

m

ge [9,11], contrasting those reports who indicated advan-age for the elderly [22—24]. When specifically addressinghe issue of the role of RT in elderly with early stage NSCLC,o study provided a firm conclusion regarding this issue16,17].

In our institution we did not put a restriction on age in these of RT alone in early stage NSCLC. This policy resulted in aumber of elderly (>70 years) patients treated with RT alonee herewith report on a single-institutional experience withxclusive RT in technically operable but medically inopera-le elderly patients with early stage (I/II) NSCLC over theeriod of 5 years.

. Material and methods

n order to investigate the effectiveness of RT in elderlyatients with early stage (I/II) NSCLC, histories of allatients treated with this treatment modality between 1995nd 1999 were reviewed. Patients were considered eligibleor this analysis if they had a histological or cytological prooff NSCLC, were staged as having stage I or II according tohe International classification [25], age >70 years, and hado previous, concurrent or adjuvant treatment (surgery orhemotherapy). In most patients (n = 25; 76%) the main rea-on for not undergoing surgery was comorbidity.

The diagnostic tests did not change over the time andhey included medical history, physical examination, fullematological profile and biochemical screening tests, chestadiographs, CT scan of the thorax and upper abdomen,berbronchoscopy, fine needle aspiration biopsy if indi-ated, pulmonary functions tests, and bone scintigraphy.

In all patients CT scanning was done for the treatmentlanning. Three-dimensional (3D) conformal RT was admin-

o the Common Toxicity Criteria (CTC) version 2.0 [28].Differences between pairs of groups in patient charac-

eristics, and incidence of toxicity were evaluated by the2 and proportional test. Survival and relapse-free survivalates were calculated from the day 1 of the treatment usinghe Kaplan-Meier method, and differences between pairs ofroups in survival curves were analyzed by the log-rank test.ll these statistical analyses were carried out using the com-uter program S-PLUS.

. Results

etween 1995 and 1999, a total of 33 patients with stage I/IISCLC were treated at our institution with RT alone. Patientharacteristics are given in Table 1. During this period, weid not encounter a single elderly woman with this diseasemong our patients, an observation worldwide. The likelyxplanation for this finding is a particular smoking trendbserved in Catalonian females. Their smoking habits werencreasingly observed in late sixties of the last century, beingn initiator of this late increase in the incidence of lungancer in women. They were coupled with increased par-icipation in the workforce and women’s improved accesso secondary and university education during seventies orighties [29,30]. Also, the high prevalence of smokers inen led to the fact that only one (3%) of all patients never

moked with two-thirds (68%) of patients being defined asx-smokers, quitting smoking in the previous 6 months. Ageanged 71—97 years, with the median age of 75 years. Squa-ous cell histology predominated, with almost three-fourths

f all patients having stage I disease.Follow up ranged from 1.9 to 83.4 months (median, 16

onths) for all patients, while the median follow-up was 41

RT in early NSCLC 151

Table 1 Patient characteristics

Characteristic N %

SexM 33 100F 0 0

Age (years)71—79 22 6880—97 11 32

PS0 13 391—2 18 55n.a. 2 6

Weight loss0—5% 23 70>5% 4 12n.a. 6 18

StageI 24 73II 9 27

TNMT1N0M0 3 9T1N1M0 0 0T2N0M0 21 64T2N1M0 2 6T3N0M0 7 21

HistologySquamous cell 22 67Adenocarcinoma 3 9Large cell 1 3Undifferentiated 4 12NOS 3 9

RT dose66—70 Gy 25 7672—78 Gy 8 24

M: male; F: female; KPS: Karnofsky performance status score;n.a.: not available; NOS: non-small cell lung cancer not other-wise specified; RT: radiation therapy.

months for living patients. Thirteen patients (36.36%) werelost to follow-up. Radiological response rate to RT alonewas not assessable in three (9%) patients. Among assessablepatients, there was a complete response (CR) in 16 (48%)patients, a partial response (PR) in seven (21%) patients,making an objective response rate of 70%. stable disease(SD) was observed in three (9%), while tumor progressionwas noted in four (12%) patients. Four (12%) patients diedof intercurrent disease.

The median survival time (MST) was 37.4 months with 1—5years survival rates of 68.8, 57.8, 50.1, 45.9, and 38.2%,respectively (Fig. 1). The median cause-specific survivaltime was 48.1 months, with 1—5 years cause-specific sur-vival rates of 79.4, 63.8, 55.3, 55.3, and 46.1%, respectively(Fig. 2). The median time to local recurrence (MTLR) was36.8 months, with 1—5 years local recurrence-free survival(LRFS) rates of 85.3, 61.3, 50.1, 43.9, and 43.9%, respec-tively (Fig. 3). The median time to distant metastasis was not

Fig. 1 Overall survival (—–) with 95% confidence intervals(· · ·).

Fig. 2 Cancer-specific survival (—–) with 95% confidence inter-vals (· · ·).

achieved yet at the time of this analysis, while 1—5 years dis-tant metastasis-free survival (DMFS) rates were 76.9, 71.4,71.4, 71.4, and 71.4%, respectively (Fig. 4). Finally, themedian disease-free survival time was 18 months, while 1—5years disease-free survival (DFS) rates were 67.4, 51, 41.7,36.5, and 36.5%, respectively (Fig. 5).

Investigation of survival according to various pretreat-ment patient and tumor characteristics and RT dose usingthe Kaplan-Meier analysis showed that neither age did

Fig. 3 Local recurrence-free survival (—–) with 95% confidenceintervals (· · ·).

152 S. San Jose et al.

Fig. 4 Distant metastasis-free survival (—–) with 95% confi-dence intervals (· · ·).

influence survival (p = 0.72), nor stage did (p = 0.57). Weightloss (p = 0.143), performance status (p = 0.55), histology(p = 0.26), and the RT dose (p = 0.30) also did not influenceoverall survival. Because the local failure was predominanttype of failure, we have used LRFS as an additional endpointinvestigating the impact of all of these potential prognosti-cators on LRFS. Kaplan-Meier survival analysis showed thatage did not influence LRFS (p = 0.94), neither stage (p = 0.52)nor weight loss (p = 0.46) did. Also, there was no influenceof performance status (p = 0.81), histology (p = 0.63) or theRT dose (p = 0.55) on LRFS (Table 2).

Analysis of the patterns of failure identified a total of17 (50%) patients failing. Twelve (35%) patients had localcomponent of failure, while eight (24%) patients had pro-

Fig. 5 Disease-free survival (—–) with 95% confidence intervals(· · ·).

gression at distant sites (some patients had more than onesite of failure).

Acute high-grade (≥3) toxicity was observed in onepatient who died of RT-induced pneumonitis 1 month post-RT. This was a 77-year-old male with a T2N0 tumor and aPS of 1, who did not experience previous weight loss, buthad concurrent chronic obstructive lung disease accompa-nied with arrhythmia. After receiving a total of 70 Gy, hedied less than 2 months after the end of RT with his tumorstatus remaining unknown since he died outside our hospitalbefore he was able to come to the first follow-up visit. Latehigh-grade toxicity was infrequent and consisted of two (6%)patients experiencing late grade 3 lung toxicity. No otherhigh-grade toxicity was observed during this study.

Table 2 Kaplan-Meier survival analysis of various prognostic factors influencing LRFS

Variable N MTLR (months) LRFS (%) p

1 year 2 years 3 years 4 years 5 years

Age (years)71—79 22 37 89.5 58.7 50.3 41.9 41.9 0.9480—97 11 n.a. 78.7 67.5 50.6 50.6 50.6

PS*0 13 37 91 67 54 40 40 0.81

5

65

66

58

647.6 23.8 23.8 23.8

1—2 18 35 80.4

Weight loss*≤5 23 36.8 90>5% 4 24.8 75

HistologySCC 22 35.5 75.8Other 11 35.1 100

StageI 24 28 85.4II 9 n.a 85.7

RT dose (Gy)66—70 25 n.a. 84.872—78 8 18 85.7

MTLR: median time to local recurrence; LRFS: local recurrence-free susquamous cell carcinoma; RT: radiation therapy; n.a.: not achieved ye

5.1 45.9 45.9 —

8.4 53.2 45.6 45.6 0.460 50 — —

0.3 43 43 — 0.631 45.7 45.7 45.7

4.9 48.1 40.1 40.1 0.525.7 57.1 57.1 57.1

5.6 51 51 51 0.55

rvival time; PS: performance status; *: some data missing; SCC:t.

RT in early NSCLC 153

4. Discussion

RT in elderly with early stage NSCLC had been reported tobe effective and well tolerated since the initial report ofAristizabal et al [22], who were first to show that patients≥70 years had significantly better 2-year survival than non-elderly (49—69 years) (35.7% versus 13.1%, p = 0.044). Asignificant trend towards better survival in older patientswith NSCLC treated with RT alone was observed also by Coyand Kennelly [23] and Newaishy and Kerr [24]. A numberof studies found no influence of age on treatment out-come, regardless of the cut-off value used in the analysisof RT characteristics [7,10,12,14], and only occasionally itwas reported that elderly fare worse than their non-elderlycounterparts. Morita et al. [9] found a survival advantagefor patients <80 years old when compared to those ≥80years old (5-year survival: 25.2% versus 7.7%; p = 0.035) with-out other endpoints used to enable better insight into thisphenomenon, while Sibley et al. [11] documented superioroutcome in younger (<60 years) patients with stage I versusolder patients, unconfirmed, however, when local progres-sion was used as an endpoint (p = 0.10). Hayakawa et al. [15]treated 97 patients ≥75 years old and 206 patients <75 yearsold, with RT doses ranging from ≥60 to >80 Gy for inoperableNSCLC. There were two groups of elderly: 75—79 years and≥80 years. No difference was found between non-elderlyand the two elderly groups (5-year survival: 12% versus 13%

must be attributed to the clinical staging and low patientnumber. The local failure remains the predominant one andrequest further investigation in the field of thoracic RT, theprincipal standard treatment in this setting.

Numerous studies dealing with this topic have attemptedto evaluate the toxicity in elderly [2—19]. When specificallyaddressing elderly with early stage NSCLC, no significantRT-related complications were found and incidence of bothacute and late high-grade toxicity was low and similar amongall age groups [17]. When RT-related deaths occurred, again,there was no difference between elderly (5%) treated withhighest dose levels (80 Gy) and their non-elderly counter-parts (4%) treated the same way [16]. Again, results of ourstudy fit well within this framework, with one treatment-related deaths and two (6%) late lung damage. We believethese figures are favorable, having in mind the high preva-lence of comorbid disease in this population.

Due to a small patient number, we did not attempt to per-form a multivariate analysis to disclose potential prognosticfactors. Kaplan-Meier survival analysis using both overallsurvival and LRFS was inconclusive, and no factor showed itsinfluence on treatment outcome. This reiterates the prob-lem of the lack of clearly identified prognostic factors in thissetting and request for further studies with more patientsneeded to gather these valuable informations.

In conclusion, this study showed that exclusive RT is aneffective treatment modality with low toxicity in elderly(>70 years) with early stage (I/II) NSCLC. Albeit of shortcom-il5c

R

versus 4%, respectively). A multivariate analysis discloseddetrimental effect of oldest age, likely to be the conse-quence of 14% treatment-related deaths in patients receiv-ing 80 Gy. Also, Gauden and Tripcony [17] investigated theeffect of age (<70 years versus ≥70 years) in patients withstage I NSCLC and found no significant difference in termsof MST, 5-year survival and recurrence-free survival betweenthe two groups.

Although with inherent limitations of retrospectivenature of reports, covering long-time periods during whichdiagnostic and treatment variables have substantiallychanged, the data from the literature, when taken together,show that RT alone is capable of producing MST of 20—27months and 5-year survivals of 15—34% in patients >70 years.

This study showed excellent results of exclusive RT inelderly patients with early stage NSCLC, at least compa-rable to those of the literature, having in mind the smallpatient number and the fact that only a few of our patientsremain at risk after 5 years. On the other side, accurate stan-dard pretreatment study, treatment characteristics and thefact that all patients were treated curatively in the 3D erawith high dose of RT (the majority of patients received, atleast 70 Gy), further strengthen our findings. Furthermore,except for the fact that we encountered no females duringthis study period, these patients were unselected and there-fore, good representative of every day’s clinical practice notonly in a large tertiary cancer center, such as ours, but ratherin smaller and not necessarily university-based teaching hos-pital departments. Results of our study are, therefore, veryapplicable to any setting.

Similarly to other studies [2—19], the analysis of the pat-tern of failure points to the predominant local nature of thefailure, with the 5-year LRFS and the 5-year DMFS being 43.9and 71.4%, respectively. While the latter figure may seemsomewhat higher than usually observed, we believe this

ngs inherent to any retrospective analysis, with somewhatow patient number and few patients remaining at risk atyears, nevertheless, it showed that exclusive RT could be

onsidered as the treatment of choice in this disease.

eferences

[1] Jeremic B, Classen J, Bamberg M. Radiation therapy alonein technically operable, medically inoperable early stage(I/II) non-small-cell lung cancer. Int J Radiat Oncol Biol Phys2002;54:119—30.

[2] Zhang HX, Yin WB, Zhang LJ. Curative radiotherapy ofearly operable non-small cell lung cancer. Radiother Oncol1989;14:89—94.

[3] Talton BM, Constable WC, Kersh CR. Curative radiotherapy fortechnically operable Stage I nonsmall cell lung cancer. Int JRadiat Oncol Biol Phys 1990;19:15—21.

[4] Dosoretz DE, Katin MJ, Blitzer PH. Radiation therapy in themanagement of medically inoperable carcinoma of the lung:results and implications for future treatment strategies. Int JRadiat Oncol Biol Phys 1992;25:3—9.

[5] Cheung PCF, Mackillop WJ, Dixon P. Involved-field radiotherapyalone for early-stage non-small-cell lung cancer. Int J RadiatOncol Biol Phys 2000;48:703—10.

[6] Haffty BG, Goldberg NB, Gerstley. Results of radical radiationtherapy in clinical Stage I, technically operable non-small celllung cancer. Int J Radiat Oncol Biol Phys 1988;15:69—73.

[7] Noordijk EM, vd Poest Clement E, Hermans J. Radiotherapy asan alternative to surgery in elderly patients with respectablelung cancer. Radiother Oncol 1988;13:83—9.

[8] Ono R, Egawa S, Suemasu K. Radiotherapy in inoperable stageI lung cancer. Jpn J Clin Oncol 1991;21:125—8.

[9] Morita K, Fuwa N, Suzuki Y. Radical radiotherapy for medi-cally inoperable non-small cell lung cancer in clinical stageI: retrospective analysis of 149 patients. Radiother Oncol1997;42:31—6.

154 S. San Jose et al.

[10] Kaskowitz L, Graham MV, Emami B. Radiation therapy alone forstage I non-small cell lung cancer. Int J Radiat Oncol Biol Phys1993;27:517—23.

[11] Sibley GS, Jamieson TA, Marks LB. Radiotherapy alone formedically inoperable stage I non-small-cell lung cancer: theDuke experience. Int J Radiat Oncol Biol Phys 1998;40:149—54.

[12] Slotman BJ, Karim ABMF. Curative radiotherapy for technicallyoperable stage I nonsmall cell lung cancer. Int J Radiat OncolBiol Phys 1994;29:33—7.

[13] Graham PH, Gebski VJ, Langlands AO. Radical radiotherapy forearly nonsmall cell lung cancer. Int J Radiat Oncol Biol Phys1995;31:261—6.

[14] Gauden S, Ramsay J, Tripcony L. The curative treatment byradiotherapy alone of stage I non-small cell carcinoma of thelung. Chest 1995;108:1278—82.

[15] Hayakawa K, Mitsuhashi N, Saito Y. Limited field irradiation formedically inoperable patients with peripheral stage I non-smallcell lung cancer. Lung Cancer 1999;26:137—42.

[16] Hayakawa K, Mitsuhashi N, Katano S. High-dose radiation ther-apy for elderly patients with inoperable or unresectable non-small cell lung cancer. Lung Cancer 2001;32:81—8.

[17] Gauden SJ, Tripcony L. The curative treatment by radiationtherapy alone of stage I non-small cell lung cancer in a geriatricpopulation. Lung Cancer 2001;32:71—9.

[18] Furuta M, Hayakawa K, Katano S. Radiation therapy for stageI—II non-small cell lung cancer in patients aged 75 years andolder. Jpn J Clin Oncol 1996;26:95—8.

[19] Dosoretz DE, Galmarini D, Rubenstein JH. Local control in med-ically inoperable lung cancer: an analysis of its importance in

outcome and factors determining the probability of tumoureradication. Int J Radiat Oncol Biol Phys 1993;27:507—16.

[20] Nugent WC, Edney MT, Hammerness PG. Non-small cell lungcancer at the extremes of age: impact on diagnosis and treat-ment. Ann Thorac Surg 1997;63:193—7.

[21] Montella M, Gridelli C, Crispo A. Has lung cancer in theelderly different characteristics at presentation? Oncol Rep2002;9:1093—6.

[22] Aristizabal SA, Meyerson M, Caldwell WL. Age as a prognosticindicator in carcinoma of the lung. Radiology 1976;121:721—3.

[23] Coy P, Kennelly GM. The role of curative radiotherapy in thetreatment of lung cancer. Cancer 1980;45:698—702.

[24] Newaishy GA, Kerr GR. Radical radiotherapy for bronchogeniccarcinoma: five year survival rates. Clin Oncol (R Coll Radiol)1989;1:80—5.

[25] Mountain CF. Revisions in the international system for staginglung cancer. Chest 1997;111:1710—7.

[26] International Commission on Radiological Units and Measure-ments. Prescribing, recording, and reporting photon beamtherapy. ICRU Report No. 50, 1993.

[27] Miller AB, Hoogstraten B, Staquet M. Reporting results of cancertreatment. Cancer 1981;47:207—14.

[28] Cancer Therapy Evaluation Program. Common Toxicity Criteria,Version 2.0. USA, Bethesda; 1999.

[29] Borras JM, Fernandez E, Schiaffino A. Pattern of smoking initia-tion in Catalonia, Spain, from 1948 to 1991. Am J Public Health2000;90:1459—62.

[30] Borras J, Borras JM, Galceran J. Trends in smoking-related can-cer incidence in Tarragona, Spain, 1980—1996. Cancer CausesControl 2001;12:903—8.