Cryptococcal antigenemia and associated risk factors among ART‐naïve and ART‐experienced HIV‐infected peoples at selected health institutions of Mekelle, Northern Ethiopia

Abstract Cryptococcal infection is a major cause of opportunistic infection in HIV/AID‐infected peoples. We determined cryptococcal antigenemia and cryptococcal meningitis among antiretroviral therapy (ART) initiated and ART‐naive HIV‐infected peoples. A cross‐sectional study was conducted at selected health facilities in Mekelle, Ethiopia. Blood was collected to determine CD4 and plasma cryptococcal antigen (CrAg). CSF CrAg and CSF culture and urease tests were also done. Socio‐demographic and clinical data were collected using a structured questionnaire and clinical chart review. From the enrolled study participants, 267 study participants had complete data, of which, 137 (51%) were females. From the study participants, 140 (52%) and 127 (48%) were ART experienced and ART naïve, respectively. The prevalence of cryptococcal antigenemia was 9 (3.4%). All the study participants, except one (CD4 = 120 cells/mm3), had CD4 count less than 100 cells/mm3. From CrAg‐positive peoples, 6 (4.7%) were ART naïve. Five CrAg‐positive peoples had cryptococcal meningitis. Being male, rural residence, and being hospitalized were associated with cryptococcal antigenemia. Cryptococcal infection poses a substantial risk of HIV‐positive peoples. This study provides relevant data for CrAg screening interventions in patients with low CD4 cell counts.

Testing for cryptococcal antigenemia is effective at reducing morbidity and mortality because patients who receive early antifungal treatment have better outcomes than those who receive delayed treatment (Kaplan et al., 2015;World Health Organization, 2011).
CrAg screening programs have begun in a number of countries with high burden of disease (Kaplan et al., 2015). However, in Ethiopia, a few ART clinics have started CrAg screening procedures.
Generating evidence through research will support the existing prevention and control programs (Beyene, Woldeamanuel, Asrat, Ayana, & Boulware, 2013;Reepalu et al., 2015). In Ethiopia, the mortality rate due to cryptococcal meningitis is high (Berhe, Melkamu, & Amare, 2012). High prevalence of cryptococcal antigenemia has also been reported in ART sites in Addis Ababa and Adama (Alemu et al., 2013;Beyene et al., 2013Beyene et al., , 2017. In contrast to these studies, one study has reported a low prevalence of cryptococcal infection in Adama (Reepalu et al., 2015). The existing studies in Ethiopia have shown a varying prevalence of cryptococcal antigenemia.
Nevertheless, further assessment in different geographical regions is needed to address associated factor differences. The previous studies were also unable to assess the magnitude of cryptococcal meningitis. Therefore, this study is aimed at determining the cryptococcal antigenemia and associated factors among ART naïve and ART-experienced HIV-infected peoples.

| Study setting, population, and design
A cross-sectional study was conducted at Mekelle, Ethiopia, from September 2016 to January 2017. Mekelle is situated 783 km north of the capital of Ethiopia, Addis Ababa. In the city, there are governmental and private health facilities providing HIV diagnostic and management clinics. This study was done among peoples who get ART-related healthcare service (both outpatients and inpatients during the study period) from Mekelle Hospital, Ayder Comprehensive Specialized Hospital, and Mekelle Health Center. All HIV-positive individuals aged 18 years and above with CD4 count less than or equal to 250cells/mm 3 were eligible for this study. Peoples who had been diagnosed with a cryptococcal infection in the previous 2 years, those who were taking antifungal treatments, and severely ill patients that did not give consent were excluded from the study. A total of 280 study participants were included, considering 84% power to detect a 10% difference, with equal group sizes and two-sided p = 0.05 and prevalence (P) from Adama (Sawadogo et al., 2016).

| Data collection
A structured questionnaire was prepared to collect socio-demographic data (age, gender, and residence) and clinical data (a headache, fever, nausea, cough, vomiting, visual changes and neck stiffness). History record cards were also reviewed for length of ART stay.

| Ethics approval and consent to participate
The study protocol was evaluated and approved by the Research Ethics Review Committee of the College of Health Sciences, Mekelle University, and ethical clearance was obtained. Official support letters were obtained from Mekelle University and Tigray Regional Health Bureau. Moreover, prior to conducting the study, the purpose and objective of the study were described to the participants and a written informed consent was obtained. Laboratory examinations with a direct benefit in the health of the study participants were informed to physicians. The consent involves permission to disseminate the findings of the study through the scientific workshop and publish in reputable journals.

| Laboratory testing
All laboratory tests for plasma CrAg test, CD4 testing, CSF CrAg test, and culture test were done at the time of data collection.

| Rapid CrAg lateral flow assay (LFA) test
Plasma and CSF samples were used to detect cryptococcal antigenemia and cryptococcal meningitis, respectively. Cryptococcal Antigen Lateral Flow test (ImmunoMycologics, Inc., Oklahoma, USA) was performed, and results were observed within 10 min and interpreted as per the manufacturer's instruction.

| Culture for definitive diagnosis of cryptococcal meningitis
CSF was centrifuged at about 250 g for 10 min. The sediment was inoculated to Sabouraud Dextrose agar on two sterile petri dishes and was incubated at room temperature and at 37°C for up to 14 days.
A moist white-cream colored mucoid colonies appeared 3-4 days of incubation. For positive cultures, urease test was carried out on urea agar in a sterile test tube. A colony from the culture was inoculated and incubated at 37°C for 3-5 hr.

| Data quality assurance
Reagents were checked for proper functioning using provided control reagents. CSF was collected aseptically to prevent contamination using 70% alcohol and 2% tincture of iodine. Sterility of media was checked by incubating random samples of prepared culture media at room temperature and at 37°C for 14 days.

| Statistical analysis
The data were entered into the EPI Info version 7 (CDC, USA) every day. The data were imported from EPI Info and analyzed using    Figure 1).

| Prevalence of cryptococcal antigenemia
The prevalence of cryptococcal antigenemia was 9 (3.4%). All the study participants, except one (CD4 = 120 cells/mm 3 ), had CD4 count less than 100 cells/mm 3 . The cryptococcal antigenemia-positive peoples had one or more clinical sign and symptoms, of which, 6 (6.9%) had a headache, 4 (5.2%) had a fever, and 3 (9.7%) had neck stiffness. Six of the CrAg positives were ART naive, and 3 were ART experienced. From the total study participants with a CD4 count less than or equal to 50 cells/mm 3 , 6 (9.8%) were with cryptococcal antigenemia. Considering their ART status, among the ART-naïve groups from those with a CD4 count of less than or equal to 50 cells/ mm 3 , 4 (12.9) were positive for Cryptococcus. However, only one study participant with a CD4 count of 120 cells/mm 3 was positive to Cryptococcus. Eight of the nine CrAg positive were on WHO stage IV of HIV/AIDS (Table 1).

| Cryptococcal meningitis among cryptococcal antigenemia-positive peoples
Cryptococcal antigen-positive peoples were examined for signs and symptoms of meningitis. From 9 peoples with antigenemia, 5 from ART naïve and 1 from ART-experienced groups had one or more signs and symptoms indicative for cryptococcal meningitis. Lumbar puncture was performed on those 6 study participant, of which, only 5 were positive for CSF LFA CrAg test. All CrAg positives were confirmed by culture and biochemical test (Figure 1).

| Factors associated with cryptococcal infection
Among risk factors, being male (p = 0.017), living in rural areas (p = 0.008), and being hospitalized (p = 0.001) were statistically associated with cryptococcal antigenemia (Table 2). The prevalence of cryptococcal antigenemia in the current study is relatively higher than the prevalence reported from Adama (1.6%) (Reepalu et al., 2015) and much lower than the findings from Addis Ababa (8.4%) (Alemu et al., 2013)  TA B L E 2 Associated risk factors with cryptococcal antigenemia (n = 267) (10.2%) (Beyene et al., 2013). The difference in the prevalence rate might be due to improvements in a time of ART initiation for HIV-infected peoples, which might reduce the incidence and high mortality associated with cryptococcal meningitis and other opportunistic infections (Alemu et al., 2013). However, in contrast to these, in a recently conducted study on HIV-initiated peoples with a CD4 count of less than 150 cells/mm 3 showed a prevalence of 6.2% (Beyene et al., 2017). The difference could be due to a large number of study participants with lower CD4 count are included (Beyene et al., 2017).

| D ISCUSS I ON
We compared our findings with other studies conducted outside Ethiopia. The findings from this study were similar to recent prevalence reports from Namibia (3.3%) (Sawadogo et al., 2016) and Tanzania (3.7%) (Letang et al., 2015) and slightly lower than the prevalence reported in South Africa (4.3%) (Longley et al., 2016).
Considering cryptococcal meningitis (CM), in this study, high CM rates among ART-naïve study participants were found. This is similar to other studies in Sub-Saharan Africa. For example, in South Africa, CM was found 4/10 antigenemia-positive cases (Longley et al., 2016), in Cameroon 41/146 of patients who had clinical signs of meningitis (Ngouana et al., 2015), and in Tanzania 15/17 cryptococcal antigenemia-positive hospitalized patients (Wajanga et al., 2011).
But we are unable to discuss further the findings from a statistical point of view due to the few numbers of study participants with CM.
However, in this study males, residing in rural areas, CD4 < 100 cells/mm 3 , and being hospitalized were statistically associated with cryptococcal antigenemia. Although cryptococcal infections occur in both sexes, a statistically significant difference in gender was also observed in other studies; this is likely due to poor health-seeking behavior of men to present later to care (Alemu et al., 2013;Assogoba et al., 2015;Ngouana et al., 2015) and the interaction Cryptococcus with testosterone, which results in increased capsular polysaccharide release and Cryptococcus-mediated macrophage death (McClelland et al., 2013). In addition to this, rural residents were at risk for cryptococcal infection. This might be supported by genotyping studies that dealt with the ubiquitous nature of Cryptococcus species providing strong evidence for additional origins of exposure like different species of plants, soil, and house dust (Chen et al., 2015;Cogliati, 2013 Makadzange & McHugh, 2014). In addition to this, those that were hospitalized for different reasons were at a significant risk of cryptococcal infections (Kaplan et al., 2015). In this study, higher prevalence of CrAg was observed in peoples with advanced immune-suppression, which is supported by other similar studies, and WHO recommendations (Alemu et al., 2013;Ganiem et al., 2014;Osazuwa et al., 2012;Pongsai et al., 2010;Randhawa et al., 2011;Sawadogo et al., 2016;Smith et al., 2013;World Health Organization, 2011). Moreover, though a significant association was not found on their ART status, higher prevalence of Cryptococcus was observed in those that were ART naïve. Preventing new HIV infections, early HIV diagnosis, implementing early linkage to care, and ensuring timely initiation of ART with strict adherence would prevent incident cryptococcosis including Cryptococcal antigen screening and preemptive therapy with fluconazole (Asfaw et al., 2015;Smitson et al., 2014). Therefore, resource-constrained settings, including Ethiopia, should work for improving their national program for the early ART initiation and promoting a cost-effective preventive strategy to reduce the incidence and high mortality associated with cryptococcal meningitis in peoples with low CD4 count (Jarvis et al., 2013).
The limitations of this study are as follows: we were unable to see the independent effect of variables on cryptococcal antigenemia. It was also difficult to perform statistical analysis for Cryptococcal meningitis due to small sample size.

| CON CLUS ION
This study revealed that cryptococcal infection poses a substantial risk of HIV/AIDS-infected peoples despite the widespread use of HAART. High prevalence of cryptococcal infection in ART-naïve peoples was observed. Moreover, being male, living in rural areas, low CD4 count, and being hospitalized were statistically associated with cryptococcal infection. This provides relevant data for CrAg screening interventions in patients with low CD4 cell counts.

ACK N OWLED G M ENTS
We would like to thank Clinton Health Access Initiative for donating LFA CrAg test kits and to Ayder Comprehensive Specialized Hospital for allowing us to do the research in the laboratory and for the logistic support. Our great appreciation also goes to healthcare workers of the study sites and to all study participants.

CO N FLI C T O F I NTE R E S T
The authors have no competing interests to declare.

AUTH O R CO NTR I B UTI O N
HK participated in the conception of the title and study design, data collection, data analysis, data entry, and writing the manuscript. NS participated in the study during design and analysis. KH participated in data cleaning. AM was involved in the design and study monitoring. All authors read and approved the final manuscript.

DATA ACCE SS I B I LIT Y
The data will be available on request from the corresponding authors following permission from Research Ethics Review Committee of the College of Health Sciences, Mekelle University.