Leishmaniasis comprises a group of infectious diseases that may occur with cutaneous or visceral involvement. They are caused by protozoa of the genus Leishmania transmitted to humans by infected female sandflies. American cutaneous leishmaniasis (ACL), cutaneous presentantions in the New World, clinically presents quite variable lesions, from the acneiform type to ulcers with or without lymphadenopathy. The typical cutaneous lesion is an ulcer (usually a single one) with infiltrated borders in exposed areas of the body, but clinical presentation may vary depending on the immune status of the host, the parasite load, and the involved Leishmania species [1]. Treatment is indicated in confirmed cases but the recommended drugs are quite toxic [2–4], and in recent years cases of resistance have been described [5–8].

The typical ulcer of cutaneous leishmaniasis (CL) can progress to cure over a period of several months to years, if not diagnosed and treated [9]. However, some patients have early spontaneous resolution of the disease; they are diagnosed according to clinical and laboratory criteria and, before the start of drug therapy, the lesions begin the process of clinical cure, and may not require treatment [10, 11].

The Leishmaniasis Surveillance Laboratory (LAPCLIN VIGILEISH) at Evandro Chagas National Institute of Infectious Diseases (INI), Oswaldo Cruz Foundation (Fiocruz), Rio de Janeiro, receives patients with suspected ACL from the state of Rio de Janeiro and from other regions of Brazil. Patients with early clinical signs of healing without specific treatment and those treated are invited to participate in a follow-up of at least 5 years with dermatological and otorhinolaryngological annual exams, creating conditions for prompt identification of signs of recurrence of skin lesions and/or development of mucosal involvement. This study allowed the comparison of clinical, laboratory and follow-up data of the leishmaniasis patients with early spontaneous healing with those of the patients that needed specific treatment.

Four hundred and forty-five patients were analyzed, with a median age of 37 (15–80) years, and 63.40% were men. Most patients (64.50%) came from the state of Rio de Janeiro (Brazil).

Considering all the studied cases, 64.70% of the patients had a single lesion. Lesions affected different sites of the body: exclusive on the lower limbs (35.10%); exclusive on the upper limbs (30.10%); restricted to regions of head and trunk (21.80%); and multiple locations (13.00%). The presence of ulcers was prevalent at diagnosis (n = 394; 88.50%), but in 9 cases epithelialized lesions were observed in the initial clinical examination (2.00%). The median of the largest diameter of the lesions was 30 mm (4–120 mm). The median MST measure was 17 mm (2–71 mm), and 82.50% of the patients were considered positive in this exam. The median time of progression of the lesion previous to diagnosis was 60.87 days; and the median follow-up period of all cohort was 803 days. The median time between the first and the second medical appointment was 29 days in the whole studied cohort. The median follow-up period of NESH group was 990 days and in the ESH group was 631 days.

Four hundred and thirty-two patients (97.10%) were treated for ACL, while in thirteen individuals (2.90%) the spontaneous healing phenomenon was observed. In the NESH group, 430 patients were treated with meglumine antimoniate (99.53%), and 2 cases were treated with amphotericin B deoxycholate (0.47%).

When compared with the NESH group, the ESH group showed a higher percentage of single lesions (p = 0.027) and epithelialized lesions in the first clinical examination (p = 0.001). In 30.80% of the patients in the ESH group the lesion was located in the dorsal trunk, and this was observed in 7.60% of the cases in NESH group (p = 0.017). There were no significant differences between the median times to wound healing in patients who healed spontaneously with lower limb injuries and those with lesions in other locations, p = 0.534.

In the NESH group, there was a predominance of ulcerated lesions (89.60%, p = 0.002), the finding of amastigotes in the histopathological exam (62.60%, p = 0.005), and positive culture (84.60%, p = 0.001). Most of these patients had a single lesion on the lower limbs (42.90%). In the NESH group, the mean time to healing of the lesions in the lower limbs (237 days) was longer than that of lesions at other locations (201 days), p = 0.012.

Considering the concordance concerning the positivity in each test, there were statistically significant differences between the NESH and ESH groups regarding the IFA (p = 0.004), ELISA (p = 0.002), MST (p = 0.05) and culture (p = 0.001). In the ESH group we observed differences between the parasitological confirmed cases (non-PCR) and the positivity in the PCR test (p = 0.002) (data not shown).

None of the 13 ESH cases presented recurrence of the skin lesions or mucosal involvement during the follow up period. From the 432 patients in the NESH group, 23 relapsed (5.32%) and 11 showed mucosal lesions (2.54%) in follow up.

Results of survival analysis

The occurrence of clinical cure of the lesion according to the healing time was evaluated through the Kaplan-Meier method (Fig. 2). After the application of censorship criteria, 10 patients were excluded and the analysis were performed on 435 patients (9 patients from ESH group and 426 patients from NESH group).

In all 9 cases of spontaneous healing analyzed by this method, wound healing occurred within approximately 3 months (91 days). As regard to the NESH group, 10 out 426 (2,34%) subjects presented prolonged follow-up (at least 256 days). All 10 cases were censored at the end of the study without presenting the event of interest (healed lesion). There was a statistically significant difference between the groups using the Kaplan-Meier method, evidenced by Peto and Log-rank tests (p-values <0.001). Thus, the ESH group showed earlier clinical cure during the observation period.

The diagnosis of ACL in the cases of early spontaneous healing evaluated in this study was confirmed by the identification of the parasite through one of the used parasitological methods. This allowed a study of cases of ACL with spontaneous resolution compared with those who needed specific treatment for the healing of their cutaneous lesions. We found a very low percentage of early spontaneous healing among a historical cohort of patients in a reference center for the disease in Rio de Janeiro, Brazil, where most of the cases are caused by L. (V.) braziliensis. Recently, a systematic review of ACL also showed a low percentage (6.0%) of spontaneous healing in CL caused by this parasite [14].

The cases of the ESH group had the highest percentage of single lesions and fewer ulcerated lesions in the initial assessment, and they had less positive results in the confirmatory parasitological tests, differences that were significant in relation to the NESH group. Furthermore, a higher positivity for the presence of parasites by histopathology and culture when the treatment was required was demonstrated, plus a greater percentage of strong reactors in MST. In medical literature, the sensitivity rates of histopathology and isolation in culture medium varies widely [15–17], but the presence of higher parasite loads enhances its detection and thus the parasitological confirmation of the case.

The number of cases of ESH positive only in PCR indirectly suggests that the parasite load tends to be lower in these patients, indicating a balance between the host inflammatory process and the parasite load. A most effective infection control would lead to spontaneous healing. As a result of the probable ability to control the parasite load and the inflammatory process, the cases of ESH had a lower average healing time than that observed in the NESH group. Moreover, the findings suggest that the higher parasite loads in the NESH patients would exert pressure on the inflammatory response, leading to stimulation of the maintenance of the activity of the inflammatory process, which slows its control and subsequently the wound healing. The involvement of the inflammatory response in the formation and maintenance of the leishmaniasis lesions has been demonstrated in a murine model [18]. In humans, although significant difference has not been demonstrated between the amount of the type 1 cytokine interferon gamma (IFNγ) in the peripheral blood of patients with cutaneous leishmaniasis cured after treatment and in the cases with spontaneous healing, in the last ones there was a tendency to higher values [19]. The balance between type 1 and type 2 immunological responses has been suggested as determinant in the evolution of ACL for self-limited or severe forms [20, 21].

The need for specific treatment and the response to the antimonial therapy have been the focus of discussions. Studies in Latin America show significant differences in cure rates with pentavalent antimony (PA) [3, 6, 8, 22–24]. It has also been reported that early treatment of ACL does not prevent the development of ulcers [25, 26] and is associated with higher treatment failure rates [26]. Short evolution times prior to treatment and weak positive MST were both associated with treatment failure [27]. Other factors have also been found to be able to influence the therapeutic response to PA. Among them we can mention the species of the involved parasite, comorbidities, presence of three or more lesions, body weight above sixty-eight kilograms and the irregularity of treatment [5, 28–33]. However, the use of second choice drugs has not shown greater efficacy, or even lower rate of adverse effects. The high toxicity of PA and the increasingly described treatment failure rates has put into discussion the directions for the treatment of ACL.

Criteria for cure, which are essentially clinical, vary according to different authors. This could influence the wide variation in the evaluation of the success of PA therapy. When there is evidence of progression to healing, the extension of the follow up without re-treatment for up to six months is suggested [34, 35]. Therefore we may question the need to propose specific treatment in cases with evidence of progression to early spontaneous healing. Many authors indicate the treatment even in cases of spontaneous regression of the lesions, due to the potential capacity of development of secondary metastatic lesions, especially to mucous membranes, in cases of infection with L. (V.) braziliensis [34, 36]. Mucosal leishmaniasis (ML) occurs in a small proportion (3 to 5%) of the ACL cases in Brazil, and can occur several years after the healing of the primary lesion [37]. It is mainly attributed to hematogenous metastases [38]. However, some of these cases share no history of prior cutaneous lesions, and ML is then considered of indeterminate origin [39]. Spontaneous cures and irregular treatments have been considered risk factors for the emergence of ML [10, 39]. However, thorough and longer term follow-up studies are needed. Moreover, in our study, ESH patients did not receive any specific treatment and yearly follow up of 5 post-treatment years did not show any sign of relapse of skin lesions or the appearance of even incipient mucosal lesions in these patients.

Some authors demonstrated that ACL lesions located in the lower limbs needed more time to heal [40, 41]; however, there were greater proportions of healed lesions in this location [16]. Our results also showed that lesions in the lower limbs in treated patients demanded higher healing time than lesions in other locations. Location in the dorsal part of the trunk most rapidly evolved into healing in all patients, and was more often observed in patients with spontaneous resolution. The location in the lower limb did not appear to influence the healing time in patients with early spontaneous healing.

In conclusion, our results suggest that cases of ACL with progression to early spontaneous healing have an apparently lower parasite load evidenced by the difficulty of parasitological diagnosis during the investigation of the case, since most of them were positive only in the PCR method. Whether this is due to an inoculation of a lower number of parasites during the sandfly bite(s) or a more efficient immune response of the infected persons cannot be defined up to now. Due to obvious signs of healing of the lesions we chose not to initiate specific treatment, since leishmanicidal drugs have great potential for toxicity. Even so, healing of lesions demanded less time than in patients who required treatment. It must be highlighted that even after a long period of follow up post-regression of the lesions, none of the ESH studied cases studied developed recurrence of the lesions or mucosal damage.

The cases of patients with spontaneous resolution in ACL are a challenge in clinical practice. Dermatological/otorhinolaryngological periodic examination of these patients, as well as clear guidance to patients about the signs and symptoms of mucosal lesions are recommended to detect the slight possibility of a relapse.