Spectrophotometric Quantification of Bilirubin in Hemorrhagic Spinal Fluid using an Innovative Algorithm

Multi-wavelength spectrophotometric analysis for detection of xanthochromia in cerebrospinal fluid and accuracy for the diagnosis of subarachnoid hemorrhage

BACKGROUND

Cerebrospinal fluid (CSF) was examined for bilirubin, an important indicator for diagnosis of subarachnoid hemorrhage (SAH).

METHODS

A multi-wavelength (340, 415, and 460 nm) spectrophotometric assay was developed for the quantitative measurement of bilirubin in CSF, enabling the mathematical correction for absorbance of hemoglobin and proteins. Bilirubin and hemoglobin results were correlated to HPLC and a standard colorimetric assay, respectively. A subset of samples was sent for an absorbance reading at 450 nm following baseline correction. The multi-wavelength bilirubin assay was validated on 70 patients with confirmed SAH and 70 patients with neurologic symptoms who ruled out for SAH.

RESULTS

The multi-wavelength spectrophometric assay demonstrated no interferences due to proteins (albumin) up to 30 g/l or oxyhemoglobin up to 260 mg/l. The assay limit of detection was 0.2 mg/l, linear to 20 mg/l, and CVs ranged from 1 to 6% at bilirubin concentrations of 0.84 and 2.1mg/l. The spectrophotometric assay correlated to HPLC and the colorimetric assay for bilirubin and hemoglobin, respectively. Results also correlated to the absorbance method (with removal of samples with high hemoglobin and proteins). The area under the ROC curve for diagnosis of SAH was 0.971 and 0.954 for the HPLC and spectrophotometric assay, respectively. At a cutoff of 0.2mg/l, the clinical specificity was 100% for both assays, and the clinical sensitivity was 94.3% and 88.6% for SAH for the HPLC and spectrophotometric asays, respectively.

CONCLUSIONS

The multi-wavelength spectrophotometric assay is an objective alternative to visual inspection, HPLC, and absorbance for CSF bilirubin.

Device for quantification of bilirubin in cerebral spinal fluid

In North America, an estimated 30,000 patients annually experience an aneurysmal subarachnoid hemorrhage (SAH). In approximately five percent of these patients, the hemorrhage is not visible on computerized tomography scans due to the inability to image blood at time intervals greater than 12 h post symptom onset. For these patients (many of which have experienced a sentinel hemorrhage that is a precursor to a more significant rupture), a method is needed for accurately analyzing cerebral spinal fluid (CSF) for evidence of SAH. Further, it is necessary to differentiate blood associated with the SAH from blood associated with the spinal tap procedure. This letter presents a point-of-care device that is capable of performing such an analysis. The stand-alone prototype device uses commercially available embedded system components to implement a point-of-care device that is capable of collecting and analyzing optical absorbance spectra. A mathematical model for the hemorrhagic CSF sample is then developed by using a partial-least-squares-regression-based regression methodology that is able to differentiate between SAH and blood associated with the spinal tap. This differentiation is achieved by quantifying bilirubin (associated with the breakdown of old blood) in the CSF. Initial testing on the prototype device suggests that the device is able to quantify bilirubin in the presence of hemoglobin over concentrations ranges that are clinically relevant to the patient population of interest.

Improved quantification of CSF bilirubin in the presence of hemoglobin using least squares curve-fitting

Subarachnoid hemorrhage (SAH) is a dangerous neurological event with a very short time window for early diagnosis. Clinical diagnoses performed in a lab seek to quantify bilirubin in cerebrospinal fluid (CSF) as a biomarker for SAHs; however laboratory assays suffer from lengthy protocols, interference from hemoglobin, and the availability of expertise. Substantial improvements in the determination of bilirubin concentration in the presence of hemoglobin in CSF are demonstrated in this work. Concentration estimates within 15% for bilirubin in the range of 0.2 to 1.6 mg /dl were determined for CSF samples containing fresh hemoglobin concentrations ranging from 0.05 to 0.25 g/dl. To demonstrate extensibility of the system with respect to more complete mock SAH samples, sample sets with one additional species of both hemoglobin and bilirubin, methemoglobin and alpha-bilirubin, respectively, were tested and yielded results within 25% of actual values, as measured by standard chemical assays of preparations prior to mixing.

Development of a point-of-care device for the quantification of bilirubin in cerebral spinal fluid

In North America, an estimated 30,000 patients annually experience an aneurysmal subarachnoid hemorrhage (SAH). In approximately five percent of these patients, the hemorrhage is not visible on computerized tomography scans due to the inability to image blood at time intervals greater than 12 hours post symptom onset. For these patients (many of which have experience a sentinel hemorrhage that is a precursor to a more significant rupture) a method is needed for accurately analyzing cerebral spinal fluid (CSF) for evidence of SAH. Further, it is necessary to differentiate blood associated with the SAH from blood associated with the spinal tap procedure. This paper presents the development of a point-of-care device that is capable of performing such an analysis. The stand alone prototype device uses commercially available embedded system components to implement a hardware platform that is capable of collecting and analyzing optical absorbance spectra. A mathematical model for the hemorrhagic CSF sample is then developed using a PLSR based regression methodology that is able to differentiate between SAH and blood associated with the spinal tap. This differentiations in achieved by quantifying bilirubin (associated with the breakdown of old blood) in the CSF. Initial testing on the prototype device suggests that the device is able to quantify bilirubin in the presence of hemoglobin over concentrations ranges that are clinically relevant to the patient population of interest.