Concurrent chemoradiotherapy for large-volume locally-advanced non-small cell lung cancer

Concurrent chemoradiotherapy for large-volume locally-advanced non-small cell lung cancer

Lung Cancer 80 (2013) 62–67 Contents lists available at SciVerse ScienceDirect Lung Cancer journal homepage: www.elsevier.com/locate/lungcan Concur...

265KB Sizes 0 Downloads 111 Views

Lung Cancer 80 (2013) 62–67

Contents lists available at SciVerse ScienceDirect

Lung Cancer journal homepage: www.elsevier.com/locate/lungcan

Concurrent chemoradiotherapy for large-volume locally-advanced non-small cell lung cancer Terry G. Wiersma a , Max Dahele a,∗ , Wilko F.A.R. Verbakel a , Peter M. van de Ven b , Patricia F. de Haan a , Egbert F. Smit c , Ellen J.F. van Reij a , Ben J. Slotman a , Suresh Senan a a

Department of Radiation Oncology, VU University Medical Center, De Boelelaan 1117, PO Box 7057, 1007 MB, Amsterdam, The Netherlands Department of Epidemiology and Biostatistics, VU University Medical Center, De Boelelaan 1117, PO Box 7057, 1007 MB, Amsterdam, The Netherlands c Department of Pulmonary Diseases, VU University Medical Center, De Boelelaan 1117, PO Box 7057, 1007 MB, Amsterdam, The Netherlands b

a r t i c l e

i n f o

Article history: Received 20 October 2012 Received in revised form 19 December 2012 Accepted 4 January 2013 Keywords: Non-small cell lung cancer Large volume Planning target volume Chemoradiotherapy Toxicity

a b s t r a c t Purpose: Patients with large volume stage III non-small cell lung cancer (NSCLC) are often excluded from concurrent chemoradiotherapy (CRT) protocols due to fears about excessive toxicity and poor survival. Patients with N3 nodal disease may be excluded for the same reason. We have routinely accepted fit patients in both the above groups for CRT if they met our planning parameters. We analyzed toxicity and survival outcomes for patients undergoing CRT with a planning target volume (PTV) exceeding 700 cc, either with or without N3 nodal disease, or a PTV less then 700 cc with N3 disease. Materials and methods: Single center, retrospective study of patients with stage III NSCLC treated with CRT between 2004 and 2011. Results: 121 patients were eligible, with 81% (98/121) having a PTV > 700 cc (of whom 33% (32/98) had N3 nodal disease) and 19% (23/121) having N3 disease and a PTV ≤ 700 cc. Grade ≥3 esophagitis and pneumonitis were recorded in respectively 34% and 4% of all patients. Median follow-up for all patients was 37.6 months (mo). Median overall (OS) and progression-free (PFS) survivals were 15.7 mo and 11.6 mo, respectively, OS for all patients with PTV > 700 cc was 14.5 mo (19.5 mo with N3 and 13.2 mo without N3), compared to 26.5 mo for PTV ≤ 700 cc with N3 (p = 0.009). About 1 in 4 patients with PTV > 700 cc died within 6 mo of starting radiotherapy (this was associated with Charlson comorbidity index [CCI] ≥ 1), while about 18% were alive at 3 years. Conclusion: Patients undergoing CRT for stage III NSCLC with a PTV > 700 cc, with or without N3 nodal disease, had a significantly shorter OS than patients with PTV ≤ 700 cc with N3. Patients with PTV > 700 cc and with CCI ≥ 1, had a significantly higher risk of early death but longer-term survivors with PTV > 700 cc are observed. The PTV and CCI should be considered in clinical decision making and used as stratification factors in future trials. © 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Concurrent chemoradiotherapy (CRT) is considered the standard of care for selected, fit, patients with locally-advanced non-small cell lung cancer (NSCLC) [1]. A recent meta-analysis reported a significant benefit of CRT over sequential treatment, with a longer progression free survival (PFS) and an absolute benefit in overall survival (OS) of 5.7% at 3 years, and 4.5% at 5 years [1]. However, an increase in grade 3 or 4 acute esophageal toxicity, from 4% in the sequential to 18% in the CRT group, was

∗ Corresponding author at: Department of Radiation Oncology, VU University Medical Center, De Boelelaan 1117, PO Box 7057, 1007 MB, Amsterdam, The Netherlands. Tel.: +31 20 44 40414; fax: +31 20 44 40410. E-mail address: [email protected] (M. Dahele). 0169-5002/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.lungcan.2013.01.006

also seen. Some authors have cautioned against the routine use of concurrent CRT for larger tumors [2,3] which have been associated with a worse prognosis [4]. In addition, for various reasons a number of study protocols have also excluded patients with large or extensive tumors (e.g. those with N3 involvement) from radical combined modality treatment [5–8]. At our institution, CRT has been routinely used in fit patients with large primary tumor volume and extensive lymph node metastases since 2004, as long as acceptable normal tissue doses can be achieved during radiotherapy treatment planning. An earlier study from our group reported that the sub-group of patients with large volume tumors, defined as those with a radiotherapy planning target volume (PTV) above the mean of 748 cc, had a significantly worse median survival than those with a smaller PTV, 13.3 and 27 months (mo) respectively [9]. However, only 33 patients had a large-volume PTV. The present, larger study focuses exclusively

T.G. Wiersma et al. / Lung Cancer 80 (2013) 62–67

on outcomes for patients with large PTV volumes and/or extensive lymph node metastases. 2. Materials and methods This retrospective study was performed with institutional approval. All patients treated at our department are registered in an oncology information management system (ARIA, Varian Medical Systems), which includes the type of tumor, disease stage and type of treatment. Using this information, an initial chart review was performed of all patients with stage IIIA/B NSCLC treated between 2004 and 2011. Based on the mean and median PTV size of the previous study [9], we defined ‘large volume’ as a PTV exceeding 700 cc. All eligible patients had to have undergone CRT alone and none underwent surgery. Clinical practice evolved during the study period. Whole body positron emission tomography was introduced routinely in 2003, and magnetic resonance imaging (MRI) of the brain in 2007. During the period 2003–10, approximately 1250 patients with NSCLC were treated in our department. Of these, about one-third had stage III disease almost three-quarters of whom received radical intent (non-surgical) treatment, and of the latter, about 60% underwent CRT, with the remainder undergoing sequential CRT and radiotherapy alone. Median overall survival (OS) for the group managed with radical non-surgical treatment was approximately 17.5 mo and about 18.5 mo in those receiving CRT. Typical practice in our institution has been to give one cycle of chemotherapy using a cisplatin-containing doublet, followed by radiotherapy delivered concurrently with cycles 2 and 3 (usually cisplatin and etoposide) of full-dose systemic therapy. Radiotherapy was planned using a 4-dimensional CT scan. During the study period, a number of treatment and planning techniques were used. The planning target volume (PTV) has typically been created from the internal target volume (ITV) plus a 1 cm isotropic margin, and treatment delivered with daily image-guidance. Since 2009, a hybrid-inverse planned intensity modulated radiotherapy (IMRT) technique has been in routine use while prior to this forward planned, field-in-field IMRT was common [10]. The prescribed radiotherapy dose varied, in part because the study covered time periods when different CRT schedules were in use (including different total radiotherapy doses and fraction sizes). Currently, the practice is to try and deliver 66 Gy in 33 fractions and to limit wherever possible the volume of total lung minus PTV receiving at least 20 Gy (V20) to 40%, to try to keep the V5 as low as possible and to keep the maximum spinal canal dose to 50 Gy [10]. At the present time radiation doses to the heart or esophagus are not used as hard constraints for plan acceptance. To assemble the data, where necessary, charts from other departments were reviewed, other hospitals and the patient’s general practitioner were contacted and radiology reports were requested. A national register of deaths was queried to establish survival. Follow-up imaging was not reviewed, thus local or distant progression was determined from the patient chart or other sources of information. The World Health Organization performance score (PS) was extracted from the patient chart and the Charlson comorbidity index (CCI, excluding the diagnosis of lung cancer) [11] was derived from information in the patient records. Primary endpoints were OS and early death, with the latter defined as death within 6 months after starting radiation therapy. The treatment plans of all patients in the early mortality group were reviewed and dosimetric parameters are reported. As we did not review all the follow-up scans, we chose to define progression-free survival (PFS) as any recurrence, both local and/or distant, after the start of radiation therapy. Secondary endpoints were CRT completion rate, delivered radiation dose and treatment-related toxicity.

63

Toxicity was retrospectively scored using the Common Terminology Criteria for Adverse Events version 4.03 (CTCAE v4.03). We acknowledge the limitation of scoring toxicity retrospectively, and therefore report only grade ≥3 adverse events here. Grade 3 esophagitis referred to severely altered eating/swallowing which required hospitalization and/or tube feeding, and grade 3 pneumonitis was identified when patients had a new requirement for oxygen use. Statistical analysis was conducted using SPSS software package version 20 (SPSS Inc. Chigaco, IL). Median follow-up was computed using the reverse Kaplan–Meier method. Univariate analysis was performed where appropriate. Rates of tumor control and patient survival were calculated using Kaplan–Meier dating from the start of radiation therapy. Comparison between groups was performed using Chi square test for nominal data and log rank analysis for Kaplan–Meier curves. Multivariate analysis was performed using Cox regression for survival outcomes and logistic regression for dichotomous outcomes. Models were built using a forward selection procedure with decisions based on the likelihood ratio statistic. Independent variables included were the PTV/N3 nodal disease subcategories (PTV ≤ 700 cc with N3 nodal disease [≤700 cc], PTV >700 cc without N3 [>700 cc without N3] or PTV >700 cc with N3 [>700 cc with N3]), age, gender, CCI = 0 versus CCI ≥ 1 and the received dose. A p-value of <0.05 was taken as being statistically significant.

3. Results 3.1. General We identified 121 patients who met the study criteria. They had a median age of 62.8 years (range 41.9–79.6 years). Of these patients, 98 (81%) had a PTV >700 cc, of whom 56 (57%) were stage IIIB and 32 (33%) had N3 nodal disease. Twenty-three patients (19%) had a PTV ≤ 700 cc with N3 involvement. Patient characteristics, including the location of N3 disease are shown in Table 1. The median follow-up (FU) for overall survival analysis was 37.6 months (mo) (95%-confidence interval [CI] = 28.4–46.8), and when subdivided into PTV/N3 subcategories, it was respectively 39.3 mo (95%-CI = 22.1–57.7) for the group ≤700 cc, 39.0 mo (95%-CI = 20.5–57.5) for the group >700 cc with N3 and 30.7 mo (95%-CI = 17.8–43.5) for the group >700 cc without N3. The median PTV for all patients was 863 cc (334–2165 cc): 932 cc (705–2165) and 546 cc (334–695) for PTV greater or less than 700 cc, respectively (no significant differences were seen between groups PTV ≥ 700 with and without N3). The mean prescribed and received dose was 6096 cGy (3900–6600) and 5859 cGy (600–7000 (with the differences in upper limits reflecting two additional fractions delivered to compensate for a treatment gap) respectively. A dose of ≥6000 cGy was delivered in 84/121 patients (69.4%), 69 of whom had a PTV >700 cc. A lower dose than the standard protocol at that time was prescribed to four patients (3.3%) in order to meet organ at risk constraints. Seventeen (14%) patients received a dose that was lower than prescribed for the following reasons: toxicity (n = 9), worsening condition (n = 3), metastasis during treatment (n = 1) and death during treatment (n = 4). A total of 16/121 patients (13.2%) did not complete treatment. Reasons for this were: in the ≤700 with N3 group toxicity in 3/23 (13.0%); in the >700 with N3 group 2/32 (6.3%) as a result of worsening condition (n = 1) and toxicity (n = 1) and in the >700 without N3 group 11/66 (16.7%) due to death (n = 4), worsening condition (n = 1), toxicity (n = 5) and metastasis (n = 1). Of the four patient deaths during treatment, two died from possible pulmonary hemorrhage and one from each of neutropenic sepsis and euthanasia.

64

T.G. Wiersma et al. / Lung Cancer 80 (2013) 62–67

Table 1 Patient characteristics. Age (y) (median, range) Gender (n) (%) Male Female Stage (n) IIIA IIIB > 700 cc + N3 Supraclavicular Contralateral mediastinal Both IIIB > 700 cc without N3 IIIB ≤ 700 cc with N3 Supraclavicular Contralateral mediastinal Both PTVa (cc) Median (mean, range) WHOb performance score (n) (%) 0 1 2 Charlson comorbidity index (n) (%) 0 1 2 3 4 5 6 Dose Prescribed (cGy): mean, range Delivered (cGy): mean, range Treatment completion Rate (%) Lung dosed Mean lung dose (cGy): mean (range) V20 (%)c : mean (range) V5 both lungs combined (%): mean (range) V5 contralateral lung (%): mean (range) Spinal cord dose Maximum point dose (cGy): mean (range)

62.8 (41.9–79.6) 84 (69.4%) 37 (30.6%) 42 (34.7%) 32 (26.4%) 14 13 5 24 (19.8%) 23 (19%) 8 13 2 863 (940, 334–2165) 26 (21.5%) 89 (73.6%) 6 (5%) 39 (32.2%) 23 (19.0%) 34 (28.1%) 18 (14.9%) 4 (3.3%) 2 (1.7%) 1 (0.8%) 6096 (3900–6600) 5859 (600–7000) 86.8 1749 (799–2641) 28.8 (10.0–41.2) 68.6 (22.1–90.7) 38.6 (4.1–87.5) 4793 (711–5393)

a

PTV: planning target volume. WHO: World Health Organization. Lung parameters are based on the structure: total lung volume minus PTV and are derived from the dose volume histogram (DVH). d Vx (%): percentage of volume receiving at least × Gy. b

c

3.2. Survival Median OS for the entire group was 15.7 mo (95%CI = 11.7–19.7), with approximately 20% of all patients alive at 4 years (Fig. 1). When subdivided into PTV/N3 subcategories, the median survival was 26.5 mo for the subgroup ≤700 cc with N3, 19.5 mo for the subgroup >700 cc with N3 and 13.2 mo for the subgroup >700 cc without N3. Overall survival differed between the three subgroups (p = 0.010) (Fig. 2). Pairwise comparison of survival curves for these subgroups showed that only curves for ≤700 cc with N3 and >700 cc without N3 differed significantly (p = 0.005). When groups with PTV >700 cc (with and without N3 nodal disease) were combined, overall survival was found to be different when compared to the group ≤700 cc with N3 (p = 0.009). Median survival in the combined group with PTV > 700 was 14.5 mo. Similar results were found using Cox regression analysis and with correction for the received dose. Uncorrected hazard ratios were 2.69 (95%-CI = 1.36–5.34, p = 0.004) for >700 without N3 and 1.86 (95%-CI = 0.88–3.97, p = 0.106) for >700 with N3. The group ≤700 cc with N3 was used as reference. After correction for the received dose, these hazard ratios became 3.56 (95%-CI = 1.77–7.17, p < 0.001) and 1.82 (95%-CI = 0.86–3.87, p = 0.121), respectively. Besides the received dose and PTV/N3 subcategory, no predictive factors for overall survival were observed on multivariate analysis.

Despite limited patient numbers, the importance of volume was also illustrated by the survival differences in patients with stage IIIB NSCLC and PTV >700 cc without N3 versus PTV <700 cc with N3 (Fig. 3). We specifically evaluated early mortality in a total of 26 (21.5%) patients who died within 6 months of starting radiotherapy. In the subgroup PTV ≤700 cc with N3, only 1 (4.3%) early death was observed. The number of early deaths in the subgroups with PTV >700 cc were 5 (15.6%) and 20 (30.3%), respectively, for the groups with and without N3. The retrospective nature of our study means that the cause of death should be treated with caution. Nevertheless, it was estimated that in 11/26 patients had developed ‘clinical’ progression (including documented disease progression and overall clinical deterioration), 9/26 were characterized as having sudden death (including myocardial infarction, pulmonary embolism and bleeding), and 2 patients with each of infection, euthanasia and unknown causes. A significant association was found between early death and PTV/N3 subgroup (p = 0.021). In line with the analysis for overall survival, a multivariate logistic regression analysis showed PTV/N3 subgroup and received dose to be the only significant predictors for early death. Patient characteristics, mode of death or dosimetry for the subgroup with PTV >700 cc are tabulated separately for the early death group and those surviving at least 6 mo (Table 2). Within the subgroup with PTV >700 cc, presence or absence of N3 nodal disease was not found to be associated with early death (chi-square test p = 0.118). Early death was more likely to occur in patients with CCI ≥ 1 when compared to CCI = 0 (32.3% versus 13.9%, p = 0.044). There was no association between early death and performance score (14.3% early deaths when PS = 0 compared to 28.6% early deaths when PS ≥ 1, chi-square test p = 0.183) nor an association between early death and gender (chi-square test p = 0.187) or age (OR = 0.973, 95%-CI = 0.92–1.03, p = 0.317). The median PFS was 11.6 mo for the whole group (95%CI = 8.6–14.6). Median PFS times for the subgroups were 20.6 mo, 9.4 and 11.6 mo for the groups ≤700 cc with N3, >700 cc with and >700 cc without N3, respectively. No significant differences were found between these groups in terms of PFS (p = 0.314). In a multivariate analysis only gender was found to be a significant predictor for PFS with men having better prognosis than women (HR women relative to men = 1.78, 95%-CI = 1.08–2.95, p = 0.025). Thirty-nine patients had documented distant recurrence outside the lungs, predominantly in the brain (18/39, 43.6%) and bone (9/39, 23.1%). The median OS for patients treated with an inverse-planned IMRT containing technique (n = 52, hybrid-IMRT = 46, volumetric arc therapy = 6, median FU 12.9 mo) and a ‘conventional’ (n = 69, median FU 47.3 mo) technique (which could include forward planned, field-in-field IMRT) was 19.2 and 15.2 mo, respectively. This was not statistically significant (p = 0.346) and 95% confidence intervals overlapped (12.3–26.1 versus 11.4–19.0 mo for IMRT and conventional, respectively). Median PTV for the IMRT and conventional groups was 903 cc (341–1820 cc) and 828 cc (334–2165 cc) respectively. 3.3. Toxicity Grade 3 esophagitis and pneumonitis were identified in 33.9% and 4.1% of all patients respectively. There was one recorded grade 5 radiation pneumonitis. Other toxicities scored as grade 5 were bronchopulmonary hemorrhage (n = 3), pulmonary embolism (n = 1), myocardial infarction (n = 1) and sepsis (n = 1). All grade 5 toxicities were in the large-volume group, giving a 7.1% incidence of grade 5 toxicity in patients with PTV >700 cc. Esophagitis grade ≥3 occurred in 47.8% of the group ≤700 cc with N3, 37.5% of the group >700 cc with N3 and in 27.3% of the group >700 cc without N3. There was no association between grade ≥3 esophagitis and the

T.G. Wiersma et al. / Lung Cancer 80 (2013) 62–67

65

1,0

Surviving fraction

0,8

0,6

0,4

0,2

0,0 0

12

24

36

48

60

72

84

96

Time (months) Fig. 1. Overall survival for all patients. No. at risk

121

78

51

42

42

41

40

40



1,0

Surviving fraction

0,8

0,6

≤700cc +N3 0,4

>700cc +N3 0,2

>700cc -N3 p=0.01

0,0

0

12

24

36

48

60

72

84

96

Time (months)

Fig. 2. Overall survival in patients with a Planning Target Volume (PTV) ≤700 cc with N3 nodal disease (≤700 cc + N3), PTV > 700 cc with N3 nodal disease (>700 cc + N3) and PTV > 700 cc without N3 nodal disease (>700 cc − N3). No. at risk PTV ≤700 cc + N3 PTV > 700 cc + N3 PTV > 700 cc − N3

23 32 66

20 23 36

14 16 21

13 12 17

13 12 17

13 12 16

13 – 16

13 – 16

– – –

1.0

Surviving fraction

0.8

0.6

PTV ≤700cc +N3 0.4

0.2

PTV >700cc -N3 p=0.009

0.0 0

12

24

36

48

60

Time (months)

72

84

96

Fig. 3. Overall survival in patients with stage IIIB and PTV ≤ 700 cc + N3 with N3 nodal disease ( ≤ 700 cc + N3) and PTV > 700 cc without N3 nodal disease (>700 cc − N3). No. at risk PTV ≤700 cc + N3 PTV > 700 cc − N3

23 24

20 12

14 6

13 5

13 –

13 –

13 –

13 –

– –

66

T.G. Wiersma et al. / Lung Cancer 80 (2013) 62–67

Table 2 Characteristics of patients with PTV > 700 cc who died within 6 months after starting radiotherapy (n = 25) and of patients with PTV > 700 cc who lived longer than 6 months (n = 73). General characteristics

<6 mo

WHOa performance score (n) (%) 0 3 (12.0) 19 (76.0) 1 3 (12.0) 2 Charlson comorbidity index (n) (%) 0 5 (20.0) 1 5 (20.0) 2 8 (32.0) 3 6 (24.0) ≥4 1 (4.0) Gender Male (n) (%) 15 (60.0) Age (years) Mean (range) 61.7 (45.3–79.6) Esophagitis Grade ≥3 (n) (%) 7 (28.0) a

>6 mo 18 (24.7) 53 (72.6) 2 (2.7) 31 (42.5) 11 (15.1) 19 (26.0) 8 (11.0) 4 (5.5) 54 (74.0) 63.6 (41.8–78.9)

Treatment characteristics Dose Prescribed (cGy): mean, range Delivered (cGy): mean, range Treatment completion Completed (n) (%) Lung Mean lung dose (cGy) V20 (%) V5 both lungs combined (%) V5 contralateral lung (%) Heart Mean heart dose (cGy) V30 (%) Spinal cord Maximum point dose (cGy): mean (range)

<6 mo 5882 (4500–6600) 4693 (600–6600) 15 (60)

> 6mo 6148 (4500–6600) 6234 (4600–7000) 70 (95.9)

1691 (928–2545) 28.3 (14.6–40.6) 54.3 (22.1–90.7) 39.0 (9.6–87.5)

1725 (799–2641) 27.6 (10–41.2) 55.3 (23.4–85.5) 39.8 (4.1–85.6)

2121 (261–4976) 30.8 (1.6–77.2)

NA NA

4668 (711–5393)

4688 (2565–5335)

23 (31.5)

WHO: World Health Organization.

three PTV/N3 subcategories (chi-square test p = 0.176). Also, there was no association between grade ≥3 esophagitis and presence of N3 nodal disease within the subgroup with PTV >700 cc (chi-square test p = 0.303). 4. Discussion This retrospective study of outcomes at a single-institution found that patients who receive CRT and have a large treatment volume, defined as PTV > 700 cc, achieve a median overall survival of 14.5 mo, with about 1 in 5 of such patients surviving 3 years. A large treatment volume that included N3 lymph nodes, did not confer a worse outcome than one without N3 nodes. The median OS was significantly better (26.5 mo) in patients with a PTV ≤ 700 cc with N3. The overall incidence of grade ≥3 esophagitis and pneumonitis were both consistent with a recent systematic review [2]. Our findings suggest that while patients should not be automatically excluded from CRT on the basis of only PTV size or N3 status these factors should be considered in order to further enhance clinical decision-making and to stratify patients in future trials. Our finding that about 1 in 4 patients with PTV >700 cc died within 6 mo of starting CRT, and that this risk was higher in patients with CCI ≥ 1, suggests that comorbidity (CCI) may be an important consideration when treating large tumors. This rate of ‘early-death’ is similar to that extrapolated from Auperin et al. [1]. The challenges of delivering concurrent CRT, even in selected patients with smaller volume tumors, are further illustrated by the dropout rates in contemporary studies. For example, more than 25% of patients dropped out before randomization to consolidation chemotherapy in the HOG-LUN study (which initially predicted a drop-out rate of 10%) and in the GILT trial Huber et al. [12,13]. In contrast to Auperin et al. [1] more patients in our large volume group with early mortality had a performance score of 1 (75% versus 46%). The majority (79%) had comorbid disease. Dutch national guidelines support the use of CRT in PS 0–1 patients [14]. It is perhaps relevant to comment that for the purposes of this study, PS was derived from the radiotherapy chart, which is typically completed after patients have begun chemotherapy. A treatment-related effect cannot be excluded despite plan dosimetry, which complied with institutional protocols. Death during treatment, for reasons related to treatment, is estimated to have occurred in 4% of all 121 patients, which is comparable with published data [2], and although the overall incidence of grade 5 toxicity in patients with a PTV >700 cc was slightly higher at 7% the study is too small to draw firmer conclusions about treatment related mortality.

Future work to better identify and exclude high-risk patients from CRT may have the potential to reduce early mortality. For example a validated quantitative-PCR-based assay has been shown to reliably identify patients with early-stage non-squamous NSCLC at high risk for mortality after surgical resection [15]. In the meantime, patients with a large PTV and a worse prognosis who are also at higher risk of early death (e.g. those with comorbid disease) might benefit from alternative treatment approaches, such as sequential chemotherapy, preceded or followed by hypofractionated radiotherapy. To the best of our knowledge, this is the largest report specifically looking at the use of CRT to treat large volume locally advanced NSCLC. However, potential limitations include: (1) retrospective studies have their limitations and cause of death could not be accurately assessed in the majority of patients – this may be important for improving patient selection and should be addressed in subsequent studies, (2) not all imaging studies were reviewed to determine location/frequency of failure/progression, (3) variations in treatment delivery and calculation of dosimetric parameters might conceivably affect outcomes and the volume of normal tissue reportedly irradiated to a given dose. Although patients with PTV >700 cc without N3 had the worst OS, we have not specifically correlated the location of the PTV relative to organs such as the heart with outcomes which may be useful for further stratification and as an aid to treatment selection; (4) the impact of toxicity related to chemotherapy was not assessed; (5) the group of patients ≤700 cc with N3 was relatively small and we have not included a ≤700 cc without N3 group; finally (6) OS was calculated from start of radiotherapy and not from date of diagnosis for example. Some of these factors may limit the strength of the conclusions that can be drawn. 5. Conclusion Patients undergoing CRT for stage III NSCLC with a PTV >700 cc, with or without N3 nodal disease, had a significantly shorter OS than patients with PTV ≤ 700 cc with N3. Although the rate of acute grade ≥3 esophageal toxicity was not prohibitive, patients with PTV > 700 cc, in particular those with CCI ≥ 1, had a significantly higher chance of early death. As longer-term survival has been observed, patients should not be automatically excluded from receiving CRT on the basis of only the PTV or N3 disease, however both the PTV and CCI should be considered in clinical decision making and used as stratification factors in future trials.

T.G. Wiersma et al. / Lung Cancer 80 (2013) 62–67

Conflict of interest statement The Department of Radiotherapy, VU University Medical Center has Research Agreements with Varian Medical Systems Inc., Palo Alto, USA; Brainlab AG, Feldkirchen, Germany. 1. TW declares no conflict of interest. 2. MD has received travel support ± honoraria from Varian Medical Systems and Brainlab. 3. WV has received travel support ± honoraria from Varian Medical Systems. 4. PvdV declares no conflict of interest. 5. PdH declares no conflict of interest. 6. ES declares no conflict of interest. 7. EvR declares no conflict of interest. 8. BS has received travel support and honorarium from Varian Medical Systems and Brainlab. 9. SS has received speakers honoraria from Varian Medical Systems, and is also a member of the Trial Management Group for a phase III study in lung cancer sponsored by Lilly Oncology. Funding No funding for this study. Acknowledgments No acknowledgments. References [1] Auperin A, Le PC, Rolland E, Curran WJ, Furuse K, Fournel P, et al. Meta-analysis of concomitant versus sequential radiochemotherapy in locally advanced nonsmall-cell lung cancer. J Clin Oncol 2010;28(13):2181–90. [2] O’Rourke N, Roque IF, Farre BN, Macbeth F. Concurrent chemoradiotherapy in non-small cell lung cancer. Cochrane Database Syst Rev 2010;(6): CD002140. [3] Zwitter M. Dutch statistics on lung cancer: sobering experience for a new approach. J Thorac Oncol 2012;7(2):269–71.

67

[4] Bradley JD, Ieumwananonthachai N, Purdy JA, Wasserman TH, Lockett MA, Graham MV, et al. Gross tumor volume, critical prognostic factor in patients treated with three-dimensional conformal radiation therapy for non-small-cell lung carcinoma. Int J Radiat Oncol Biol Phys 2002;52(1):49–57. [5] Furuse K, Fukuoka M, Kawahara M, Nishikawa H, Takada Y, Kudoh S, et al. Phase III study of concurrent versus sequential thoracic radiotherapy in combination with mitomycin, vindesine, and cisplatin in unresectable stage III non-smallcell lung cancer. J Clin Oncol 1999;17(9):2692–9. [6] Belderbos J, Uitterhoeve L, van ZN, Belderbos H, Rodrigus P, van d V, et al. Randomised trial of sequential versus concurrent chemo-radiotherapy in patients with inoperable non-small cell lung cancer (EORTC 08972-22973). Eur J Cancer 2007;43(1):114–21. [7] Uitterhoeve AL, Koolen MG, van Os RM, Koedooder K, van de KM, Pieters BR, et al. Accelerated high-dose radiotherapy alone or combined with either concomitant or sequential chemotherapy; treatments of choice in patients with non-small cell lung cancer. Radiother Oncol 2007;2:27. [8] Bradley J, Paulus R, Komaki R, Masters G, Forster K, Schild S, et al. A randomized phase III comparison of standard-dose (60 Gy) versus high-dose (74 Gy) conformal chemoradiotherapy+/− cetuximab for stage IIIa/IIIb non-small cell lung cancer: preliminary findings on radiation dose in RTOG 0617. In: Proc ASTRO. 2011 [abstract]. [9] Phernambucq EC, Spoelstra FO, Verbakel WF, Postmus PE, Melissant CF, Maassen van den Brink KI, et al. Outcomes of concurrent chemoradiotherapy in patients with stage III non-small-cell lung cancer and significant comorbidity. Ann Oncol 2011;22(1):132–8. [10] Verbakel WF, van RE, Ladenius-Lischer I, Cuijpers JP, Slotman BJ, Senan S. Clinical application of a novel hybrid intensity-modulated radiotherapy technique for stage III lung cancer and dosimetric comparison with four other techniques. Int J Radiat Oncol Biol Phys 2012;83(2):e297–303. [11] Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47(11):1245–51. [12] Huber RM, Engel-Riedel W, Kollmeier J, Andreas S, Staar S, Klautke G, et al. GILT study: oral vinorelbine (NVBo) and cisplatin (P) with concomitant radiotherapy (RT) followed by either consolidation (C) with NVBo plus P plus best supportive care (BSC) or BSC alone in stage (st) III non-small cell lung cancer (NSCLC): final results of a phase (ph) III study. J Clin Oncol 2012;30(15):7001 [abstract]. [13] Hanna N, Neubauer M, Yiannoutsos C, McGarry R, Arseneau J, Ansari R, et al. Phase III study of cisplatin, etoposide, and concurrent chest radiation with or without consolidation docetaxel in patients with inoperable stage III nonsmall-cell lung cancer: the Hoosier Oncology Group and U.S. Oncology. J Clin Oncol 2008;26(35):5755–60. [14] Landelijke Werkgroep Longtumoren. Niet-kleincellig longcarcinoom; Landelijke richtlijn versie: 2.0, http://www.oncoline.nl/niet-kleincelliglongcarcinoom [accessed 22.05.12]. [15] Kratz JR, He J, Van Den Eeden SK, Zhu ZH, Gao W, Pham PT, et al. A practical molecular assay to predict survival in resected non-squamous, non-smallcell lung cancer: development and international validation studies. Lancet 2012;379(9818):823–32.