The Evolution of Radiation Therapy Oncology Group (RTOG) protocols for nonsmall cell lung cancer

1n1. J. Radiation

Oncology

Biol.

Pergamon

Phys..

Vol. 32. No. 5. pp. 1513-1525, 1995 Copyright 0 19’35 Elsevier Science Ltd Printed in the USA. All rights reserved 0360.3016/95 $9.50 + DO

0360-3016(95)00084-4

l

Oncology THE

Intelligence

EVOLUTION OF RADIATION THERAPY ONCOLOGY GROUP PROTOCOLS FOR NONSMALL CELL LUNG CANCER ROGER W. BYHARDT,

(RTOG)

M.D.

Departmentof Radiation Oncology, Medical Collegeof Wisconsin,8700 W. WisconsinAve., Milwaukee, WI 53226 Over the past 2 decades, the Radiation Therapy Oncology Group (RTOG) has played a significant role in clarifying the role of radiation therapy (RT) in the treatment of nonsmall cell lung cancer (NSCLC). RTOG lung cancer research has evolved over this time period through a systematic succession of investigations. For unresectable NSCLC, the dependence of local tumor control and survival on total dose of standard fractionation RT, as well as pretreatment performance characteristics, was demonstrated in initial RTOG trials. Subsequently, further radiation dose intensification was tested using altered fractionation RT to total doses up to 32% higher than standard RT to 60 Gy, given as either hyperfractionation or accelerated fractionation, while attempting to retain acceptable normal tissue toxicity. These higher doses required rethinking of established RT techniques and limitations, as well as careful surveillance of acute and late toxicity. A survival advantage was shown for hyperfractionation to 69.6 Gy, in favorable performance patients, compared to 60 Gy. Further testing of high dose standard RT will use three-dimensional, conformal techniques to minimize toxicity. RTOG further extended the theme of treatment intensification for unresectable NSCLC by evaluating combined chemotherapy (CT) and RT. Improved local control and survival was shown for induction CT followed by standard RT to 60 Gy, compared to standard RT (60 Gy) and altered fractionation RT (69.6 Gy). The intent of current studies is to optimize dose and scheduling of combined CT and standard RT, as well as combined CT and altered fractionation RT. Noncytotoxic RT adjuvants, such as hypoxic cell sensitizers, nonspecific immune stimulants, and biologic response modifiers have also been studied. Resectable NSCLC has also been an RTOG focus, with studies of preoperative and postoperative RT, CT, and CT/RT, including the prognostic value of serum and tissue factors. RTOG studies have yielded incremental improvements in treatment outcome for NSCLC, better understanding of the disease dynamics, and a strong foundation for future investigations. Nonsmall

cell lung cancer, Protocols,

Radiation

therapy,

INTRODUCTION

Chemotherapy,

Combined

therapy.

The Radiation Therapy Oncology Group (RTOG) has pursued active clinical research in nonsmall cell lung (NSCLC) cancer for the past 2 decades. This narrative will summarize the evolution of clinical research in NSCLC over that period and illustrate how the results of this research has led to a paradigm shift in NSCLC treatment strategies using radiation therapy (RT) alone or combined with chemotherapy (CT) and other adjuvants. It will review currently active studies and outline perspectives for future RTOG research efforts in NSCLC. The following is an outline of the major themes of RTOG NSCLC research.

I. Unresectable NSCLC A. Dose intensification of radiation alone 1. Optimal schedule of standard fractionation RT a. Conventional techniques b. Three-dimensional (3D)/conformal techniques 2. Optimal schedule of altered fractionation RT a. Hyperfractionation b. Accelerated fractionation B. Treatment intensification using combined modality RT and CT 1. Induction CT followed by standard RT ? CT

work was supportedin part by the National CancerInstitute, National Institutesof Health, Department of Health and Human Services under Grants CA21661 and CA32115.The author wishesto thank JamesD. Cox, M.D., for his editorial advice and years of wise counsel.In addition, the author recognizesthe years of hard work and effort put

forth by all the RTOG clinical investigators,statisticians,administrators,clinical researchassociates,and many others responsiblefor the successfulcompletionof the studiesdescribed in this manuscript. Accepted for publication 17 February 1995.

Acknowledgements-This

1513

I. J. Radiation

1514

Oncology

0 Biology

l Physics

2. Concurrent CT and standard fractionation RT 3. Concurrent CT and hyperfractionated RT C. Noncytotoxic adjuvants to RT 1. Hypoxic cell sensitizer (misonidazole) 2. Nonspecific immune stimulant (levamisole) 3. Biologic response modifier (recombinant interferon-beta) D. Ancillary studies 1. Pulmonary effects of RT 2. Quality of life issues 3. Prognostic factors II. Resectable NSCLC A. RT ? CT adjuvant to surgery 1. Preoperative RT ? CT 2. Postoperative RT ? CT B. Biomarkers as prognostic indicators C. Chemoprevention of second primary NSCLC

UNRESECTABLE

NONSMALL CANCER

CELL

LUNG

The therapeutic dilemma posed by unresectable nonsmall cell lung cancer (NSCLC) is well known. About one-third of the approximately 135,000 new cases of NSCLC now diagnosed yearly, will have nonmetastatic, but locally extensive, tumor or be medically inoperable (3). It remains a brutal fact that the outlook for unresectable NSCLC, treated with radiation therapy alone, has been rather poor with reported 5year survival rates of 3 to 9%. This patient group has been and continues to be a major research focus for RTOG. The accompanying Table 1 describes each protocol in tabular form and can be used as a reference guide to protocols for both resectable and unresectable NSCLC.

DOSE

INTENSIFICATION ALONE

OF RADIATION

Many of the RTOG unresectable NSCLC trials have focused on intensification of RT alone, either with standard or altered fractionation. A protocol “flow” diagram is depicted in Fig. 1 showing how the results of Phase I/ II trials logically directed the design of subsequent Phase III studies. Reference to this diagram during the discussion below will help clarify the evolution of research strategy. Optimal schedule of standard fractionation RT The first RTOG NSCLC trial, 73-01, evaluated standard fractionation and showed the importance of local tumor control in achieving survival enhancement (37). The 73-01 trial established a dose-response relationship between local control and survival using escalating total doses from 40 to 60 Gy, delivered as 2.0 Gy daily fractions, 5 days per week (40). With over 500 cases studied,

Volume

32,

Number

5. 1995

this trial became the standard of comparison for future lung trials. In failure pattern analysis conducted after long term follow-up, there were fewer in-field failures as components of initial failure for the 50 and 60 Gy dose groups and the median time to failure was longer for 60 Gy (19 months) than for 40 Gy (8 months) with p = 0.017 (39). Looking at ultimate in-field failures, there was a significant dose-response effect. In-field failure was 58% for 40 Gy and only 35% for 60 Gy. This benefit persisted at least to 60 months of follow-up. A short-term survival advantage was observed for 60 Gy compared to the lower dose arms, but this did not translate into a long-term survival benefit. Survival up to 3 years after treatment was better for the patients receiving 60 Gy (approximately 15%) than for those treated to lower doses; 3-year survival was 10% for 50 Gy, and 6% for 40 Gy (split or continuous). After 3 years, all dose groups had the same survival (6%). With the demonstration of the relationship of dose to local control and local control to survival, a major goal of subsequent RTOG trials of RT over the next 15 years was improvement of local tumor control in inoperable NCSCL through intensification of the biologic effectiveness of the radiation treatment schedule. Optimal schedule of altered fractionation After RTOG 73-01, one approach to dose intensification was the use of altered fractionation (11). Radiobiologic elucidation of both acute and late radiation effects on normal tissue and tumor cell kinetics during treatment laid the groundwork for such departures from “standard” fractionation (10, 22-24, 27, 32, 41, 51, 54). A crucial element of strategies to deliver effectively higher total doses of radiation without increasing normal tissue toxicity has been recognition that there is not a one to one relationship between acute and late normal tissue effects, and that, of the two, the most significant is late effects. Altered fractionation implies any deviation from “standard” fractionation of 1.8 to 2.0 Gy once daily, 5 days per week, for 6-7 weeks. Two forms of altered fractionation, hyperfractionation (Hfx) and accelerated fractionation (Afx), or variations of these two, were studied by RTOG. Hype@-actionation. Hyperfractionation is characterized by more than one fraction daily using a fractional dose smaller (l.l- 1.5 Gy per fraction) than used with “standard” fractionation and with the total dose delivered in about the same overall time as standard fractionation, namely, 6 to 8 weeks. Most tumor cell populations, as well as normal tissues with a rapid turnover rate (mucous membrane, gastrointestinal epithelium, bone marrow, etc.), display enhanced cell kill with Hfx. Normal tissues, such as spinal cord, subcutaneous tissues, and bone, are responsible for late normal tissue effects (myelitis, fibrosis, necrosis, etc.), which appear beyond 3 months after radiation. With hyperfractionation, the increased total

RTOG

NSCLC

protocols l

dose is intended to improve tumor cell kill, without increasing late toxicity, but while expecting and accepting increased, but recoverable, acute toxicity. Withers and others have shown a significant, reproducible difference in the capacity of late and early responding tissue to repair radiation cellular damage (32, 53, 54). The late-responding tissues efficiently repair sublethal radiation changes, that is, damage that could lead to cell death, if not repaired. This repair can be accomplished in 6-8 h, especially when the fraction size is smaller than standard fractionation. The multiple small fractions per day deliver a higher total daily dose and final total dose while sparing increased late complications. To avoid increasing the overall treatment time beyond 6-7 weeks, the smaller fractions are usually delivered two or three times daily. This avoids the problem of tumor cell repopulation that can occur during a protracted course of radiation, resulting in reduced tumor control (27,53). Although the clinically detectable tumor may be diminishing in size during treatment, this may be a poor reflection of biologic activity of the viable tumor population. Repopulation of tumor cells may overcome radiation tumor cell killing effects if treatment is prolonged, either by alteration of the fractionation schedule or unplanned treatment interruptions. Fractionation schedules with total treatment times longer than “standard” fractionation, may result in reduced tumor control rates as a consequence of accelerated repopulation. Thus, several different types of multiple daily fraction schedules with shorter total treatment times, defined as accelerated fractionation, were designed to overcome repopulation and result in enhanced tumor control. Both Hfx and Afx exploit reassortment of tumor cells into more radiosensitive phases of the cell cycle, as well as reoxygenation of less sensitive hypoxic tumor cells into more radiosensitive euoxic cells. Hyperfractionation may overcome accelerated repopulation to some extent by enabling a higher total dose to be delivered to the tumor in the same total treatment time as “standard” fractionation. Accelerated fractionation delivers the same total dose in a shorter overall time and should, in theory, overcome accelerated repopulation even more efficiently. Pilot studies of Hfx began in the late 1970s. A two institution pilot (RTOG 77-04) was designed to identify the appropriate fraction size of twice daily treatment of squamous cell carcinoma of the upper aero-digestive tract (33). It revealed that 1.5 Gy per fraction, twice daily, produced severe acute radioepithelitis (mucositis) requiring treatment interruption, whereas, 1.25 Gy twice daily was tolerated to 60 Gy without a break. Based on these observations, the initial Phase I/II study of two fractions of 1.2 Gy per day for NSCLC (RTOG 81-08) evaluated the acute toxicity of Hfx by step-wise escalation of the total dose from 50.4 to 60, 69.6, and 74.4 Gy. The primary tumor and regional nodes received 50.4 Gy at 1.2 Gy twice daily, with 4-6 h between frac-

R. W. BYHARDT

1515

tions, then the primary tumor was boosted through a reduced volume to the assigned total dose. The preliminary results of the study demonstrated successful multiinstitutional participation in a Hfx trial and an acceptable level of acute and late effects (48). After minimum follow-up of 5 years, in a final report of RTOG 81-08, the control arms of two RTOG Phase III studies conducted during the same period, protocols 78-l 1 and 79-17, were used for comparison to the 120 Hfx patients (13). Of 275 cases, 245, receiving 60 Gy in 6 weeks at 2.0 Gy per fraction, were evaluable and matched the stage distribution and pretreatment characteristics of 81-08. Only 79 patients that had been assigned to 69.6 Gy constituted a large enough cohort in which to analyze survival; 5 (6.3%) lived beyond 5 years, but if 30 Stage IV patients were eliminated 8.3 + 4.0% survived 5 years. All had Kamofsky performance status (KPS) of 80 or better. This compared to 5.6 5 1.5% for Stage II and III patients treated with standard once-daily irradiation. In 1983, RTOG began a Phase I/II total dose seeking hyperfractionation trial for unresectable NSCLC designed to identify the maximum tolerable dose of hyperfractionated irradiation and to evaluate tumor control at each dose level (14). Twice daily doses of 1.2 Gy were selected based on the pilot studies noted above and on single institution experience with twice daily fractionation for locally advanced squamous cell carcinoma of the upper aerodigestive tract, as there were no other pilot trials of hyperfractionation for NSCLC. Because important end points for RT are late effects and prolonged tumor control, the typical approach to dose escalation used in Phase I drug trials seeking a maximum tolerated dose was not used because it would have been too time consuming. Instead, RTOG developed a new dose-escalation design, Phase I&I dose searching, to determine: (a) the maximum tolerated dose of Hfx RT for normal tissues and (b) the effect of escalating total doses on local tumor control (17). Accrual of sufficient numbers of patients at several doses above the standard dose level permitted an adequate number of patients to be at risk over a long enough time to assess late effects. Previously untreated patients with American Joint Committee on Cancer (AJCC) Stage II-IIIB nonsmall cell lung cancer with a KPS of 50 or greater were eligible for RTOG 83-l 1. By 1987,884 patients had been randomized to receive either 60.0, 64.8, or 69.6 Gy. After sufficient time had elapsed to observe for late effects at the lowest dose arm, accrual was closed to that arm and patients were randomized to 74.4 Gy and then to 79.2 Gy. Fractions of 1.2 Gy were given twice daily 4 to 8 h apart, 5 days per week. The primary tumor and regional lymphatics were treated to 50.4 Gy and then the primary tumor only was boosted to the total dose assigned at randomization. All deaths were scored as treatment failure. Acute toxicities were defined as those occurring within 90 days of

tc &

Q: How do 60 Gy StdRT, best Hfx regimen, and best CT fb StdRT regimen compare? Closed l/92; 490 pts Q: Which is better for N2 operable: ind. CT fb SG or ind. CT fb RT? Closed 4194; 79 pts

Q: Does total dose effect tumor control/survival? Closed 8/78; 481 pts Q: What is tolerable dose of RT given bid? Q: Does addition of a hypoxic cell sensitizer improve local control/survival? Closed 7/83; 117 pts Q: Can bid hyperfractionation be done groupwide? What doses tolerable? Closed 5/83; 125 pts Q: What is maximum tolerable bid dose? What is impact local control and survival? Closed 1l/87; 884 pts Q: What is tolerance and efficacy of concomitant boost to high dose qd? Closed 5/85; 59 pts Q: Does levamisole improve survival when added to Std. RT vs. RT alone? Closed 2183; 285 pts Q: Is concomitant boost accelerated fractionation (Afx) tolerable?: Does it impact survival? Closed l/89; 363 pts Q: What is tolerance/efficacy of CT fb CT/StdRT to 6 1.2 Gy? Closed 5/89; 30 pts

Study objectives

60 Gy vs. 60 Gy + levamisole (nonspecific immune stimulant) tolerated OK, but no improvement in survival vs. RT alone. Dose intensity escalation study; 63 Gyl5 Wks, 70.2 Gyl5.5 Wks, 70.2 Gy/5 Wks Acute toxicity increased; late toxicity same. Best survival: MeS = 9 Mos; 1 YrS = 40%; 2 YrS = 22%. C x 5 with StdRT to 61.2 Gy startingDay 50. Acute toxicity; 21/30 (70%) had grade3 or higher hematologic; 1 grade4 esophagitis. MeS = 16 Mos; 1 YrS = 68%; 2 YrS = 34%. ComparedStd.RT (60 Gy) to bestarm from 83-11 (69.6 Gy bid) to CALGB 84-33 CT/RT arm (inductionVbl/C fb Std. RT to 60 Gy). Toxicity acceptable;statisticallysuperior1 year survival (60%) andmediansurvival (14 m) with induction CT fb RT. CpNbl/Mito x 2 given prior to either SG or RT fb two more cycles of CT. Accrual slow; replacedby a different strategy intergrouptrial (93-09).

RTOG 83-21 Phase III cII-IIIB (inop) RTOG 84-07 Phase I/IIde cII-IIIB (inop) RTOG 88-04 PhaseIA cII-BIB (inop) RTOG 88-08 PhaseIIIig cII-IIIB (inop) fvPS RTOG 89-01 PhaseIIIig pIIIA (N2)

RTOG 83-12 Phase I/II cII-IIIB (inop)

Dose escalation study; started at 60, 63, and 69.6 Gy. As acute toxicity shown to be OK, to 74.4 and 79.2 Gy. Best results 69.6 Gy, favPS: MeS = 13 Mos; 1 YrS = 56%; 2 YrS = 29%. Gave 75 Gyl28fxl5.5 Wks; 1.8 Gyld to primary and nodes + concomitant boost of 8.8 Gy to primary. MeS = 10 Mos; 1 YrS = 41%; 2 YrS = 25%; 3 YrS = 18%; 5 YrS = 4%.

Tested large fraction RT of 6 Gy twice weekly to 36 Gy/3 Wks vs. 36 Gy/3 Wks + misonidazole; showed no difference in local control or survival with misonidazole added to hypofractionation. HfxRT to 60 Gy vs. 63.Gy vs. 69.6 Gy bid; all three dose levels tolerable; doable groupwide.

40 Gy vs. 50 Gy vs. 60 Gy Local control: 48%, 65%, 61% Median survival: 7 Mos, 9 Mos, 10 Mos 1.25 Gy bid vs. 1.5 Gy bid; severe acute reactions with 1.5 Gy bid

Study and Findings

RTOG 83-l 1 Phase I/IIde cII-IIIB (inop)

RTOG 81-08 Phase I/II cII-BIB (inop)

RTOG 73-01 Phase III cII-IIIB (inop) RTOG 77-04 Phase I/II RTOG 79- 17 Phase III cII-IIIB (inop)

Study, phase, stage, perf. S.

Table 1. Evolution of progress in RTOG nonsmall cell lung cancer protocols

A: Afx concomitantboostgives acceptable increasein acute toxicity, with no increasein late toxicity comparedto qd; no increasein short term S; long term S improved. A: PreRT CT plus concurrentCT and StdRT causesa moderateincreasein acute toxicity, but with encouragingincreasein survival. Basisfor future PhaseIII trial. A: Early resultsshowinduction CT fb RT superiorto standardRT and Hfx RT, with acceptableacute/latetoxicity. Need longer FU. A: Incompletestudy. Toxicity analysis ongoing.

A: Acute toxicity OK. 3/46 (6%) severelate effects; rest Grade 1 or 2. Survivals comparefavorably with Hfx RT and chemoradiotherapy. A: Levamisoledoesnot improve resultsof RT for inoperableNSCLC.

A: Acute and late toxicity of bid acceptable; local control and survival appearbetter than qd RT.

A: Up to 69.6 Gy doableand tolerableand testablegroupwide.

A: Misonidazolenot effective when addedto hypofractionatedRT.

A: 1.25Gy bid toleratedbest.

A: Yes. 60 Gy better than <60 Gy.

Study conclusions

2

Tested 69.6 Gy bid concurrent with oral etoposide and cisplatin; evaluated tolerance and efficacy. 57% Grade-4 hematologic-toxicity; MeS 17 mos.; 1 YrS = 67%; PFS = 52%.

RTOG 91-06 Phase I/II cII-IIIB (inop) fvPS

RT

Compares Vbl/C fb StdRT (60 Gy) to Vbl/C + Std. RT (60 Gy) to C/oral E + HfxRT (69.6 Gy). Will measure survival improvement.

RTOG 94-10 Phase III cIIIA-B (inop) fvPS

accruing OK.

accruing slowly.

future arm for

regEnYrS for

Afx = accelerated fractionation; fx vinblastine; E = etoposide; Mito = = months; Yr = year; S = survival; = number; Q = question; and CS =

A: Open 7194.

A: Open 6194.

A: Open by 7/95 (est.).

A: Open 4194.

A: Open 10/94.

A: Pilot study; near completion; an RTOG Phase III trial?

A: Increased acute toxicity with combined imen. Too early to assess late effects. couraging early survival results 67% 1 and 17 Mos. at MeS. Potential arm phase III. A: Pending longer FU and analysis.

A: Study ongoing;

A: Study ongoing;

A: Increased, but acceptable acute toxicity. Late toxicity not increased. MeS does not appear better than StdRT, but 1 and 2 YrS compares well to induction CT fb RT for CALGB 8433 (55% and 24%) and for RTOG 88-08 (60% and ?). Use for next phase III? A: Study ongoing; accruing well. Results pending.

Abbreviations: NSCLC = nonsmall cell lung carcinoma; RT = radiation therapy; Std. = standard fractionation; Hfx = hyperfractionation; = fractions; d = day(s); Wks = weeks; qd = once daily; bid = twice daily; tid = thrice daily; CT = chemotherapy; C = cisplatin; Vbl = mitomycin; Cra = 13 cis-retinoic acid; FU = follow-up; pts = patients; PS = performance status; f v = favorable; ufv = unfavorable; mos. Me = median; fb = followed by; inop = inoperable; c = clinical; p = pathologic; r = randomized; de = dose escalation; ig = intergroup; n companion study.

Will test various tissue derived tumor markers (K-ras, ~53, etc.) for NSCLC. Can be on other RTOG studies or not.

40 patients to be tested at 60, 66, 72, and 78 Gy. Will look at toxicity and local control/survival.

Uses StdRT to 45 Gy + CT (C/E) fb either surgery or additional to 60 Gy.

Looks at combined concomitant boost and accelerated hyperfractionation; tid RT to 79.2 Gy; looks at acute and late tolerance and tumor effects. Will test StdRT to 60 Gy with and without recombinant interferonbeta IV three times/week in weeks 1, 3, and 5 of RT.

RTOG 94-09 CSig

RTOG 92-05 Phase I/II cI-IIIB (inop) RTOG 93-04 Phase III cII-IIIB (inop) UfVPS RTOG 93-09 Phase IIIig pIIIA (N2) RTOG 93-l 1 Phase I/IIde ~11-111 (inop)

Compares induction CT plus concurrent CT/StdRT (60 Gy) with concurrent CT/HfxRT (69.6 Gy). Study ongoing; to close 3/94.

CT(C/E) plus postop StdRT to 50.4 Gy vs. 50.4 Gy alone for all operated pts with Nl or N2 disease.

RTOG 91-05 Phase IIIig pII/IIIA

RTOG 92-04 Phase IIr cII-IIIB (inop)

PIT’s and serum markers (surfactant and procollagen III) are obtained pre-, post-RT and q 3 mos. x 1 year, then yearly.

RTOG 91-03 cs

Q: What is the comparative efficacy of the 88-04 and 91-06 regimens? Closed 3/94; 168 pts Q: Is tid AHfx tolerable and efficacious? Open 12193; 25143 pts Q: What is effect on survival of adding recombinant interferonbeta to Std. RT? Open 10/94; target n = 168 pts Q: Compare induction CRx + RT fb either surgery or more RT. Open 4/94; target n = 612 pts Q: What is toxicity and efficacy of escalating doses of RT to primary via 3-D/conformal RT? Pending final revision. Q: Evaluate prognostic import of marker studies in NSCLC. Open 6/94; target n = 465 pts Q: Compare sequential CT fb RT to concurrent CT/RT and to concurrent CT/HfxRT. Open 7/94; target n = 597 pts

Compares 13 Cra to placebo in Stage I lung ca 6 Wks to 6 mos after curative resection. Testing concept of “chemoprevention.”

RTOG 91-01 Phase III PI

Q: Does 13-cisretinoic acid prevent second primary lung cancers cured lung cancer? Open 12/93; 315/1290 pts Q: Evaluate pulmonary fts. before, during and after RT. Open 12/93; 44/187 pts Q: Is CT + RT postop in operated Nl-2 pts better than postop. RT alone? Open 12/93; 185/465 pts Q: What is tolerance of concurrent HfxRT and oral E + C? Closed 7192; 79 pts

Tests 69.6 Gy bid (best arm from 83-l 1) with concurrent cisplatin + vinblastine (CALGB 84-33). 44% grade 4 or worse acute hematologic toxicity. No increase late toxycity. MeS = 12.2 Mos.; 1 YrS = 54%; 2 YrS = 28%; 76% ufvPS.

RTOG 90- 15 Phase I/II cII-IIIB (inop)

Q: Is concurrent CT and HfxRT tolerable? Closed 10/91; 42 pts

1518

I. J. Radiation Oncology 0 Biology 0 Physics

Standard

Frsctionation

73-01; Phase III 40 vs. 50 vs. 60 Gy Std.RT :. 60 Gy best LC

Hyperfractionation

Accelerated

fractionation

77-04: Phase I/II tested fx. size bid :. bid fx of 1.25 Gy

83-12; Phase I/II synch. c. boost to 75 Gy :. act. toxicity; 2 YS = 25%

81-08; Phase I/II DE to 74.4 Gyi1.2 Gy bid :. WC. toxicity of bid

84-07; Phase I/II asynch. c. boost to 70.2 Gy :. ax. toxicity; 2 YS = 21%

u 78-11179-17 conml :. additionaJ data for Std. RT (60 Gy) u 93-11; Phase VII DE (in planning)

arms

11 83-11; Phase U/-II DE to 79.2 Gyl1.2 Gy bid :. xc. toxicity and best LC at 69.6 Gy favorable performance status

II 92-05; Phase VTI 79.2 Gy at 1.1 Gy tid (results pending)

II 88-08; Phase III Std. RT (60 Gy) f CT vs Hfx RT (69.6 Gy) . survival best with Std. RTandCT

Fig. 1. “Flow“ diagram of RTOG protocols investigating dose intensification of radiation therapy alone in unresectable nonsmall cell carcinoma of the lung using variations of standard and altered fractionation. Abbreviations: Std. = standard; fx = fraction; bid = twice daily; synch. synchronous; c. boost = concornmitant boost; :. = conclusion; LC = local control; act. = acceptable; YS = year survival; DE = dose escalation; asynch. = asynchronous; tid = three times daily; CT = chemotherapy; Hfx = hyperfractionation.

the start of treatment and late if they appearedor persisted past 90 days. Toxicity grades ranged from 0 (none) to 5 (fatal) and 3,4, and 5 were scored as “major”. To facilitate comparison to the results of the Cancer and Leukemia Group B (CALGB) protocol 84-33 (18), all patients were restaged according to the 1984 AJCC classification, and 350 patients were identified who had CALGB “favorable” performance status. The favorable subgroup had regional stage III (T3 and/or N2, MO) disease, KPS 70100, and weight loss < 5% in the 3 months prior to diagnosis. There was no significant difference between the five arms in pretreatment characteristics or incidence of acute or late toxicity, despite a 30% difference in total doses. There was no difference in incidence of all types of grade 3 or greater late toxicity among the three highest dose arms and no difference in grade 3 or greater toxicity according to total dosesbetween the favorable and unfavorable subgroups. There was no difference in survival according to assigned total dose. The “favorable” subgroup of 350 patients had higher 12- and 24-month survival rates with 69.6 Gy than with lower doses(p = 0.02), but there were no survival differences among the three highest doses. The median (13 months), l-year (56%) and 2-year (29%) survival rates for Hfx to 69.6 Gy, in the favorable subgroup, appeared superior to the benchmark standard fractionation results. A dose-survival benefit was suggestedfor 69.6 Gy Hfx in patients with favorable pretreat-

Volume 32, Number 5, 1995

ment characteristics. In comparison, favorable performance patients treated with standard RT (60 Gy in 30 fractions) on the control arm of RTOG protocol 83-21 had a median survival of 8.9 months, l-year survival of 34%, and 2-year survival of 10% (38). The samestandard RT was used in the control arm of CALGB protocol 8433 and had nearly identical median (8.7 months), l-year (40%), and 2-year (13%) survival rates (18). Long-term follow-up of the entire group of 83-11 patients reveals that the 5-year survival ranged between 6 to 8% for all 5 dose groups (47). The favorable performance subset showed 5-year survival of 12%, 8%, lo%, lo%, and 9% for 60, 64, 69.6, 74.4, and 79.2 Gy, respectively. There was no longer a significant difference in survival between the three lower dose arms (p = 0.09). The survival advantage seen for 69.6 Gy was no longer evident by the 3-year interval. This outcome, especially in light of the 73-01 long-term survival, suggests,again, that intensification of RT hasprolonged short-term survival, perhaps by improving local tumor response, but local and/ or distant failures still occur, although delayed. Maximum local control benefit from Hfx may have been reached at 69.6 Gy to explain the absenceof further survival improvement beyond this dose. Alternatively, lethal treatment-related toxicity mimicking tumor progression may have been scored as tumor-related death and not scored as toxicity. RTOG 83-11 was a notable achievement in terms of the large number of patients studied, the ability of multiple participating hospitals to adopt new and more complicated treatment techniques, and the courage of the investigators to “push the envelope” of total administered dose. Adverse effect of treatment interruptions during Hfx. In a supplemental review of RTOG 83-11, Cox et al. found that 90/848 (11%) patients had treatment interruptions that increased the overall treatment time (12). Of these, 40 were judged “minor,” 21 “major-acceptable,” and 29 “major-unacceptable.” As the assigneddose increased, so did the proportion of patients with treatment interruptions (10% for 60 Gy and 21% for 79.2 Gy). Compared to patients whose treatment duration was “per protocol,” survivals were significantly shorter (p = 0.016) for Hfx patients with any interruption-related lengthening of treatment time. Estimated 2- and 5-year survival rates were 24 and 10% for the “per protocol” patients vs. 13 and 3% for those with treatment interruptions. This difference, like the survival benefit for Hfx treated to 69.6 Gy, was seenonly for favorable prognosis patients (KPS 90-100; weight loss < 5%, and no N3). In addition, the effect of interruption was seenonly with Hfx to 2 69.6 Gy. As noted above and in the prognostic factors section to follow, prognostic factors have an important impact on outcome. Even if a suitably high, uninterrupted total dose is given using Hfx, improved survival, compared to “standard” fractionation, accrues only to “favorable”

RTOG

NSCLC

protocols

prognosis patients. If interruptions prolong delivery of the hyperfractionated total dose, even a suitable total dose and the favorable prognostic group status will not “save the regimen.” This is also relevant to schedules combining CT and RT, because either modality can stimulate a proliferative response in surviving clonogens. If accelerated repopulation does occur, any toxicity-related interruptions of therapy could reduce local-regional tumor control. This would be especially true for sequential treatment, as with induction CT followed by RT, and less likely for concurrent CT and RT. In the latter regimen, the RT may eradicate the proliferative clonogens stimulated by CT and vice versa. However, the greater acute toxicity of the concurrent approach may produce interruptions that obviate the gains provided by concurrent use of both. Relationship of $eld size and pulmonary toxicity using Hfx. As treatment regimens intensify, acute normal tissue toxicity becomes an increasingly significant factor. This may make other variables of the radiation therapy technique, such as field size, take on greater importance. In another supplemental review of RTOG 83-l 1 the risk of overall z grade 2 pulmonary toxicity (greater of either acute or late) was increased for patients treated with field sizes in excess of the protocol-specified margin of 2 cm beyond primary tumor (5). This effect was seen only when the area of the lung treated beyond protocol margins exceeded 35 cm* and when the overall field size was below 180 cm’. Field sizes beyond 180 cm2, either as necessitated by tumor size or anatomic considerations, may already treat a volume of lung beyond a pulmonary toxicity threshold. For smaller fields, < 180 cm2, small increases in the volume of irradiated lung may increase the chances of irradiating a critical threshold number of functional subunits (FSU). When this threshold is exceeded, the incidence of acute toxicity may increase rapidly. This may be more pronounced for regimens like Hfx, which have greater acute pulmonary toxicity effects. Unfortunately, field size is determined by the size of the tumor and many patients with advanced local disease have large tumors requiring field sizes beyond 180 cm2 if adequate margins are to be used to cover microscopic tumor extensions that extend beyond the radiographically visible disease. In this setting, the threshold number of FSUs of normal lung may already be exceeded. Pulmonary toxicity risk may already be high and further enhanced by tumor-related cardiopulmonary effects. A possible solution to the problem of increased pulmonary toxicity risks with large fields may be to combine radiation with systemic cytotoxic treatment to address microscopic tumor extensions and reduce the radiation fields to cover only the gross primary and nodal disease with a minimal margin. This, of course, depends on the CT being effective enough to predictably sterilize microdeposits and microextensions of tumor. Then the CT may

0 R. W.

BYHARDT

1519

also sterilize microextensions outside small radiation fields drawn with narrow margins. Although there has been a recent trend towards using smaller fields of coverage in NSCLC, this concept has not yet been universally accepted in NSCLC. Smaller fields may be useful, however, especially in view of the increased acute toxicity observed when CT is given concurrently with RT. Acceleratedfractionation (Afx). Synchronous Concomitant Boost. RTOG 83- 12 was a Phase I/II trial designed to test a concomitant boost technique as a means to deliver a large daily dose to the tumor volume, with sparing of normal tissue, while giving a higher total dose in a shorter overall time than standard RT (20). Using CT-aided dosimetry, 59 patients with clinical Stage T3-T4, Nl-3 NSCLC received 1.8 Gy to the primary tumor and localregional nodes (to 50.4 Gy) followed immediately by a concomitant boost dose of 0.88 Gy to only the primary tumor and involved lymph nodes (to 75 Gy in 5.5 weeks). Acute toxicity was higher than standard RT, but acceptable, and late toxicity was not increased. In a final report, 2-year survival (25%) and 3-year survival (18%) was comparable to some of the best reports in the literature for altered fractionation (26). It was concluded that with careful dosimetry the length of treatment can be shortened, compared to standard RT, and a biologically high dose of RT can be delivered with acceptable acute and late toxicity. Asynchronous concomitant boost. RTOG 84-07 was a Phase VII trial also testing Afx via concomitant boost, but the “boost” field was treated 4 to 6 h after the “large” field. Between 1984 to 1989, 355 patients with unresectable NSCLC were studied (6). Large fields, encompassing the primary tumor and loco-regional nodes, received 1.8 Gy. The boost field, encompassing only the primary tumor and involved nodes, also received 1.8 Gy. The boost field was treated initially twice weekly to 63 Gy in 5 weeks (61 patients). After sufficient follow-up to determine acceptable acute toxicity, the total dose was escalated to 70.2 Gy in 5.5 weeks (180 patients) and then, later, was further “shortened” to 70.2 Gy in 5 weeks (114 patients). Acute lung and esophagus toxicity 2 grade 3 (7 to 17%) seemed higher than with standard RT, and late toxicity 2 grade 3 was 5 to 9%, which was not different between the three dose schedules and did not seem significantly different from standard RT. There was no difference in survival between the three dose schedules. Although median survival (9 months) and l-year survival (39 to 44%) seemed similar to standard RT, the 2-year survival for 70.2 Gy in 5 weeks was 21%, which seemed somewhat better than that usually reported for standard RT (10 to 13%). For CALGB favorable patients, the 2-year survival rate was 22%, relatively close to the 24% 2-year survival observed in the CALGB CT and RT arm (18). Accelerated hyperfractionation. Fractionation studies for unresectable NSCLC continued with RTOG 92-05,

1520

I. J. Radiation Oncology 0 Biology 0 Physics

which further intensified the biologic dose and shortened the overall treatment time by combining concomitant boost, Hfx, and Afx. It was a logical next step in evaluating escalating dose/time intensity. This approach represented an ‘ ‘Americanized’ ’ adaptation from the Con(CHART) regimen,which was first testedin the United Kingdom, for both headandneck cancerandunresectableNSCLC, with apparent good tolerance and improvement in survival comparedto historical controls (43). In the CHART regimen, RT is given three times per day with 6-h separationover 12 consecutive days to a total dose of 54 Gy in 36 fractions. Instead,92-05 gave 1.1 Gy three timesdaily, 4 h apart (0800, 1200, and 1600 h), 5 days per week, to 79.2 Gy delivered in 5 weeks. The 0800 h and 1600 h treatments covered the primary tumor and both involved and uninvolved locoregional nodes, which received 52.8 Gy, while the 1200 h treatment covered only the primary tumor and any involved nodes,boosting the dose to 79.2 Gy. The regimen was first tested in a limited participation pilot study (28). In early results,the acute toxicity 2 grade 3 hasbeen < 10%. Longer follow-up is required to assesslate effects and survival, but the regimen seemsremarkably well tolerated acutely. 3DKonfomzal radiation therapy. In a return to intensification of standard fractionation RT, a developing Phase I/II trial (93-l 1) will evaluate the use of 3DKonformal treatment techniques using single daily fractions of RT. The main objective of the study will be to evaluate if this technique allows delivery of a high total dose to the primary tumor and involved nodes (Gross Tumor Volume or GTV) with sparing of treatment-related morbidity. The main dose-limiting normal tissue in treatment of lung cancer is the normal lung, which has a very low threshold for permanent injury (approximately 20 Gy with standard RT). Single institution evaluations done with dose-volume histograms and similar studies show that 3D technology provides better coverage of the tumor volume while reducing the volume of normal lung exposed to RT. Preliminary clinical analysesof NSCLC patients treated with 3D technology have shown that reduction in the volume of lung exposed to RT correlates to a reduction in the risk of RT pneumonitis. In 93- 11, conformal fields defined by three-dimesional techniques will cover the primary tumor and loco-regional nodes (Clinical Target Volume 1 or CTVl) with a margin to compensate for variabilities in treatment setup, breathing, or motion during treatment (Planning Target Volume 1 or PTV,) and will receive 50 Gy. Smaller conformal fields will be designed with a minimum l-cm margin around the GTV (CTVJ and a further margin around the CTVZ (PTV2). The PTV? will receive stepwise, escalating boost doses to total dosesas high as 78 to 80 Gy. Escalation to the next highest boost dose will be determined by acute tolerance. Because the evidence is fairly good that total doses biologically more intense than standard RT to 60 Gy are associated with improved local tumor control, study 9311 is important because it may provide a way to deliver

Volume 32, Number 5, 1995 G=

(1

88-04; Phase II ind. CT* th cow. CT* and Std. RT I63 Gy) :. act. toxicity and efficacy

(1 88-08; Phase III Std. RT i CT* vs. Hfx RT (69.6 Gy) _‘_ Std. RT and Cf best survival

u

* u

(1

u

u

u

u

u

u

90-U; Phase IA1 Cont. CT* and Hfx RT (69.6 Gy) :. ?a. toxicity: 2 YS = 28%

91-06; Phase I/TI Cont. CF’ and Hfx RT (69.6 Cry) :. act. toxicity: 1 YS = 67%

u

II

II

u

=a

a

II

I

u u

e=

u u

92-04; Phase II randomized 91-06 vs. 88.04 (To,, early. 1

u 94-10; Phase III Ind. CT’ and Std. RT (60 Gy) vs. Cont. CT’ and Std. RT (60 Gy) vs. Cont. Cl-” and Hfx RT (69.6 Gy) (O”gOl”g,

Fig. 2. “Flow“ diagram of RTOG protocols investigating treatment intensification using radiation therapy and chemotherapy in unresectable nonsmall cell lung cancer. Abbreviations: ind. = induction; CT = chemotherapy; fb = followed by; cont. = concurrent; Std. = standard; RT = radiation therapy; dot triangle = conclusion; act. = acceptable; Hfx = hyperfractionation; YS = year survival; *vinblastine/cisplatin; de1 1 of 2 #‘s cisplatin/ oral etoposide.

those higher doseswith acceptable acute toxicity. This is especially true when one considers that future regimens for unresectable NSCLC will likely include cisplatinbased CT regimens. Induction or concurrent CT plus RT increases the probability and severity of acute toxicity; thus, reducing the volume of normal tissue exposed to RT makes good sense.A major effort in designing 93-l 1 has been standardization of 3D terminology and technique through development of a Quality Assurance document to assure uniform application of 3D/conformal RT at all participating institutions. Treatment intensijkation using combined RT and CT As evidence of effectiveness in NSCLC of cisplatinbasedCT regimens began to accumulate in the late 198Os, RTOG began testing radiation and adjuvant CT for unresectable NSCLC. The evolutionary “flow“ of RT/CT protocols, discussedin the following paragraphs, can be traced by reference to Fig. 2. Induction CT followed by standard RT + CT. RTOG 88-04 was a PhaseII trial that demonstrated that induction vinblastine/cisplatin (VblK) followed by concurrent (VblK) and standard RT to 63 Gy had an acceptable increase in acute toxicity, no increase in late toxicity, and suggestedan improvement in 2-year survival (34%) (45). Thus, RTOG 88-04 establishedthe tolerance of induction and concurrent VbYC plus standard RT to 63 Gy.

RTOG

NSCLC

protocols

Shortly after RTOG 88-04 was reported, the Cancer and Leukemia Group B (CALGB) reported, in a randomized Phase III trial (8433), that induction VbYC followed by standard RT to 60 Gy in 6 weeks provided superior median (13.8 months) and 2-year survival (24%) compared to standard RT alone (8.7 months and 13%, respectively). On the basis of these two trials and the results of 831 1, RTOG launched a three-arm Phase III trial (88-08) comparing hyperfractionated RT (69.6 Gy) and standard RT (60 Gy), alone, or preceded by Vbl/C. A preliminary report of the results of RTOG 88-08 showed median survival (13.8 months) and l-year survival (60%) to be superior with the induction CT regimen to both of the other two arms, alone or combined (46). Taxol and Topotecan, given as single agents or with cisplatin, concurrent with standard RT, are also undergoing Phase I testing at individual RTOG institutions as possible options for future Phase II trials. Concurrent CT and hyperfractionated RT. While 8808 was in progress, RTOG began Phase I/II testing of several strategies combining hyperfractionated irradiation and simultaneous cisplatin-based chemotherapy. The aim was to identify potential regimens for future comparison to the best treatment arm resulting from 88-08. The rationale for concurrent CT/RT was to further enhance local tumor cell kill by taking advantage of both the radiosensitizing and cytotoxic properties of cisplatin. It also tests the hypothesis that accelerated repopulation, which can be stimulated by induction chemotherapy cell kill, might be minimized by concurrent administration of chemotherapy and hyperfractionated irradiation. RTOG 90-15 was the first of a series of studies of hyperfractionated RT, instead of standard RT, concurrent with CT. It showed that concurrent Vbl/C and hyperfractionated RT to 69.6 Gy in patients with inoperable NSCLC, unselected for favorable performance characteristics, resulted in increased, but acceptable, acute toxicity and no increase in late toxicity. Median survival was not higher than results of standard irradiation alone (12.2 months), but l-year (54%) and 2-year (28%) survival rates were similar to those reported for CALGB 8433 (7). RTOG 91-06 was nearly identical in design to 90-15 except Vbl was replaced with oral etoposide (E). Early survival results are encouraging, with a median survival of 17 months and a l-year survival of 67%, the best Phase II RTOG results observed, thus far, for unresectable NSCLC (30). Grade 4 acute hematologic toxicity was high at 60%, but etoposide dose/scheduling adjustments made in the recently completed RTOG 92-04, a randomized Phase II trial comparing the regimens of 91-06 and 88-04, shows improved acute tolerance, but there is not yet sufficient follow-up of 92-04 for survival data. A newly activated, high priority, RTOG three-arm Phase III trial (94- 10) for favorable performance patients evaluates whether concurrent CT and RT are more effi-

0

R. W.

BYHARDI

1521

cacious than induction CT followed by RT. Several reported trials of cisplatin containing regimens plus at least 60 Gy suggest median survivals from 13.5 to 17 months (4, 8, 19, 21,29). RTOG 94- 10 compares Vbl/C followed by standard RT to 60 Gy (the CALGB 8433 regimen), in arm 1, to concurrent Vbl/C and standard RT to 60 Gy (modified 90-15 regimen), in arm 2, and to concurrent C/oral E and hyperfractionated RT to 69.6 Gy (91-06 regimen), in arm 3. Chemotherapy-induced accelerated repopulation has been shown in the laboratory (42, 52). The hypothesis under study in 94- 10 is that the concurrent use of CT and RT in arms 2 and 3 may avoid CT-induced accelerated repopulation by shortening the overall treatment time and may provide some radiosensitization. The 91-06-type regimen used in the third arm further intensifies the concurrent approach with the use of Hfx to a higher total dose (69.6 Gy) than the standard RT to 60 Gy used in arm 2. Noncytotoxic adjuvants to RT Hypoxic cell sensitizer (Msonidazole). In an effort to improve local control in locally advanced NSCLC, in the mid- 1970s RTOG undertook a Phase I/II trial of large fraction irradiation (hypofractionation), 6 Gy twice weekly to 36 Gyl3 weeks, combined with the hypoxic cell sensitizer misonidazole (50). This treatment schedule was chosen to reduce the risk of misonidazole neurotoxicity, by dosing with misonidazole only twice a week, and to be able to give misonidazole with each dose of RT. This trial showed the regimen to be tolerable with objective responses in two-thirds of the patients. On this basis, a Phase III randomized trial was undertaken in 1979, comparing radiation to 36 Gy/3 weeks with either misonidazole or placebo in 117 patients It showed no advantage to the misonidazole plus 36 Gy/3 weeks regimen over 36 Gy/ 3 weeks alone or over standard RT to 60 Gy/6 weeks (49). The incidence of esophageal morbidity with the hypofractionation schedule was significant, producing severe or worse acute reactions in 10% of the patients. Although economic of time, hypofractionation yielded increased morbidity with no improvement in local control. Based on this experience and a review of the results of other hypofractionation trials, RTOG abandoned hypofractionation for NSCLC (9). NonspeciJc immune stimulant (Levamisole). With recognition in the early 1970s of the importance of host immune competence in various cancers, manipulation of patients’ immune status to enhance cell-mediated immunity or host cytotoxic humoral antibody formation against their tumors appeared to be an attractive means of improving local control of irradiated NSCLC. The antihelminthic drug levamisole was found to have immunotropic properties. In animal models, it induced an immunostimulatory response, enhanced the effect of antitumor drugs, and improved survival (25). In 1978, RTOG undertook a single-blind evaluation of the immunostimulatory effect of

1522

I. J. Radiation

Oncology

0 Biology

0 Physics

levamisole, 2.5 mg/kg, vs. placebo on days 1 and 2 of each week of standard RT to 60 Gy/6 weeks for inoperable NSCLC with Kamofsky performance status of 60 or greater. A Phase I trial of levamisole in unresectable NSCLC had shown the 2.5 m&g dose to be within the tolerable range (31). No significant prolongation of survival, progression-free survival, or differences in the patterns of failure between levamisole and placebo was seen. It was concluded that levamisole yielded no benefit to RT for NSCLC (37). Although neither misonidazole nor levamisole showed a benefit, the control arms of the two trials did provide useful outcome data for standard RT to 60 Gy/6 weeks delivered with more modem equipment, custom blocking, and sophisticated dosimetry techniques than had been used in RTOG 73-01. The outcome data from these two trials was pooled with the older data for greater statistical power and used for comparison to subsequent RTOG trials of altered fractionation or standard RT plus chemotherapy. Biologic response modifier (recombinant interferon beta). Another new two arm Phase III trial (RTOG 9304) for unfavorable performance patients will compare standard RT to 60 Gy with and without concurrent intravenous recombinant interferon beta (rIFN-beta,,,). This study attempts to confirm an apparent improvement in survival in IIIA/B patients noted in a pilot study of concurrent rIFN-beta,,, and standard RT in patients unselected for favorable performance status (35). Other RTOG institutions are testing another form of rIFN-beta given concurrently with RT after induction Vbl/Cl in a limited participation pilot study in patients with IIIA/B disease. Ancillary studies in unresectable NSCLC Pulmonary function effects of RT. A current study, RTOG 91-03, evaluates the pulmonary function effects of curative irradiation and two serum markers of radiation pulmonary injury, surfactant apoprotein, and procollagen type III. Prognostic factors related to treatment outcome. Several analyses of completed RTOG NSCLC studies have provided valuable insight into the complex relationship of prognostic factors and various treatment outcome endpoints. RTOG investigators were among the first to summarize the factors related to survival by the use of multivariate regression analysis of both RTOG and other published clinical trials to identify variables jointly related to survival within disease subgroups. They identified that prognostic factors could be used to define subsets of patients with similar durations of survival. Using these homogeneous subsets, more sensitive treatment comparisons are possible. For example, using this approach, long-term follow-up data from RTOG 73-01 was analyzed (2). Although the overall duration of survival was similar for each of the four assigned treatment schedules, with a median survival

Volume

32. Number

5, 1995

of 9 months, logistic regression analysis showed KPS and prior weight loss as significantly predicting survival at 6 months. These two factors divide patients into a “poor“ risk group (KPS < 90; weight loss > 5%) with a 6month median survival, and a “good“ risk group (KPS 2 90; weight loss 5 5%), with a median survival of 10 months @ = 0.001). Further, the factors significantly related to survival at 12 months were both local and distant tumor progression before 6 months. Only histologic type and primary site failure were related to survival at 18 months. Adenocarcinoma, regardless of local control status, had the best prognosis and large cell, with local progression, had the worst prognosis. Recursive partitioning analysis according to prognostic variables was performed on 1,592 inoperable NSCLC patients treated on four RTOG studies (83-l 1, 83-21, 8403, and 84-07) (44). Homogeneous survival subsets were identified. Variables analyzed included eight pretreatment characteristics and one treatment parameter. Univariate analysis suggested KPS (5 70 vs. > 70), pleural effusion, weight loss (5 5% vs. > 5%), age (2 60 vs. < 60), T stage, (Tl and T2 vs. T3 and T4), and N stage (NO vs. N+), were important prognostic factors. Grade, radiation dose, and histology were not univariate prognostic factors. By recursive partitioning, KPS was the most significant covariate; for KPS 5 70, median survival was 5.9 months, whereas, for KPS > 70, median survival was 9.9 months. Overall, median survival for the whole group was 8.9 months with 16% alive at 2 years. For patients with KPS 5 70, N stage, age, weight loss, and “RT dose not received“ emerged as important prognostic factors, while for KPS I 70, only pleural effusion was prognostic. Because median survival ranged from 5.9 to 12.2 months, depending on prognostic grouping, it is quite apparent that prognostic factors play as important a role in outcome as does choice of therapy. Thus, for proper design of clinical trials that test any new treatment in inoperable NSCLC, appropriate stratification of prognostic factors is essential to assure their uniform distribution in all treatment groups. For example, if the experimental arm of a Phase III NSCLC trial has more “poor“ risk patients than the control arm, a potentially effective regimen might be rejected (false negative). Conversely, if it has more “good“ risk patients than the control arm, then a potentially ineffective regimen might be accepted (false positive). As more has been learned about the pathophysiology and molecular biology of NSCLC, important biomarkers have been identified that have the potential to refine our prognostic abilities. Mutational inactivation of tumor suppressor genes, such as ~53, or expression or activation of certain oncogenes, such as K-ras, are important in the pathogenesis of NSCLC. Also, expression of blood group antigen A and the epidermal growth factor receptor (EGFR) can correlate to a favorable prognosis. Other potentially important factors include neuroendocrine differ-

RTOG

NSCLC

protocols

entiation, ~105 nuclear antigen levels, and tumor vascularity. As noted below, RTOG is participating in a large intergroup trial (0 115; RTOG 94-09) evaluating these factors by analysis of tissue obtained at surgery in resectable NSCLC. RESECTABLE

NSCLC

RT 2 CT as an adjuvant to surgery Preoperative RT + CT in stage IIIA NSCLC. Several recently activated RTOG studies evaluate the role of adjuvant CT/RT in NSCLC patients with potentially resectable disease.RTOG; 93-09 is participating in a Phase III intergroup trial for patients with IIIA disease; compares induction CT with concurrent Vbl/C and standard RT to 45 Gy followed by either surgical resection or boost RT to bring the total dose to 60 Gy. The trial includes a quality of life assessmentand was developed by combining two similar prior studies of SWOG and RTOG. The SWOG study, 8805, of 126 patients showed acceptable tolerance with the trimodality approach. There was a 30month median survival in patients with no pathologic node involvement at surgery vs. 10 months with involved nodes (1). The RTOG study, 89-01, was a PhaseIII comparison in 75 patients of induction chemotherapy (vinblastine/cisplatin/mitomycin) followed by either surgery or radiation therapy. It also showed acceptable tolerance. Survival analysis is ongoing. Postoperative RT + CT in stage II or IIIA NSCLC. RTOG is also participating in a Phase III intergroup trial (91-05) comparing postoperative RT (50.4 Gy) to postoperative RT (50.4 Gy) and CT with C/E for patients with resected Stage II or IIIA NSCLC. The role of RT alone in borderline-resectable (N2) disease. Prior to the 1988 reclassification of Stage III into IIIA and IIIB in the AJCC (ASCC) staging system, some of the clinical Stage III (N2) NSCLC patients, now considered marginally resectable, were treated with RT alone, someon RTOG protocols. To place current trials of adjuvant RT/CT before or after surgery into perspective, RTOG 83-l 1 results with Hfx and earlier RTOG studies of standard RT to 60 Gy were reanalyzed only for the patients in Stages Tl-3N2. The 3-year survival rate for IIIA patients with favorable performance status was 7%

0 R. W.

1523

BYHARDT

for standard RT compared to 20% for Hfx to 69.6 Gy (15). Three-year survival rates for surgery alone were 26% in 426 patients reported by Naruke et al. (36) and for CT and surgery were 34% in 41 patients reported by Martini et al. (34). However, the Martini et al. study had only 22% T3 tumors, whereas the Naruke et al. study and the two RT series had around 40% T3 tumors. Also, weight loss and performance status data were not given in the surgical series. Biomarkers as prognostic indicators. RTOG 94-09 examines the prognostic import of tissue markers (~53, Kras, etc.) from tissue specimens obtained at surgery for patients entered on the postoperative CT/RT trial, RTOG 91-05, noted above. Chemoprevention of second primary NSCLC Based on observations of the role of retinoids in differentiation (16), an intergroup study is exploring chemoprevention. This Phase III trial (RTOG 91-01) evaluates the role of 13 Cis-retinoic acid in prevention of second primary lung cancers following curative treatment of early stage disease.This trial has randomized over half of the targetted 1200 patients necessary to determine if the second malignancy rate can be altered. CONCLUSIONS Over the past 20 years, RTOG has been responsible for a succession of clinical studies that have logically evolved, one from the other, and have yielded important information regarding the factors influencing the successful treatment of NSCLC. For unresectable NSCLC, the major themes have included dose intensification of radiation alone, using both standard and altered fractionation, treatment intensification using combined modality RT and CT, and noncytotoxic adjuvants to RT. For resectable NSCLC, both pre- and postoperative RT + CT have formed the basisfor studies attempting to reduce local and distant failures in high risk patients. Chemoprevention of second primary NSCLC in curatively resected patients is also being tested. There are currently nine active protocols for NSCLC covering a range of NSCLC stagesin patients with both favorable and unfavorable performance status characteristics.

REFERENCES 1. Albain, K.; Rusch, V.; Crowley, A.; Turrisi, T.; Rice, T.; Livingston, R. Concurrent cisplatin/etoposide (PE) + chest radiation (CRT) followed by surgery for stages 3A (N2) and 3 B nonsmall cell lung cancer. Completed analysis of SWOG-8805 (Abst. 1120). Proc. ASCO 13:337; 1994. 2. Bauer, M.; Birch, R. Pajak, T. F.; Perez, C. A.; Weiner, J. M. Prognostic factors in cancer of the lung. RSNA Categorical Course in Radiation Therapy, 87-l 12; 1984. 3. Boring, C. C.; Squires, T. S.; Tong, T.; Montgomery, S. Cancer statistics, 1994. CA-A Cancer J. Clin. 44:7-26; 1994.

4. Breneman, J. C.; Mitchell, S. E.; Hawley, D. K.; Aron, B. S.; Schroder, L. E. Concurrent radiotherapy and chemotherapy for locally advanced nonsmall-cell cancer of the lung. Am. J. Clin. Oncol. 14:9-15; 1991. 5. Byhardt, R. W.; Martin, L.; Pajak, T. F.; Shin, K. H.; Emami, B.; Cox, J. D. The influence of field size and other treatment factors on pulmonary toxicity following hyperfractionated irradiation for inoperable nonsmall cell lung cancer (NSCLC)-Analysis of a Radiation Therapy Oncology Group (RTOG) protocol. Int. J. Radiat. Oncol. Biol. Phys. 27537-544; 1993.

1524

I. J. Radiation

Oncology

l Biology

l Physics

6. Byhardt, R. W.; Pajak, T. F.; Emami, B.; Herskovic, A.; Doggett. R. S.; Olsen, L. A. A phase I/II study to evaluate accelerated fractionation via concomitant boost for squamous, adeno, and large cell carcinoma of the lung: Report of Radiation Therapy Oncology Group 84-07. Int. J. Radiat. Oncol. Biol. Phys. 26:459-468; 1993. 7. Byhardt, R. W.; Scott, C. B.; Ettinger, D. S.; Curran, W. J.; Doggett, R. L. S.; Coughlin, C.; Scarantino, C.; Rotman, M.; Emami, B. Concurrent hyperfractionated irradiation and chemotherapy for unresectable non-small cell lung cancer. Results of Radiation Therapy Oncology Group (RTOG) 90-15 (Abst.). Lung Cancer 11:188; 1994. 8. Coughlin, C. T.; Del Prete, S. A.; Grace, M. P.; O’Donnell, J. F.; Quackenbush, L. Cisplatin and radiation therapy for locally advanced squamous cell carcinoma of the lung. Cancer Treat. Rep. 70643-645; 1986. 9. Cox, J. D. Large-dose fractionation (hypofractionation). Cancer 55:2105-2111; 1985. 10. Cox, J. D. Presidential address: Fractionation: A paradigm for clinical research in radiation oncology. Int. J. Radiat. Oncol. Biol. Phys. 13:1271-1281; 1987. 11. Cox, J. D.; Azarnia, N.; Byhardt, R. W.; Perez, C. A.; Fu, K.; Spunberg, J.; Sause, W. T. Altered fractionation for nonsmall cell carcinoma of the lung. Chest 96:68S; 1989. 12. Cox, J. D.; Pajak, T. F.; Asbell, S.; Russell, A. H.; Pederson, J.; Byhardt, R. W.; Emami, B.; Roach, M.. Interruptions of high-dose radiation therapy decrease long-term survival of favorable patients with unresectable nonsmall cell carcinoma of the lung: Analysis of 1244 cases from 3 Radiation Therapy Oncology Group (RTOG) trials. Int. J. Radiat. Oncol. Biol. Phys. 27:493-498; 1993. 13. Cox, J. D.; Pajak, T. F.; Herskovic, A.; Urtasun, R.; Podolsky, W. J.; Seydel, H. G. Five year survival after hyperfractionated radiation therapy for nonsmall cell carcinoma of the lung (NSCCL): Results of RTOG protocol 81-08. Am. J. Clin. Oncol. 12:280-284; 1991. 14. Cox, J. D.; Azamia, N.; Byhardt, R. W.; Shin, K. H.; Emami, B.; Pajak, T. A randomized phase I/II trial of hyperfractionated radiation therapy with total doses of 60.0 to 79.2 Gy-Possible survival benefit with 2 69.6 Gy in favorable patients with Radiation Therapy Oncology Group Stage III nonsmall cell lung carcinoma. Report of Radiation Therapy Oncology Group 83- 11. J. Clin. Oncol. 8: 15431555; 1990. 15. Cox, J. D.; Azamia, N.; Byhardt, R. W.; Shin, K. H.; Emami, B.; Perez, C. A. N2 (clinical) nonsmall cell carcinoma of the lung: Prospective trials with total doses 60 Gy by the Radiation Therapy Oncology Group. Int. J. Radiat. Oncol. Biol. Phys. 20:7- 12; 1991. 16. De luca, L. M. Retinoids and their receptors in differentiation, embryogenesis, and neoplasia. FASEB J. 5~2924; 1991. 17. Diener-West, M.; Pajak, T. F.; Bauer, M.; Cox, J. D. Randomized dose-searching phase ILti/II trials of fractionation in radiation therapy for cancer. J. Natl. Cancer Inst. 83:1065-1071; 1991. 18. Dillman, R. 0.; Seagren, S. L.; Propert, K. J.; Guerra, J.; Eaton, W. L.; Perry, M. C.; Carey, R. W.; Frei, E. F.; Green, M. R. A randomized trial of induction chemotherapy plus high-dose radiation vs. radiation alone in stage III nonsmall cell lung cancer. N. Engl. J. Med. 323:940-945; 1990. 19. Elson, D.; Brindle, J.; Rice, B. 5 FU and cisplatin plus radiation therapy for limited inoperable nonsmall cell lung cancer (Abst.). Proc. Am. Sot. Clin. Oncol. 6:172; 1987. 20. Emami, B.; Perez, C. A.; Herskovic, A.; Hederman, M. A. Phase I/II study of treatment of locally advanced (T3/T4)

Volume

21.

22. 23. 24. 25.

26.

27. 28.

29.

30.

31. 32.

33. 34.

35.

36.

32, Number

5, 1995

nonoat cell lung cancer with high dose radiotherapy (rapid fractionation): Radiation Therapy Oncology Group study. Int. J. Radiat. Oncol. Biol. Phys. 15:1021-1025; 1988. Farley, P.; Giri, J.; Ronquiollo, A. Treatment of nonsmall cell lung cancer with simultaneous chemotherapy and radiation therapy (Abst.). Proc Am Sot. Clin. Oncol. 3:21 1; 1984. Fowler, J. F. Carcinoma of the lung: Hyperfractionation or resection and chemotherapy? Int. J. Radiat. Oncol. Biol. Phys. 20:169-171; 1991. Fowler, J. F. Fractionation in radiation therapy. Semin. Radiat. Oncol. 2:16, 67; 1992. Fowler, J. F.; Harari, P. M. Hyperfractionation’s promise in cancer treatment. Contemp. Oncol. March: 14-20; 1993. Gordon, D. S.; Hall, L. S.; McDougal, J. S. Studies on the in vivo and in vitro effects of levamisole in murine chemoimmunotherapy model. 2nd Conference on Modulation of Host Resistance in the Prevention and Treatment of Induced Neoplasia. Washington, DC: DHEW Publication; 1974. Graham, M. V.; Pajak, T. E.; Herskovic, A.; Emami, B.; Perez, C. A. Phase I/II study of the treatment of locally advanced (T3/T4) nonoat cell lung cancer with concomitant boost radiotherapy by the Radiation Therapy Oncology Group (RTOG 83-12): Long term results. Int. J. Radiat. Oncol. Biol. Phys. (submitted). Hall, E. J. The Janeway Lecture 1992: Nine decades of radiobiology: Is Radiation Therapy any the better for it? Cancer 71:3753-3766; 1993. Herskovic, A.; Orton, C.; Seyedsadr, M.; Lattin, P.; Kraut, M.; Han, I.; Ahmad, K.; Holt, J. Initial experience with a practical hyperfractionated accelerated radiotherapy regime. Int. J. Radiat. Oncol. Biol Phys. 21:1275-1281; 1991. Langer, C. J.; Curran, W. J.; Keller, S. M.; Catalano. R.; Fowler. W.; Blankstein, K.; Litwin, S.; Bagchi, P.; Nash, S.; Comis, R. Report of a phase II trial of concurrent chemoradiotherapy with radical thoracic irradiation (60 Gy), infusional fluorouracil, bolus cisplatin and etoposide for clinical stage IIIB and bulky IIIA nonsmall cell lung cancer. Int. J. Radiat. Oncol. Biol. Phys. 26:469-478; 1993. Lee, J. S.; Scott, C.; Komaki, R.; Fossella, F.; Dundas, G. S.; McDonald, S.; Palmer, M.; Curran, W. J.; Byhardt, R. W. Concurrent chemoradiation therapy with oral VP-16 and cisplatin for locally advanced inoperable nonsmall cell lung cancer (NSCLC): RTOG protocol 91-06 (Abst. 1222). Proc. Am. Sot. Clin. Oncol. 13:363; 1994. Litchenfeld, J. L.; Wiemik, P. H. Levamisole in a phase I trial. Cancer Treat. Rep. 60:963-964; 1976. Maciejewski, B.; Withers, H. R.; Taylor, J. M. G.; Hliniak, A. Dose fractionation and regeneration in radiotherapy for cancer of the oral cavity and oropharynx: II. Normal tissue response: Acute and late effects. Int. J. Radiat. Oncol. Biol. Phys. l&101-111; 1990. Marks, R.; Witherspoon, B.; Davis, L. Hyperfractionation-Where we stand-A preliminary RTOG report. Proc. Am. Sot. Ther. Radiol. 4:139- 140; 1978. Martini, N.; Kris, M. G.; Gralle, R. J.; Bains, M. S.; McCormack, P. M.; Kaiser, L. R.; Burt, M. E.; Zaman, M. B. The effects of preoperative chemotherapy on the resectability of nonsmall cell lung carcinoma with mediastinal node metastases (N2MO). Ann. Thorac. Surg. 45:370-379; 1988. McDonald, S.; Chang, A. Y.; Rubin, P.; Wallenberg, J.; Kim, I. S.; Sobel, S.; Keng, P.; Muhs, A. Combined betaseron R (Recombinant Human Interferon Beta) and radiation for inoperable nonsmall-cell lung cancer. Int. J. Radiat. Oncol. Biol. Phys. 27:613-619; 1993. Naruke, T.; Goya, T.; Tsuchiya, R.; Suemasu, K. The im-

RTOG

37.

38.

39.

40.

41.

42.

43.

44.

45.

NSCLC

protocols

portance of surgery to nonsmall cell cancer of the lung with mediastinal node metastases. Ann. Thorac. Surg. 46:603610; 1988. Perez, C. A.; Bauer, M.; Edelstein, S.; Gillespie, B. W.; Birch, R. Impact of tumor control on survival in carcinoma of the lung treated with irradiation. Int. J. Radiat. Oncol. Biol Phys. 125399547; 1986. Perez, C. A.; Bauer, M.; Emami, B.; Byhardt, R. W.; Doggett, R. L. S.; Gardner, P.; Zinninger, M. Thoracic irradiation with or without levarnisole (NSC # 177023) in unresectable nonsmall cell carcinoma of the lung: A phase III randomized trial of the RTOG. Int. J. Radiat. Oncol. Biol. Phys. 15:1337-1346; 1988. Perez, C. A.; Pajak, T. F.; Rubin, P.; Simpson, J.; Mohiuddin, M.; Brady, L. W.; Perez-Tamayo, R.; Rotman, M. Long-term observations of the patterns of failure in patients with unresectable nonoat cell carcinoma of the lung treated with definitive radiotherapy. Cancer 59: 1874-1881; 1987. Perez, C. A.; Stanley, K.; Rubin, P.; Kramer, S.; Brady, L. W.; Marks, J. E.; Perez-Tamayo, R.; Brown, G. S.; Concannon, J. P.; Rotman, M. A prospective randomized study of various irradiation dose fractionation schedules in the treatment of inoperative nonsmall cell lung cancer. Cancer 2412744-2753; 1980. Peters, L. J.; Ang, K. K.; Thames, H. D. Altered fractionation schedules. In: Perez, C. A.; Brady, L. W., eds. Principles and practise of radiation oncology. Philadelphia: J. B. Lippincott; 1991:97- 113. Rockwell, S.; Frindel, E.; Valleron, A. J.; Tubiana, M. Cell proliferation in EMT6 tumors treated with single doses of x-rays or hydroxyurea. I. Experimental results. Cell Tissue Kinet. 11:279-289; 1978. Saunders, M. I.; Dische, S. Continuous hyperfractionated, accelerated radiotherapy (CHART) in nonsmall cell carcinoma of the bronchus. Int. J. Radiat. Oncol. Biol. Phys. 19:1211-1215; 1990. Sause, W.; Scott, C. B.; Byhardt, R. W.; Emami, B.; Marcial, V.; Herskovic, A. Recursive partitioning analysis of 1,592 patients on four RTOG studies in nonsmall cell lung cancer. J. Clin. Oncol. 12:336 [1123]; 1993. Sause, W. T.; Scott, C.; Taylor, S.; Byhardt, R. W.; Banker,

0

R. W.

46.

47. 48.

49.

50.

5 1.

52. 53. 54.

BYHARDT

1525

F. L.; Thomson, H. W.; Jones, T. K.; Cooper, J. S.; Lindberg, R.. Phase II trial of combination chemotherapy and irradiation in nonsmall cell lung cancer, Radiation Therapy Oncology Group 88-04. Am. J. Clin. Oncol. 15:163-167; 1992. Sause, W.; Scott, C.; Taylor, S.; Johnson, D.; Livingston, R.; Komaki, R.; Emami, B.; Curran, W. J.; Byhardt, R. W.; Turrisi, A.; Dar, A. R.; Cox, J. D. Radiation Therapy Oncology Group (RTOG) 88-08 Eastern Cooperative Oncology Group (ECOG 4588): Preliminary results of a phase III trial in regionally advanced unresectable nonsmall cell lung cancer J. Natl. Cancer Inst. 87:198-205; 1995. Scott, C. B. Personal communication, 2/8/95. Seydel, H. G.; Diener-West, M.; Urtasun, R.; Podolsky, J.; Cox, J. D.; Zinninger, M.; Francis, M. Hyperfractionation in the radiation therapy of unresectable nonoat cell carcinoma of the lung: Preliminary report of a RTOG pilot study. Int. J. Radiat. Oncol. Biol. Phys. 11:1841-1847; 1985. Simpson, J. R.; Bauer, M.; Wasserman, T. H.; Perez, C. A.; Emarni, B.; Wiegensberg, I.; Zinninger, M.; Durbin, L. M. Large fraction irradiation with or without misonidazole in advanced nonoat cell carcinoma of the lung: A phase III randomized trial of the RTOG. Int. J. Radiat. Oncol. Biol. Phys. 13:861-867; 1987. Simpson, J. R.; Perez, C. A.; Philips, T. L.; Concannon, J. P.; Carella, R. J. Large fraction radiotherapy plus misonidazole for treatment of advanced lung cancer: Report of a phase I/II trial. Int. J. Radiat. Oncol. Biol. Phys. 8:303308; 1982. Thames, H. D.; Withers, H. R.; Peters, L. J.; Fletcher, G. H. Changes in early and late radiation responses with altered dose fractionation: Implications for dose-survival relationship. Int. J. Radiat. Oncol. Biol. Phys. 8:219-226; 1982. Tubiana, M. The kinetics of tumor cell proliferation and radiotherapy. Br. J. Radiol. 44:325-347; 1971. Withers, H. R.; Taylor, J. M. G.; Maciejewski, B. The hazard of accelerated tumor clonogen repopulation during radiotherapy. Acta. Oncol. 27:131; 1988. Withers, H. R.; Thames, H. D. Dose fractionation and volume effects in normal tissues and tumors. Am. J. Clin. Oncol. 11:313-329; 1988.