This study reports the immediate outcomes in terms of mortality and complications of patients with Tetralogy of Fallot who had primary intracardiac repair from Uganda heart Institute between February 2012 and August 2022.
Demographics, pre -operative, intraoperative characteristics of participants
In this study 85% (n = 75) of the participants were more than two years of age. The median age at operation was 4.0 years. Phenotypically syndromic patients were only 5.7%(n = 5) and the commonest syndromes were Downs and Noonan with an equal presentation. It’s however well known that the most common syndrome associated with TOF is 22q11.2 microdeletion (di-George) (14), although in our study only 1.1%(n = 1) was found. This could be an underestimate since only phenotypic features were used without chromosomal studies. According to literature some children with DiGeorge may look phenotypically normal (14). It’s therefore recommended that these children are screened for 22q11.2 microdeletion so that the anticipated problems are properly managed.
Pulmonary valve annular diameter is of great significancy prior to surgery as it’s one of the major determinants of method of surgery and outcome especially pulmonary regurgitation. Trans annular patch repair (TAP) is indicated for patients with inadequate pulmonary valve annulus usually Z score <-2 (11, 12). In our study only 8%(n = 7) had an inadequate pulmonary valve anulus requiring a trans annular patch. This could have been affected by bias in patient selection.
Pre -operative high hematocrit levels and low oxygen saturations have been found in other studies as predictors of complicated surgical course (15). In our study, 36.3% (32/88) of the participants had higher hematocrit levels than the upper normal of 53.7%. At multivariate analysis, low hematocrit < 35% was shown to predict good outcome with reduction in arrhythmias at a p-value of 0.02. There was however no association between baseline hematocrit and mortality.
Only 31.8% (28/88) of participants had optimum Oxygen saturations of > 92%. The remaining 68.2% (60/88) had sub optimal oxygen saturations. There was no correlation between preoperative oxygen saturations and mortality in our study.
All participants had TOF with favorable anatomy qualifying them for primary intracardiac repair without prior palliative surgery. This could have been affected by bias at patient selection since they were targeting patients who qualified for primary repair.
The commonly found VSD type was mal aligned peri-membranous accounting for 86.4%(n = 76). Mal-aligned VSDs are the commonest, and they are known to be because of persistent patency of the interventricular foramen.
Right ventricular outflow tract obstruction (RVOTO) was mainly in the infundibular area with 97.7% (n = 86). Some few patients had additional stenosis in other areas including pulmonary valve, main pulmonary artery, and hypoplastic branch pulmonary arteries.
The commonest associated extracardiac lesions in this study included right aortic arch 14.8%(n = 13), bilateral superior vena cava 8%(n = 7), patent ductus arteriosus 8.0%(n = 7), coronary artery crossing right ventricular outflow tract 3.4%(n = 3) and double outlet right ventricle 2.3% (n = 2). Intracardiac associations included isolated secundum atrial septal defects /patent foramen ovale (PFO) at 26.1%(n = 23) and additional tiny muscular VSDS 4.5% (n = 4).
Patients with associated lesions were however fewer as compared to other studies, this could have been because of the superiority of the methods used in those studies as compared to ours where only a transthoracic echocardiography was used. In the study by Hu B-y (16), they used computed tomography. Low dose dual source computed tomography was found to have a higher accuracy compared to transthoracic echocardiography at detecting associated lesions.
Mortality among patients who had primary intracardiac repair at UHI.
The overall mortality was 8%(n = 7).
Our mortality was however slightly lower and a bit comparable with the Cameroonian study where they reported a mortality of 9.0%(17). This study also looked at older children and it was an African setting like ours.
Another study carried out in Iran from 2008 to 2010 found a mortality rate of 12.2% which is higher than our mortality (18). In this study they assessed 74 cases of TOF with pulmonary stenosis for early post operative mortality and their mean age was 5.74yrs comparable with ours which was 5.72yrs.
A study done in Addis Ababa-in Ethiopia, the mortality rate was 12.9%, higher than ours. The study looked retrospectively at 62 children and adolescents who underwent tetralogy of Fallot (TOF) repair. In their study, 91.9% underwent primary TOF repair while 6.5% first underwent initial palliative surgery(19). We note that these studies, which compare with ours were in developing countries.
Some of the studies conducted in developed countries found mortality quite lower and some with no mortality recorded in the early post operative period. A study conducted in Saudi Arabia by Sameh e t al (20) found no mortality in the early post operative period. They looked at 64 cases of children who had undergone surgical repair and compared those who had undergone trans annular patch with those who didn’t. In this study there was no surgical mortality recorded in both groups. It’s noted that these two groups had lower median ages compared to ours.
Other early outcomes were associated with patients who had primary TOF repair.
The outcome with highest frequency included residual VSDs at 39.8%(n = 35). Those with trivial to small VSDs were 36.4% (n = 32), moderate 2.3% (n = 2) and large 1.1%(n = 1). Trivial to small VSD’s are usually expected findings following TOF repair which may close with time and pose no hemodynamic significancy post operatively. Moderate-large VSDs usually need follow-up and may require closure with time either surgically or using devices. The large residual VSD in this study was a result of patch dehiscence secondary to septicemia and had redo surgery within that period.
The rest of the early post operative outcomes included residual right ventricular (RV) dysfunction 37.5% (33/88), residual pulmonary regurgitation 30.7%(27/88), residual right ventricular outflow tract obstruction (RVOTO) 30.7%(27/88), pleural effusion 27.3%(24/88), arrhythmias 27.3%(24/88), post operative infections 26.1%(23/88), left ventricular (LV) systolic dysfunction 10.2% (9/88), acute renal failure requiring dialysis 8% (7/88), low cardiac output syndrome 6.8%(6/88), pneumothorax 4.5%(4/88), pericardial effusion 3.4%(3/88), re-operation due to bleeding 1.1% (1/88), new onset seizures 1.1%(1/88), chylothorax 1.1%(1/88).
In the Cameroonian study by Jct tchoumi e tal (17) acute post-surgical complications included mild pericardial effusion detected in four cases, pleural effusion in three cases ,one case with chylothorax and no cases of arrhythmia. These were fewer complications than the ones reported in our study though they had a slightly higher mortality of 9%. These differences could have been because of fewer numbers in their study where they looked at only 22 patients.
Factors associated with mortality and morbidity of patients who underwent primary TOF repair.
Cumulative prolonged cardiopulmonary bypass time of > 180 minutes and prolonged mechanical ventilation time of > 24hrs post procedure were statistically associated with mortality with a p-value of 0.001 at bivariant analysis. At multivariate analysis however, only mechanical ventilation time was significantly associated with mortality p-value of 0.014. This could be because patients who died in our study were sicker and required more mechanical ventilatory support.
In the Ethiopian study, cardiopulmonary bypass time/ aortic cross-clamp time and pulmonary valve annulus diameter less than three standard deviation( SD) were independently associated with perioperative mortality (19). This could have been because some had prior palliative surgeries.
Acute kidney injury (AKI) requiring dialysis was one of the common denominators among the patients who died. This can be attributed to limited access to hemodialysis at UHI with only peritoneal dialysis available by then. AKI can also be attributed to right ventricular (RV) failure and prolonged cumulative cardiopulmonary bypass time which was common to all who died.
Factors associated with residual VSD in this study were gender and performing augmentation of RVOT.
Factor associated with residual PR in this study were gender status, duration of admission preoperatively (days), body surface area (m2), aortic clamp time (hrs), mechanical ventilation time (hrs) and surgical technique applied specifically the commissural approach.
Factors associated with risk of getting arrhythmias in our study were increase in age at operation (yrs.), female gender. Having an arrhythmia prolonged patient’s stay in the ICU.
Lower preoperative haematocrit (%), not doing commissurotomy or augmentation of right ventricular outflow tract (RVOT) in our study were found as good indicators of surgery. A study done in Berne Switzerland found age, cardiopulmonary bypass time and aortic cross clamp time and higher maximum post operative troponin serum levels(21). These factors were quite different from those found in our study apart from age. The difference could be because they looked at 3 different heart conditions i.e. isolated ventricular septal defects, tetralogy of Fallot and atrial ventricular septal defects.
Right ventricular dysfunction in our study was associated with absence of a genetic syndrome and prolonged cardiopulmonary bypass time. Children with obvious phenotypic syndromes were however very few. It’s possible that we had more children with genetic syndromes but with subtle physical features.
Study strengths and limitations.
This study provides the first documented report of early surgical outcome of Tetralogy of Fallot patients following primary intracardiac repair in Uganda.
The disadvantage of this study design is the inferior level of evidence compared with prospective studies.
It was a single center study /only center where open-heart surgery is done in the country.
There was bias at patient selection for surgery during nascent period of the program.
Limited follow-up of patients and some key statistics could not be measured given the retrospective design e.g. right ventricular function assessment post operatively was by (tricuspid annular plane systolic excursion (TAPSE) and cardiologists’ subjective assessment. Fractional area change is the preferred method of assessment for more accurate results.
Missing data and confounders were difficult to cater for.
A larger sample size was needed to be able to describe rare outcomes.