To the Editor.
—In their study of CNS invasion by B burgdorferi, Luft et al1 have taken great care to minimize the risk of contamination of CSF samples in the laboratory, pointing out the extreme sensitivity of the PCR they used to detect B burgdorferi DNA. However, they fail to address the issue of sample contamination during collection. A lumbar puncture is an invasive diagnostic intervention that frequently results in sample contamination with blood. Clinically, we are familiar with the observation that erythrocytes present in the first tube of CSF are often absent in the fourth tube, suggesting dilution of contaminating blood. Contamination of CSF samples with even a minute volume of blood containing target DNA could lead to the erroneous conclusion that the CSF was the source of the sequences amplified. Although we are given the values of CSF fluid protein, glucose, and leukocyte counts in Table 1,
Donnenberg MS. Borrelia burgdorferi in the Central Nervous System. JAMA. 1992;268(7):872-873. doi:10.1001/jama.1992.03490070050033