We present the case of a patient with cirrhosis who had an extremely accelerated disease time course, as death occurred one and a half days after high fever was detected. This shows that CRBSIs can become fatal for patients with cirrhosis rapidly. Additionally, in the present case, liver cirrhosis complicated and delayed the detection of the initial sign of infection as well as the intervention. As CRBSIs accompanied by other organ complications might be fatal, preventative measures against CRBSI should be applied rigorously.
CRBSI in patients with stroke, CNS injury-induced immunodepression syndrome (CIDS), and cirrhosis-associated immune dysfunction (CAID)
CRBSI is common and can be fatal, especially in patients with accompanying organ failure [2]. Currently, it is not clear whether stroke patients are more susceptible to CRBSI. Some reports have described a few CRBSI cases in the stroke care unit (SCU) [5, 6]. However, those data may not be accurate, and there could have been many false negative cases because blood cultures are not performed for all febrile patients in the SCU, even though they may have a strong risk of developing CRBSI. In contrast, patients with severe stroke receive several medications, such as parental nutrition for massive diarrhea or complicated constipation; antibiotics for aspiration pneumonia, urinary tract infection (UTI) or meningitis; hemodynamic volume therapy for vasospasm following SAH; or antiplatelet therapy for cerebral infarction and vasospasm. Therefore, the use of central line catheters is necessary for these patients, and these should be frequently changed in the SCU and intensive care unit, in order to prevent infection.
Stroke and other CNS injury are known to create a systemic immunosuppression, rendering patients more prone to infections. CIDS is characterized by lymphopenia and immune dysfunctions; in particular, it is characterized by functional deactivation of T-helper (Th) 1 and natural killer (NK) cells, and monocytes/macrophages. Catecholamines have been shown to suppress Th1 activity, promote Th2 cell differentiation, and trigger interleukin (IL)-10 secretion. A correlation between elevated levels of metanephrine and the incidence of post-stroke infections has been demonstrated. Experimental data indicated that an impaired early catecholamine-mediated lymphocyte response, particularly a reduced interferon gamma (IFNγ) secretion by T and NK cells, could be an essential cause of the impaired antibacterial defense after brain ischemia [7]. As a form of post-stroke immunodepression, CIDS could have contributed to the CRBSI observed in the present case.
CAID alters both innate and acquired immunity [8] and results in immunodeficiency as well as systemic inflammation [9]. Endotoxemia could enhance the inflammatory host response and expansive systemic bacterial infection by CAID, resulting in septic shock, multi-organ dysfunction, and death [10,11,12]. It has also been reported that cirrhosis increases levels of IL-6, tumor necrosis factor alpha (TNF-α), endotoxin, and nitric oxide that induce vasodilatation [13] and reduce C-protein as well as coagulation factor levels [12].
CAID status and infection marker sensitivity
With a mortality rate of approximately 30% [14], CAID patients have a higher mortality risk than do patients with normal immunity. In the present case, the patient seemed to have CAID that might have contributed to the fatality. Thus, her immunity must have been severely weakened. A late intervention, such as an infection source control or a vital-supported therapy, could be fatal for immune-compromised patients.
Generally, CRP and WBC indicate the inflammation induced by an infection. Thus, CRP level monitoring might be beneficial in detecting the infection, especially in cirrhosis patients [15]. However, Silvestre et al. showed that CRP should not be used as marker of infection in septic patients with severe hepatic failure, as the liver cannot produce enough CRP in response to the inflammation [16]. In cirrhosis patients, the blood overflow into spleen destroys blood cells, resulting in pancytopenia [17]. Therefore, in cirrhosis patients with liver failure, CRP and WBC might not reflect the severity of the inflammation, even in a state of severe sepsis. In fact, the WBC count, hemoglobin levels, and platelet count observed after the initial blood test were relatively low (4,300 cells/μL, 9.3 g/dL, and 42,000 cells/μL, respectively) in our patient. We believed that the CNS damage had caused the fever and that the infection was not severe. The timing of antibiotic treatment was delayed by 24 h after initial onset of symptoms. These data suggest that pseudo-values of inflammation could precipitate patient death. Recently, it has been reported that the thyroid marker procalcitonin (PCT) is a trustful biomarker for bacterial infection, with a sensitivity and specificity of 81.5% and 87.3%, respectively. This biomarker could detect bacterial infection in cirrhosis patients after a relatively short period [18]. Although they were not examined in our case, PCT levels should be examined in patients with cirrhosis who develop high fever.
CRBSI prevention and case consideration
Prevention measures such as hand hygiene and the use of fully sterile barriers should be implemented during catheter insertion; application of chlorhexidine gluconate for skin antisepsis, maintenance of the sterile technique, and removal of catheters when they are not necessary are effective in preventing infections. In our case, the blood and catheter cultures revealed that Klebsiella pneumoniae, a gram-negative rod bacterium that is part of the enteric flora and produces endotoxins, caused a severe infection. The central line had been inserted through the left femoral vein in a sterile fashion and was placed for 2 weeks (Fig. 3). Stools could have possibly contaminated the catheter and thus enabled the inhabiting Klebsiella pneumoniae to infect the blood stream. This situation could have been managed using two strategies: an early empirical and over-indicative administration of antibiotics or a strict application of CRBSI preventive measures. The single most important prognostic factor in sepsis is timing of antibiotic administration. In our case, there was a delay of nearly 24 h before antibiotic administration was started. Although the differential diagnosis was challenging and the laboratory results were seemingly normal, infection should always be considered a reason for fever, especially when further signs of systemic inflammatory response syndrome are observed (in this case, hemodynamic compromise and disseminated intravascular coagulation). Moreover, preventive guidelines for CRBSI recommend that CVC lines should be inserted through subclavian or jugular veins; this is not currently obligatory in general practice [19, 20]. We routinely try not to place the catheter at the femoral site. In this case, we had initially inserted the catheter through the right neck and then reinserted it through the left femoral vein because the left neck was infected and not hygienic. Currently, we do not understand whether removal and replacement of CVCs through the left femoral vein is useful in avoiding CRBSI, though it is not recommended to routinely switch the position of or to replace CVCs [19, 21]. Recent randomized control studies also showed no differences in outcome between early CVC removal and watchful waiting [22]. When CRBSI was suspected, fever was mildly high (37.5 °C) (Fig. 2a) and other infections, such as pneumonia or UTI, were also suspected. In this case, watchful waiting instead of a daily CVC removal could be another option. We cannot conclude that long-term CVC insertion causes CRBSI and death, and whether early replacement or watchful waiting is better to prevent CRBSI.
Additionally, the blood endotoxin level was notably high (428.8 pg/mL) in our case, suggesting a probable endotoxin shock. The presence of endotoxemia in patients with septic shock and positive blood culture are related to higher mortality (39%) than that in patients without endotoxemia (7%) [23]. This implies that endotoxemia with sepsis might have worsened the patient’s condition and indicated a poor prognosis in our case. Further investigations are needed.
Strategy for CRBSI
Patients with severe stroke have a high probability of developing fatal infections, including CRBSI, which is associated with a mortality rate of approximately 10% [24, 25]. However, a tremendous reduction in CRBSI rates by the application of preventative measures has recently been reported in several studies [26,27,28]. Theoretically and epidemiologically, it is possible and important to prevent CRBSIs, as emphasized by guidelines [19, 20]. However, it is highly challenging to cure CRBSIs once the infection occurs, especially in patients with accompanying organs failure, as in our case. Thus, these patients have a higher risk of death once they develop severe sepsis, including CRBSI.
Conclusion
CRBSI patients with liver cirrhosis are very challenging to treat. Clinicians should be aware of the possibility of CRBSI in patients with severe stroke and pay more attention to preventive measures, such as avoiding the insertion of CVCs through femoral veins, particularly in patients with accompanying organ failure, as in our case.