Use of 18F-FDG PET/CT as a Predictive Biomarker of Outcome in Patients With Head-and-Neck Non–Squamous Cell Carcinoma
Abstract
OBJECTIVE. The purpose of this article is to establish whether pretreatment 18F-FDG uptake predicts disease-free survival (DFS) and overall survival in patients with head-and-neck non–squamous cell carcinoma (SCC).
MATERIALS AND METHODS. Eighteen patients (six women and 12 men; mean [± SD] age at diagnosis, 57.89 ± 13.54 years) with head-and-neck non-SCC were included. Tumor FDG uptake was measured by the maximum standardized uptake value (SUVmax) and was corrected for background liver FDG uptake to derive the corrected SUVmax. Receiver operating characteristic analyses were used to predict the optimal corrected SUVmax cutoffs for respective outcomes of DFS (i.e., absence of recurrence) and death.
RESULTS. The mean corrected SUVmax of the 18 head-and-neck tumors was 5.63 ± 3.94 (range, 1.14–14.29). The optimal corrected SUVmax cutoff for predicting DFS and overall survival was 5.79. DFS and overall survival were significantly higher among patients with corrected SUVmax < 6 than among patients with corrected SUVmax ≥ 6. The mean DFS for patients with corrected SUVmax < 6 was 25.7 ± 11.14 months, and the mean DFS for patients with corrected SUVmax ≥ 6 was 7.88 ± 7.1 months (p < 0.018). Among patients with corrected SUVmax < 6, none died, and the mean length of follow-up for this group was 35.2 ± 9.96 months. All of the patients who died had corrected SUVmax ≥ 6, and the overall survival for this group was 13.28 ± 12.89 months (p < 0.001).
CONCLUSION. FDG uptake, as measured by corrected SUVmax, may be a predictive imaging biomarker for DFS and overall survival in patients with head-and-neck non-SCC.
Over 40,000 people are diagnosed with head-and-neck cancer each year in the United States [1]. Approximately 95% of these tumors are squamous cell carcinomas (SCCs) [2]. The remaining 5% are referred to collectively as non-SCCs and include cancers derived from neuroendocrine tissue, connective tissue, soft tissue, salivary, thyroid and parathyroid glands, epithelial cells, and immune cells [2]. The clinical presentation of these lesions, their treatment, and the prognosis is highly variable, depending on the type of tumor, its histologic variant and grade, and the extent of spread at the time of biopsy [3, 4].
For patients with head-and-neck SCC, 18F-FDG PET has been a useful prognostic marker to assess treatment outcome, regardless of the primary treatment modality [5, 6]. It was found that tumors with higher FDG uptake were at greater risk of treatment failure, and it was concluded that these lesions should be considered for more aggressive therapy [5]. Tumor uptake of FDG has been associated with various cellular characteristics, such as cell viability and proliferative activity [5, 7].
The purpose of this study is to establish whether pretreatment FDG uptake of head-and-neck non-SCC predicts disease-free survival (DFS) and overall survival.
Materials and Methods
The study was conducted as an observational retrospective review of patient records and a prospective review of PET/CT images. Prior approval from our institutional review board was obtained with an approval of waiver of informed consent. HIPAA guidelines were followed. Patients were identified from the institutional tumor registry between August 2004 and January 2007 with newly diagnosed biopsy-proven head-and-neck non-SCC. Patients with thyroid carcinoma were excluded: 85–95% of thyroid carcinoma is papillary thyroid carcinoma, which, compared with other head-and-neck non-SCCs, has a much more favorable prognosis (10-year survival, > 90%) [8]. Patients who underwent a pretreatment FDG PET/CT study at our institution were included in this study. Patients with a history of prior head-and-neck malignancy and those with a history of non–head-and-neck cancer who were not disease free for at least 3 years were excluded.
One hundred fifty-six patients diagnosed with head-and-neck cancers who underwent a baseline PET/CT study at our institution were identified from the institutional data warehouse. One hundred twelve patients with SCCs were excluded. Of the 44 patients with non-SCC and a baseline PET/CT, 21 patients with thyroid carcinomas were excluded. From the remaining 23 patients, five additional patients were excluded for the following reasons: two patients had a history of head-and-neck SCC within 3 years of identification of the primary tumor, two patients were being treated for another cancer not in the head and neck, and one patient had no follow-up or treatment history.
Patients
The study population of 18 patients with a mean (± SD) age at diagnosis of 57.89 ± 13.54 years forms the basis for this analysis. Of these patients, six were women (62.50 ± 10.67 years old at diagnosis) and 12 were men (55.98 ± 14.64 years old at diagnosis). The number of patients in this study group with known T stages was 16; four patients had T1 lesions, three patients had T2 lesions, none had T3 lesions, and nine patients had T4 lesions. Two patients did not have a designated T stage: the first of the two had stage II Hodgkins lymphoma, and the second had stage IV large B cell non-Hodgkins lymphoma. For all patients, including those who died of disease, the mean length of follow-up was 20.06 ± 12.74 months, and the median length of follow-up was 22 months (range, 1–40 months). Among patients who did not die of disease, the mean length of follow-up was 24.31 ± 10.17 months, and the median length of follow-up was 26 months (range, 6–40 months).
Data Collection
Data collection included patient age, sex, and disease stage at presentation using the American Joint Committee on Cancer 2002 TNM staging [9]; risk factors; treatment received; local, regional and distant disease control; and mortality data at last follow-up. Overall mortality data were obtained from the Social Security database in July 2009, and survival was measured from the date of the original positive PET/CT to July 2009 in patients who were still alive, or to the date of death for the patients who had died [10].
PET/CT
All PET/CT studies were performed on a 16-MDCT PET/CT scanner (GE Discovery, GE Healthcare) according to the institutional standard clinical protocol. Weight, height, and blood glucose levels were recorded for all patients. All patients had a blood glucose level less than 200 mg/dL and were injected with 10–15 mCi of FDG with an incubation period of about 60 minutes. The amount of injected radioactivity was routinely measured by quantification of the radioactivity of the syringe before and after injection. All patients were scanned from skull to thigh in supine position, according to the institutional clinical protocol. All patients underwent a dedicated head-and-neck acquisition. The body scans were performed first in the caudocranial direction to avoid accumulation of FDG in the bladder with the arms up, and then dedicated head-and-neck scans were performed from skull base to aortic arch with the arms down. This protocol avoids scanning the head and neck twice to reduce the radiation dose from CT.
The dedicated head-and-neck PET scans were done using 2D imaging with emission scans lasting 5–6 minutes and an FOV of 30 cm. The matrix size was 128 × 128, and slice thickness was 3.3 mm. The contrast-enhanced CT scan of the head and neck was done with an injection of 60 mL of ioversol (Optiray 320, Tyco Health Care) at a rate of 3 mL/s and was chased with 30 mL of saline at the same rate. Imaging started 40 seconds after injection. The pitch was 0.984, and the tube current was 440 mA. Slices were reconstructed at 1.25-mm thickness and 1.25-mm spacing for review and also at 3.3 mm to match the PET images for fusion. Two-dimensional imaging with longer bed time per position, higher FOV, and IV contrast agent improves the diagnostic quality of the dedicated head-and-neck PET/CT.
Trunk (clavicle to mid thigh) PET scans were done using 3D imaging with emission scans that ranged from 2 to 4 minutes with an FOV of 50 cm. The CT scans were performed to match the PET scans’ FOV and slice thickness. Two types of CT scans were done: low-dose scans for only attenuation correction and lesion localization of the PET data, or high-dose scans for attenuation correction, lesion localization, and diagnostic level of the CT scan. Both scans were done using a 512 × 512 matrix. The low-dose scans were done without IV contrast agent. The pitch was approximately 1.75, and collimation was 10 mm. Slices were reconstructed at 3.75-mm thickness and with a 3.27-mm spacing. The x-ray beam was set at 120 kV, and the tube current was modulated but ranged from low-dose levels (maximum, 150 mA) to full-dose levels. The high-dose scans of the trunk were done similarly to the low-dose scans except for the use of oral and IV contrast material and higher x-ray dose. A total of 100 mL of ioversol (Optiray 320, Tyco Health Care) was given by IV injection at a rate of 3 mL/s and was followed by a saline chase of 30 mL at the same rate. Commercial software (SmartPrep, GE Healthcare) was used to trigger imaging once the celiac trunk reached a density of 180–200 HU.
Standardized Uptake Value Measurements
All PET/CT studies were retrieved from our institutional electronic archival system and were prospectively reviewed on a workstation (GE Advantage, GE Healthcare) by the authors for image analysis. One author, who is a board-certified radiologist and is fellowship-trained in nuclear medicine and neuroradiology and has 18 months of experience as faculty, performed all the measurements. For the purposes of this study, the relevant standardized uptake values (SUVs) for calculation and reporting were maximum SUV (SUVmax) and mean SUV within the primary tumor. These were determined by visually identifying the region or regions on the PET images (in axial plane) that qualitatively appeared to have the most intense FDG uptake and that corresponded to a known tumor on the basis of other data (CT scan). For SUVmax and mean SUV determination, a region of interest (ROI) incorporating the gross tumor volume was identified, and the manufacturer’s (GE Healthcare) algorithm was used to calculate the SUVmax within this ROI. The manual ROI diameter included the area of FDG avidity of the primary tumor or the most avid lymph node in the case of patients for whom no primary tumor was identified. The background SUVmax and mean SUV were measured from the liver. The constant manual ROI diameter of 30 mm was used to measure the liver SUV parameters in the anterior aspect of the right lobe of the liver. All tumor SUV parameters were corrected for background FDG uptake by dividing the measured SUVmax and mean SUV of the primary tumor by the mean SUV of the liver to derive the corrected SUVmax (Fig. 1).
Statistical Methods
All analyses were performed using the R Statistical System (version 2.8.0, R Foundation). All hypothesis tests performed were two-sided with significance level of p equals 0.05. Summary statistics are presented as mean and SD for continuous variables, or frequency and percentage for categoric variables. Observed SUVmax, mean SUV, and corrected SUVmax values were distributed approximately lognormally, and so these values were log-transformed before analysis to satisfy distributional assumptions. The independent samples t test or analysis of variance was used for between-group comparisons of mean corrected SUVmax values. For respective outcomes of progression to disease and death, separate receiver operating characteristic analyses were used to estimate sensitivity and specificity for each corrected SUVmax cutoff. Finally, Kaplan-Meier curves were used to plot estimated survival curves for DFS and overall survival, and the log-rank test was used for between-group comparisons of survival time in the presence of censored data.
Fig. 1 —49-year-old man with left parotid adenocarcinoma.
A, Region of interest (ROI) was drawn to incorporate FDG uptake in single axial slice of left parotid lesion to establish uncorrected maximum standardized uptake value (SUVmax).
B, Standard 3-cm ROI was drawn in right lobe of liver to establish background FDG uptake (mean SUV) to calculate corrected SUVmax of primary left parotid lesion (corrected SUVmax = SUVmax of primary tumor / mean SUV of liver).
Results
SUV Measurements
The mean SUVmax of the 18 primary head-and-neck lesions was 12.52 ± 8.69 (range, 0.80–28.50). The mean SUV of the 18 primary head-and-neck lesions was 9.37 ± 6.28 (range, 0.70–22.60). The mean liver SUV was 2.20 ± 0.67 (range, 0.70–3.40). There was a significant positive correlation between the liver SUVmax and liver mean SUV (r2 = 0.97; p < 0.0001). The mean SUVmax of the 18 head-and-neck lesions corrected for their respective livers (corrected SUVmax = tumor SUVmax / liver mean SUV) was 5.63 ± 3.94 (range, 1.14–14.29).
Corrected SUVmax and Disease-Free Survival
Nine of 18 (50%) patients in this study had disease progression. Nine patients were disease free with a median follow-up of 26 months (range, 6–32 months). The mean corrected SUVmax of the nine patients with disease progression was 7.47 ± 4.48 (range, 1.14–14.29). The mean corrected SUVmax of the nine patients who were disease free was 3.80 ± 2.34 (range, 1.70–7.96). The corrected SUVmax of individuals’ lesions with disease progression was higher than the corrected SUVmax of the lesions of those without progression, but not significantly so (7.47 ± 4.48 vs 3.80 ± 2.34; p = 0.054) (Fig. 2).
The optimal corrected SUVmax cutoff for predicting disease progression in our patients was 5.79 (sensitivity, 67%; specificity, 78%). DFS was significantly higher among patients with corrected SUVmax < 6 than among patients with corrected SUVmax ≥ 6 (p = 0.0186). The mean DFS for patients with corrected SUVmax < 6 was 25.7 ± 11.14 months, and that for patients with corrected SUVmax ≥ 6 was 7.88 ± 7.1 months (Fig. 3).
Corrected SUVmax and Overall Survival
One of 18 patients left the United States and survival data are unknown. Of the remaining 17 patients, five (29.4%) died during the follow-up period. The mean corrected SUVmax of the five patients who died was 10.16 ± 3.82 (range, 6.27–14.29). The mean corrected SUVmax of the 12 patients who survived was 3.62 ± 2.19 (range, 1.14–7.96). The corrected SUVmax of individuals’ lesions who died was significantly higher than that of the lesions of those who survived (10.16 ± 3.82 vs 3.62 ± 2.19; p = 0.01) (Fig. 4).
The optimal corrected SUVmax cutoff for predicting death was 5.79, which was the same as the value for predicting DFS with corrected SUVmax, with sensitivity of 100% (95% CI, 48–100%) and specificity of 83.3% (95% CI, 51–97%). Overall survival was significantly higher among patients with corrected SUVmax < 6 than among patients with corrected SUVmax ≥ 6 (p = 0.001). Among patients with corrected SUVmax < 6, none died of disease, and the mean length of follow-up for this group was 35.2 ± 9.96 months. All of the patients who died had corrected SUVmax ≥ 6, and the overall survival for this group was 13.28 ± 12.89 months (Fig. 5).
Discussion
Tumor staging, as defined by measuring the extent of local, regional lymph node, and distant metastases, are significant predictors of survival in patients with head-and-neck non-SCC [1, 3]. However, it is currently difficult to predict the reliability of the outcome of treatment, even in patients within the same tumor or nodal category [11]. Therefore, determining additional prognostic factors is of great interest [12].
FDG PET has been shown to be useful in the diagnosis and management of head-and-neck cancer [13–18]. The intensity of FDG uptake is emerging as a valuable predictive factor regarding treatment outcome [5, 11, 19, 20]. FDG PET has been shown to accurately predict the responsiveness of cancer to the initial phase of radiotherapy or chemotherapy [21]. Furthermore, pretreatment tumor FDG uptake has been shown to represent an independent prognostic factor in patients with head-and-neck cancers, whatever the primary treatment modality [7]. A higher pretreatment SUV has been shown to correlate with more residual viable tumor cells documented histologically 4 weeks after treatment of head-and-neck cancer [22]. In a prospective study, Allal et al. [5] found that advanced T category (T3–T4) correlated significantly with local control of head-and-neck SCC, whereas the SUV category had a significantly negative effect on both local control and DFS.
A corrected SUVmax cutoff of 1.48 has been shown to differentiate tonsillar SCC from physiologic tonsillar FDG uptake with 100% specificity and sensitivity [23]. For head-and-neck non-SCC, the evidence for FDG uptake as a prognostic indicator is limited. In a study by Roh et al. [24], no significant difference in SUV was found between high- and low-grade salivary tumors. FDG uptake was not found to be associated with local tumor invasion, FDG uptake was not found to be a prognostic indicator in salivary gland cancer, and using an SUV cutoff of 4 did not show a statistically significant difference in DFS. However, Suh et al. [25] found that FDG uptake by primary extranodal natural killer/T cell lymphoma, another non-SCC, was higher in nonresponders to therapy and in patients with poor survival outcomes (7.0 ± 3.0 vs 4.0 ± 1.7). Differences in SUV relative to histologic grade were significant, but there were considerable overlaps between subgroups. The findings of that study indicated that FDG uptake may reflect aggressiveness of extranodal natural killer/T cell lymphoma and that SUV was a prognostic indicator, because an SUV greater than 5.5 was virtually indicative of aggressive disease and resistance to therapy.
Our results show that corrected SUVmax measurements from FDG PET of head-and-neck non-SCC can be used as a tool for predicting patient outcome, despite varying tumor types and treatment modalities, similar to the way that staging drives prediction of survival. This may identify patients who have more aggressive tumors so that their treatments can be intensified or tailored to improve therapeutic outcome. In our analysis, SUVs were corrected for background uptake by dividing the measured SUVmax of the primary tumor by the mean SUV of the liver to derive the corrected SUVmax. This correction was performed because the SUV parameters can be influenced by factors such as weight, plasma glucose level, incubation time, partial volume effect, and reconstruction algorithm, leading to variability between patients [26–30]. Comparison of uptake values between patients also assumes that PET scans are performed after comparable doses of FDG and incubation times before scanning, which is not always the case [31]. Thus, by comparing lesion SUVmax to background FDG accumulation using the SUV of the liver, variability in SUV may be reduced so that reliable intrabiologic and interbiologic comparisons can be made. This is useful in clinical trial settings where standardization between scanners and patients is necessary to arrive at meaningful results.
Our study results need to be interpreted within the context of the study design. Our study involved a small number of patients, with variation in different histologic subtypes. However, it was useful to establish that corrected SUVmax is a potential predictive imaging biomarker, regardless of histologic type, among head-and-neck non-SCCs. Because this was an observational study with prospective image analysis and retrospective chart review, we could not control the incubation period of FDG uptake or scanning protocol variations over the study period because these studies were performed as a part of routine clinical studies. However, use of the corrected SUVmax would have largely eliminated the errors related to variable dose, incubation period, scanner protocol variation, and interbiologic variability. The overall mortality rather than disease-specific mortality was established. Overall mortality is a better outcome parameter than disease-specific mortality. However, we did not risk adjust for disease stage or therapy variations, and DFS may be a better outcome in this study context.
In conclusion, FDG uptake measured by corrected SUVmax is a potential predictive imaging biomarker for DFS and overall survival in patients with head-and-neck non-SCC.
References
1.
Carvalho AL, Nishimoto IN, Califano JA, Kowalski LP. Trends in incidence and prognosis for head and neck cancer in the United States: a site-specific analysis of the SEER database. Int J Cancer 2005; 114:806–816
2.
Daley T, Darling M. Nonsquamous cell malignant tumours of the oral cavity: an overview. J Can Dent Assoc 2003; 69:577–582
3.
Lin HW, Bhattacharyya N. Staging and survival analysis for nonsquamous cell carcinomas of the larynx. Laryngoscope 2008; 118:1003–1013
4.
Benchaou M, Lehmann W, Slosman DO, et al. The role of FDG-PET in the preoperative assessment of N-staging in head and neck cancer. Acta Otolaryngol 1996; 116:332–335
5.
Allal AS, Slosman DO, Kebdani T, Allaoua M, Lehmann W, Dulguerov P. Prediction of outcome in head-and-neck cancer patients using the standardized uptake value of 2-[F-18]fluoro-2-deoxy-D-glucose. Int J Radiat Oncol Biol Phys 2004; 59:1295–1300
6.
Rege SD, Chaiken L, Hoh CK, et al. Change induced by radiation therapy in FDG uptake in normal and malignant structures of the head and neck: quantitation with PET. Radiology 1993; 189:807–812
7.
Allal AS, Dulguerov P, Allaoua M, et al. Standardized uptake value of 2-[F-18] fluoro-2-deoxy-D-glucose in predicting outcome in head and neck carcinomas treated by radiotherapy with or without chemotherapy. J Clin Oncol 2002; 20:1398–1404
8.
Pelizzo MR, Boschin IM, Toniato A, et al. Diagnosis, treatment, prognostic factors and long-term outcome in papillary thyroid carcinoma. Minerva Endocrinol 2008; 33:359–379
9.
Greene FL, Page DL, Fleming ID, et al., eds. AJCC cancer staging manual, 6th ed. Chicago, IL: American Joint Committee on Cancer, 2002
10.
RootsWeb. www.rootsweb.ancestry.com. Accessed August 15, 2009
11.
Vansteenkiste JF, Stroobants SG, Dupont PJ, et al. Prognostic importance of the standardized uptake value on (18)F-fluoro-2-deoxy-glucose-positron emission tomography scan in non-small-cell lung cancer: an analysis of 125 cases. Leuven Lung Cancer Group. J Clin Oncol 1999; 17:3201–3206
12.
Antoch G, Saoudi N, Kuehl H, et al. Accuracy of whole-body dual-modality fluorine-18-2-fluoro-2-deoxy-D-glucose positron emission tomography and computed tomography (FDG-PET/CT) for tumor staging in solid tumors: comparison with CT and PET. J Clin Oncol 2004; 22:4357–4368
13.
Dresel S, Grammerstorff J, Schwenzer K, et al. [F-18]FDG imaging of head and neck tumours: comparison of hybrid PET and morphological methods. Eur J Nucl Med Mol Imaging 2003; 30:995–1003
14.
Slevin NJ, Collins CD, Hastings DL, et al. The diagnostic value of positron emission tomography (PET) with radiolabelled fluorodeoxyglucose (18F-FDG) in head and neck cancer. J Laryngol Otol 1999; 113:548–554
15.
Nowak B, Di Martino E, Janicke S, et al. Diagnostic evaluation of malignant head and neck cancer by F-18-FDG PET compared to CT/MRI. Nuklearmedizin 1999; 38:312–318
16.
Agarwal V, Branstetter BF, Johnson JT. Indications for PET/CT in the head and neck. Otolaryngol Clin North Am 2008; 41:23–49
17.
Paul SAM, Stoeckli SJ, von Schulthess GK, Goerres GW. FDG PET and PET/CT for the detection of the primary tumour in patients with cervical non-squamous cell carcinoma metastasis of an unknown primary. Eur Arch Otorhinolaryngol 2007; 264:189–195
18.
Jabour BA, Choi Y, Hoh CK, et al. Extracranial head and neck: PET imaging with 2-[F-18]fluoro-2-deoxy-D-glucose and MR imaging correlation. Radiology 1993; 186:27–35
19.
Oshida M, Uno K, Suzuki M, et al. Predicting the prognoses of breast carcinoma patients with positron emission tomography using 2-deoxy-2-fluoro [F-18]-D-glucose. Cancer 1998; 82:2227–2234
20.
Vranjesevic D, Filmont JE, Meta J, et al. Whole-body (18)F-FDG PET and conventional imaging for predicting outcome in previously treated breast cancer patients. J Nucl Med 2002; 43:325–329
21.
Snow GB, Annyas AA, van Slooten EA, Bartelink H, Hart AA. Prognostic factors of neck node metastasis. Clin Otolaryngol Allied Sci 1982; 7:185–192
22.
Kitagawa Y, Sadato N, Azuma H, et al. FDG PET to evaluate combined intra-arterial chemotherapy and radiotherapy of head and neck neoplasms. J Nucl Med 1999; 40:1132–1137
23.
Davison JM, Ozonoff A, Imsande HM, Grillone GA, Subramaniam RM. Squamous cell carcinoma of the palatine tonsils: FDG standardized uptake value ratio as a biomarker to differentiate tonsillar carcinoma from physiologic uptake. Radiology 2010; 255:578–585
24.
Roh JL, Ryu CH, Choi SH, et al. Clinical utility of F-18-FDG PET for patients with salivary gland malignancies. J Nucl Med 2007; 48:240–246
25.
Suh C, Kang YK, Roh JL, et al. Prognostic value of tumor F-18-FDG uptake in patients with untreated extranodal natural killer/T-cell lymphomas of the head and neck. J Nucl Med 2008; 49:1783–1789
26.
Cook GJ, Maisey MN, Fogelman I. Normal variants, artefacts and interpretative pitfalls in PET imaging with 18-fluoro-2-deoxyglucose and carbon-11 methionine. Eur J Nucl Med 1999; 26:1363–1378
27.
Zasadny KR, Wahl RL. Standardized uptake values of normal tissues at PET with 2-[fluorine-18]-fluoro-2-deoxy-D-glucose: variations with body weight and a method for correction. Radiology 1993; 189:847–850
28.
Lee JR, Madsen MT, Bushnel D, Menda Y. A threshold method to improve standardized uptake value reproducibility. Nucl Med Commun 2000; 21:685–690
29.
Ramos CD, Erdi YE, Gonen M, et al. FDG-PET standardized uptake values in normal anatomical structures using iterative reconstruction segmented attenuation correction and filtered back-projection. Eur J Nucl Med 2001; 28:155–164
30.
Boellaard R. Standards for PET image acquisition and quantitative data analysis. J Nucl Med 2009; 50(suppl 1):11S–20S
31.
Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to PERCIST: evolving considerations for PET response criteria in solid tumors. J Nucl Med 2009; 50(suppl 1):122S–150S
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: April 27, 2010
Accepted: March 22, 2011
First published: November 20, 2012
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.