Once Daily High-dose Radiation (≥60 Gy) Treatment in Limited Stage Small Cell Lung Cancer

Objective: To investigate outcomes and prognostic factors in patients treated with once daily high-dose (≥ 60 Gy) radiation therapy (HDRT) and concurrent platinum-based chemotherapy in limited stage small cell lung cancer (LSSCLC). While we await current phase III trials to determine optimal radiation dose fractionation schemes in LS-SCLC, we report our experience in LS-SCLC with once daily HDRT. We hypothesized that HDRT would achieve similar efficacy and tolerability as twice daily therapy. Methods: We conducted a single institution retrospective review of all patients with LS-SCLC who underwent curative intent treatment from 2005-2013. Patients treated with HDRT (≥ 60 Gy) and concurrent chemotherapy (cisplatin or carboplatin and etoposide) were included in our analysis. Clinicopathologic variables assessed include gender, performance status, time to treatment, response to treatment, toxicity, volumetric tumor response at 3 months and use of prophylactic cranial irradiation (PCI). Results: 42 patients with LS-SCLC who initiated concurrent chemoradiation from 2005 to 2013 were included in the analysis. 38 patients (90%) completed definitive treatment to the lung; 16 (38%) also completed PCI. Median failure free survival (FFS) and overall survival (OS) were 11.9 and 23.1 months, respectively. Two-year and 5 year OS rates were 47% (CI=30-62%) and 21% (CI=7-38%), respectively. On univariate analysis, PCI was associated with improved FFS but this was not significant (p=0.18). Gender was the only co-variate significantly associated with statistical differences in FFS (p=0.03) and OS (p=0.02). Grade 3 and 4 esophagitis were 10.5% and 2.6%, respectively. Pre-HDRT tumor volume and 3-month post-treatment tumor volume were both associated with FFS (p<0.01) but not OS. Conclusion: In this single institution series, daily HDRT demonstrated a 2 year OS of 47% in LS-SCLC. This compares well to the historical survival of daily fractionation (47%) from INT 0096 reported by Turrisi et al. Male gender was predictive of significantly worse FFS and OS. Once daily HDRT has similar OS to twice-daily radiation schemes; however, further studies assessing once daily HDRT for LS-SCLC are warranted.


Introduction
Small cell lung cancer (SCLC) accounts 3% of all new cases of lung cancer in the United States [1,2]. Most patients at diagnosis have extensive stage disease; however, in a third of new cases, disease is confined to a hemithorax, mediastinum or supraclavicular lymph nodes, which is categorized as limited stage SCLC (LS-SCLC). Five year survival rate for patient with LS-SCLC is approximately 25% [3,4].
Multiple thoracic radiation treatment strategies have been investigated in an effort to find the best dose-fractionation scheme. Cancer and Leukemia Group B (CALGB) 8837 determined the maximum tolerated dose (MTD) of 70 Gy/35 fractions administering radiation daily whereas in twice daily treatment MTD was 45 Gy in 30 fractions [11]. Analysis of CALGB trials (CALGB 30206, 30002 and 39808) utilizing 70 Gy in 35 daily-fractions demonstrated 2 year survival rate of 37% [12]. Examining accelerated high-dose radiation therapy (HDRT) via concomitant boost, a phase II study by the Radiation Therapy Oncology Group (RTOG), administered 61.2 Gy in 34 fractions and showed comparable 2 year survival rate of 36.6% with 18% grade ≥ 3 esophagitis [13].
Preclinical and a few pilot clinical studies have demonstrated that the small cell lung cancer dose-response curve lacks a shoulder so a relatively low dose of radiation per fraction can kill malignant cells exponentially, while sparing normal tissues. Based on these findings, Turrisi et al. conducted a phase III intergroup trial (INT 0096), comparing twice daily vs. once daily radiation to 45 Gy concurrently with cisplatin and etoposide. This study resulted in 2 year survival rate of 47% in patients who received twice daily radiation, and 10% improvement in 5 year OS compared to patients who received once daily treatment. Incidence of grade ≥ 3 esophagitis, infectious and pulmonary complications were 32%, 9% and 6% respectively [14]. This trial set the standard for managing LS-SCLC; however, patterns of care studies revealed that only 21% of LS-SCLC patients received twice-daily radiation therapy in 2006-2007 [15,16]. Under-utilization 3-month post treatment CT scans were utilized to determine tumor volume response. Tumor and involved lymph nodes (GTV) were contoured on each individual CT slice to determine total tumor burden volume in cubic centimeters and response was calculated using the equation: Tumor Volume Response = pretreatment tumor volumeposttreatment tumor volume / pretreatment tumor volume * 100%.

End points
Clinical end points, OS and FFS, were determined through assessment of post-treatment imaging and clinical follow-up. All thoracic failures on imaging were assessed with a pre-treatment CT in order to categorize failure as regional in-field, regional in-and-out of field, regional and distant, or distant only. OS was measured from the date of treatment initiation to the date of death from any cause. FFS was measured from date of treatment initiation to date of recurrence or date of death from any cause. Patients alive or, alive without recurrence were censored at date of last follow-up for OS and FFS, respectively.

Statistical Analysis
Of the 42 patients identified, 4 patients did not complete primary treatment and were excluded from the analysis. Survival probabilities for OS and FFS were estimated and plotted using the Kaplan-Meier method. Estimates along with 95% point wise confidence intervals were reported. For comparisons of the survival distributions, the log-rank test was used. Univariable Cox proportional hazards regression models were used to assess the effects of prognostic variables on the outcomes of interest. Prognostic variables included: gender, age at diagnosis, T stage, Eastern Cooperative Oncology Group (ECOG) performance status, receipt of PCI duration of radiation, treatment duration, time between biopsy and treatment completion, and treatment response. All statistical testing was two-sided and assessed for significance at the 5% levels using SAS v9.4 (SAS Institute; Cary, NC).

Patient population
64 patients underwent workup and treatment for LS-SCLC at our institution from 2005 to 2013. Of those, 22 were excluded due to sequential chemoradiation, radiation initiation after 2 cycles of chemotherapy, and planned radiation <60 Gy [3,12]. 42 underwent treatment planning with curative intent consisting of HDRT (≥ 60 Gy) and concurrent chemotherapy. Staging studies included endobronchial ultrasound or mediastinoscopy for lymph node evaluation in all patients and PET imaging was completed in 88% of these patients. The staging CT was evaluable in 40 patients. Thirty-eight of the 42 patients completed concurrent HDRT with curative intent and underwent outcomes analysis. Four patients terminated definitive therapy between 5 and 25 fractions and went to hospice care, including one patient who suffered from Lambert-Eaton Myasthenic Syndrome. Two patients were unable to tolerate platinum based chemotherapy due to impaired renal function, one of whom was unable to complete definitive HDRT.
The median age of patients was 63 with 58% males and ECOG performance status was 2 or less in 97%. The most common radiation regimen prescribed was 61.2 Gy in 34 fractions (82% of patients). The most common chemotherapy regimen prescribed was carboplatin (AUC 5), etoposide (100 mg/m 2 ), being utilized approximately 3:1 compared to cisplatin, etoposide. PCI was administered to 16 patients (38%). Radiation was initiated at a median of 9.5 days after the first chemotherapy cycle. Reasons for not pursing PCI in 22 patients of this effective regimen is mainly due to practical issues of twice daily regimen and its perceived toxicities [17,18].
Since once daily radiation is widely adopted, CALGB 30610/RTOG 0538 and Concurrent ONce-daily VErsus twice-daily RadioTherapy (CONVERT) studies are looking at direct comparison of various once daily radiation schedules. However, final results of these trials are not anticipated for some time. In an attempt to guide patient's decision making, we performed a retrospective analysis of patients who underwent daily HDRT concurrently with standard chemotherapy at our institution for LS-SCLC and report their outcomes in this manuscript. We hypothesized that daily single-fraction HDRT with concurrent chemotherapy would achieve similar efficacy and tolerability as twice daily radiation therapy.

Patient selection
After institutional review board approval, a retrospective chart review of all patients treated at our institution for LS-SCLC was carried out. Patients were selected if they initiated curative intent therapy with concurrent chemotherapy and HDRT ≥ 60 Gy from 2005 to 2013.

Treatment and follow-up
Treatment planning for all patients was carried out using the Pinnacle treatment planning system (Phillips Medical Systems, Fitchburg, WI) and beginning in 2006, 4D-CT-planning software (Siemens Medical Solutions, Concord, CA). Target volume delineation consisted of gross tumor volume (GTV) including clinically or pathologically involved lymph nodes identified by Positron Emission Tomography/Computed Tomography (PET/CT) scans without elective nodal targeting. CTV and ITV were created to encompass the tumor motion at the discretion of the treating radiation oncologist. Planning target volume (PTV) expansion of 0.5 cm was added to account for daily set up errors. Respiratory gating after 4D-CT was utilized when tumor motion was greater than 1 cm. Treatment was delivered using 6 or 10 MV photons using either 3D conformal or intensity modulated radiation therapy depending on optimal dose distribution as determined by the attending physician. Chemotherapy regimen including dose and timing were confirmed in treatment records. Patients were seen at least weekly during radiotherapy to determine tolerance to treatment. PCI was administered after appropriate response to definitive therapy and post treatment brain MRI confirmed absence of intracranial disease. On and post-treatment records were reviewed for toxicity, hospitalizations, and disease recurrence. Toxicity was determined retrospectively by reviewing on-treatment assessments, hospital records, laboratory values and the administration of narcotic medications according to the Common Toxicity Criteria for Adverse Events (CTCAE v4).

Tumor response assessment
The pre-treatment staging CT and the 3 month post-treatment CT were imported into Velocity treatment planning software (Varian; Palo Alto, CA). Pre-treatment tumor and involved lymphadenopathy (GTV) was measured in its largest dimension and the sum of these values recorded. Three-month post-treatment GTV measurements were repeated in the same manner and RECIST (response evaluation criteria in solid tumors) was recorded using previously established methods [19].
The pre-treatment staging CT with PET (88%) followed by the were patient refusal, poor performance status, vascular disease and recurrence with intracranial disease on MRI, and other unknown reasons [2,3,5,9].

Recist response
Post treatment imaging assessment was completed in 34 patients for whom post-treatment imaging was available. Response assessment using RECIST at three months post treatment classified response as complete response in 10 patients (26.3%), partial response in 20 patients (52.6%), and stable disease (SD) in 4 patients (10.5%). In patients who completed concurrent HDRT, there were no instances of progressive disease on imaging. Treatment response was not associated with FFS or OS.

Volumetric response
Imaging was available for three-month post-treatment tumor burden quantification in 34 patients. Tumor response to definitive therapy was measured in cubic centimeters with an average reduction in volume of 91% (SD=10.6). Pre-HDRT tumor volume and residual tumor volume at 3 months were both associated with FFS (p<0.01) but not OS.

Toxicity
Treatment-related toxicities are summarized in Tables 1 and 2 and Figure 1. Grade ≥ 3 esophagitis was seen in 13.1% of patients. No grade ≥ 3 pneumonitis was observed but 8% of patients did develop grade 2 pneumonitis. Twelve patients (31.6%) required hospitalization for any cause during HDRT. Four of those were admitted for grade 3 febrile neutropenia. Other causes for hospitalization included odynophagia, pneumonia, dehydration, bowel obstruction, and severe weakness. There were no treatment-related deaths associated with concurrent chemoradiation.

Patterns of treatment failure
Thirty-eight patients completed definitive treatment and could be evaluated for disease recurrence. Four had post-treatment imaging completed at an outside institution and were not available for analysis. Twenty patients (52.6%) were free of disease at last known follow-up. Most recurrences were 'distant only' (9 patients, 23.7%) and this was most commonly seen in brain (6 patients). The remaining failures were 'regional in-field' (3 patients, 7.9%), 'regional in-and-out of field' (1 patient, 2.6%), as well as 'regional and distant' recurrence (5 patients, 13%). Hence 36.7% of patients had some component of distant failure.
Although many studies have published outcomes in LS-SCLC, few have investigated HDRT with concurrent platinum based chemotherapy in LS-SCLC. Rutter et al. analyzed outcomes of 1228 LS-SCLC patients from the National Cancer Database and did not find any statistically significant differences in the median survival of patients according to the radiation dose-regimen received (70 Gy daily, 61.2 Gy daily, 45 Gy twice daily) [29]. They do not describe toxicity data. It is possible that these regimens may not differ from each other in terms of efficacy; however, difference in toxicities and ease of treatment delivery can make one regimen superior to the other.
The CALGB 30610-RTOG 0538 study was designed to evaluate three different radiation regimens in a randomized fashion: 70 Gy in 35 single daily fractions, 61.2 Gy in 34 fractions (16 once daily fractions followed by 18 twice daily fractions as concomitant boost) and 45 Gy in 30 twice-daily fractions. The hybrid single/twice daily treatment arm consisting of 61.2 Gy/34 fractions was dropped at interim analysis due to toxicity. A similar effort is ongoing in Europe where the CONVERT study is randomizing good performance status patients to once daily HDRT (66 Gy/33 fractions) versus twice-daily treatment (45 Gy/30 fractions) [36].

Characteristics of limited stage small cell lung cancer patients (N=38)
Age-yr     Hospital admission for any cause for those completing treatment was recorded as 12 (31.6%).
Hematologic lab values were unavailable for 5 subjects. No grade 5 toxicities occurred. A platinum-based chemotherapy combined with etoposide and radiation is the standard treatment for LS-SCLC; however, the benefit of using cisplatin versus carboplatin is unresolved. A meta-analysis showed similar survival outcomes for cisplatin versus carboplatinbased regimens in small cell lung cancer [37]. A phase III trial compared cisplatin to carboplatin in SCLC and showed no significant difference in response rate or survival between the two chemotherapies, and the carboplatin regimen was associated with a better toxicity profile [38]. In regards to esophagitis, studies in NSCLC have shown that cisplatin is more likely to be associated with esophagitis than carboplatin [39]. The CALGB 30610/RTOG 0538 study is allowing either cisplatin or carboplatin based chemotherapy with radiation whereas the CONVERT study is restricted to cisplatin. These studies will likely shed more light on the optimal chemotherapy regimen to be used and the incidence of radiation esophagitis in the cases of cisplatin and carboplatin.
Results of ongoing studies will not be available for some time; however, knowledge of patient outcomes who underwent HDRT remains important in guiding treatment selection for LS-SCLC patients. Our single institution retrospective review of HDRT (≥ 60 Gy) with concurrent platinum based chemotherapy demonstrated comparable efficacy (2year OS 47%, median overall survival of 23.1 months) with the best-reported twice daily regimen [14]. Toxicity profile revealed grade ≥ 3 esophagitis rate was 13.1%, grade 2 pneumonitis rate was 8% and no grade ≥ 3 pneumonitis was seen. In addition, our volumetric tumor assessment showed that pre-HDRT tumor volume was associated with worse FFS (p<0.01) as has been reported by Reymen et al. This may be a useful metric to be utilized in future studies of LS-SCLC for patient's stratification [40,41].
The limitations of our study include those inherent to the retrospective analysis which may include incomplete patient information or imaging as described. In addition, limitations in toxicity assessment from the available treatment information and follow up records is noted.

Conclusion
Daily HDRT treatment (≥ 60 Gy) with concurrent platinum based chemotherapy in LS-SCLC compares well in efficacy to twicedaily regimen with a better toxicity profile. Prospective studies assessing once-daily HDRT in LS-SCLC and the optimal concomitant chemotherapy regimen are yet to be reported.  Selected studies investigating dose fractionation in LS-SCLC treated with concurrent chemotherapy and radiation. Most common concurrent chemotherapy regimen specified (*) cisplatin/etoposide, (**) carboplatin/etoposide, (***) platinum based, (NR) not reported.