Brentlinger PE, Behrens CB, Kublin JG. Challenges in the Prevention, Diagnosis, and Treatment of Malaria in Human Immunodeficiency Virus–Infected Adults in Sub-Saharan Africa. Arch Intern Med. 2007;167(17):1827-1836. doi:10.1001/archinte.167.17.1827
Copyright 2007 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2007
Many countries in sub-Saharan Africa currently report high prevalences of both human immunodeficiency virus (HIV) and Plasmodium falciparum malaria. The likelihood of HIV-malaria coinfection may affect clinical management of patients. The extent to which standard clinical guidelines address HIV-malaria coinfection is unclear.
We reviewed standard World Health Organization and other guidelines for diagnosis and treatment of malaria and/or HIV-related illness. We also searched PubMed (1990 to present) for literature on HIV-malaria interactions and treatment of coinfection. We restricted our review to the situation of the nonpregnant HIV-infected adult.
We found only 6 articles describing the clinical presentation of HIV-malaria coinfection in adults. We also identified 10 clinical or laboratory syndromes that are shared by malaria and AIDS-related conditions and that might provoke diagnostic confusion. We identified 12 antimalarial medications whose coadministration with antiretrovirals is known or suspected to result in drug-drug interactions or overlapping toxicities.
Substantial overlap in the clinical and laboratory characteristics of malaria and HIV-related syndromes generates potential difficulties in AIDS staging and in diagnosis and management of patients at risk for coinfection. Significant drug-drug interactions and overlapping drug toxicity profiles further complicate concurrent management of malaria and HIV. Standard clinical guidelines do not reflect the full complexity of the interactions and overlaps between the 2 infections. Clinicians who manage HIV-infected patients in malaria-affected regions should systematically consider malaria when evaluating patients with a broad spectrum of symptoms. Further research is urgently needed to define best practices for prevention, diagnosis, and management of HIV-malaria coinfection in this region.
Plasmodium falciparum and human immunodeficiency virus type 1 (HIV) intersect widely. Each pathogen infects millions annually in Africa alone (Table 1).1,2 Coinfection may accelerate the spread of both diseases.3
In our own field experience, the differential diagnosis of malaria and HIV-related syndromes has proven challenging, and neither textbooks nor local guidelines have been of sufficient help. Therefore, we conducted a literature review with the primary aim of describing the challenges that HIV-malaria coinfection presents to clinical care. We focused on P falciparum in the nonpregnant adult because children, pregnant women, and patients with non–P falciparum malaria often require different approaches to diagnosis and treatment.4- 6 Because most available clinical data on HIV-malaria coinfection were acquired in sub-Saharan Africa, our discussion will center on this region.
We reviewed and compared World Health Organization (WHO) and 2 other widely used guidelines (Médecins sans Frontières; International Center for AIDS Care and Treatment Programs) for management of malaria and AIDS in low-income settings.7- 15 We also searched the primary English-language scientific literature on HIV-malaria coinfection, using the terms HIV, AIDS, malaria, plasmodium, viral load, CD4, antiretroviral, and antimalarial, using the PubMed database for 1990 to the present. We also explored references cited in articles and guidelines. We categorized key clinical syndromes on the basis of internal consensus among ourselves. We consulted review articles and standard guidelines for information on drug toxicities and drug-drug interactions, but we did not review the primary pharmacologic literature.
We identified 6 studies describing the clinical presentation of acute malaria in HIV-infected adults.16- 21 We also identified 10 syndromes that are shared by malaria and AIDS-related conditions. We identified 12 antimalarial medications whose coadministration with antiretrovirals is known or suspected to result in drug-drug interactions or overlapping toxicities. Relevant clinical concerns are summarized in subsequent sections and in Table 2.
Acute malaria elevates HIV viral load significantly, approaching 1 log in some subgroups and persisting for as long as 8 weeks, even after effective antimalarial treatment.22,45 Potential mechanisms include increased HIV replication, up-regulation of HIV coreceptors, and dendritic cell activation.46- 50
Malaria may increase the risk of HIV transmission. Although increased sexual transmission of HIV resulting from malaria-related viral load elevations has not yet been characterized prospectively, it is likely to occur.23,24 The public health impact may be substantial because of the high prevalence of malaria and the prolonged nature of observed HIV viral load elevations.3 Malaria-associated immune activation may also render individuals with malaria more susceptible to the acquisition of HIV.
Malaria appears to reduce the CD4 count. One observational study26 described an excess decline in CD4 count of approximately 40/μL per year for each episode of acute malaria in HIV-infected persons. A prospective study51 found significant increases in mean CD4 counts (297 to 447/μL) and decreases in the proportion of patients with a CD4 count less than 200/μL (28.7% to 13.2%) after successful malaria treatment in HIV-infected individuals. Findings in adults not infected with HIV have been similar.52
Considerations are as follows:
CD4 count measurement during or shortly after acute malaria may lead to misclassification of disease stage and premature initiation of antiretroviral treatment (ART). No guideline recommended postponement of routine CD4 measurements in the aftermath of malaria.
Malaria-induced declines in CD4 count and elevations in viral load might also result in erroneous diagnosis of ART regimen failure and premature switching to second-line regimens. However, no available data describe CD4 and viral load response to acute malaria in patients taking antiretroviral drugs.
Early studies of malaria incidence in HIV-infected persons failed to find significant interactions, but these studies were thought to have methodologic limitations.53,54 More recent prospective cohort studies have demonstrated rising malaria incidence as CD4 counts decline in HIV-infected adults living in regions of stable malaria transmission.55,56 The observed increase is greatest in the lowest CD4 strata. A study in Uganda documented an odds ratio of 6.1 for clinical malaria in persons with a CD4 count less than 200/μL, as compared with those with a CD4 count of 500/μL or more.56 Increased malaria severity has also been observed in HIV-infected persons living in regions of unstable malaria transmission.16,17,57 In a study in South Africa, the odds ratio for association of HIV infection and fatal malaria was 7.5.16 The incidence of malaria in HIV-infected persons is estimated to be higher than that of cryptococcal disease, invasive pneumococcal disease, herpes zoster, toxoplasmosis, and Pneumocystis pneumonia.58
The association between malaria and advancing immunosuppression in the HIV-infected adult is not reflected in current guidelines for staging of AIDS. Recurrent respiratory tract infections and severe bacterial infections are WHO stage 2 and 3 criteria, respectively, but recurrent and/or severe malaria are not considered.7 We are aware of only 1 published study that regarded severe malaria as an AIDS-defining opportunistic disease (OD).59
Increased incidence and severity of malaria infection are likely to augment the overall burden of morbidity and mortality in HIV-infected persons. The consequences of malaria may include anemia, lost work and school time, chronic neurologic deficits, and death.12 Estimated country-specific HIV-related increases in malaria mortality in sub-Saharan Africa range from 0.65% to 114%.60
Frequent prescribing of antimalarial medications may increase the risk of adverse drug reactions and the difficulty of diagnosing them. There is increased risk of adverse reactions to sulfonamides (eg, sulfadoxine-pyrimethamine) with advancing AIDS.61 Data on the pharmacodynamics of antimalarial compounds in adults with severe malnutrition or wasting are unavailable, although data on severely malnourished children suggest that alterations in metabolism do occur.62,63 Adverse reactions to antimalarials may be difficult to distinguish from opportunistic infections (OIs), from adverse reactions to antiretrovirals, and from the symptoms of malaria itself.64 Some overlapping drug toxicity profiles that may cause diagnostic confusion are described in Table 3. However, few published studies have quantified the likelihood of adverse effects of antimalarial medications in adult patients with AIDS, and one recent review observed that the methods used in antimalarial drug trials did not consistently support rigorous reporting of adverse drug effects.73 The safety of antimalarial regimens for HIV-infected infants and pregnant women is even less well established.74 Suggested guidelines for pharmacovigilance in Africa reiterate the need for further research.14,75
Frequent prescribing of antimalarial medications may increase the risk of drug-drug interactions. For example, because HIV protease inhibitors inhibit cytochrome P450 enzymes, coadministration of these agents with halofantrine hydrochloride may result in cardiotoxic effects. Coadministration of amodiaquine, artesunate, and efavirenz has resulted in hepatotoxic effects in healthy volunteers; significantly impaired metabolism of amodiaquine (likely to increase risk of adverse effects) has been observed in the presence of efavirenz, saquinavir, tipranavir, and ritonavir in clinically relevant concentrations and in individuals with common polymorphisms in CYP2C8.71,76 Conversely, the nonnucleoside reverse transcriptase inhibitors may compromise the therapeutic efficacy of quinine through induction of the cytochrome P450 system. While the predicted outcomes of some pharmacokinetic interactions have been described, the full extent of this problem is unknown because published data on interactions between antiretrovirals and antimalarials—particularly artemisinin-based combination regimens—are scarce.77,78 Indeed, standard WHO guidelines for conduct of antimalarial drug efficacy trials suggest exclusion of patients taking other medications on a long-term basis.79 In Table 4, we describe selected drug-drug interactions involving important antimalarials, antiretrovirals, and other agents. Again, the published data seldom provide quantitative estimates of risk.
Frequent prescribing of antimalarial medications may speed the development of antimalarial drug resistance.86- 88 Malaria resistance to preventive co-trimoxazole (a combination of trimethoprim and sulfamethoxazole) may also develop, although it has not yet been documented in vivo.85,89,90
Antiretroviral agents may play a future role in malaria prevention and treatment. Recent investigations suggest that some antiretrovirals may possess antimalarial properties and may even act synergistically with antimalarials against P falciparum.91- 96 If so, continued ART may provide some antimalarial protection for patients whose co-trimoxazole prophylaxis is suspended after immune reconstitution occurs. However, the effectiveness of ART for long-term malaria prophylaxis (or treatment) has not yet been established in clinical trials.
Fever is the most widely recognized manifestation of malaria. Fever may also be caused by HIV itself, OI, or adverse drug reaction. Several studies of the causes of fever in HIV-infected persons living in regions of stable malaria transmission have found that the majority of episodes of fever are not caused by malaria.30,97,98
Related concerns are as follows:
The common practice of prescribing antimalarials for all episodes of fever in regions where malaria is endemic is likely to lead to both overtreatment of malaria and underdiagnosis of other treatable causes of fever.11,38,99,100 Presumptive malaria treatment has been justified by the scarcity of clinical laboratory facilities, the potential lethality of malaria, and the poor perceived predictive value of negative results of malaria smears in high-prevalence areas. However, WHO now recommends that all health facilities engaged in HIV/AIDS care be equipped to test for malaria. The guidelines we reviewed gave varying recommendations: one suggested laboratory confirmation of malaria in patients with fever and no other identifiable source11; another recommended presumptive treatment of all fevers with antimalarials in areas of high-level malaria transmission.8 No AIDS-specific guideline mentioned that asymptomatic malaria parasitemia may coexist with other febrile conditions, and that the presence of malaria parasitemia therefore does not confirm malaria as the sole source of a patient's fever.
Malaria causes anemia through several mechanisms, including increased destruction of infected erythrocytes and postinfection impairment of erythropoiesis.101,102 Ten percent of adults with severe malaria are reported to have hemoglobin levels of 7 g/dL or less on hospital admission overall.12 Infection with HIV itself may cause anemia, as do many OIs and ODs.32,103 Finally, several medications used to treat AIDS and its complications can cause anemia.36,104
Concerns are as follows:
Where malaria transmission occurs, the overall burden of anemia is likely to be greater, with resultant decreases in work capacity, increases in maternal mortality, and, perhaps, faster progression of HIV disease.105,106 Anemia is an independent predictor of mortality in patients with AIDS in both developed-country and low-income settings.103,107
Neither the absence of parasitemia nor the persistence of anemia after antimalarial treatment rules out malaria as a cause of anemia. Malaria-related anemia can progress for up to 2 weeks after clearance of parasitemia.102 Unexplained anemia unresponsive to malaria treatment is now an indicator of stage 3 AIDS,7 but no guideline defined the anticipated time course for hematologic recovery after a malaria episode. The standard primary care–oriented guideline8 did recommend presumptive treatment for malaria if otherwise unexplained anemia occurs in a patient at risk for malaria, and the standard WHO reference on severe malaria12 discusses severe malaria-related anemia in detail.
The presence of malaria parasitemia does not rule out other causes of anemia. The etiology of anemia is commonly multifactorial.108 Some,9,109 but not all, current guidelines acknowledge the difficulty of distinguishing anemia caused by medication toxicity (eg, zidovudine) from anemia caused by infection.
Pulmonary symptoms have been described in 3% to 10% of patients with P falciparum malaria.110 Cough occurs in up to 50% of adults with malaria, and pulmonary edema has been called “a grave and usually fatal manifestation of severe falciparum malaria in adults.”12,111 However, cough and/or respiratory distress may also be signs of OI/OD or immune reconstitution inflammatory syndrome.
Malaria is not included in common clinical guidelines or algorithms that address the differential diagnosis of respiratory symptoms in HIV-infected persons. Only the monograph on severe malaria12 discussed the association between malaria and respiratory distress in detail, and no HIV-specific algorithm for diagnosis of respiratory complaints mentioned malaria. In contrast, there is ample pediatric literature describing overlapping presentations of malaria and pneumonia in children younger than 5 years.112
Nausea, vomiting, splenomegaly, hepatomegaly, jaundice, abdominal pain, and transaminase level elevations have all been reported in patients with acute malaria. Similar gastrointestinal signs and symptoms may be the result of OI/OD, adverse reactions to antiretrovirals, or immune reconstitution inflammatory syndrome.
Malaria is not included in common clinical algorithms or guidelines that address the differential diagnosis of gastrointestinal symptoms in HIV-infected persons. The standard monograph on severe malaria12 was the only document to describe the association between gastrointestinal symptoms (eg, vomiting and jaundice) and malaria in detail. One other guideline9 recommended (in a footnote only) consideration of malaria in patients with high fever and abdominal pain. The other documents we reviewed did not consider malaria in the differential diagnosis of gastrointestinal or hepatobiliary complaints.
Neurologic signs and symptoms of malaria include headache, mental status changes, coma, and seizures.43 Headache occurs with acute malaria episodes in regions of both stable and unstable transmission; the more serious syndromes—hallmarks of severe malaria—occur (in adults) more frequently in regions where malaria transmission is unstable. Most8,9,11,12 of the guidelines we consulted recommended consideration of malaria in the HIV-infected person with severe neurologic symptoms.
Other concerns are as follows:
Severe cerebral malaria is commonly treated with medications that may be incompatible with antiretrovirals, or whose safety in this context is unknown. Quinine is considered incompatible with certain antiretrovirals.77 Artesunate (usually in combination with a second agent) is likely to replace quinine soon for the treatment of severe malaria, but no published studies describe its compatibility with antiretrovirals.113
Immunosuppression related to HIV may render previously immune adults vulnerable to severe malaria. The immune response to malaria is mediated by CD4 cells.114,115 Thus, adults who reside in areas of stable transmission may lose their acquired immunity with advancing HIV-related immunosuppression. Patients in AIDS stages 3 and 4 who are not taking co-trimoxazole prophylaxis, are not using insecticide-treated bed nets, and do not have reliable access to medical attention would presumably be at highest risk. Early studies that failed to discover associations between cerebral malaria and HIV in regions of stable transmission did not control for HIV disease stage.116,117
Lactic acidosis is a complication of severe malaria and a poor prognostic indicator.12,118
No HIV-specific guideline mentioned that the differential diagnosis of lactic acidosis should include severe malaria. Lactic acidosis was described only in the standard monograph on severe malaria,12 although the primary-care guideline8 mentioned that rapid, deep respirations could be a sign of malaria. The HIV-specific guidelines9,10,109 mentioned lactic acidosis only in the context of adverse drug reaction. Widespread use of stavudine in resource-limited settings increases the likelihood that this potentially fatal complication will occur.39,40
In one recent Kenyan study,18 87.7% of adults not infected with HIV had an adequate clinical and parasitologic response to sulfadoxine-pyrimethamine monotherapy at 28 days, compared with only 68.0% of HIV-infected patients with a CD4 count less than 200/μL (P < .001). Molecular methods were used to confirm that approximately one-third of the treatment failures were caused by reinfection. A study in Zambia observed similar outcomes, but without quite achieving statistical significance.19
Thus, HIV-infected patients with acute malaria may require increased posttreatment vigilance to detect treatment failure. Higher likelihood of treatment failure also implies greater exposure of HIV-infected persons to second-line and even third-line antimalarials, with the attendant risks of drug interaction or adverse reaction. Because exposure to malaria is so frequent in regions of intense perennial transmission, the existing WHO recommendation that patients be screened and treated for malaria109 before starting ART does not address this question. The increases in HIV viral load attendant to malaria may also be sustained for longer periods in the presence of decreased antimalarial efficacy, resulting in increased transmission of HIV.3
Our review of existing guidelines and pertinent literature leads us to conclude that the HIV-infected patient residing where malaria is endemic or epidemic may be at substantial risk of misdiagnosis and mismanagement for 5 reasons. First, the heterogeneous clinical manifestations of both malaria and HIV-related disease render clinical decision making difficult even for experienced clinicians. Second, both asymptomatic and symptomatic malaria may occur concurrently with OIs and/or adverse drug reactions, thus complicating diagnosis and management. Third, the nonoverlapping and oversimplified nature of some vertical guidelines is likely to lead to errors of commission (eg, overtreatment of malaria in febrile patients) and omission (eg, failure to consider malaria in the patient with respiratory or gastrointestinal symptoms or lactic acidosis). Fourth, the published literature is as yet inadequate to guide clinicians and patients who wish to estimate the risks of concurrent administration of antimalarials and antiretrovirals (or other agents). Finally, inadequate access to clinical laboratory facilities places patients with AIDS at high risk of receiving antimalarial medications presumptively, even when their symptoms are caused by other entities.
We concur with WHO's recommendations for harmonizing and improving HIV-malaria policy,14 but we would propose amplifying and/or modifying them as follows.
Intensified efforts to reduce malaria transmission should be conducted where both malaria and HIV infection are prevalent. Insecticide-treated bed nets should be made widely available at the lowest possible cost (preferably none).119
Malaria-HIV research priorities should include the following:
Description of longitudinal response of CD4 count and HIV viral load to malaria infection in persons taking highly active ART
Description of the clinical effectiveness of antiretrovirals for prevention and/or treatment of P falciparum infection
Assessment of the effectiveness, safety, and pharmacodynamics of artemisinin-based combination therapy when coadministered with antiretrovirals and other medications commonly used in AIDS care
Description of antimalarial pharmacodynamics in malnourished and/or wasted HIV-infected persons
Longitudinal surveillance of malaria susceptibility to preventive co-trimoxazole
Description of clinical presentation of acute malaria in HIV-infected persons (stratified by CD4 count and/or clinical disease stage), in regions of stable and unstable transmission, with a particular focus on syndromes other than acute fever
Standard vertical guidelines for prevention, diagnosis, and treatment of HIV and malaria should be systematically harmonized, so that patients receive adequate attention regardless of whether the attending clinician was trained by the local malaria or HIV/AIDS program.
Guidelines for diagnosis and management of common clinical syndromes in HIV-infected persons should be rigorously validated, with particular attention to the role of laboratory testing for malaria.
Persons infected with HIV who are taking highly active ART or other long-term medication regimens should be urged to seek care at laboratory-equipped health facilities for malarialike symptoms, rather than self-medicating, whenever feasible; lower-level health units must be systematically reinforced to expand capacity for acute care of patients with AIDS.
Future iterations of the WHO staging criteria should clarify the association between recurrent malaria, severe malaria, anemia, clinical disease stage, and timing of postmalaria CD4 and viral load measurement.
In the preantiretroviral era, marked gains in life expectancy for patients with AIDS were achieved through aggressive OI management.120 The same rigor that led to effective guidelines for the control of Pneumocystis jiroveci in developed countries must now be brought to bear on the control of malaria and other important opportunistic pathogens endemic to sub-Saharan Africa.34
Correspondence: Paula E. Brentlinger, MD, MPH, Department of Health Services, School of Public Health and Community Medicine, Box 357660, University of Washington, Seattle, WA 98195 (firstname.lastname@example.org).
Accepted for Publication: April 17, 2007.
Author Contributions: Dr Brentlinger had full access to all of the documents reviewed and takes responsibility for the integrity and accuracy of the review. Study concept and design: Brentlinger, Behrens, and Kublin. Acquisition of data: Brentlinger. Analysis and interpretation of data: Brentlinger, Behrens, and Kublin. Drafting of the manuscript: Brentlinger. Critical revision of the manuscript for important intellectual content: Behrens and Kublin. Administrative, technical, and material support: Brentlinger, Behrens, and Kublin.
Financial Disclosure: None reported.
Funding/Support: Drs Brentlinger and Behrens were partially supported by Health Resources and Services Administration grant 6U91HA06801-01-01. Dr Kublin was supported by the Fred Hutchinson Cancer Research Center.
Role of the Sponsor: The funders had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.