Fatores epidemiológicos, clínicos e microbiológicos relacionados ao prognóstico de adultos com candidemia: estudo de coortes retrospectivas Brasil - Espanha
Data
2024-05-14
Autores
Agnelli , Caroline 
Orientadores
Colombo, Arnaldo
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
A infecção da corrente sanguínea por espécies de Candida é a principal causa de micose invasiva nosocomial, conferindo altas taxas de morbidade e mortalidade entre pacientes adultos hospitalizados em todo o mundo. Fatores prognósticos associados à mortalidade têm sido estudados nos últimos anos, e variam de acordo com a região geográfica devido principalmente às diferenças epidemiológicas da população sob maior risco, distribuição das espécies de Candida com distintos fatores de virulência, agilidade diagnóstica, e evolução das práticas terapêuticas no manejo da infecção. Apesar dos avanços na área, estudos de vigilância sugerem uma maior incidência global de candidemia decorrente de mudanças demográficas como o envelhecimento populacional, e do aumento de procedimentos médicos invasivos mais complexos. No entanto, os dados mais recentes variam consideravelmente entre países, com incidência de até 2,5 casos por 1.000 admissões no Brasil, em comparação com 0,76 na Espanha. Além disso, as taxas de mortalidade geral entre adultos com candidemia também são notavelmente mais altas nos países da América Latina quando comparadas às da Europa Ocidental, podendo chegar a mais de 75%, principalmente entre os pacientes críticos. Esta linha de pesquisa objetivou investigar fatores epidemiológicos, clínicos, e microbiológicos relacionados ao prognóstico de adultos com candidemia, colocando sob perspectivas distintas algumas coortes de pacientes tanto do Brasil quanto da Espanha, a fim de aprofundar o estudo dessas variáveis e da história natural da candidemia, além de propor soluções ajustadas à realidade brasileira para os desafios de melhoria do manejo terapêutico e de redução da mortalidade relacionada a esta micose. Artigo 1: Estudo de coorte retrospectivo incluindo todos os casos consecutivos de adultos com candidemia persistente diagnosticados entre 2010-2018 em um hospital terciário na Espanha com programa de racionalização de antifúngicos através de equipe multidisciplinar especializada. Definimos candidemia persistente como o crescimento da mesma espécie de Candida no sangue após cinco ou mais dias da coleta da primeira hemocultura positiva, a despeito de terapia antifúngica. Um total de 255/322 (79,2%) adultos com candidemia possuíam hemoculturas de controle, sendo 35/255 (13,7%) casos de candidemia persistente. Comparando os casos de candidemia persistente com os demais, não encontramos diferenças no manejo terapêutico, seja na adequação do antifúngico (p=0,084), controle de foco primário de infecção (p=0,172), tempo para início do tratamento (p=0,311), ou para o controle desse foco (p=0,748). Apesar da tendência à maior proporção de cepas produtoras de biofilme entre os casos de candidemia persistente, não encontramos diferenças com significância estatística neste estudo (p=0,064). Ademais, embora o tempo de hospitalização tenha sido mais prolongado entre os pacientes com candidemia persistente (p=0,021), não houve diferença entre as taxas de mortalidade geral aos 30 dias (31,4% vs. 22,3%, p=0,238). O único fator associado à candidemia persistente foi a presença de foco de infecção ainda oculto no momento da hemocultura de controle positiva [odds ratio (OR) 4,28, IC 95%:1,77-10,34, p=0,001], com destaque para as complicações não oculares (p=0,002), incluindo tromboflebite séptica, endocardite, e acometimento de outros órgãos profundos. Portanto, concluímos que a candidemia persistente em um centro com programa de racionalização de antifúngicos não está associada ao manejo terapêutico inadequado, ou a um pior prognóstico dos pacientes, mas a focos ainda ocultos de infecção metastática que exigem propedêutica diagnóstica específica e manejo direcionado. Artigo 2: Estudo de coorte retrospectivo incluindo todos os pacientes adultos diagnosticados com candidemia entre 2012-2017 em um hospital terciário na Espanha, com determinação de (1,3)-ß-D-Glucano (BDG - Fungitell, positivo se ≥80pg/mL), em ao menos duas amostras clínicas durante o episódio de candidemia. Comparamos a epidemiologia e o desfecho de pacientes com todas as amostras de BDG negativas em relação aos casos com ao menos uma amostra de BDG positiva. Um total de 148/236 (62,7%) dos episódios de candidemia apresentavam detecção de BDG em ao menos 2 amostras séricas, sendo 26/148 (17,6%) deles com todos os testes de BDG negativos. Neste grupo, o crescimento de Candida em todos os três pares de hemoculturas coletadas no momento do diagnóstico foi menos frequente (p=0,005), e a apresentação clínica foi menos grave quando avaliada pela mediana do Pitt Score (p=0,039). Além disso, apesar de não haver diferença entre as taxas de manejo terapêutico adequado (p=0,599), o clareamento da fungemia foi mais rápido, já na primeira hemocultura de controle (p=0,005), a proporção de complicações foi menor (p=0,008), assim como a taxa de mortalidade por todas as causas em 30 dias (3,8% vs. 23,8%, p=0,004), e o tempo de hospitalização (p=0,003). Concluímos através da análise multivariada, que ter todos os testes de BDG negativos durante o episódio de candidemia mostrou-se um fator individualmente relacionado a um melhor prognóstico (OR 0,12, IC 95%: 0,03-0,49,p=0,003), possivelmente atribuído a uma menor carga fúngica circulante. Nesse contexto, o valor prognóstico desse biomarcador parece ser promissor, podendo contribuir com a individualização do manejo da candidemia em um futuro próximo. Artigo 3: Estudo de coorte retrospectivo multicêntrico, no qual avaliamos 720 episódios consecutivos de candidemia em adultos diagnosticados em hospitais terciários entre 2010-2018, sendo 323 provenientes da Espanha, e os demais do Brasil. Comparando os casos de candidemia entre os países, notamos que os pacientes espanhóis receberam tratamento mais precoce (p<0,001), com uso mais frequente de equinocandinas (p=0,001), e retirada mais rápida do cateter venoso central (CVC), p=0,012. Ademais, a taxa de mortalidade geral foi menor na Espanha tanto aos 14 dias (20,1% vs. 35,8%, p<0,001), quanto aos 30 dias (31,6% vs. 51,9%, p<0,001). Entre os fatores prognósticos associados à mortalidade de maneira independente, encontramos a idade avançada (p=0,001), neutropenia (p=0,001), doença pulmonar crônica (p<0,001), uso de corticoide (p=0,039), e manejo terapêutico inadequado (p<0,001). Já entre os fatores protetores, encontramos candidemia por Candida parapsilosis (p<0,001) e ser proveniente da Espanha (p<0,001). Portanto, este estudo confirma taxas de mortalidade realmente mais baixas entre adultos com candidemia na Espanha, e sugere que há fatores modificáveis do manejo terapêutico associados ao prognóstico desses pacientes com margem de melhoria nos centros brasileiros. Artigo 4: Estudo retrospectivo de tendência histórica comparando duas coortes de vigilância epidemiológica incluindo adultos com candidemia provenientes de onze centros brasileiros, diagnosticados em dois períodos distintos: 2010-2011 (período I) vs. 2017-2018 (período II). Identificamos 616 casos consecutivos, sendo 247 do período II. Os pacientes do período mais recente tinham mais comorbidades (p<0,001), e mais internações hospitalares prévias (p=0,001). Além disso, receberam equinocandinas com maior frequência (p=0,001), porém sem mudanças significativas no tempo para início do antifúngico (p=0,369) ou para retirada do CVC (p=0,644). Apesar da tendência à queda, ainda chama a atenção a porcentagem de pacientes sem tratamento em ambos os períodos I e II (23,6% vs. 17,4%, p=0,07), respectivamente. Infelizmente, não houve avanço na redução das taxas de mortalidade geral aos 14 dias (33,6% vs. 37,7%,p=0,343), ou aos 30 dias (51,4% vs. 48,6%,p=0.511). Em conclusão, evidenciamos taxas de mortalidade ainda elevadas, sem redução significativa ao longo do tempo, apesar do maior acesso às equinocandinas, provavelmente decorrente do aumento da complexidade dos pacientes, e do atraso nas intervenções terapêuticas. Nossos dados sugerem que as estratégias do manejo de candidemia devem ser adaptadas às mudanças epidemiológicas constatadas, objetivando maior agilidade diagnóstica em prol da redução do número de pacientes elegíveis não tratados, favorecendo um início mais precoce de antifúngico eficaz e do controle do foco de infecção.
Bloodstream infection by Candida species is the main cause of nosocomial invasive mycosis, conferring high rates of morbidity and mortality among hospitalized adult patients worldwide. Prognostic factors associated with mortality have been studied in recent years, and variations according to geographic regions are mainly due to epidemiological differences in the population at greatest risk, distribution of Candida species with different virulence factors, diagnostic agility, and evolution of therapeutic practices. Despite advances in the area, surveillance studies suggest a higher global incidence of candidemia resulting from demographic changes such as population aging, and an increase in more complex invasive medical procedures. However, such data varies considerably between countries, with an incidence of up to 2.5 cases per 1,000 admissions in Brazil, compared with 0.76 in Spain. Furthermore, all-cause mortality rates among adults with candidemia are also notably higher in Latin American countries when compared to those in Western Europe, reaching more than 75%, especially among critically ill patients. This research project aimed at investigating epidemiological, clinical, and microbiological factors related to the prognosis of adults with candidemia, by analyzing different cohorts of patients from both Brazil and Spain from different perspectives. Our studies were conducted in order to further understand these variables and the natural history of candidemia, aside from proposing solutions adapted to the Brazilian reality for the challenges of improving therapeutic management, and reducing the mortality rates related to this mycosis. Article 1: Retrospective cohort study including all consecutive cases of adults with persistent candidemia diagnosed between 2010-2018 in a tertiary hospital in Spain with an antifungal rationalization program through a specialized multidisciplinary team. We defined persistent candidemia as the growth of the same species of Candida in blood cultures within five or more days after the first positive blood culture was collected, despite receiving antifungal therapy. A total of 255/322 (79.2%) adults with candidemia had subsequent control blood cultures, being 35/255 (13.7%) cases of persistent candidemia. Comparing the cases of persistent candidemia with the others, we found no differences in therapeutic management, whether in the adequacy of the antifungal (p=0.084), control of the primary source of infection (p=0.172), time to start treatment (p=0.311), or to accomplish source control (p=0.748). Despite the tendency towards a higher proportion of biofilm-producing strains among cases of persistent candidemia, we did not find statistically significant differences in this study (p=0.064). Furthermore, although hospitalization time was longer among patients with persistent candidemia (p=0.021), there was no difference between overall mortality rates at 30 days (31.4% vs. 22.3%, p=0.238). The only factor independently associated with persistent candidemia was the presence of an undiagnosed source of infection at the time of the positive control blood culture (OR 4.28, 95% CI:1.77-10.34, p=0.001), with emphasis on non-ocular complications (p=0.002), including septic thrombophlebitis, endocarditis, and involvement of other deep organs. Therefore, persistent candidemia in a center with an antifungal rationalization program is not associated with inadequate therapeutic management or a worse prognosis, but with hidden foci of metastatic infection that require specific diagnostic workup and proper management. Article 2: Retrospective cohort study including all adult patients diagnosed with candidemia between 2012-2017 in a tertiary hospital in Spain, with determination of (1,3)-ß-D-Glucan (BDG Fungitell, positive cut-off if ≥80pg/mL), in at least two clinical samples during the candidemia episode. We compared the epidemiology and outcomes of patients with all negative BDG samples versus cases with at least one positive BDG sample. A total of 148/236 (62.7%) candidemic patients had at least two BDG tests, being 26/148 (17.6%) persistently negative. In this group, Candida growth in all three sets of blood cultures collected at the time of diagnosis was less frequent (p=0.005), and the clinical presentation was less severe when assessed by their median Pitt Score (p=0.039). Furthermore, although no differences between the rates of adequate therapeutic management were observed (p=0.599), the clearing of fungemia was faster (p=0.005), the proportion of complications was lower (p=0.008), as was all-cause mortality within 30 days from fungemia (3.8% vs. 23.8%, p=0.004), and the hospital length of stay was shorter (p=0.003). In the multivariate analysis, having persistently negative BDG tests during the candidemia episode was associated with a better prognosis (OR 0.12, 95% CI: 0.03-0.49, p=0.003), possibly due to lower fungal burden. In this context, the prognostic value of this biomarker appears to be promising and may contribute to the individualization of candidemia management in the near future. Article 3: Retrospective, multicenter cohort study, evaluating 720 episodes of candidemia among adults diagnosed in tertiary hospitals between 2010-2018, being 323 patients from Spain, and the rest from Brazil. By comparing cases of candidemia between countries, we noticed that Spanish patients received treatment earlier (p<0.001), with more frequent use of echinocandins (p=0.001), and faster removal of the central venous catheter (p= 0.012). Furthermore, the overall mortality rate was lower in Spain both at 14 days (20.1% vs. 35.8%, p<0.001) and at 30 days from fungemia (31.6% vs. 51.9%, p<0.001). Among the prognostic factors independently associated with mortality, we found advanced age (p=0.001), neutropenia (p=0.001), chronic lung disease (p<0.001), use of corticosteroids (p=0.039), and inadequate therapeutic management (p<0.001). Among the protective factors, we found candidemia due to Candida parapsilosis (p<0.001), or being from Spain (p<0.001). Therefore, this study confirms lower mortality rates among adults with candidemia in Spain, and suggests that there are modifiable therapeutic management factors associated with the prognosis of these patients with room for improvement in Brazilian centers. Article 4: Retrospective multicenter study of historical trends comparing two epidemiological surveillance cohorts including adults with candidemia from eleven Brazilian centers, diagnosed in two different periods: 2010-2011 (period I) vs. 2017-2018 (period II). We identified 616 consecutive cases, being 247 from period II. Patients from the most recent period had more comorbidities (p<0.001), and more frequent previous hospital admissions (p=0.001). Although they received echinocandins more frequently (p=0.001), no changes were observed in the time to start the antifungal agent (p=0.369) or to remove the CVC (p=0.644). Despite the downward trend, the percentage of patients without treatment in both periods I and II (23.6% vs. 17.4%, p=0.07), respectively, still draws attention. Unfortunately, there was no progress over time in reducing overall mortality rates at 14 days (33.6% vs. 37.7%, p=0.343), or at 30 days from fungemia (51.4% vs. 48.6%, p=0.511). In conclusion, we found mortality rates to be still high despite greater access to echinocandins, probably due to an increase in patient complexity and delays in therapeutic interventions. Therefore, candidemia management strategies must be adapted to epidemiological changes, aiming for diagnostic agility in favor of reducing the number of untreated eligible patients, anticipating earlier effective antifungals and source control.
Bloodstream infection by Candida species is the main cause of nosocomial invasive mycosis, conferring high rates of morbidity and mortality among hospitalized adult patients worldwide. Prognostic factors associated with mortality have been studied in recent years, and variations according to geographic regions are mainly due to epidemiological differences in the population at greatest risk, distribution of Candida species with different virulence factors, diagnostic agility, and evolution of therapeutic practices. Despite advances in the area, surveillance studies suggest a higher global incidence of candidemia resulting from demographic changes such as population aging, and an increase in more complex invasive medical procedures. However, such data varies considerably between countries, with an incidence of up to 2.5 cases per 1,000 admissions in Brazil, compared with 0.76 in Spain. Furthermore, all-cause mortality rates among adults with candidemia are also notably higher in Latin American countries when compared to those in Western Europe, reaching more than 75%, especially among critically ill patients. This research project aimed at investigating epidemiological, clinical, and microbiological factors related to the prognosis of adults with candidemia, by analyzing different cohorts of patients from both Brazil and Spain from different perspectives. Our studies were conducted in order to further understand these variables and the natural history of candidemia, aside from proposing solutions adapted to the Brazilian reality for the challenges of improving therapeutic management, and reducing the mortality rates related to this mycosis. Article 1: Retrospective cohort study including all consecutive cases of adults with persistent candidemia diagnosed between 2010-2018 in a tertiary hospital in Spain with an antifungal rationalization program through a specialized multidisciplinary team. We defined persistent candidemia as the growth of the same species of Candida in blood cultures within five or more days after the first positive blood culture was collected, despite receiving antifungal therapy. A total of 255/322 (79.2%) adults with candidemia had subsequent control blood cultures, being 35/255 (13.7%) cases of persistent candidemia. Comparing the cases of persistent candidemia with the others, we found no differences in therapeutic management, whether in the adequacy of the antifungal (p=0.084), control of the primary source of infection (p=0.172), time to start treatment (p=0.311), or to accomplish source control (p=0.748). Despite the tendency towards a higher proportion of biofilm-producing strains among cases of persistent candidemia, we did not find statistically significant differences in this study (p=0.064). Furthermore, although hospitalization time was longer among patients with persistent candidemia (p=0.021), there was no difference between overall mortality rates at 30 days (31.4% vs. 22.3%, p=0.238). The only factor independently associated with persistent candidemia was the presence of an undiagnosed source of infection at the time of the positive control blood culture (OR 4.28, 95% CI:1.77-10.34, p=0.001), with emphasis on non-ocular complications (p=0.002), including septic thrombophlebitis, endocarditis, and involvement of other deep organs. Therefore, persistent candidemia in a center with an antifungal rationalization program is not associated with inadequate therapeutic management or a worse prognosis, but with hidden foci of metastatic infection that require specific diagnostic workup and proper management. Article 2: Retrospective cohort study including all adult patients diagnosed with candidemia between 2012-2017 in a tertiary hospital in Spain, with determination of (1,3)-ß-D-Glucan (BDG Fungitell, positive cut-off if ≥80pg/mL), in at least two clinical samples during the candidemia episode. We compared the epidemiology and outcomes of patients with all negative BDG samples versus cases with at least one positive BDG sample. A total of 148/236 (62.7%) candidemic patients had at least two BDG tests, being 26/148 (17.6%) persistently negative. In this group, Candida growth in all three sets of blood cultures collected at the time of diagnosis was less frequent (p=0.005), and the clinical presentation was less severe when assessed by their median Pitt Score (p=0.039). Furthermore, although no differences between the rates of adequate therapeutic management were observed (p=0.599), the clearing of fungemia was faster (p=0.005), the proportion of complications was lower (p=0.008), as was all-cause mortality within 30 days from fungemia (3.8% vs. 23.8%, p=0.004), and the hospital length of stay was shorter (p=0.003). In the multivariate analysis, having persistently negative BDG tests during the candidemia episode was associated with a better prognosis (OR 0.12, 95% CI: 0.03-0.49, p=0.003), possibly due to lower fungal burden. In this context, the prognostic value of this biomarker appears to be promising and may contribute to the individualization of candidemia management in the near future. Article 3: Retrospective, multicenter cohort study, evaluating 720 episodes of candidemia among adults diagnosed in tertiary hospitals between 2010-2018, being 323 patients from Spain, and the rest from Brazil. By comparing cases of candidemia between countries, we noticed that Spanish patients received treatment earlier (p<0.001), with more frequent use of echinocandins (p=0.001), and faster removal of the central venous catheter (p= 0.012). Furthermore, the overall mortality rate was lower in Spain both at 14 days (20.1% vs. 35.8%, p<0.001) and at 30 days from fungemia (31.6% vs. 51.9%, p<0.001). Among the prognostic factors independently associated with mortality, we found advanced age (p=0.001), neutropenia (p=0.001), chronic lung disease (p<0.001), use of corticosteroids (p=0.039), and inadequate therapeutic management (p<0.001). Among the protective factors, we found candidemia due to Candida parapsilosis (p<0.001), or being from Spain (p<0.001). Therefore, this study confirms lower mortality rates among adults with candidemia in Spain, and suggests that there are modifiable therapeutic management factors associated with the prognosis of these patients with room for improvement in Brazilian centers. Article 4: Retrospective multicenter study of historical trends comparing two epidemiological surveillance cohorts including adults with candidemia from eleven Brazilian centers, diagnosed in two different periods: 2010-2011 (period I) vs. 2017-2018 (period II). We identified 616 consecutive cases, being 247 from period II. Patients from the most recent period had more comorbidities (p<0.001), and more frequent previous hospital admissions (p=0.001). Although they received echinocandins more frequently (p=0.001), no changes were observed in the time to start the antifungal agent (p=0.369) or to remove the CVC (p=0.644). Despite the downward trend, the percentage of patients without treatment in both periods I and II (23.6% vs. 17.4%, p=0.07), respectively, still draws attention. Unfortunately, there was no progress over time in reducing overall mortality rates at 14 days (33.6% vs. 37.7%, p=0.343), or at 30 days from fungemia (51.4% vs. 48.6%, p=0.511). In conclusion, we found mortality rates to be still high despite greater access to echinocandins, probably due to an increase in patient complexity and delays in therapeutic interventions. Therefore, candidemia management strategies must be adapted to epidemiological changes, aiming for diagnostic agility in favor of reducing the number of untreated eligible patients, anticipating earlier effective antifungals and source control.
Descrição
Citação
AGNELLI, Caroline. Fatores epidemiológicos, clínicos e microbiológicos relacionados ao prognóstico de adultos com candidemia: estudo de coortes retrospectivas Brasil - Espanha. 2024. 163 f. Tese (Doutorado em Infectologia) - Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, 2024.