REPORT Case Report : Three ' s a crowd : a case report examining the diagnostic and pharmacokinetic challenges in HIV-tuberculous meningitis-malaria co-infection

In 2016, 10.4 million cases of tuberculosis (TB) were reported globally. Malaria also continues to be a global public health threat. Due to marked epidemiological overlap in the global burden of TB and malaria, co-infection does occur. An HIV-infected, 32-year-old male presented with a two-week history of headache with fevers to Mulago National Referral Hospital, Uganda. Five months prior, he was diagnosed with pulmonary TB. He endorsed poor adherence to anti-tuberculous medications. in Mycobacterium tuberculosis CSF was confirmed on Xpert MTB/RIF Ultra. On day 2, he was initiated on dexamethasone at 0.4mg/kg/day and induction TB-medications were re-commenced (rifampicin, isoniazid, ethambutol, pyrazinamide) for TBM. He continued to spike high-grade fevers, a peripheral blood smear showed P. parasites despite a negative malaria rapid diagnostic test (RDT). He falciparum received three doses of IV artesunate and then completed 3 days of oral artemether/lumefantrine. To our knowledge this is the first published case of HIV-TBM-malaria co-infection. TBM/malaria co-infection poses a number of management challenges. Due to potential overlap in symptoms between TBM and malaria, it is important to remain vigilant for co-infection. Access to accurate parasitological diagnostics is essential, as RDT use continues to expand, it is essential that clinicians are aware of the potential for false negative results. Anti-malarial therapeutic options are limited due to important drug-drug interactions (DDIs). Rifampicin is a potent enzyme inducer of several hepatic cytochrome P450 enzymes, this induction results in reduced plasma concentrations of several anti-malarial medications. Despite recognition of potential DDIs between rifampicin and artemisinin compounds, and rifampicin and quinine, no treatment guidelines 1,2 3 1 1 1 1,4 1,5 4 1 2 3 4 5 Referee Status: Invited Referees version 2 published 15 Jan 2019 version 1 published 06 Sep 2018 1 2 report report , University of the Tom Boyles Witwatersrand, South Africa 1 , University of Catriona Waitt Liverpool, UK Makerere University College of Health Sciences, Uganda 2 06 Sep 2018, :111 ( First published: 3 ) https://doi.org/10.12688/wellcomeopenres.14726.1 15 Jan 2019, :111 ( Latest published: 3 ) https://doi.org/10.12688/wellcomeopenres.14726.2 v1 Page 1 of 11 Wellcome Open Research 2018, 3:111 Last updated: 15 JAN 2019

potential DDIs between rifampicin and artemisinin compounds, and rifampicin and quinine, no treatment guidelines currently exist for managing patients with co-infection. There is both an urgent need for the development of new anti-malarial drugs which do not interact with rifampicin and for pharmacokinetic studies to guide dose modification of existing anti-malarial drugs to inform clinical practice guidelines.

Introduction
In 2016, 10.4 million cases of tuberculosis (TB) were reported globally 1 . Tuberculous meningitis (TBM) accounts for 1-5% of these 2 . Although TBM can occur in immunocompetent persons, the disease disproportionately affects persons living with HIV and children. Malaria also continues to be a global public health threat. In 2016, an estimated 216 million cases occurred, with 90% of those in Africa 3 . Due to marked epidemiological overlap in the global burden of TB and malaria, co-infection does occur. In an Angolan retrospective study of 1,906 TB inpatients (37% HIV-infected), Plasmodium falciparum co-infection occurred in 38% during hospitalization 4 .
TBM/malaria co-infection poses a number of management challenges. Rifampicin, the cornerstone for drug-sensitive TB treatment, is a potent enzyme inducer that increases the expression of several hepatic cytochrome P450 (CYP450) enzymes, including CYP2A6, CYP2B6, CYP2C, and CYP3A isoenzymes, as well as the efflux drug transporter P-glycoprotein 5 . This induction alters the pharmacokinetics of drugs metabolized by these pathways, reducing plasma concentrations of several anti-malarial medications, including artemisinin-based drugs, quinine and atovaquone/proguanil. In HIV/TBM/malaria co-infection, there are additional interactions between antiretroviral therapy (ART) and anti-malarial medications 6 .
Here; we present the case of a hospitalized HIV-infected adult with HIV/TBM/malaria co-infection, highlighting important diagnostic and pharmacokinetic challenges.

Case report
An HIV-infected 32-year-old male presented to Mulago National Referral Hospital, Uganda with a 2-week history of headache with fevers and a 1-day history of confusion ( Figure 1).
He had been on ART (zidovudine, lamivudine, efavirenz) and co-trimoxazole prophylaxis for 5 years. At 5 months prior, he was diagnosed with pulmonary TB by positive sputum Xpert MTB/RIF (Cepheid, Sunnyvale, CA, USA). He had completed 2 months of induction TB therapy (rifampicin, isoniazid, ethambutol, pyrazinamide) and was 3 months into continuation phase (rifampicin, isoniazid). He endorsed poor adherence to both ART and anti-tuberculous medications.
On examination, the patient was febrile (38.6°C). His blood pressure was 112/71 mmHg, pulse 94 beats/minute, respiratory rate 48, and oxygen saturation 98%. He was wasted, dehydrated, and had overt rigors. His Glasgow Coma Scale was 14/15 with nuchal rigidity and positive Kernig's sign. Cranial nerves were intact. He had normal tone and power in all limbs. A clinical diagnosis of HIV-associated meningitis was suspected and he was recruited into the 'Improving Diagnostics and Neurocognitive Outcomes in HIV/AIDS-related Meningitis' study (registration: ISRCTN42218549). Whilst awaiting further investigations, he received empiric therapy of ceftriaxone 2 g twice daily for possible bacterial meningitis.
A finger stick cryptococcal antigen lateral flow assay (CrAg LFA) (IMMY, Norman, Oklahoma, USA) was negative. Liver and renal function tests were normal. Cerebrospinal fluid (CSF) opening pressure was elevated to 33 cm H 2 O (normal <20 cm H 2 O), CSF white cells 590 /µl, protein 419 mg/dl (normal range 15-45 mg/dl), CSF lactate 9.5 mmol/L (normal range <2.5 mmol/l). CSF glucose was unavailable. Mycobacterium tuberculosis in CSF was confirmed on Xpert MTB/RIF Ultra. On day 2, he was initiated on dexamethasone at 0.4 mg/kg/day and induction TB-medications were re-commenced (rifampicin, isoniazid, ethambutol, pyrazinamide) for TBM. His ART was continued. He continued to spike high-grade fevers (39.6°C.) with tachycardia (pulse 118 beats/min). A peripheral blood smear showed P. falciparum parasites (1+ trophozoites), despite a negative malaria histidine rich protein-2 (PfHPR2)based rapid diagnostic test (Malaria Plasmodium falciparum Rapid Test Cassette, Vaxpert, Florida, USA). Given his ongoing neurological symptoms, which could be compatible with cerebral malaria, the decision was made to treat for severe malaria. Drug-drug interactions (DDIs) between rifampicin and artemisinin compounds, and rifampicin and quinine are recognized (Table 1); a decision was made to treat with IV artesunate as the most efficacious anti-malarial for severe malaria 6 . He received three doses of IV artesunate (3 mg/kg), after which a repeat peripheral blood smear showed no malaria parasites. He then completed 3 days of oral artemether/lumefantrine. His fevers subsided on day 6 and he was discharged on day 8.

Discussion
This case demonstrates the diagnostic and treatment challenges encountered when managing patients with advanced HIV and intercurrent infections. Protracted high-grade fevers are not an uncommon feature of TBM, even after appropriate antituberculous treatment has been commenced, and it is important to remain vigilant for co-infections. Co-infection with HIV-TB-malaria is well recognized 4 ; however, to our knowledge this is the first published case of HIV-TBM-malaria co-infection.
Due to the potential overlap in symptoms and signs between TBM and malaria (fever, confusion, reduced level of consciousness, seizures, sepsis), access to accurate parasitological diagnostics is essential. Light microscopy (Giemsa stain) remains the gold standard parasitological diagnostic, but the World Health Organization (WHO) recommends immunochromatographic rapid diagnostic testing (RDT) in settings with limited laboratory facilities 6 . RDTs-which detect Plasmodium antigens such as PfHPR2, Plasmodium lactate dehydrogenase (pLDH) or plasmodial aldolase-are the backbone of expanding access to malaria diagnostics in resource-limited settings. Since their introduction in the late 1990s, the number of RDTs available, and the scale of their use, has increased rapidly. A meta-analysis of 74 studies assessing accuracy of PfHRP-2 RDTs for diagnosis of uncomplicated P. falciparum malaria in endemic settings reported average sensitivity and specificity (95% CI) of 95.0% (93.5-96.2%) and 95.2% (93.4-99.4%), respectively 7 . However RDTs do have several limitations: poor sensitivity at low parasite densities; susceptibility to the prozone effect (PfHRP2-detecting RDTs); false-negative results due to PfHRP2 gene deletions; false-positive results caused by other infections, and susceptibility to heat and humidity 8 . The largest problem historically has been poor manufacturing quality, which the World Health Organization (WHO) malaria product testing programme has addressed. However, Vaxpert does not participate in this WHO quality assurance program. As demonstrated by this case, false negatives can occur.
The WHO recommends treatment with an artemisinin derivative for both uncomplicated (oral artemisinin combination therapy (ACT) for 3 days) and complicated falciparum malaria infections (intravenous or intramuscular artesunate for at least 24 hours and until the patient can tolerate oral medications, at which stage 3 days of ACT should be completed) 6 . Despite the recognition of potential DDIs between first-line anti-malarial drugs and rifampicin, no treatment guidelines currently exist for managing patients with TB/malaria co-infection. The WHO states that there is currently a lack of evidence to recommend dosage modifications but advises clinicians of increased risk of recrudescent infections due to DDIs in co-infected patients 6 .
In a DDI study in HIV-infected Ugandans to investigate the pharmacokinetics of artemether, dihydroartemisinin (DHA) and lumefantrine during rifampicin intake, co-administration with rifampicin resulted in a significantly lower exposure (area under the curve between 0 and 12 hours post-dosing) to artemether (89% lower) and DHA (85% lower) 9 . Co-administration of artemether-lumefantrine and rifampicin should therefore be avoided. There are no published studies examining concomitant administration of IV artesunate and rifampicin. However, as DHA (the active metabolite of artesunate) is metabolized by CYP450 enzymes (Table 1), there is a theoretical risk that co-administration with rifampicin would result in reduced plasma DHA concentrations and a reduction in efficacy 5 . Oral agents may undergo metabolism in the gut and the liver prior to reaching the systemic circulation, but intravenous drugs are directly administered to the systemic circulation and so reductions in DHA exposure with intravenous artesunate could be potentially of lower magnitude than what was seen with oral artemether.
Similarly, the bioavailability of quinine-metabolized almost exclusively via CYP450 (CYP3A4 and CYP2C19) enzymes-is significantly reduced when co-administered with rifampicin and has been associated with a clinically significant reduction in efficacy. Adults treated for uncomplicated falciparum malaria were randomized to receive oral quinine either alone or in combination with rifampin; recrudescence rates were five times higher (15/23; 65%) in the rifampicin arm than those treated with quinine alone (3/25; 12%, P<0.001) 10 . It is recommended that for patients already receiving rifampicin, quinine doses should be increased. However, to date, no guidance on doseadjustment strategies have been published, and although it is advised that therapeutic drug monitoring may be useful, this is not feasible in most parts of the world where co-infection occurs 5 .
A range of anti-malarial drugs used in the treatment of non-severe malaria and for prophylaxis also have documented DDIs with rifampicin including: atovaquone; chloroquine; piperaquine; mefloquine and doxycycline ( Table 1). As no clinical guidelines currently exist regarding dose modification, these drugs should be used with caution in patients concurrently receiving rifampicin.
Many antiretrovirals also interact with the CYP450 system (Table 1). For example, co-administration with efavirenz reduces exposure to the active components of artemether and lumefantrine 46% and 21%, respectively. However, the net induction effects of the use of both rifampicin and efavirenz on antimalarial compounds has not been quantified. Similarly, protease inhibitors are potent CYP450 inhibitors and their combined effect with rifampicin induction is unknown.

Conclusions
In conclusion, patients with HIV-TBM-malaria co-infection present a number of management challenges. The potential symptom overlap in clinical presentation means that clinicians must remain vigilant for co-infection and access to reliable parasitological diagnostics is imperative. As malaria RDT use continues to expand, it is essential that clinicians are aware of the potential for false negative results. Therapeutic options for TB-malaria co-infection are limited due to DDIs. There is both an urgent need for the development of new anti-malarial drugs which do not interact with rifampicin and for pharmacokinetic studies to guide dose modification of existing anti-malarial drugs to inform clinical practice guidelines.

Data availability
All data underlying the results are available as part of the article and no additional source data are required.

Consent
The Improving Diagnostics and Neurocognitive Outcomes in HIV/AIDS-related Meningitis study protocol was approved by Ugandan and Minnesota IRBs. Written informed consent from the patient's next of kin was given for publication of that patient's clinical details as the patient himself lacked the mental capacity to consent. Thank you very much for this report. Whilst this may be the first published report of the three infections occurring concurrently, it is likely that in practice these comorbidities and resulting drug-drug interactions are much more common, and hence it is an important contribution to the literature. Both the diagnostic challenges in this case, and the limitations of the best available therapeutic options given the drug-drug interactions are clearly presented.

Grant information
The table summarising the drug interactions is very helpful.
For those who may be less familiar with managing such patients, was the case unusual? Is it common to present with TB meningitis five months into therapy where there has been apparent adherence?
Was the ceftriaxone discontinued once the positive TB result from CSF was obtained? (I presume so, but it is not explicitly stated) I would be interested to know about any follow-up post discharge. Was anything done differently, for example support for adherence etc?
The conclusion finishes with reiterating the 'urgent need for the development of new anti-malarial drugs which do not interact with rifampicin and for pharmacokinetic studies to guide dose modification of existing anti-malarial drugs to inform clinical practice guidelines' -I entirely agree with this statement, but given the challenges and diversity in the population in question which has been described, I would also have liked a couple of sentences in the discussion about how such studies could best be designed. What would be ideal? Logistically, how could such a population be drawn together to provide meaningful data? Do healthy volunteer studies have a part to play here? Are the authors involved in any existing work to try and address some of these gaps? Are they aware of any other work that is ongoing?
Minor points CSF opening pressure should be measured in cm CSF, not cm H20. 1 2 was clearly at risk of DR-TB.
-It would be worth pointing out that it takes around 2 weeks for the full effects of enzyme induction by rifampicin to be felt. If malaria is diagnosed at a similar time to TB, as in this case, the initial effects of enzyme induction might be limited, but increase over time which would increase the challenge of finding the appropriate dose.
-It is never a bad thing to remind people of the need for a lipid meal before taking lumafantrine in order to achieve adequate absorption, something which is often neglected.

Is the background of the case's history and progression described in sufficient detail? Yes
Are enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes Is sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes

Is the case presented with sufficient detail to be useful for other practitioners? Yes
No competing interests were disclosed.

Competing Interests:
Reviewer Expertise: Infectious diseases in low resource settings I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.