Stability of archived liquid-based cytologic specimens

摘要

In our recent article,1 we reported that long-term storage of cervical Papanicolaou specimens in PreservCyt (Cytyc Corporation, Boxborough, MA), a methanol-based medium used in liquid cytology, resulted in significant losses in DNA stability and nuclear preservation. Because of our interest in using these specimens as an alternative source of DNA for measurements related to our studies of human papillomavirus (HPV) and cervical neoplasia,2, 3 we further examined whether these archived specimens would support polymerase chain reaction (PCR) amplification assays aimed at detecting other infectious agents that might be etiologic cofactors. To readdress this issue, we tested enrollment cervical specimens from 70 consenting women in our natural history study (approved by the National Cancer Institute and INCIENSA [Instituto costarricense de Investigacion y ensenanza en salud y Nutricion], Costa Rican institutional research boards), initiated in 1993–1994, of HPV and cervical neoplasia in Guanacaste, Costa Rica.2 PCR amplification was used to test specimens for the presence of Chlamydia trachomatis and Neisseria gonorrhoeae DNA. Paired cervical specimens were collected at the same enrollment visit, one stored in PreservCyt at ambient temperature and the other stored in specimen transport medium (STM; Digene, Gaithersburg, MD) at −70 °C, were tested for C. trachomatis plasmid DNA and N. gonorrhoeae genomic DNA using a commercial PCR assay (AMPLICOR; Roche Diagnostics, Indianapolis, IN) to amplify fragments of 207 and 201 base pairs (bp), respectively.4, 5 These target fragments were smaller than the smallest β-globin fragment (268 bp) that we measured in our previous report.1 For increased prevalence of these sexually transmitted infections, we selected a subset of women who had either at least 3 different sexual partners before age 30 years or at least 5 different sexual partners before age 48 years. As controls, PreservCyt specimens that were positive for either C. trachomatis or N. gonorrhoeae (n = 10; it was unknown for which infection each specimen was positive) and specimens that were negative for C. trachomatis and N. gonorrhoeae (n = 10) were provided by Digene and included in the testing, which was performed in a masked fashion. PreservCyt aliquots of 4.5 mL were pelleted by centrifugation and suspended in 100 μL TE buffer (10 mM Tris, 1 mM ethylenediamine tetraacetic acid, pH 8.0) for testing; 100 μL of the STM specimen was used without additional manipulation. Specimens were lysed and DNA was extracted using the MagNA Pure LC Robot (Roche Diagnostics), and magnesium chloride concentration was adjusted to 1.5 mM for PCR amplification. DNA was amplified using biotinylated primers specific for either C. trachomatis or N. gonorrhoeae; amplicons were captured by hybridization to microwells coated with oligonucleotides that were specific for either C. trachomatis or N. gonorrhoeae; and bound, biotinylated amplicons were detected using avidin–horseradish peroxidase and a colorometric substrate.4, 5 Results from paired specimens were compared using a symmetry chi-square test. Among the 70 pairs of specimens selected, a PreservCyt specimen was missing from 1 pair and an STM specimen from another pair did not amplify; these pairs therefore were excluded from the analysis, leaving a final set of 68 paired specimens. The results of the testing are shown in Table 1. We found that only 8 of 67 (11.6%) PreservCyt specimens were positive for C. trachomatis, significantly less than the 19 of 67 (28.4%) STM specimens found to be positive (P = 0.012; symmetry chi-square test). (One pair was excluded from the C. trachomatis analysis because the result for the STM specimen was indeterminate and the corresponding PreservCyt specimen was negative.) Similarly, we found that only 1 of 68 (1.5%) PreservCyt specimens was positive for N. gonorrhoeae, significantly less than the 9 of 67 (13.2%) STM specimens found to be positive (P = 0.011; symmetry c