- Email: [email protected]
Increased frequency of detection of human papillomavirus deoxyribonucleic acid in exfoliated cervical cells during pregnancy
Increased frequency of detection of human papillomavirus deoxyribonucleic acid in exfoliated cervical cells during pregnancy Robert F. Rando, PhD,"·b Steven Lindheim, MD,b Lisa Hasty, MD,b Thomas V. Sedlacek, MD,b Mark Woodland, MD,b and Catherine Eder, BAa Philadelphia, Pennsylvania Exfoliated cells of the uterine cervix obtained from women during pregnancy and at the time of their first postpartum examination were used to monitor the prevalence of human papillomavirus infections in this population and to study the natural fluctuations in viral expression. When deoxyribonucleic acid hybridization analysis alone was used to monitor the presence of human papillomavirus infection, 20.9% of our study population had results that were positive for human papillomavirus deoxyribonucleic acid during their first-trimester examinations. A dramatic increase in the percentage of women with positive results for human papillomavirus deoxyribonucleic acid was observed at the time of the patients' third-trimester examinations (46%). The overall increase in human papillomavirus-positive patients was a combination of a small number of patients who had positive results on their first examination and negative results on their second examination, and a larger number of patients who had negative results on their first-trimester examination and positive results for human papillomavirus deoxyribonucleic acid in the exfoliated cervical cells at the time of their third-trimester examination. The total percentage of patients with positive results for human papillomavirus deoxyribonucleic acid in their cervical cells at one or both assay points during pregnancy was 52.5%. Samples obtained at the postpartum examination demonstrated a dramatic decrease in the number of samples positive for human papillomavirus deoxyribonucleic acid (17.5%). This result was a combination of a large decrease in human papillomavirus-positive patients coupled with a small increase in detectable levels of human papillomavirus deoxyribonucleic acid in cervical samples from patients who had negative results on their previous examination. This study demonstrates a very high level of detectable human papillomavirus deoxyribonucleic acid in exfoliated cervical cells obtained during pregnancy and shows that the detectable levels of human papillomavirus deoxyribonucleic acid fluctuate during pregnancy. (AM J OSSTET GYNECOL 1989;161 :50-5.)
Key words: Human papilloma virus (HPV), deoxyribonucleic acid (DNA), uterine cervix, pregnancy, hybridization Surveillance of reported cases of clinically apparent anogenital human papillomavirus (HPV) infections indicated a steady rise in the incidence of new cases between 1966 to 1983. 1 In 1976 Meisels and Fortin 2 helped to heighten interest in human papillomaviruses when they reinterpreted the morphologic features of dysplasia of the uterine cervix as manifestations of HPV infections. Interest was further stimulated when HPV deoxyribonucleic acid (DNA) was identified in premalignant and malignant anogenitallesions. 3 To date more than 50 distinct types of HPVs have been described. At least 12 distinct types have been detected
From the Departments of Pathology" and Obstetrics and Gynecology,' Pennsylvama Hospital. Received for publication July 22, 1988; revised October 9, 1988; accepted January 10, 1989. Reprint requests: Robert F. Rando, PhD, Department of Pathology, Preston Bldg., Rm. 758, Pennsylvania Hospital, 8th and Spruce St., PhIladelphia, PA 19107.
50
in anogenital tract and oral cavity lesions. Several studies have reported that HPV type 16 and HPV type 18 are found in association with a majority of high-grade dysplastic lesions of the cervix, carcinoma in situ, and cervical carcinomas.' Two recent studies estimated that 10% to 20% of random populations that had cytologically "normal" cervices were in fact infected with HPV. 4 . 5 Burk et a\.6 detected HPV DNA in 29% of a population of women who were referred to a colposcopy clinic and had normal Papanicolaou smears. Wickenden et a\.7 detected HPV infections in 10.5% of a population from a sexually transmitted disease clinic who had cytologically normal cervices. It is interesting that a study conducted among pregnant women8 detected HPV DNA in 28% of the study population. The enhanced detection of viral infections has been reported in patients with impaired cell-mediated immunity.g· 10 It has also been noted that HPV infections and HPV -associated neoplasia of the anogenital tract
HPV detection during pregnancy
Volume 161 Number I
are more readily detected in immunodeficient patients than in the general population. 9 • 10 During pregnancy, a natural altered or depressed immunologic state has been suggested. s It has also been reported that clinically apparent HPV infections of the female genital tract seem to occur more frequently during pregnancy and regress rapidly after delivery. I I Schneider et al. B reported the detection of HPV DNA in exfoliated cervical cells in 28% of a pregnant population compared with 12.5% for nonpregnant control patients; a higher detection level occurred in samples obtained later in pregnancy. In addition to an immunodepressed condition, hormonal effects during pregnancy may playa role in the enhancement of viral infection or expression. Longterm users of oral contraceptives show an increased risk for the development of HPV-associated neoplasia. 12 13 There are also reports of hormonal stimulation of HPV-directed transcription in vitro 14 and proliferation of viral growth with [3-estradiol in vivo (Krieder J, personal communication). This article describes a prospective study designed to evaluate the basal level of cervical HPV infections in a population of pregnant women, and to determine whether there are fluctuations in the detectable levels of HPV DNA over the course of pregnancies through the first postpartum examination.
A
23
4
5
6
7
8
51
c
B
Material and methods
Sample collections. From April 1987 until August 1987 all new patients at Pennsylvania Hospital's obstetric clinic were recruited for this study. This population of women was young (mean age = 23.45 years), sexually active, and from a low socioeconomic group. At the time of the patient's first-trimester examination, a sample of exfoliated cervical cells was obtained and used for both cytologic and DNA hybridization analysis. Similar specimens of exfoliated cervical cells were obtained for DNA hybridization analysis during a thirdtrimester visit (28 to 34 weeks) and at the time of the patient'S first postpartum checkup (no later than 6 weeks after delivery). DNA extraction. Exfoliated cervical cells were suspended in phosphate-buffered saline solution, concentrated by low-speed centrifugation (1500 rpm), resuspended in 1 ml of cell lysis buffer (10 mmoll L Tris hydrochloride, pH 8.0, 0.6% sodium dodecyl sulfate, 10 mmollL ethylenediaminetetraacetic acid). Ribonuclease A (10 (J.g/ml) and proteinase K (50 (J.g/ml) were added to the cell lysis buffer and the samples were incubated at 37° C for 2 hours. Total cellular DNA was purified by extracting the lysed cellular material twice with an equal volume of phenollchloroform/isoamyl alcohol (25: 24 : 1). The extracted nucleic acids were then precipitated by adjusting the salt concentration to
Fig. 1. Detection of HPV DNA in samples obtained from exfoliated cervical cells by Southern blot hybridization. DNA was extracted, cleaved with restriction enzymes. electrophoresed, and transferred onto nylon filter membranes as described in "Material and methods." The procedure described was performed in triplicate. One of each of the three triplicate filters was hybridized with a combination of 32P-labeled molecular probes corresponding to HPV-16118 (A) or HPV-31/33/35 (data not shown) or HPV-6111 (B). DNA extracted from the samples analyzed in lane 3 hybridized better with the HPV-16118 probes. DNA extracted from the sample in lane 5 hybridized better with the HPV-6111 probes. Lane C contained 20 pg of an HPV -16 recombinant clone that was cleaved with Pst I.
0.3 mollL sodium acetate and adding 2.5 volumes of cold ( - 20° C) ethanol. The DNA was concentrated by centrifugation, washed twice with 70% ethanol, dried under vacuum, resuspended in 1 ml of a buffer containing 10 mmollL Tris hydrochloride (pH 7.5) and 1
52
Rando et al.
July 1989 Am J Obstet Gynecol
n=80
40 30
20
o
1st T r imester _
3rd T rimester
Init ia l
Total
~ New Positive 3rd T rimest.
Posit ive
Fig. 2. Graphic representation of HPV-positive cervical samples obtained from patients compliant through their third-trimester examination. The number of patients with positive results in the third trimester is a combination of patients who had positive results in the first trimester (minus those patients who had positive results at their first-trimester examination but negative results for HPV DNA in their cervical cells during their third-trimester examination) combined with the number of new patients with negative results in the first trimester but who had HPV detected during their third-trimester examination. The bar labeled total represents patients who were HPV DNA-positive in one or both of their examinations.
Table I. Correlation of Papanicolaou smear versus DNA hybridization analysis for the detection of HPV during first trimester of pregnancy Papamcolaou smear Positive*
DNA pos. DNA neg. TOTAL
6 8 14
I
Negativet
Total
17 79 96
87
23
110
*Papanicolaou-positive smears included those described as condylomatous, hyperkeratosis, parakeratosis, or squamous atypia. t Fourteen patients were classified as having benign atypia. Of those 14 patients, 5 had positive results for HPV DNA.
mmol/L ethylenediaminetetraacetic acid; the concentration was determined by spectrophotometric analysis at 260 nm. Hybridization analysis. Total DNA (10 to 15 J.l.g) extracted from patient samples was cleaved with the restriction endonuclease P st I as recommended by the manufacturer (New England Biolabs). The resulting DNA fragments were fractionated by electrophoresis in a O.B% agarose gel and then transferred to nylon filter membranes (Amersham Hybond-N).
DNA hybridization analysis was performed with modifications of the Southern blot technique as previously described. I5 Molecular probes corresponding to HPV types 6, 11, 16, IB, 31, 33, and 35 were amplified and purified using standard laboratory methods. I6 The HPV genomes were purified from flanking bacterial plasmid DNA and then radiolabeled with 32P-deoxycytidine triphosphate. IS. 16 Results
Exfoliated cervical cells obtained during the firsttrimester examination from 110 patients were examined by cytologic and DNA hybridization analysis for the presence of HPV infection. A comparison of the Papanicolaou smear and DNA hybridization analysis for the detection of HPV is described in Table I. Thirty-one of the 110 patients (28.2%) had positive results for the presence of HPV in their cervical cells. Of these 31 HPV-positive patients, eight (7.2 %) had positive results by Papanicolaou smears alone, six (5.4%) had positive results by both criteria, and 17 (15.4%) had only positive hybridization analysis results (Table J). A total of 23 of 110 (20.9%) had positive DNA hybridization analysis results. DNA hybridization analysis was performed in triplicate for the first-trimester patient samples. The tripli-
HPV detection during pregnancy
Volume 161 Number 1
53
A.
Table II. Correlation of sexual partners with detection of HPV DNA No. of sexual partners
1-2 n
I
2!3
%
n
I
%
HPV DNA pos.*
13/38
34.2
25/38
65.7
HPV DNA neg.
11/28
39.2
17/28
60.7
(n = 38) (n = 28)
x2
=
0.187
p> 0.05
*Any patient who had HPV DNA detected during their first-trimester, third-trimester, or both examinations were considered HPV DNA pos. cate hybridization reactions used mixtures of HPV types 16 and 18, HPV types 31, 33, and 35, or HPV types 6 and 11 for molecular probes (Fig. 1). Of the 23 patient samples that contained HPV DNA at the first-trimester examination, HPV-16118 was detected in 14 (60.8%) samples, HPV-31133/35 was detected in six (26%), and HPV-6111 was detected in one (4.3%) sample. In the remaining two samples (8.7%), HPV DNA was detected but not typable with the available molecular probes. Samples of exfoliated cervical cells were obtained from 80 of the original 110 patients during their third trimester-visit. Thirteen of these 80 patients (16.2%) had positive HPV DNA results during their firsttrimester examination. DNA hybridization analysis was performed on these samples. The results indicate an increase in the percentage of patients with positive results for HPV DNA in the third trimester, compared with the first-trimester results (Fig. 2). This increase was reflected in a moderate regression of detectable levels of HPV DNA in patients with positive results for HPV DNA during the first-trimester visit and negative results at the third-trimester visit (7.5%), coupled with a large increase of patients with negative results for HPV DNA during the first trimester and positive results at the third trimester (37.5%). The total number of patients with positive HPV DNA results during their third-trimester of pregnancy was 37 of80 (46.2%). The total number of patients who had positive results for HPV DNA at either or both examinations was 43 of 80 (53.7%). No significant correlation was observed when information concerning the smoking habits (obtained from 64 patients who remained compliant through their third-trimester examination) was compared with the presence of HPV DNA in the patients cervical cells (data not shown). In addition, information concerning the number of sexual partners (obtained from 66 patients compliant through the third trimester) was compared with the detection of HPV DNA in the patients
let Ttl.
~
Trl
Poet Pwtun
Tol8l
n=57
B.
i i
I o
-~
Fig. 3. Graphic representation of HPV DNA analysis for those patients who were compliant through their postpartum examination. The total percentage of patients positive at the three assay points is depicted in the top panel (A). The percentage of patients who changed with respect to the detection of HPV DNA in their cervical cells from first to third trimester or from third trimester to postpartum examinations is depicted in the bottom panel (B).
cervical cells (Table II). No significant difference was observed between patients who reported having one or two sexual partners compared with those who reported more than two partners and the detection of HPV DNA in the patients cervical cells. Exfoliated cervical cells were obtained from 57 patients who remained compliant through their first postpartum examination (Fig. 3). DNA hybridization results indicate a decrease in the number of patients with detectable HPV infections between the third trimester of pregnancy and the first postpartum examination. The observed decrease was a combination of a small increase (10.2%) of HPV-positive patients between the third trimester and first postpartum examination coupled with a large decrease (45.6%) over this same time period. The total percentage of patients with positive
54
Rando et al.
results at the time of their postpartum examination was 17.5%; 66% of patients had positive results at one of the three assay points. Comment
Use of DNA hybridization (Southern blot) analysis as a diagnostic tool has confirmed the observations of several recent studies!' 2, 3 that routine cytologic examination alone will not detect all types of cervical HPV infections. However, neither will DNA hybridization analysis alone detect all HPV infections. A significant number of patients (eight of 110, or 7.2%) with cytologically abnormal features consistent with HPV infection did not have HPV detected by DNA hybridization analysis in the first trimester. The HPV infection level detected with DNA hybridization criteria alone for samples obtained during the first trimester (23 of 110 or 20.9%) was lower than the infection level of 29% reported for a population that was cytologically normal but referred to a colposcopy clinic,6 and much higher than the level of 10.5% reported for women with normal cytologic and col poscopic examinations in a sexually transmitted disease clinic. 7 This level was also lower than the level reported by Schneider et aLB for a general population of pregnant women (28%). The difference between the firsttrimester examination results reported here (Table I) and the 28% reported by Schneider et aLB could be a result of differences in the population size, the sensitivities of the different assay techniques used, the nature of the Pennsylvania Hospital clinic population, and the period during pregnancy at which samples were obtained. Of the 80 patients who were compliant until their third-trimester examination, 37 (46.2%) had positive results for cervical HPV infection and 42 (52.5%) had positive HPV DNA results in one or both of their analyzed samples (Fig. 2). It has been reported that detection of HPV DNA is greater in cervical samples obtained later in pregnancy than in those obtained in the early stages. s In addition to the large progression in the number of HPV DNA-positive patients from the first to the third trimester, there was a concomitant decrease of detectable HPV DNA levels (6.2%) in the study population. It is possible that experimental error accounted for some of the samples that changed categories over time, but the number of cases that did change indicates that a certain level of natural fluctuation in viral expression occurs in these patients. The observed decrease in detection of HPV DNA in the samples obtained from patients during their postpartum examinations (Fig. 3) reflects a difference in detectable levels of HPV DNA in the pregnant and nonpregnant cervical samples. We cannot determine from this study whether the difference in detectable
July 1989 Am J Obstet Gynecol
levels of HPV DNA was a result of a higher number of HPV -infected cells during pregnancy or a higher level of viral DNA per infected cell. The increase in detection during pregnancy could represent an increase in viral expression in response to host immunosuppression during pregnancy, or to changes in the hormonal melieu bathing the epithelial cells of the cervix during pregnancy, or both. It has been reported that regulatory elements on the viral genome are responsive to cellular factors 17 and to glucocorticoid hormones in vitro. 14 It also has been observed that J3-estradiol can stimulate the growth of experimental condylomas induced with HPV-II in nude mice (Krieder j, personal communication). Obvious fluctuations of the HPV status in maternal epithelial cells were noted in this study. The regression and progression patterns, although not well defined, appear to correlate with the natural immunologic and hormonal environment of the pregnant and postpartum patient. The actual percentage of the population infected with anogenital HPV s could be much higher than described in other studies of nonpregnant women. Further studies are needed to elucidate the natural history and true prevalence of anogenital HPV infections. We thank J. C. Cohn (Pennsylvania Hospital, Department of Orthopedics) and K. Kushner (Pennsylvania Hospital, Department of Pathology) for assistance with data analysis and presentation. We also thank R. Unger (Hospital of the University of Pennsylvania, Department of Obstetrics and Gynecology) for critically reviewing this manuscript, and Microprobe Corp. of Bothell, Wash., for providing HPV -33 and HPV -35 molecular probes. REFERENCES I. Centers for Disease Control. Condylomata acuminata-
United States. MMWR 1983;32:306-8. 2. Meisels A, Fortin R. Condylomatous lesions of the cervix and vagina. I. Acta Cytol 1976;20:505-9. 3. Gissmann L, Schwarz E. Persistence and expression of human papillomavirus DNA in genital cancer. In: Evered D, Clark C, eds. Papillomaviruses. Chichester: Wiley, 1986: 190-7. 4. Lorintz AT, Temple GF, Patterson JA, Jenson AB, Kurman RJ, Lancaster WD. Correlation of cellular atypia and human papillomavirus deoxyribonucleic acid sequences in exfoliated cells of the uterine cervix. Obstet Gynecol 1986;68:508-12. 5. De Villiers EM, Schneider A, Miklaw H, et al. Human papillomavirus infections in women with and without abnormal cervical cytology. Lancet 1987;2:703-6. 6. Burk RD, Kadish AS, Calderin S, Romney SL. Human papillomavirus infection of the cervix detected by cervicovaginallavage and molecular hybridization: correlation with biopsy results and Papanicolaou smear. AMJ OBSTET GYNECOL 1986;982-9. 7. Wickenden C, Malcolm ADB, Steele A, et al. Screening for wart virus infection in normal and abnormal cervices by DNA hybridisation of cervical scrapes. Lancet 1987; 1:65-7. 8. Schneider A, Hotz M, Gissmann L. Increased prevalence
Volume 161 Number 1
9.
10. II. 12. 13.
of human papilloma viruses in the lower genital tract of pregnant women. IntJ Cancer 1987;40:198-201. Sillman FH, Stanek A, Sed lis A, et al. The relationship between human papilloma virus and lower genital intraepithelial neoplasia in immunosuppressed women. AM J OBSTET GYNECOL 1984; 150:300-8. Halpert R, Frichter TG, Sed lis A, et al. Human papillomavirus and lower genital neoplasia in renal transplant patients. Obstet Gynecol 1986;68:251-8. Gary R, Jones R. Relationship between cervical condylomata, pregnancy and subclinical papilloma virus infection. J Reprod Med 1985;30:393-9. World Health Organization. Invasive cervical cancer and combined oral contraceptives. Br Med J 1985;290:961-5. Brinton LA. Current epidemiological studies-emerging hypothesis. In: Peto R, zur Hausen H, eds. Viral etiology of cervical cancer. New York: Cold Spring Harbor Laboratory Press, 1986: 17 -28.
HPV detection during pregnancy
14. Gloss B, Bernard HU, Seedorf K, Klock G. The upstream regulatory region of the human papillomavirus-16 contains an E2 protein-independent enhancer which is specific for cervical carcinoma cells and regulated by glucocorticoid hormones. EMBO J 1987;6:3735-43. 15. Rando RF, Grodd DE, ChirikjianJG, Lancaster WD. Isolation and characterization of a novel human papillomavirus type 6 DNA from an invasive vulvar carcinoma. J Virol 1986;57:353-6. 16. Maniatis T, Fritsch EF, SambrookJ. Molecular cloning: a laboratory manual. New York: Cold Spring Harbor Laboratory Press, 1982. 17. Guis D, Grossman S, Bedell MA, Laimons LA. Inducible and constitutive enhancer domains in the noncoding region of human papillomavirus type 18. J Virol 1988; 62:665-72.
Mechanism for human papillomavirus transmission at birth Thomas V. Sedlacek, MD: Steven Lindheim, MD: Catherine Eder, BA,b Lisa Hasty, MD: Mark Woodland, MD: Avi Ludomirsky, MD: and Robert F. Rando, PhDa.b Philadephia, Pennsylvania We attempted to investigate mechanisms, in addition to sexual contact, by which human papillomaviruses associated with anogenital tract lesions could be transmitted. Samples of exfoliated cervical cells were obtained from 45 pregnant women and were assayed by Southern blot hybridization analysis for the presence of human papillomavirus nucleic acids. Twenty-five of the 45 women had cells positive for human papillomavirus deoxyribonucleic acid. A neonatal nasopharyngeal aspirate was obtained at term and analyzed for the presence of human papillomavirus deoxyribonucleic acid. We documented the presence of human papillomavirus deoxyribonucleic acid in the oral pharyngeal cavity of the neonates in 15 of 45 nasopharyngeal samples analyzed. Amniotic fluid was obtained from 13 patients when their membranes were artificially ruptured. These samples were assayed for the presence of human papillomavirus deoxyribonucleic acid; two of the 13 amniotic fluid samples contained human papillomavirus deoxyribonucleic acid. The detection of human papillomavirus deoxyribonucleic acid in the oral cavity of neonates is indicative of a perinatal mechanism of viral transmission. The detection of human papillomavirus deoxyribonucleic acid in the amniotic fluid may suggest an in utero mechanism of transmission. However, problems encountered in collecting the amniotic fluid samples preclude us from definitive interpretation of these data. (AM J OBSTET GVNECOL 1989;161 :55-9.)
Key words: Human papillomavirus (HPV), deoxyribonucleic acid (DNA), uterine cervix, pregnancy, hybridization Several different human papillomavirus (HPV) types have been identified in benign'" and malignant4 . 5 lesions of the respiratory tract. One of the easiest HPVinduced lesions to identify is laryngeal papillomatosis. Benign laryngeal papillomatosis can lead to hoarseness From the Departments of Obstetrics and Gynecology' and Pathology, b Pennsylvania Hospital. ReceIved for publtcation July 22, 1988; revISed October 9, 1988; accepted January 10, 1989. Repnnt requests: Robert F. Rando, PhD, Department of Pathology, Preston Bldg., Rm. 758, Pennsylvania Hospital, 8th and Spruce St., Philadelphia, PA 19107.
and progressive enlargement that results in obstruction of the airway.6 In the majority of cases, papillomatous growths in the respiratory tract can be controlled with a limited number of ablative procedures. However, in some instances the papillomas in the larynx rapidly regrow after apparent complete surgical excision. In juvenile-onset laryngeal papillomatosis (clinical onset before puberty), 75% of cases are diagnosed before 5 years of age. 6 Several laboratories have established the causative relationship between human papillomaviruses (HPVs) [especially HPV type 6 (HPV-6) and HPV type 11 (HPV-11)] and respiratory papillo-
55
Key words: Human papilloma virus (HPV), deoxyribonucleic acid (DNA), uterine cervix, pregnancy, hybridization Surveillance of reported cases of clinically apparent anogenital human papillomavirus (HPV) infections indicated a steady rise in the incidence of new cases between 1966 to 1983. 1 In 1976 Meisels and Fortin 2 helped to heighten interest in human papillomaviruses when they reinterpreted the morphologic features of dysplasia of the uterine cervix as manifestations of HPV infections. Interest was further stimulated when HPV deoxyribonucleic acid (DNA) was identified in premalignant and malignant anogenitallesions. 3 To date more than 50 distinct types of HPVs have been described. At least 12 distinct types have been detected
From the Departments of Pathology" and Obstetrics and Gynecology,' Pennsylvama Hospital. Received for publication July 22, 1988; revised October 9, 1988; accepted January 10, 1989. Reprint requests: Robert F. Rando, PhD, Department of Pathology, Preston Bldg., Rm. 758, Pennsylvania Hospital, 8th and Spruce St., PhIladelphia, PA 19107.
50
in anogenital tract and oral cavity lesions. Several studies have reported that HPV type 16 and HPV type 18 are found in association with a majority of high-grade dysplastic lesions of the cervix, carcinoma in situ, and cervical carcinomas.' Two recent studies estimated that 10% to 20% of random populations that had cytologically "normal" cervices were in fact infected with HPV. 4 . 5 Burk et a\.6 detected HPV DNA in 29% of a population of women who were referred to a colposcopy clinic and had normal Papanicolaou smears. Wickenden et a\.7 detected HPV infections in 10.5% of a population from a sexually transmitted disease clinic who had cytologically normal cervices. It is interesting that a study conducted among pregnant women8 detected HPV DNA in 28% of the study population. The enhanced detection of viral infections has been reported in patients with impaired cell-mediated immunity.g· 10 It has also been noted that HPV infections and HPV -associated neoplasia of the anogenital tract
HPV detection during pregnancy
Volume 161 Number I
are more readily detected in immunodeficient patients than in the general population. 9 • 10 During pregnancy, a natural altered or depressed immunologic state has been suggested. s It has also been reported that clinically apparent HPV infections of the female genital tract seem to occur more frequently during pregnancy and regress rapidly after delivery. I I Schneider et al. B reported the detection of HPV DNA in exfoliated cervical cells in 28% of a pregnant population compared with 12.5% for nonpregnant control patients; a higher detection level occurred in samples obtained later in pregnancy. In addition to an immunodepressed condition, hormonal effects during pregnancy may playa role in the enhancement of viral infection or expression. Longterm users of oral contraceptives show an increased risk for the development of HPV-associated neoplasia. 12 13 There are also reports of hormonal stimulation of HPV-directed transcription in vitro 14 and proliferation of viral growth with [3-estradiol in vivo (Krieder J, personal communication). This article describes a prospective study designed to evaluate the basal level of cervical HPV infections in a population of pregnant women, and to determine whether there are fluctuations in the detectable levels of HPV DNA over the course of pregnancies through the first postpartum examination.
A
23
4
5
6
7
8
51
c
B
Material and methods
Sample collections. From April 1987 until August 1987 all new patients at Pennsylvania Hospital's obstetric clinic were recruited for this study. This population of women was young (mean age = 23.45 years), sexually active, and from a low socioeconomic group. At the time of the patient's first-trimester examination, a sample of exfoliated cervical cells was obtained and used for both cytologic and DNA hybridization analysis. Similar specimens of exfoliated cervical cells were obtained for DNA hybridization analysis during a thirdtrimester visit (28 to 34 weeks) and at the time of the patient'S first postpartum checkup (no later than 6 weeks after delivery). DNA extraction. Exfoliated cervical cells were suspended in phosphate-buffered saline solution, concentrated by low-speed centrifugation (1500 rpm), resuspended in 1 ml of cell lysis buffer (10 mmoll L Tris hydrochloride, pH 8.0, 0.6% sodium dodecyl sulfate, 10 mmollL ethylenediaminetetraacetic acid). Ribonuclease A (10 (J.g/ml) and proteinase K (50 (J.g/ml) were added to the cell lysis buffer and the samples were incubated at 37° C for 2 hours. Total cellular DNA was purified by extracting the lysed cellular material twice with an equal volume of phenollchloroform/isoamyl alcohol (25: 24 : 1). The extracted nucleic acids were then precipitated by adjusting the salt concentration to
Fig. 1. Detection of HPV DNA in samples obtained from exfoliated cervical cells by Southern blot hybridization. DNA was extracted, cleaved with restriction enzymes. electrophoresed, and transferred onto nylon filter membranes as described in "Material and methods." The procedure described was performed in triplicate. One of each of the three triplicate filters was hybridized with a combination of 32P-labeled molecular probes corresponding to HPV-16118 (A) or HPV-31/33/35 (data not shown) or HPV-6111 (B). DNA extracted from the samples analyzed in lane 3 hybridized better with the HPV-16118 probes. DNA extracted from the sample in lane 5 hybridized better with the HPV-6111 probes. Lane C contained 20 pg of an HPV -16 recombinant clone that was cleaved with Pst I.
0.3 mollL sodium acetate and adding 2.5 volumes of cold ( - 20° C) ethanol. The DNA was concentrated by centrifugation, washed twice with 70% ethanol, dried under vacuum, resuspended in 1 ml of a buffer containing 10 mmollL Tris hydrochloride (pH 7.5) and 1
52
Rando et al.
July 1989 Am J Obstet Gynecol
n=80
40 30
20
o
1st T r imester _
3rd T rimester
Init ia l
Total
~ New Positive 3rd T rimest.
Posit ive
Fig. 2. Graphic representation of HPV-positive cervical samples obtained from patients compliant through their third-trimester examination. The number of patients with positive results in the third trimester is a combination of patients who had positive results in the first trimester (minus those patients who had positive results at their first-trimester examination but negative results for HPV DNA in their cervical cells during their third-trimester examination) combined with the number of new patients with negative results in the first trimester but who had HPV detected during their third-trimester examination. The bar labeled total represents patients who were HPV DNA-positive in one or both of their examinations.
Table I. Correlation of Papanicolaou smear versus DNA hybridization analysis for the detection of HPV during first trimester of pregnancy Papamcolaou smear Positive*
DNA pos. DNA neg. TOTAL
6 8 14
I
Negativet
Total
17 79 96
87
23
110
*Papanicolaou-positive smears included those described as condylomatous, hyperkeratosis, parakeratosis, or squamous atypia. t Fourteen patients were classified as having benign atypia. Of those 14 patients, 5 had positive results for HPV DNA.
mmol/L ethylenediaminetetraacetic acid; the concentration was determined by spectrophotometric analysis at 260 nm. Hybridization analysis. Total DNA (10 to 15 J.l.g) extracted from patient samples was cleaved with the restriction endonuclease P st I as recommended by the manufacturer (New England Biolabs). The resulting DNA fragments were fractionated by electrophoresis in a O.B% agarose gel and then transferred to nylon filter membranes (Amersham Hybond-N).
DNA hybridization analysis was performed with modifications of the Southern blot technique as previously described. I5 Molecular probes corresponding to HPV types 6, 11, 16, IB, 31, 33, and 35 were amplified and purified using standard laboratory methods. I6 The HPV genomes were purified from flanking bacterial plasmid DNA and then radiolabeled with 32P-deoxycytidine triphosphate. IS. 16 Results
Exfoliated cervical cells obtained during the firsttrimester examination from 110 patients were examined by cytologic and DNA hybridization analysis for the presence of HPV infection. A comparison of the Papanicolaou smear and DNA hybridization analysis for the detection of HPV is described in Table I. Thirty-one of the 110 patients (28.2%) had positive results for the presence of HPV in their cervical cells. Of these 31 HPV-positive patients, eight (7.2 %) had positive results by Papanicolaou smears alone, six (5.4%) had positive results by both criteria, and 17 (15.4%) had only positive hybridization analysis results (Table J). A total of 23 of 110 (20.9%) had positive DNA hybridization analysis results. DNA hybridization analysis was performed in triplicate for the first-trimester patient samples. The tripli-
HPV detection during pregnancy
Volume 161 Number 1
53
A.
Table II. Correlation of sexual partners with detection of HPV DNA No. of sexual partners
1-2 n
I
2!3
%
n
I
%
HPV DNA pos.*
13/38
34.2
25/38
65.7
HPV DNA neg.
11/28
39.2
17/28
60.7
(n = 38) (n = 28)
x2
=
0.187
p> 0.05
*Any patient who had HPV DNA detected during their first-trimester, third-trimester, or both examinations were considered HPV DNA pos. cate hybridization reactions used mixtures of HPV types 16 and 18, HPV types 31, 33, and 35, or HPV types 6 and 11 for molecular probes (Fig. 1). Of the 23 patient samples that contained HPV DNA at the first-trimester examination, HPV-16118 was detected in 14 (60.8%) samples, HPV-31133/35 was detected in six (26%), and HPV-6111 was detected in one (4.3%) sample. In the remaining two samples (8.7%), HPV DNA was detected but not typable with the available molecular probes. Samples of exfoliated cervical cells were obtained from 80 of the original 110 patients during their third trimester-visit. Thirteen of these 80 patients (16.2%) had positive HPV DNA results during their firsttrimester examination. DNA hybridization analysis was performed on these samples. The results indicate an increase in the percentage of patients with positive results for HPV DNA in the third trimester, compared with the first-trimester results (Fig. 2). This increase was reflected in a moderate regression of detectable levels of HPV DNA in patients with positive results for HPV DNA during the first-trimester visit and negative results at the third-trimester visit (7.5%), coupled with a large increase of patients with negative results for HPV DNA during the first trimester and positive results at the third trimester (37.5%). The total number of patients with positive HPV DNA results during their third-trimester of pregnancy was 37 of80 (46.2%). The total number of patients who had positive results for HPV DNA at either or both examinations was 43 of 80 (53.7%). No significant correlation was observed when information concerning the smoking habits (obtained from 64 patients who remained compliant through their third-trimester examination) was compared with the presence of HPV DNA in the patients cervical cells (data not shown). In addition, information concerning the number of sexual partners (obtained from 66 patients compliant through the third trimester) was compared with the detection of HPV DNA in the patients
let Ttl.
~
Trl
Poet Pwtun
Tol8l
n=57
B.
i i
I o
-~
Fig. 3. Graphic representation of HPV DNA analysis for those patients who were compliant through their postpartum examination. The total percentage of patients positive at the three assay points is depicted in the top panel (A). The percentage of patients who changed with respect to the detection of HPV DNA in their cervical cells from first to third trimester or from third trimester to postpartum examinations is depicted in the bottom panel (B).
cervical cells (Table II). No significant difference was observed between patients who reported having one or two sexual partners compared with those who reported more than two partners and the detection of HPV DNA in the patients cervical cells. Exfoliated cervical cells were obtained from 57 patients who remained compliant through their first postpartum examination (Fig. 3). DNA hybridization results indicate a decrease in the number of patients with detectable HPV infections between the third trimester of pregnancy and the first postpartum examination. The observed decrease was a combination of a small increase (10.2%) of HPV-positive patients between the third trimester and first postpartum examination coupled with a large decrease (45.6%) over this same time period. The total percentage of patients with positive
54
Rando et al.
results at the time of their postpartum examination was 17.5%; 66% of patients had positive results at one of the three assay points. Comment
Use of DNA hybridization (Southern blot) analysis as a diagnostic tool has confirmed the observations of several recent studies!' 2, 3 that routine cytologic examination alone will not detect all types of cervical HPV infections. However, neither will DNA hybridization analysis alone detect all HPV infections. A significant number of patients (eight of 110, or 7.2%) with cytologically abnormal features consistent with HPV infection did not have HPV detected by DNA hybridization analysis in the first trimester. The HPV infection level detected with DNA hybridization criteria alone for samples obtained during the first trimester (23 of 110 or 20.9%) was lower than the infection level of 29% reported for a population that was cytologically normal but referred to a colposcopy clinic,6 and much higher than the level of 10.5% reported for women with normal cytologic and col poscopic examinations in a sexually transmitted disease clinic. 7 This level was also lower than the level reported by Schneider et aLB for a general population of pregnant women (28%). The difference between the firsttrimester examination results reported here (Table I) and the 28% reported by Schneider et aLB could be a result of differences in the population size, the sensitivities of the different assay techniques used, the nature of the Pennsylvania Hospital clinic population, and the period during pregnancy at which samples were obtained. Of the 80 patients who were compliant until their third-trimester examination, 37 (46.2%) had positive results for cervical HPV infection and 42 (52.5%) had positive HPV DNA results in one or both of their analyzed samples (Fig. 2). It has been reported that detection of HPV DNA is greater in cervical samples obtained later in pregnancy than in those obtained in the early stages. s In addition to the large progression in the number of HPV DNA-positive patients from the first to the third trimester, there was a concomitant decrease of detectable HPV DNA levels (6.2%) in the study population. It is possible that experimental error accounted for some of the samples that changed categories over time, but the number of cases that did change indicates that a certain level of natural fluctuation in viral expression occurs in these patients. The observed decrease in detection of HPV DNA in the samples obtained from patients during their postpartum examinations (Fig. 3) reflects a difference in detectable levels of HPV DNA in the pregnant and nonpregnant cervical samples. We cannot determine from this study whether the difference in detectable
July 1989 Am J Obstet Gynecol
levels of HPV DNA was a result of a higher number of HPV -infected cells during pregnancy or a higher level of viral DNA per infected cell. The increase in detection during pregnancy could represent an increase in viral expression in response to host immunosuppression during pregnancy, or to changes in the hormonal melieu bathing the epithelial cells of the cervix during pregnancy, or both. It has been reported that regulatory elements on the viral genome are responsive to cellular factors 17 and to glucocorticoid hormones in vitro. 14 It also has been observed that J3-estradiol can stimulate the growth of experimental condylomas induced with HPV-II in nude mice (Krieder j, personal communication). Obvious fluctuations of the HPV status in maternal epithelial cells were noted in this study. The regression and progression patterns, although not well defined, appear to correlate with the natural immunologic and hormonal environment of the pregnant and postpartum patient. The actual percentage of the population infected with anogenital HPV s could be much higher than described in other studies of nonpregnant women. Further studies are needed to elucidate the natural history and true prevalence of anogenital HPV infections. We thank J. C. Cohn (Pennsylvania Hospital, Department of Orthopedics) and K. Kushner (Pennsylvania Hospital, Department of Pathology) for assistance with data analysis and presentation. We also thank R. Unger (Hospital of the University of Pennsylvania, Department of Obstetrics and Gynecology) for critically reviewing this manuscript, and Microprobe Corp. of Bothell, Wash., for providing HPV -33 and HPV -35 molecular probes. REFERENCES I. Centers for Disease Control. Condylomata acuminata-
United States. MMWR 1983;32:306-8. 2. Meisels A, Fortin R. Condylomatous lesions of the cervix and vagina. I. Acta Cytol 1976;20:505-9. 3. Gissmann L, Schwarz E. Persistence and expression of human papillomavirus DNA in genital cancer. In: Evered D, Clark C, eds. Papillomaviruses. Chichester: Wiley, 1986: 190-7. 4. Lorintz AT, Temple GF, Patterson JA, Jenson AB, Kurman RJ, Lancaster WD. Correlation of cellular atypia and human papillomavirus deoxyribonucleic acid sequences in exfoliated cells of the uterine cervix. Obstet Gynecol 1986;68:508-12. 5. De Villiers EM, Schneider A, Miklaw H, et al. Human papillomavirus infections in women with and without abnormal cervical cytology. Lancet 1987;2:703-6. 6. Burk RD, Kadish AS, Calderin S, Romney SL. Human papillomavirus infection of the cervix detected by cervicovaginallavage and molecular hybridization: correlation with biopsy results and Papanicolaou smear. AMJ OBSTET GYNECOL 1986;982-9. 7. Wickenden C, Malcolm ADB, Steele A, et al. Screening for wart virus infection in normal and abnormal cervices by DNA hybridisation of cervical scrapes. Lancet 1987; 1:65-7. 8. Schneider A, Hotz M, Gissmann L. Increased prevalence
Volume 161 Number 1
9.
10. II. 12. 13.
of human papilloma viruses in the lower genital tract of pregnant women. IntJ Cancer 1987;40:198-201. Sillman FH, Stanek A, Sed lis A, et al. The relationship between human papilloma virus and lower genital intraepithelial neoplasia in immunosuppressed women. AM J OBSTET GYNECOL 1984; 150:300-8. Halpert R, Frichter TG, Sed lis A, et al. Human papillomavirus and lower genital neoplasia in renal transplant patients. Obstet Gynecol 1986;68:251-8. Gary R, Jones R. Relationship between cervical condylomata, pregnancy and subclinical papilloma virus infection. J Reprod Med 1985;30:393-9. World Health Organization. Invasive cervical cancer and combined oral contraceptives. Br Med J 1985;290:961-5. Brinton LA. Current epidemiological studies-emerging hypothesis. In: Peto R, zur Hausen H, eds. Viral etiology of cervical cancer. New York: Cold Spring Harbor Laboratory Press, 1986: 17 -28.
HPV detection during pregnancy
14. Gloss B, Bernard HU, Seedorf K, Klock G. The upstream regulatory region of the human papillomavirus-16 contains an E2 protein-independent enhancer which is specific for cervical carcinoma cells and regulated by glucocorticoid hormones. EMBO J 1987;6:3735-43. 15. Rando RF, Grodd DE, ChirikjianJG, Lancaster WD. Isolation and characterization of a novel human papillomavirus type 6 DNA from an invasive vulvar carcinoma. J Virol 1986;57:353-6. 16. Maniatis T, Fritsch EF, SambrookJ. Molecular cloning: a laboratory manual. New York: Cold Spring Harbor Laboratory Press, 1982. 17. Guis D, Grossman S, Bedell MA, Laimons LA. Inducible and constitutive enhancer domains in the noncoding region of human papillomavirus type 18. J Virol 1988; 62:665-72.
Mechanism for human papillomavirus transmission at birth Thomas V. Sedlacek, MD: Steven Lindheim, MD: Catherine Eder, BA,b Lisa Hasty, MD: Mark Woodland, MD: Avi Ludomirsky, MD: and Robert F. Rando, PhDa.b Philadephia, Pennsylvania We attempted to investigate mechanisms, in addition to sexual contact, by which human papillomaviruses associated with anogenital tract lesions could be transmitted. Samples of exfoliated cervical cells were obtained from 45 pregnant women and were assayed by Southern blot hybridization analysis for the presence of human papillomavirus nucleic acids. Twenty-five of the 45 women had cells positive for human papillomavirus deoxyribonucleic acid. A neonatal nasopharyngeal aspirate was obtained at term and analyzed for the presence of human papillomavirus deoxyribonucleic acid. We documented the presence of human papillomavirus deoxyribonucleic acid in the oral pharyngeal cavity of the neonates in 15 of 45 nasopharyngeal samples analyzed. Amniotic fluid was obtained from 13 patients when their membranes were artificially ruptured. These samples were assayed for the presence of human papillomavirus deoxyribonucleic acid; two of the 13 amniotic fluid samples contained human papillomavirus deoxyribonucleic acid. The detection of human papillomavirus deoxyribonucleic acid in the oral cavity of neonates is indicative of a perinatal mechanism of viral transmission. The detection of human papillomavirus deoxyribonucleic acid in the amniotic fluid may suggest an in utero mechanism of transmission. However, problems encountered in collecting the amniotic fluid samples preclude us from definitive interpretation of these data. (AM J OBSTET GVNECOL 1989;161 :55-9.)
Key words: Human papillomavirus (HPV), deoxyribonucleic acid (DNA), uterine cervix, pregnancy, hybridization Several different human papillomavirus (HPV) types have been identified in benign'" and malignant4 . 5 lesions of the respiratory tract. One of the easiest HPVinduced lesions to identify is laryngeal papillomatosis. Benign laryngeal papillomatosis can lead to hoarseness From the Departments of Obstetrics and Gynecology' and Pathology, b Pennsylvania Hospital. ReceIved for publtcation July 22, 1988; revISed October 9, 1988; accepted January 10, 1989. Repnnt requests: Robert F. Rando, PhD, Department of Pathology, Preston Bldg., Rm. 758, Pennsylvania Hospital, 8th and Spruce St., Philadelphia, PA 19107.
and progressive enlargement that results in obstruction of the airway.6 In the majority of cases, papillomatous growths in the respiratory tract can be controlled with a limited number of ablative procedures. However, in some instances the papillomas in the larynx rapidly regrow after apparent complete surgical excision. In juvenile-onset laryngeal papillomatosis (clinical onset before puberty), 75% of cases are diagnosed before 5 years of age. 6 Several laboratories have established the causative relationship between human papillomaviruses (HPVs) [especially HPV type 6 (HPV-6) and HPV type 11 (HPV-11)] and respiratory papillo-
55