CARDIOVASCULAR JOURNAL OF AFRICA • Volume 33, No 4, July/August 2022 198 AFRICA In a more recent update from the Western Cape, de Villiers et al. documented a rise in the cases of IV drug use (14.2%) and congenital heart disease (10.5%), with underlying RHD present in only 34.3% of their study population.1 Meel et al. also recently reported 68 cases of IE in IV drug users over a two-year period (December 2014 to February 2017) in Gauteng, South Africa.16 Most of the subjects of Meel et al. were male (97%) and HIV positive (76.1%), not unlike Western series, which reported IE with HIV infection almost exclusively in IV drug users.16 This changing profile of IE, characterised by an increase in IV drug use, corrective surgery and device-related sepsis, probably explains why Staphylococcus aureus infection has replaced Streptococcus viridians as the most frequent pathogen isolated in recent studies.17 These factors, however, do not explain the emergence of staphylococcal infection as the predominant pathogen in our study. The pattern of Staphylococcus aureus as the dominant pathogen, followed by Streptococcus viridans, has been documented in other local studies.1,18-20 Importantly, a third of subjects had negative blood cultures, which we attributed in part to prior antibiotic therapy administered at the referring hospital. Another explanation for culture negativity, which emerged from our study, was the lessthan-ideal specificity of the modified Duke criteria, since the surgeon found no evidence of IE in 10 of the 73 subjects who underwent surgery, five of whom were culture negative. Among our 12 HIV-positive subjects, infection with Staphylococcusaureus and Streptococcusviridans was notedinonly one case each. Opportunistic organisms not typically associated with IE were present in the remaining seven HIV-positive cases. This list of organisms included Bacillus cereus, which is rarely cultured in native valve endocarditis and is associated with immunocompromised states.21 Klebsiella pneumonia is another atypical organism cultured in IE.22 This organism is extremely rare in community-acquired infections, but Klebsiella pneumonia endocarditis can be present when a pneumonia leads to progressive bacteraemia, resulting in complications such as endocarditis, especially in immunocompromised states.22 We documented one such case of Klebsiella pneumonia IE in the HIV-positive group. This organism typically involves the aortic valve, which was evident in the HIV-positive patient in our study.22 We also documented Micrococcus in one HIV-positive patient with native valve endocarditis. This is another organism that is not typically associated with IE and is extremely rare in native valve endocarditis. This organism is also more commonly associated with immunocompromised states.23 The atypical organisms cultured in the HIV-positive group prove that these patients had opportunistic infections due to their immunocompromised state. Commencement and compliance with antiretroviral therapy may prevent such infections in the future. All the patients in whom an atypical organism was cultured received directed antibiotic therapy for six weeks. In our study, we documented an 87.5% survival rate in the HIV-positive group who underwent surgery. Even though the HIV-positive group was small, these findings demonstrate that patients who are HIV-positive can negotiate surgery well. In a review of cardiac surgery at our centre, Blythe et al. demonstrated that HIV-positive non-IV drug users with a CD4 count of > 400 copies had outcomes that were comparable to HIV-negative patients undergoing cardiac surgery.24 The patients in our study who underwent surgery had CD4 counts ranging from 90–970 copies (mean 471.4 copies) and our outcomes are similar to that of Filsoufi et al., who described a three-year survival rate of 86% in 25 patients undergoing cardiac surgery with a CD4 count ranging from 51–1 050 copies (mean 440 copies).24 The 26.4% adjusted mortality rate in our study is high, compared to in-hospital mortality rates of 16–20% for community-acquired IE in Western series,4 but consistent with local studies that reported mortality rates of 23.4–35.6%,9,25 and other studies from developing countries that reported rates of 19–46%.1 De Villiers et al. attributed the low mortality rate of 16.2% in their most recent report to the liberal use of cardiac surgery for patients, according to international guideline-based indications for surgery.1 The overall mortality rate with HIV infection was high (33.3%) but the few cases in this group did not allow an analysis of the role of immune suppression in the development of complications/death related to co-existing HIV infection. In our study, the main indications for surgery were intractable heart failure, large vegetations with increased risk for embolisation and ongoing sepsis despite appropriate antibiotic therapy. Although these indications for cardiac surgery followed established guidelines,15 several factors contributed to the high mortality rate in our study. Firstly, our pre-operative mortality rate was high because eight patients presented late in a critically ill state, precluding surgery. Delay due to difficulties in access to care as well as late diagnosis and appropriate treatment at the referring hospital led to many cases presenting with complications and in advanced heart failure, resulting in most deaths occurring prior to surgery, even in patients with rightsided IE, which is known to be associated with a lower mortality rate.16 A very sobering finding was that seven patients died while awaiting surgery, which may have been preventable had surgery been expedited in these cases. In this respect, the recent article by de Villiers et al. bears relevance since these authors were able to show a much lower mortality rate with early surgical intervention.1 Multivariate regression analysis showed that acute onset of IE, vegetation size of > 15 mm and medical management without surgical intervention were associated with increased in-hospital mortality rates. It is clear that aggressive disease led to valve Table 6. Univariate and multivariate logistic regression analyses Variables Unadjusted odds ratio (95% CI) p-value Adjusted odds ratio (95% CI) p-value Age > 50 years* 0.76 (0.10–5.96) 0.796 0.09 (0.00–32.28) 0.262 Acute-onset IE** 2.39 (0.90–6.38) 0.081 251.46 (1.18–5343.63) 0.043 Prosthetic valve IE** 3.78 (0.99–14.49) 0.053 0.40 (0.00–516.70) 0.800 Fever 3.45 (1.31–9.11) 0.012 7.56 (0.04–1601.26) 0.459 Clubbing** 0.54 (0.20–1.48) 0.232 0.75 (0.02–23.84) 0.868 Haematuria 3.76 (1.42–9.98) 0.008 6.86 (0.23–206.08) 0.267 Haemoglobin < 10 g/dl 1.30 (1.03–1.64) 0.012 0.32 (0.01–11.93) 0.537 Staphylococcus aureus* 1.11 (0.33–3.77) 0.869 2.74 (0.03–216.45) 0.651 Heart failure* 1.10 (0.67–1.79) 0.704 10.57 (0.50–22.99) 0.129 Embolic events* 1.03 (0.64–1.65) 0.893 4.33 (0.22–86.53) 0.338 Vegetation size > 15 mm* 2.12 (0.72–6.73) 0.009 222.60 (1.04–4730.34) 0.043 Medical management only 0.03 (0.00–0.11) 0.000 20.89 (2.12–200.06) 0.037 *Known predictors of mortality. **Variables included in the multivariate analysis if p < 0.25 in the univariate analysis.
RkJQdWJsaXNoZXIy NDIzNzc=