2016 年 39 巻 7 号 p. 1195-1200
It is well known that methicillin-resistant Staphylococcus aureus (MRSA) produces many virulence factors, such as hemolysins, leukocidins, proteases, enterotoxins, exfoliative toxins, and immune-modulatory factors. The aim of study was to identify staphylococcal pathogenicity that may affect the prognosis of patients with MRSA bacteremia. We obtained 149 MRSA strains from blood cultures between January 2009 and December 2014 in our institution. We collected information on patient characteristics, laboratory data, staphylococcal toxin genes, and susceptibility of the strain toward anti-MRSA agent and analyzed them as factors contributing to 30-d mortality. The “survival” and “dead” groups consisted of 103 and 46 patients, respectively. Multiple logistic regression analysis showed a four-fold increase in the risk of mortality in patients exhibiting isolated MRSA with staphylococcal enterotoxins (SEs) genes as well as toxic shock syndrome toxin-1 (TSST-1) genes [odds ratio: 3.89; 95% confidence interval: 1.20–12.60; p=0.024]. Kaplan–Meier analysis also showed significantly higher mortality in patient with isolated MRSA with SEs and TSST-1 genes. After adjusting for confounders, the coexistence of SEs and TSST-1 were independently associated with the 30-d mortality compared with treatment and susceptibility. The coexistence of superantigenic toxin genes greatly affects the clinical course and prognosis of patients with MRSA bacteremia.
Vancomycin (VCM) has been the standard treatment for MRSA infections for over 50 years. The area under the concentration–time curve [AUC/minimum inhibitory concentration (MIC) ratio] is a representative parameter of the pharmacokinetics–pharmacodynamics (PK–PD) of VCM.3) The Infectious Diseases Society of America recommends a target AUC/MIC>400 for the treatment of severe MRSA infections.4) However, this is often difficult to achieve for MRSA strains with high VCM MICs, thus increasing the risk of insufficient therapeutic response. Previous meta-analyses have attempted to address the association between MIC and clinical outcomes.5–7) In contrast to these studies, a recent meta-analysis examining 8291 episodes reported no differences in the risk of death between patients with S. aureus high and low VCM MICs.8) These results highlight the current lack of consensus regarding the relationship between VCM MICs and the prognosis of patients with MRSA infections.
Our previous study measured MICs at 0.25 µg/mL intervals using the broth microdilution method and found that VCM MICs≥1.5 µg/mL affected the prognosis for MRSA bacteremia.9) However, there were no significant differences in VCM AUC/MIC and clinical prognosis between the groups.
The pathological mechanism of S. aureus infections includes infectious pathogenicity such as pneumonia, cellulitis, osteomyelitis, and infective endocarditis as well as toxic pathogenicity such as food poisoning and toxic shock syndrome.10,11) Several virulence factors have been reported to be associated with the clinical outcomes of MRSA infections. The expression of major virulence factors, such as Panton–Valentine leukocidin (PVL), is known to be correlated with higher mortality in community-associated MRSA infections.12,13) We believe that toxic pathogenicity is closely associated with patient prognosis. S. aureus is known to produce various virulence factors,14) such as toxic shock syndrome toxin-1 (TSST-1) and staphylococcal enterotoxins (SEs), which are superantigenic toxins (SAgT) associated with severe infections.11,15,16) A previous study reported that TSST-1 and SEC1 caused shock when intravenously injected into an animal model.17) In addition, recent experimental study showed that SAgT play a critical role in the development and progression of S. aureus infective endocarditis and sepsis.18) However, the association between SAgT and the prognosis of patients with MRSA bacteremia has not been sufficiently studied. Therefore, here we aimed to examine the effects of staphylococcal virulence factors on the prognosis of patients with MRSA bacteremia.
This single-center, retrospective, observational study was performed at the Showa University Hospital, which is a 1014-bed teaching hospital in Tokyo, Japan. Clinical MRSA strains (n=149) from blood cultures were collected from inpatients older than 15 years of age at our hospital between January 2009 and December 2014. They were divided into two groups: “survival” and “dead” (n=103 and 46, respectively). Blood culture was performed using the BD BACTEC FX System (Becton, Dickinson and Company, Franklin Lakes, NJ, U.S.A.). All strains were screened using the MicroScan® WalkAway® system and those identified as MRSA were subjected to oxacillin or cefoxicin testing. Patients with a fever of ≥38°C who tested positive for at least one strain per blood culture bottle were included in this study. In patients who tested positive more than one strain, only the first one was tested. We collected information regarding the age, gender, underlying diseases, Charlson comorbidity index (CCI),19) systemic inflammatory response syndrome criteria and sequential organ failure assessment (SOFA) score,20) laboratory data, colonization of MRSA, duration of central venous catheter, source of infection (soft tissue and bone, central nervous system, respiratory tract, intra-abdominal, urinary tract, and intravascular or medical devices), and administration of anti-MRSA agents and analyzed their contribution to the 30-d mortality.
We performed antimicrobial susceptibility test on all MRSA strains using the Sensititre® microbroth dilution system (Trek Diagnostic Systems Inc., Cleveland, OH, U.S.A.). Susceptibility toward anti-MRSA agents were tested using original MIC plates (Trek Diagnostic Systems Inc.) with the following antibiotics: VCM, teicoplanin (TEIC), arbekacin, daptomycin, linezolid (LZD), and tigecycline (TIGE) as previously described.9) Resistance break points published by the Clinical and Laboratory Standards Institute were used.21)
We performed a molecular analysis to evaluate the effect of the following staphylococcal virulence factors on patient prognosis: TSST-1 (tst), SE (sea, seb, and sec), leukotoxin [lukSF-PV (PVL) and lukED], hemolysin (hla, hlb, hld, hlg, and hlg-2), exfoliative toxin (eta and etb). The presence of these virulence factors was determined using PCR methods, as previously described.22) Staphylococcal cassette chromosome mec (SCCmec) types were also identified using previously described methods.23)
Continuous variables have been presented as the mean and standard deviation (S.D.) or medians and interquartile ranges and were compared between groups using the independent t-test or Mann–Whitney U tests. Categorical variables have been presented as the number and percentage of patients within each group and compared using the χ2 or Fisher’s exact test. Survival was analyzed using the Kaplan–Meier method. Plots were compared using the log-rank test. We used forward stepwise logistic regression to identify independent variables associated with the 30-d mortality and included those variables with p<0.05 in the univariate analysis.
All statistical tests were two-sided; a p value of <0.05 was considered statistically significant, and all analyses were performed using SPSS statistical software (version 23; IBM Japan Ltd.).
This study was approved by the research ethics committee of Showa University (Approval No. 1476). Informed consent was exempted according to the Ethical Guidelines for Medical and Health Research Involving Human Subjects by the Ministry of Education, Culture, Sports, Science and Technology and the Ministry of Health, Labour and Welfare of Japan.
We analyzed potential factors that affected the 30-d mortality in patients with MRSA bacteremia (Table 1). The analysis revealed significantly higher CCI (p=0.002), SOFA score (p=0.002), insertion of central venous catheter (p=0.033), and lower administrated of VCM (p=0.019) in the “dead” group than in the “survival” group.
SIRS, systemic inflammatory response syndrome; SOFA, sequential organ failure assessment; MRSA, methicillin-resistant Staphylococcus aureus; VCM, vancomycin; TEIC, teicoplanin; ABK, arbekacin; LZD, linezolid; DAP, daptomycin. a) Values represent the mean±S.D. or median (IQR), unless otherwise indicated. b) Data from “survival” (N=79) and “dead” (N=43) groups.
We explored the virulence factor profiles of MRSA genotypes according to the 30-d mortality (Table 2); no differences were found between MRSA strains genotypes. Exfoliative toxins (eta and etb) and PVL genes were absent, whereas hemolysins (hla, hlb, hld, hlg, and hlg-2) and lukE-lukD were present. Because differences were observed in the presence of SEs (sea, seb, and sec) and tst, we explored the combination of these SAgT genes.
Figure 1 shows the combination of SAgT genes (tst, sea, seb, and sec) according to the 30-d mortality. Significantly higher mortality was observed in patients infected with coexistent strains of tst, sea, seb, and sec (TABC) (10.7 versus 23.9%; p=0.035).
Significantly higher mortality observed in patients infected with coexistent strains of tst, sea, seb, and sec (TABC). A, sea; B, seb; C, sec; T, tst.
The microbiological characteristics and antibiotic susceptibilities of MRSA strains according to the 30-d mortality are shown in Table 3. The analysis revealed significantly higher VCM MICs and the presence of TABC strains in the “dead” group. There were no significant differences in SCCmec types and 30-d mortality between the two groups.
MIC: minimum inhibitory concentration; MRSA: methicillin-resistant Staphylococcus aureus; VCM: vancomycin; TEIC: teicoplanin; ABK: arbekacin; LZD: linezolid; DAP: daptomycin; TIGE: tigecycline; SCCmec: staphylococcal cassette chromosome mec. a) Values represent the median (IQR) unless otherwise indicated.
Multiple logistic regression analyses were performed to identify independent risk factors for 30-d mortality (Table 4). The following factors were found to be significantly associated in the univariate analysis: CCI, SOFA score, insertion of central venous catheter, first administration of VCM, VCM MICs, and strains with TABC (Tables 1–3). In the multivariate analysis controlling for several confounders, strains with TABC were independently associated with higher mortality [odds ratio (OR): 3.89; 95% confidence interval (CI): 1.20–12.6; p=0.024].
SOFA: sequential organ failure assessment; MRSA: methicillin-resistant Staphylococcus aureus; TABC: coexistence of tst, sea, seb, and sec. Hosmer and Lemeshow test=0.303, discrimination rate=71.3%. Variables identified as significant (p<0.05) were entered into multiple logistic regression models.
Kaplan–Meier analysis showed significantly higher 30-d mortality in a subset of patients with MRSA exhibiting the presence of SAgT (Fig. 2). The clinical status of patients with MRSA bacteremia and TABC strain is shown in Table 5. The analysis revealed significantly higher catecholamine administration (p=0.039), SOFA score (p=0.012), and lower platelet count (p=0.010) in the MRSA bacteremia with TABC strain. No significant differences in the CCI were observed between the groups.
TABC, coexistence of tst, sea, seb, and sec.
MRSA: methicillin-resistant Staphylococcus aureus; TABC: coexistence of tst, sea, seb, and sec; SOFA: sequential organ failure assessment. a) Values represent the mean±S.D. or median (IQR) unless otherwise indicated.
The susceptibility of the TABC strain toward the anti-MRSA agent is shown in Table 6. The TABC strain had significantly higher VCM and TEIC MICs (p=0.008 and 0.001, respectively), and the common genotypes were SCCmec type II (95.5%).
MRSA: methicillin-resistant Staphylococcus aureus; MIC: minimum inhibitory concentration; TABC: coexistence of tst, sea, seb, and sec; VCM: vancomycin; TEIC: teicoplanin; ABK: arbekacin; LZD: linezolid; DAP: daptomycin; TIGE: tigecycline.
This retrospective observational study is the first to examine the contribution of MRSA bacteremia combined with SAgT toward patient prognosis.
Previously, several studies have evaluated factors influencing the prognosis of patients with S. aureus bacteremia.24) Host factors, pathogen factors, clinical management, and microbiological factors, such as methicillin resistance and VCM MICs, were found to be associated with mortality.5–7,9,24) However, a recent meta-analysis reported no differences in the risk of death in patients with S. aureus high and low VCM MICs.8) Our previous study showed that there was no significant difference between the survival and dead group in the VCM AUC/MIC.9) Although the reason for this inconsistent finding remains unclear, our data suggest that strain-specific pathogenicity, rather than vancomycin susceptibility, contributes to the outcome of MRSA bacteremia. Furthermore, the present study showed higher VCM and TEIC MICs in strains with TABC, suggesting that virulence factors and susceptibilities contribute to differences in strain-dependent mortality. The association between strain-specific virulence and clinical outcomes of MRSA bacteremia has also been previously reported.25–27) These factors may help explain the contradictory results of the meta-analysis.
This study shows that the combination of SAgT genes is associated with the clinical status and prognosis for MRSA bacteremia. The superantigens bind to class II major histocompatibility complex on the surfaces of antigen-presenting cells and T-cell receptors of cytotoxic lymphocytes. As a result, large amounts of cytokines are released, causing systemic symptoms such as shock, pancytopenia, and multi-organ failure.16) Our results demonstrated that significantly lower platelet count, higher catecholamine administration, and higher SOFA score were observed in patients infected with TABC, suggesting severe sepsis or septic shock in patients with MRSA bacteremia. Thus, SAgT induced systemic symptoms that potentially caused poor patient prognosis. In a previous study, Japanese women in Tokyo exhibited a lower prevalence of TSST-1 antibody titers than did Caucasians in the United States.28) We believe that this finding was associated with systemic symptoms caused by the overproduction of SAgT in Japanese. Human-origin intravenous immunoglobulin has a neutralizing effect on TSST-1 and SE produced by MRSA.29) It may be useful for passive immunization therapy in patients with MRSA bacteremia. Protein synthesis inhibitors may also be useful for the treatment of severe staphylococcal infections associated with the production of SAgT. Previous studies have shown that the level of TSST-1 produced decreases in cells treated with the protein inhibitors clindamycin, LZD, gentamicin, and TIGE.30,31) In the clinical setting, these protein inhibitors attenuate SAgT production in life-threatening cases of toxic shock. However, these are adjuvant therapeutic measures that require further clinical research.
Our study had a number of limitations. First, this was a small, single-center study conducted in Tokyo, Japan, and our findings cannot be generalized to other countries with different circulating MRSA strains. Moreover, it was a retrospective study with a limited sample size; therefore, potential and unmeasured confounding factors were not controlled for. Because there was potentially good prognosis in patients treated with primary VCM administration, the lower administration of VCM was observed in “dead” group. The “dead” group includes patients with supportive and palliative care (e.g. end-stage cancer). These potential confounding factors were uncontrolled in this study. Second, we were unable to analyze several staphylococcus virulence factors, particularly various SE genotypes. Therefore, it is possible that the differences in mortality are a result of other virulence factors that we did not examine. A previous study reported that SEA play a key role in sea-positive S. aureus sepsis by triggering overexpression of inflammatory mediators associated with shock.32) However, the role of each virulence factor in the prognosis of patients with MRSA bacteremia is still unclear. A larger study is required to examine combinations of various staphylococcal virulence factors. Finally, the PCR-based evaluation of virulence determinants should be interpreted with some caution as it may not directly reflect in vivo expression. Although our study has several limitations, the results allow us to conclude that combination of SAgT genes is potentially associated with higher risk of mortality associated with MRSA bacteremia.
In conclusion, after adjustment for several confounders, combinations of SAgT, such as TSST-1 and SEs, were independently associated with the 30-d mortality in MRSA bacteremia. Our results suggest that strain-specific virulence factors, rather than treatment and susceptibility, contributes to the outcome of MRSA bacteremia. This reinforces the importance of adjuvant therapy, such as clindamycin, LZD, gentamicin, TIGE, and immunoglobulin, in severe MRSA infections. Further genome-wide studies and clinical trials should be performed to improve the outcome of MRSA bacteremia.
We thank our students for providing technical support. The authors would like to thank Enago (www.enago.jp) for the English language review. This study was presented at the Interscience Conference of Antimicrobial Agents and Chemotherapy (ICAAC) 2015, San Diego CA (abstract number B-522).
Shoji H, Takuma T, and Niki Y, received donations for this research from Takeda Pharmaceutical Co., Ltd., Pfizer Japan Inc., Daiichi Sankyo Co., Ltd., Bayer Yakuhin, Ltd., Taisho Toyama Pharmaceutical Co., Ltd., Dainippon Sumitomo Pharma Co., Ltd., Mitsubishi Tanabe Pharma Corporation, Kyorin Pharmaceutical Co., Ltd., Astellas Pharma Inc., Chugai Pharmaceutical Co., Ltd., and GlaxoSmithKline K.K. Shoji H, Takuma T, and Niki Y are endowed chairs and funded by Shionogi & Co., Ltd., Meiji Seika Pharma Co., Ltd., and Toyama Chemical Co., Ltd., respectively.