Elevated prevalence of multidrug-resistant gram-negative organisms in HIV positive men

Multidrug–resistant gram–negative organisms (MDRGN), which have previously been defined as Enterobacteriaceae with extended spectrum beta–lactamase (ESBL)–phenotype, and Enterobacteriaceae, Acinetobacter baumannii and Pseudomonas aeruginosa resistant against piperacillin, any 3rd/4th generation cephalosporin, and fluoroquinolones +/– carbapenems [1], are frequent causes of community and healthcare-associated infections [2, 3]. In recent years, carbapenems became the drugs of choice for the treatment of invasive MDRGN infections [4]. At the same time, the consumption of carbapenems has been increased dramatically [5] and the worldwide prevalence of carbapenemase-producing organisms soon after was found to be elevated [6, 7]. While the explosive spread of carbapenemases has previoulsy been attributed to several factors, such as increasing administration of carbapenems or contact to the health care system in high-prevalence countries (HPC) [8‐10], global spread might also be favored by plasmid encoded resistance genes transfer [11]. However, comprehensive knowledge of the ways of transmission is still lacking. Future concise considerations on the spread of MDRGN should therefore also evaluate routes of transmission that have not been noticed so far. We hypothesize that microbiota, with MDRGN might be part of it, is exchanged during mucous membranes contacts, e.g. sexual intercourse, Although the primary route of MDRGN transmission might be different, sexual transmission should therefore not dismissed. This hypothesis has previously been confirmed by investigations on the sexual transmission of Human Herpes Virus Type 8, E. coli O117:H7 and Shigella spp., although the primary route of these agents’ transmission is via saliva or feco-orally, respectively [12‐14]. It additionally should be taken into account that it has previously been reported that the risk to acquire “classical” sexually transmitted diseases (STD) such as gonorrhea or syphilis, or infection with the Human Immunodeficiency Virus (HIV) is higher in cohorts of men who have sex with men (MSM) than for the general population [15, 16]. This has also been demonstrated for Germany [17, 18]. With regard to these aspects, pre– or co–existing STD (e.g. HIV infection) have been demonstrated to be suitable parameters to identify individuals with previous or current unprotected, “risky” sexual intercourse. We therefore hypothesize that MDRGN prevalence might be higher in patient groups with HIV infection compared to patient groups tested negative for HIV. In light of the persistent global spread of MDRGN, possibly with a number of different routes, the role of sexual transmission has to be clarified. We present the first study addressing this possible transmissibility of MDRGN by sexual intercourse.

In the current study, we evaluated n = 109 male HP and n = 109 male HN, respectively, admitted to UHF between November 2014 and March 2016. Both the male HPs’ and male HNs’ median age was 46 years, the standard deviation being 10.8 and 11.0 years, respectively.

Table 1 shows the general groups of diseases HP and HN they were admitted for to UHF. In HP, the most frequent group were patients admitted to the department of Internal Medicine II/Infectious Diseases with 82.6% (n = 90/109). HN were most frequently admitted to UHF due to oncological diseases with 39.4% (n = 43/109). Among male HP and male HN, a total of n = 35 MDRGN isolates were detected. Thereof, n = 26/109 and n = 9/109 rectal swabs were positive for any MDRGN in male HP and male HN, respectively, which resulted in an overall MDRGN prevalence of 23.9% (16.2–32.9) in male HP and a significantly (p = 0.002) lower MDRGN prevalence in male HN with 8.3% (3.8–15.1). Looking into the single proportions of the detected MDRGN, no significant differences were detected between male HP and HN patients, respectively (Fig. 1).

The most frequently detected species within the MDRGN isolates (n = 35 in total) was E. coli with resistance due to ESBL expression and additional resistance to fluoroquinolones (E. coli ESBL/FQ) with n = 25/35 (71.4%; 53.7–85.3). Thereof, E. coli ESBL/FQ was detected in n = 19/26 (73.1%; 52.2–88.4%) of male HP and in n = 6/9 (66.7%; 29.9–92.5%) of male HN (p = 0.713), respectively. All MDRGN detected in male HP and male HN are depicted by in Table 2. None of the bacterial species P. aeruginosa or A. baumannii and neither, no resistance to carbapenems were detected in any of the isolates from male HN nor any of HP.

In light of the explosive global spread of MDRGN, attention must be directed to potential routes of transmission that have not been recognized so far. Several routes of transmission, e.g. traveling in HPC [8], antibiotic selective pressure [5], or transmission by livestock [21] have previously been demonstrated. However, to our knowledge this is the first study focusing on the potential transmission of MDRGN in male HIV positive individuals.

The prevalence of MDRGN in male HN amounted to 8.3% (3.8–15.1; Table 2). This value does not substantially deviate from the prevalence of ESBL of about 6.3% in clinical specimen, which has formerly been described to be 6.3% by a German cross–sectional study [22]. However, the prevalence of MDRGN in rectal swabs of male HP (23.9%; 16.2–32.9) exceeded the value investigated by Valenza et al. [22] by almost three-fold. These findings suggest that HIV positive men might be more susceptible to acquire MDRGN. Furthermore, our findings indicate a pre–dominance of E. coli ESBL/FQ (Table 1), it might therefore be interesting to characterize these E. coli-isolates isolated from male HP further for expression of pathogenicity factors, e.g. type 1 fimbriae (fimH), pili associated with pyelonephritis (pap), S and F1C fimbriae (sfa and foc), afimbrial adhesins (afa), cytotoxic necrotizing factor (cnf), hemolysin (hly) or aerobactin (aer) using molecular techniques [23] and subsequently compare these findings with isolates collected from HN. We therefore feel it is warranted to further characterize MDRGN isolates isolated from HP and HN by molecularbiological methods, preferentially by whole genome sequencing (WGS).

Although no significant differences in proportion of any single MDRGN species were observed (Table 2), this might indicate that the sexual route of transmission is apparently not preferred by any particular MDRGN species. It rather might be interpreted as a general inherent risk to acquire a MDRGN. Since the sexual route of transmission of MDRGN has not been suggested so far, this is as important finding of this study. However, several epidemiological events give an impression that sexual spread of organisms is possible, although their preferred way of transmission is reported to be non-sexual: Shigella sonnei, which is either none of the “classical” STD-causing organisms, has been shown to spread among men who have sex with men (MSM) in Montréal, Canada, and Germany [24, 25] the verocytotoxin-producing EHEC O117:H7 has also been documented to spread among MSM in Great Britain [16] and, as well, no “classical” STD, meningococcal serogroup C diseases has newly been shown to spread among MSM [26, 27]. This outlines that sexual transmission of Enterobacteriaceae is distinctly possible via translocation of intestinal bacteria. Sexual transmission of MDRGN is therefore suggested not to be dismissed in future comprehensive considerations on MDRGN transmission prevention.

However, the link between HIV positivity and risky sexual behavior must be interpreted diligently: if positive HIV serology might indicate that the patient has previously had a risky behavior (and HIV was not contracted e.g. by blood transfusion), this may have lasted for years up to now, or it may have ended in the meanwhile. The time of carriage of MDRGN might be short [28] and we therefore cannot ascertain that the timeframes between the risky behavior and MDRGN acquisition could coincide. However, at this time we did not see any alternative assessment available to clarify the potential route of sexual transmission.

Concerning that data regarding the history of previous hospital stays nor antibiotic pre–treatment for male HP and male HN were not evaluated, it is not possible to confirm that male HP in the UHF–setting might have had a stronger history of antibiotic pre–treatment than male HN. Considering that all patients enrolled in this investigation matched the requirement to be screened for MDRGN (which is usually limited to patients admitted to ICU or IMC at UHF), this might indicate that both male HP and male HN had a recent history of antibiotic treatment. However, HIV patients in particular might have undergone a prexisiting antibiotic pressure due to their HIV infection (e.g. administration of cotrimoxazole to treat Pneumocystis jirovecii pneumonia) and might therefore have a higher risk for admittance to hospitals and health care facilities where nosocomial MDRGN transmissions can occur [29]. These aspects might therefore have introduced a source of biases in the selection of patients.

However, not also antibiotic treatment, but also pre–existing co-morbidities also might be an aspect that has to be evaluated in a future setting to round off investigations on conditions of sexual transmission of MDRGN. Since gastrointestinal diseases or co-infections might compromise the gastrointestinal mucosa, this might promote the transmission of MDRGN. As such interrelation has formerly been shown for amebiasis and HIV [30], this aspect should not be neglected.

Furthermore, it remains open to question whether these findings are also valid for women. As urinary tract infections (UTI) are one of the most common (usually endogenous) bacterial infections often triggered by sexual intercourse, and many women suffer from at least one UTI during their lifetime [31], in a future setting the prevalence of MDRGN should be investigated in urogenital and/or vaginal swabs in female cohorts.

Our study demonstrated that prevalence of MDRGN in HIV positive male individuals is significantly higher than in HIV negative male individuals. Although no significant differences in proportion of any single MDRGN species were observed, this indicates that sexual transmission of MDRGN within the male HIV positive population is possible. With regard to the results of this study, practicing safer sex should be propagated actively further on to reduce not only the spread of many STD-associated pathogens but most likely also MDRGN.