We searched the literature in PubMed (www.ncbi.nlm.nih.gov/pubmed), EMBASE (www.elsevier.com/online-tools/embase), and Google Scholar (http://scholar.google.com/) for all published articles until January 2013, using the terms “breast-feeding” and “Chagas,” “breastfeeding” and “Trypanosoma cruzi,” “milk” and “Chagas,” and “milk” and “Trypanosoma cruzi.” No language, age, or sex restrictions were used. Both animal and human studies were considered. Direct access to some of the older publications was not possible, and information about results and conclusions was obtained from other articles.

As early as 1913, Nattan-Larrier reported finding T. cruzi in the milk of laboratory animals (12). Disko and Krampitz found T. cruzi trypomastigotes in the milk of experimentally infected mice, which was not attributable to contamination of milk with blood. However, they did not subsequently demonstrate infection of suckling mice (13). In 1972, Miles examined the milk of mice in the acute phase of the disease and detected T. cruzi trypomastigotes and antibodies against T. cruzi, but concluded that, experimentally, even in the acute phase of the disease, transmission through breast-feeding was rare (14). In another study, Campos et al. found that nursing mice born to lactating females in the acute phase of the infection were not infected (15). Another study on vertical transmission of T. cruzi among chronically infected laboratory rats did not demonstrate transmission through breast-feeding, and histopathologic examination of the mammary glands of the infected rats did not reveal the parasites (16). More recently, Martins et al. inoculated 15 female mice with Y strain T. cruzi trypomastigote forms a few hours after they gave birth, and breast-feeding of offspring was continued. Despite confirmed infection in female mice, no infection through breast-feeding was documented in the 142 offspring examined (examination of fresh blood by microscopy) (17). Other authors have not been able to confirm experimental transmission through breast-feeding during the acute phase of the disease (18–21), and researchers have hypothesized that maternal antibodies transferred with the milk may play a protective role (14,22).

Studies that examined pasteurization and microwave treatment of human milk to prevent transmission of Chagas disease have demonstrated transmission when human milk contaminated with Y strain trypomastigotes was orally administered to mice. However, such transmission occurred under experimental conditions, not natural breast-feeding, and the milk was artificially contaminated. In 1 experiment, batches of laboratory mice were orally or intraperitoneally inoculated with human milk that was contaminated with T. cruzi trypomastigotes and pasteurized, contaminated with T. cruzi and not pasteurized, or not contaminated with T. cruzi. Only mice that were given contaminated nonpasteurized milk (orally or intraperitoneally) were infected (detected by examination of peripheral blood) (23). In a similar study, mice that received either infected or noninfected microwave-treated milk (63°C, domestic microwave oven 2,450 MHz, 700 W), orally or intraperitoneally, were not infected (negative parasitologic and serologic results), whereas mice that received contaminated milk that had not been microwaved acquired the infection (24).

Studies involving metacyclic trypomastigotes from triatomine feces have found that contaminated milk seems to be an adequate medium (when compared with other types of food and water) for oral transmission of T. cruzi to mice, and this fact has been attributed to its moisture and nutrient content (25). In certain animal experiments, infection through the oral route (not breast-feeding) has been established by using blood trypomastigotes, and the success of this mode of transmission appeared dependent on the size of the inoculum used (26,27). Transmission of T cruzi by breast-feeding is possible in animal models, but oral infections may be difficult to establish unless metacyclic forms of the parasite are involved.

Two reports from 1936 and 1983 describe finding trypomastigotes in the milk of mothers in the acute phase of Chagas disease. Mazza et al. (5) reported transmission through lactation; in the other study, the newborn was not breast-fed (28). However, another publication later indicated that in the study by Mazza et al., the collected milk was contaminated by blood (29). Medina-Lopes reported 2 cases of acute Chagas disease in infants that had been acquired through breast-feeding from mothers in the chronic phase. In 1 of these cases, no parasites were found in the milk, and the source of the infection was postulated to be infected blood because the mother had bleeding nipples. In the other case, congenital, vector-borne, and transfusion-associated transmission were excluded, and transmission was suspected to have occurred through breast-feeding (6,7).

In 1988, Bittencourt et al. were unable to demonstrate transmission of infection through breast-feeding. Samples of milk/colostrum from 78 mothers in the chronic phase of the disease were studied (5 mothers had parasitemia at the time of milk collection). Results of the following were negative: the parasitologic study of all sample-inoculated mice and the serologic tests of serum samples from 97 breast-fed children born free of infection (22).

In 1990, Shikanai-Yasuda et al. published a series of reports of cases of acute Chagas disease, and breast-feeding was considered a possible route of transmission for 2 patients (30). Amato Neto et al. (31) could not document the presence of T. cruzi in the colostrum and breast milk of 40 women with Chagas disease through direct observation, culture, and inoculation. The possibility of using a larger study population and/or more efficient diagnostic techniques was highlighted (31).

Other studies that examined maternal transmission of T. cruzi included breast-feeding as a possible contributing factor, but once again, this mode of transmission could not be demonstrated. In 2004, Rassi et al. found 2 cases of maternal transmission of infection in 2 patients who had also been breast-fed, but this mechanism could not be established, and the infections may have resulted from vertical transmission (32).

The data on transmission of T. cruzi through breast-feeding in humans are scarce; reports are not recent and the findings have several limitations. Additional data on methods used to diagnose infection and exclude other forms of transmission in these studies are summarized in the Table.

In mice, oral transmission of T. cruzi infection through human milk contaminated with trypomastigotes is possible, although natural transmission through breast-feeding has not been clearly demonstrated in recent studies. Infection in laboratory mice acquired after oral or intraperitoneal administration of human milk contaminated with T. cruzi trypomastigotes may be prevented by thermal treatment (pasteurization or microwave treatment at 63°C) of milk, which can inactivate the trypomastigotes. Such treatment of milk may therefore be an option for lactating mothers who have the acute form of the infection or bleeding nipples.

In humans, the oral route is an efficient mode of transmission for T. cruzi, as has been demonstrated in recent outbreaks of infection acquired through contaminated food and fruit juices (33–35). Ingestion of food or drink contaminated with vector feces (containing metacyclic trypomastigotes) or crushed insects is the most likely mode of oral infection (36). The concentration of the ingested inoculum may result in the different clinical manifestations and variations in attack rates reported in outbreaks (34).

Milk appears to be an adequate medium for transmission of T. cruzi because of its moisture and nutrient content and, in the case of human milk, body temperature would also be favorable (the parasite can be destroyed in dry vector feces and by high temperatures) (23,24,36). Despite this, except for some dated cases, in humans, transmission through breast-feeding has not been reported.

A further consideration to account for the relative lack of infections transmitted through breast-feeding is that, if they occur at all, blood-form trypomastigotes would be expected to occur in human milk and not metacyclic trypomastigotes, which are found in the vector. Blood, metacyclic, and culture trypomastigotes are similar in form and structure, but the different forms may have different surface glycoproteins, which are used for cell invasion (17,37). Metacyclic trypomastigotes may have specialized surface molecules (such as gp82, a member of the gp85/ trans-sialidase superfamily, or pepsin-susceptible gp90 isoforms, which make parasites highly invasive against target cells after contact with gastric juice) with functions for mucosal invasion that are not present in blood-form trypomastigotes. The latter, in turn, express other molecules that facilitate dissemination within the host (38–40). Even though trypomastigote forms of the Y strain of T. cruzi, which expresses glycoprotein 82 in metacyclic forms and should have high infectivity, were used in some experiments, transmission through breast-feeding could not be demonstrated. Blood-form trypomastigotes may therefore not be as efficient as metacyclic trypomastigotes in establishing mucosal invasion of the gastrointestinal tract (partly because of their decreased capacity to migrate through the gastric mucin layer), even though oral infections with blood trypomastigotes have been demonstrated experimentally with the possibility of infection depending on amount of inoculum. Peptic digestion may also destroy some of the infective forms, mainly blood trypomastigotes (40).

On the basis of the data examined, oral infection through breast-feeding does not appear to be an efficient mode of transmission for T. cruzi. In humans, milk contamination by T. cruzi has been described only in isolated cases. In the acute phase, patent parasitemia occurs, but T. cruzi has been demonstrated in breast milk only in 1 case (transmission did not occur because the newborn was not breast-fed) (27). In the chronic phase, parasitemia is usually low grade and tends to fluctuate. A couple of cases of acute infection in breast-feeding infants of mothers in the chronic phase have been reported, but transmission may have resulted from the ingestion of blood from bleeding nipples (6,7). In older studies, it is not always clear how accurately vertical and other modes of transmission had been excluded with the diagnostic tools available at the time, and in other cases, breast-feeding is contemplated only as a possible route of infection. Many of the studies reported were performed some years ago, and current novel techniques, such as PCR, may aid in detecting low parasite loads. On the other hand, new techniques may even yield equivocal results (e.g., because of the presence of infected blood but not true contamination of milk by T. cruzi) if contamination with infected blood were below the level of detection found with other methods. In humans, transmission of T. cruzi through maternal milk (not contaminated with infected blood) may not be considered clearly proven.

As previously postulated in laboratory studies, in humans, a combination of several factors likely account for the low efficiency of breast-feeding as a possible mode of transmission of T. cruzi infection. These include the low number of trypomastigotes ingested by the lactating infant, the biologic forms and strains of T. cruzi involved, the passage of protective maternal antibodies, and the destruction of some parasites by gastric juice.

A limitation of this review was that some of the articles dated back several decades and were not readily available. Some information about these studies was obtained from other articles.

Published studies on the transmission of T. cruzi through breast-feeding are scarce. After reviewing the various studies, and given the demonstrated benefits of breast-feeding (especially in lower income disease-endemic countries where a safe and inexpensive alternative may not be available), we conclude that discontinuing or interrupting breast-feeding by mothers with chronic Chagas disease is not recommended. If the mother has fissures or bleeding nipples, temporary discontinuation of breast-feeding may be recommended, although thermal treatment (pasteurization or microwaving) of expressed milk before feeding the infant may be a safe alternative. When mothers are in the acute phase of the disease (an infrequent situation that will occur almost exclusively in disease-endemic countries) or have reactivated disease (resulting from immunosupression, in which circulating parasitemia level may be high), breast-feeding may pose a risk for the infant. The benefits of using thermally treated expressed milk should be considered on an individual basis. To avoid any possibility of transmission, human milk banks should exclude mothers with Chagas disease as donors. As other modes of transmission of Chagas disease come under control (vector-borne transmission and transmission through infected blood/organs), further studies are needed on the possibility of transmission through breast-feeding, as this could occur even in countries where the disease is nonendemic.