Feasibility and Acceptability of Cryptococcal Antigen Screening and Prevalence of Cryptocococcemia in Patients Attending a Resource-Limited HIV/AIDS Clinic in Malawi

Introduction

Cryptococcal meningitis causes over 500 000 deaths annually and more AIDS-related deaths than tuberculosis in sub-Saharan Africa. ¹ In sub-Saharan Africa, the majority of patients with cryptococcal meningitis die because they are diagnosed late in the course of infection. ² Yet, cryptococcal infection always begins with a preclinical stage weeks before the devastating syndrome of meningitis develops. ³ Early detection and preemptive therapy for cryptococcal infection could save thousands of lives, but the cost and complexity of available tests (latex agglutination or enzyme immunoassay) as well as the limited data on the feasibility and impact of routine screening have limited enthusiasm for widespread screening programs. ⁴ Recently, the US Food and Drug Administration cleared a rapid, low-cost, easy-to-use serum antigen test that can detect Crytococcus infection before the onset of meningitis. ⁵ Preliminary performance studies of this test have shown strong agreement with the gold standard cryptococcal antigen (CrAg) assays. ^5,6 The World Health Organization (WHO) recommends treatment for asymptomatic cryptococcal infection with oral fluconazole. ⁷ Previous studies suggest a cryptococcal screening and treatment program is cost-effective and even cost-saving (less than US$140 per life saved) in populations of >2% CrAg positivity. ⁸

Partners in Hope in Lilongwe, Malawi, conducted an evaluation to determine the prevalence of cryptococcal infection in new HIV-infected patients starting antiretroviral therapy (ART) with CD4 count <100 cells/µL or WHO clinical HIV/AIDS stage III or IV disease and to determine the acceptability and feasibility of starting a CrAg screening program.

Methods

All patients initiating ART at Partners in Hope clinic between November 4, 2013, and May 31, 2014, 18 years or older who had CD4 count <100 µL or were WHO clinical HIV/AIDS stage III or IV, and able to consent for screening were offered cryptococcal screening. All new patients were staged following guidelines for initiating ART in Malawi and tested for CD4 count. Patients were prescreened to exclude already treated transferred patients. The ART-naïve patients were screened by a nurse with an eligibility checklist. If a patient was eligible and consented for the CrAg screening, a CrAg screening test using the lateral flow assay (IMMY Diagnostics, Norman, Oklahoma) was performed on the same whole blood specimen submitted for CD4 testing. A research nurse in clinic collected baseline data including age, gender, WHO clinical HIV/AIDS disease stage, symptoms (deafness, change in vision, confusion, headache for >3 days, or fever). ⁹ Patients were not offered any incentives for screening. All patients with positive CrAg test were assessed by a clinician for symptoms of cryptococcal meningitis and if asymptomatic were offered fluconazole 800 mg orally × 2 weeks followed by 400 mg orally × 8 weeks followed by 200 mg orally until CD4 count is >200 cells/µL for 6 months in accordance with WHO guidelines. ⁷ All participants were followed for 3 months to monitor for mortality, ART use, and cryptococcal meningitis treatment. The study received ethical approval from the Malawi National Health Sciences Research Committee as well as from the University of California Los Angeles Institutional Review Board.

Results

During the 7-month study period, 552 patients presented to Partners in Hope clinic to initiate HIV care. Of those, 176 were transfer-ins already on treatment and were not offered screening. Of the 376 treatment-naïve new patients, 258 were screened for the study and 113 were eligible. All (100%) accepted CrAg screening. One patient was incorrectly staged at initial visit as WHO clinical HIV/AIDS stage III or IV. He was restaged at a subsequent visit, and was found to be WHO clinical HIV/AIDS stage I and was not included in the analysis (see Figure 1).

OPEN IN VIEWER

Figure 1.

Flowchart of patient screening, eligibility, and enrollment, Partners in Hope clinic, Malawi, 2013-2014. *Patients not screened due to unavailability of research nurse.

Of the 113 patients who underwent CrAg screening, there were 53 females (46.9%) and 60 males (53.1%). Of those, 8 were WHO clinical HIV/AIDS stage I patients (7.1%), 5 WHO clinical HIV/AIDS stage II (4.4%), 80 WHO clinical HIV/AIDS stage III (70.8%), and 20 WHO clinical HIV/AIDS stage IV (17.7%; Table 1). The mean age was 36.4 years (standard deviation [SD] 8.5); the mean CD4 count was 124 cells/µL (SD 144; Table 2).

OPEN IN VIEWER

Table 1.

Conditions Associated with WHO Clinical HIV/AIDS Grading, Cryptococcal Screening Program Patients, Partners in Hope clinic, Malawi, 2013 to 2014.^a

Reason for WHO Clinical HIV/AIDS Stage as Determined by Malawian Clinical Officers in Treatment-Naïve Patients Eligible for Cryptococcal Screening
Diagnosis	# Patients (total = 113)
Severe weight loss	40
Chronic diarrhea	15
Severe bacterial infection	16
Persistent fever	8
Tuberculosis (extrapulmonary)	5
Anemia	4
Moderate weight loss	3
Persistent oral thrush	3
Anal fistula—is this WHO ¾?	2
PCP	2
Kaposi sarcoma	2
Otherb	7
Not specifiedc	13

Abbreviations: PCP, Pneumocystis Pneumonia; WHO, World Health Organization.

^aIndividuals can have more than 1 WHO clinical HIV/AIDS stage III or IV condition.

^bPersistent oral thrush, herpes zoster, pulmonary tuberculosis, cerebral or B-cell Hodgkin’s lymphoma, esophageal candidiasis, lymphadenopathy, and necrotizing ulcerative stomatitis.

^cThis includes patients included in the study for CD4 count <100 cells/µL, who were not WHO clinical HIV/AIDS stage III or IV.

OPEN IN VIEWER

Table 2.

Characteristics of Treatment-Naïve, Newly Enrolled Patients at a Malawian HIV/AIDS Clinic and CrAg Prevalence with CD4 Count <100 Cells/µL or WHO Clinical HIV/AIDS Stage III or IV.

	#	%	Frequency of CrAg+	CrAg + prevalence, %
Total enrolled	113	100	2	1.8
Sex
Female	53	46.9	1	1.9
Male	60	53.1	1	1.7
Age, years
<30	22	19.5	1	4.5
≥30	91	80.5	1	1.1
WHO clinical HIV/AIDS stage
I	8	7.1	0	0
II	5	4.4	0	0
III	80	70.8	1	1.3
III	20	17.1	1	5.0
III and IV	100	88.4	2	2.0
CD4 count ≥100 cells/µL	56	49.6	0	0
CD4 count <100 cells/µL	57	50.4	2	3.5
Symptoms
Deafness	3	2.7	0	0
Vision changes	10	8.9	1	10.0
Confusion	3	2.7	0	0
Headache	22	19.5	1	4.5
Fever	38	33.6	0	0

Abbreviations: CrAg, cryptococcal antigen; WHO, World Health Organization.

Of the 113 screened patients, 2 patients tested CrAg positive. One patient reported a change in vision and headache to the screening nurse and the other was asymptomatic. Both patients had a CD4 count <100 cells/µL and both were started on oral fluconazole after being evaluated by a clinician (the originally symptomatic patient was determined to be asymptomatic by the clinician). Both patients were retained in the 6-month follow-up visit, at which point, the patient initially presenting with headache developed cryptococcal meningitis and was hospitalized for intravenous amphotericin treatment.

The overall CrAg prevalence for all patients in the study was 2/113 or 1.8% (95% confidence interval [CI]: 0-4.24), the prevalence for patients with CD4 count < 100 cells/µL was 2/57 (3.5%; 95% CI: 0-8.4), and the prevalence among those with WHO clinical HIV/AIDS stage III was 1/80 (1.3%) and with WHO clinical HIV/AIDS stage IV was 1/20 (5.0%).

Discussion

We implemented a routine screening program for CrAg detection in a resource-limited clinic in Malawi. Patients had excellent uptake of CrAg screening: 100% of patients who were identified as eligible and offered screening accepted CrAg testing. A basic screening questionnaire administered by nurses allowed simple, risk stratification of patients and initiation of fluconazole. Our findings suggest CrAg screening can be implemented in settings with nurses and minimal backup from higher level clinicians like clinical officers and medical doctors.

We found a higher CrAg prevalence among patients with CD4 count < 100 cells/µL as compared to those with CD4 count ≥100 cells/µL (3.5% vs 0%) and combined higher prevalence for WHO clinical HIV/AIDS stages III and IV versus stages I and II (2% vs 0%). In our study, we found 56 patients who had CD4 count >100 cells/µL yet were WHO clinical HIV/AIDS stage III or IV. That is a surprisingly high number of patients with advanced WHO clinical HIV/AIDS stages with CD4 counts >100 cells/µL and may explain why the cryptococcal prevalence was lower in advanced WHO clinical HIV/AIDS stage patients than expected in our study. Of note, our laboratory does participate in United Kingdom National External Quality Assessment Service with successful performance, and therefore our CD4 counts are likely a valid result. Most of those advanced WHO clinical HIV/AIDS stage patients were staged based on weight loss (22 patients), diarrhea (8 patients), severe bacterial infection (10 patients), and tuberculosis. ⁵ The rest of the patients were found to have anemia (2 patients), anal fistula (2 patients), persistent oral thrush (1 patient), lymphoma (1 patient), and necrotizing ulcerative stomatitis (1 patient). Countries like Malawi may move toward elimination of all CD4 testing in the future and adopt a “test-and-treat” strategy where all newly identified HIV-infected patients are offered treatment. It may be necessary to use clinical staging to determine the cost-effectiveness of CrAg screening among groups at risk based on clinical criteria alone. The prevalence of CrAg positivity in this study for WHO clinical HIV/AIDS stage III and IV was 2%, and the prevalence rate for only WHO clinical HIV/AIDS stage IV was 5%, and both frequencies are above the thresholds found to be cost-effective in other studies. ⁸

By screening and treating those who are asymptomatic but CrAg positive, the morbidity and mortality of cryptococcal infection were shown in prior studies to be reduced. In Uganda, death from cryptococcal meningitis was prevented in 71% of those treated preemptively as compared to 0% in those not receiving preemptive treatment. ⁴ In another study done in Kenya, benefit was not seen in early intervention among individuals with asymptomatic cryptococcemia, however, that lack of benefit may be secondary to lack of adequate screening for cryptococcal meningitis and low rates of initiation of fluconazole, thus highlighting the need for continued studies on optimal models for implementation. ¹⁰ The consequences of correct treatment for symptomatic patients were highlighted in the 1 patient in our study, who presented with central nervous system complaints but was deemed on clinical evaluation to be asymptomatic and treated with oral fluconazole.

In Malawi, the cost of screening 1 individual is approximately US$2.50 and includes the cost of the test and human resources to perform the testing. Using the prevalence we found in Malawi of 3.5% in clients with CD4 count < 100 cells/µL, the number needed to test would be 29 patients with CD4 count < 100 cells/µL to find one positive result. That would cost about US$72.50. Using the prevalence of 2.0% for patients with WHO clinical HIV/AIDS stages III and IV, 50 patients would need to be screened to find one positive result, which would cost about US$125. In contrast, the cost of treatment for 1 patient with cryptococcal meningitis in Malawi is US$734. That cost includes amphotericin and admission with standard room charges for a paying patient. In many resource-limited settings, amphotericin is not available, so screening in these locations has the added benefit of avoiding death due to lack of optimal therapy. The cost to identify early infection and therefore save one life if all patients with CD4 count < 100 cells/µL were screened at a prevalence rate of 3.5% would be US$100.60, and for patients with WHO clinical HIV/AIDS stages III and IV at a prevalence rate of 2.0% would be US$176.06. Even at a prevalence rate of 2%, there is cost savings due to the high cost of treating advanced cases of cryptococcal meningitis.

Conclusion

Cryptococcal screening in a resource-limited HIV clinic was acceptable and feasible, with high uptake by patients. The prevalence for asymptomatic cryptococcal infection was higher in patients screened based on their CD4 count than those screened based on WHO clinical HIV/AIDS staging.