Analysis of Imaging and Pathological Features in 12 Cases of Schwannomatosis and Literature Review

Objective We analyzed US, CT, and MR images plus the pathological characteristics of 12 patients with pathologically conrmed schwannomatosis to explore the relationship between imaging and pathology. Method A retrospective analysis was undertaken over a 10-year period (01/01/2000-31/12/2019) of 12 patients undergoing Imaging and pathological examination for schwannomatosis, and the relevant literature was reviewed. Results The median age at diagnosis was 39.4 years (range, 22–56 years). We included the mean diameter of 36 lesions (5.10 ± 0.84 cm; range, 0.5–9.1 cm) in the focal observation. US showed that the mass boundary was clear, and most of the masses were low echo. CT showed that tumors were low density; the plain-scan CT value was 22–35 HU and enhancement amplitude 10–30 HU. MRI showed that most of the tumors were low or medium signal on T1-weighted image (T1WI) and high signal on T2WI; enhancement could be homogeneous or inhomogeneous. We could see the cellular Antoni A and myxoid Antoni B areas of the microscopic morphology. According to the proportions of the two areas, masses could be divided into four types: I, II, III and IV. Immunohistochemical (IHC) staining showed that the expression of S-100 protein was diffusely positive (36/36). Conclusions Imaging examination in schwannomatosis is helpful for localization and characterization of focus, as well as observation of the relationship between the tumor and the surrounding tissue structure. However, the specic diagnosis should also be based on pathological manifestations, IHC results and genetic analysis. Characteristics of mass imaging correlate with pathological changes in the mass.


Introduction
Schwannomatosis is a syndrome that is characterized by multiple peripheral schwannomas and can be sporadic or familial in nature. It is reported that the incidence rate is approximately one in 140,000-150,000 [1,2]. Painless masses are the most common clinical manifestation. In one study, 89% of schwannomatosis patients had peripheral schwannoma, 74% had spinal schwannoma and 9% had intracranial tumor that did not affect the vestibular nerve [3]. The highest priority in the diagnosis of schwannomatosis is to distinguish it from neuro bromatosis type 2 (NF2). About one-third of schwannomatosis patients carry a mutation of leucine zipper transcription regulator 1 (LZTR1), which is located at 22q11.2 and centromeric to smarcb1, a gene that also factors into differential diagnosis between schwannomatosis and NF2, but in many clinical cases the mutant gene cannot be found. Diagnosis of schwannomatosis is therefore usually based on clinical examination, speci cally the clinical criteria of ≥2 non-intradermal schwannomas and the absence of bilateral vestibular schwannoma. Imaging and pathology are part of the diagnostic criteria to some degree for this syndrome. The use of imaging examination in the diagnosis of schwannomatosis has been well described in the extant literature [4]. However, reports cover both imaging-ultrasound (US), computed tomography (CT) and magnetic resonance (MR)-and pathological characteristics of schwannomatosis. In this study, we reviewed 12 cases of schwannomatosis seen over the past two decades at our institution, analyzing their clinical, imaging and pathological characteristics, and attempted to elucidate the internal relationship between imaging and pathology in schwannomatosis.

Clinical Data And Methods
This study was a retrospective analysis of 12 cases of schwannomatosis con rmed by pathology after complete imaging examination (US, CT and MRI) at our hospital during January 2000-December 2019. The median age at diagnosis was 39.4 years (range, 22-56 years), with a median delay from the initial symptoms to diagnosis of 7.4 years (range, 0.1-40 years). All 12 patients received imaging examinations.
Two cases were intracranial, 7 involved the intraspinal nerves, 6 were in the torso, 11 were in the limbs, 5 were in the chest and 7 were abdominal. The total number of masses was >60 (some masses had fused with one another, making them di cult to accurately count). The focus of imaging observation and pathological diagnosis was the masses that underwent surgical resection or puncture biopsy. (Intracranial and intraspinal masses are not easily examined by US and were therefore not included in the selective analysis). Ultimately, we included a total of 36 masses in our analysis (Table 1). MHz; for the abdominal convex array probe, the frequency was 2-5 MHz. Two deputy chief US specialists observed schwannomatosis patients' US imaging data. We recorded tumor location, quantity, size, shape, boundary, internal and rear echo, relationship with peripheral nerves and blood vessels, and blood ow signals. When opinions differed, a consensus was reached after discussion. Evaluation criteria were as follows: Blood ow signal in the mass was graded on the Adler scale: grade 0, no blood ow signal found in the tumor; grade I, small amount of blood ow, with one to two punctate or rod-shaped tumor blood vessels visible; grade II, medium amount of blood ow, with three to four punctate blood vessels visible; and grade III, large amount of blood ow, with at least ve punctate blood vessels or two longer blood vessels.

CT examination method
We used a Siemens SOMATOM Emotion multi-row spiral CT scanner (Siemens AG, Munich, Germany) with layer thickness of 5-10 mm, layer spacing of 5 mm, window width of 200-350 HU, window position of 35-50 HU, 1.25-mm thin-layer reconstruction of the focus layer, 120 kV and 220 mAs. Meglumine dimeglumine (300 mg/ml) was used as the contrast agent; we injected it into the elbow vein using a high-pressure syringe at a dosage of 80-120 ml or 1.5 ml/kg body weight and an injection rate of 2.0-4.0 ml/s. Times of twophase contrast-enhanced scanning were 25-30 s for the arterial phase and 1 min for the venous phase after injection of contrast medium. Two senior radiologists analyzed the CT manifestations of the lesions, including location, size, shape, number, growth mode, edge, scan density and enhancement characteristics.

Pathological-examination method
All of the specimens were xed with 4% neutral formaldehyde, embedded in para n and stained with hematoxylin and eosin (H&E) for histological observation. Section thickness was 4 μm. We observed the histological morphology of the tumor under an Olympus BX53 light microscope (Olympus, Tokyo, Japan). We used the SP method for immunohistochemical (IHC) staining and the heat repair method for antigen repair. The rst antibodies used included S-100, vimentin, calretinin, epithelial-membrane antigen (EMA), Clusters of Differentiation 34 and 56 (CD34, CD56), Ki-67 labeling index (LI), dendritic-cell sarcoma (DCS), smooth-muscle actin (SMA) and mindbomb E3 ubiquitin protein ligase 1 (MIB1).

Statistical Methods
We used SPSS statistical software version 20.0 (IBM Corp., Armonk, New York, USA) for analysis.
Considering the small sample size, data for patient age and course of disease were represented by the median. We used the Kolmogorov-Smirnov method to test the normal distribution of quantitative data (tumor diameter). Measurement data conforming to a normal distribution were expressed by.

Ethics Statement
This retrospective study was approved by our hospital's institutional review board. We obtained signed consent from each patient.

Ultrasonic-sonogram features
According to their US characteristics, the 36 masses could be roughly divided into 3 categories: (1) Solid, with a uniform elliptical or circular shape and a clear edge. Nineteen masses were solid, of which 16 had an envelope and 3 did not. Eleven of the 19 had uneven low echo; The other 8 masses had a uniform low echo.
(2) Cystic. Five masses fell into this category. The wall of the cyst was smooth and most of the non-echoic areas were in the cyst. A small number of point and occule hyperechoic areas were seen, some accompanied with strip hyperechoic separation, and the echo at the rear was obviously enhanced. (3) Cystic-solid (mixed) type, of which there were 12 masses. Tumor volume was generally large, the mass was round or oval and its membrane was intact. There were gridlike or irregular ake-like anechoic areas in the interior, and the echo in the rear was slightly enhanced. US features of the 36 masses are shown in Table 2.

CT imaging features
Internal mass density on CT plain scan was uniform in 9 masses and uneven in 27. On CT plain scan, density was mainly medium and low, slightly lower than that of muscle. CT value was 15-30 HU. After intravenous injection of contrast agent, the enhanced scan was mainly light and medium inhomogeneous, and enhanced CT value was 20-45 higher than on plain scan. We found heterogeneous reticular enhancement, clear boundary.
Mass size ranged from 1.5 to 14 cm (average, 5.2 ± 0.8 cm). Masses were mainly low and equal signal on T1WI but were dominated by an uneven high signal on T2WI. Tumors showed obvious uniform or uneven enhancement. Most of the capsule was present, and enhancement of the tumor capsule was not obvious on MRI enhanced scan. The long axis of the tumor ran along the direction of the nerve, from the proximal to the distal end; the surrounding tissues were compressed and changed, and the destruction was not obvious.

Pathological features
Most of the masses were round or nodular, tough and with complete capsules. Cystic change and necrosis were seen in 17 masses. Under the light microscope, we saw 2 kinds of tumor tissue component manifestation: the fusiform-sheath tumor cells were in a paliform or interlaced arrangement, and some cells in the Antoni A area, loose distribution of cells in the Antoni B area, vascular proliferation and cyst formation.

Relationship between imaging and histopathology features see
Cystic (5) Grade II (17) Grade III (7) Abbreviations: Y = YES; N = NO II (5) III (7) CT features Pre-contrast: Inhomogeneous Post-contrast enhancement: Internal mild (4) Internal moderate (4) Internal severe (2) Pre-contrast: Homogeneous Post-contrast enhancement: Mild (4) Moderate (3) Severe (2) Pre-contrast: Inhomogeneous (high external/low internal) Post-contrast enhancement: Peripheral mild (1) Peripheral moderate (2) Unenhanced central low-density area (2) Pre  [7]. The tumorigenesis of schwannomatosis is a response to a somatic vehit/three-step mechanism, resulting in the loss of function of genes adjacent to LZTR1 and the contiguous genes of locus 22q11.2q12.2[8]. Paganini et al. [9] immunostained for LZTR1 protein in 22 masses from nine unrelated patients. It is suggested that LZTR1 molecular analysis might help elucidate the molecular characteristics of schwannomatosis patients. Ding et al. [10] suggest that the structural and functional abnormalities of the SMARCB1 gene might be the molecular basis of familial schwannomatosis. If the focus involves peripheral-nerve tissue, it can be accompanied by pain [11]. Jordan et al. [12], using a simple 10-point pain scale (SF-36) to explore the relationship between pain and mutation of the SMARCB1 and LZTR1 genes in schwannomatosis, found that the median pain score of the LZTR1 group was 3.9 and that of the SMARCB1 group was 0.5 (P = 0.0414). Not only was the pain of LZTR1 mutation patients signi cantly higher than that of SMARCB1 mutation patients, but their pain-related quality of life as assessed by SF-36 was signi cantly worse (P = 0.0106). Pain score was correlated with tumor volume (rho = 0.32471, P = 0.0499) but not with number of tumors (rho = 0.23065, P = 0.1696). common. The opposite ends of each mass were connected by nerve bers, and a tail sign (20/36) was found in some masses.

3.2.2CT characteristics
Masses were located in the muscle space; there was a slightly low-density, oval-shaped block shadow running longitudinally; and density could be equal or unequal. Enhancement scanning varied according to the pathological components of the tumor. In type II tumors, CT enhancement masses were almost uniformly enhanced; there was a mild enhancement in four cases, moderate and severe enhancement in 1 case each.
The CT enhancement of type I, III and IV tumors was mainly unequal. Type I showed mainly central enhancement. Type III showed mainly peripheral light-to-moderate enhancement, and two of these ve masses had no enhancement due to the cystic area in the center. All of the type IV tumors showed inhomogeneous enhancement, and their CT values increased by 20-45 HU.

3.2.3MRI characteristics
The main MRI manifestations of masses in schwannomatosis were low and equal signal on T1W1, medium and high signal on T2W1 and mixed high signal on diffusion-weighted imaging (DWI). Edge and internal continuous enhancement can be seen on the enhanced scan, and the boundary between benign tumors and surrounding tissue is relatively clear [19]. The results of this study were similar to those described in the extant literature. MRI is the most effective imaging diagnostic method for intraspinal tumors, as it can clearly distinguish various tissues and structures in the spinal canal and it has certain use for determining the origin, shape, size, quantity and adjacent structures of tumors. It can also be used to guide the formulation of the surgical plan. The whole spine must be scanned via MRI before intraspinal-tumor surgery; doctors must carefully observe the small foci to avoid a missed diagnosis.

3.2.4Advantages and disadvantages of different imaging modalities
Because US examination is inexpensive, very safe, convenient, easy to operate and without risk of radiation damage, it has become the rst choice for screening and postoperative reexamination of schwannomatosis.

However, because of the bone block, US examination is limited in the detection of intracranial and intraspinal
lesions. CT has high spatial resolution and no overlapping of tissue structure images, allowing it to accurately judge the relationship between masses and surrounding tissues, which provides an important basis for clinical surgery; however, its soft-tissue resolution is poor. MRI has the advantages of high softtissue resolution and multi-directional imaging, which can better show the source of the tumor, as well as the relationship between masses and blood vessels/surrounding tissues; it is more conducive to differentiating between benign and malignant tumors, so it has become a prerequisite examination before surgery. Godel et al. [20] compared the volumes of dorsal-root ganglia in 16 patients with schwannomatosis, 14 patients with NF2 and 26 healthy controls by MR neurography. Their study found that dorsal-root ganglion volume was the same in NF2 as in healthy controls, but not in schwannomatosis. Dorsal-root ganglia might be vulnerable sites in terms of origination of are exia and sensory loss, as well as useful diagnostic markers in NF2. MRI examination is costlier than ultrasound and CT, and it is not suitable for patients who have nondemagnetized metal implants or claustrophobia. In addition, in recent years, some researchers have used positron emission tomography (PET) for imaging diagnosis of schwannomatosis [21], but PET imaging might not be a reliable predictor of malignant transformation in schwannoma, which reduces the enthusiasm for tumor surgery without obvious clinical symptoms or signs [22]. This imaging modality is also expensive and di cult to promote.

Pathological features of schwannomatosis
In general, specimens have a capsule, and there are no nerve bers in the tumor body. In terms of path morphology, two types of schwannoma tissue can be seen under the microscope. In type A tissue, tumor cells are arranged in bundles; cells within the tissue are spindle-shaped and arranged parallel to each other into palisades; the nucleus is thin and long; and there are abundant slender reticular bers between the cells.
Type B has few cells in the tissue. The arrangement of tumor cells is disordered and loose, with liquid between cells and bers that can collect in the capsule. There is a large extracellular space between tissues, and abundant capillaries and blood sinuses between cells. Tumor blood vessels often show sinusoidal, spongy, or capillary-like dilation and hyaline change of blood vessel walls. In terms of immunophenotype, schwannomatosis is the same as classic soft-tissue schwannoma. Our 12 cases all showed diffuse positive S-100 protein expression, which could be used as the rst marker. NF was generally not expressed, Ki-67 was below 3%, and patients were mostly positive for CD34 and vimentin.

Relationship between imaging manifestations and pathology
This study showed a certain correlation between tumor imaging and pathological properties in schwannomatosis. Because of the density of the cells in Antoni type A tissue, CT plain scan and US ndings in lesions mainly composed of this tissue showed mostly moderate and slightly low density or echo. The uniformity of the internal density or echo depended on the amount of Antoni type B tissue. The distribution of Antoni type B tissue in the tumor was consistent with that of the low-density and low-echo areas. Antoni B areas easily become cystic, and their imaging features are round low-density areas with smooth inner walls.
The enhancement range of the tumor varied greatly, from light to signi cant; the enhancement could be irregular or more uniform, or there could be interstitial or cystic low density in the enhancement area. This change was consistent with the distribution of multicystic space and hemorrhagic degeneration in Antoni type B tissue, similar to the characteristic manifestation of schwannoma on T2WI: target sign. In this study, 25 masses had enhancement amplitudes >20 HU, 5 of which were schwannomas with active growth. There were abundant blood sinuses in the tumors. The blood-rich Antoni type A tissues were enhanced, while enhancement of Antoni type B tissues was not obvious. In this study, 11 masses with enhancement <20 HU were found to have a large number of collagen bers. Other causes of tumor enhancement included hyaline degeneration, hemorrhage and necrosis. CDFI revealed no signi cant blood ow signal in most lesions, only a small amount of linear color blood ow signal. The reason was that the wall of the tumor vessel was brosis and accompanied by thrombosis.

Differential diagnosis
Neuro bromatosis type 2 NF2 NF2 is an autosomal-dominant inherited disease that is characterized by multiple benign tumors of the nervous system and caused by a gene located on the q12 band of chromosome 22. Bilateral acoustic neuroma, the rst symptom of which is bilateral progressive hearing loss, is the most common clinical characteristic of NF2. Clinical diagnosis of NF2 is based on the Manchester standard [23]. There is a certain overlap between the diagnoses of schwannomatosis and NF2, the characteristics of the 2 conditions differ

3.4Treatment and prognosis
Surgical resection is the most reliable treatment for schwannomatosis. Ten of our 12 patients underwent surgery. The other two declined the operation after communication with us because they had too many tumors and surgery would have been di cult.

Study limitations
The limitations of this study were as follows: (1) This was a retrospective study with a small number of cases. Because some of the diagnoses were con rmed at an early age and/or before the new diagnostic standard came into being [26], not all of the patients underwent lztr1 and smarcb1 gene examination for clinical diagnosis. We cannot guarantee that any of the patients with schwannomatosis did not also have mosaic NF2. Some patients with mosaic NF2 will be included in this diagnosis when they are at a young age. Some patients with schwannomatosis might have unilateral vestibular schwannoma. (2) These patients did not have 3-mm thin slice scans or whole-body MRI examinations, and it is not excluded that some masses were not found in some areas without scanning. (3) It was di cult for some patients to con rm their family histories or for us to obtain samples from their family members for genetic analysis. (4) We did not compare the imaging characteristics seen in this study with those of NF2. (5) Because treatment plan and e cacy evaluation were not the focus of this study, no corresponding discussion and case follow-up analysis was conducted.
Declarations Data Availability Statement Some or all data, models, or code generated or used during the study are available from the corresponding author by request. (liuinsistence@163.com).
Author Funding Not applicable.
Compliance with ethical standards Con ict of interest We declare that we have no nancial and personal relationships with other people or organizations that can inappropriately in uence our work, there is no professional or other personal interest of any nature or kind in any product, service, and/or company that could be construed as in uencing the position presented in this article.