Clinical Characteristic and Prognosis of Epithelioid Hemangioendothelioma: 35 Cases from Single Center in Recent Decade

Backgrounds: Epithelioid hemangioendothelioma (EHE) is a rare carcinoma worldwide with low-grade malignancy. Few reports have evaluated large case series of EHE. And we tried to describe different treatment and the overall survival rate of EHE, exploring prognostic the factors of outcome. Methods: This was a retrospective study enrolling patients diagnosed with EHE in our center from June 2009 to March 2020. Data including demographic characteristics, laboratory date, treatments, imaging data, immunohistochemical results, follow-up results were collected, and a retrospective database was constructed for analysis. Results: This study enrolled 35 patients with EHE pathologically diagnosed in our center. We reported the mean onset age was 41 years old, ranging from 3 to 70. EHE was more common in female patients (60%). The most commonly affected organ was liver (63%). The 5-year survival rate was 62.86%. After we operated COX regression analyse to test the effect of age ( ≥ 55 years old or <55 years old), sex, position (multiple organs or single organ involved), symptoms (symptomatic or asymptomatic) and Ki-67 ( ≥ 10% or <10%) on disease outcome, we only found that Ki-67 was the independent factor affecting the prognosis, with signicant P value equal to 0.034 and hazard ratio equal to 5.809. Conclusions: In conclusion, EHE has relatively low-grade malignancy and its 5-year survival rate is 62.86%. EHE patients whose Ki-67 ≥ 10% tended to experience poor outcome.


Introduction
Epithelioid hemangioendothelioma (EHE) is a rare carcinoma worldwide with low-grade malignancy. The incidence of EHE is around one in a million and it represents less than 1% of all the vascular tumors [1]. As it is reported, the median age of onset is 36 years old, with a male to female ratio of 1:4. And there has also several reports about children and the elderly. Reported sites of EHE involves lung(12%), liver(21%), liver plus lung (18%), and bones(14%), although this kind of tumor may involve any site of body to be its target [2][3][4]. As for the prognosis, a survey of 264 EHE patients reported that the overall survival was 73% at 5 years [5].
Due to the rarity of EHE, there is still no optimal treatment strategy. Hence, we enrolled patients pathologically con rmed with EHE in our center in recent decade to describe different treatment and the overall survival rate. And we tried to explore the prognostic factors in outcome.

Material And Methods
This was a retrospective study enrolling patients diagnosed with EHE in our center from June 2009 to March 2020. All patients underwent lesion resect or puncture biopsy procedure and were con rmed EHE by pathological biopsy. If a patient had multiple lesions on several organs and was con rmed EHE on one organ, after we reviewed the imaging data, we were able to conclude clinically that the lesions on the other organs were EHE.
We collected patients' demographic characteristics data like age, height, weight, body mass index (BMI), the duration of course and history of some underlying diseases such as hypertension and diabetes. Laboratory data like the concentration of hemoglobin was also recorded to determine if the patient had anemia, which is an indicator of general health condition. The imaging data like CT, PETCT or MRI scan was also retrieved if they had the examination. We recorded some lesion-related data, for example the position, the size, and the number. The immunohistochemical results of pathological biopsy were gathered to identify some speci c markers like Ki-67 labeling index.
We completed the follow-up on the phone call or at our outpatient building and inquired about patients' health condition and some other subsequent treatment. The survival period started from the time at diagnosis to the date of death or last clinical follow-up. We de ned endpoint event as death or discharge without medical order (usually in poor health condition) or lost to follow-up. Progressed disease was considered if a patient had recurrence of EHE on the same organ or we identi ed new metastatic lesion on distant positions after treatment.
The main outcome of this study was to identify the effect of some factors on prognostic outcome and to describe the overall survival rate of EHE. All the data analysis is operated on SPSS 25.0. Continue variables are described as mean ± standard deviation (SD) and category variables as composition.
Kaplan-Meier survival curve is utilized to represent the overall survival rate of EHE in the whole cases and in subgroup patients. The difference of overall survival rate between subgroups is examined by Log-Rank Test. We also operate COX regression analyse to identify independent factors affecting the prognostic outcome of EHE.

Results
We collected 35 patients con rmed EHE eligible for our study. The mean age was 41 years old, ranging from 3 to 70. Male and female incidence ratio was 1:1.5. At the time of the last follow-up, the lesion was mainly located at liver (n = 22) and lung (n = 14). Thirteen patients were found the lesions at multiple places, with lung and liver (n = 7) accounting for the most part. Even for the lesion at only single place, there could be multiple nodules. Nearly half of the patients were conscious of the disease when they underwent physical examination, which meant they did not present with any symptom or discomfort.
Clinical and laboratory data were detailed in Table 1. Among the 35 cases, 6 underwent observation treatment. Three of them lost to follow-up and another 2 patients, who were not in good condition and refused to receive any further therapy, nally discharged from our hospital without medical order. Nine of the whole 35 patients received palliative treatment, with  and 62.86%, respectively. (Fig. 2) According to the main complaint at the diagnosis of EHE, we found that the survival curve was poorer in symptomatic patients than that in asymptomatic ones. However, the difference was not signi cant. (P = 0.252) (Fig. 3A) The situation was similar when we divided patients into male and female. (Fig. 3B) As for the pathological biopsy, 26 samples had Immunohistochemistry results about Ki-67 and we classi ed them into two groups, ≥ 10% and <10%. After we utilize Kaplan-Meier Survival Curve to calculate the overall survival rate, we found patients with Ki-67 ≥ 10% experienced worse consequence than that with Ki-67<10%. P value was 0.016 and the difference is statistically signi cant. (Fig. 3C) Finally, we operated COX regression analyse to test the effect of age (≥ 55 years old or <55 years old), sex, position (multiple organs or single organ involved), symptoms (symptomatic or asymptomatic) and Ki-67 (≥ 10% or <10%) on disease outcome. We only found that Ki-67 was the independent factor affecting the prognosis, with signi cant P value equal to 0.034 and hazard ratio equal to 5.809.

Discussion
In this study, we collected 35 pathologically con rmed EHE patients and analyzed the impact of several factors on outcome. EHE was initially described by Dail and Liebow in 1975 as intravascular sclerosing bronchioalveolar tumor [6]. In 1982, Weiss and Enzinger proposed the name of EHE, whose malignant degree is between hemangioma and angiosarcoma [7]. EHE is prone to occur in soft tissue and multiple organs, and originates from vascular endothelial or pre-endothelial cells.
Studies had reported several factors affecting prognosis: multiple organs involvement, disease progression, more than 55 years old, male patients, patients with obvious symptoms of vessels invaded (for example: hemoptysis and anemia), pleural involvement with pleural effusion and Ki67 greater than 10% [12,13]. EHE has no speci c clinical manifestations. All symptoms are related to the sites of tumor. Of the cases we collected, 22 patients had corresponding clinical manifestations, including cough, chest pain, backache abdominal pain and some other symptoms. The study showed that patients with clinical manifestations had poorer prognostic outcome than those without. Although there was no statistical difference in the p value, there was a tendency for symptomatic patients to be worse than asymptomatic patients. This was similar with the study by Satoshi Shiba [14], however, the author found the difference to be signi cant. We suggest that larger samples of EHE are needed to analyze the relationship between the main complaint and the overall survival rate.
Previous study considered Ki-67 ≥ 10% to be an important factor with poorer prognosis in angiosarcoma, which originates from vascular endothelial cell and is similar with EHE [15]. We also compared the overall survival rate between EHE with Ki-67 ≥ 10% and Ki-67<10%, and we found that the difference was statistically signi cant, which meant EHE patients with Ki-67 ≥ 10% had worse outcome than those with Ki-67<10%. It was frustrating that only 26 samples had Ki-67 labeling index, and as a consequence, the role of Ki-67 needed to be further explored.
If feasible, surgical removal is the best curative strategy. For pulmonary EHE, surgery can be proposed in cases of unilateral single or multiple nodules. Lung transplantation should be evaluated in patients with vascular aggressivity and pleural effusion. Besides, unresectable hepatic EHE without extrahepatic metastases is an excellent indication for liver transplantation [16]. We had a special patient. She was diagnosed as multiple hepatic EHE with main complaint of epigastric pain at her age of 34 and had the rst liver transplantation in December 2013. After 64 months, tumor recurred and she had a second liver transplantation. Now she is having a regular follow-up, and taking immunosuppressants such as Sirolimus, and Everolimus. These pills are the mTOR inhibitors which have been proved to inhibit the growth of tumor cells [17]. Previous studies had described the use of the mTOR inhibitor Everolimus in combination with Sirolimus after transplantation, which not only suppressed the immune response, but also effectively prevented tumor recurrence and improved survival after liver transplantation [18][19][20][21]. This case explains that repeated liver transplantation is a feasible and effective treatment strategy for livercon ned EHE [17,[22][23][24]. On the other hand, follow-up is also considered a reasonable strategy for asymptomatic patients with diffuse lesions due to the low degree of malignancy of EHE, as spontaneous degeneration of EHE has also been reported [25].

Conclusions
In conclusion, EHE has relatively low-grade malignancy and its 5-year survival rate is 62.86%. EHE patients whose Ki-67 ≥ 10% tended to experience poor outcome.

Consent for publication
Written informed consent for publication was obtained from all participants.

Availability of data and materials
The datasets used or analysed during the current study are available from the corresponding author on reasonable request.