Interruptions Of Once-Daily Thoracic Radiotherapy Do Not Correlate With Outcomes In Limited Stage Small Cell Lung Cancer: Analysis of CALGB Phase III Trial 9235

Summary

Purpose

Retrospective data suggests prolonging the time to complete thoracic radiotherapy (TRT) may negatively impact tumor control and survival in limited stage small cell lung cancer (LSCLC). We examined the association between TRT duration and outcomes on a prospective phase III study.

Material and Methods

This review included 267 patients who received protocol TRT on a phase III CALGB LSCLC study assessing the addition of tamoxifen to standard chemo-radiotherapy. TRT, to a planned dose of 50 Gy in 2 Gy daily fractions, was initiated with the 4^th chemotherapy cycle. TRT interruptions were mandated for hematologic toxicity (granulocytes < 1000/mm³ or platelets < 75,000/mm³) and esophageal toxicity (dysphagia necessitating intravenous hydration).

Results

TRT interruptions ≥ 3 days occurred in 115 patients (43%), most frequently during the 4^th week of TRT, and did not differ between treatment arms. Hematologic toxicity and esophageal toxicity were the most frequent indications for interrupting TRT. Variables including advanced age (>70 years), gender, race, or radiotherapy treatment volume did not predict for TRT interruptions. Overall survival (OS) and local tumor control did not correlate with the administration of TRT interruptions or with TRT duration.

Conclusion

Toxicity mandated interruptions of conventional dose, once-daily, TRT may not adversely affect outcomes for patients receiving TRT concurrent with chemotherapy (cycle 4) for LSCLC. The implications for accelerated or high dose TRT regimens are not clear.

Introduction

Standard treatment for LSCLC has evolved to include two-agent systemic chemotherapy in combination with thoracic radiotherapy (TRT). The utility of TRT was widely studied in the 1970’s and 1980’s with contrasting results, but two meta-analyses published in the early 1990’s demonstrated a significant survival benefit with the addition of TRT to systemic chemotherapy (1, 2). Patterns of Care data suggests that the majority of practitioners in North America have adopted a concurrent chemo-radiotherapy approach for fit patients with LSCLC (3–5). Although few modern trials have compared sequential and concurrent chemoradiotherapy, superior results have generally been reported from trials administering thoracic radiotherapy concurrent with cisplatin based chemotherapy. Morevoer, recent prospective evidence from Japanese investigators suggests a potential survival benefit for concurrent therapy compared with a sequential approach (6).

While several issues regarding the integration of TRT and chemotherapy for LSCLC remain unresolved, a landmark United States Intergroup phase III trial showed that altering the manner in which radiotherapy is administered can ultimately impact survival in LSCLC. In that study, accelerated hyperfractionated TRT (AH-TRT) proved superior to conventionally fractionated TRT (CF-TRT) when initiated with the first cycle of cisplatin and etoposide chemotherapy, demonstrating that dose intensity is of critical importance in LSCLC (7). Despite this result, CF-TRT remains, by far, the most frequently utilized regimen in the United States. In fact, more than 80% of patients received CF-TRT (median dose 50.4 Gy) in the Patterns of Care survey of patients treated between 1998 and 1999, while only 6% of patients received AH-TRT (3).

Given the impact of survival with accelerated twice daily radiotherapy in LSCLC, practitioners are often strongly discouraged from palliating the toxic effects of therapy by interrupting therapy and protracting the radiotherapy course. Whether toxicity mandated radiotherapy interruptions are detrimental in LSCLC, for patients receiving CF-TRT, is not clear as there is little published data. A recent retrospective review from Canada indicated that overall treatment time may also be critical for patients receiving once-daily CF-TRT. In that report, radiotherapy treatment interruptions were associated with poorer long-term survival and local tumor control in patients receiving combined therapy for LSCLC (6). We report the details of TRT delivery from a prospective phase III CALGB trial that employed conventionally fractionated TRT.

Materials and Methods

Patient Selection

The primary objective of CALGB 9235 was to determine whether the addition of tamoxifen to standard chemo-radiotherapy improves OS for patients with LSCLC. Eligible patients included those individuals (1) with histologically or cytologically confirmed LSCLC, (2) who were 18 years of age and older, (3) with a performance status of 0–2, and (4) who had received no prior chemotherapy, radiation therapy, or immunotherapy. LSCLC was defined as disease confined to one hemithorax including hilar, mediastinal, and supraclavicular nodes whether ipsilateral or contralateral. Specific limits for acceptable levels of hematologic, hepatic, and renal function were stipulated. Patients with ipsilateral pleural effusion, whether or not they were cytologically positive, were excluded. All patients were required to sign an informed consent document approved by local institutional review boards.

Treatment

Patients were randomized at the CALGB Statistical Center to Arm 1 (cisplatin and VP-16) or Arm 2 (TAM plus cisplatin and VP-16). Treatment on Arm 1 was cisplatin 80 mg/m² intravenously on day 1 and VP-16 80 mg/m² intravenously on days 1–3; cycles were repeated every 3 weeks for 5 planned cycles. In Arm 2, TAM was administered at a dose of 80 mg orally twice daily on days 1–5 along with the same dose of cisplatin and VP-16 as on Arm 1. However, on Arm 2 cisplatin was given on day 2 and VP-16 was administered on days 2–4 of each cycle.

After completion of 3 cycles of treatment all patients were evaluated to determine the response to the induction chemotherapy regimen. During chemotherapy cycles 4 and 5, patients received concurrent TRT. Following the completion of the concurrent chemoradiotherapy, patients were re-evaluated to determine the best response to the treatment. Those patients who achieved a complete remission (CR) or “near” CR were offered the option of prophylactic cranial irradiation (PCI). Determination of near “CR” was left to the treating oncologist for patients that had a very good “PR” with residual indeterminate abnormality on chest imaging. Patients who agreed to PCI received a total dose of 36 Gy in 18 fractions given on 18 treatment days over 3.5 weeks beginning approximately 4 weeks after completion of all chemotherapy.

Patients in complete response (CR) were followed every 2 months for 2 years with laboratory studies and CT scans of the chest. Thereafter, patients were followed every 3 months for an additional 3 years and then on an every 6 month basis. Patients who demonstrated a partial response (PR) were eligible for observation or additional treatment at the discretion of the treating physician.

Thoracic Radiotherapy (TRT)

TRT began on the first day of the fourth cycle of chemotherapy (treatment day 64) and continued through cycle 5. The treatment volume included the primary tumor, ipsilateral hilar, and bilateral mediastinal lymph nodes including those at the thoracic inlet and the lower mediastinum. A shrinking field was used to treat two consecutive ports of treatment volume: the original volume defined by the pre-treatment studies before chemotherapy and the boost volume outlined by CT scans taken immediately before starting the fourth cycle of chemotherapy. The total dose to the original volume was 4000 cGy in 20 fractions over 4 weeks while the boost volume was 1000 cGy in 5 fractions over 1 week. Radiotherapy data, including diagnostic imaging, simulation films, portal films, and dosimetric calculations, was reviewed by the Quality Assurance Review Center (QARC; Providence, RI).

Statistical Considerations

CALGB 9235 was reviewed semi-annually by the CALGB Data and Safety Monitoring Board. Patient data collection was managed by the CALGB Statistical Center. Data quality was ensured by careful review of data by CALGB Statistical Center staff and by the study chairperson. Statistical analyses were performed by CALGB statisticians.

There were 2 distinct survival end points of interest. If the effect variable of interest was a variable associated with radiation, then survival was calculated as the time between the last dose of chest radiation and death. If the effect variable of interest was not a variable associated with radiation, then survival was calculated as the time between patient going off treatment and death. The Kaplan-Meier product limit estimator was used to describe the distribution of survival time. The Cox proportional hazards model was used to assess the effect of interruptions, total days of radiation to the chest, grade 3+ hematologic toxicity, grade 3+ hematologic toxicity during radiation, and esophageal toxicity after adjustment for other potentially confounding factors such as weight loss, symptom duration, performance status, and study arm. Logistic regression was used to assess the effect of length and width of the radiotherapy field on esophageal toxicity after an adjustment for potentially confounding factors such as weight loss, study arm, and symptom duration. The median potential follow-up for all patients at the time of analysis is 54 months.

Demographics

Three hundred and nineteen (319) patients accrued between August 6, 1993 and January 15, 1999. Twelve patients either never started treatment or were found to be ineligible. Although 307 initiated protocol chemotherapy, 40 patients (13%) did not proceed to protocol TRT (due primarily to disease progression) leaving 267 patients who initiated protocol TRT.

Table 1 summarizes patient characteristics. 139 patients were treated with chemotherapy and radiation and 128 patients with chemotherapy and radiation plus tamoxifen. The overall median age was 62, ranging from 34 to 81. The patients were generally male (57%), white (90%), with PS=1 (52%), weight loss less than 5% (77%), and duration of symptoms less than 3 months (60%).

Table 1.

Patient characteristics

	N (267)	Percentage (%)
Gender
Male	152	57
Female	115	43
Age
<49	30	11
50–59	71	27
60–69	110	41
70–79	53	20
80+	3	1
Median (range)		62 (34,81)
Race
White	239	90
Hispanic	1	0
Black	23	9
Oriental	2	1
Performance Status
0	129	48
1	138	52
Weight loss?
Missing	5	2
<5%	206	77
5%–10%	31	12
>10%	25	9
Duration of symptom(s)
Missing	25	9
0	1	0
<3 months	161	60
3–6 months	54	20
>6 months	26	10

As part of the quality assurance program of the CALGB, members of the Data Audit Committee visit all participating institutions at least once every three years to review source documents. The auditors verify compliance with federal regulations and protocol requirements, including those pertaining to eligibility, treatment, adverse events, tumor response, and outcome in a sample of protocols at each institution. Such on-site review of medical records was performed for a subgroup of 127 patients (39.8%) of the 319 patients under this study.

Results

The addition of tamoxifen to standard chemoradiotherapy did not result in improved patient outcomes and a detailed analysis of the primary study endpoint has been published (9).

One hundred and fifteen patients experienced TRT interruptions ≥3 days (52 on arm 1 and 63 on arm 2). TRT was most frequently interrupted during the 4^th week of therapy. The reason stated for interrupting therapy by the individual investigator included hematologic toxicity, esophageal toxicity, both hematologic toxicity and esophageal toxicity, or other in 57%, 4%, 5% and 34% of patients respectively. During concurrent TRT and chemotherapy, grade 3–4 granulocytopenia, thrombocytopenia, and esophageal toxicity occurred in 73%, 35%, and 24% of patients respectively.

Factors Predicting for TRT interruptions

Factors analyzed with respect to TRT interruptions included treatment arm, weight loss, performance status, gender, age, race, radiotherapy field size, and radiotherapy treatment area. Elderly patients (age ≥ 70 years) did not experience increased TRT interruptions compared with younger patients when taken as a whole. However, patients ≤50 years of age were less likely to have TRT interruptions than patients > 50 years of age (p=0.012). The remaining factors did not predict for TRT interruptions. The same variables were also assessed in relationship to hematologic toxicity and esophageal toxicity. Patients ≥ 70 years were more likely to experience hematologic toxicity and there was a trend towards increased grade 4 hematologic toxicity (p= .10) with increasing radiation treatment volume. Hematologic toxicity did not correlate with the length of the radiotherapy field, however. None of the variables examined predicted for esophageal toxicity.

Correlation between TRT interruptions and outcome

The Cox proportional hazards model was used to assess the effect of interruptions and the duration of TRT on survival after the termination of RT. Other variables included in this analysis include weight loss prior to diagnosis, performance status, symptom duration, and treatment arm (table 2). Weight loss prior to diagnosis and pre-treatment performance status correlated with overall survival. There was no association between the presence of TRT interruptions and overall survival, and the duration of TRT interruption (e.g. 0–3 days vs 4–10 days vs > 10 days) also did not predict for survival or local relapse free survival. (figure 1, table 3). A separate analysis limited to patients with excellent performance status (e.g. PS = 0) did not reveal a detrimental effect of TRT interruptions (figure 2).

Table 2.

Overall Survival According to Select Patient and Treatment Variables

Variable	Coefficient	se(coefficient)	p-value
Weight loss: 5%–10% versus <5%	−0.0309	0.244	0.9
Weight loss: >10% versus <5%	0.4744	0.281	0.091
Performance Status 1 versus 0	0.4124	0.163	0.011
Performance Status 2 versus 0	0.5927	0.314	0.059
Symptom duration: 3–6 months versus <3 months	−0.2329	0.202	0.25
Symptom duration: >6 months versus <3 months	−0.1506	0.273	0.58
With tamoxifen	0.0318	0.149	0.83
Interruptions for 4–10 days versus 0–3 days.	0.1891	0.194	0.33*
Interruptions for > 10 days versus 0–3 days.	−0.1427	0.224	0.52*

Fig 1.

Overall survival stratified by days of radiotherapy interruption for all patients.

Table 3.

Overall Survival and Local Relapse Free Survival vs TRT Interruptions

Overall Survival
TRT Interruptions	N	Median Survival time in Months (95% CI)	Two Years estimate (95% CI)	Three Years estimate (95% CI)
0–3 Days	159	21.7 (20.1,26.2)	43% (36%, 52%)	31% (25%, 39%)
4–10 Days	49	18.1(14.5,26.5)	41% (29%, 57%)	29% (18%, 45%)
>10 Days	37	23.8(16.1,39.3)	49% (35%, 68%)	35% (23%, 54%)
Local Relapse Free Survival
0–3 Days	159	46.2 (21.4, NA)	55% (47%, 65%)	52% (43%, 62%)
4–10 Days	49	38.1 (14.8, NA)	51% (37%, 70%)	47% (33%, 67%)
>10 Days	37	98.2 (20.6, NA)	59% (44%, 80%)	NA

Fig 2.

Overall survival stratified by days of radiotherapy interruption for patients with ECOG performance status = 0.

An exploratory analysis was performed to assess whether the impact of TRT interruptions might vary according to response to induction chemotherapy (e.g. CR vs less than CR). A significant impact of TRT interruptions was not observed for either local relapse free survival (table 4) or overall survival when stratified by response to induction chemotherapy.

Table 4.

Local Relapse Free Survival by Response to Induction Chemotherapy

Response < CR
TRT interrupted?	n	Median Survival time in Month (95% CI)	Two Years estimate (95% CI)	Three Years estimate (95% CI)
No	61	15.6 (11.5 NA)	37% (25%, 58%)	28% (14%, 56%)
Yes	56	14.8 (11.5, NA)	30% (15%, 59%)	30% (15%, 59%)
Response = CR
No	90	81.0 (47.7, NA)	65% (55%, 77%)	61% (51%, 74%)
Yes	60	98.2 (28.8, NA)	62% (50%, 78%)	57% (44%, 73%)

CR, complete response; TRT, thoracic radiotherapy

Relapse within the radiotherapy volume was observed in 37 patients without TRT interruptions and 27 patients with TRT interruptions (p=0.71). Relapse rates outside the irradiated volume but within the chest also did not differ.

Discussion

Reducing radiotherapy dose intensity by prolonging the time to complete therapy has been hypothesized to result in diminished outcomes after treatment for several epithelial tumors. Retrospective reviews of prospective trials conducted in both the United States and Europe, for example, have strongly suggested an inverse relationship between treatment duration and local tumor control for both non-small lung cancer and squamous cell carcinoma of the head and neck (10–13). One generally postulated explanation is that the beneficial effects of therapy may be offset by the accelerated repopulation of tumor clonogens, which may become clinically relevant as early as 3 weeks following the initiation of radiotherapy (14). Bese et. al. recently reviewed the literature regarding the effect of unplanned radiotherapy interruptions and suggested methods to compensate for treatment prolongation (15).

While less information is available regarding the duration of radiotherapy and outcome for patients with small cell lung cancer, the rapid doubling time of this tumor suggests delaying the time to complete a course of radiotherapy may be detrimental. The discordant results of two randomized phase III trials comparing radiotherapy schedules with CFTRT indirectly suggest that a planned treatment break and the resultant reduction in dose intensity, may reduce the efficacy of twice-daily TRT when administered with chemotherapy. Intergroup Trial 0096 demonstrated a significant survival benefit for patients receiving 45 Gy in 1.5 Gy twice-daily fractions over 3 weeks compared with patients receiving the same nominal dose in 1.8 Gy daily fractions over 5 weeks (5). Conversely, a trial conducted by the North Central Cancer Treatment Group found no difference in outcomes between patients receiving 1.5 Gy twice-daily TRT and patients treated with conventionally fractionated TRT (13). However, the NCCTG trial had several important differences, including a planned TRT break of 2.5 weeks such that the treatment intensity in the BID arm was diminished. In addition, TRT was given with the 3^rd cycle of chemotherapy on the NCCTG trial and patients with progressive during induction chemotherapy were not eligible to continue on study.

The impact of interrupting once-daily fractionated TRT in the context of combined modality therapy for LSCLC was recently reported by investigators from the London Regional Cancer Center (Ontario, Canada). In that retrospective analysis, daily TRT, 40 Gy in 15 fractions over 3 weeks or 50 Gy in 25 fractions over 5 weeks, was initiated with etoposide and cisplatin chemotherapy at either the second or third cycle in 215 patients with LSCLC (6). Fifty-six patients (26%) experienced TRT interruptions and the vast majority were secondary to hematologic toxicity. Median and 5 year overall survivals were 13.8 months and 4.2% versus 15.6 months and 8.3%, for interrupted and uninterrupted radiotherapy courses, respectively. The presence of treatment breaks was the most significant factor influencing survival in multivariable analysis (P=0.006).

While our findings are in contrast to the Canadian report, the CALGB database may allow for a more valid assessment of the impact of TRT interruptions. A larger and more homogeneously treated patient population was included in the current study, radiotherapy data was collected prospectively, and both median and long-term survival figures are superior to the Canadian results. Although the results of this analysis may appear counterintuitive, there are several possible explanations for the lack of correlation between dose intensity and outcome in the CALGB trial. For example, the majority of treatment interruptions occurred during the fourth week of TRT and there may not be substantial differential effect of 30 – 40 Gy compared with 50 Gy when delivered concurrent with chemotherapy. Perhaps if the total given dose was greater, as is typically the case in non-small cell lung cancer, then a detrimental effect of interrupting treatment may have been observed. Phase II trials have shown QD TRT doses as high as 70 Gy can be given safely in limited small cell lung cancer(17), and phase III trials comparing high dose QD TRT and 45 Gy BID TRT were recently initiated in Europe and the United States. It is also possible that underlying biologic factors may have played a role and the efficacy of therapy may have been enhanced for patients experiencing toxic effects of treatment.

The optimal timing of TRT in LSCLC is not completely clear, despite publication of recent meta-analyses, although most investigators now agree that TRT should start with the first or second cycle of chemotherapy (18,19). Meta-analyses indicate that early TRT may be more important when accelerated TRT is administered with less of an impact when conventionally fractionated TRT is used to modest doses of 45–50 Gy (18,19). For example, the median overall survival on the reported CALGB study (cycle 4 chemotherapy) is similar to the once daily treatment arm form Intergroup Study 0096, which started TRT with the first cycle of chemotherapy (7,9).

As the administration of radiotherapy interruptions were not randomly assigned, it is not possible to assess whether minimizing treatment breaks, via supportive measures, might result in improved outcomes. Hematologic toxic effects are the most frequent cause of treatment interruptions during combined thoracic radiotherapy and chemotherapy for LSCLC. Hematologic growth factors have been of proven value in lessening the hematologic toxic effects of systemic chemotherapy, but their use in conjunction with radiotherapy is not currently accepted practice.

We were not able to identify significant patient or treatment related factors that predict for treatment interruptions. Our analysis did not demonstrate a relationship between radiation field size and TRT interruptions or esophageal toxicity, although increasing hematologic toxicity was noted with increasing radiotherapy field size. This evaluation is limited as patients in the current trial were treated prior to the routine utilization of conformal radiotherapy in cooperative group trials. Therefore, specific dose-volume relationships cannot be assessed. Current trials collecting such data will be important in determining precise relationships, which may help guide the treatment of patients in the future.

In summary, CALGB data does not support a relationship time to complete TRT and outcome when once daily TRT is employed as part of combined modality therapy for LSCLC. The examination of TRT interruptions during other prospective trials is warranted to further explore the potential impact of dose-intensity in LSCLC. It is emphasized that this analysis applies only to the administration of modest dose conventionally fractionated TRT, administered with the 4^th cycle of chemotherapy, and the results should not be extrapolated to instances where high dose TRT or accelerated TRT is administered. Moreover, the CALGB experience should not be taken as an invitation to indiscriminately interrupt therapy, but may support the judicious administration of treatment breaks when necessitated by the toxic effects of therapy.