High levels of anti-Leishmania IgG3 and low CD4⁺ T cells count were associated with relapses in visceral leishmaniasis

Visceral leishmaniasis (VL) is caused by Leishmania (L.) infantum in Brazil, being transmitted to mammals by Lutzomyia longipalpis sand fly [1]. VL is endemic in more than 90 countries or territories, however, in 2017, few countries as Brazil, Ethiopia, India, Kenya, South Sudan and Sudan have concentrated more than 90% of the cases [2]. In the Americas, 3562 new cases of VL were diagnosed in 2018 and Brazil was responsible for 97% of them [3]. Belo Horizonte, located in Minas Gerais state, is one of the Brazilian cities with the highest number of VL patients, with 2378 reported cases from 2007 to 2019 [4].

Pathogen and host’s immune system interaction leads to different clinical presentation, predisposing a high variety of outcomes, since asymptomatic disease to high severity and risk of death. Commonly, most patients affected by VL respond well to anti-Leishmania treatment and evolve to the remission. However, there are those who evolve to the severe form of VL, with a high lethality rate [5]. In addition, relapse is also observed in medical practice [6, 7].

Clinical and laboratory markers such as age, bleeding, edema, jaundice, dyspnea, bacterial infection, HIV/AIDS co-infection, leukocyte count below 1500 cells/mm³, thrombocytopenia below 50,000 cells/mm³ and renal failure have also been linked to VL severity [8, 9]. Also, several soluble molecules, such as elevated levels of IL-2R, IL-1β, IL-6, IL-8, IL-27, TNF, soluble CD14 (sCD14) [10‐13], soluble CD163 (sCD163) [14] and specific anti-Leishmania immunoglobulins [15, 16], as well as low leptin levels [17] have already been associated with severity of VL. Interestingly, IL-6 levels were also associated with risk of death in VL [12].

Indeed, besides parasite specific immunosuppression, the exuberant inflammatory condition constitutes the key mechanism in the physiopathology of L. infantum infection. Similar to what is seen in sepsis, severe dengue and severe malaria, high degree of cell activation and high levels of cytokines are seen in the active VL [12, 18‐20]. These mechanisms contribute to the immune response impairment, which in turn have a negative impact on the effector capacity to control the parasite. Therefore, these features can influence the clinical evolution of VL patients, in terms of clinical cure status, severity/death and disease recurrences/relapses.

Relapses are characterized by the resurgence of signs and symptoms after an initial improvement of a disease manifestation [21]. It is considered a risk factor for death in VL, especially in HIV-co-infected patients [22, 23]. Recently, our group demonstrated that the maintenance of high levels of cell activation, microbial translocation products, anti-Leishmania IgG3 and a low CD4⁺ T cell reconstitution could be associated with VL relapse in VL/HIV coinfected patients [23]. Moreover, this low immune reconstitution has been related to a greater impairment of the thymic output among VL/HIV-relapsing patients [24]. The inability to reconstitute the effector response seems to exert a key role in the VL relapses.

In Brazil, the official frequency of relapses after VL treatment among non HIV-infected patients is underestimated, although the compulsory notification form has a specific field to inform this clinical presentation [25]. In the literature, the vast majority of studies are related to therapeutic failure reports [26]. There are few VL relapses reports and those which have it, present a great variability in the incidence of relapses according to the population and regions studied [27‐30]. In a Spanish cohort, VL relapse rate was 12% [27] while in Sudan, the clinical follow-up during an epidemic in Babar showed 5.7% of relapses, most of patients concentrated in few families [31]. A previous Brazilian study showed relapse in 2.3% of children diagnosed with VL between 2006 and 2011 [30]. Based on secondary data, an analysis of VL cases also in Brazil from 2001 to 2010 showed that relapses occurred in 3.1% of patients not co-infected with HIV [29].

Factors underlying clinical relapses in VL alone have not been deeply addressed. Clinical characteristics as male gender, extreme of age and discreet reduction of splenomegaly were risk factors associated with relapses in Indian VL patients [32, 33]. In Georgia, a pediatric cohort showed that children younger than one-year-old had a higher chance of relapse as the ones who were late diagnosed [34]. More recently, a Brazilian study showed that in addition to HIV infection, thrombocytopenia, lower limb edema and secondary pneumonia were factors independently associated with relapse [28].

VL relapsing patients should not achieve an effector immune response able to maintain the remission of the disease. Herein, we actively monitored patients with VL, followed from the active phase of the disease  up to 12 months after treatment and confirmed the presence of two distinct groups: those with primary VL without relapse along 6 months (assumed as single/life episode) and those with two or more VL episodes/life, either with relapse identified at the cohort enrollment or during the follow-up period. Our aim was to identify possible differences in clinical and immunological parameters that could infer mechanisms involved in recurrence of VL. We evidenced that after therapy, both groups improved hematological and biochemical parameters and reduced IL-6 levels. In contrast, VL relapsing patients maintained increased IgG3 levels besides lower CD4⁺ T cell counts in comparison to non-relapsing VL patients. These results suggest that VL relapsing patients keep a B cellular activation status along with a deficient T cell compartment reconstitution.

There are many studies carried out in India [33], Sudan [32], Georgia [34] and Brazil [28] addressing retrospectively non-HIV infected patients presenting a VL relapsing course, but divergent findings in terms of clinical and laboratorial aspects associated with relapses were found. In the present study, the patients’ symptoms were quite similar in the active phase, and it was not possible to identify significant differences between R- and NR-VL groups. However, more intense laboratorial abnormalities were observed in NR-VL compared to R-VL group. As R-VL patients have previously undergone anti-Leishmania treatment, we believe that it may have contributed to a partial clinical recovery, with a reduction in the damage caused by the infection. However, this was not enough to restore the R-VL organic functions.

Splenomegaly at study enrollment was observed in both groups, being a usual manifestation of VL [9]. Splenomegaly and thrombocytopenia have already been identified as markers of VL relapses [28, 32, 33]. NR-VL patients showed significant platelet elevation after treatment, different from observed for R-VL group. It is known that spleen histological disorganization affects its functionality [35] which implies in higher consumption of platelets, as a confounding factor, resulting in elevated parasitemia [9], and immunodepression related to its immunity function [32].

The responsible medical team decided each patient treatment in an individualized way, based on clinic characteristics and side effects of the avaiable drugs. It is important to highlight that although Amph. B Lipid Complex is not considered a drug of first choice in VL treatment in international and Brazilian official guidelines, there are studies that were able to show the effectiveness of this drug in VL cases, including in HIV coinfected patients [36‐39]. Therefore, we do not believe that the treatment with this specific Amph. B formulation is related with the occurrence of new VL episodes in these patients.

After VL treatment, a recovery of bone marrow cellularity was observed in NR-VL patients, especially with an increase in leukocytes levels compared to the active phase of VL. A significant and faster gain of CD4⁺ T cells was observed in NR-VL patients immediately after anti-Leishmania treatment, but not in R-VL patients. Similar results were observed in VL/HIV coinfection [23]. This deficit in the recovery of CD4⁺ T lymphocytes can be related to an impaired input of cells originated from bone marrow, deficient replication rate or thymic dysfunction [24, 40, 41]. Low CD4⁺ T cell counts in VL/HIV coinfected patients during the active phase of VL was a predictor of a poor prognosis: death or recurrence [42, 43]. Thus, patients who are unable to restore the CD4⁺ T counts may be prone to relapse. Interestingly, in the current series, all VL patients maintained low CD8⁺ T lymphocyte counts up to 6 months of follow-up, in relapsing and not relapsing groups. Whether CD8⁺ T cells contribute to protection, immunopathogenesis or even to the immunosuppression in VL is still unclear. Likewise, the function of these cells has already been described for human cutaneous leishmaniasis during the active and healing phases [44‐47]. Some studies have showed that CD8⁺ T lymphocyte cytotoxic activity, secretion of cytokines and chemokines, granzyme B levels and lymphoproliferation may contribute to the parasite control in experimental and human VL [48‐51]. At the same time, there was a positive correlation between the CD8⁺ T lymphocytes and transaminases levels, which raises the hypothesis that liver damage could be associated with cytotoxic activity. Herein, this occurred regardless of the VL clinical outcome. Such an association has already been demonstrated in hepatitis B [52] and in infectious mononucleosis by Epstein-Barr virus [53].

Also, endogenous IL-10 secretion has already been ascribed to the CD8⁺ T cells in VL which may contribute to the immunosuppression. Simultaneously, the chronic immune activation may lead CD8⁺ T lymphocytes to express molecules with inhibitory function on their surface, such as PD-1, CTLA-4 [54, 55], TIM-3 and LAG3 [16]. In these situations of exhaustion of the cellular response, CD8⁺ T lymphocytes function may be severely impaired in active phase of VL [56].

In parallel, the levels of specific anti-Leishmania IgG and their subclasses after active VL may have an important role as marker of cure and/or predisposition to relapse [17, 57]. Here, IgG3 levels were persistently increased in R-VL group in relation to NR-VL, which may indicate  the presence of continuous parasite stimulation, without a complete control by the immune system. L. donovani Indian relapsing patients [58, 59] have shown similar behavior for IgG1, whose levels remained high after the active episode of VL. In other studies [60‐62], a reduction in IgG1 and IgG3 levels was found in patients infected either by L. infantum or by L. donovani, who were considered clinically cured few months after the treatment. In addition, we have found that a negative association between IgG3 and CD4⁺ T-cell count in R-VL patients, corroborating the link between persistence of high levels of IgG3 and relapse.

Finally, we observed that IL-6, a cytokine known to be inflammatory and already associated with a more severe prognosis in the evolution of VL [12, 17], was correlated to laboratory severity markers, such as hypoalbuminemia, leukopenia and thrombocytopenia. As expected, CRP levels also correlated positively with IL-6 since its binding with its receptor activates immunocompetent and hematological cells leading to the production of acute phase responses, as CRP [63]. However, there were patients that despite presenting clinical signs and laboratorial data associated with severity did not present elevated IL-6 levels, showing that other mechanisms in VL can trigger immunopathological mechanisms related to severity. This study design and the analysis performed until now are not sufficient to state the role of IL-6 may in the VL relapsing course. This study has limitations, including the small sample, the short follow-up time and mainly the comparison between patients enrolled at the primary but also at a relapsing VL episode, which can add differences in the total time of infection and in previous exposure to anti-Leishmania treatments. However, our results are hypothesis generator and rise important questions to be evaluated in a prospective study, as the potential of CD4⁺ T-cell count or IgG3 as biomarkers for VL relapses.

Therefore, the VL relapsing course among patients not infected with HIV and without other recognized associated immunosuppression is a challenging reality considering the lack of understanding of the pathophysiological mechanisms involved, of reliable prognostic markers and clinical protocols for addressing the condition. Our findings suggest a deficit in T-cell reconstitution and maintenance of B-cell compartment activation as possible immunomechanisms underlying the VL relapse.