Management of granulomatous amebic encephalitis: Laboratory diagnosis and treatment

Abstract

Granulomatous amebic encephalitis is a life-threatening central nervous system (CNS) infection caused by the free-living amoebae Acanthamoeba spp., Balamuthia mandrillaris and Sappinia pedata. The disease has a subacute to chronic onset affecting commonly the immunocompromised population with high mortality rate. The diagnosis of this disease entity requires high suspicion with appropriate sample collection and testing by the laboratory experts. Radiological investigations are nonspecific and commonly confused with CNS tuberculosis, neurocysticercosis, disseminated encephalomyelitis, viral encephalitis etc., delaying the accurate diagnosis of these cases. Early diagnosis plays a crucial role in the survival of these cases since appropriate management can be initiated. No single drug is effective; hence multiple antibiotics targeting various proteins or receptors are required for successful treatment. A combination of surgical and medical interventions involving multiple specialty experts is required to prevent death and morbidity in survivors.

INTRODUCTION

Granulomatous amebic encephalitis (GAE) is a subacute to chronic granulomatous infection of the central nervous system (CNS) caused by the free-living amoebae, Acanthamoeba spp., Balamuthia mandrillaris and Sappinia pedata (previously called as Sappinia diploidea), unlike the acute CNS infection (primary amebic encephalitis [PAM]) caused by the free-living amoebae, Naegleria fowleri and Paravahlkampfia francinae.[1,2] These CNS infections are reported to occur mostly in immunocompromised individuals like posttransplantation, HIV infection etc., and occasionally in immunocompetent hosts.[3,4,5] These GAE cases are infrequent, but high rates of mortality have been observed.[6,7] Very few cases of survival have been reported, particularly in CNS balamuthiasis.[8] These amoebae exist as saprophytes (free-living) in the environment and harbor endosymbiotic pathogenic bacteria like Legionella pneumophila, Mycobacterium avium complex, Listeria monocytogenes, Burkholderia pseudomallei, Vibrio cholerae, EHEC O157:H7 etc., potentiating their environmental survival and virulence.[9] These amoebae spread by hematogenous route from the skin, sinuses, or lungs, depending on their portal of entry. Many of these cases have an antecedent skin lesion/lung infection occurring few months earlier predicting the upcoming CNS infection. Approximately 50% of the survivors of GAE cases due to B. mandrillaris have an antecedent skin lesion and hence a better prognosis exists in these individuals due to the possibility of early detection. Radiological investigation is nonspecific showing a space-occupying lesion commonly as hypodense/ring-enhancing lesion in most cases. These GAE cases are commonly misdiagnosed as cases of neurotuberculosis, neurocysticercosis, viral encephalitis etc., delaying the accurate diagnosis and hence the high mortality observed.[10] High index of suspicion by the clinicians and the laboratory experts is required to detect this life-threatening infection and start appropriate therapy. Although a protozoa, this parasite does not respond to any antiparasitic agents. A combination of timely neurosurgical intervention (surgical resection of the lesion) and commencing the patient on a regimen containing multiple antibiotics plays a crucial role in the survival of these cases.

LABORATORY DIAGNOSIS

Microscopy

Microscopic detection of the morphological forms of the parasite (trophozoite and cyst) has been the conventional and common method utilized for the diagnosis of GAE. Cerebrospinal fluid (CSF), skin/sinus/lung biopsy, and brain tissue biopsy (obtained postsurgical resection/postmortem) have been the samples of choice for diagnosis of these GAE infections. Usually, skin lesions are found in the face or the extremities as erythematous plaques/ulcers.[11] In CNS infections, the most common affected regions are the cerebral hemispheres, thalamus, cerebellum, and midbrain. Lesions are usually necrotic and hemorrhagic granulomatous material during resection. CSF examination reveals nonspecific lymphocytic pleocytosis with elevated protein and normal/low glucose levels in GAE cases whereas in PAM cases, abundant neutrophils are observed resembling bacterial meningitis.[12] Wet mount examination can reveal the motile trophozoite forms for which multiple aliquots of CSF are centrifuged at low-speed (250 ×g ) to avoid disruption of the trophozoite forms. CSF needs to be collected in adequate amounts (approximately 3-5 ml minimum) for detection of the trophozoites. In GAE cases due to Acanthameoba/Balamuthia, sparse or no trophozoites are seen in the wet mount examination whereas in PAM caused by N. fowleri, plenty of trophozoites are visualized in the CSF wet mount examination. Hence, CSF examination is not the method of choice to rule out GAE infections. When trophozoites are seen in the CSF examination, genera differentiation can be done with the following features: N. fowleri trophozoites are comparatively small measuring around 10-25 μm with typical limacine/eruptive amoeboid movement and gives a positive enflagellation test (i.e.,) few drops of CSF is mixed with 1 ml of distilled water and examined after 1 h for flagellated forms, typical of N. fowleri. Trophozoites of Acanthamoeba spp. measure 15–40 μm with spiny projections called acanthopodia and gives a negative enflagellation test whereas Balamuthia trophozoites are larger (60 μm) are characterized by extensively branched pseudopodia (pleomorphic appearance) with sluggish ameboid movement and negative enflagellation test. Cysts are usually not observed in the CSF examination.[13,14]

Histopathological examination (HPE) of tissue biopsies (skin/brain) has a high yield of detection of GAE infections. Both morphological forms (cyst and trophozoites) can be demonstrated in HPE of formalin-fixed parafinised samples. Various stains employed in the demonstration of these morphological forms are hematoxylin and eosin, trichrome, wright, Grocott's GMS modification, acridine orange, calcofluor white, etc.[15] In HPE, the trophozoites and cysts are seen cuffing the perivascular region. Trophozoites/cyst stages of the parasite are commonly mistaken as macrophages/necrosed keratinocytes. Trophozoites/cysts of Acanthamoeba spp. have a nucleus with single nucleoli whereas Balamuthia cysts/trophozoites are characterized by the presence of 2 nucleoli in the nucleus.[16] The pleomorphic morphology of Balamuthia and the acanthopodia of Acanthamoeba spp. are usually not appreciable in HPE. S. pedata can be identified in the HPE by the characteristic presence of 2 nuclei with the opposing/adjacent sided flattened and each of these nuclei possessing 2 nucleoli. Cysts of Acanthamoeba spp. can be classified into three groups based on their size and morphology. Group 1 constitutes Acanthamoeba spp. with cyst size measuring 15-30 μm, Group 2 cysts measuring 18 μm and Group 3 cysts measuring less than 18 μm. Group 1 and 2 cysts usually have the wrinkled outer morphology whereas, in Group 3, the outer covering is smooth. Around 24 species of Acanthamoeba have been categorized into these three groups and most pathogenic species of Acanthamoeba fall in Group 2.[1,17,18] Under light microscopic examination, genera differentiation is difficult since the layers of the cyst are not differentiated. In electron microscopy, Acanthamoeba cysts are seen as double layered cysts – outer wrinkled ectocyst and inner smooth round/oval/polygonal shaped endocyst. Whereas cysts of B. mandrillaris are seen as 3 layered – outer irregular/wrinkled thin ectocyst, middle amorphous/fibrillar mesocyst, and a smooth thick endocyst.

Immunohistochemistry (IHC) has been widely used in the detection of the morphological forms of the parasite. Specific antibodies targeting the amebic antigens are added to the specimen, subsequently enzyme conjugate and substrate added to detect the antigens. This technique has been helpful in differentiation of tissue macrophages and necrosed keratinocytes from amoebae by targeting CD68 and pancytokeratin, respectively. This technique also reliably differentiates Acanthamoeba spp. from B. mandrillaris, which is difficult in direct HPE. In IHC, the cysts are stained weakly than the trophozoite stages. Granular antigen staining represents the amebic fragments/antigens and is predominantly seen scattered in the macrophages, endothelial cells and vessel walls showing an angitis-like picture.[19] Similarly, indirect immunofluorescence (IIF) has also been used to differentiate the genera causing GAE.

Culture

Three methods of cultivation exist to culture these free-living amoebae causing GAE. They are xenic media (bacterized media), axenic media (media containing no actively metabolizing cells other than the parasite intended) and tissue culture. Xenic culture is performed on a nonnutrient/low nutrient agar in the presence of bacteria. Bacterial isolates usually chosen for the xenic culture are nonmucoid colonies of Escherichia coli or Enterobacter aerogenes since presence of the capsule can hinder phagocytosis and thereby the cultivation of these amoebae. The reason for the choice of non/low-nutrient agar is to prevent overgrowth of bacteria. Axenic is performed on enriched culture medium with added antibiotics such as penicillin and aminoglycosides (gentamicin/streptomycin). Antifungals are to be avoided particularly amphotericin B and azoles since they can hamper the growth/retrieval of these amoebae. Tissue culture has been recently introduced for effective retrieval/cultivation of these amoebae. Cell lines that are used are monkey kidney cell line (ATCC CRL 1586), rat glioma cell line (ATCC CCL 107), human lung fibroblasts, and human brain microvascular endothelial cell line.[20]

Although CSF specimens have been widely used for culturing N. fowleri, it has limited the role in the diagnosis of GAE cases since trophozoites are rarely or never seen in cases of CNS acanthamoebiasis or balamuthiasis respectively. Brain/skin tissue biopsy has been widely used for culture as a diagnostic tool. The initial choice of media for primary isolation of agents causing GAE should be xenic (bacterized) media, and then subculture done in axenic media. The only exception is B. mandrillaris, which does not grow in xenic media. Hence, always xenic and axenic to be inoculated simultaneously or tissue culture can be used since it supports the growth of all the agents causing GAE. Sappinia grows well in xenic media and tissue culture cells.[21] The biopsy specimen is kept in the nonnutrient agar containing bacteria and incubated for 1-2 days at 37°C. The presence of these amoebae can be deduced by examining the culture plate under low power (×10) objective of an inverted microscope. The microscopic features and enflagellation test can make further differentiation of the genera as mentioned earlier. Tissue culture flasks need to be incubated for extended periods since Balamuthia are slow growing, and once they grow, they are observed as ameba with spider-like locomotion on the surface of the feeder cells. Examples of axenic media widely used for cultivation of Acanthamoeba are Byer's DGM-21A, Shukla's M-11 etc., and for B. mandrillaris is BM-3 medium (Shuster and Visvesvara et al.). Although the procedure is simple and easy to perform, the time consumed and the reliability of the test is questionable.[13,20]

Serological diagnosis

In PAM caused by N. fowleri, the onset and progression of the disease are rapid and hence antibody detection plays no role in this disease. Whereas in cases of GAE, the onset is subacute to chronic, lasting weeks to months before progressing to severe disease. In these GAE cases, detection of antibodies is reliable and preferable since it's a noninvasive procedure and helps in early detection of cases. Various methods such as IIF, ELISA, and flow cytometry, have been employed in the detection of antibodies.[22,23,24,25] The report from California Encephalitis project shows that antibodies are detectable in healthy individuals and confirmed CNS infections. Low titers (<1:64) were observed in healthy hispanic individuals and blood donors suggesting their exposure to these environmental free-living amoebae, whereas high antibody titers (1:128–1:1024) were observed in GAE cases confirmed by IHC and molecular techniques.[26] In IIF, the amoebic antigenic extracts are fixed in aliquots onto a slide and serial dilutions of the patient sera added and subsequently antihuman antibody tagged with fluorescein isothiocyanate is added to detect emission of fluorescence. No cross-reactions were observed between cases of Acanthamoeba or Balamuthia infections.[27] Fluorescence associated cell sorting shows higher sensitivity of antibody detection and hence higher titer cut-off are to be followed to differentiate healthy from diseased individuals. Although noninvasive and the detection method is easy, only presumptive diagnosis can be made particularly in cases of subacute onset (lasting few weeks). Antigen detection methods are currently not available for diagnostic purposes. Several recombinant protein antigens such as 25 kDa, 50 kDa, and 75 kDa have been identified by western blot and silver staining analysis, which could be a promising target for early detection of these GAE cases.[28]

Molecular diagnosis

Molecular assays have been emerging as the most promising tool in the diagnosis of infectious diseases. Recently, multiple assays detecting different targets have been reported for the identification of agents causing GAE. Common targets are nuclear and mitochondrial 18S rDNA (SSU rRNA gene), RNAse P gene, etc.[29,30,31] Samples that are usually tested for diagnosis of these infections are brain/skin tissue biopsy, archived parafinised tissue samples, etc.[32,33] CSF can be used, but the sensitivity is expected to be low since the trophozoites/cysts are rarely seen in this sample. Reports exist in the detection of Acanthamoeba and Balamuthia DNA from CSF samples in GAE cases.[34] Commercial tissue extraction kits/conventional manual extraction can be used for DNA extraction. Cysts forms of these parasites are hardy and require treatment with proteinase K for obtaining optimum DNA quantity. Polymerase chain reaction (PCR) analysis of nuclear 18S rDNA have revealed multiple genotypes in Acanthamoeba spp.[35] Currently, around 18 genotypes have been described of which T4 genotype predominates in both keratitis and nonkeratitis infections. Certain rare genotypes such as T1, T10, and T12 have been reported from cases of CNS infections but yet an environmental niche has not been identified, unlike the T4 genotype, which is found both in clinical infections and in the environment.[36] In contrast, analysis of nuclear and mitochondrial 18S rDNA analysis of Balamuthia has revealed a single genotype causing human infections worldwide.[37] No such genotypes have been identified yet for S. pedata. No cross-reactions have been observed between Acanthamoeba and Balamuthia although both amoebae have been shown closely related by phylogenetic analysis. The limits of detection for most assays have been determined using cultured amoebae. PCR targeting the nuclear SSU rRNA and RNAse P gene have showed a detection limit of approximately 1-2 amoebae/specimen. PCR detecting mitochondrial 18S rDNA of B. mandrillaris have been reported a detection limit of 0.2 ameba/specimen since multiple copies of mitochondrial DNA may be present in a single ameba.[38] Qvarnstrom et al., had described a Taqman based multiplex real-time PCR targeting the 18S rRNA gene in the detection of N. fowleri, Acanthamoeba spp. and B. mandrillaris.[39] Molecular assay detecting S. diploidea (now S. pedata) employed amplification initially with universal eukaryotic SSU rRNA gene followed by amplification of specific internal transcriber spacer regions.[40,41] Newer assay targeting the 18S rDNA have been described and can be added to the multiplex PCR assay described by Qvanstrom et al., These molecular assays are currently the investigation of choice because of its high sensitivity and rapid result generation that helps in appropriate patient management.

Treatment

The mainstay of successful treatment depends on early diagnosis of these GAE cases. Most cases are detected at late stages and hence high mortality has been observed in these cases. Many survivors of this disease (approximately 50%) were diagnosed early since they had skin manifestations that predicted and helped in the prevention and early detection of these fatal amebic infections. Most of these skin lesions precede CNS infections by weeks to months. Retrospective analysis of survival cases reveals a combination of surgical resection of the affected lesion and a regimen of multiple antibiotics.[42] The regimens of antibiotics chosen contain drugs acting at various levels and receptors. The commonly used antibiotics are pentamidine, cotrimoxazole, propamidine isethionate, azoles like fluconazole, itraconazole and voriconazole, amphotericin B, flucytosine, rifampin, azithromycin, amikacin, etc.[43] Recently introduced drugs that have been successfully used in the treatment of GAE infections are the anticancer drug miltefosine, antipsychiatric agents phenothiazines and thioridazine, although the latter 2 drugs had to be stopped during treatment because of severe toxicity. Usually, a combination of multiple antibiotics (approximately 4-5) have been used in successfully treated cases of GAE infections, except for 2 CNS acanthamoebiasis cases where a single drug (cotrimoxazole) had been effective in treatment of these life-threatening infections.[3,44,45] Initiation of antiretroviral agents in a patient with HIV immunosuppression and topical application of chlorhexidine/miltefosine in cases of infections with skin manifestations showed a better outcome. Certain studies show that the initiation of the miltefosine reduced mortality risk in CNS acanthamoebiasis and balamuthiasis. Miltefosine, azoles, pentamidine, and cotrimoxazole were used in the treatment of >90% successfully treated GAE cases. Only 10 survivors of CNS balamuthiasis have been reported based on the PubMed database search.[46] Although survivors are known in GAE cases, postinfectious sequelae of neurocognitive disorders have been noticed. This sequelae is due to the extensive cerebral edema that develops during the course of the disease and hence needs to be managed meticulously to prevent this complication in survivors. Better survival and prognosis have been observed in elderly when compared to younger/middle age group individuals.[47] Several in vitro and in vivo studies have tested the efficacy of these antibiotics against these free-living amoebae.[48,49,50,51,52] Both tissue culture and cell-free growth system had been utilized to screen amebic cultures for antibiotic susceptibility testing.[53] These studies show that the cyst forms of the parasite require higher drug concentrations for the amebicidal action than the trophozoite forms. Voriconazole has been proven ineffective against B. mandrillaris and hence should not be employed in the treatment regimen of CNS balamuthiasis. Most drugs used for treatment are amebistatic with few being amebicidal at therapeutic concentrations. Hence, the duration of treatment duration is for months to completely eradicate the parasite. Recent in vitro studies show loperamide, haloperidol, apomorphine, procyclidine, and amiodarone as promising drugs that can be utilized in the treatment of GAE infections.[54,55]