Discussion
Minimally invasive cardiac surgery has been reported since the 1990’s14. Described benefits includes: reduced blood loss and pain, shorter LoS, and generally, superior patient satisfaction20. MI was also found beneficial in high-risk patients to reduce surgical trauma7.
Due to the rarity of intra cardiac masses few studies comparing the outcomes of MI and MS surgery have been published. The present meta-analysis, by aggregating data from 11 studies, confirms that the MI approach for cardiac tumor resection is as safe as MS, with excellent early and late outcomes, very low recurrence rates, and rare need for reoperation at follow-up.
Among the total number of 653 tumors described, 601 (92%) were myxomas, 22 (3.3%) papillary fibroelastoma, 13 (1.9%) thrombus, 6 (0.9%) vascular malformation, 2 (0.3%) lipoma, 1 (0.1%) fibroma, 1 (0.1%) rhabdomyoma, 1 (0.1%) chondrosarcoma, 1 (0.1%) hamartoma, and 5 (0.7%) classified as ‘other’. Only Ravikumar9included a case of secondary atrial chondrosarcoma.
This study confirms that the MI approach requires prolonged operative time due to longer CPB and CC time. However, there was no demonstrable effect on clinical outcomes and MI had similar rates of post-operative complications as the MS approach. In line with previous studies, LOS was significantly shorter in the MI group20. The same results were observed in the subgroup analysis of the high-quality studies10-14, 16.
Notably, we did not find significant differences in terms of tumor size between groups. We may speculate that large masses may preclude the MI approach. Nevertheless, even in cases of large size of the tumor, the MI approach may be still feasible21.
Regarding the incidence of recurrences, our results are in line with previously published literature. A retrospective study from the Mayo Clinic spanning over 50 years reported a recurrence rate of 5.6% with a MS approach22. Similarly, Keeling et al reported a rate of 2%23 and in a prospective single cohort series of patients treated with a minimally invasive approach, Bianchi et al24 reported an incidence of recurrence of 3.3%.
With cardiac myxoma, while a stalk base resection is generally indicated to avoid tumor recurrence, its superiority over endocardial resection is under debate. The Mayo Clinic22 study demonstrated that there were no differences in tumor recurrence based in resection margin. Nevertheless, we could not analyze the impact of different surgical techniques (e.g. resection with patch, endocardial resection, single vs double atrial approach) on late recurrences.
In our meta-analysis, there were no conversions to sternotomy and the rate of reopening for bleeding was similar between the two groups (safety end-point). The presence of a right atrial tumor has been reported as possible contraindication for a minimally invasive approach due to fragmentation during cannulation procedures25. Importantly, there was no difference in the occurrence of post-operative neurological events and no clinically relevant embolization events were observed in the MI group.
All series included in the analysis were low-medium volume, likely due to the rarity of cardiac tumors. With this in mind, we could not evaluate the effect of the surgical cumulative volume on clinical outcomes4.
While a majority of patients included in this study had benign cardiac tumors, it worth noting that the use of MI has been described in literature in the context of primary malignant tumors26. The MI approach has largely been utilized with benign tumors rather than malignant tumors probably due to the greater ease of resection, the less invasive nature of benign tumors, and the lack of need for very complex cardiac reconstructions, which would be difficult with MI access.
Sensitivity analysis of the robotic studies was undermined by the limited sample size; no analysis in terms of survival and relapse at follow-up could be carried out; while we may conclude that the robotic approach may be as safe as MS, we cannot validate the long-term results.
A main limitation inherent to the study design stems from the use of retrospective cohort studies in our pooled analyses. There was a certain degree of clinical heterogeneity; while most of the studies included exclusively myxoma at the level of the left or right atrium, others have included masses at the level of the aortic valve or ventricles. Follow-up times in this analysis were short with respect to tumor recurrence/survival and long-term follow-up was not always available. Finally, no cost-analysis could be performed.