AJR ARRS: Your Link to CME
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lell, M. M.
Right arrow Articles by Greess, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lell, M. M.
Right arrow Articles by Greess, H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?
DOI:10.2214/AJR.07.3081
AJR 2008; 190:785-789
© American Roentgen Ray Society

Original Research

Thin-Slice MDCT of the Neck: Impact on Cancer Staging

Michael M. Lell1,2, Christian Gmelin2, Christoph Panknin1,3, Karin T. Eckel4, Matthias Schmid4, Werner A. Bautz2 and Holger Greess2

1 Department of Radiological Sciences, David Geffen School of Medicine at UCLA, Peter V. Ueberroth Bldg., Ste. 3371, 10945 LeConte Ave., Los Angeles, CA 90095.
2 Department of Radiology, University Erlangen-Nuremberg, Erlangen, Germany.
3 Siemens Medical Solutions, Forchheim, Germany.
4 Department of Medical Informatics, Biometry, and Epidemiology, University Erlangen-Nuremberg, Erlangen, Germany.

Received August 30, 2007; accepted after revision September 25, 2007.

 
Address correspondence to M. M. Lell (michael.lell{at}uk-erlangen.de).


Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
OBJECTIVE. The objective of this study was to compare thin-slice multiplanar evaluation and conventional 3-mm axial evaluation of head and neck MDCT in tumor staging.

MATERIALS AND METHODS. Ninety-six patients with histologically proven squamous cell carcinoma were evaluated independently, once using 3-mm axial images and once using 1-mm interactive multiplanar reformation (MPR) images. Tumor stage was assessed with both methods; histology served as the reference. Thirty-seven patients with hypopharyngeal and laryngeal tumors had en bloc resection, allowing direct comparison of tumor infiltration into designated anatomic structures. Two examiners independently assessed the data sets. Interobserver agreement was tested with a modified kappa test. The Wilcoxon signed rank test with continuity correction was applied to test the null hypothesis, which postulates the equality of both methods. The chi-square test was applied to compare the number of correctly classified tumors for the two methods and readers.

RESULTS. Interobserver agreement was high ({kappa} = 0.88–0.91). Both methods allowed accurate tumor staging, and no significant differences between the two methods were found (reader A, p = 0.61; reader B, p = 1). With MPR assessment, more anatomic structures were rated positive for tumor infiltration, but diagnostic accuracy did not differ significantly in the subgroup of patients with histologic correlation from en bloc resection.

CONCLUSION. Conventional 3-mm axial evaluation of head and neck MDCT proved to be sufficient in tumor staging.

Keywords: CT • larynx • MDCT • neoplasm • pharynx


Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Squamous cell carcinomas (SCCs) are the most frequent of all head and neck malignancies [1]. According to U.S. estimates of new cancers in 2006, about 31,000 occurred in the oral cavity and pharynx, and 9,500 in the larynx [2]. Tumor volume and lymph node infiltration are important factors that influence the therapeutic approach and the prognosis of the patient with SCC [3–7]. Exact tumor staging is necessary for treatment planning, leading to reduced postoperative morbidity and tumor recurrence-associated mortality. Direct laryngoscopy is most accurate in evaluating the mucosal surface of the aerodigestive tract [8, 9]. In fact, such examinations frequently identify superficial carcinomas that cannot be detected on MRI or CT. Submucosal extension cannot be sufficiently assessed by endoscopy and physical examination but can be evaluated with MRI or CT [10, 11]. Clinical examination alone frequently underestimates the extent of disease.

Axial images with a slice thickness of 3–5 mm are advocated in various imaging protocols, although MDCT technology is capable of acquiring high-resolution (submillimeter) studies of the whole neck in less than 20 seconds. When overlapping images are reconstructed from raw data with a nominal slice thickness of 0.5–1.25 mm, multiplanar reformation (MPR) images of the tumor can be viewed interactively in arbitrarily chosen imaging planes [12]. The price for this is a significant increase in reconstruction and data transfer time and storage demands. In addition, hundreds of images to review can diminish the productivity of a radiologist.

The goal of our study was to assess and compare two reconstruction and image reading strategies for MDCT for staging primary SCC of the neck: conventional 3-mm axial images versus interactive multiplanar evaluation of high-resolution thin-slice images.


Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Patients
Ninety-six patients (87 men, 9 women; age range, 34–84 years; mean age, 56 years; median age, 54 years) with SCC of the head and neck were enrolled in this study over a 20-month period. Written informed consent from all patients and institutional review board approval were obtained. The primary tumor locations were nasopharynx (n = 8), oral cavity (n = 23), oropharynx (n = 21), hypopharynx (n = 18), and larynx (n = 26). Tumor stage, on the basis of surgical and histopathologic reports, is shown in Table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 1: Tumor Localization and Final Tumor Stage

 

CT
Patients were instructed to take shallow breaths and refrain from swallowing during scanning. MDCT was performed on a 16-MDCT scanner (Sensation 16, Siemens Medical Solutions) with tube voltage, 120 kV; effective tube current, 150 mAs; collimation, 0.75 mm; table feed, 12 mm/rotation; and rotation time, 0.5 second. The effective radiation dose for a typical scanning range of 250 mm was 3.6 mSv for men and 4.1 mSv for women. A nonionic contrast agent (120 mL of iomeprol [Iomeron 300, Bracco]) was injected at a flow rate of 3 mL/s and scanning was started 80 seconds after the start of contrast injection. The scanning range was individually adapted and included the skull base to the upper mediastinum; in patients with nasopharyngeal cancer, the range was extended to the level above the frontal sinuses. Two image data sets were reconstructed using a standard soft-tissue (B 40) convolution kernel: one data set with a slice thickness of 3 mm (3-mm reconstruction increment), the other with a slice thickness of 1 mm (0.7-mm reconstruction increment). For the assessment of bone and cartilage, additional data sets with identical parameters were reconstructed using a sharp (bone) convolution kernel (B 70). Reconstruction time was measured for a scanning range of 250 mm using a phantom scan.

Data Analysis
CT data were anonymized and independently evaluated by two radiologists with more than 5 years of experience in head and neck radiology who were blinded to the clinical data and to each other's results. Local tumor infiltration of specific structures (Table 2) and tumor stage [13, 14] were evaluated. To reduce bias introduced by a learning curve or recall of corresponding examinations, data were evaluated in random order and with a time interval of at least 1 month between each pair of corresponding examinations. A workstation (Leonardo, Siemens Medical Solutions) was used to read the images. Signs of malignant tumors on CT were evaluated as previously described in the literature [15–19]. Vascular infiltration was diagnosed if the vessel was surrounded by tumor more than 180° or an intraluminal mass was seen. Thirty-two anatomic structures were evaluated for tumor infiltration.


View this table:
[in this window]
[in a new window]

  		TABLE 2: Positive Diagnosis of Tumor Infiltration

 

Histologic and surgical reports were reviewed for the final T stage to validate the results from CT. Tumor resection was performed in 82 patients; in 14 patients only data from panendoscopy and multiple biopsies were available. This was the case in patients with T3 and T4 nasopharyngeal carcinoma (n = 3) and other T4 tumors in which no primary surgery was performed.

Statistical Analysis
A modified extension of Cohen's kappa test by Janson and Olsson [20] was used to measure the agreement between the two readers for both methods (MPR, 3-mm axial) for assessment of the different anatomic structures. Because histopathologic results were not available for all patients and all anatomic sites, the results of both methods could not be tested against a gold standard. The Wilcoxon signed rank test with continuity correction was applied to test the null hypothesis, which postulates the equality of both methods. Therefore, a score over all anatomic structures was calculated for each patient, resulting in 96 score values ranging from 0 to 64 (two readers, 32 anatomic structures). The scores were defined as follows: the value 1 was assigned if there was evidence of tumor infiltration and the value 0 was assigned if there was not. Perfect agreement would result in a difference between the corresponding scores of zero.

In 37 patients with carcinomas of the larynx and hypopharynx, pathohistologic reports were available with explicit details on tumor infiltration of the following 10 structures: thyroid cartilage; arytenoid cartilage; cricoid cartilage; aryepiglottic fold; paralaryngeal space; preepiglottic space; anterior commissure; and the supraglottic, glottic, and subglottic mucosa. For this subgroup, sensitivity, specificity, and accuracy for detecting tumor infiltration were computed for each observer and method. To compare the methods, an extension of the McNemar test [21] was applied. The local significance level of 0.05 was adjusted using the Holm method [22] for multiple testing.

A chi-square test was used to compare the number of tumors classified according to the International Union Against Cancer (UICC) TNM system correctly for the two methods and each reader. Fisher's exact test was applied to investigate the relationship between misclassification and the method applied.


Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
A typical scanning range of 250 mm results in 84 images with 3-mm slice thickness and 358 thin-slice images with our reconstruction protocol. Reconstruction time was 11 seconds for the 3-mm set and 1 minute 55 seconds for the thin-slice data set (scanner software version VB20, Siemens Medical Solutions).

A high level of agreement was found for both readers (MPR, {kappa} = 0.88; 3-mm, {kappa} = 0.91). Table 2 shows the numbers of cases in which tumor spread to the specific structures was diagnosed by the readers with respect to the image evaluation technique. The most frequent findings were infiltration of the parapharyngeal space, the supraglottic region, and the tongue base.

The differences of the scores for both methods showed that thin-slice image evaluation resulted in a higher score than image evaluation with 3-mm axial images (p < 0.01, Wilcoxon signed rank test with continuity correction).

Sensitivity, specificity, and accuracy values for the 10 structures of the larynx in the subgroup of 37 patients with resected laryngeal and hypopharyngeal carcinomas were high (sensitivity, 58–100%; specificity, 75–100%; accuracy, 87–100%). The McNemar test did not indicate significant differences between the readers or methods.

Agreement between CT-based T stage and the final T stage was high. No significant differences between the two methods for either reader (reader A, p = 0.61; reader B, p = 1; chi-square test) were detected. In Figure 1A, 1B, 1C, 1D, the numbers of correct and incorrect CT classifications are plotted against the final tumor stage. The plot shows no apparent differences between the methods; Fisher's exact test (Table 3) reflects this finding (reader A, p = 0.87; reader B, p = 0.45).


Figure 1
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1A —Correctly classified tumors (black) and misclassified tumors (white) according to histopathologic results. Graphs show results of reader A using multiplanar reformation (MPR) (A), reader B using MPR (B), reader A using 3-mm axial MDCT images (C), and reader B using 3-mm axial MDCT images (D).

 

Figure 2
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1B —Correctly classified tumors (black) and misclassified tumors (white) according to histopathologic results. Graphs show results of reader A using multiplanar reformation (MPR) (A), reader B using MPR (B), reader A using 3-mm axial MDCT images (C), and reader B using 3-mm axial MDCT images (D).

 

Figure 3
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1C —Correctly classified tumors (black) and misclassified tumors (white) according to histopathologic results. Graphs show results of reader A using multiplanar reformation (MPR) (A), reader B using MPR (B), reader A using 3-mm axial MDCT images (C), and reader B using 3-mm axial MDCT images (D).

 

Figure 4
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1D —Correctly classified tumors (black) and misclassified tumors (white) according to histopathologic results. Graphs show results of reader A using multiplanar reformation (MPR) (A), reader B using MPR (B), reader A using 3-mm axial MDCT images (C), and reader B using 3-mm axial MDCT images (D).

 

View this table:
[in this window]
[in a new window]

  		TABLE 3: Fisher's Exact Test for Number of Misclassified Stages Between Multiplanar Reconstruction (MPR) and Axial Method

 


Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Head and neck surgeons often rely on imaging to determine if a neoplasm is resectable. In the evaluation of a patient with head and neck cancer, a conflict exists between what it would take to completely resect an advanced cancer versus the impact such a resection would have on the patient's quality of life and self-image. Critical structures, which include arterial encasement, prevertebral fascia involvement, mediastinal infiltration, tracheal and esophageal extension, laryngeal cartilage penetration, preepiglottic fat involvement, dural spread, bone infiltration, perineural spread, orbital involvement, and brachial plexus invasion, have to be assessed to model an appropriate therapy concept [23]. Isotropic data sets can be acquired with modern MDCT scanners in a couple of seconds and MPR images can be generated in excellent quality at the price of increased demands on the PACS and reading radiologist. The aim of our study was to compare a conventional imaging approach with contiguous 3-mm images and a more-advanced high-resolution multiplanar image reading approach to find out if the increased expenses are justified.


Figure 5
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2A —54-year-old woman with squamous cell carcinoma of soft palate. Because of beam-hardening artifacts and partial volume effects, this lesion (arrow) was detectable only on single 3-mm axial MDCT image.

 


Figure 6
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2B —54-year-old woman with squamous cell carcinoma of soft palate. Multiplanar reformation image in coronal plane improves tumor (arrow) delineation.

 
Both of our readers showed a high level of agreement for each image evaluation method. Previous reports showed that MPR can facilitate the assessment of the craniocaudal extension of tumors and the assessment of certain structures such as the paraglottic space [12, 24]. We hypothesized that special tumor locations (Fig. 2A, 2B) as well as the previously mentioned critical structures demand additional imaging planes, especially if infiltrated structures are oriented parallel to the scanning, such as the orbital floor or skull base as seen frequently with nasopharyngeal carcinoma and tumors of the paranasal sinuses. In agreement with this hypothesis, more anatomic structures were rated to be invaded by tumor with thin-slice MPR evaluation compared with evaluation with 3-mm axial slices.

On the other hand, both readers were able to extract all essential information for skull base infiltration from the axial 3-mm slices. Lau et al. [25] also reported that out of a typical MR protocol, axial T1-weighted contrast-enhanced fat-saturated images were the most accurate to stage patients with nasopharyngeal tumors. The subgroup analysis of the 37 patients for whom histologic correlation from en bloc resection was available also did not reveal significant differences between the two analysis methods in terms of diagnostic accuracy. T staging, according to UICC in all 96 patients, did not differ significantly between the two methods. These findings are in agreement with those reported by Keberle et al. [26], who found no statistically significant difference in tumor staging comparing 3-mm axial images and MPR from a 4 x 1 mm helical data set but noted that in 19% of the patients studied, modifications in the surgical procedure resulted from additional information of the MPR.

A limitation of our study is that we did not have direct pathohistologic correlation for all anatomic structures evaluated in this study, so only the results of a subgroup could be validated. Eighty-two patients were treated surgically, and the final T stage was histologically verified; 14 patients were treated with radiochemotherapy and the final T stage was based on clinical findings and results from multiple biopsies. Although MPR images may show pathology more clearly, our results indicate that 3-mm axial image evaluation is adequate for staging head and neck tumors. As a compromise, we recommend data acquisition with thin-slice detector collimation and image reconstruction in 3-mm slices. This conserves the opportunity to perform further reconstructions and MPR assessment in selected equivocal cases. Further studies are required to confirm this strategy.

In conclusion, because thin-slice image reconstruction places a huge burden on the reading radiologist as well as network and PACS systems, the justification for performing it should be based on objective evidence of benefit. Our results indicate that thin-slice multiplanar image evaluation is not superior to axial 3-mm image evaluation in staging tumors according to the UICC classification.


References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Vokes EE, Weichselbaum RR, Lippman SM, Hong WK. Head and neck cancer. N Engl J Med 1993;328 : 184–194[Free Full Text]
  2. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. C A Cancer J Clin 2006;56 : 106–130
  3. Chiesa F, Mauri S, Tradati N, et al. Surfing prognostic factors in head and neck cancer at the millennium. Oral Oncol1999; 35:590 –596[CrossRef][Medline]
  4. Plataniotis GA, Theofanopoulou ME, Kalogera-Fountzila A, et al. Prognostic impact of tumor volumetry in patients with locally advanced head-and-neck carcinoma (non-nasopharyngeal) treated by radiotherapy alone or combined radiochemotherapy in a randomized trial. Int J Radiat Oncol Biol Phys 2004; 59:1018 –1026[CrossRef][Medline]
  5. Vokes E. Current treatments and promising investigations in a multidisciplinary setting. Ann Oncol2005; 16[suppl 6]:vi25 –vi30[Abstract/Free Full Text]
  6. Liu M, Weynand B, Delos M, Marbaix E. Prognostic factors in squamous cell carcinomas of the head and neck. Acta Otorhinolaryngol Belg 1999;53 : 155–160[Medline]
  7. Magnano M, Bongioannini G, Lerda W, et al. Lymph node metastasis in head and neck squamous cells carcinoma: multivariate analysis of prognostic variables. J Exp Clin Cancer Res 1999;18 : 79–83[Medline]
  8. Madison MT, Remley KB, Latchaw RE, Mitchell SL. Radiologic diagnosis and staging of head and neck squamous cell carcinoma. Radiol Clin North Am 1994;32 : 163–181[Medline]
  9. van den Brekel MW, Castelijns JA, Snow GB. The role of modern imaging studies in staging and therapy of head and neck neoplasms. Semin Oncol 1994;21 : 340–348[Medline]
  10. Mafee MF, Schild JA, Valvassori GE, Capek V. Computed tomography of the larynx: correlation with anatomic and pathologic studies in cases of laryngeal carcinoma. Radiology 1983;147 : 123–128[Abstract/Free Full Text]
  11. Hsu WC, Loevner LA, Karpati R, et al. Accuracy of magnetic resonance imaging in predicting absence of fixation of head and neck cancer to the prevertebral space. Head Neck 2005;27 : 95–100[CrossRef][Medline]
  12. Lell MM, Greess H, Hothorn T, Janka R, Bautz WA, Baum U. Multiplanar functional imaging of the larynx and hypopharynx with multislice spiral CT. Eur Radiol 2004;14 :2198 –2205[CrossRef][Medline]
  13. Sobin LH, Wittekind CL. TNM classification of malignant tumors, 6th ed. New York, NY: Wiley,2002
  14. Collaborative Staging Task Force of the American Joint Committee on Cancer. Collaborative staging manual and coding instructions, version 01.03.00. Jointly published by American Joint Committee on Cancer (Chicago, IL) and U.S. Department of Health and Human Services (Bethesda, MD). NIH Publication Number 04-5496, 2004
  15. Becker M. Neoplastic invasion of laryngeal cartilage: radiologic diagnosis and therapeutic implications. Eur J Radiol2000; 33:216 –229[CrossRef][Medline]
  16. Harnsberger HR, Wiggins RH, Hudgins PA, et al. Diagnostic imaging: head and neck. Salt Lake City, UT: AMIRSYS, 2004
  17. Lenz M, Greess H, Baum U, Dobritz M, Kersting-Sommerhoff B. Oropharynx, oral cavity, floor of the mouth: CT and MRI. Eur J Radiol 2000; 33:203 –215[CrossRef][Medline]
  18. Curtin HD, Williams R, Johnson J. CT of perineural tumor extension: pterygopalatine fossa. AJR 1985;144 : 163–169[Abstract/Free Full Text]
  19. Mukherji SK, Pillsbury HR, Castillo M. Imaging squamous cell carcinomas of the upper aerodigestive tract: what clinicians need to know. Radiology 1997;205 : 629–646[Free Full Text]
  20. Janson H, Olsson U. A measure of agreement for interval or nominal multivariate observations. Educ Psychol Meas2001; 61:277 –289[Abstract/Free Full Text]
  21. Lachenbruch PA, Lynch CJ. Assessing screening tests: extensions of McNemar's test. Stat Med 1998;17 :2207 –2217[CrossRef][Medline]
  22. Aickin M, Gensler H. Adjusting for multiple testing when reporting research results: the Bonferroni vs Holm methods. Am J Public Health 1996; 86:726 –728[Abstract/Free Full Text]
  23. Yousem DM, Gad K, Tufano RP. Resectability issues with head and neck cancer. Am J Neuroradiol 2006;27 :2024 –2036[Abstract/Free Full Text]
  24. Toyoda K, Kawakami G, Kanehira C, et al. Enhanced four-detector row computed tomography imaging of laryngeal and hypopharyngeal cancers. J Comput Assist Tomogr 2002;26 : 912–921[CrossRef][Medline]
  25. Lau KY, Kan WK, Sze WM, et al. Magnetic resonance for T-staging of nasopharyngeal carcinoma: the most informative pair of sequences. Jpn J Clin Oncol 2004;34 : 171–175[Abstract/Free Full Text]
  26. Keberle M, Sandstede J, Hoppe F, Fischer M, Hahn D. Diagnostic impact of multiplanar reformations in multi-slice CT of laryngeal and hypopharyngeal carcinomas [in German]. Rofo2003; 175:1079 –1085[Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lell, M. M.
Right arrow Articles by Greess, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lell, M. M.
Right arrow Articles by Greess, H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS