MRI findings and demographics of elastofibroma dorsi: assessment of diffusion-weighted imaging and contrast enhancement patterns

Abstract

Background

Elastofibroma dorsi is a rare pseudotumoral lesion. Thus, there is no report of magnetic resonance imaging (MRI) findings that investigates multiple patients particularly with respect to diffusion-weighted imaging (DWI) findings and contrast enhancement patterns.

Purpose

To describe the imaging findings of elastofibroma on MRI, particularly DWI findings and contrast enhancement patterns, and to further investigate patient demographics.

Material and Methods

Forty-four patients with elastofibroma that underwent MRI were enrolled in this retrospective study. All images were evaluated by two radiologists to visually assess the signal intensity for each sequence. Enhanced elastofibromas were classified into four categories to assess the enhancement pattern. Differences in gender and laterality were also assessed statistically.

Results

An equal number of men and women were included (n = 22 each). There was no significant difference in laterality (P = 0.783). All lesions (73 lesions) had low signal intensity on both T1-weighted (T1W) and T2-weighted (T2W) images: heterogeneous in 56, homogeneous in 17. None of the 41 lesions with DWI had true abnormal diffusion restriction. The average ADC value was 1.36 × 10^–3 ± 0.29 mm²/s. All 31 lesions that had contrast-enhanced MRI were classified according to enhancement pattern: homogeneous (three lesions, 9.7%); heterogeneous (15 lesions, 48.4%); streak-like (three lesions, 9.7%); and rim-like (ten lesions, 32.2%).

Conclusion

There were no statistically significant differences in gender or laterality. Elastofibroma showed homogeneous to heterogeneous low signal intensity on T1W and T2W images. No lesion showed abnormal diffusion restriction, and all lesions demonstrated enhancement on MRI.

Keywords

Elastofibroma magnetic resonance contrast enhancement diffusion-weighted imaging laterality gender difference

Introduction

Elastofibroma dorsi is a rare pseudotumoral fibroproliferative lesion arising predominantly from the soft tissues between the lower scapula and chest wall, which was first described by Jarvi and Saxen in 1961 (1). It is categorized as a benign fibroblastic/myofibroblastic tumor in the WHO classification published in 2013 (2). Elastofibroma dorsi is histologically characterized by the accumulations of bundle-like collagen fibers and degenerating elastic fibers forming discs and globules (3). Malignant transformation and invasive growth have not been reported. Typically located in the subscapular region, less common elastofibroma sites have been reported including the olecranon, orbits, oral mucosa, deltoid muscle, foot, inguinal region, tricuspid valve, stomach, sigmoid colon, greater omentum, axilla, intraspinal space, and adjacent to the greater trochanter (4 –7). Some groups have reported elastofibroma dorsi as frequently seen in elderly women, typically in the seventh decade (age range = 6–94 years) (8 –10). Geographic and familial clusters have also been reported, particularly in Okinawa, Japan (8). The laterality and gender gap of elastofibroma have been reported in many studies (8,9,11,12), although no report has statically examined these tendencies. The clinical symptoms of elastofibroma are palpable mass, discomfort, functional restriction, occasional pain, and clicking on motion; however, it is often asymptomatic and incidentally found on computed tomography (CT) or magnetic resonance imaging (MRI) (9,13).

Recently, it has been proposed to diagnose elastofibroma by radiological findings alone if typical clinical or imaging features are present because they should be managed conservatively (13 –15). Needle biopsy or surgical resection has been increasingly avoided for non-growth, less symptomatic elastofibromas (13 –15). Thus, it is essential for radiologists to be familiar with typical and atypical imaging findings of elastofibroma dorsi on MRI and other imaging modalities. There are some previous reports regarding MRI findings of elastofibroma dorsi (9,16). However, to the best of our knowledge, there is no report of MRI findings of elastofibroma investigating multiple patients particularly with regard to diffusion-weighted imaging (DWI) findings and contrast enhancement patterns. Thus, it is still meaningful to clarify the typical and atypical MRI patterns of elastofibroma dorsi, with regard to DWI findings and contrast enhancement patterns, and further, to investigate gender differences and laterality of occurrence.

Material and Methods

This retrospective study was approved by the Institutional Review Board at each institution and written informed consent was waived.

Patient selection

All MR reports (from 2007 to 2014) at University of the Ryukyus Hospital, Yonabaru Chuo Hospital, and St. Marianna University School of Medicine Hospital were initially reviewed by two radiologists (MT and TY, with eight and 15 years of experience in diagnostic radiology, respectively) in order to identify all MR examinations of elastofibroma dorsi. Forty-four patients with 73 subscapular lesions who were diagnosed histologically with elastofibroma (n = 11, 17 lesions; ten patients with 16 lesions were diagnosed by tumor resection and one patient with one lesion by biopsy) or by a combination of typical imaging findings such as subscapular fibrous semilunar shaped lesion and clinical presentation (n = 34) were enrolled in this study. Operations were performed to alleviate symptoms such as clicking and pain, and for biopsy to exclude malignant or aggressive tumors. No patients were excluded for any reason, such as poor image quality or other coexisting diseases.

The clinical records of these patients were re-evaluated to identify patient characteristics, histopathology, clinical symptoms, and clinical course.

MR scanning

Four patients underwent MR scanning using a 3.0-T scanner, and all other patients were examined on 1.5-T MR scanners. The MR scanners used in this study were a Discovery MR750 3.0-T (GE Healthcare, Milwaukee, WI, USA) for four patients, a MAGNETOM Avanto 1.5-T (Siemens Medical Solutions, Erlangen, Germany) for 24 patients, a SignaHDe1.5-T (GE Healthcare) for 13 patients, and an Achieva1.5-T (Philips Medical Systems, Best, The Netherlands) for three patients.

Spin echo T1-weighted (T1W) imaging (TR/TE =443–572/6.71–12 ms) was performed in 12 patients and gradient echo T1W imaging (TR/TE = 110/4.76 ms, flip angle [FA] = 70°) was performed in 32 patients. Spin echo T2-weighted (T2W) imaging (TR/TE = 2500–4025/72–180 ms) was acquired in 43 patients and gradient echo T2W imaging (TR/TE = 4/1.98 ms, FA = 70°) in one patient. Fat-suppressed spin echo T2W imaging (TR/TE = 2500–9107/72–84 ms) was obtained in 32 patients, fat-suppressed gradient echo T2W imaging (TR/TE = 3.9–4.1/1.93–2.04 ms, FA = 70°) in four patients, and short tau inversion recovery (STIR) imaging (TR/TE = 655–6846/11–60 ms, FA = 150–180°, inversion time = 150–180) in six patients. DWI was obtained in 24 patients using a single-shot echo-planar imaging sequence (TR/TE =1400–9000/56–95 ms, FA = 90°) with fat suppression. The corresponding b-values to the diffusion-sensitizing gradient were 0 and 1000 s/mm² except for one patient who had b-values of 0 and 800 s/mm².

In 19 patients, gadolinium-enhanced T1W fat suppression imaging was obtained by the same imaging parameters used in the unenhanced T1W imaging and 0.2 mL/kg of contrast: gadoterate meglumine (Magnescope, Guerbet, Tokyo, Japan); gadopentetate dimeglumine (Magnevist, Bayer HealthCare, Osaka, Japan); or gadodiamide (Omniscan, Daiichi Sankyo, Tokyo, Japan) was administrated intravenously.

Image analysis

MR images from all cases were evaluated cooperatively by the two radiologists. The radiologists recorded the signal intensity for each sequence, enhancement pattern, tumor location, approximate volume, and the number of lesions. The approximate lesion volume was calculated by an ellipsoid formula (sagittal length × transversal length × antero-posterior length × 0.5). Regions of interest (ROIs) were positioned in the tumor to calculate the apparent diffusion coefficient (ADC) values. A circular ROI was placed to be as large as possible within the elastofibromas and to include the lowest ADC value by visual inspection.

To assess the enhancement pattern, the elastofibromas were classified into four categories: homogeneous, heterogeneous, streak-like, and rim-like enhancement patterns. The enhancement ratio of elastofibroma to the greater pectoral muscle was also calculated from the same image.

Statistical analysis

The laterality of the elastofibroma and any gender differences were assessed by the chi-square test. The difference in tumor size between the right and left sides was assessed by the Wilcoxon test. P values < 0.05 were considered statistically significant. All statistical analyses were performed using JMP 11.0 software (SAS Institute, Cary, NC, USA).

Results

Patient characteristics

Among the 44 patients with elastofibroma, the mean age was 66 years (age range = 35–83 years), and an equal number of men and women were included (n = 22 each). All elastofibromas were located in the subscapular region. Thirty-eight patients’ symptoms were included in the clinical records. The most common symptom was a palpable mass and other major symptoms included clicking and pain (Table 1).

Table 1.

Clinical presentations in our institutions.

Total patients (n)		44
Total elastofibroma (n)		73
Gender (n)			P = 1.000
	Male	22
	Female	22
Age (years)
	Mean age (range)	66 (35–83)
Complaint (n)
	Palpable mass	37
	Clicking with motion	6
	Pain	5
	Incidentally found on CT or MRI	2
Location (n)	Laterality		P = 0.783
	Right	37
	Left	36
	Unilateral subscapular region (right/left)	15 (8/7)
	Bilateral subscapular region	29
	Extrasubscapular region	1

MRI findings

On MRI, 29 patients (65.9 %) had bilateral lesions and 15 patients (34.1 %) unilateral. Among patients with a unilateral lesion, eight lesions (53.3 %) were in the right subscapular region and seven (46.7 %) in the left. In total, there were 73 elastofibromas found by MRI in the 44 patients. There was no significant difference in laterality by chi-square test (P = 0.783). The mean approximate volume of the right subscapular elastofibroma was 56.4 ± 52.3 cm³ and that of the left was 47.3 ± 48.5 cm³. There was also no significant difference in tumor volume between the right and left side (P = 0.562).

All lesions had low signal intensity in most of the volume on both T1W and T2W images, which reflects the fibrous tissue. Fifty-six lesions in 32 patients showed a variable number of heterogeneous high signal streaks on T1W and T2W images, which were mainly caused by the fat component (Fig. 1). The remaining 17 lesions in 12 patients demonstrated homogeneously low signal intensity on T1W and T2W images, because they were purely fibrotic lesions and had less fat streaks. One of these lesions was histologically proven to have a decreased fat component (Fig. 2). Among the 66 lesions that were scanned on T2FS or STIR, 48 lesions showed partial or regional high signal intensity.

Fig. 1.

A 61-year-old man diagnosed with right subscapular elastofibroma. MRI shows the typical heterogeneous pattern on T2W imaging (a) and T1W imaging (b) by fibrotic tissue with a streak-like fat component.

Fig. 2.

A 63-year-old female patient diagnosed with right subscapular elastofibroma underwent surgery because of pain and clicking. The lesion appears as homogeneous and low-intensity on T2W imaging (a) and T1W imaging (b). A homogeneous enhancement pattern was also seen (c). The gross specimen was primarily fibrous tissue, and just a few thin, linear streaks of fat (arrowheads) were seen (d). A photomicrograph image (original magnification × 20; elastica-van Gieson stain) (e) showed accumulations of collagen fibers and degenerating elastic fibers, which were confirmed as elastofibroma. The low signal intensity on MRI is considered to represent the fibrous component. The difference in signal heterogeneity may be attributed to a difference in fat quantity.

DWI was obtained in 41 lesions from 24 patients. Seven lesions in four patients were histologically proven. Four lesions had high signal intensity on DWI, although they also had high signal intensity on the ADC map and a high ADC value (1.22–1.93 mm²/s), suggesting a T2-shine through phenomenon. There were no lesions that showed high signal intensity on DWI accompanied with corresponding low signal intensity in the lesion on the ADC map. The average ADC value was 1.36 × 10^–3± 0.29 mm²/s (0.77–1.93 ×10^–3 mm²/s) (Fig. 3).

Fig. 3.

A 70-year-old man with left elastofibroma dorsi. The (a) T2W imaging, (b) DWI (b = 1000 s/mm²), and (c) ADC map are shown. The elastofibroma has low signal intensity on DWI and a high ADC value (1.31 ± 0.21 [SD] × 10^–3 mm²/s).

A total of 31 lesions in 19 patients who underwent contrast-enhanced MRI showed various enhancement patterns. Nine lesions in five patients were histologically confirmed. Three lesions had a homogeneous enhancement pattern (9.7%), 15 lesions were heterogeneous (48.4%), three lesions were streak-like (9.7%), and ten lesions were rim-like (32.2%) (Table 2) (Fig. 4). In nine histologically proven lesions, one lesion had a homogeneous pattern, six lesions were heterogeneous, and two lesions were streak-like. The mean enhancement ratio of elastofibroma to greater pectoral muscle was 2.1 ± 0.81.

Table 2.

Enhancement patterns on MR images.

	n
Patients (lesions)	19 (31)
Negative for enhanced lesion	0
Enhanced lesion	31
Homogeneous	3
Heterogeneous	15
Streak-like	3
Rim-like	10

Fig. 4.

(a–d) Male patients aged (a) 67 years, (b) 73 years, (c) 61 years), and (d) 46 years diagnosed with elastofibroma show enhancement patterns on MR. All elastofibromas in this study were enhanced in homogeneous (a), heterogeneous (b), streak-like (c), or rim-like patterns (d).

Discussion

MRI indication for elastofibroma is basically low signal intensity on T1W and T2W images, which reflects the histopathologic features of the fibro-elastic component. The signal heterogeneity can have a range from homogeneous to heterogeneous due to a varying degree of layered fat streaks. All elastofibromas in this study were enhanced in various patterns and no lesion showed abnormal diffusion restriction on the DWI and ADC maps. It is essential for diagnostic radiologists to know that elastofibromas can have these various imaging features on MRI.

Although the female-to-male ratio was 1:1 and there was no significant difference in gender in our study, it has been reported in multiple studies that elastofibroma tends to be more common in women (female-to-male ratio = 3.7–13.2) (8,9,11,12). However, Giebel et al. reported that elastofibroma was detected in men more than twice as much as in women (ten men [16.9%] vs. three women [7.3%] in 100 autopsy specimens, female-to-male ratio = 0.3) (17). Considering these studies and our result, we suggest that the true female-to-male ratio may be much closer to equal than it has been reported previously.

In the current study, there was also no significant difference in laterality between the right and left sides, with regard to frequency and approximate volume. Thus, although this finding also differs from previous reports, which indicated that elastofibroma was frequently found on the right side (8,9,18), we believe that there are no laterality or gender differences in the occurrence of elastofibroma.

The typical MRI appearance of elastofibroma dorsi is a semilunar-shaped mass in the subscapular region, known to show heterogeneous, low to iso-signal intensity compared to skeletal muscle with high-signal intensity streaks in parallel to the longest axis both on T1W and on T2W images. The signal intensity reflects the histopathologic features, fibro-elastic components, and layered fat streaks (9,14,16).

However, 17 homogenous solid lesions without fat streaks were detected in our study. We speculate that these homogenous lesions were mainly composed of elastic tissue with less fat, compared to the typical elastofibroma (Figs. 1 and 2), and that the heterogeneity of the signal depends on the amount of the fatty component.

In our study, all elastofibromas were enhanced in various ways: heterogeneous; homogenous; rim-like; and streak-like patterns. It has been reported previously that elastofibroma can be enhanced heterogeneously to various degrees (9,14). However, to the best of our knowledge, this is the first study to indicate the different enhancement patterns and the frequency of each pattern. These enhancement patterns seemed to reflect the enhancement of the fibrous component in the lesion. The fact that all elastofibromas were enhanced in our study indicates that elastofibromas are enhanced with very high frequency despite previous studies, which reported the presence of unenhanced elastofibromas (9).

On DWI, most of the lesions had low signal intensity with low to high ADC values. There were no lesions with high signal intensity on DWI accompanied with a low ADC value. The average ADC value of the elastofibromas in this study was lower than that for benign soft tissue tumors reported previously (19). Although some lesions had a low ADC value (lowest 0.77 × 10⁻³± 0.34 mm²/s), no lesion showed high signal intensity on DWI. We considered that the low ADC values of elastofibromas could be explained by a “T2 blackout effect” caused by hypointensity of the fibrous component in the elastofibroma on a T2W image (20,21). To the best of our knowledge, this is the first report evaluating the ADC values of elastofibromas from multiple patients.

The differential diagnosis of soft tissue tumors in periscapular regions, which show low to iso-signal intensity on T1W and T2W images is limited. Elastofibroma is a fibrous mass with some adipose tissue, thus fibrous and fibroblastic tumors such as extra-abdominal desmoids and fibromas would be the most suspicious differential diagnosis. A second possibility would be an adipocytic tissue tumor such as liposarcoma. Scapulothoracic bursitis could also be a differential diagnosis because the location and shape are similar to those of an elastofibroma. In addition, a projecting bone tumor of the scapula such as an osteochondroma might appear similar on MRI. Although there are multiple differential diagnoses described above, the characteristic location and imaging findings make it easy to differentiate elastofibroma from other lesions. However, as indicated in our study, elastofibroma may show various and sometimes strong enhancement patterns and a low ADC value. In addition, it is reported that uptake of FDG is frequently observed in elastofibroma. From a previous study, it is known that elastofibroma is often incidentally detected as a FDG-avid mass on FDG-PET/CT (22,23). It was reported that frequent mild to moderate uptake, and occasional high or no uptake of FDG may be observed with elastofibroma (the mean SUV = 2.0 ± 0.63; range = 0–5.1) (22). It is important to understand these findings to avoid misdiagnosing elastofibroma as a malignant tumor.

There were several limitations to our study. First, this was a retrospective study using MRI data from multiple institutions. MR images were obtained in different scan settings. Second, only ten cases of elastofibroma were diagnosed histologically. If typical clinical and imaging features are present, further biopsy or resection is generally avoided in patients with elastofibroma. Because all cases had a typical clinical course and imaging findings, we consider that the clinical diagnosis of elastofibroma is reliable in all enrolled patients. Third, although MR images were evaluated cooperatively by two radiologists, independent reporting by each radiologist and then a comparison of the findings would have strengthened the reliability of the outcome.

In conclusion, MRI findings of elastofibroma are considered to have low signal intensity on T1W and T2W images with various signal heterogeneities, no abnormal diffusion restriction, and enhanced in various patterns. In addition, there were no significant differences in gender or lesion laterality found by our study.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Jarvi

Saxen

. Elastofibroma dorse. Acta Pathol Microbiol Scand Suppl 1961; 51(Suppl 144): 83–84.

Christopher

Fletcher

Bridge

. WHO classification of tumours of soft tissue and bone , 4th edn. Lyon: IARC, 2013, pp. 110–111.

Fukuda

Miyake

Masuda

et al.

Histogenesis of unique elastinophilic fibers of elastofibroma: ultrastructural and immunohistochemical studies. Hum Pathol 1987; 18: 424–429.

El Hammoumi

Qtaibi

Arsalane

et al.

Elastofibroma dorsi: clinicopathological analysis of 76 cases. Korean J Thorac Cardiovasc Surg 2014; 47: 111–116.

Miranda Zambrano

Rivero-Garvia

Galbarriatu

et al.

Spinal elastofibroma: a case report. Pediatr Neurosurg 2013; 49: 240–242.

Sakatani

Shomori

Adachi

et al.

Elastofibroma of the sigmoid colon. Pathol Res Pract 2000; 196: 205–207.

Potter

Summerlin

D-J

Rodgers

. Elastofibroma: the initial report in the oral mucosa. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2004; 97: 64–67.

Nagamine

Nohara

Ito

. Elastofibroma in Okinawa. A clinicopathologic study of 170 cases. Cancer 1982; 50: 1794–1805.

Battaglia

Vanel

Pollastri

et al.

Imaging patterns in elastofibroma dorsi. Eur J Radiol 2009; 72: 16–21.

10.

Marin

Perzin

Markowitz

. Elastofibroma dorsi: benign chest wall tumor. J Thorac Cardiovasc Surg 1989; 98: 234–238.

11.

Tamimi Marino

Sesma Solis

Perez Lara

et al.

Sensitivity and positive predictive value of magnetic resonance imaging in the diagnosis of elastofibroma dorsi: review of fourteen cases. J Shoulder Elbow Surg 2013; 22: 57–63.

12.

Naylor

Nascimento

Sherrick

et al.

Elastofibroma dorsi: radiologic findings in 12 patients. Am J Roentgenol 1996; 167: 683–687.

13.

Nishio

Isayama

Iwasaki

et al.

Elastofibroma dorsi: diagnostic and therapeutic algorithm. J Shoulder Elbow Surg 2012; 21: 77–81.

14.

Malghem

Baudrez

Lecouvet

et al.

Imaging study findings in elastofibroma dorsi. Joint Bone Spine 2004; 71: 536–541.

15.

Parratt

Donaldson

Flanagan

et al.

Elastofibroma dorsi: management, outcome and review of the literature. J Bone Joint Surg Br 2010; 92: 262–266.

16.

Faccioli

Foti

Comai

et al.

MR imaging findings of elastofibroma dorsi in correlation with pathological features: our experience. Radiol Med 2009; 114: 1283–1291.

17.

Giebel

Bierhoff

Vogel

. Elastofibroma and pre-elastofibroma—a biopsy and autopsy study. Eur J Surg Oncol 1996; 22: 93–96.

18.

Kransdorf

Meis

Montgomery

. Elastofibroma: MR and CT appearance with radiologic-pathologic correlation. Am J Roentgenol 1992; 159: 575–579.

19.

van Rijswijk

Kunz

Hogendoorn

et al.

Diffusion-weighted MRI in the characterization of soft-tissue tumors. J Magn Reson Imaging 2002; 15: 302–307.

20.

Hiwatashi

Kinoshita

Moritani

et al.

Hypointensity on diffusion-weighted MRI of the brain related to T2 shortening and susceptibility effects. Am J Roentgenol 2003; 181: 1705–1709.

21.

Silvera

Oppenheim

Touzé

et al.

Spontaneous intracerebral hematoma on diffusion-weighted images: influence of T2-shine-through and T2-blackout effects. Am J Neuroradiol 2005; 26: 236–241.

22.

Onishi

Kitajima

Senda

et al.

FDG-PET/CT imaging of elastofibroma dorsi. Skeletal Radiol 2011; 40: 849–853.

23.

Blumenkrantz

Bruno

González

et al.

Characterization of elastofibroma dorsi with 18FDG PET/CT: a retrospective study. Rev Esp Med Nucl 2011; 30: 342–345.