**4. Discussion**

In this study, the most frequently isolated microorganisms from *B. lanceolatus* oral cavity in Martinique were *A. hydrophila*, followed by other members of the *Enterobacteriaceae* family. The cultures

were polymicrobial and bacteria were susceptible to third-generation cephalosporins in most cases, coinciding with the majority of studies that focused on other snakes from different regions of the Americas [8,10–12].

*B. lanceolatus* is endemic to Martinique in the French West Indies. Its venom can cause severe local damage, ranging from blistering and tissue necrosis to secondary cellulitis and abscess in severe cases [1]. As previously shown [2–7], the oral microbiota of snakes is known to contain a wide range of microorganisms. Conformingly, *B. lanceolatus* oral flora is polymicrobial, combining Gram-positive, Gram-negative bacteria were isolated, with varying proportions across the Martinique sub-regions.

*A. hydrophila* was the most commonly isolated bacterial species, especially when the snake was captured from a "wet" zone. This bacterium is known to cause tissue damage and necrotizing fasciitis [13–15], raising the hypothesis that the tissue damage observed after *B. lanceolatus* bite may result from the tissue-damaging properties of the venom combined with the necrosis potency of *A. hydrophila*. It is obvious that tissue damage induced by the venom is responsible for perfusion abnormalities in the zone of the bite. This is due to microvascular damage and thrombosis induced by the venom itself, in addition to microcirculatory compression due to local edema and compartment syndrome in some cases. In these conditions, tissue damage and ischemia, with an associated decrease in the local innate immunity, is an excellent medium for the growth of bacteria ejected from the snake's oral cavity at the moment of the bite, as previously demonstrated for *Staphylococcus aureus* in an experimental model [16].

In the case of a moderate to severe snakebite, most authors and guidelines recommend the use of antibiotics to reduce complications by preventing secondary infection. However, the 2010 World Health Organization statements advised against the use of preemptive antibiotics in snake bites except in certain circumstances [17]. Our study clearly showed that envenomings by *B. lanceolatus* species have a high incidence of bite abscesses and, thus, represent an ideal candidate to use antibiotics, if they are given when there is the evidence of infection soon enough after the incident and when appropriate antibiotics are employed.

The most recommended antibiotic for snakebite treatment is amoxicillin/clavulanate, albeit without strong evidence supporting this recommendation. The Infectious Diseases Society of America (IDSA) guidelines for the diagnosis and managemen<sup>t</sup> of skin and soft-tissue infections recommended amoxicillin/clavulanate as the preemptive antibiotic of choice to reduce complications by preventing secondary infection from animal bites other than snakes [18]. However, preemptive amoxicillin/clavulanate was reported to be ineffective in preventing secondary infections from *Bothrops* snakebites in the Western Brazilian Amazon region [19]. Similarly, in another study based on a 10-year experience in a northern Taiwan medical center, amoxicillin/clavulanate alone appeared non-convenient for the empirical or definitive treatment of soft tissue infections after snakebite [20]. In this setting, the authors advised to use the combination of amoxicillin/clavulanate with ciprofloxacin or to choose parenteral piperacillin/tazobactam.

Our findings were consistent with these two studies [19,20], since we showed that amoxicillin/clavulanate was not effective against 66.7% of the isolated bacteria. By contrast, we found that the most appropriate first-line antibiotics were the third-generation cephalosporins that should be preferred in *B. lanceolatus*-bitten patients with signs of local infection. However, since the systematic prophylactic antibiotic use in snakebite patients remains a matter of debate [21], the use of antibiotics in *B. lanceolatus* cases should be restricted to cases where there is evidence of infection.
