Systemic therapies in advanced epithelioid haemangioendothelioma: A retrospective international case series from the World Sarcoma Network and a review of literature

Abstract Background This observational, retrospective effort across Europe, US, Australia, and Asia aimed to assess the activity of systemic therapies in EHE, an ultra‐rare sarcoma, marked by WWTR1‐CAMTA1 or YAP1‐TFE3 fusions. Methods Twenty sarcoma reference centres contributed data. Patients with advanced EHE diagnosed from 2000 onwards and treated with systemic therapies, were selected. Local pathologic review and molecular confirmation were required. Radiological response was retrospectively assessed by local investigators according to RECIST. Progression free survival (PFS) and overall survival (OS) were estimated by Kaplan‐Meier method. Results Overall, 73 patients were included; 21 had more than one treatment. Thirty‐three patients received anthracyclines regimens, achieving 1 (3%) partial response (PR), 25 (76%) stable disease (SD), 7 (21%) progressive disease (PD). The median (m‐) PFS and m‐OS were 5.5 and 14.3 months respectively. Eleven patients received paclitaxel, achieving 1 (9%) PR, 6 (55%) SD, 4 (36%) PD. The m‐PFS and m‐OS were 2.9 and 18.6 months, respectively. Twelve patients received pazopanib, achieving 3 (25%) SD, 9 (75%) PD. The m‐PFS and m‐OS were.2.9 and 8.5 months, respectively. Fifteen patients received INF‐α 2b, achieving 1 (7%) PR, 11 (73%) SD, 3 (20%) PD. The m‐PFS and m‐OS were 8.9 months and 64.3, respectively. Among 27 patients treated with other regimens, 1 PR (ifosfamide) and 9 SD (5 gemcitabine +docetaxel, 2 oral cyclophosphamide, 2 others) were reported. Conclusion Systemic therapies available for advanced sarcomas have limited activity in EHE. The identification of new active compounds, especially for rapidly progressive cases, is acutely needed.


| INTRODUCTION
Compared to other vascular sarcomas, epithelioid haemangioendothelioma (EHE) is marked by peculiar clinical features, defined molecular characteristics and specific treatment challenges. 1 EHE is rarer than angiosarcoma or other vascular sarcomas (incidence rate <1/1,000,000), is more common in females and is often multifocal and/or multicentric at presentation, with lung, liver and bone being the typically involved sites. 2 Despite its rarity, it encompasses a wide heterogeneity in clinical behaviour, with cases naturally stable over time as opposed to others which are slowly progressive. A rapidly evolving third variant does exist, which behaves as a high-grade sarcoma. Similarly, the symptoms burden in this disease can be variable, but overall, a significant impact on quality of life and psychological distress have been recently reported in EHE patients. 3 Serosal effusion and bone metastases have been reported as adverse prognostic factors. 2,4,5 From the molecular point of view, the WWTR1 (WW Domain Containing Transcription Regulator 1) -CAMTA1 (Calmodulin Binding Transcription Activator 1) and YAP (Yes-associated protein 1)-TFE3 (Transcription Factor Binding To IGHM Enhancer 3) rearrangements are found in approximately 90% and 10% of all cases, respectively. 6,7 Recently, gene fusions involving WWTR1 with a partner different from CAMTA1 have also been reported. 8 The detection of these rearrangements is today a hallmark in diagnosis, all the more in those cases marked by aggressive morphological features, where the differential diagnosis with angiosarcoma can be challenging, or atypical presentations. Both WWTR1 and YAP are downstream effectors in the Hippo pathway, which could be therefore involved in the pathogenesis of this rare condition. 9 In patients with asymptomatic, naturally stable, advanced EHE, watchful waiting is today a reasonable approach, as prolonged stabilities over time or even spontaneous regressions have been reported. [10][11][12] Conversely, for symptomatic patients or patients with progressive multifocal or multicentric disease, systemic treatments are the standard of care.

FREZZA Et Al.
Liver transplant in patients with multifocal disease confined to the liver has also been investigated, but results are controversial. 11,12 Data currently available on systemic agents are limited to case reports or small single-institution case series. Anecdotal responses have been reported with pazopanib, which is currently approved in the treatment of advanced soft tissue sarcomas refractory to anthracycline, as well as with other anti-angiogenic compounds including sorafenib, bevacizumab and apatinib. [13][14][15][16][17][18] mTOR inhibitors (i.e. sirolimus) can control the disease in slowly progressive cases, whereas their activity seems to be limited in more aggressive variants, such as those presenting with serosal effusions. 5,19,20 In the absence of any alternative potentially active treatment approved other that pazopanib, EHE patients with advanced progressive disease are treated with conventional chemotherapy, which is assumed to have very limited activity. 13,21 In this international, observational retrospective effort, we collected all the cases of histologically proven, advanced EHE, diagnosed from 2000 onwards and treated with systemic therapies at the major sarcoma reference centres within the framework of the World Sarcoma Network, with the aim of studying their potential. The results are reported herein.

| Patient population
We considered all patients of any age with advanced EHE, diagnosed from 2000 onwards, and treated with systemic therapies (anthracycline-based regimens, weekly paclitaxel, pazopanib, INF-α 2b or other), in front or further line. Patients treated sirolimus were not included in this series, as most of them are the subject of a separate report. 5 The positivity of either WWTR1-CAMTA1 or YAP1-TFE3 rearrangement, determined by FISH or by positive immunostaining for CAMTA1 or TFE3, was required. Patients were treated at 20 sarcoma reference centres in EU, US, Australia, and Japan, within the World Sarcoma Network. Written informed consent to the treatment was obtained as required by local regulation. Approval of this retrospective case series analysis by the Institutional Review Board of each participant institution was required.

| Study design and data collection
Data were extracted from clinical databases and confirmed through a review of patient files. A questionnaire was circulated among participating institutions, in order to explore the frequency of radiological assessment for patients on treatment, at participating centres. All institutions evaluated disease response radiologically every 2-4 months. All participating sites were asked to provide a retrospective local assessment of treatment response based on RECIST version 1.1. 22 Data on radiologic evidence of any disease progression prior to commencing treatment were also collected for all cases with available scans.
Together with data collection, a literature search was conducted. We searched for articles written in English, published in PubMed from 1995 until September 2020 with the terms 'epithelioid haemangioendothelioma', 'chemotherapy', 'anthracycline', 'paclitaxel', 'pazopanib', 'antiangiogenic', 'treatment'. The final reference list was defined on the basis of originality and relevance to the scope of this article.

| Statistical analyses
Descriptive statistics and frequency tabulation were used to summarize patient and tumor characteristics. Overall response rate (ORR) was defined as the proportion of patients who achieved complete response (CR) or partial response (PR) by RECIST 1.1. The corresponding 95% confidence intervals (CIs) were calculated based on the binomial distribution.
Progression-free survival (PFS) and overall survival (OS) were estimated using the Kaplan-Meier method, and survival distributions by group were compared using the log-rank test. PFS was calculated as the interval from the start of treatment to the date of the first documented evidence of progressive disease (PD) or death due to any cause or to the date of the last follow-up. OS was calculated as the interval from the start of treatment to the time of death from any cause or to the date of the last follow-up. A two-sided p < 0.05 was considered statistically significant.
Statistical analyses were carried out with SAS (version 9.4, SAS Institute, Cary, NC, USA) and R software (version 3.6.1, R Foundation for Statistical Computing, Vienna, Austria).

| Patient population
Seventy-three patients with advanced EHE treated with systemic therapies were identified. The median follow-up was 35.8 months (interquartile range, IQR, 17.5-93.3) and the median OS was 17.4 months (IQR, 9.3 -not evaluable, NE). Thirty-three patients received anthracycline-based regimens, 11 weekly paclitaxel, 12 pazopanib, 15 INF-α 2b and 27 other agents. Twenty-one patients received more than one of the selected treatments, and therefore are included in more than one group. The population characteristics are summarized in Table 1.

| Anthracycline-based group
Thirty-three patients were included, all evaluable for response. The median follow-up was 33 months (IQR, 19.4-89.5). In 19 patients (58%) there was evidence of PD prior to commencing treatment, in 14 (42%) there was not. Fifteen patients (45%) received anthracycline as single agent, 9 (27%) in combination with ifosfamide and 9 (27%) in combination with different compounds. Twenty-nine patients (88%) received anthracycline as a first-line, 4 (12%) as a second line treatment. At the time of the present analysis, 32 patients completed their treatment. Treatment was discontinued for PD in 15 (47%) cases, for toxicity in 2 (6%), for maximum cumulative anthracycline dose in 14 (44%), and for other reasons in 1 (3%).

| Weekly-paclitaxel group
Eleven patients were included, all evaluable for response. The median follow-up was 80 months (IQR, 27.2 -93.3). In 5 patients (45%) there was evidence of PD prior to commencing treatment, in 6 (55%) there was not. Eight patients (73%) received weekly paclitaxel as a first-line, 3 (27%) as a second-line treatment. At the time of the present analysis,  all patients completed their treatment. Treatment was discontinued for PD in 7 (64%) cases, for toxicity in 1 (9%), for agreement on a treatment holiday in 2 (18%), and for other reasons in 1 (9%). The best RECIST response with weekly paclitaxel was 1 (9%) PR, 6 (55%) SD, and 4 (36%) PD. The ORR was 9% (95%CI: 0%-41%). The patient who achieved a response was a 38-year old male, with a YAP1-TFE3 positive EHE of the lungs with pleural and lymph nodal involvement, treated with weekly paclitaxel as a first-line treatment. Details on the evidence or radiological disease progression prior to commencing treatment are reported in Tables 1 and  2.
The best RECIST response with pazopanib was 3 (25%) SD, and 9 (75%) PD. Details on the evidence or radiological disease progression prior to commencing treatment are reported in Tables 1 and 2. The m-PFS in this group was 2.9 (IQR, 2.1 -7.1, Figure 1C), and the m-OS was 8.5 (IQR, 3.4 -NE, Figure 2C). The 2-year PFS and OS rates were 17% (95% CI: 5-59%) and 38% (95% CI: 17-84%). No significant differences were noticed between patients with and without evidence of previous progression in terms of m-PFS (2.9 months and NE, respectively, p = 0.47) whereas a trend toward a better m-OS (8.5 months and NE, respectively, p = 0.86) was reported. clinical choice after prolonged stability in 2 (13%) and for drug unavailability in 2 (13%). The best RECIST response with INF-α 2b was 1 (7%) PR, 11 (73%) SD, and 3 (20%) PD. The ORR was 7% (95%CI: 0%-32%). The patient who achieved a PR was a 48-year old male, with a WWTR1-CAMTA positive EHE involving liver and lungs treated with INF-α 2b as a first-line treatment. Details on the evidence or radiological disease progression prior to commencing treatment are reported in Tables 1 and  2.
Kaplan-Meier curves for PFS and OS on anthracyclinebased regimes, paclitaxel, pazopanib and interferon are reported in Figures 1 and 2, respectively. Results on treatment and outcome for each regimen, and association with prior evidence of PD, are summarized in Table 2.

| DISCUSSION
This academic, multi-institutional, international, retrospective study collected the largest series ever of patients affected by advanced EHE treated with systemic therapies. Seventythree patients (33 treated with anthracycline-based regimens, 11 with weekly paclitaxel, 12 with pazopanib, 15 with INF-α 2b and 27 with other agents were included. Anthracyclinebased regimens, the standard first-line therapy in advanced soft tissue sarcomas, showed a minor activity in advanced EHE (ORR of 3%, m-PFS of 5.5 months). Similarly, a limited activity was seen with weekly-paclitaxel (ORR of 9%, m-PFS of 2.9 months) and pazopanib (no responses, m-PFS of 2.9 months), while INF-α 2b, which is currently not approved in the disease, resulted in an ORR of 7% and a m-PFS of 8.9 months. Our study is limited by its retrospective nature, both in the collection of clinical data and assessment of tumor response. The retrospective design limited a robust assessment of disease progression prior to commencing treatment, which would have been extremely valuable given the unpredictable natural history of EHE, and did not allow a reliable collection of data on the impact of systemic therapies in symptoms control and quality of life. Also, the number of patients included in each treatment group is limited. Despite these limitations, this is the largest series of EHE published so far and it was     confirmation of diagnosis by an expert sarcoma pathologist, together with the determination of CAMTA-1 / TFE-3 status is a major strength of this study. Unfortunately, we could not observe any relevant antitumor activity from any of the agents used. Notably, the m-PFS for most systemic therapies in this series was lower than 6 months. This points to a patient population selected for an aggressive behaviour at the time of commencing systemic therapy. This is in line with the tendency of all expert centres to delay any treatment until the development of clearcut symptomatic or radiological progression. Indeed, in the absence of any approved treatment option, patients with advanced, progressive EHE are currently managed predominantly with conventional chemotherapy and pazopanib, similar to most advanced STS subtypes. The results of previous studies on systemic treatments in advanced EHE are summarized in Table 3.
Actually, anthracyclines, as single agents or in combination with alkylating agents such as ifosfamide, are the standard first-line treatment in advanced STS, with an ORR reported in the range of 20% and a m-PFS in the range of 6 months across different studies, which is regarded as clinically meaningful in this setting, taking into consideration the biological aggressiveness of most STSs and their tendency toward a rapid progression. [23][24][25] In advanced EHE, data on the activity of anthracycline-based regimens are limited. To the best of our knowledge, only 2 responses have been reported with doxorubicin in 1997 and 2010, in two metastatic EHE patients, however, the diagnosis was not confirmed by either molecular testing or IHC. 26, 27 Cioffi A. et al. did not record any radiological responses in 16 advanced EHE patients treated with anthracyclines, with a m-PFS of 4.8 months. 19 Similarly, no PR were reported by Yousaf N. et al and in most case reports available. 13 Consistent with these findings and on a larger scale, our series confirmed the very limited activity of anthracycline in advanced EHE, with only 1 PR observed (ORR: 3%) and a m-PFS of 5.5 months. Four patients were treated with liposomal doxorubicin, all with prior evidence of PD, and a PR and 2 SD were observed. In the literature, 3 case reports highlighting a PR and a prolonged disease control in advanced, rapidly progressive EHE treated with liposomal doxorubicin, are available. All this would suggest a possible role for this agent in the management of progressive EHE. [28][29][30] Paclitaxel has a recognized activity in vascular sarcomas, especially in angiosarcoma, in which an ORR in the range of 20-50% across different series has been reported. 31,32 In advanced EHE, no responses to paclitaxel have been described in the few case-reports available, nor in the retrospective series by Yousaf N. et al, where 6 patients received paclitaxel across different lines with a median treatment duration of 3 months. 13 Consistently, only 1 PR was observed (ORR=9%) over 11 patients treated with weekly paclitaxel in this series, predominantly (73%) as a front-line, with a m-PFS of 2.9 months.
Responses to pazopanib in EHE have been described in 4 different clinical reports on previously progressive patients, among which a retrospective study from EORTC exploring the activity of pazopanib in a series of vascular sarcomas. 14,15,33,34 In this last study, 1 CR and 1 PR were observed over 10 advanced EHE treated. 15 Conversely, no response was seen in the single case described by Yousaf N et al. 13 Similarly, no responses were seen in our study of 12 EHE patients who received pazopanib, mostly in first and second line, with a short m-PFS (2.9 months). Notably, in these previously published studies molecular confirmation of diagnosis is lacking.
The activity of INF-α in advanced EHE has already been reported, being used both in the paediatric and adult population. [35][36][37] The agent was used assuming an anti-angiogenic effect. Though in our series only 1 PR was observed in 15 patients, m-PFS was longer than any other agent, e.g. 8.9 months. Taken together, available results seem to confirm the potential role of interferon in the management of this rare sarcoma subtype, though the drug is not approved for this indication.
Similarly, sirolimus is also potentially active in the treatment of advanced EHE, with a reported ORR of 10% and a m-PFS of 13 months in previously progressive patients. 5 There are some patients with an aggressive behavior, often marked by a serosal effusion, who respond less to therapy. 5 In this series, we did not to include patients treated with sirolimus, as they are the subject of a separate paper.
In brief, it is clear from this series, and the literature, that we lack any standard medical therapy in EHE. In particular, active drugs in STS are essentially inactive in EHE. An old agent such as IFN has some activity but is not approved for this disease. The same is true of other agents such as mTOR inhibitors, for which retrospective and anecdotal evidence points to slightly less than one half of previously progressing patients free of progression at two years. 5,20 Undoubtedly, the natural history and uncertain prognosis of these patients make clinical studies very difficult, while a proportion of them does not require any treatment. Indeed, controlled studies are virtually impossible due to the rarity of disease, but a distinct number of patients present with an aggressive clinical behavior and are in need for front-line active treatments. 5 Clearly, any prospective study on any new therapy would need a clear background landscape in terms of definition of the disease and its natural history. To this aim, a global consensus development process is in place. At least, it may clarify which subgroups of patients are more in need for any medical therapy and which is their expected prognosis. Indeed, the biomolecular profile of the disease is now better understood and at least is able to single out true EHE from other vascular sarcomas, primarily angiosarcoma. Failing to | 2657 FREZZA Et Al. do so in the past was a major obstacle even to developing new therapies. Further understanding of the molecular pathogenesis might obviously help find new therapeutic targets. The Sarcoma Alliance for Research through Collaboration (SARC) have conducted a phase 2 trial of trametinib in unresectable/ metastatic EHE. Other global efforts are ongoing, and it is worth highlighting that they involve patient groups.

ETHICS STATEMENT
Approval of this retrospective case series analysis by the Institutional Review Board of each participant institution was required.