[Skip to Content]
[Skip to Content Landing]
Clinicopathologic Reports, Case Reports, and Small Case Series
January 1, 2005

Fluorescein Interference With Homocysteine Testing

Arch Ophthalmol. 2005;123(1):114-116. doi:10.1001/archopht.123.1.114

Measuring plasma homocysteine levels is increasingly common in the evaluation of patients with retinal vascular disease for a possible hypercoagulable state. Fluorescein administered in angiographic studies may interfere with blood tests using fluorescein-labeled reagents, as well as other methods, measuring near its 493.5-nm absorption maximum or 525-nm emission wavelengths or involving fluorescence quenching. To our knowledge, this is the first report of fluorescein interference with plasma homocysteine testing.

Report of a Case

A 37-year-old man was referred to the Lahey Clinic in Burlington, Mass, for decreased vision of 2 months’ duration in the right eye. The patient had a history of hypertension and Crohn disease for which he took multiple medications, including prednisone.

At initial examination, his best-corrected visual acuity was 20/80 OD and 20/25 OS. Results from dilated funduscopic examination, fluorescein angiography, and optical coherence tomography were consistent with findings of a central retinal vein occlusion in the right eye.

Workup for a hypercoagulable state was initiated. Approximately 1 hour after administration of fluorescein dye, the patient’s blood was drawn. The workup revealed an iron deficiency anemia with a hematocrit level of 25% (reference range, 42%-52%). The homocysteine level could not be measured because of high background fluorescence. One week later, his homocysteine level was normal at 1.58 mg/L (11.7 μmol/L) (reference range, 0.54-1.62 mg/L [4-12 μmol/L]).

Plasma homocysteine level is measured at the Lahey Clinic laboratory by fluorescence polarization immunoassay (Abbott AxSYM; Abbott Laboratories, Chicago, Ill). The high background fluorescence, demonstrable with UV illumination (Figure), interfered with detection of the fluorescein label used in this assay, resulting in an error code. This is a competitive, homogeneous immunoassay. Assays of this type do not have wash steps or phase separations, so all of the constituents of the sample, including the fluorescein contaminant, are present in the final reaction mixture. The likelihood of interference is influenced both by the concentration of fluorescein in the patient’s sample and the sample volume for a particular test relative to the total reaction volume.1

From left to right, normal serum, the patient’s serum, and lipemic serum under ambient (A) and UV light (B).

From left to right, normal serum, the patient’s serum, and lipemic serum under ambient (A) and UV light (B).

Heterogeneous immunoassays, a much more common method of analysis, include 1 or more separation steps. For example, the analyte of interest may be bound by antibody to a solid phase, or they may be precipitated together. Other sample components do not bind, are washed away, and so are no longer present as potential interferences when the eventual result, fluorescent or otherwise, is measured.

Fluorescein interference after angiography has been reported for a number of fluorescence polarization assays, including those for serum creatinine, total protein, cortisol, thyroxine, digoxin, quinidine, and cyclosporine levels.2,3 Of these, cyclosporine level is commonly measured by fluorescence polarization immunoassay. Fluorescein interference has also been reported with tests performed on the Kodak Ektachem (Kodak, Rochester, NY), including measurements of serum amylase, unconjugated bilirubin, and conjugated bilirubin levels; minor spurious elevations are reported for aspartate aminotransferase and alkaline phosphatase measures. Urinary creatinine and protein, serum total protein, magnesium, and chloride measurements using the DuPont aca(DuPont, Research Triangle Park, NC) and Beckman Astra analyzers (Beckman Coulter, Fullerton, Calif) are also reported to have interference by fluorescein following retinal angiography.4,5

Elimination of fluorescein predominantly occurs through the kidneys within 36 to 48 hours if renal function is normal. Interference with serum testing has been noted up to 12 hours later in individuals with normal renal function.2 Interference may persist much longer in individuals with renal impairment, common among patients with diabetes mellitus. Ophthalmologists should be aware of the potential interference of fluorescein from retinal angiography in serum and urinary tests and manage patients accordingly.

Correspondence: Dr Marx, Department of Ophthalmology, Lahey Clinic, 41 Mall Rd, Burlington, MA 01805 (jeffrey.l.marx@lahey.org).

Financial Disclosure: None.

Schneider  MCRayburn  BKFrye  R Interference by intravenous fluorescein with drug assays in the Abbott TDx. Clin Chem 1983;292001
Bloom  JNHerman  DCElin  RJ  et al.  Intravenous fluorescein interference with clinical laboratory tests. Am J Ophthalmol 1989;108375- 379
Rutzky  LPAmador  AWelsh  MKahan  BD Production of interference in the TDx polarization immunoassay by fluorescein angiography. Transplantation 1993;56491- 492
Koumantakis  GWyndham  L Fluorescein interference with urinary creatinine and protein measurements. Clin Chem 1991;371799
McClellan  SLPerry  BWDoumas  BT Effect of fluorescein on clinical laboratory tests. Clin Chem 1982;282335- 2336