Abstract
Aim: We evaluated the relationship of early metabolic responses on 18F-fluorodeoxyglucose–positron-emission tomography/computed tomography (PET/CT) performed within one month after concurrent chemoradiotherapy (CCRT) with local tumor control and survival in patients with advanced stage III non-small cell lung cancer (NSCLC). Patients and Methods: One hundred and nineteen patients with unresectable stage III NSCLC who completed definitive CCRT were included. PET/CT was performed 2-4 weeks after completion of radiotherapy. Results: The maximum standardized uptake value (SUVmax) reduction ratio of the primary lesion (primary SRR, 80%, p<0.001), gross tumor volume (150 cm3, p=0.036), and pre-radiotherapy ratio of SUVmax of the metastatic lymph node to that of the primary lesion (60%, p=0.05) were significantly associated with OS in multivariate analysis. The primary SRR was the only statistically significant parameter for local control. Conclusion: Early metabolic response of the primary lesion after CCRT correlated with local control and overall survival in patients with unresectable stage III NSCLC.
Concurrent chemoradiotherapy (CCRT) has been proven as the treatment of choice for patients with advanced stage III non-small cell lung cancer (NSCLC). However, some patients do not benefit from CCRT, and a high percentage of patients experience disease progression within the radiation field (1). The importance of tumor response and ultimate local control has been revealed to be associated with survival of patients with unresectable NSCLC (1-3). Immediate assessment of tumor response could help identify patients who would benefit from adjuvant therapy at a time when the disease course could be influenced by an alternative therapy. The combination of CCRT and surgery to treat stage III NSCLC was reported to improve overall survival (OS), and surgical benefit is generally limited to a subset of patients with adequate down-staging (4-7).
The role of 18F-fluorodeoxyglucose (18F-FDG)–positron-emission tomography/computed tomography (PET/CT) in radiation oncology has been widely studied and well-outlined in recent articles (8, 9). PET/CT can provide biological and physiological information about residual tumor viability that is available before information on morphological changes and could thus provide more accurate information for the selection of patients who would benefit from intensive treatment after chemotherapy or radiotherapy (RT) (10, 11). In the present study, we aimed to determine the relationship between early tumor responses shown by PET data and local tumor control or survival in patients with advanced stage III NSCLC who were treated with curative CCRT. If the accuracy of the predictive value of metabolic response is high, it could be used to guide patient stratification when designing clinical trials that evaluate the benefits of tri-modal therapy or consolidation chemotherapy after CCRT.
Patients and Methods
This was a retrospective cohort study of a prospective database. The Institutional Review Board of Chonnam National University Hwasun Hospital approved this study, 2013-159. A total of 119 consecutive patients with unresectable stage III NSCLC who had completed definitive CCRT between January 2005 and December 2010 were included. Unresectability was determined by the attending thoracic surgeons, and CCRT eligibility was determined by the radiation oncologists. Tumors were histologically-documented as NSCLC, and the staging work-up included brain magnetic resonance imaging and PET/computed tomographic, in addition to baseline laboratory studies, pulmonary function tests, and chest CT. The tumor stage was classified according to the American Joint Committee on Cancer, seventh edition, staging system. Eligible criteria for CCRT included an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1 and a weight loss of <10% within the six months preceding diagnosis.
The gross tumor volume (GTV) comprised the primary tumor and metabolically-positive lymph nodes on PET/computed tomographic, or a lymph node of which the short axis diameter exceeded 14 mm on CT. The clinical target volume (CTV) included the GTV with a 6-8 mm margin and the involved mediastinal lymph node stations. The CTV with a 5-7 mm margin was defined as the planning target volume (PTV). RT was performed according to the CT simulation with 10-MV X-rays. The planned radiation dose was 60-66 Gy, given in 2-2.4 Gy fractions over five or six weeks. A CCRT doublet regimen of weekly cisplatin (20 mg/m2) and paclitaxel (45 mg/m2) was administered for five or six weeks.
All PET/CT studies were performed using a Discovery LT scanner (GE Healthcare, Milwaukee, WI, USA) after at least 6 h of fasting and 60 min after the intravenous injection of 370-740 MBq of 18F-FDG, depending on the patient's weight (7.5 MBq/kg). CT images were acquired from the thigh to the head with an 8-slice helical CT scanner and without the administration of contrast agents for attenuation mapping and anatomical information. PET images were then acquired over the same anatomical area after FDG administration, with an imaging time of 3 min per bed position. CT-based attenuation-corrected PET images were reconstructed using an ordered-subset expectation maximization algorithm. Accurate co-registration of the CT and PET images was performed using commercially available software (Advantage Workstation; GE Healthcare). Abnormal FDG uptake was defined as an uptake rate greater than the background activity in the surrounding tissue, and the FDG uptake intensity was quantified by calculating the standard uptake value (SUV). The SUV was calculated from the amount of FDG injected, the total body weight, and soft-tissue uptake in the attenuation-corrected regional images as follows: SUV=(activity/unit volume)/(injected dose/total body weight). The SUVmax was defined as the peak SUV of the pixels with the highest counts within the region of interest. We assessed four PET/CT parameters to analyze the association with local tumor control and survival: the pre and post-RT SUVmax, the change in SUVmax (ΔSUVmax=pre-RT SUVmax–post-RT SUVmax), the SUV reduction ratio (SRR; SRR=[pre-RT SUVmax–post-RT SUVmax]×100/pre-RT SUVmax), and the pre-RT ratio of the SUVmax of metastatic lymph node to that of the primary lesion (pre-RT N/P ratio; pre-RT N/P ratio=pre-RT SUVmax of the primary lesion/pre-RT SUVmax of the metastatic lymph node).
After RT completion, PET/CT was performed within one month, and regular follow-ups were scheduled at 3-month intervals with alternating chest CT and PET-CT scans up to two years after RT and at 4-6-month intervals for the following three years. The median follow-up duration for these patients was 21 months (range=3-44 months).
Local control (LC) was assessed using a combination of clinical assessments, CT, PET/CT, or pathology. Local failure was defined as an increase in radiological abnormalities on CT or a serial escalation of the SUVmax on PET/CT within the RT field that was not judged to be radiation-induced pneumonitis. The duration of LC was defined as the duration from the date of start of RT to the date of local failure or last follow-up for censored patients. OS was defined as the elapsed time from the RT start date to the date of death from any cause or the last follow-up for censored patients.
The Kaplan–Meier method was used to plot survival curves, which were analyzed using the log-rank test to determine prognostic factors in univariate analysis. The Cox regression model was used to identify the prognostic factors in multivariate analysis. All statistical analyses were performed using SPSS, version 19.0 (SPSS Inc., Chicago, IL, USA). Null hypotheses with no difference were rejected if the p-values were less than 0.05.
Results
The median patient age in the study was 68 years (range=37-84 years), and 113 patients were male. The histological diagnosis was squamous cell carcinoma in 82 patients. The other patient characteristics are shown in Table I. The median interval between the date of the initial diagnostic PET/CT and RT initiation was 13 days, and RT was started within four weeks after PET/CT scanning. The median pre-RT primary SUVmax and pre-RT nodal SUVmax before CCRT were 14.5 and 7.7, respectively. The pre-RT N/P ratio was 61.07%. The median interval from the date of RT completion to follow-up PET/CT imaging was 10 days, and all follow-up PET/CT were performed within four weeks after RT completion. The median post-RT primary SUVmax and primary ΔSUVmax values were 3.8 and 9.6, respectively (Table II).
At the time of analysis, 56 patients (47.3%) were alive. The 2- and 3-year OS rates for all 119 patients were 54.1% and 47.3%, respectively. The median survival time was 28 months. The GTV (≤150 cm3 vs. >150 cm3; p=0.011), CTV (≤450 cm3 vs. >450 cm3; p=0.037), post-RT primary SUVmax (≤4 vs. >4; p=0.002), primary ΔSUVmax (≥10 vs. <10; p=0.006), primary SRR (≥80% vs. <80%; p<0.001), and pre-RT N/P ratio (≥60% vs. <60%; p<0.037) were significantly associated with OS in univariate analysis (Table III). However, only GTV (p=0.036), primary SRR (p=0.001), and pre-RT N/P ratio (p=0.050) were statistically significant prognostic factors in multivariate analysis (Figure 1, Table III).
In the first failure pattern, isolated local recurrences were noted in 48 (40.3%) patients and isolated distant recurrences were observed in 19 (15.9%) patients; both local and distant failures occurred simultaneously in 9 (7.6%) patients. The actuarial 2- and 3-year LC rates were 46.3% and 41.8%, respectively. The median time-to-local progression was 18 months.
Patient characteristics.
The post-RT SUVmax (p=0.013) and primary SRR (p=0.010) were statistically significant prognostic factors for LC in univariate analysis. However, the primary SRR (≥80% vs. <80%; p=0.010) was the only statistically significant parameter in multivariate analysis (Figure 2, Table III). Eight (22.9%) out of 35 patients with primary SRR values ≥ 80% and 40 (47.7%) out of 84 patients with primary SRR values <80% experienced local failure (Table IV). Among the 48 patients with isolated local failure, recurrence occurred at the primary tumor lesion-alone in 29 patients (60.4%), at lymph nodes alone in five patients (10.4%), and simultaneously in the primary tumor lesion and lymph nodes within the RT field in the remaining 14 patients (29.2%). In patients with primary SRR values ≥80%, five patients (14.3%) had isolated recurrences at the primary tumor site. However, 24 patients (28.6%) with primary SRR values <80% had isolated recurrences at the primary tumor site.
18F-fluorodeoxyglucose–positron-emission tomography/computed tomography characteristics.
Discussion
PET/CT is a tool to gain biological information from tumors and the various PET/CT parameters associated with clinical outcomes in patients with NSCLC, with inconsistent results (12-16). Primary SRR, which means the change in the SUVmax of the primary tumor, has been studied as a prognostic or predictive factor and has a near-linear relationship with the percentage of non-viable tumor cells (13, 15, 16). There was a report that SRR value after neoadjuvant chemoradiation predicted pathologically complete regression with an accuracy of 95% in surgical NSCLC specimens. When the SUVmax decreases by 80% or more, the patient is likely to show a complete pathological response, irrespective of cell type, mediastinal lymph node stage, neoadjuvant treatment, or the final absolute SUVmax (15). SRR might be a potent predictor of radiation effects on tumors and a predictive factor for treatment outcomes in patients with advanced NSCLC. Among patients with stage III NSCLC treated with induction chemotherapy and CCRT followed by complete surgical resection, those with a SRR value >60% after induction therapy had longer survival than did patients below this threshold (16). In a study by MacManus et al. of metabolic responses after radical RT/chemoradiotherapy for NSCLC, complete metabolic response was shown to be a positive prognostic factor that correlated with treatment failure and survival (17).
Predictive factors on overall survival rate and local control rate (n=119).
Kaplan–Meier curve of overall survival rate according to primary maximum standardized uptake value reduction ratio (SRR).
The other consideration is that follow-up PET/CT was performed at a median of 70 days after treatment, which seems rather late for decisions about adjuvant therapy after CCRT. Our study evaluates the association between early responses on PET/CT with LC and survival, a method that could be used to guide the decision regarding adjuvant therapy immediately without a clinically significant delay after CCRT in patients with advanced-stage NSCLC. In the present study, primary SRR was the only statistically significant PET/CT parameter that affected LC and OS in multivariate analysis. Patients with primary SRR values ≥80% showed improvements in both LC and OS. This can be explained by the fact that patients with primary SRR values ≥80% might have a high probability of pathologically-complete tumor cell clearance and a greater probability of achieving LC, resulting in a survival benefit.
Cerfolio et al. reported that an N/P ratio >0.56 could predict nodal malignancy with 94% sensitivity and 72% specificity (18). This suggests that the N/P ratio could be a predictor of nodal malignancy. In our study, distant failure occurred in 15 out of 59 patients (25.5%) with pre-RT N/P ratios >60% and in 13 out of 60 patients (21.7%) with pre-RT N/P ratios ≤60%. Simultaneous local and distant failure was more frequent in patients with pre-RT N/P ratios >60% (six patients;10.2%) than in patients with pre-RT N/P ratios ≤60% (three patients; 5.0%, data not shown). In the present study, the N/P ratio could be used to predict disease progression beyond the RT field in RT-treated patients as well as the survival impact.
Kaplan–Meier curve of local control rate according to primary maximum standardized uptake value reduction ratio (SRR).
Relationship between SUVmax reduction ratio (SRR) of the primary tumor and the pattern of first failure.
PET/CT during or after RT might be useful for evaluating treatment responses, but no consensus has been reached regarding the optimal timing during or after RT. Cerfolio et al. reported that the optimal time to perform a repeat FDG-PET/CT procedure after completing neoadjuvant chemotherapy and high-dose RT in order to maximize its accuracy for re-staging of patient with NSCLC is approximately one month after the last dose of radiation (19). Huang et al. reported that SUVmax changes between two serial PET/CT procedures, taken before and four weeks after chemoradiotherapy, allow for early prediction of treatment responses in patients with advanced NSCLC (20).
A limitation of our study was the timing of the follow-up PET procedure. PET was performed at between eight days and one month after CCRT completion, and >70% of patients underwent follow-up PET/CT within two weeks after RT completion. Therefore, an SRR cut-off point of 80%, suggested to be predictive of clinical outcomes, should cautiously be used as a tumor response criterion when designing clinical trials to determine which patients with NSCLC might benefit from further treatment after CCRT.
In conclusion, the primary SRR (≥80% vs. <80%) was a statistically significant parameter associated with both LC and OS. Our data showed a relationship between early metabolic responses and LC and OS. We suggest follow-up PET/CT evaluations at one month when designing clinical trials in order to make early decisions, thus enabling the provision of additional consolidation chemotherapy or surgery without treatment delays. Prospective clinical trials for the verification of SRR as an indicator for surgical resection or systemic consolidation chemotherapy after CCRT might be warranted in order to guide individualized treatment in patients with advanced stage III NSCLC.
- Received December 16, 2013.
- Revision received February 3, 2014.
- Accepted February 5, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.