Surgical and multimodality treatment of cardiac sarcomas: A systematic review and meta‐analysis

Primary cardiac sarcomas (PCSs) are an extremely rare and aggressive type of malignancies that have been described only by a limited number of observational studies. This study aimed to evaluate the currently existing evidence comparing surgical to multimodality treatment of PCS.


| INTRODUCTION
Primary cardiac neoplasms are a rare entity with an incidence of 1.38 per 100 000, 1 thereby being far less common than metastatic cardiac tumors. Malignancies account for a quarter of primary cardiac neoplasms, with sarcomas comprising between 75% and 95% of these. 2,3 Usually found within the right atrium, primary cardiac sarcomas (PCSs) derive from mesenchymal cells and aggressively impact younger patients with a dismal prognosis. 4,5 Presenting aggressively but initially asymptomatically, PCS can advance rapidly into a plethora of manifestations such as pulmonary hypertension and edema, congestive heart failure, chest pain, dyspnea, pericardial effusions, arrhythmias, and systemic symptoms of malignancy. 5,6 There exists an absence of conclusive clinical evidence for the optimal management of PCS due to the rarity of the disease, and, therefore, multimodal management has been largely guided by retrospective studies limited by small patient numbers and it is often principled by the therapeutic strategies for sarcomas of noncardiac origin. When feasible, complete surgical resection is central to the therapeutic strategies for PCS, as conservative therapy without surgery, radiotherapy, or chemotherapy has shown only a 10% survival rate at 9-12 months. 4 However, even the surgical management of PCS is largely provisional with inconclusive efficacy as it is plagued by dismal multifactorial complications and contraindications, such as severe postsurgical morbidity, potential unresectability of the sarcoma and metastatic spread, and concerns of relatively insignificant increases in patient survival with a disproportionately greater fall in the quality of life. 5 Complete surgical resection can also be accompanied by subsequent radiotherapy and chemotherapy, but despite this, PCS patients continue to have a poor prognosis with a median survival of less than 1 year. 4,5 We aim, for the first time, to assess the currently available evidence on the surgical and multimodality treatment of PCS.  Figure 1). The search terms used were: ("cardiac sarcoma" OR "heart sarcoma" OR "primary cardiac sarcoma" OR "cardiac malignancy" OR "heart malignancy") AND ("surgery" OR "cardiac surgery" OR "cardiothoracic surgery" OR "surgical treatment" OR "surgical resection" OR "resection" OR "multimodality treatment" OR "Treatment" OR "radiotherapy" OR "chemotherapy"). Further articles were F I G U R E 1 PRISMA flow chart. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses identified through the use of the "related articles" function on MEDLINE and a manual search of the reference lists of articles found through the original search. The only limits used were the English language and the mentioned time frame. Patient consent and Institutional Review Board (IRB) approval were not necessary for this study as no patients were deployed.

| Study inclusion and exclusion criteria
Both randomized and observational studies of patients undergoing surgical or multimodality treatment for PCSs were included. Studies were excluded from the review if (1) inconsistencies in the data precluded valid extraction, (2) all patients included in the studies did not receive a diagnosis of PCS, (3) the study was performed in an animal model, (4) studies did not have a comparison group, or (5) the size of the study population was small (<10 patients). Case reports, reviews, abstracts from meetings, and preclinical studies were excluded. By using the following criteria two reviewers (A.A.R. and A.T.L.) independently selected articles for further assessment after the title and abstract review. Disagreements between the two reviewers were resolved by a third independent reviewer (S.T). Potentially eligible studies were then retrieved for full-text assessment.

| Data extraction and critical appraisal of evidence
All full texts of retrieved articles were read and reviewed by two authors (A.A.R. and A.T.L.) and inclusion or exclusion of studies was decided unanimously. When there was disagreement, a third reviewer (S.T.) made the final decision. Using a pre-established protocol, the following data were extracted: first author, study type and characteristics, number of patients, population demographics, tumor characteristics, and survival outcomes. For this review, a data extraction sheet was developed and pilot-tested on three randomly selected included studies, whereupon the sheet was refined accordingly. Data extraction was performed by two review authors (A.A.R. and A.L.). A third author (R.V.) validated the correctness of the tabulated data. Potential inter-reviewer disagreements were resolved by consensus.

| Data analysis
Odds ratios (ORs) with 95% confidence interval (CI) and p values for mortality at 1, 2, 3, 4, and 5years were calculated. Forest plots were created to represent the clinical outcomes. χ 2 tests and I 2 tests were performed for assessment of statistical heterogeneity. 7 The OR were combined across the studies using a Mantel-Haenszel randomeffects model. 8 Funnel plots were constructed to assess publication bias.
All analyses were completed using the "metafor" package of R Statistical Software (version 4.0.2 (2020-06-22), Foundation for Statistical Computing, Vienna, Austria). A two-tailed p-value < .05 was considered statistically significant.

| Tumor and treatment characteristics
Characteristics regarding the PCS and the respective treatment details are provided in Table 2. On average the origin of the PCS was found: 36.2 ± 6.6% in the right atrium (RA), 34.9 ± 8.4% in the left atrium (LA), 9.1 ± 8.1% in the right ventricle (RV), and 7.2 ± 4.1% in the left ventricle (LV). Surgical resection of the cardiac sarcoma (alone or in conjunction with another treatment) was performed on an average in 68.8 ± 7.9% of the patients.

| Mortality: Surgical treatment versus no surgery
Surgery (alone or in conjunction with other treatments) was compared to no surgery (conservative, chemotherapy, or radio-chemotherapy), with three studies 11,17,18 reporting mortality at 1, 2, and 3 years, and four studies 10,11,17,18
There was no evidence of heterogeneity of treatment among studies at 1,

| Risk of bias across studies
Funnel plot analysis ( Figures S1, S2, and S3) did not disclose asymmetry around the axis for the treatment effect in any of the studied outcomes. Consequently, publication bias related to these outcomes is unlikely. We found no evidence for publication bias, Egger's as well as Begg and Mazumbdar tests were not significant for any of the meta-analyses. Funnel plots are available in the Supporting Information section.  (36.2 ± 6.6% in RA; 34.9 ± 8.4% in LA). The current literature found cardiac sarcomas situated on the right side of the heart to be more bulky, infiltrative, and to lead to earlier metastasis when compared to left-sided PCS, thus also leading to worse outcomes. 13

| Surgery or no surgery?
Surgical resection of PCS has been long considered the gold standard treatment for PCS in eligible patients without metastasis at diagnosis. 20 The operation is usually carried out through a median sternotomy, minimal touch technique, and with the use of cardiopulmonary bypass with cardioplegia. 23

| Surgery versus multimodality treatment
The evidence with regard to the pre-and postoperative treatment of cardiac sarcoma remains scarce and highly controversial. Due to its extremely rare prevalence, no prospective study currently exists reporting on the benefits of radiotherapy, chemotherapy, or chemotherapy plus radiotherapy. Furthermore, the marked differences between PCS and noncardiac sarcoma make data extrapolation from larger studies on noncardiac sarcoma difficult and not always a feasible option. However, postoperative chemotherapy after surgical resection tends to be recommended even in cases of complete resection with clear surgical margins, mainly due to the possibility of malignant tissue being left. 29 The data on the chemotherapeutic regimens used usually tends to be derived from prospective clinical trials carried out on noncardiac sarcoma patients.
On the one hand, studies have been advocating for a multimodal approach, including surgery and chemotherapy and/or radiotherapy, referring to advantages in terms of survival in the multimodality group. Randhawa et al. 13

| Mortality in angiosarcoma
Another area of controversy around PCS is the severity of angiosarcoma and its association with a worse prognosis.
Angiosarcomas, an endothelial cell tumor, have been found to be the most common type of PCS according to the literature, also supported by the results of our analysis. Their prevalence has been found to range anywhere from 21% to 48% of all PCSs. [9][10][11][12][13][14][15][16][17][18] They mostly originate in the right atrium and appear as a multicentric mass. 21 Similar to other PCSs, they tend to infiltrate into surrounding cardiac tissues, and rapidly metastasize into the lungs, thus making surgical resection a complex option. 30

| Further comments
It is of utmost importance to consider the fact that the outcomes of PCS can be the result of diverse factors, not all of them being necessarily associated with the nature of the treatment itself. Similarly, the degree of impact of covariates and of patient characteristics should not be underestimated. Selection bias remains a point of concern with regards to the data available within this field of research. Indeed, it could be argued that more fragile patients would have been assigned to conservative or nonsurgical treatment, while healthier patients would have undergone surgical treatment of PCS.
Therefore, the results of this meta-analysis should be interpreted with caution and consideration should be made on the effects of potential selection bias. In our meta-analysis, we found a mortality advantage from Year 1 to Year 3 with surgery when compared to no surgery; confirming these findings, Yin et al. 18 found that the aHR for surgery was significant. However, these findings could not be confirmed in the adjusted Cox regression by Chen et al. 16 Similarly, a 1-year mortality advantage was found with multimodality treatment, as confirmed by the aHR by Hendriksen et al., 17 but not confirmed by Chen et al. 16

| Limitations
A major limitation of the evidence included in our review is the nonrandomized and retrospective nature of all the studies, with most of them having a small number of patients, thus increases the risk of biases. Moreover, current data were too limited to run a sensitivity analysis based on adjusted estimates from the multivariate regression. Furthermore, no current consensus and standardized chemotherapeutic or radiotherapeutic regimen currently exists for PCS, with differences among them between studies being present. The same was valid with regard to differences in surgical resection and reconstruction techniques used for PCS. Furthermore, due to the nature of the data and inconsistencies in reporting it was not possible to compare the outcomes of the different subtypes in the multimodality treatment group. Considerable statistical heterogeneity is present for some outcomes, which we counterbalanced by using the random-effects model.

| CONCLUSION
Surgery has been found, in line with the previously published evidence, to provide a mortality advantage when compared to no surgery at 1-3 years in patients with a diagnosis of PCS. Multimodality treatment was only found to provide lower mortality rates at 1 year, with no significant difference at 2, 3, 4, and 5 years. Although the data of this metanalysis should be interpreted with caution due to numerous outlined limitations, they serve as a baseline for future prospective studies to be carried out to assess differences in treatment outcomes for PCS patients.