Neurocysticercosis and HIV Infection: what can we learn from the published literature?

ABSTRACT Infections caused by the human immunodeficiency virus (HIV) and by the larvae of Taenia solium (i.e., cysticercosis) are still widespread in many developing countries. Both pathologies modify host immune status and it is possible that HIV infection may modulate the frequency and pathogeny of cysticercosis of the central nervous system (i.e., neurocysticercosis [NCC]). Objective: To describe published cases of NCC among HIV-positive patients and to evaluate whether the characteristics of NCC, including frequency, symptoms, radiological appearance, and response to treatment differed between HIV-positive and HIV-negative patients. Methods: Forty cases of NCC/HIV co-infected patients were identified in the literature. Clinical and radiological characteristics, as well as response to treatment, were compared with non-matching historical series of NCC patients without HIV infection. Results: Most of these patients had seizures and multiple vesicular parasites located in parenchyma. Clinical and radiological characteristics were similar between HIV-positive and HIV-negative patients with NCC, as well as between immunocompromised and non-immunocompromised HIV-positive patients. Conclusion: Our review did not reveal clear interactions between HIV and NCC. This may be partially due to the small number of cases and reliance on published research. A systematic, multi-institutional effort aiming to report all the cases of this dual pathology is needed to confirm this finding and to clarify the possible relationship between both pathogens.

by different pathogens and also enables easier transmission from one compartment of the body to another. The most frequent opportunistic infections of the central nervous system (CNS) are toxoplasmosis, cryptococcosis, and tuberculosis 2, 3 .
Neurocysticercosis (NCC) is an endemic infection in many countries in Latin America, Africa, and Asia and results from the localization of the larvae of Taenia solium in the CNS 3 . In the year 2010 it was estimated that there were 372,000 incident cases of cysticercosis, that NCC was the cause of the death of about 28,000 people, and that the disease burden estimated to be caused by cysticercosis was 2.79 million (95%UI 2.14-3.61 million) DALYs 4 . The presence of T. solium in humans induces an immunological response, and it has been shown that the different stages of the parasite (vesicular, colloidal, and calcified) are each associated with a particular immunological profile. In particular, calcified parenchymal NCC is associated with a Th2 response that involves an increase of IL-4, IL-5, and IL-13 in peripheral blood, while severe NCC is associated with an increase of IL-5, IL-6 and IL-10 in CSF, along with a decrease in the lymphoproliferation response 5 .
NCC and HIV are infections that might have some type of interaction. Both diseases are prevalent in many of the same countries, and a recent report from Brazil found that HIV/AIDS was recorded as the most frequent underlying cause of death when NCC was listed as an associated cause of death 6 . Also, although the lifecycle of T. solium occurs mainly in rural areas, different case series show that most patients lived in urban settings 7 , similar to patients with HIV 8 . At an immunological level, although both diseases alter the immunological system in different ways, we cannot rule out the possibility that the HIV-associated immunosuppression can modulate the T. solium infection. In this sense, it is interesting to note that infection with the related parasite Taenia crassiceps seems to be more frequent in HIV-positive individuals than in HIV-negative ones 9 .
Considering these antecedents, we described all of the NCC-HIV-positive cases we could identify in the published medical literature. Also, using control studies describing NCC-HIV-negative cases, we assessed whether interactions between the two pathogens could exist, influencing NCC prevalence, symptoms, imaging characteristics, and treatment response, using the medical literature. We specifically hypothesized that immunosuppressed individuals with HIV might have increased NCC prevalence (possibly related to opportunism of NCC) and radiological severity, as well as a decrease of NCC symptoms and response to treatment.

METHODS
We conducted a scoping review of the literature by mapping the medical literature on HIV and NCC, as our research questions were broad and complex, and could not be condensed into targeted questions that could be answered by a systematic review.
We  11 . Reports not meeting these case definitions (e.g., NCC ascertained using serological tests only) or reports completely lacking information on the characteristics of NCC infection (number, stage, and localization of parasites, and clinical features) and on CD4 status were excluded.
The results of our search are depicted in Figure. Detailed information (symptoms, number, location and stage of parasites, CD4 count and outcome) regarding each NCC/HIV case is provided in supplementary Table 1.
Information regarding control groups (i.e., HIVnegative individuals with NCC) is detailed in Table 1. A comparison was made using matched population 12 . The remaining control studies were included based on the following criteria: 1) Similar institutions where NCC/HIVpositive patients were evaluated 13 ; 2) pooled estimates of neurological symptoms associated with NCC obtained by a systematic review of the literature 14 ; and 3) series of patients from hospital settings in which radiological findings and outcomes were described 13,15,16,17,18,19,20,21,44 .
Regarding symptoms, among the 30 cases with these characteristics clearly described, most of the patients presented with seizures (62.1%), followed by focal deficit (20.7%), intracranial hypertension (10.3%), and headache (6.9%) ( Table 2   and supplementary table). At a radiological level (Table 3 and  supplementary table), most of the patients presented with multiple parasites (82.9%), located in parenchyma (79.3%), and in the vesicular stage (36.0%). It is interesting to note that in all of the patients with only extraparenchymal parasites, parasites were located in the subarachnoid space, and 50% of them (2 patients) presented with parasites at the spinal level. After treatment, most of the patients improved (88.9%)    (Table 4). Two patients died; for one of them the cause of death was not NC (thrombocytopenia), and for the other, the cause was not clearly described.

Effect of the intensity of HIV-related immunosuppression on NCC characteristics
Individual CD4 count was described for only 18 NCC/HIV cases, 10 patients (55.6%) had CD4 ≥ 200 cells/mm 3 while eight patients (44.4%) had CD4 < 200 cells/mm 3 23,24,25,28,29,30, 31,32,33,35,36,37,38,40,41 . A comparison of the main characteristics of NCC between the two groups is presented in Table 5. As seen, in the two groups, the main symptom was seizure, most of the parasites were located in the parenchyma and degenerating (colloidal or calcified). Patients improved with treatment in the two groups.

Effect of HIV on NCC prevalence
Two comparative studies allowed us to evaluate the effect of HIV on NCC prevalence, since it is possible that NCC represents an opportunistic infection.   The first was a study where the prevalence of NCC (with CT-based diagnosis) in 170 matched HIV-positive and HIV-negative patients in northern Tanzania were evaluated and compared 12 . The authors did not find any significant differences in NCC prevalence and clinical manifestations between the two populations.
The second comparison was made using data from two studies that examined the frequency of NCC among HIV-positive and HIV-negative individuals at the Mexican National Institute of Neurology 13,24 . In both studies, NCC diagnosis was made using radiological parameters. The first evaluated the prevalence of NCC among HIV-negative patients admitted to this institution in 1994 and 2004. The authors reported an NCC prevalence of 2.4% (100/4098) in 1994, and of 2.5% (120/4706) in 2004. The second study described the prevalence of NCC in 91 HIV patients diagnosed between 1987 and 1993. Two of the participants in this study had NCC, with an estimated 2.2% prevalence of NCC in HIV-positive individuals. Thus, no major difference in NCC prevalence between HIV+ and HIV-populations was found (2.2% vs. 2.4%).

Effect of HIV on NCC symptoms, imaging characteristics, and treatment response
We compiled the symptoms of published NCC/HIV cases 22,23,24,25,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43 , and compared them with a published meta-analysis providing estimates of the manifestations among HIV-negative patients with symptomatic NCC 14 . The results are presented in Table 2. Within the two groups of patients, seizure was the symptom most frequently reported. The frequency of focal deficit and intracranial hypertension were similar between the two groups of patients and only headache was lower in the HIV-positive group.
We compiled HIV/NCC cases in which response to treatment was specified 22,24,25,26,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43 . Treatment consisted of surgery and/or cysticidal drugs. It was frequently not described whether the criteria of improvement were clinical or radiological. Period of follow-up was also highly variable and was frequently short after treatment. In cases of no improvement or death, it was difficult to know whether the cause was NCC or an opportunistic infection associated with HIV. The data of NCC/HIV patients were compared with different studies that evaluated the outcomes of the treatment in NCC patients without HIV 17,19,20,21 . As shown in Table 5, most patients of both groups improved with specific treatment (surgery or cysticidal drugs).

DISCUSSION
HIV infection and NCC are endemic in many of the same geographical regions, which include predominantly developing countries. It is well known that both diseases modulate the immune system and it is possible that some interactions may occur between them. Particularly, immunosuppression could favor the invasion and growth of parasites in CNS, increasing the prevalence and radiological severity of NCC. On the other hand, due to the decrease in inflammatory reaction, symptoms and response to treatment could be diminished.  To evaluate these hypotheses, we compared different characteristics between HIV-negative and HIV-positive patients, both with NCC. Only 40 individual cases of NCC and HIV coinfection were identified in the literature. The descriptions of the patients in the case reports were frequently not complete and we are conscious that this could generate bias in our results. Publication bias is also especially relevant to consider for the individual cases of HIV and NCC co-infection. Indeed, authors and journals might be more likely to publish case reports that present especially unique or severe disease presentations, making case reports an inaccurate representation of the true nature of the disease presentation. Another limitation that warrants mentioning is that the HIV status of the individuals with NCC in the studies used for controls was not known; however, given the low overall prevalence of HIV in these settings, it seems very unlikely to have influenced the comparisons we made.
In spite of all these limitations, we think that this exercise is interesting to show the current state of the literature and the gaps that future research should fill.
We could not identify any clear differences between HIV-positive and HIV-negative patients regarding the frequency of NCC infection, the clinical-radiological presentations nor the response to treatment.
The frequency of NCC infection was similar in HIV-positive and HIV-negative individuals, consistent with an earlier autopsy study made in Mexico 45 . Furthermore, NCC diagnosis did not appear to be more frequent in HIV individuals with CD4 counts <200 cell/µl compared to those with higher counts. It is interesting to note that there were only a few reported cases of NCC among immunosuppressed individuals without HIV 46,47,48,49,50 , supporting the idea that immunosuppression does not increase susceptibility to NCC. Moreover, we did not find any clear differences in NCC characteristics and outcomes between immunosuppressed and non-immunosuppressed HIV patients (Table 5), data that contradict previous reports suggesting that immunosuppression can increase the frequency of severe radiological presentation of NCC 25,50,51 . Furthermore, the main symptoms of NCC in HIV-positive and HIV-negative individuals were moderate (seizure, focal deficit), and the frequency of intracranial hypertension (the severest symptom) was similar between both groups ( Table 2). It is known that symptoms in NCC mainly depend on the intensity of the inflammatory reaction 52 . Thus, although it was suggested that immunosuppressed patients infected with NCC might present with fewer symptoms compared to HIV-negative immunocompetent patients infected with NCC [45][46][47][48] , and that symptoms might increase when CD4 count is recovered, as occurs in the immune reconstitution inflammatory syndrome (IRIS) and as it was shown in some published cases 33,53 , our data do not enable us to confirm this hypothesis.
Multiple vesicular parenchymal parasites were the most frequent radiological presentation in both groups of patients (HIV-positive and HIV-negative) ( Table 3). Some authors reported in preliminary studies an increased frequency of unusual NCC presentations in HIV-positive patients, such as giant cysts, racemose cysticerci or spinal localization (6% vs. 1% reported in HIV-negative patients). These unusual NCC presentations might be due to a parasite overgrowth allowed by the reduction in immunological response 25,53 . In our study, the only relevant difference was a higher prevalence of extraparenchymal parasites located at a spinal level. Indeed, in our NCC-HIV-positive sample, 2 of the 6 patients (33.3%) with extraparenchymal cysts (associated or not with parenchymal cysts) presented with spinal cysts. In one study describing NCC in HIV-negative patients, prevalence of spinal cysts was much lower at 3.36% (8/238) 7 . This difference is striking, and more studies are necessary to confirm this result and to, eventually, explore the factors possibly implied.
Regarding NCC patients' response to treatment, we did not find differences between HIV-positive and HIV-negative patients, nor between immunosuppressed and non-immunosuppressed HIV patients (Tables 4 and 5). This result is intriguing when considered in the context of the possible relevance of inflammation (particularly of Th17 response) in response to NCC treatment 54 . In HIV-positive patients, a decrease of Th17 cells, reflecting the depletion of total CD4+ T cells occurs 55 , which could modify patients' response to NCC treatment. The apparent contradiction between these data could be attributed to the low number of HIV/NCC subjects included in these studies, as well as to the lack of knowledge regarding the CD4 status of many of the patients.
In conclusion, the description of NCC-HIV-positive cases was quite similar to the results obtained in a previous study gathering 27 cases of NCC-HIV co-infection 33 . As in our study, the authors found that the most frequent presentation was multiple parenchymal lesions, and most of those who were treated responded to cysticidal therapy.
We could not identify any interaction between the two pathologies. However, as said before, it is clear that most of the published cases lack precise information, particularly regarding the intensity of the immunosuppression of the patients, the clinical-radiological and inflammatory status of NCC.
Despite the evident limitations of this study, the information reported herein is original and of interest. Implementing a multi-institutional study that may further increase the available data on individuals afflicted with both NCC and HIV and enabling stronger analysis and conclusions is required. Future research should also seek to answer questions about the clinical management of NCC/HIV co-infection, especially related to known drug-drug interactions between antiretroviral, antiparasitic, and antiepileptic treatments.