| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From Dermatology (SMS) and Nuclear Medicine (GMS) Services, Veterans Affairs Palo Alto Health Care System, Palo Alto, California; and Departments of Dermatology (SMS, LAC), Radiology (GMS), and Surgery (DLJ), Stanford University Medical Center, Stanford, California.
Correspondence: Address correspondence and reprint requests to: Susan M. Swetter, MD, Department of Dermatology, Stanford University Medical Center, 900 Blake Wilbur Dr., W0069, Stanford, CA 94305; Fax: 650-723-7796; E-mail: sswetter{at}stanford.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: A retrospective cohort review was performed of 104 patients with primary or recurrent melanoma who underwent PET to determine sensitivity/specificity for metastatic detection compared with body computed tomography (CT). One hundred fifty-seven PET and 70 CT scans were analyzed, with a median patient follow-up of 24 months. Metastases were confirmed with positive histology (87.5%) or documented disease progression (12.5%). Fifty-three patients prospectively underwent consecutive studies within a mean 3-week interval for direct comparative analysis.
Results: PET demonstrated 84% sensitivity (95% confidence interval [CI], .78 to .89) and 97% specificity (95% CI, .91 to .99), whereas CT showed 58% sensitivity (95% CI, .49 to .66) and 70% specificity (95% CI, .51 to .84). Exclusion of areas not evaluated on CT (head, neck/supraclavicular, extremities) increased CT sensitivity to 69% (95% CI, .59 to .77). Sixty-six consecutive PET and CT scans were performed with 81% and 57% of metastases detected, respectively.
Conclusions: PET is more sensitive and specific than CT for detection of melanoma metastasis and should be considered the primary staging study for recurrent disease. PET shows greater ability to detect soft tissue, small-bowel, and lymph node metastasis that do not meet criteria designated as abnormal by CT. PET is superior to CT even when sites not routinely evaluated by CT are excluded from comparative analysis.
Key Words: Melanoma • FDG-PET • CT • Staging • Comparative studies
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Whole-body PET does not provide the same degree of anatomical detail as computed tomography (CT) or magnetic resonance imaging (MRI). However, its advantages include routine evaluation of the entire body, including areas not evaluated on body CT (head, neck/supraclavicular areas, and extremities); increased ability to detect malignant tumors because of greater glucose metabolism than in normal tissues; and absence of CT contrast-related problems.1,14,15 Disadvantages of PET include limited ability to detect brain metastasis (because of inherently high metabolic activity and central nervous system FDG uptake),2,3 lesion diameter 
5 to 7 mm necessary for malignant tumor detection,3,6 and, most important, reduced availability compared with CT scan. However, the increased use of multipurpose nuclear medicine cameras modified for PET as well as the emergence of mobile PET scanners have greatly increased the availability of PET nationwide.
We compared the metastatic detection ability of PET with that of CT through a retrospective analysis of 108 patients with primary or recurrent melanoma who underwent staging PET within our institution from 1995 to 2000. Four patients were excluded because of PET detection of concurrent nonmelanoma internal malignancies (pulmonary adenocarcinoma, oropharyngeal squamous cell carcinoma, and preexisting chronic B-cell lymphoma). Of the remaining 104 patients, 157 PET scans, 70 body CT scans, and 48 brain MRIs were performed. Furthermore, we directly compared the diagnostic efficacy of PET with that of CT in 66 consecutive studies performed within a 3-week mean interval in a subgroup of 53 patients.
|
PATIENTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the remaining 104 melanoma patients, 157 PET scans were performed for 3 clinical scenarios: staging at the time of diagnosis of primary cutaneous melanoma without clinical evidence of metastasis (18%); staging for clinical suspicion or documentation of disease progression, including presentation with nodal or disseminated metastasis with an unknown primary tumor (65%); or routine melanoma surveillance in patients with high-risk tumors or nodal or distant disease (17%). Initial PET scans in the 104 patients were performed with the following AJCC melanoma stage groupings at presentation: 5% stage I (
1.5-mm depth), 37% stage II (>1.5-mm depth), 29% stage III (regional nodal metastasis), and 29% stage IV (disseminated disease). Seventy-seven patients were men and 27 were women, with a mean age of 54 years (range, 19 to 87 years; Table 1).
|
1 mm, 1.01–2 mm, 2.01–4 mm, and >4 mm), 16 and number of metastases identified on PET versus CT within these subgroups are listed in Table 2.
|
PET was performed after intravenous administration of FDG (average dose, 15 mCi). An emission scan was acquired from the top of the head to the feet in 10 to 12 bed positions at 4 minutes per position. Attenuation-corrected regional emission scans were performed on suspicious regions. In the absence of suspicious lesions on initial PET scan, attenuation-corrected regional emission scans were performed on specific areas of interest, such as the primary melanoma region or the site of any previous or newly diagnosed recurrences. Seventy body CTs were performed in 54 of 104 patients and included imaging of the chest, abdomen, and pelvis. MRI comparison to PET was limited to the brain, and 43 of 104 patients in the study cohort underwent brain MRI. The PET brain data were obtained during acquisition of the body scan rather than with use of a PET protocol optimized for brain imaging.
In the noncomparative analysis of 157 PET and 70 CT scans, sensitivity and specificity of metastatic detection for each modality were defined as follows: 
|
| (1) |
|
| (2) |
TP findings were based on histological examination of individual metastases (or adjacent metastases in unresectable patients) in 87.5% of cases and by disease progression in 12.5% of cases and were documented according to anatomical location. False-negative (FN) findings for metastases on CT or PET were subsequently identified by other imaging studies and confirmed histologically in 89% of cases or by disease progression in 11%. By necessity, true-negative totals for PET, CT, and MRI were based on the number of scans performed that correctly showed the absence of metastasis.
Direct comparative analysis of consecutive PET and CT scans was prospectively undertaken in a subgroup of 53 patients in the study cohort. In most cases, PET and CT were ordered simultaneously to avoid referral bias, and scans were initially reviewed independently. All PET scans were performed at the Palo Alto Veterans Affairs Medical Center and interpreted by the same nuclear medicine physician (G.M.S.). All CT scans were reviewed or performed at Stanford University Medical Center or the Veterans Affairs Palo Alto Health Care System. TP metastatic detection in the consecutive series was confirmed histologically in 88% of cases and via clinical disease progression in 12%. FNs identified in the discordant studies, i.e., metastases detected on the alternative imaging study, were confirmed histologically in 79% of cases and by clinical progression or other positive tests in 21% of cases. Of the eight CT false-positive (FP) findings, three (37.5%) were confirmed by pathology and five (62.5%) by other negative studies, lack of clinical disease or melanoma progression, or both. No FP PET findings were noted in the consecutive series.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Brain MRI was compared with whole-body PET to confirm the inability of nondedicated PET brain views to detect central nervous system metastasis. Forty-eight brain MRIs performed in 43 patients in the study cohort detected 7 metastases in 3 individuals, with no FNs. MRI correctly predicted the absence of brain metastasis in 44 studies, and there were no FP scans. As expected, brain MRI was 100% sensitive and specific for brain metastasis (95% CI, .59 to 1.0 for sensitivity; 95% CI, .92 to 1.0 for specificity; both one sided) compared with 0% PET detection of brain metastasis in whole-body scans that included imaging of the head.
Sixty-six consecutive PET and CT studies were performed to directly compare the efficacy of PET with that of CT (regardless of prior test findings) in 53 of 104 patients. Thirty individuals in this subgroup had 132 identifiable sites of metastasis detected by PET, CT, or both. Thirteen patients in this subgroup underwent an additional PET/CT comparison for further evaluation of prior questionable findings or suspicion of disease recurrence, and one patient had two additional PET/CT comparisons for melanoma surveillance. Histological confirmation was obtained in 88% of detected metastases, and 12% of metastatic lesions were documented by disease progression or death as a result of melanoma. The mean interval between consecutive PET and CT scans was 22 days (minimum, 0 days; maximum 78 days) for all patients and was 15 days (minimum, 0 days; maximum, 54 days) for those patients with radiologically identified metastases. CT was performed before PET in 40 (61%) of the comparative studies, on the same day as PET in 8 (12%) of the comparisons, and after PET scan in 18 (27%) comparison scans.
Overall, PET detected 81% (107 of 132) of metastases, whereas CT detected 57% (75 of 132). When brain, neck/supraclavicular, and nontruncal subcutaneous sites were excluded (because they were not visualized on body CT), CT detection increased to 68% (75 of 110) and PET detection to 84% (92 of 110). PET detection increased to 86% (107 of 125) with exclusion of missed brain metastasis alone.
Further analysis of the 66 direct PET/CT comparisons was performed according to anatomical site of metastasis. PET detected 79% (15 of 19) of subcutaneous metastases, whereas CT detected only 47% (9 of 19). Specifically, PET detected 73% of truncal (11 of 15) and 100% of nontruncal subcutaneous metastases (4 of 4). CT detected 60% of truncal (9 of 15) and none of the nontruncal soft tissue metastasis, because it does not routinely evaluate the extremities or head. PET detected 100% (11 of 11) of neck and supraclavicular metastases; CT detected none of these, because again, regions above the clavicle are not routinely imaged on body CT. CT also failed to identify two infraclavicular metastases that were identified on PET scan. PET detected 88% (15 of 17) of metastases to the peripheral lymph nodes (groin, axillae, extremities), whereas CT detected 65% (11 of 17). PET detected 100% of hilar/mediastinal metastases (15 of 15), whereas CT detected 67% (10 of 15). Both PET and CT detected 84% (19 of 25) of parenchymal lung metastases. PET detected 86% (25 of 29) of intra-abdominal metastases (including liver, pancreas, spleen, bowel, kidney, adrenal, and unspecified intra-abdominal metastases), whereas CT detected 72% (21 of 20) overall. Finally, PET and CT showed similar ability to detect skeletal metastasis (in involved truncal sites), with 71% detection for both modalities (five of seven bony metastases determined with two FNs on PET and two FNs on CT, each detected by the comparative imaging modality). Subsite analysis revealed that PET was clearly superior to CT for detection of melanoma metastases in the subcutis, peripheral lymph nodes, mediastinum/hilum, and intra-abdominal regions. PET and CT were comparable for detection of parenchymal lung metastasis (Table 4).
|
Forty-five brain MRI and PET scans were directly compared in 38 patients in the study cohort, with a mean interval of 16 days between studies for all patients and 8 days for those with brain metastasis. Six metastases were detected by MRI in two patients and none by PET scan. In the third patient with brain metastasis, PET scan was performed from the base of skull downward and so was not included in the previous comparison. However, despite PET scan’s poor ability to detect brain metastasis, it did correctly identify one subcutaneous metastasis in the facial region, suggesting an additional benefit of whole-body PET scan.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Multiple studies have demonstrated the efficacy of PET for detection of melanoma metastasis over the past decade.1–10 Sensitivity for detecting individual metastatic lesions or abnormal regions has ranged from 87% to 97% in four studies,2,5,9,10 whereas sensitivity for identifying individual patients with metastatic disease ranged from 87% to 100% in three studies.5,6,8 FN findings were usually due to metastasis smaller than 10 mm. The specificity of PET for areas without metastatic disease has ranged from 44% to 94% and from 78% to 96% for patients without metastatic disease in the previous studies. FPs were most often due to FDG uptake in surgical wounds, inflammatory sites, and benign tumors. FPs were reported to decrease when PET scans were interpreted with knowledge of the clinical history.2,9
Two studies compared PET with CT for metastatic staging in melanoma patients and found PET to be superior to CT. Specifically, Holder et al.6 conducted a prospective study of 103 PET scans performed on 76 patients with AJCC stage II to IV melanoma and showed increased PET sensitivity compared with CT (94% vs. 55%, respectively) and similar specificity (83% vs. 84%, respectively). PET was superior for metastatic detection within the regional and mediastinal lymph nodes, liver, and soft tissue and was comparable to CT in the lung parenchyma. Rinne et al.5 prospectively studied 100 patients with melanoma, comparing PET with conventional diagnostic tests, including chest x-ray, abdominal ultrasound, ultrasound of lymph nodes, chest/abdomen CT, brain MRI, and bone scintigraphy. PET had a sensitivity of 92% and a specificity of 94%, whereas conventional diagnostic tests had a sensitivity of 58% and a specificity of 45%. However, CT had a greater sensitivity than PET for detecting small lung metastasis in this series (87% vs. 70%, respectively).
The goal of our study was to compare the efficacy of PET and CT for the detection of individual melanoma metastases. Our data suggest an advantage for PET over CT in terms of greater sensitivity (84% vs. 58%, respectively) and specificity (97% vs. 70%, respectively) in the 104-patient study cohort, in which 87.5% of metastases were confirmed histologically. Exclusion of sites not evaluated on body CT (head, neck, and extremities) increased the sensitivity of CT to 69%. The conclusions of most published PET studies are limited by pretest selection bias and/or posttest evaluation bias, which tends to overestimate the sensitivity and underestimate the specificity of PET. Therefore, we prospectively studied a subgroup of 53 patients in the study cohort who underwent 66 consecutive PET and CT scans within a 3-week mean time interval, so that both scans were obtained independently of the findings of the prior test.
Direct comparison of consecutive scans showed an increased percentage of metastases detected with PET versus CT (81% and 57%, respectively). Exclusion of areas not evaluated on CT increased CT detection to 68% and PET to 84%. Our subsite analysis in the consecutive series showed that PET was clearly superior to CT for detection of melanoma metastasis in subcutaneous sites, peripheral lymph nodes, the mediastinum, and the abdomen. Notably, PET detected 100% of individual mediastinal metastases in nine patients, whereas CT detected only 67%. In most comparative PET/CT series, CT has been superior in detecting small pulmonary metastases.1,5,17 In our study, PET and CT detected the same number of metastases in the lung parenchyma (76%); thus, we consider them equivalent for evaluation of this region in patients with melanoma. It should be emphasized that no FP PET results were noted in the 34 discordant PET-versus-CT comparisons, as opposed to 8 CT scans that showed FP findings compared with PET.
We re-evaluated the data in the consecutive series of PET versus CT and determined that the superiority of PET was not likely due to a temporal phenomenon, because 39% of the CTs in this group were performed on the same day or after the PET scan. Furthermore, for the 27 discordant PET-versus-CT comparisons in which either PET or CT was falsely negative for metastasis, the maximum time interval for TP detection after an initial FN result was only 39 days (mean, 12 days), again arguing against the increased ability of PET to detect metastasis because of the timing of studies obtained. However, the difference between metastatic detection of CT and PET may be explained in part by metastases that do not meet size criteria designated as abnormal by CT. Although CT scanning provides enhanced anatomical detail and is sensitive for morphologically enlarged lymph nodes, it is insensitive for small tumor foci within normal-sized lymph nodes.6
CT is further limited by a smaller field of view compared with PET, and our anatomical subsite analysis demonstrated that whole-body PET is useful for detecting metastasis in regions not routinely evaluated by body CT, including the neck and extremities. It may even be useful in the head region for detection of soft tissue metastasis, as evidenced by one patient in our study with a subcutaneous facial metastasis detected by PET. As expected, routine whole-body PET was not useful for the detection of central nervous system metastasis, and brain MRI and head CT remain the imaging studies of choice in this region.
Study limitations included the retrospective nature of the analysis, referral bias in the original study cohort of 104 patients undergoing PET, and the absence of histological confirmation in 12.5% of detected metastatic lesions, although these metastases were documented by other imaging modalities, clinical disease progression, or patient death as a result of melanoma. In our larger series assessing the sensitivity and specificity of PET and CT in a noncomparative fashion, CT showed both increased FP and FN rates for detection of melanoma metastasis compared with PET. However, the positive predictive value of CT or PET may have been affected by prior imaging results, as well as subsequent referral for either imaging modality. Our prospective subgroup of 53 patients who underwent consecutive PET and CT scans was designed to avoid this bias. Results from this direct comparison confirm the higher accuracy of whole-body PET. Likewise, our clinical experience supports prior data suggesting that the increased sensitivity/specificity of PET for staging melanoma patients may favorably affect surgical and therapeutic management.2,7,9,18,19
It is important to note that the majority of our study population (65%) underwent PET for suspicion of recurrent melanoma and that our conclusions are not necessarily generalizable to the 35% of patients who had PET for staging of primary melanoma or for routine melanoma surveillance, without suspicion of metastatic disease. Furthermore, other studies have not supported the use of PET for staging or surveillance of clinically localized disease or as a replacement for sentinel lymph node biopsy in patients with occult regional metastasis.8,20,21 The majority of our patients in the subgroup who underwent consecutive PET and CT also had clinical findings suggestive of recurrent melanoma, and this remains our primary indication to obtain PET in the management of melanoma patients.
There is no proven role for PET in the initial staging of primary cutaneous melanoma for detection of either regional nodal or distant metastatic disease. PET may be used in certain situations along with sentinel lymph node biopsy to stage patients with high-risk primary tumors, although the utility of PET in detecting visceral metastasis in this setting has not been formally explored. We do not advocate PET in place of sentinel lymph node biopsy for staging the regional nodal basins, although it may help to guide management of patients in whom sentinel node biopsy is not possible or is technically problematic (e.g., failure of radioactive or blue dye to migrate).
Although recent studies have shown that whole-body FDG-PET imaging is highly sensitive for metastatic melanoma, it is not entirely specific for melanoma, as evidenced by the 4 patients in the original 108 with other concurrent internal malignancies, including newly diagnosed pulmonary adenocarcinoma, oropharyngeal squamous cell carcinoma, and previously documented chronic B-cell lymphoma. Other studies have confirmed the usefulness of PET in detecting non–small-cell lung cancer and lymphoma. PET scan for melanoma staging may be difficult to interpret in patients with concomitant malignancies.2,15,22,23 In addition, PET requires a lesion diameter of 5 to 7 mm and is not reliable for detecting metastasis smaller than 5 mm.3,6
Krug et al.17 recently performed a retrospective analysis of FDG-PET for 94 stage III and IV melanoma patients for staging purposes. The use of PET was not found to influence staging (compared with CT or other imaging modalities) and was deemed inferior to CT for diagnosing lung and liver metastases. However, confirmation of diagnosis was possible in only a minority of cases in this study. Our results and the majority of published data differ dramatically and demonstrate that PET adds value to CT for detection of melanoma metastasis in multiple anatomical sites. Overall, PET shows greater sensitivity, specificity, and accuracy than CT for detection of melanoma metastasis. We suggest that PET, where available, be used in conjunction with CT for optimal staging in melanoma patients, particularly when there is suspicion of disease recurrence or documented progression.
In 1999, the Health Care Financing Administration (now the Centers for Medicare and Medicaid Services) approved the use of PET for evaluating "recurrent melanoma prior to surgical intervention" and broadened the reimbursement policy in 2000 for diagnosis, staging, and restaging of both primary and recurrent melanoma.24 This measure, as well as the increased availability of standard and mobile PET scanners nationwide, makes the routine use of PET in melanoma patients more feasible.
In addition, the recent emergence of combined PET/CT scanners may offer great promise for detection and staging of metastatic melanoma. This technology obtains both anatomical and functional images in a single scanning session and is mainly used to precisely locate abnormalities seen on PET.25,26 We have recently begun to use combined PET/CT imaging at our academic center in selected melanoma patients in whom CT did not show any structural abnormality that would correspond to the metabolic abnormality observed on PET. It remains to be seen whether this combined modality improves the diagnostic accuracy of PET or CT when performed independently or whether it affects patient management.
|
Footnotes |
|---|
Whole-body positron emission tomography was superior to body computed tomography (CT) in detected melanoma metastasis in patients with suspicion of recurrent disease. Consecutive scans performed in 53 of 104 patients revealed a metastatic detection rate of 81%, compared with 57% with CT.
Received for publication January 22, 2002. Accepted for publication May 15, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Strobel, R. Dummer, D. B. Husarik, M. Perez Lago, T. F. Hany, and H. C. Steinert High-Risk Melanoma: Accuracy of FDG PET/CT with Added CT Morphologic Information for Detection of Metastases Radiology, August 1, 2007; 244(2): 566 - 574. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Aloia, J. E. Gershenwald, R. H. Andtbacka, M. M. Johnson, C. W. Schacherer, C. S. Ng, J. N. Cormier, J. E. Lee, M. I. Ross, and P. F. Mansfield Utility of Computed Tomography and Magnetic Resonance Imaging Staging Before Completion Lymphadenectomy in Patients With Sentinel Lymph Node-Positive Melanoma J. Clin. Oncol., June 20, 2006; 24(18): 2858 - 2865. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Z. Belhocine, A. M. Scott, E. Even-Sapir, J.-L. Urbain, and R. Essner Role of Nuclear Medicine in the Management of Cutaneous Malignant Melanoma J. Nucl. Med., June 1, 2006; 47(6): 957 - 967. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Wagner Fluorodeoxyglucose Positron Emission Tomography for Melanoma Staging: Refining the Indications Ann. Surg. Oncol., April 1, 2006; 13(4): 444 - 446. [Full Text] [PDF]
|
|
|
|
|
|
M. S. Brady, T. Akhurst, K. Spanknebel, S. Hilton, M. Gonen, A. Patel, and S. Larson Utility of Preoperative [(18)]F Fluorodeoxyglucose-Positron Emission Tomography Scanning in High-Risk Melanoma Patients Ann. Surg. Oncol., April 1, 2006; 13(4): 525 - 532. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Reinhardt, A. Y. Joe, U. Jaeger, A. Huber, A. Matthies, J. Bucerius, R. Roedel, H. Strunk, T. Bieber, H.-J. Biersack, et al. Diagnostic Performance of Whole Body Dual Modality 18F-FDG PET/CT Imaging for N- and M-Staging of Malignant Melanoma: Experience With 250 Consecutive Patients J. Clin. Oncol., March 1, 2006; 24(7): 1178 - 1187. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. E. Finkelstein, J. A. Carrasquillo, J. M. Hoffman, B. Galen, P. Choyke, D. E. White, S. A. Rosenberg, and R. M. Sherry A Prospective Analysis of Positron Emission Tomography and Conventional Imaging for Detection of Stage IV Metastatic Melanoma in Patients Undergoing Metastasectomy Ann. Surg. Oncol., August 1, 2004; 11(8): 731 - 738. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Wagner Editorial: A Role for FDG-PET in the Surgical Management of Stage IV Melanoma Ann. Surg. Oncol., August 1, 2004; 11(8): 721 - 722. [Full Text] [PDF]
|
|
|
|
 |
|
 P T Finger, M Kurli, P Wesley, L Tena, K R Kerr, and A Pavlick Whole body PET/CT imaging for detection of metastatic choroidal melanoma Br. J. Ophthalmol., August 1, 2004; 88(8): 1095 - 1097. [Full Text] [PDF]
|
|
|
|
|
|
O. Alonso, M. Martinez, L. Delgado, A. De Leon, D. De Boni, G. Lago, M. Garces, F. Fontes, J. Espasandin, and J. Priario Staging of Regional Lymph Nodes in Melanoma Patients by Means of 99mTc-MIBI Scintigraphy J. Nucl. Med., October 1, 2003; 44(10): 1561 - 1565. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
