Detection of skeletal muscle metastasis: torso FDG PET-CT versus contrast-enhanced chest or abdomen CT

Abstract

Background

Skeletal muscle metastasis (SMM) in cancer patients has not been sufficiently evaluated regarding prevalence and proper method of detection.

Purpose

To determine the prevalence of SMM and compare the diagnostic competencies for SMM of torso F-18 fluorodeoxyglucose (FDG) positron emission tomography-computed tomography (PET-CT) and contrast-enhanced chest or abdomen CT.

Material and Methods

We investigated 18,225 PET-CT studies of 6359 cancer patients performed from 2005 to 2012. The PET-CT studies describing potential SMM were retrieved and the corresponding medical records were reviewed. The gold standard for SMM was histopathologically-proven SMM or imaging study-based disease progression. The detectability of SMM was compared between PET-CT and contrast-enhanced CT.

Results

Twenty-six patients had 84 SMM lesions, representing a SMM prevalence of 0.41%. Lung cancer was the most common SMM-associated malignancy (54%) and the gluteal/pelvic girdle muscle was the most frequently involved SMM site (37%). All 84 SMM lesions were visualized on PET-CT (100%). Of these PET-CT positive 84 SMM lesions, 51 lesions were in the CT field of view (FOV) (61%), whereas 33 lesions were out of the CT FOV (39%). Among these 51 lesions, 17 lesions showed rim-enhancing nodules/masses (33%), eight lesions showed homogeneously enhancing nodules (16%), three lesions showed heterogeneously enhancing nodules (6%), and 23 SMM lesions (45%) were non-diagnostic by CT. All 51 SMM lesions within CT FOV were detected on PET-CT (100%), whereas only 28 were visualized on CT (54.9%), resulting in a significant difference (P < 0.005). On average, 2.6 more organs with concomitant metastases were found when SMM was revealed by PET-CT.

Conclusion

The prevalence of SMM was as low as 0.41% in the current large cohort of cancer patients. Torso PET-CT was a more competent modality than contrast-enhanced CT in the detection of SMM.

Keywords

Computed tomography (CT)positron emission tomography-computed tomography (PET-CT)malignancy neoplasm metastasis skeletal muscle

Introduction

Skeletal muscles are believed to be rare sites of cancer metastasis although the skeletal muscles receive a substantial proportion of cardiac output and constitute approximately half of the total body mass (1). The reported prevalence of skeletal muscle metastasis (SMM) among cancer patients varies, in the range of 0.03–17.5%, with a relatively higher prevalence of 6–17.5% in autopsy studies (2,3) compared to 0.03–5.6% in clinical imaging-based studies (4 –6), indicating that more sensitive imaging modalities are required to detect SMM (7).

It has been suggested that the diagnostic approach using only computed tomography (CT) for the detection of SMM may reduce the detection rate of SMM compared with F-18 fluorodeoxyglucose (FDG) positron emission tomography (PET)-CT or magnetic resonance imaging (MRI) (3,5,7). Although MRI is considered the gold standard test in the evaluation of muscle disease due to its excellent soft-tissue contrast, many SMM lesions are discovered on chest or abdomen CT, which is a routine work-up procedure for cancer patients (8).

PET-CT has become a major staging and re-staging method for cancer patients, and is superior to CT in M staging in a variety of cancers (5,9). The more routine application of PET-CT in cancer patients has resulted in the detection of more distant metastases, such as SMM, that can be missed using CT (7). Recently, there have been many case reports about SMM detected by PET-CT (10 –12) and a few comprehensive clinical studies (6,13). However, PET-CT and contrast-enhanced CT have not been directly compared regarding the detection of SMM.

This study investigated the prevalence of SMM among cancer patients using PET-CT and compared the detectability of SMM between torso PET-CT and contrast-enhanced chest or abdomen CT in the clinical setting of staging or re-staging of cancer patients.

Material and Methods

Patients

This retrospective study was approved by the Konkuk University Medical Center Institutional Review Board (KUH 1140069) in Korea and informed consent was waived. Inclusion criteria were the presence of pathologically-proven primary malignancy and the presence of SMM confirmed by histopathology or by other confirmatory imaging modality, such as follow-up CT, PET-CT, or MRI and clinical findings. Axilla or pectoralis muscle lesions in breast cancer patients were excluded since many were from local recurrence or local tumor invasion, rather than true SMM. Skin and subcutaneous fatty tissue metastases were carefully evaluated and excluded. In total, 18,225 PET-CT study reports of 6359 cancer patients in our Picture Archiving and Communication System (GE Healthcare, Milwaukee, WI, USA) acquired between 1 March 2005 and 12 June 2012 were searched. Using the search terms “muscle”, “soft tissue”, “muscular”, and “intramuscular,” 390 PET-CT studies were identified. After meticulous review of the PET-CT images and medical records of these 390 cases, 26 patients were proven to have SMM. Their PET-CT and CT were retrospectively analyzed concerning detection and location of SMM as well as primary malignancies and other concomitant metastases. There were 11 patients with chest CT, eight patients with abdomen CT, and seven patients with both chest and abdomen CT.

We retrospectively compared the PET-CT and CT for SMM detection, analyzed the SMM location and primary malignancies, and searched for the concomitant additional metastasis.

PET-CT scanning protocol

After the patients fasted for 6 h, FDG (4.8 MBq/kg) was intravenously injected in the resting state. Intravenous CT contrast agent was not administered. Blood glucose level was below 120 mg/dl. PET-CT images were acquired 60 min after FDG injection with a GEMINI scanner (Philips Medical System, Cleveland, OH, USA). The PET-CT scanner consisted of a dual-slice CT. Scan field of view (FOV) was from the skull base to the mid-thigh level (torso imaging). CT parameters were tube voltage 120 kV, current 50 mA, rotation time 0.75 s per rotation, pitch 1.5, and 5 mm section thickness. Immediately after the CT scan, PET images were acquired for 2.5 min per frame using a conventional three-dimensional protocol.

CT scanning protocol

CT was performed using one of three multidetector (MD) CT scanners: 64-channel Somatom Definition (Siemens Medical Systems, Erlangen, Germany), 64-channel LightSpeed VCT XT, and 16-channel LightSpeed Pro 16 (GE Healthcare, Milwaukee, WI, USA). For chest CT, unenhanced scans were obtained initially followed by enhanced scans with a 50-s delay after intravenous injection of 70 mL of iomeprol (Iomeron 350; Bracco Imaging Italia S.R.L, Milano, Italy) and 20 mL normal saline infused at a rate of 2 mL/s with an automatic injector. Scan images were obtained from the thoracic inlet to the kidney. The scan parameters for the LightSpeed VCT XT scanner were detector collimation 64 × 0.625 mm, pitch 0.984, 120 kVp, 200 mAs. The Somatom Definition scanner had detector collimation of 64 × 0.6 mm, pitch 1.2, 120 kVp, 200 mAs. The LightSpeed Pro 16 scanner had detector collimation of 16 × 1.25 mm, pitch 0.938, 120 kVp, and 200 mAs. The automated dose modulation using the maximum allowable tube current was set at 200 mAs, and section thickness/reconstruction intervals of 2.5 mm/2.5 mm were applied for all three scanners. For abdomen CT, unenhanced scans were obtained initially followed by enhanced scans with 15-s and 80-s delays, respectively, after intravenous injection of 150 mL contrast at a rate of 3 mL/s. Images were obtained from the lower chest to the iliac crest. The scan parameters were a detector collimation of 64 × 0.625 mm, pitch 0.984, 120 kVp, and automated dose modulation with the maximum tube current set at 200 mAs; section thickness/reconstruction interval of 3 mm/3 mm (Somatom Definition) or 3.75 mm/3.75 mm (LightSpeed VCT XT). The parameters for the LightSpeed Pro 16 were a detector collimation of 16 × 1.25 mm, pitch 0.938, 120 kV, 200 mAs, 3.75 mm/3.75 mm.

Image analysis

PET-CT images were reviewed by two nuclear medicine physicians (YS and HWC) with 19 and 10 years of clinical expertise, respectively. Chest or abdomen CT images were retrospectively reviewed by two radiologists (JGY and IS) with 20 and 3 years of clinical experience in chest radiology, respectively. PET-CT and CT data results of SMM were obtained by consensus of the two nuclear medicine physicians and the two radiologists. Initial appearances of SMM on PET-CT were evaluated and compared with CT images.

PET-CT was usually performed for initial staging and re-staging for malignancy in the current study. Contrast-enhanced CT was usually performed at the site of primary malignancy, such as chest CT for lung cancer and abdomen CT for abdominal cancer. When liver metastases were suspected in lung cancer patient, we performed abdomen CT additionally and when lung metastases were suspected in abdominal cancer patient, we added the chest CT. The additional CT examinations were performed separately; a combined chest and abdominal CT was not used. Direct comparison regarding SMM was limited to the same scan field area.

Statistical analysis

Data are represented as mean ± standard deviation. McNemar chi-square test was employed to test the detectability of the imaging studies. P value < 0.05 was considered statistically significant.

Results

Patients with SMM

Among 6359 cancer patients, 26 patients (16 men, 10 women; age range, 41–89 years; mean age, 64 ± 11 years) had 84 SMM lesions (1–27 SMM lesions/patient, 3.23 ± 5.6), representing a SMM prevalence of 0.41%. Seven SMM lesions from seven patients were confirmed by tissue biopsy and the lesions in 19 patients were confirmed after clinical evaluation and assessment by follow-up chest or abdomen CT, PET-CT, or MRI. Maximum standardized uptake value (SUVmax) of the SMM lesions were in the range of 2.5–24.3 (7.2 ± 4.5). Time interval between the enhanced CT and PET-CT was 1–12 days (4.6 ± 3.7 days). SMMs were derived from seven different primary malignancies and lung cancer was the most common primary malignancy (54 %) (Table 1). Among the 84 SMM lesions, gluteal/pelvic girdle muscle and chest/abdomen wall muscle were the common sites of SMM (Table 2).

Table 1.

Primary malignancies with skeletal muscle metastasis (n = 26).

Primary malignancy	n = 26
Lung cancer	14 (54.0%)
Pancreatic cancer	5 (19%)
Gastric cancer	2 (7.7%)
Liver cancer	2 (7.7%)
Ovarian cancer	1 (3.8%)
Uterus cancer	1 (3.8%)
Melanoma	1 (3.8%)

Table 2.

Sites of individual skeletal muscle metastasis lesions (n = 84).

Sites	n = 84
Gluteal or pelvic girdle	31 (37%)
Chest or abdominal wall	21 (25%)
Paraspinal	17 (20%)
Arm or leg	10 (12%)
Iliopsoas	5 (6%)

FDG PET-CT vs. contrast-enhanced CT for detection of SMM

Among the 84 SMM lesions, 51 lesions were in the scan field of chest or abdomen CT and 33 SMM lesions were out of CT FOV. These 33 SMM lesions were usually in the arm or leg muscles, or such as a thigh muscle metastasis studied with abdomen CT in a hepatocellular carcinoma patient (Fig. 1). A total of 51 SMM lesions within the CT FOV were analyzed in terms of detectability by PET-CT or contrast-enhanced CT. All 51 lesions were positive on PET-CT (51/51 = 100%), whereas only 28 SMM lesions were found on CT (28/51 = 54.9%, P < 0.005). The most common positive CT appearance of SMM was rim-enhancing intramuscular nodules or masses (17 lesions, 33%) (Fig. 2). Eight lesions showed homogeneously enhancing nodules (16%). Three lesions showed heterogeneously enhancing nodules (6%). However, 23 SMM lesions (45%) in the CT FOV were non-diagnostic by CT due to the lack of significant contrast enhancement of the SMM lesions (Fig. 3).

At the patient level, 26 (100%) patients were proven to have SMM using PET-CT, but only 15 (57.7%) patients were SMM positive using CT (P < 0.005), mostly due to the larger FOV of PET-CT. Among the 18 patients who had SMM in the FOV of CT, PET-CT showed a higher detectability (18/18, 100%) than CT (15/18, 83.3%) without a significant difference (P > 0.05).

Concomitant distant metastases and upstaging after detection of SMM

Concomitant distant metastatic lesions other than SMM occurred in 25 patients (25/26, 96%). An average of 2.6 other sites were subject to metastases when SMMs were first diagnosed using PET-CT. The most common site of concomitant distant metastases was bone (16/26, 62%) followed by lymph node metastasis (14/26, 54%) and lung metastasis (13/26, 50%). Distant metastases were also found in adrenal gland (6/26, 23%), liver (6/26, 23%), skin or subcutaneous fat (5/26, 19%), pancreas (4/26, 15%), pleura (3/26, 12%), peritoneum (3/26, 12%), and spleen (1/26, 4%).

Discussion

In the current study, 26 of 6359 (0.41%) cancer patients developed SMM initially and during follow-up. This result is not inconsistent with previous imaging studies reporting the prevalence of SMM of various cancer types in the range of 0.2–1.6% with PET-CT or CT (4 –6,14). SMM has long been considered as a very rare form of hematogenous metastasis in several CT and MRI studies (3,4,8,15). One of the issues regarding the detection of SMM is that they are easily missed by clinical evaluation because most SMM lesions are asymptomatic. Moreover, routine CT is usually focused on a single body part, rather than whole body. Skeletal muscles are usually at the periphery of the FOV of chest or abdomen CT scans, and are apt to receive less attention. Recently SMM from various malignancies have been reported in several case reports and a few clinical studies of PET-CT due to its almost complete covering of the whole body and higher sensitivity (6,10 –14).

The higher detectability of PET-CT versus contrast-enhanced CT may be explained by the extent of FOV and the tumor-to-background contrast. In the current study, the major reason for the lower detectability of CT was the limited FOV of CT. For instance, of the 56 CT negative SMM lesions, 33 (58.9%) were located outside the CT FOV. The FOV of torso PET-CT could have been extended further cranially to the top of the skull and caudally to the bottom of the feet (so-called true whole body PET-CT). Nguyen et al. (14) found 22 soft tissue (muscle and subcutaneous tissue) metastases in the scan field from the skull base to the upper thigh (“limited” whole body = torso) and detected 19 more lesions in the scan field from the top of the skull to the bottom of the feet (“true” whole body) among 500 patients with PET-CT. Given that, these authors detected 46% more soft tissue metastases after extension of PET-CT scan field from limited to true whole body. The more extended FOV (torso, limited whole body) of our PET-CT compared with chest or abdomen CT FOV must have contributed to the identification of more SMM by PET-CT than CT in our study. Furthermore, we might have missed some silent SMM lesions located in the scalp and lower leg muscles; therefore the true rate of SMM might have increased from 0.41% to a higher value if the FOV has been extended to a true whole body PET-CT.

Arpaci et al. (15) reported that the most common CT findings among SMM lesions were rim-enhancing nodules. Surov et al. (4) described the homogeneous enhancement as the most common CT finding (52.5%) followed by rim enhancement (32.5%). In the current study, 17 SMM lesions showed rim enhancing nodules (17/51, 33%) and eight lesion showed homogeneous enhancement (8/51, 16%) among the lesions in the CT FOV. Focusing on the diagnostic CT lesions (excluding non-diagnostic), most of the SMM lesions showed rim enhancement (17/28, 61%) followed by homogeneous enhancement (8/28, 29%).

With regard to the tumor-to-background contrast, FDG avidity of tumors on PET-CT may play a more crucial role than iodine contrast enhancement on CT in terms of identification of SMM (7). This is true because the analysis of 51 SMM lesions within the CT FOV showed higher detectability by PET-CT than contrast-enhanced CT. The same tendency was also observed in the patient level analysis. In the current study, all CT-negative SMM lesions within the CT FOV were readily visualized using PET-CT. In this regard, one should be aware of the low tumor-to-background contrast of SMM lesions when evaluating cancer patients with CT. However, it is difficult to draw definitive conclusions regarding the diagnostic accuracy for SMM, because the specificity issue was not relevantly addressed in the current retrospective study. Higher FDG uptake indicate higher glucose metabolism which can be seen not only in cancer but also in inflammation (6,7,9) and contrast-enhancement on CT does not always indicate malignancy. Further well-designed studies are needed to address the accuracy issue of imaging studies.

Since, in general, SMMs are present in the advanced stage of disease, such as stage IV disease, detection of SMM would be less likely to up-stage the clinical status (6). On average, 2.6 other sites were subject to metastases at the time when the SMMs were found in the current study. Bone, lymph nodes, or lung metastases were present in more than 50% of the patients, in order of frequency. Arpaci et al. reported that 90% of the patients had concomitant metastases, with lymph nodes, bone, and lung being the common sites (15). All of the patients were at stage IV with more than one metastatic site when SMM was detected, except for one uterine cancer patient. The latter developed two pelvic girdle muscle metastases after hysterectomy without other metastatic foci (Fig. 4). The SMM of pelvic muscles were missed at initial abdomen CT, although visible retrospectively. More frequent application of PET-CT may lead to more frequent detection of SMM, but further study is required to address this issue.

One limitation of this study is that not all the SMM lesions were pathologically proven. However, it is not always justifiable to perform tissue diagnosis of metastatic lesions including SMM especially when the patients are in the advanced stage of malignant diseases. In addition, the stated sensitivity of PET-CT in detecting SMM may be too high. Since not all 6359 patients without SMM on PET-CT were followed in the same way as the study patients, there is a small possibility that SMM may have been missed, and that the detectability of SMM by PET-CT may be less than 100%.

In conclusion, the prevalence of SMM was as low as 0.41% in a large cohort of cancer patients in a tertiary referral university hospital. Lung cancer was the most common primary malignancy associated with SMM and the gluteal/pelvic girdle muscle was the most frequently involved SMM site. PET-CT was a more competent modality for detection of SMM than contrast-enhanced CT, when staging and re-staging the cancer patients.

Footnotes

Conflict of interest

None declared.

Funding

This work was supported by Konkuk University in 2013.

References

Qvarfordt

Christenson

Eklof

. Intramuscular pressure, muscle blood flow, and skeletal muscle metabolism in chronic anterior tibial compartment syndrome. Clin Orthop Relat Res 1983; 179: 284–290.

Acinas Garcia

Fernandez

Satue

. Metastasis of malignant neoplasms to skeletal muscle. Rev Esp Oncol 1984; 31: 57–67.

Lee

Choi

Park

. Various image findings of skeletal muscle metastases with clinical correlation. Skeletal Radiol 2008; 37: 923–928.

Surov

Hainz

Holzhausen

. Skeletal muscle metastases: primary tumours, prevalence, and radiological features. Eur Radiol 2010; 20: 649–658.

Haygood

Wong

Lin

. Skeletal muscle metastases: a three-part study of a not-so-rare entity. Skeletal Radiol 2012; 41: 899–909.

Khandelwal

Takalkar

Lilien

. Skeletal muscle metastases on FDG PET/CT imaging. Clin Nucl Med 2012; 37: 575–579.

Emmering

Vogel

Stokkel

. Intramuscular metastases on FDG PET-CT: a review of the literature. Nucl Med Commun 2012; 33: 117–120.

Pretorius

Fishman

. Helical CT of skeletal muscle metastases from primary carcinomas. Am J Roentgenol 2000; 174: 401–404.

Truong

Viswanathan

Erasmus

. Positron emission tomography/computed tomography in lung cancer staging, prognosis, and assessment of therapeutic response. J Thorac Imaging 2011; 26: 132–146.

10.

Heffernan

Fennelly

Collins

. Multiple metastases to skeletal muscle from carcinoma of the esophagus detected by FDG PET-CT imaging. Clin Nucl Med 2006; 31: 810–811.

11.

Liu

Ghesani

Mirani

. PET-CT demonstration of extensive muscle metastases from breast cancer. Clin Nucl Med 2006; 31: 266–268.

12.

Yilmaz

Elboga

Celen

. Multiple muscle metastases from lung cancer detected by FDG PET/CT. Clin Nucl Med 2011; 36: 245–247.

13.

Surov

Pawelka

Wienke

. PET/CT imaging of skeletal muscle metastases. Acta Radiol 2014; 55: 101–106.

14.

Nguyen

Chaar

Osman

. Prevalence and patterns of soft tissue metastasis: detection with true whole-body F-18 FDG PET/CT. BMC Med Imaging 2007; 7: 8–8.

15.

Arpaci

Ugurluer

Akbas

. Imaging of the skeletal muscle metastases. Eur Rev Med Pharmacol Sci 2012; 16: 2057–2063.