Intramuscular hemangiomas (IMHs) are unique vascular malformations that are benign and most commonly occur in the trunk and extremities. In previous study, only 0.8% of hemangiomas were located in the muscle (1, 2). Of the reported cases, approximately 45% are located in the lower extremities (3). In general, about 14% of IMHs are originated from the muscles of the oral and maxillofacial region (4). In 1982, Mulliken et al. classified vascular lesions based on clinical observation and histopathology into hemangiomas and vascular malformations (5). Hemangiomas that feature with endothelial cell proliferation are true tumors, whereas vascular malformation is not since it simply manifests as abnormal vasculatures. Current opinions suggest that IMHs is more close to a vascular malformation rather than a hemangioma (6). However, IMHs can often be accompanied by endothelial cell proliferation. In 1972, Allen and Enzinger proposed the currently well-known categorization, which divides IMHs into three types based on the size of vessel diameter, namely, venous (diameter >140 μm), capillary (diameter <140 μm), and mixed type (7). Different sizes of blood vessels in the lesions develop different blood supplies, which can be roughly classified into 2 types, the poor (smaller volume and slower blood flow in the lesions) and the rich blood circulation type (larger volume and faster blood flow in the lesions).

IMHs are usually asymptomatic, inconspicuous, and difficult to detect in the early stages. When the tumor grows larger, it may compress and push adjacent nerves and muscles, causing various symptoms. It is reported in the literature that the preoperative diagnosis rate is only 8 to 19% (8). IMHs are often misdiagnosed because of the variable blood supply and uncertain symptoms (9). For IMHs with a rich blood supply, arterial embolization before surgery would be needed to reduce intraoperative bleeding; however, for IMHs with a poor blood supply, simple embolization would be sufficient enough before surgery (10). In addition, IMHs become more difficult to treat as they enlarge.

Evaluation of blood supply of IMHs through preoperative imaging provides practical information for the selection of treatment methods. Different pathological types of IMHs express distinct images. Among all of the publications, results of computed tomography (CT) and magnetic resonance imaging (MRI) depicted IMHs as venous malformations since the non-contrast CT scanning density of the lesions was relatively lower than surrounded muscles and the enhancement are not clear (poor blood circulation type) (4). Meanwhile, a few studies have also shown that IMHs lesions were obviously enhanced in contrast CT and MRI (rich blood circulation type) (11, 12). Currently no study has been proposed to identify the relationship between the 3 pathological types of IMHs and the blood supply status. In addition, due to the various blood supply status, IMHs clinically may be confused with malignant tumors (9). Therefore, familiarity with IMHs imaging features and the preoperative evaluation of the blood supply status are of great significance for the precise diagnosis and treatment of IMHs.

Medical records of 32 patients with oral and maxillofacial IMHs from January 2001 to December 2020 (average age: 36.28 ± 14.49 years), namely, 7 males (21.8%) and 25 females (78.1%), were analyzed in this study. Before surgery, 10 patients received contrast CT scanning, 27 patients conducted contrast MRI scanning, 5 patients received both scanning.

CT scanning was performed with Aquillion ONE (Canon Medical Systems, Otawara, Japan) or Brilliance iCT (Philips Healthcare, Cleveland, OH, USA) with the following parameters: tube voltage of 120 kV/250 mAs, slice thickness of 5 mm, non-contrast scanning, and contrast scanning. CT images with contrast were 1.25-mm reconstructed using Philips Extended Brilliance Workspace software. A nonionic contrast agent (350 mg I/ml) was intravenously injected at a rate of 1.5–2.0 ml/s.

MRI scanning were performed on a GE 750 3.0-T or 1.5-T Sigma Twin Speed MR scanner (General Electric Medical Systems, Milwaukee, WI, U.S.) with a head and neck array coil. In the beginning, subjects were examined by conventional MRI scanning. Scanning sequences and parameters were described as follows: 1) Fast recovery fast spin echo T2WI (FRFSE T2WI) with repetition time (TR) of 4,020–4,820 ms, echo time (TE) of 84.8 ms, for 3 times; 2) Fast spin echo T1WI (FSE T1WI) with TR of 400–600 ms, TE of 8–9.9 ms, for 2 times. 3) Diffusion-weighted imaging (DWI) with using a single-shot spin-echo echoplanar imaging sequence (TR/TE, 2,200/70 ms; matrix, 128 × 128 mm; FOV, 240 × 240 mm; and b = 1,000 s/mm²). ADC maps were generated. 4) a single-voxel 1H-MRS localizer scan was first performed, which involved a single or multichannel surface head coil for T1W or T2W spin-echo sequence transverse scans (T1W: TR/TE, 460/7.9 ms; T2W: TR/TE, 4,700/85 ms).

Dynamic contrast enhanced (DCE)-MRI was performed using a gradient recalled echo–echo planar imaging (GRE) with TR of 400–600 ms, TE of 58–9.9 ms, thickness of 5 mm, gap of 0 mm, FOV of 24–26 cm, and matrix of 256 × 192 mm. The gradient strength was set as 30 Mt/m. The MRI contrast agent, gadopentetic acid (Gd-DPTA), was injected at a rate of 2.5 ml/s and a dose of 0.2 ml/kg. A single scan took 5 s, and the total scan time was 220–280 s.

The section with the maximum diameter of lesions was selected as the central section. The mean lesion ADC was evaluated by manually drawing an ROI on the b = 1,000 diffusion images and copying this ROI to the corresponding ADC map by two radiologists, who had more than 3 years of experience in the MRI of head and neck lesions and were blinded to the clinical outcomes. The ROI of ¹H-MRS was set according to the location and size of the lesion, and a prescan was performed on the ROI, which included receiver gain transference, water center frequency adjustment, ROI shimming, and water suppression. The size and location of the ROIs in multiple ¹H-MRS images were held constant within the same lesion. This procedure was followed by point-resolved spectroscopy (PRESS) (TR/TE = 1,500/144 ms) and eight excitations and a ¹H-MRS spatial localizer scan with a water linewidth <15 Hz. Metabolites, which included choline (Cho), creatine (Cr), inositol (MI), water, and N-acetyl-aspartate (NAA), were automatically integrated using PROBE-SV software. A positive result was defined as the detection of a Cho peak (3.2 ppm) in ¹H-MRS; a negative result was defined as no detection of a Cho peak.

Clinical variables including gender, age, location of lesions, treatment, recurrence and imaging variables such as the visibility of phleboliths or vermicular vessels, the lesion morphology on MRI, and MRI signal characteristics of each IMHs patients were collected and summarized to evaluate the correlation with pathological types.

IMHs consist of vascular and nonvascular components; the latter includes fat, fibers, smooth muscles, etc., with fat as the largest element (13). However, there was no obvious decrease in the signal of fat-saturation sequences of all the observed oral and maxillofacial IMHs patients in this study. This phenomenon was different from the IMHs in other body parts. Pathologically, cavernous type of IMHs accounted for the majority and was expressed similar to venous malformations while the other 2 types were revealed as rich blood supply lesions. Most of the studies suggested that the imaging findings of oral and maxillofacial IMHs are consistent with venous malformations (6, 14–17), and only a few studies indicated that the signal of IMHs lesions can be greatly enhanced (11, 12); however, none of them neither has put forward the concept of rich blood supply status nor combined the enhancement with pathological types of IMHs for analysis. Through the comprehensive analysis of CT and MR images, we found that the capillary and mixed type of IMHs match with the characteristics of rich blood supply lesions, while cavernous IMHs were consistent with previous study results as relatively poor blood supply lesions.

Although there were recurrences in 4 IMHs patients, it is difficult to clearly explain the relationship between recurrence and pathological types, morphological features, or treatment. Even with the confirmation of the lesion boundary by MRI, it was still difficult to remove the lesion completely. Currently we are working on the improvement of IMHs treatments. Distinct from the traditional method, the new treatment confirmed both lesion boundary and blood supply status, through MRI. Patients with a rich blood supply received TAE then flap operation to expose the lesion, followed by electrocoagulation to cause malformed vascular fibrosis. Contrarily, patients with a poor blood supply were directly injected with sclerosing agents if the lesions are small. For diffuse lesions, electrocoagulation can be performed without TAE. This treatment method can reduce the recurrence rate due to incomplete resection and retain as much of the soft tissue as possible. Therefore, an accurate preoperative evaluation of the blood supply and lesion boundary of oral and maxillofacial IMHs is of great value for guiding clinical treatment (18).

This study showed that all pathological types of IMHs mainly yielded an intermediate signal intensity on T1WI, and most of the patients had slightly hyperintense on T2WI (only 2 cases of cavernous IMHs were hyperintense). This phenomenon might be due to the larger vessel diameter in cavernous IMHs causing larger amount of free fluid in blood vessels. On the contrary, smaller vessel cavities were observed in capillary type of IMHs, which might be associated with the slightly high signal intensity on T2WI.

IMHs may involve a single muscle (localized) as well as multiple muscles including masticatory and expression muscles (diffuse). We identified that localized oral and maxillofacial IMHs were more likely to onset at tongue and masseter. The 5 misdiagnosed cases were all localized IMHs patients. The reason for the misdiagnosis may be because diffuse IMHs morphologically have larger difference from tumors, which are easier to be diagnosed; localized IMHs have clearer boundary of lesions with intermediate intensity on T1WI, slightly high intensity on T2WI, and even clearer with contrast, which makes it more likely to be confused with mesenchymal tumors. Extra functional MRI (fMRI) was performed on these 5 misdiagnosed patients, which indicated that imaging doctors have tried to distinguish the malignancy of lesions through TIC, ADC or MRS. Although the average ADC of the 5 patients was 1.50 × 10⁻³ mm²/s, of 3 cases were not high enough for distinguishing between localized IMHs or low-grade malignant tumors. TIC results showed that 4 cases were type I, and 1 case was type II. Unfortunately, considering the variance of blood supply in different IMHs, TIC was not able to provide stable diagnostic reference. Only two misdiagnosed patient has received MRS, and Cho peaks were not observed in both cases. Obvious Cho peaks usually can be seen in sarcomas at the site of 3.2 ppm of the H spectrum (19). Therefore, we suggested that the presence of Cho peaks in MRS might be an important criterion for differentiating localized IMHs from low-grade malignant mesenchymal tumors in the maxillofacial region. We need further support from a large number of clinical trials.

Among the 4 patients with embolization before resection, 3 had vermicular vessels in the lesions; however, among the 28 patients who received resection directly without embolization, only 7 (25%) had vermicular vessels in the lesions. We inferred that surgeons tended to determine whether preoperative embolization was necessary based on the presence of vermicular vessels. The percentage of having vermicular vessels was apparently different among the 3 IMHs types, which were 17.6, 33.3, and 66.7% in cavernous IMHs, capillary IMHs, and mixed IMHs cases, respectively ( Table 1 ). This observation suggested that the probability of observing vermicular vessels was higher in capillary and mixed IMHs images than that of in cavernous IMHs images. However, it is still too risky to use presence of vermicular vessels alone as a judgement for non-cavernous IMHs since it was also identified in 3 cavernous IMHs cases in this study. No obvious difference was found in the presence of phleboliths among 3 pathological types in this study, except for cavernous IMHs cases. Two mixed IMHs cases had phleboliths as well. The fMRI might be more reliable than morphological features such as presence of vermicular vessels or phleboliths for determining the blood supply status of IMHs. Combined evaluation of presence of vermicular vessels and fMRI might be more accurately for determining the IMHs.

Our findings were inherently restricted to small sample size due to the rarity of the maxillofacial IMHs, we only performed a meaningful analysis on the fMRI analysis in early stage. Although we need prospective surgical pathologic analysis to confirm these findings, it still provides useful clues for unveiling the clinical mechanism of IMHs. From the imaging and pathological findings of IMHs, we suggested that MRS and TIC might be able to provide more information on diagnosis of IMHs and determination of blood supply status.

In conclusion, multi-muscular diffuse IMHs are relatively easy to be diagnosed, and single muscular localized IMHs, with the lack of enlarged venous reflux vessel and the obvious border of mass-like lesions, are tended to be confused with low-grade malignant mesenchymal tumors. Presence of Cho peaks in MRS can be used as a potential marker for differential diagnosis. Treatment of IMHs is highly related to its blood supply status, and simply uses the presence of vermicular vessels for determining blood supply status seem to be far from enough. TIC may be used as criteria for identifying IMHs with rich blood supply. Understanding the morphological and functional imaging features will be a key role in determining the clinical treatment plan for patients with IMHs.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by the 9th People’s Hospital affiliated to Shanghai Jiao Tong university School of Medicine. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin. Written informed consent was obtained from the individual(s), and minor(s)’ legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

DZ and JW drafted the work or revised it critically for important intellectual content. CZ provided pathological information. XT is responsible for the final approval of the version to be published. LZ is responsible for the substantial contributions to the conception or design of the work. All authors contributed to the article and approved the submitted version.

This study has received funding by the research of the “Clinical +Plan” Project of the Shanghai Ninth People’s Hospital, Shanghai JiaoTong University School of Medicine (JYLJ201918), and the Technology Transfer Project of Shanghai Jiao Tong University School of Medicine (ZT202108).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.