CARDIOVASCULAR JOURNAL OF AFRICA • Volume 35, No 2, May – August 2024 AFRICA 77 AF, whereas a tachycardia-dependent right bundle branch block was present in five patients at the time of the operation. Three patients had had previous cardiac surgery and 14 (50%) needed a concomitant procedure. Carpentier type B anomalies were the most frequent type (n = 12), followed by type A (n = 11) and C anomalies (n = 5). Mean intensive care unit and hospital stays were 6.3 ± 7.9 and 13.2 ± 10.2 days, respectively. Early mortality was observed in three patients [tricuspid valve replacement (TVR): two, Danielson: one] (10.7%) due to low cardiac output syndrome, where one patient underwent a re-do TVR. The other patients already had pre-operative biventricular dysfunction and developed renal failure. Intra-aortic balloon pump support was provided in all three patients. Two patients (TVR: one, modified Danielson: one) (7.1%) died during the late postoperative period. Causes of mortality were cerebrovascular accident and multi-organ failure (the patient who underwent Danielson repair). Postoperative infection involving the lungs was observed in both patients. Median follow up was 126 months [interquartile range 105–192]. Overall survival rate was 89.3, 85.4, 85.4 and 68.3% at five, 10, 15 and 20 years, respectively (Fig. 1). At long-term follow up, the NYHA class was II in five patients and III in two patients. Further analyses were reviewed to see how subgroup movement affected the results, even if the number per group was small, and the patients were divided into two groups (group I: tricuspid reconstruction, group II: TVR) to assess the impact of procedural intervention and residual TI on survival. We concluded that there was no statistically significant difference at survival according to the type of surgery (group I: 94.7% at five years, 89.2% at 10, 15 and 20 years; group II: 77.8% at five, 10 and 15 years) (p = 0.094, Fig. 2) and the presence of residual TI (≥ 2) (p = 0.57, Fig. 3). None of the patients required early re-operation for recurrent TI. Freedom from late re-operation was 71.1% at 20 years (Fig. 4). Reasons for re-operation were prosthetic dysfunction in one patient at 17.1 years and recurrent TI in two patients who underwent valve replacement at 7.3 and 16 years. One patient was found to have immobile bioprosthesis cuspis three years after surgery, and was treated with recombinant tissue plasminogen activator and discharged with a normo-functional bioprosthesis. There was no known early or late bioprosthetic valve endocarditis reported from our institution. Consecutive echocardiographic assessment showeda reduction in the RA size (6.1 ± 0.6 to 4.4 ± 0.5 cm, p < 0.001) with good functioning of the reconstructed TV. Likewise, improvement in NYHA functional class was observed (p < 0.001). Pre- and postoperative echocardiographic evaluation showed that there were significant declines in TI, PHT and right ventricular ejection function as well (p < 0.05). At a recent follow up, only one patient was found with moderate TI, whereas the others had no or mild regurgitation. Comparative echocardiographic evaluation of the patients can be seen in Table 4. Integrity of anterior leaflet + attachment location TVR (n = 9) Carpentier (n = 3) Hunter–Lillehei–Hardy (n = 2) Classical/modified Danielson (n = 12) Targeting RV Repair was not feasible Targeting leaflet Fig. 1. Consort flow chart of interventions. RV, right ventricle; TVR, tricuspid valve repair. 0 100 200 300 Time (months) Number at risk 1.0 0.75 0.50 0.25 0.00 Survival probabilities 0 100 200 300 All strata 26 21 5 1 Time (months) Fig. 2. Overall survival.
RkJQdWJsaXNoZXIy NDIzNzc=