Histidine-rich Glycoprotein as a Novel Predictive Biomarker of Postoperative Complications in Intensive Care Unit Patients

A Prospective Observational Study

Masahiko Oiwa; Kosuke Kuroda; Naoya Kawanoue; Hiroshi Morimatsu

Disclosures

BMC Anesthesiol. 2022;22(232) 

In This Article

Methods

Study Design and Ethical Considerations

This single-center, prospective, observational study was approved by the Institutional Review Board of the Okayama University Hospital (Okayama, Japan) on August 14, 2020 (approval number: 2007–006). The need for registration of the study was waived because this was an observational investigation. The requirement for written informed consent was waived by the Institutional Review Board because this was a non-invasive study using residual blood samples collected from routine blood tests performed on POD 1. We described the study protocol to the all patients and obtained verbal informed consent for study participation and publication were obtained from them. This information was preserved as an electronic medical record before their inclusion in the study. The patients received a copy of the study description and were provided with contact information, in case additional questions or concerns arose. In addition, the study protocol was published on the website. We followed the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.[24]

Patients and Data Collection

Patients admitted to the intensive care unit (ICU) after surgery at Okayama University Hospital (Okayama, Japan) during consecutive periods were prospectively included. At our institution, all patients post respiratory surgery, neurosurgery, hepato–biliary–pancreatic surgery, esophageal surgery, cardiovascular surgery, and highly invasive oral and otolaryngological surgery are admitted to the intensive care unit. In other departments, patients are admitted to the intensive care unit post-surgery at the discretion of the physician. According to previous studies, HRG levels are higher in adults than in children[15] and decrease during pregnancy;[22] thus, patients who were pregnant or < 20 years old were excluded. We planned to enroll 150 patients based on a power calculation. According to our previous study,[16] we expected that the HRG levels would vary by 20 μg/mL between patients with and without postoperative complications; this calculation was based on the number of patients required for an 80% power to detect a 20 μg/mL difference in HRG levels. A two-sided type I error of 0.05 was considered for the 10% incidence of postoperative complications and loss to follow up.

All enrolled patients' information was collected from electronic medical records. Preoperative comorbid cardiovascular diseases included arrhythmia, coronary artery disease, heart failure, and macrovascular diseases. Chronic kidney disease was classified with an estimated glomerular filtration rate < 50 mL/min. The surgical Apgar score (SAS) was calculated using anesthesia records. Preoperative and postoperative sequential organ failure assessment scores and acute physiology and chronic evaluation II scores on admission to the ICU were calculated using clinical variables and blood-test results.

Postoperative complications were defined as an extended Clavien–Dindo classification[25] grade II or higher, occurring within 7 days after surgery. Among the postoperative complications, we defined infectious complications as those that required antibiotic therapy or drainage due to infection. The mortality rate was assessed 28 days postoperatively. The enrolled patients were followed up to the day of discharge or 28 days postoperatively.

Measurement Methods

To measure HRG levels, we used the residual blood samples collected for routine blood tests in tubes containing K2-EDTA in the morning of POD 1. The samples were then centrifuged at 3,000 rpm for 10 min. Plasma components were transferred to polypropylene tubes with a pipette, and a protease inhibitor cocktail (Complete mini EDTA-free; Roche Diagnostics, Basel, Switzerland) was added. The samples were stored at -80 °C.

Plasma HRG levels were measured using a modified quantitative sandwich enzyme-linked immunosorbent assay, in which the detection and chromogenic reagents were changed from those previously described[16] because of discontinuation of the reagent. In brief, a rat monoclonal antibody (mAb) against human HRG (made in-house, number 75–14) was used as the capture antibody, and a nickel (Ni 2+)-activated derivative of horseradish peroxidase (HisProbe™-HRP Conjugate; Thermo Fisher Scientific, Waltham, MA, USA) was used for detection. Plasma samples were diluted 200-fold and 400-fold in phosphate-buffered saline containing 1% bovine serum albumin and 0.1% K2-EDTA and pipetted into mAb-coated 96-well plates (Clear Flat-Bottom Immuno Nonsterile 96-Well Plates, Thermo Fisher Scientific). A microplate washer (Immuno Wash™ 1575 Microplate Washer; Bio-Rad Laboratories, Hercules, CA, USA) was used for the washing process. Subsequently, o-Phenylenediamine (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) and 30% H2O2 were used for the chromogenic reaction; the reaction was stopped with 3 M H2SO4. Plasma HRG levels were measured using a 96-well plate reader (Nivo™ 5S Multimode Plate Reader; PerkinElmer, Waltham, MA, USA) at an absorbance of 492 nm. A standard curve was established using serial dilutions of known amounts of purified HRG (prepared in-house). Each plasma sample was measured in duplicate, and plasma HRG levels were determined by averaging two independent assays. The intra and inter-assay coefficients of variability were 7.4% and 13%, respectively. WBC, CRP, PCT, and P-SEP levels were measured from the same blood used for the HRG-level measurements. PCT and P-SEP levels were determined using a chemiluminescent enzyme immunoassay (SRL, Tokyo, Japan). WBC and CRP levels were measured at the Clinical Chemistry Laboratory of Okayama University Hospital.

Outcomes

The primary outcome was the HRG levels on POD 1 in the patients with and without postoperative complications. The secondary outcomes were the WBC, CRP, PCT, and P-SEP levels on POD 1 in the patients with and without postoperative complications, the association of HRG, WBC, CRP, PCT, and P-SEP with postoperative complications, and their ability to predict postoperative complications.

Statistical Analysis

The statistical approach was designed a priori. Multivariate, receiver operating characteristic (ROC) curve, and subgroup analyses were designed as post-hoc analyses. Categorical variables are expressed as numbers (percentiles) and compared using Fisher's exact test. Continuous variables are expressed as median and interquartile ranges (IQRs, 25–75th percentiles) and compared using the Mann–Whitney U test or Kruskal–Wallis test. Furthermore, the Steel–Dwass test was used to compare the medians of continuous variables for the post-hoc analysis among the three groups. The differences in the means of continuous variables are expressed as differences in means and 95% confidence intervals (CIs) and were compared using t-tests. Cox proportional hazards models and ROC curve analysis were used to assess the ability of each biomarker to predict postoperative complications. The results of the Cox proportional hazards models are expressed as hazard ratio (HR), 95% CI, and Harrel C-index score. In the multivariate analysis, we adjusted for the presence of preoperative cardiovascular comorbidities, age, American Society of Anesthesiologists Physical Status Classification (ASA-PS), operative time, and the volume of intraoperative bleeding, which have been reported to be associated with postoperative complications.[1,2,26–28] To assess the association between HRG levels and postoperative complications, we utilized the Kaplan–Meier method and log-rank test by classifying patients into two groups using the cut-off levels obtained from the logistic regression ROC curve analysis. A two–sided P-value < 0.05 was considered statistically significant. Data were analyzed using JMP Pro 14.0.0 (SAS Institute Inc., Cary, NC, USA) and STATA 16.1 and 17.0 (Stata Corp LLC, College Station, TX, USA).

processing....