Discussion
cKP and hvKP
Klebsiella pneumoniae , a gram-negative bacillus, has been implicated as a cause of pneumonia, urinary tract infection, abdominal cavity infection, and intravascular device infection; this type ofKlebsiella pneumoniae is recognized as classical Klebsiella pneumoniae (cKP) [1]. However, different types of Klebsiella pneumoniae have been recognized, starting in East Asia. The first report of a unique Klebsiella pneumoniae was reported in Taiwan in the 1980s, with characteristic of community-acquired infection with complications of pyogenic liver abscess and endophthalmitis [3]. Similar cases were reported afterward worldwide. This new and unusual type of Klebsiella pneumoniae has been studied and named hvKP based on its malignancy. The features of hvKP have been described as community-acquired infection among healthy young individuals, which tends to be associated with endophthalmitis, multiple abscesses, and especially liver and septic arthritis. It has been reported that 90.9% of the pathogens causing liver abscesses with Klebsiella pneumoniae infection are hvKP [4]. The string test, which uses an inoculation loop to generate a viscous string from a bacterial colony, is used for differential diagnosis between cKP and hvKP. If a string length of > 5 mm is generated because of hypermucosviscosity, the test is considered positive [5]. However, the specificity of the string test and hvKP varies between 51% and 98%, and even cKP shows positive string test rates of 17% to 23% [6]. Although one retrospective study reported no clinical relevance of a 30-day mortality rate between cKP and hvKP [7], a review article demonstrated that hvKP is associated with a significant mortality rate of 3 to 55% compared to cKP [1]. Iron acquisition has been described as the decisive factor of hvKP’s hypervirulence compared to cKP [1].
Our patient had hvKP infection with a positive string test, and this infection induced multiple abscesses, such as liver abscess, retropharyngeal abscess, and psoas major abscess. He had septic shock because of severe hvKP infection. Therefore, endotoxin adsorption therapy with a longer PMX-DHP duration was performed 2 times to treat his septic shock. PMX-DHP has been shown to adsorb not only endotoxin but also monocytes and anandamide, resulting in a reduction in blood levels of inflammatory cytokines such as interleukin-6, tumor necrosis factor-alpha and interleukin-17A [8]. In addition, we have shown that a longer duration of PMX-DHP improves hemodynamics and decreases vasopressors rapidly compared to the conventional duration, and a ≥ 8-hour duration of PMX-DHP was associated with an improvement in pulmonary oxygenation [9]. As a result, our patient’s hemodynamics and pulmonary oxygenation improved after PMX-DHP treatment, and he recovered from septic shock. It is important for clinicians to be aware of hvKP, and whole-body examination must be conducted, with or without the positive string test, when suspected.
Infectious rhabdomyolysis due to hvKP
At first, our patient presented complained of weakness in the lower limbs, slurred speech, and lower back pain. He was diagnosed with rhabdomyolysis because of a high serum CK level (37370 U/L) and high myoglobin level (11850 ng/mL). Rhabdomyolysis is a fatal clinical condition caused by the release of toxic intracellular content from damaged skeletal muscles. Although there are no diagnostic criteria for rhabdomyolysis, a serum CK level higher than 10 times the normal value in the absence of heart or brain diseases is used for clinical indication [10]. Despite multiple of rhabdomyolysis, rhabdomyolysis secondary to infection is uncommon. Furthermore, infectious rhabdomyolysis is reported to have a poorer prognosis (such as incidence of acute kidney failure) than rhabdomyolysis secondary to exercise events, though peak CK and myoglobin levels are lower in the former [11]. The most commonly reported organisms causing infectious rhabdomyolysis are Legionella spp., Streptococcus spp.,Salmonella spp., and influenza virus, and the main site of primary infection is the respiratory tract [12,13]. There are several hypotheses regarding the mechanism of infectious rhabdomyolysis, but none of them are well established: direct muscle invasion of bacteria or viruses, immunologic reactions such as ‘cytokine storms’, toxic effects from bacteria or viruses, and muscle damage caused by high fever and hypoxia [13]. Significant mortality of 38% in infectious rhabdomyolysis has been reported [14].
Because our patient did not have any initial imaging findings except for the hypodense lesion in the liver, we suspect that the primary source of infection was a liver abscess. To the best of our knowledge, there are only two English reports describing cases of rhabdomyolysis secondary to liver abscess [4,15], but there are no reported rhabdomyolysis cases due to hvKP infection. Our case was rhabdomyolysis secondary to hvKP infection, making liver abscess the primary focus. Overall, this was an extremely rare case with a good course despite the high mortality rate.