Detection of Chlamydial Cervicitis by Papanicolaou Stained Smears and Culture

Size: px

Start display at page:

Download "Detection of Chlamydial Cervicitis by Papanicolaou Stained Smears and Culture"

James Gibbs
6 years ago
Views:

Detection of Chlamydial Cervicitis by Papanicolaou Stained Smears and SANDY A. DORMAN, M.D., LEIGH M. DANOS, CMIAC, DEBBIE J. WILSON, KENNETH L. NOLLER, M.D., GEORGE D. MALKASIAN, M.D., JOHN R.

1 Detection of Chlamydial Cervicitis by Papanicolaou Stained Smears and SANDY A. DORMAN, M.D., LEIGH M. DANOS, CMIAC, DEBBIE J. WILSON, KENNETH L. NOLLER, M.D., GEORGE D. MALKASIAN, M.D., JOHN R. GOELLNER, M.D., AND THOMAS F. SMITH, PH.D. The ability of the cervical Papanicolaou (Pap) smear to detect cervicitis associated with Chlamydia trachomatis was investigated. s and cervical cytology samples were obtained from 487 women seen at the Mayo Clinic Department of Obstetrics and Gynecology and at the Sexually Transmitted Disease clinic of the Olmsted County Health Department. Adequate pap smears contained endocervical or metaplastic cells. Thirtyseven patients had positive Chlamydia cultures; of these, 4 were cytologically suggestive of Chlamydia, were negative but satisfactory, and 3 were unsatisfactory. Of the 45 patients with negative cultures, 2 had cytologic findings suggesting infection. Thus, of 35 cases suggestive cytologically, 4 (4%) were confirmed by culture. Of the 2 falsepositive cytologies, eight were in postpartum women. Because of problems with specificity and inadequate smears, cervical cytology specimens should not replace culture as a means of detection, but can identify women who should be cultured for C. trachomatis. (Key words: Chlamydia trachomatis; Cervical cytology; Pap smear; Cytoplasmic inclusions; McCoy cell culture) Am J Clin Pathol 983; 79: CHLAMYDIA TRACHOMATIS is wellrecognized as a common cause of infection of the female genital tract. Of women attending sexually transmitted disease (STD) clinics, 23% yielded chlamydiae from the cervix. 63 ' 8 An isolation rate of 22% also has been noted in sexually active female adolescents. 6 Women with this infection are at risk for a variety of diseases, including cervicitis, 3 salpingitis, 9 endometritis, 4 vulvovaginitis, 3 and they are a potential source of infection to male partners. A neonate of an infected mother is at risk of acquiring inclusion conjunctivitis 2 and pneumonia. While the identification of women harboring C trachomatis is desirable, cell culture procedures are inapplicable to the screening of large numbers of women because of the cost and difficulty of the technic. Serologic tests are not practical because of a high prevalence of chlamydial antibodies in culturenegative women. 7 Naib" originally described Chlamydiainduced changes noted in Papanicolaou (Pap) stained smears of newborns with neonatal conjunctivitis and from their mothers. Gupta and coworkers 5 emphasized and refined these Received June 7, 982; received revised manuscript and accepted for publication August 5, 982. Address reprint requests to Dr. Dorman: Department of Pathology, Easton Hospital, 2st and Lehigh Streets, Easton, Pennsylvania 842. Sections of Clinical Microbiology, and Obstetrics and Gynecology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota findings as they pertained to women with chlamydial cervicitis. Despite these cytologic descriptions, considerable skepticism remains about the usefulness and validity of these observations. 84 The purpose of the study was to investigate the specificity and sensitivity of cervical cytology for the detection of chlamydial infections. A comparison of this technic to culture methods was made. Patient Population Materials and Methods Study participants were recruited from three sources: () those seeking medical care in the Division of Gynecology of the Mayo Clinic, which serves the women of Olmsted County as well as a large referral population; (2) women seen by the Division of Obstetrics of the Mayo Clinic, a population which consists primarily of women native to the area; and (3) women attending a small STD clinic at the Olmsted County Health Department. All women who had a cervical smear also had a culture obtained at the same time. In addition, women who were seen primarily because of vaginal discharge and who had specimens taken for chlamydial culture also had a cervical smear obtained. A total of,596 women had both samples obtained during the study period; however, only 487 specimens were cultured immediately, with the remainder frozen at 7 C for evaluation at a later date. Those cultured immediately included 375 patients with discharge plus 2 samples from asymptomatic women. The latter group included all asymptomatic women seen at the STD clinic (58); women with positive cervical smears (8); and two controls with negative cytologies agematched to each of the women with positive cervical smears (36). Based on findings from this group, an additional 332 (3%) of the remaining,9 frozen specimens were chosen randomly, thawed, and cultured. 2973/83/4/42 $.5 American Society of Clinical Pathologists 42 on 6 April 28

2 422 DORMAN ET AL. A.J.C.P. April 983 Methods for C. trachomatis was performed as previously described. 9 In brief, specimens were obtained by inserting a cotton swab 2 centimeters into the cervical canal in nonpregnant women, or at the os without insertion in pregnant women. Swabs were extracted into 2SP transport medium, and stored at 2 C during transportation to the laboratory. Aliquots (O.i ml) of the 2SP extracts were inoculated into glass shell vials containing coverslips seeded with McCoy's cells. The shell vials were centrifuged, incubated with medium containing cycloheximide, and the monolayers stained with iodine to identify typical chlamydial inclusions. Because of the large number of specimens to be cultured, only one vial was inoculated rather than the four vials usually inoculated. All cytologic smears were obtained from the cervix using wooden Argis spatulas. The cells were spread onto a glass slide and immediately fixed in 95% ethanol. The slides were stained using a modified Papanicolaou method, and all slides were screened without knowledge of the culture results. Specimens satisfactory for evaluation included those in which either endocervical or metaplastic cells were present. The slides from the 487 patients specified above were screened by one of the authors (L.D.). The remainder were screened by staff cytotechnologists. All slides positive on screening for chlamydiajlike inclusions were examined by staff pathologists. Gupta and associates 5 have described three distinct cytologic patterns associated with Chlamydia. Stage included infected cells that were squamoid in appearance, often had slight nuclear alterations, and perinuclear or a diffuse, finely granular cytoplasm. Stage 2 cells had finely granular cytoplasm in which single or multiple inclusion bodies could be seen. The cytoplasm was described as "frothy" in appearance. In the third identifiable stage, eosinophilic chlamydial particles and/or dense aggregates were seen in the cytoplasmic inclusions. The minimum criteria for the diagnosis of a chlamydial infection was the presence of Stage 2 and 3 cytoplasmic inclusions (Figs. and 2). Results Positive cultures were obtained from 37 women: women were seen in the Gynecology Department; five were from Obstetrics, four of whom were pregnant; and 22 cases were identified from the STD clinic, two of whom were pregnant. Twentyfour of the 37 culturepositive women had satisfactory smears, and chlamydial inclusions were noted in 4 (58%). An additional 2 women had cytologic specimens positive for Chlamydia but were not confirmed by culture (Table ). Of the 487 patients cultured, 38 (63%) had adequate cervical cytology as described above. All 36 agematched controls were negative. Cytologic examination detected chlamydial infection in none of four culturepositive pregnant women seen at the Division of Obstetrics (Table 2) and only one of two examined at the STD clinic. In this population, only 3% of cervical smears were adequate for examination. Evaluation of postpartum women revealed a significant number (8/2) of positive cytologies, which were not confirmed by culture (Table 3). Postpartum women accounted for only 8.4% of all patients cultured. Ninetyfive per cent of cytologic smears were satisfactory for examination in this group of patients. Of the 37 culturepositiye patients, nine (25%) were asymptomatic. Six of nine had adequate cytologic specimens, and four were detected by the cervical smear prior to culture. The remainder were detected because all women seen at the STD clinic were cultured. s of the 332 asymptomatic women with negative cytology revealed only two (.6%) that were positive for C. trachomatis. Discussion Although the frequency of infections with C. trachomatis is increasing rapidly, 7 the laboratory diagnosis of the organism remains difficult. The most reliable procedure, growth in cell culture, is timeconsuming, technically demanding, and expensive. Consequently, this capability is generally available only in large laboratories. Serologic tests are valuable only in acute primary infections, and most useful in systemic infections, such as pelvic inflammatory disease, in which high titers FIG. I (upper). Chlarhydiainfected cells. These metaplastic cells dep Stage and Stage 3 inclusions. Metaplastic cells with Stage inclusions demonstrate "porous" cytoplasm. Finely granular "coccoid bodies" m be seen within the "porous" areas. Stage 3 inclusions, seen in two cells, are perinuclear inclusions with distinct borders, molding, and central rget formation. Papanicolaou stain (original magnification X4). FIG. 2 (lower). Chlamydiainfected cells. Metaplastic cells demonstrate Stage 2 inclusions consisting of larger granular elementary bodies. Several larger inclusion bodies also may be seen. Papanicolaou stain (original magnification X53). > on 6 April 28

3 Vol. 79 No. 4 CERVICAL CYTOLOGY AND CHLAMYDIA «* on 6 April

4 424 DORMAN ET AL. A.J.C.P. April 983 Table. Comparison of Papanicolaou Stain with in Specimens from Women Attending the Obstetrics and Gynecology Departments at the Mayo Clinic and a Sexually Transmitted Disease Clinic Unsatisfactory Table 2. Comparison of Papanicolaou Stain with in Specimens from Pregnant Women Attending the Obstetrics Department at the Mayo Clinic 9 Unsatisfactory (> :256) develop. The majority of patients with chlamydial infections are seen at STD clinics, where specimens for the recovery of this organism are rarely collected because of the complexity and expense of the test. However, the public health problem is significant, since the prevalence of chlamydial cervicitis approaches 2 3% in these clinics. 63 A simple, inexpensive screening method that would identify those women who should be cultured would be of great benefit and obviate the Table 3. Comparison of Papanicolaou Stain with in Specimens from Postpartum Women Attending the Obstetrics Department at the Mayo Clinic Unsatisfactory necessity of performing cultures on all women with discharge. Gupta and associates 5 reported that Papanicolaoustained smears contain three stages of inclusions when the patient had cervicitis due to C. trachomatis. However, cultural proof of infection was obtained in only four of 6 (2.5%) cases. Subsequent studies have noted varying degrees of inflammation or dysplasia but no inclusions. These studies dispute Gupta's findings. 2 ' 8,l4 One study 6 demonstrated chlamydial inclusions in cervical smears. However, only six cases were reviewed, with two positive for inclusions. These numbers are too small to be meaningful. Our findings, in part, corroborate those of Gupta. Fourteen of 37 culturepositive women were detected by cervical smear. Twentyone patients were culturenegative yet had positive cytologies. In our experience, a positive cervical smear was confirmed by culture 4% of the time. Several problems arise when using the cervical smear to make the diagnosis of chlamydial cervicitis. Because chlamydiae infect endocervical or metaplastic cells, an adequate sample of the squamocolumnar junction must be obtained. In our study, only about 63% of our cervical smears contained an appropriate sample of cells in which to make the diagnosis. Reasons for the absence of these cells would include variation of specimen collection and the location of the squamocolumnar junction. Notably, in pregnant women the percentage of adequate cytologic smears decreased to 3 % (Table 2) possibly because obstetricians were reluctant to enter the endocervical canal during pregnancy. Because so few smears were adequate, detection of chlamydial infection by cervical cytology during pregnancy rarely was accomplished. The problem of adequate cytology is illustrated further by four smears from the STD clinic. Each smear had classic Stage and 2 inclusions but without Stage 3 were not called positive. Possibly, if more endocervical cells had been present, stage 3 inclusions may have been seen. Perhaps more significant is the problem of specificity. Sixty per cent of our smears were not confirmed by cultural methods. The sensitivity of the cultural method is not known 7 ; however, it appears to be 89%, based on one study of both male and female patients. 8 Chlamydial inactivation during storage, transit, and processing may have precluded detection of C. trachomatis using cell culture technics. Some specimens were not inoculated for 48 hours, which has been shown to reduce the isolation rate by approximately 2%. 5 The sensitivity of the culture methodology employed in this study has been shown to be good, although not optimal. By examining two coverslips rather than four, the test sensitivity was 98% 2 ; when only one was inoculated the on 6 April 28

Vol. 79 No. 4 CERVICAL CYTOLOGY AND CHLAMYDIA 425 sensitivity was 92%. These problems, together with presumed subclinical infection, may have precluded detection in at least some of these women.

5 Vol. 79 No. 4 CERVICAL CYTOLOGY AND CHLAMYDIA 425 sensitivity was 92%. These problems, together with presumed subclinical infection, may have precluded detection in at least some of these women. An alternative view is that these inclusions may not be specific for Chlamydia. This is particularly evident in the postpartum patients evaluated. In this small group of women, we found 38% (Table 2) of the unconfirmed positive cytologic smears. It is possible that in postpartum women, the changes seen in cervical smears may be due to something else. Followup of seven of eight women reveals that they remain asymptomatic, as they were at the time of the original culture. The eighth woman has vaginitis and has been culturenegative for C. trachomatis on two separate occasions. Of particular interest is the large number of asymptomatic, culturepositive women who were seen. This finding is consistent with data from other studies. 38 It was encouraging to note that of the six adequate smears from these patients, chlamydial inclusions were noted in four. The finding of such a large percentage of asymptomatic women who were culturepositive prompted us to investigate more asymptomatic women with negative cytologies to determine if a significant number of culturepositive patients were missed. As noted, only.6% of subsequent specimens grew C. trachomatis. In this population, asymptomatic women who do not have positive cytologic results are unlikely to harbor Chlamydia. It must be emphasized that the majority of patients seen at the Mayo Clinic are from an upper socioeconomic population, and many are referral patients who do not come primarily for evaluation of vaginitis or vaginal discharge. Therefore, these results may not be extrapolated to areas of the country where chlamydial infections are more endemic. In summary, because of problems with specificity and inadequate smears, cervical cytology specimens (Papanicolaou smears) should not replace cultural methods for the diagnosis of Chlamydia. The Mayo Clinic cytology laboratory reports cervical smears with chlamydial inclusions as "changes suggestive of chlamydial infection." Thus, when the etiology of the vaginal discharge or other genital tract symptomatology is unknown, cervical cytology may serve to identify those women who should be cultured for Chlamydia. This approach could effect considerable savings for both individual patients and STD clinics. References. Beem MO, Saxon EM: Respiratorytract colonization and a distinctive pneumonia syndrome in infants infected with Chlamydia trachomatis. N Engl J Med 977; 296:36 2. Briggs RM, Holmes KK, Kiviat N, Barker E, Eschenbach DA, DeJong R: High prevalence of cervical dysplasia in STD clinic patients warrants routine cytologic screening. Am J Public Health 98;7: Golmeier D, Ridgeway GL, Oriel JD: Chlamydial vulvovaginitis in a postmenopausal woman. Lancet 98; 2: Gump DW, Dickstein S, Gibson M: Endometritis related to Chlamydia trachomatis infection. Ann Intern Med 98; 95: Gupta PK, Lee EF, Erozan YS, Frost JK, Geddes ST, Donovan PA: Cytologic investigations in Chlamydia infection. Acta Cytol 979;23: Hilton AL, Richmond SJ, Milne JD, Hindley F, and Clarke SKR: Chlamydia A in the female genital tract. Br J Vener Dis 974; 5: 7. Holmes KK. The Chlamydia epidemic. JAMA 98; 245: Kiviat N, Paavonen J, Brockway J, et al: High prevalence of cytologic atypias in cervical chlamydia infection (Abstract). st Sexually Transmitted Disease World Congress. Puerto Rico, Nov. 52, Mardh PA, Ripa T, Svensson L, Westrom L: Chlamydia trachomatis infection in patients with acute salpingitis. N Engl J Med 977;296: Mardh PA, Westrom L, Colleen S, WolnerHanssen P: Sampling, specimen handling and isolation techniques in the diagnosis of Chlamydial and other genital tract infections. Sex Transm Dis 98; 8:2885. Naib ZM: of TRIC agent infection of the eye of newborn infants and their mothers' genital tracts. Acta Cytol 97; 4: Naib ZM: Oculogenital infection with TRIC agents. Clin Obstet Gynecol 972; 5: Oriel JD, Powis PA, Reeve P, Miller A, Nicol CS: Chlamydial infection of the cervix. Br J Vener Dis 974; 5:6 4. Paavonen J, Purola E: Cytologic findings in cervical chlamydial infection. Med Biol 98; 58: Reeve P, Owen J, Oriel JD: Laboratory procedures for the isolation of Chlamydia trachomatis from the human genital tract. J Clin Pathol 975; 28: Saltz GR, Linneman CC, Brookman RR, Rauk JL: Chlamydia trachomatis cervical infections in female adolescents. J Pediatr 98; 98: Schachter J, Cles L, Ray R, Hines P: Failure of serology in diagnosing Chlamydial infections of the female genital tract. J Clin Microbiol 979; : Schachter J, Hanna L, Hill EC, et al: Are chlamydial infections the most prevalent venereal disease? JAMA 975; 23: Smith TF: Role of the diagnostic virology laboratory in clinical microbiology: Tests for Chlamydia trachomatis and enteric toxins in cell cultures. Medical Virology. Edited by de la Maza, LM, Peterson EM. New York, Elsevier Biomedical, 982, pp Smith TF, Brown SD, Weed LA: Diagnosis of Chlamydia trachomatis infections by cell cultures and serology. Lab Med 982; 3:92 on 6 April 28

Immunodiagnosis of Sexually Transmitted Disease

THE YALE JOURNAL OF BIOLOGY AND MEDICINE 58 (1985), 443-452 Immunodiagnosis of Sexually Transmitted Disease JULIUS SCHACHTER, Ph.D. Department oflaboratory Medicine, University of California, San Francisco,