Adipose tissue is a predictor of 30-days mortality in patients with bloodstream infection caused by carbapenem-resistant Klebsiella pneumoniae

This study was a single-center retrospective cohort research, approved by the ethics committee of the First Affiliated Hospital of Wenzhou Medical University, one of the largest health care centers in the southern province of Zhejiang, China. Patients, both suffered from CRKP BSI supported by clinical symptoms and etiological evidence during their hospitalization and performed abdominopelvic computed tomography imaging from 2016 January 1st to 2020 June 1st, were eligible for the study. CRKP BSI were nosocomial infections and was first detected during hospitalization in our hospital. Meanwhile, Abdominopelvic CT examinations were performed in these infectious patients due to abdominal trauma, enteric or urinary tract infection, severe sepsis, multiple organ dysfunction syndromes, etc. Patients who did not detect CRKP BSI for the first time in our hospital, less than 16 years old, or refused further treatment were excluded. CRKP was defined as minimum inhibitory concentration (MIC) ≥ 4 mg/L to both imipenem and meropenem according to Clinical and Laboratory Standards Institute guidelines [12].

Patients' baseline data, including demographic characteristics, presence of comorbid conditions, acute complications (shock, acute respiratory failure, acute renal failure), Sequential Organ Failure Assessment (SOFA) Scores, exposure to invasive intervention surgery, renal replacement therapy, infection site, combined viral and fungal infection, Pitt bacteremia scores, and details on therapy were reviewed and obtained from electronic medical records. The observational onset of our research was defined as the date of microbiology specimen collection in the first cultivated CRKP. The date: SOFA scores, body temperature, and inflammation indicators of blood were recorded within 48 h before positive blood culture with CRKP infection. Antibiotic treatment, exposure to tracheostomy and mechanical ventilation, renal replacement therapy, and acute complications (respiratory failure, kidney failure, and shock) maintained for more than 48 h were considered for analysis. Surgery was defined as invasive procedures which could be found in the surgical record in the electronic medical records within 1 month before and after the CRKP BSI. Cultivated fungal, detection of 1,3-β-d-glucan or galactomannan antigen-positive, combined with the patient’s symptoms, signs, and antifungal drug treatment was referred to as fungal infection. Equally, the related viral nucleic acid tests and the clinical conditions of infectious patients were treated as viral infections. The average Chinese weight, 66 kg for males and 57 kg for females, was an alternative to some of these seriously infected patients in case of lacking weight records. Based on the recommended medication and antimicrobial susceptibility test of Klebsiella pneumoniae, antibiotic treatment options were divided into polymyxin B-based (PMB-based) or tigecycline-based (TGC-based) combination therapy or others (aminoglycosides, Fosfomycin, carbapenem, etc.) [13]. Early appropriate antibiotic therapy was the regard of administered 48 h or less by the prescribing physicians after the first culture of CRKP and included at least two in-vitro active drugs [14‐16].

Area-based quantifications of adipose tissue and muscle mass compartments were performed on the L3/4 spinal level using a semiautomatic software tool (Syngo Volume tool, Siemens Healthcare, Munich, Berlin, Germany) [17]. Two experienced radiologists drew the eligible CT planar. Records of multidetector-CT scans with quantification of body composition were retrieved 7 days before or after positive blood culture with CRKP. If the patient has undergone multiple abdominal CT examinations, the time point closest to the starting point of our experiment was chosen. The manually determined specific region of interest (ROI) includes VAT (the fascial plane of the abdominal muscle wall, using standard Hounsfield Unit (HU) range − 190 to − 30, Fig. 1) area expressed in mm², TAT (whole abdominal circumference, using HU range − 190 to − 30), SM (sum of M. psoas major, M. erector spinae, M. quadratus lumborum, M. latissimus dorsi, M. transversus abdominis, M. obliquus internus abdominis and externus abdominis, and M. rectus abdominis, measured by limiting the attenuation threshold between 40 and 100 HU, Fig. 1) area expressed in mm², the subcutaneous adipose tissue (SAT) area was defined as the subtraction between total adipose tissue (TAT) area and VAT area [18]. All CT examinations were performed using scanners: Brilliance-64, Philips Medical Systems, Eindhoven, The Netherlands; 128-MDCT scanner Somatom Definition, Siemens Health-care Sector, Forchheim, Germany.

R software, GraphPad Prism, and Stats were conducted for statistical analyses. The 14-day and 30-day survival nomograms were constructed based on the prognostic factors derived from multivariate Cox regression analysis to predict 14-day and 30-day survival possibilities. Continuous variables were exhibited for means, medians, range, and standard deviation (SD) and compared using an independent t-test or Wilcoxon test; Spearman's correlation coefficient was used for variable correlation; Chi-square test was used to analyze categorical variables; log-rank survival analysis was employed to determine the effect of various variables on patients’ 30-day survival. The optimal cut-off values for VAT and TAT were calculated by the X-tile program (constructed by Yale University, New Haven, CT, USA). The cut-off of high TAT and VAT were 36.95 mm² and 18.96 mm², respectively. All statistical tests were two-sided and P < 0.05 was considered statistically significant.

A total of 90 eligible patients, both suffered from CRKP bloodstream infection and had abdominopelvic computed tomography within 7 days were recruited from the First Affiliated Hospital of Wenzhou Medical University, a 4100-bed general teaching hospital; of these, 2 were excluded from the analysis due to incomplete information, 6 were excluded owing to cultivated CRKP before entering our hospital, 3 were excluded because of only 1 day of hospitalization, 7 were excluded for failed abdominopelvic CT image. Abdominopelvic CT examination was routinely implemented for these severe infectious patients with some complex or critical complications according to doctors' experience. As of discharge time, 27 patients died during 30 days follow-up, and the average survival time was 24.14 days. Additional baseline clinicopathological parameters were presented in Table 1.

As for the association of adipose fat and muscle tissue with 30-day mortality, we observed that TAT and VAT were dependently related to 30-day mortality while SM had no differences (Table 1). Furthermore, high-TAT was closely associated with worse clinical outcomes, after adjusting for comorbid conditions and other differences in baseline characteristics (Table 2). Compared with low-VAT, patients with high-VAT showed a strikingly poor prognosis in both 30-day all-cause mortality (P = 0.0108, Fig. 2A) and 30-day CRKP BSI mortality (P = 0.0049, Fig. 2C). The results of TAT were similar to VAT (Fig. 2B, D).

In the multivariable analysis (Table 2), factors associated with 30-day all-in hospital mortality included total adipose tissue (TAT) [adjusted hazard ratio (HR) = 1.028, 95% confidence interval (CI), 1.003–1.053; P = 0.025], age [HR = 1.030, 95% CI, 1.000–1.061; P = 0.047] and SOFA scores [HR = 1.138, 95% CI 1.049–1.263; P = 0.002].

A nomogram was constructed to predict 14- and 30-day survival of patients with CRKP bloodstream infectious (Fig. 3). Total scores were summations of each variable based on the intersection of the vertical lines. As shown in Fig. 3, VAT, TAT, age, and SOFA scores contributed the most risk points (ranged 0–100), whereas the other clinical information contributed much less. By using a nomogram, we could precisely convert each patient's clinical index to the corresponding point, and then evaluate the likelihood of survival. HR values for therapy progression derived from Cox models suggested that patients with lower TAT are more likely to survive from CRKP bloodstream infection. The combination of body composition, age, and SOFA showed a good ability to predict survival.