Differential diagnosis of fibroma of tendon sheath and giant cell tumor of tendon sheath in the finger using signal intensity on T2 magnetic resonance imaging

Abstract

Background

The differential diagnosis of fibroma of tendon sheath (FTS) and giant cell tumor of tendon sheath (GCTTS) on the basis of clinical and radiographic characteristics remains difficult.

Purpose

To evaluate the quantitative measurement of signal intensity (SI) obtained by magnetic resonance imaging (MRI) for the differential diagnosis of FTS and GCTTS in just the finger.

Material and Methods

We retrospectively identified patients with FTS (n = 6) and GCTTS (n = 22) of the finger who were treated at our hospitals between April 2011 and August 2019. Two researchers independently reviewed the MRIs and measured the regions of interest (ROIs) in the tumor and flexor tendon from the same image. The SI ratio obtained for the tumor and tendon ROIs was measured and compared using receiver-operating characteristic curve analyses. Sensitivity and specificity analyses were performed.

Results

The SI ratios (mean ± SD) of FTS and GCTTS were 1.83 ± 0.64 and 6.34 ± 3.16 for researcher 1 and 1.82 ± 0.60 and 6.10 ± 3.22 for researcher 2, respectively. The areas under the curve were 0.970 and 0.970 for researchers 1 and 2, respectively. The cut-off values of the SI ratio as determined by researchers 1 and 2 for differentiating FTS from GCTTS were 3.00 and 3.00, respectively (sensitivity = 95.5%, specificity = 100%).

Conclusions

The SI ratio is useful for differentiating FTS from GCTTS independent of a combination of tumor signal and shape.

Keywords

Differential diagnosis magnetic resonance imaging tumor hand

Introduction

Fibroma of tendon sheath (FTS) is a benign, slow-growing fibrous tumor. FTS was first characterized in 1979 by Chung and Enzinger in a series of 138 cases (1). Characteristic features include well-circumscribed lesions located predominantly on the upper extremities, particularly on the fingers (49%) and hands (12%) (1). FTS commonly occurs as a painless mass on the flexor aspect of the hand (2). The clinical presentation of FTS mimics that of giant cell tumor of tendon sheath (GCTTS). The lesions that arise as a result of the two conditions, especially in the fingers, are similar in size, location, and gross morphological features. Moreover, magnetic resonance imaging (MRI) manifestations are similar between FTS and GCTTS in previous studies (3 –7). These tumors typically have a signal intensity (SI) that is lower than or equal to that of skeletal muscle on T1-weighted (T1W) and T2-weighted (T2W) imaging (8). Therefore, the differential diagnosis of these tumors on the basis of clinical and MRI characteristics remains difficult. As reported in a recent study, these tumors can be classified as either FTS or GCTTS based on a combination of two features: homogenous signal and hypointensity shape on proton density-weighted imaging (PDWI) (9). However, this method of discrimination cannot be applied to finger tumors because that study did not include any patients with FTS of the finger (9).

Ideally, the method of differentiation should be simple and accurate. We searched for a more simple and accurate technique of discriminating these tumors other than tumor shape or signal. The aim of the present study was to evaluate the quantitative measurement of the SI of MRI for the differential diagnosis of FTS and GCTTS in only the finger.

Material and Methods

Our university’s Institutional Review Board for Clinical Research approved this retrospective study (no. 285-65), and the requirement for informed consent was waived.

Patients and study design

Inclusion criteria were as follows: (i) soft-tissue tumor arising in the finger; (ii) histology of FTS or GCTTS; and (iii) primary tumor treated in our hospitals between April 2011 and August 2019. A total of 43 finger tumors underwent resection during the period specified. Thirteen patients were excluded because of histology. Two patients were excluded because of recurrence (Fig. 1). Finally, six patients with FTS (4 men, 2 women; median age = 52.5 years) and 22 patients with GCTTS (8 men, 14 women; median age = 56 years) were identified and included in this study. The tumors, located in the hand, were confirmed pathologically from the specimens obtained after surgery.

Fig. 1.

Inclusion criteria were as follows: (i) soft-tissue tumor arising in the finger; (ii) histology of FTS or GCTTS; and (iii) inclusion of the primary tumor. Thirteen patients were excluded because of histology. Two patients were excluded because of recurrence. Six patients with FTS and 22 patients with GCTTS were identified and included in this study. FTS, fibroma of tendon sheath; GCTTS, giant cell tumor of tendon sheath.

MRI protocol

All MRI examinations were performed using the multivendor 1.5- or 3.0-T system. We analyzed the slice of the tumor center obtained using fast spin-echo T2W cross-sectional images (repetition time = 3000–4800 ms, echo time = 80–100 ms, field of view = 40–150 mm, slice thickness/gap = 1–3/0.1–0.5 mm) before the surgery.

MRI analysis

All images were independently analyzed by two readers (R1, radiologist [HT] with 23 years of clinical experience; R2, orthopedic surgeon [ME] with 15 years of clinical experience) who were blinded to specimen identification. The evaluated MRI features included lesion morphology and the imaging features on T1W and T2W imaging. Regions of interest (ROIs) were measured in the tumor and flexor tendon on the same MR axial image, which showed the maximum tumor diameter. The ROIs in the tumors and the tendon were hand-drawn and set with the surrounding tumor and tendon. The mean SI values of the tumor (SI tumor) and tendon (SI tendon) were calculated using SYNAPSE5 (v. 5.4.001; FUJIFILM Medical Systems, USA, Inc.). The SI ratio was calculated using the following formula:

SI ratio = SI tumor / SI tendon

The ROIs of the tumor and tendon were measured by two researchers, who were provided with representative examples of ROI placement. The ROIs of the same lesions as set by the two researchers were subsequently compared (Fig. 2). All patients had normal tendons that showed homogenous low intensity on both T1W and T2W imaging.

Fig. 2.

Image analysis of representative cases. In the T2W axial images, the two researchers independently set the ROIs of the tumor (solid line) and flexor tendon (dotted line) and calculated the SI ratios of these lesions. (a) A 43-year-old male patient with FTS. SI ratio = 2.42 (researcher 1) and 2.61 (researcher 2). (b) A 60-year-old female patient with GCTTS. SI ratio = 7.94 (researcher 1) and 8.91 (researcher 2). FTS, fibroma of tendon sheath; GCTTS, giant cell tumor of tendon sheath; ROI, region of interest; SI, signal intensity; T2W, T2-weighted.

Statistical analysis

Fisher’s exact test and Mann–Whitney U tests were used to compare the associations between the clinicopathological factors. The Mann–Whitney U tests were used to compare the SI between patients with FTS and GCTTS. Interclass correlation coefficients (ICCs) and Bland–Altman statistics were calculated to evaluate the interrater reliability between the two researchers, using MedCalc (v. 18.11.6; MedCalc Software, Ostend, Belgium). The 95% confidence interval (CI) of the mean difference was used to determine systematic bias. Significant systematic bias between measurements cannot be inferred if zero is included within the 95% CI (10). To investigate the presence of proportional bias, a regression equation was established from the Bland–Altman plot and the significance of the regression was tested. A significant regression (P < 0.05) was indicative of proportional bias. Statistical analyses were performed using SPSS, version 23 (IBM Corp., Armonk, NY, USA). Receiver-operating characteristic (ROC) curve analyses were performed using MedCalc. Optimal cut-off values were determined by Youden index, and the sensitivity, speciﬁcity, positive predictive value, and negative predictive value were calculated. P values < 0.05 were considered statistically signiﬁcant.

Results

Clinicopathological characteristics

Six patients with FTS (4 men, 2 women; median age = 52.5 years) and 22 patients with GCTTS (8 men, 14 women; median age = 56 years) treated in our hospitals between April 2011 and August 2019 were identified and included in this study. The mean tumor size was 20.0 mm (range = 11.2–30 mm) for FTS and 14.1 mm (range = 10.0–21.0 mm) for GCTTS.

MRI features

MRI findings are shown in Table 1. The shapes of the tumors were as follows: five FTS and 10 GCTTS were oval; four GCTTS were lobulated; five GCTTS were nodular; and one FTS and three GCTTS had a casting mold appearance. We evaluated bone erosion on MRI. One of the six lesions with FTS showed bone absorption. Of the 22 lesions with GCTTS, six showed bone erosion. Of the six FTS lesions, T1W imaging showed heterogenous low isointensity relative to muscle in two lesions and homogenous isointensity in four lesions. Of the 22 GCTTS lesions, T1W imaging showed heterogenous low isointensity relative to muscle in four lesions and homogenous isointensity in 18 lesions. All FTS and GCTTS lesions showed heterogenous low–high intensity on T2W MRI.

Table 1.

MRI features of patients with FTS and GCTTS.

MRI feature	FTS (n = 6)	GCTTS (n = 22)
Morphology
Oval	5	10
Lobulated	0	4
Nodule	0	5
Casting mold	1	3
Bone absorption
Yes	1	6
No	5	16
T1W imaging
Heterogenous low isointensity	2	4
Homogenous isointensity	4	18
T2W imaging
Heterogenous low–high intensity	6	22

FTS, fibroma of tendon sheath; GCTTS, giant cell tumor of tendon sheath; MRI, magnetic resonance imaging; T1W, T1-weighted; T2W, T2-weighted.

SI ratio

The SI ratios (mean ± SD) of the FTS and GCTTS lesions were 1.83 ± 0.64 and 6.34 ± 3.16 for researcher 1 and 1.82 ± 0.60 and 6.10 ± 3.22 for researcher 2, respectively (P = 0.001; Fig. 3).

Fig. 3.

Comparison of SI ratios and coefficient of variation (CV) with FTS and GCTTS. (a) The CVs of FTS and GCTTS in the ROIs were not significantly different. (b) The mean values of the SI ratios of FTS and GCTTS were significantly different (P = 0.001). CV, coefficient of variation; FTS, fibroma of tendon sheath; GCTTS, giant cell tumor of tendon sheath; ROI, region of interest; SI, signal intensity.

ICCs and Bland–Altman analysis

ICCs were calculated between the two researchers for SI tumor (r = 0.9998; 95% CI = 0.9995–0.9999), SI tendon (r = 0.9994; 95% CI = 0.9987–0.9997), and SI ratio (r = 0.9633; 95% CI = 0.9202–0.9834). Moreover, the mean difference ± 1.96 SD with Bland–Altman statistics were on the SI tumor (–3.44; 95% CI = –8.47 to 1.58), the SI tendon (–2.07; 95% CI = –5.91 to 1.78), and the SI ratio (0.11; 95% CI = – 0.23 to 0.45) (Fig. 4). Significant systematic bias between measurements was not observed because of the inclusion of zero within the 95% CI. Moreover, the regression was not significant and no proportional bias was observed.

Fig. 4.

The ROC curve between FTS and GCTTS for each researcher. The AUCs by researcher 1 and 2 were 0.970 (95% CI = 0.825–0.992) and 0.970 (95% CI = 0.825–0.992). The cut-off values of the SI ratio by researchers 1 and 2 for differentiating between FTS and GCTTS were 3.00 and 3.00, respectively, on the most accurate point of the ROC curve (sensitivity = 85.7%, specificity = 100%, positive predictive value = 100%, negative predictive value = 95.5%). AUC, area under the ROC curve; CI, confidence interval; FTS, fibroma of tendon sheath; GCTTS, giant cell tumor of tendon sheath; ROC, receiver-operating characteristic; SI, signal intensity.

ROC analysis

The results of the ROC curve analysis for each researcher are shown in Fig. 5. The area under the curve for researchers 1 and 2 was 0.970 (95% CI = 0.825–0.992) and 0.970 (95% CI = 0.825–0.992), respectively. The cut-off values of the SI ratio obtained by the researchers to differentiate FTS from GCTTS were 3.00 and 3.00, respectively, on the most accurate point from the ROC curve (sensitivity = 85.7%; specificity = 100%; positive predictive value = 100%; negative predictive value = 95.5%).

Fig. 5.

ICCs and Brand–Altman analysis of each SI between the two researchers. (a) The regression was not significant for SI tumor (P = 0.273; ICCs demonstrated excellent reliability, and zero was included within the 95% CI). (b) The regression was not significant for SI tendon (P = 0.640; ICCs demonstrated excellent reliability, and zero was included within the 95% CI). (c) The regression was not significant for SI ratio (P = 0.526; ICCs demonstrated excellent reliability, and zero was included within the 95% CI). CI, confidence interval; ICC, interclass correlation coefficient; SI, signal intensity.

Discussion

Soft-tissue tumors in the finger with decreased SI in T2W images are often FTS or GCTTS. The SI ratio calculated for the tumor and flexor tendon by T2W MRI helps to differentiate FTS from GCTTS. The cutoff values of the SI ratio for differentiating FTS and GCTTS were 3.00 and 3.00 for researchers 1 and 2, respectively, which was obtained using the most accurate point from the ROC curves.

FTS is a rare, benign tumor manifesting as slow-growing lesions in adults aged 20–50 years. These lesions predominantly arise in the upper extremities (82% of all lesions) and particularly in the fingers, hands, and wrist. The MRI signal characteristics of FTS are varied and not specific. T1W MR images mostly present as isointense or with low SI (8). However, on T2W images, FTS exhibits various signal changes. Generally, tumors with low SI on T2W images have significant fibrous elements with marked hypocellularity. This variability is attributed to the variable degree of cellularity and hyalinization (11). However, GCTTS, possibly because of its hemosiderin content, may have a low T2 signal. Therefore, it is difficult to differentiate tumors with low SI on T2W images.

A recent study reported that the homogeneity of the signal and shape of hypointensity on PDWI is useful for discriminating between FTS and GCTTS; however, that study did not include patients with FTS or GCTTS of the finger. Thus, the results cannot be applied to finger tumors (9). Morphologically, most FTS have been reported to be round or oval, whereas GCTTS are lobulated or exhibit a casting mold pattern (9). The morphology of FTS in this study was consistent with that reported in the literature (8); however, most GCTTS in this study had an oval-like pattern. That is because GCTTS of the finger is relatively smaller than that of other sites and does not result in a lobulated appearance.

Although the incidence of adjacent bone absorption is higher with GCTTS than with FTS (9,12), both FTS and GCTTS had the same incidence of bone erosion in this study. The differentiation between GCTTs and FTS in the fingers may be difficult because of bone erosion. Finger tumors are detected earlier than in other sites, and thus, bone destruction is unlikely to occur.

The aim of the present study was to determine the optimal method of discriminating these tumors. Therefore, we examined the potential of SI values, obtained from T2W MRI, for differentiating FTS from GCTTS. The SI ratio calculated for the tumor and flexor tendon by T2W MRI helps to differentiate FTS from GCTTS, enabling an accurate diagnosis independent of a combination of the signal and shape.

However, the SI value obtained using MRI is not an absolute value and may vary with various factors, including signal detection and signal amplification. Therefore, in the current study, we measured the SI of the tendon and calculated the SI ratio, because it existed constantly within the same cross-sectional images as those of the finger tumors. The use of SI, which does not vary with the MR scanner or scan technique, makes it possible to apply the current method for T2W images obtained elsewhere.

The present study has some limitations. First, the sample size of the FTS group was limited. Second, MRI was performed using different scanners with two different magnetic field strengths and a varied imaging technique.

In conclusion, the cut-off value of 3.00 of the SI ratios, calculated from the SI values of the tumor and flexor tendon on T2W images, is useful for differentiating FTS from GCTTS.

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 the following financial support for the research, authorship, and/or publication of this article: This work was partly supported by grants from JSPS KAKENHI (17K10974 to ME).

ORCID iD

Makoto Emori

References

Chung

Enzinger

FM.

Fibroma of tendon sheath. Cancer 1979; 44:1945–1954.

Al-Qattan

MM.

Fibroma of tendon sheath of the hand: a series of 20 patients with 23 tumours. J Hand Surg Eur 2014; 39:300–305.

Glover

Chebib

Simeone

FJ.

Intra-articular fibroma of tendon sheath arising in the acromioclavicular joint. Skeletal Radiol 2014; 43:681–686.

Hitora

Yamamoto

Akisue

, et al. Fibroma of tendon sheath originating from the knee joint capsule. Clin Imaging 2002; 26:280–283.

Kundangar

Pandey

Acharya

, et al. An intraarticular fibroma of the tendon sheath in the knee joint. Knee Surg Sports Traumatol Arthrosc 2011; 19:1830–1833.

Solomou

Kraniotis

Giant cell tumor of the tendon sheath of the tendinous insertion in pes anserinus. Radiol Case Rep 2017; 12:353–356.

Gouin

Noailles

Localized and diffuse forms of tenosynovial giant cell tumor (formerly giant cell tumor of the tendon sheath and pigmented villonodular synovitis). Orthop Traumatol Surg Res 2017; 103(Suppl. 1):91–97.

Fox

Kransdorf

Bancroft

, et al. MR imaging of fibroma of the tendon sheath. AJR Am J Roentgenol 2003; 180:1449–1453.

Guo

You

, et al. Magnetic resonance imaging features of fibromas and giant cell tumors of the tendon sheath: differential diagnosis. Eur Radiol 2019; 29:3441–3449.

10.

Bland

Altman

DG.

Measuring agreement in method comparison studies. Stat Methods Med Res 1999; 8:135–160.

11.

Athwal

Bueno

Bansal

, et al. Intra-articular fibroma of tendon sheath involving the scapholunate and radiocarpal joints. Skeletal Radiol 2006; 35:599–602.

12.

Palmerini

Staals

Maki

, et al. Tenosynovial giant cell tumour/pigmented villonodular synovitis: outcome of 294 patients before the era of kinase inhibitors. Eur J Cancer 2015; 51:210–217.