Evaluation of Sexually Abused and Nonabused Young Girls for Intravaginal Human Papillomavirus Infection Laura T.
Gutman, MD; Karen St. Claire, MD; Marcia E. Herman-Giddens, PA, MPH; William W. Johnston, MD; William C. Phelps, PhD
\s=b\ Objective.\p=m-\The
objective of this study was to compare the prevalence of intravaginal human papillomavirus\x=req-\ associated disease in two groups of girls to develop information regarding the means of transmission of anal-genital human papillomavirus disease. Design.\p=m-\Apair of parallel studies of prevalences of human papillomavirus infections in two populations of prospectively enrolled girls. Patients.\p=m-\Index patients consisted of 15 consecutive girls aged 11 years or younger who were confirmed to have been sexually abused, had signs or symptoms of vaginal disease, and required generalized anesthesia for evaluation. Selection of nonabused control patients was based on negative findings from screening evaluations and physical examinations. Main Outcome Measures.\p=m-\Prevalencesof cervical\x=req-\ vaginal human papillomavirus infections in the two populations were compared. Vaginal wash samples from index and control patients were assayed for human papillomavirus 1, 2, 4, 6, 11, and 16 by reverse-blot and Southern substantial body of epidemiologie data indicates that human papillomavirus (HPV)-related diseases of the adult genital tract are transmitted primarily through sexual contact.1 In contrast, there have been conflicting reports about the means of transmission of anal-genital warts to postneonatal children. The major area of uncer¬ tainty is whether the primary mode of transmission of genital warts to children is similar to that in adults, ie, sexual contact. This question has been studied primarily through the assessment of children who have been iden¬ tified as having genital warts and were evaluated for pos¬ sible sexual abuse. However, the methods of evaluating
A
Accepted for publication
March 25, 1992. From the Departments of Pediatrics (Drs Gutman and St. Claire and Ms Herman-Giddens) and Pathology (Dr Johnston), Duke University Medical Center, Durham, NC, and the Burroughs-Wellcome Research Laboratories, Research Triangle Park, NC (Drs Gutman and Phelps). Presented in part at the 1991 Papillomavirus Workshop, Seattle, Wash, July 23,1991, and at the joint plenary session for the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy and the Infectious Diseases Society of America, Chicago, Ill, September 30, 1991. Reprint requests to Box 3971, Duke University Medical Center, Durham, NC 27710 (Dr Gutman).
Downloaded From: by a Midwestern University User on 01/22/2019
transfer hybridization methods. Papanicolaou smears were examined from index patients. Results.\p=m-\Vaginalwash samples from five (33%) of 15 index patients were positive for human papillomavirus 6, 11, or 16, compared with none of 17 controls. The presence or absence of external anal-genital warts was not correlated with results from the assay of intravaginal samples. Blinded readings of vaginal exfoliative cytologic findings of the index patients showed koilocytosis, atypia, or inflammatory reactions in four of five human papillomavirus\x=req-\ positive girls, and normal cytologic findings in one human
papillomavirus-positive girl. Conclusion.\p=m-\Thesefindings support
other studies that indicate that sexual contact is a major route in the transmission of anal-genital human papillomavirus-related disease in children. Evaluation of intravaginal specimens was
required to identify human papillomavirus-infected girls since the results of the wash samples were not correlated with the presence or absence of external anal-genital warts. (AJDC. 1992;146:694-699)
for sexual abuse have varied greatly, and, thus, reported rates of confirmed child sexual abuse in children with anal-genital warts have ranged from 0% to 91%.2j3 Evalu¬ ation of abused girls for intravaginal HPV-related infec¬ tion has not been recommended, and no data on the prevalence of intravaginal HPV infection in children have been published. The intravaginal microbiologie findings of nonabused young girls have not been well characterized because of the social and technical limitations in acquiring samples. Prevalence studies in defined populations represent a means of investigating the relationship between sexual abuse and anal-genital HPV infection in children. How¬ ever, because of the limitations on sample collection, studies to identify intravaginal HPV infection in either sexually abused girls or nonabused girls have not been
reported.
The present study documents the prevalence of a selected number of HPV types from vaginal wash samples of two groups of young girls whose status regarding sex¬ ual abuse was fully characterized. The first group com¬ prised girls who had been victims of invasive child sexual abuse, while girls in the second group had not been sex¬ ually abused. This study differs in two respects from oth¬ ers that have examined the means of transmission of anal-genital HPV disease in children. In this study, two
of girls were compared, and the study focused on the diagnosis of intravaginal HPV disease as manifested by HPV, DNA, and cytologie findings, rather than on external genital warts.
populations
SUBJECTS AND METHODS Selection of Index Patients
The Duke Child Protection Team examines approximately 250 girls per year for suspected child sexual abuse, and abuse is con¬ firmed in about 40% of examined girls. The index population comprised 15 girls aged 11 years or younger who were Tanner stage 1 or 2 in sexual development and who had been referred to the Duke Child Protection Team for evaluation of suspected child sexual abuse. The index population included all girls from a 41/4-year period beginning in 1986 who fulfilled the following criteria: (1) The Child Protection Team completed the evaluation for suspected child sexual abuse. (2) The girls had a history of symptomatic vaginal disease (ie, discharge, bleeding, or pain) that required a medical examination. (3) Generalized anesthesia was used because the girls were unable to tolerate the examina¬ tion while awake. (4) The girls were confirmed to have been abused by the Child Protection Team. Examinations using anes¬ thesia were indicated when there were persistent medical signs or symptoms and the child could not tolerate the examination while awake. Index patients were referred for an examination with anesthesia for evaluation of acute injury and/or recurrent or chronic symptomatic vaginal disease manifested as bleeding, discharge, pain, or lesions (Table 1).
Table 1.—Genital Signs and Symptoms of Index Patients (n 15) =
No. of Patients
Presenting complaints
Vaginal discharge Chronic Acute
or
recurrent
Vaginal bleeding
Labial abrasions Perivaginal lesions
Perivaginal condyloma acuminatum Perivaginal molluscum contagiosum
Perianal lesions
Additional findings during anesthesia administration Scars, tears, notches, or distortions of hymen
Hymenal opening transverse diameter 3=8
mm
Changing external examination results with serial evaluations Abnormal external genital examination results
7 2 5 3 3 1 1
8
10 2
15
Table Characteristic
Selection of Control Patients Control patients were recruited from the pediatrie clinics at the Duke Eye Center, Durham, NC. They were girls aged 11 years or younger and Tanner stage 1 in sexual development who had been scheduled to receive ophthalmic examinations or proce¬ dures while receiving generalized anesthesia. Control patients underwent a screening assessment for sexual abuse because child sexual abuse is a common problem and nonabuse of this population could not be assumed." The median age of the index population was 59 months, with a range of 23 to 131 months. The median age of the control pop¬ ulation was 34 months, with a range of 3 to 68 months (Table 2). The control population was significantly younger than the index population (P=.006; two-sample Wilcoxon's rank sum test). Eight (53%) of the index children were black and seven (47%) were white. Of 17 controls, six (35%) were black and 11 (65%) were white. The groups did not differ significantly in the percentage of black children (P= .47; Fisher's Exact Test).
Evaluation for Child Sexual Abuse Methods used in evaluating the index patients for sexual abuse have been described7 and follow currently recommended proce¬ dures, including parent interviews and child interviews and ex¬ aminations.8 Five groups of data were obtained: (1) a review of the behavioral history of the child, (2) a review of the medical history of the child, (3) multiple diagnostic interviews with the child, (4) a forensic medical examination, and (5) assessment for other sexually transmitted diseases. In the present study, "dis¬ closure" refers to the detailed verbalization by the child or other witness of a specific sexually abusive act. Perpetrators were "identified" if the child or a caretaker had provided specific de¬ scription of the person as an assailant. Child sexual abuse was confirmed on the basis of any of the following: (1) the child or a witness disclosed the abuse; (2) gen¬ ital gonorrhea or acquired syphilis was diagnosed; (3) the gen¬ ital examination results were abnormal, with findings clearly in¬ dicative of abuse as outlined at the 1985 National Child Sexual Abuse Summit meeting and subsequently updated9"12; (4) serial genital examinations showed changing abnormalities that were indicative of active sexual abuse; or (5) there was acknowledg¬ ment by the perpetrator. The following methods were used in evaluating control patients for concerns of sexual abuse. At an eye center clinic visit, the study and procedure were described to the parents and child by a member of the Child Protection Team and informed consent was obtained. An intake questionnaire was administered by one of us (K.S.C.) that included questions about each of the follow¬ ing medical areas: chronic illness and previous medical inter¬ ventions (five questions asked); medications (two questions); genitourinary signs and symptoms (nine questions); gastrointes¬ tinal signs and symptoms (three questions); vaginal signs and symptoms (five questions); and behavioral disturbances (10 questions). Open questions were asked regarding any concerns the parents may have had concerning physical or sexual abuse. The physical examination of the external genitalia was per¬ formed at the time of administering anesthesia. The evaluation did not include a separate diagnostic interview with the child.
2.—Study Populations
Index Patients (n = 15)
(range)
Controls (n = 17)
59 (23-131) 34 (3-68) .006 (Wilcoxon's rank sum test) age, mo Race, No. (%) of patients .47 (Fisher's Exact Test) Black 8(53) 6(35) 11 (65) White 7(47) Assays positive for HPV6, 11, or 16, 5 (33)* .015 (Fisher's Exact Test) No. (%) of patients *HPV indicates human papillomavirus infection. The HPV types identified included HPV6 in two patients and HPV11, HPV 6/11, and HPV16 in one patient each.
Median
Downloaded From: by a Midwestern University User on 01/22/2019
The
study protocol required that parents be informed of find¬
ings that raised concerns about possible child sexual abuse and
the parents were aware that a report of suspected sexual abuse would be made should findings strongly indicate abuse. The study was approved by the Duke Institutional Review Board and written parental consent was obtained.
Physical
Examination and Collection of Specimens
The external genital examination for both groups of children included inspection of the labia and introitai area for redness, discharge, lesions, or injuries; identification of hymenal type and contour; and measurement of the horizontal diameter of the hy¬ menal opening.13 Samples obtained included vaginal swabs and washes. Intravaginal samples from index patients were obtained during direct visualization of the vagina and cervix, and cervicalvaginal Papanicolaou smears were obtained. Vaginal wash ma¬ terial was collected through a flexible, small-bore plastic cathe¬ ter that was fitted to the end of a 10-mL syringe. The catheter was inserted through the hymenal opening, and 10 mL of Dulbecco's phosphate-buffered saline was washed in and out of the vagina two or three times. Direct intravaginal visualization was not at¬ tempted in the control children and therefore Papanicolaou smears were not obtained.
Virologie Diagnosis Plasmid-cloned isolates of HPV1, 2, and 4 have been previ¬ ously described.14 The HPV6 probe consists of the Fnu4Hl (b. 509) fragment to the Mstll (b. 4520).15 The HPV16 consists of the full genome from a pUC18 clone derived from the proto¬ type.16 The HPVll/pUC18 was cloned from a patient specimen
(L.T.G, unpublished data, 1991).
The HPV DNA fragments were isolated electrophoretically and labeled with 32P using random oligonucleotide primers for use as
probes.
HPV DNA Analysis Cellular DNA was isolated from the pellet of vaginal wash material by overnight digestion, phenol extraction, and alcohol precipitation. The extracted DNA was digested with RNAse, ex¬ tracted with phenol, and alcohol precipitated a second time. The concentration of DNA was determined with a spectrophotometer. The range of DNA extracted from samples from index patients was 630 ng to 2.25 µg and that from control patients was 315 to 630 ng. A minimum of 300 ng was required to complete the study protocol. Reverse Southern
Hybridization
Reverse Southern hybridization for HPV1, 2, 4, 6, 11, and 16 has been previously described.17-18 Briefly, 200 ng of each of HPV types 1, 2, 4, 6,11, and 16 was immobilized on a nylon filter. The probe consisted of 50 to 100 ng of patient DNA that was labeled with 32P using random nucleotide primers. Hybridization was performed at 65°C. The sensitivity of this assay was less than or equal to 1 pg of HPV DNA in 1 µg of cellular DNA.
Southern Transfer
Hybridization
Two hundred nanograms of cellular DNA from each index and control patient was digested with the restriction endonuclease enzymes, BamHI or Psf I, and analyzed with Southern transfer
hybridization procedures.19 Samples were separated by electro¬ phoresis, transferred to nylon filter, and probed with 32P-labeled HPV6/11 and HPV16. Hybridization produced a readily visible signal with less than 0.2 pg of HPV DNA per 200 ng of cellular DNA.
Cervical-Vaginal Cytologie Findings Cervical-vaginal swabs were collected from index patients for
cytologie examination. The preparations were interpreted by a single observer (W.W.J.), who was blinded to the results of the HPV assays. Descriptive observations were made and findings were categorized according to the Bethesda System.20 Koilocytes are mature squamous cells characterized by a large perinuclear cavity writh sharp borders. The cellular cytoplasm is dense and the nucleus shows varying degrees of degeneration. These cells are commonly present in HPV-infected lesions. RESULTS
Control and Index Patients Fifteen index patients met the criteria for entry into the study. Thirteen girls (87%) gave a full disclosure of sexual abuse and two (13%) gave a partial disclosure of sexual abuse. Thirteen girls had abnormal examination findings of the introitai area or hymen that were consistent with healing or healed trauma from sexual abuse (Table 1). The 15 index patients came from 15 separate households and no index patient was related to any other patient. Seventeen control patients underwent a behavioral, medical, and physical screening evaluation for child sex¬ ual abuse and had negative findings. A vaginal wash sample was obtained from each of these girls for evalua¬ tion for HPV. Two potential control patients were omit¬ ted from the study because of concerns of possible abuse that were identified during the screening evaluation. Both children disclosed their abuse and identified the perpe¬ trators.
Perpetrators (33%) of the index children had been assaulted by single perpetrators, four (27%) had been assaulted by multiple perpetrators, three (20%) disclosed abuse but did not identify a specific perpetrator(s), and three (20%) identified one perpetrator with at least one additional perpetrator suspected. All identified perpetrators were Five
either related to the child or had access to the child because of a relationship with another member of the
family.
The types of abuse that were disclosed by the index pa¬ are listed in Table 3. Eleven children (73%) de¬ scribed penetrating penile-vaginal or penile-anal abuse, and two (13%) described digital abuse. Two children (13%) were unable to give details of their assault. tients
Table
3.—Type of Sexual Abuse Disclosed (n 15)* =
Site of Contactt I-1
Assailant
No. of Index
Child
Children
Penis
Vagina
9
Penis
Anus
4
Penis
Mouth
2
Finger
Vagina Vagina multiple types
2
Mouth
1
*Five children disclosed of abuse. tThe site of contact was unknown for two children.
Downloaded From: by a Midwestern University User on 01/22/2019
HPV DNA
Assay Vaginal samples from five (33%) of the 15 index patients were positive for HPV using Southern transfer hybridization. Four (80%) of these five were HPV types 6 wash
11, and one was HPV type 16 (Table 2). The results of the reverse-blot assays corroborated the findings of the Southern transfer hybridization assays (Figs 1 and 2). La¬ beled DNA from index patients did not hybridize with HPV1, 2, or 4 on reverse blots. The prevalences of HPV or
Fig 1.—Restriction endonuclease digestion results from a cervicalvaginal wash specimen. BamHI and Pst / demonstrated fragments
characteristic of HPV6a.
DNA from the two study populations, 0 of 17 controls vs five of 15 index cases, differed significantly (P= .015) us¬ ing Fisher's Exact Test. The 95% confidence interval for HPV positivity in the index population was 9% to 57%.
External Genital Warts The relationship between the presence or absence of external genital lesions and intravaginal infection is shown in Table 4. Four of the index patients had clinically evident external anal-genital warts, one in the perianal area and three in the perivaginal area. Two girls also had an intravaginal specimen that was HPV positive and two did not. Three (60%) of the five girls who had HPVpositive intravaginal specimens did not have identifiable external genital warts. Biopsy material of external lesions from one patient was examined for HPV DNA, and both wash and biopsy results showed HPV6a.
Cytologie Findings Of the five girls with HPV-positive wash specimens, only one had cytologie findings that included koilocytosis (Fig 3). Cytologie findings from a second HPV-positive child showed a severe acute inflammatory reaction in which the parabasal cells were surrounded by a dense aggregation of polymorphonuclear leukocytes. The nuclei of parabasal cells showed atypia with darkening of the nuclear chromatin, thickening of the nuclear envelope, and increased nuclear to cytoplasmic ratio. For the third child, cytologie findings were restricted to chronic inflam¬ matory reactions showing lymphocytic and monocytic cellular predominance and relatively fewer parabasal cells. Cytologie findings for the fourth child included a severe acute inflammatory reaction with a predominance Downloaded From: by a Midwestern University User on 01/22/2019
Fig 2.—Restriction endonuclease digestion results from a cervicalvaginal wash specimen. BamHI and Pst / demonstrated fragments characteristic of HPV11.
Table 4.—Relationship Between Presence or Absence of External Genital Lesions and Positive or Negative Assay Results From Intravaginal Wash Samples For Index Patients With Human Papillomavirus (HPV) Infection
Vaginal Wash Assay Results, No. of Patients I
Clinical Evaluations External
I
HPV DNA
HPV DNA
Positive
Negative
Total
anogenital warts
present
Perianal
Perivaginal External anogenital warts
absent Total No. of index
patients
0 2
1 1 13
3
8
11
10
15
5
of polymorphonuclear leukocytes and parabasal cells that showed mild perinuclear clearing but not overt koilocytosis. The fifth child had normal cytologie findings, with¬ out evidence of inflammatory reaction, atypia, or koilo-
cytosis.
Cytologie examinations from the 10 index girls who had
negative HPV DNA vaginal washes showed neither cel¬ lular reactions suggestive of HPV infection nor acute or
chronic
inflammatory reactions.
COMMENT The index patients represent a subpopulation of sexu¬ ally abused girls who were particularly symptomatic from the trauma that they had experienced and whose abuse
other
studies,
more than 60% of adults in established re¬ with partners who have genital warts were concordant for the presence of lesions.24-25 Similar data for child victims of sexual abuse are not available because, in these cases, perpetrators identified by children are rarely charged with the offense. This effectively prohibits any evaluation for sexually transmitted diseases. Other evidence that HPV infection is sexually trans¬ mitted in adults includes data showing a relationship be¬ tween prevalence of HPV infection and numbers of sex¬ ual partners. In adults, these two factors increase proportionally.26-27 Similarly for children, a recent study has shown a significant association between the presence of multiple, other sexually associated organisms in chil¬ dren who have genital warts and who have been abused
lationships
by multiple perpetrators.3
Fig 3.—Papanicolaou
smear
from human
papillomavirus-positive
child, demonstrating koilocytic findings in squamous cell. The nu¬ cleus is marginated and condensed, the cell is enlarged, anda halo surrounds the nucleus
(original magnification 320).
probably unusually severe. At least four (27%) of the 15 patients had been abused by multiple perpetrators. All had signs or symptoms of in tra vaginal disease. Most of the types of disclosed abuse that they experienced were invasive. The 33% rate of recovery of HPV DNA from pa¬ tient samples is similar to the recovery rate of HPV from a sexually at-risk adult population that was studied using was
dot-blot hybridization method.21 Using the Southern transfer hybridization method, the recovery rate of HPV DNA from index cases was slightly higher than the 15% rate of recovery from another at-risk adult population.22 Results from the HPV assays of the index children therefore closely resemble results from adults who are highly sexually active. Even so, the 33% prevalence rate of HPV from the index population is ex¬ pected to be an underrepresentation of the true preva¬ lence of this infection in children who have been severely abused because sample sizes were small, assays for only six HPV types were performed, and only a single sample was taken from each child. In adults, the primary means of transmission of HPVrelated genital disease is through sexual contact. Evidence of sexual transmission in adults comes from studies of couples; 90% of new sexual contacts of persons with gen¬ ital warts result in the development of genital lesions.23 In a
Downloaded From: by a Midwestern University User on 01/22/2019
Medical literature on children with anal-genital papil¬ lomavirus disease contains conflicting information about sexual contact as the means of transmission of this infec¬ tion to children. Unfortunately, many of these reports and studies were done prior to the development of currently established methods for evaluating child sexual abuse. Most of these earlier reports did not use a sexual abuse evaluation process, which must include a diagnostic interview of the child and parent by a skilled interviewer and a forensic genital examination by a qualified exam¬ iner. Rather, most of the earlier reports relied on the child to volunteer the disclosure and categorized nondisclosing children who had not received a thorough evaluation as nonabused. Several of the more recent studies have also omitted any description of the evaluation methods and findings that they used to diagnose sexual abuse, or they have catego¬ rized children as nonabused when the child's status was inadequately documented or unknown.2'2829 Conse¬ quently, reported associations of sexual abuse with HPV transmission to children are likely to be inaccurately low. Despite the limitations in their diagnosis of sexual abuse, however, these earlier reports document that at least 50% of children with genital warts disclosed that they had been sexually abused.30-31 The only published study using cur¬ rent diagnostic methods for child sexual abuse in girls having genital warts showed that 10 (91%) of 11 had been sexually abused. In that study disclosure of abuse was provided in 71% of cases, and in a similar study a high disclosure rate was also recorded.3,32 Sexual abuse was confirmed in four of the five children aged 3 years or younger.3 This strongly suggests that even in children who are preverbal or only marginally verbal the presence of genital tract lesions due to HPV is frequently associated with a confirmed diagnosis of sexual abuse. The current study used the Southern transfer hybrid¬ ization assay for several reasons. First, previous studies using this method have demonstrated that most genital papillomas of children are caused by HPV6 or 11 and only infrequently by HPV2,16, or 18.33-34 Other types are rarely or never identified in genital lesions of children. Conse¬ quently, our assay was designed to examine specimens for a limited number of HPV types. Second, because nonamplified DNA was measured, quantitative informa¬ tion was provided, and identification of the type was precise. Third, the method is the criterion standard. In the present study, racial distribution of the index and control children did not differ significantly. The mean age of the control population, however, was significantly
younger than that of the index population. The control patients were therefore closer in time to possible neona¬ tal transmission of HPV, and they were also more likely to have difficulty in evaluations for sexual abuse due to their preverbal or poorly verbal developmental stages. These factors resulted in a conservative bias for the study. Of particular concern for the HPV-positive children is that the long-term consequences of early exposure to this infection are unknown. Of note are reports from several medical centers about the increased incidence of genitaltract papillomavirus infections in female adolescents. Be¬ nign, dysplastic, and malignant HPV-associated lesions are also now being reported in this age range.35"39 Because the length of time between HPV infection and progression to dysplastic and malignant stages may be quite long, some of these female adolescents may have acquired their infection prior to the legal age of consent for intercourse and through abusive sexual contact as a child. The present data indicated that our index population of abused girls resembles populations of sexually active adults in the prevalence of intravaginal HPV DNA, in the presence of anal-genital lesions, and in cytologie findings. These children have acquired their internal genital HPV disease at very immature ages. Data regarding their prognosis and appropriate treatment are needed. Study findings also support other data that children presenting with anal-genital papillomavirus disease are at significant risk of having been abused and should be evaluated for possible sexual abuse and other sexually transmitted in¬ fections. These studies were supported in part by a grant from the Mellon Foundation. The authors gratefully acknowledge the active participation of Robert Machemer, MD, and Edward Buckley, MD, and the faculty and personnel of the Duke Eye Center, Durham, NC. Gerald Hajian, PhD, Burroughs-Wellcome Co, Research Triangle Park, NC, pro¬ vided expert assistance with data analysis and statistical evaluation of the results, and David Barry, MD, provided editorial review. Julia Barnes, Burroughs-Wellcome Co, provided technical advice. Nancy Berson, MSW, Carmenza Salgado, PA, and Jeanne Niemeyer, MSW, Duke University Medical Center, conducted skillful interviews and examinations of index patients, and Gale Hill, PhD, Duke Univer¬
sity, provided microbiologie support.
References 1. Koutsky LA, Galloway DA, Holmes KK. Epidemiology of genital human papillomavirus infection. Epidemiol Rev. 1988;10:122-163. 2. Obalek S, Jablonska S, Favre M, Walczak L, Orth C. Condylomata acuminata in children: frequent association with human papillomaviruses responsible for cutaneous warts. J Am Acad Dermatol.
1990;23:205-213.
3. Herman-Giddens ME, Gutman LT, Berson NL, Duke Child Protection Team. Association of coexisting vaginal infections and multiple abusers in female children with genital warts. Sex Transm Dis.
1988;15:63-67.
4. Russell DEH. The incidence and
of intrafamilial and extrafamilial sexual abuse of female children. Child Abuse Negl.
prevalence
1983;7:133-146. 5. Moore KA, Nord CW, Peterson JL. Nonvoluntary sexual activity among adolescents. Fam Plann Perspect. 1989;21:110-114. 6. Siegel JM, Sorenson SB, Golding JM, Burnam MA, Stein JA. The
prevalence of childhood sexual assault: the Los Angeles Epidemiologic Catchment Area Project. Am J Epidemiol. 1987;126:1141-1153. 7. Gutman LT, St Clair KK, Weedy C, et al. Human immunodeficiency virus transmission by child sexual abuse. AJDC. 1991;145:137-141. 8. Krugman RD, Bays JA, Chadwick DL, Levitt CJ, McHugh MT, Whitworth JM. Guidelines for the evaluation of sexual abuse of children. Pediatrics. 1991;87:254-260. 9. Tipton AC. Child sexual abuse: physical examination techniques and interpretation of findings. Adolesc Pediatr Gynecol. 1989;2:10-25. 10. White ST, Ingram DL, Lyna PR. Vaginal introital diameter in the evaluation of sexual abuse. Child Abuse Negl. 1989;13:217-224.
Downloaded From: by a Midwestern University User on 01/22/2019
11. Paradise JE. The medical evaluation of the sexually abused child. Pediatr Clin North Am. 1990;37:839-862. 12. Muram D. Child sexual abuse-genital tract findings in prepubertal girls, I: the unaided medical examination. Am J Obstet Gynecol.
1989;60:328-333. 13. McCann J, Voris J, Simon M, Wells R. Comparison of genital examination techniques in prepubertal girls. Pediatrics. 1990;85:182-187. 14. Heilman C, Law M-F, Israel MA, Howley PM. Cloning of human
papilloma virus genomic DNAs and analysis of homologous polynucleotide sequences. J Virol. 1980;36:395-407. 15. Schwarz E, Durst M, Demankowski C, et al. DNA sequence and genome organization of genital human papillomavirus type 6b. EMBO
J. 1983;2:2341-2348. 16. Durst M, Gissman L,
Ikenberg H, Zur Hausen H. A papillomaviDNA from a cervical carcinoma and its prevalence in cancer biopsy samples from different geographic regions. Proc Natl Acad Sci U S A. rus
1983;80:3812-3815. 17. deVilliers E-M, Schneider A, Gross G, zur Hausen H. Analysis of
benign and malignant urogenital tumors for human papillomavirus infection by labelling cellular DNA. Med Microbiol Immunol. 1986; 174:281-286. 18. Webb
DH, Rogers RE, Fife KH. A one-step method for detecting and typing human papillomavirus DNA in cervical scrape specimens from women with cervical dysplasia. J Infect Dis. 1987;156:912-919. 19. Sambrook J, Fritsch EF, Maniatis T, eds. Molecular Cloning. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1989. 20. Solomon D. The 1988 Bethesda system for reporting cervical/ vaginal cytologic diagnoses. Anal Quant Cytol Histol. 1989;11:291\x=req-\ 297. 21. Burmer GC, Parker JD, Bates J, East
K, Kulander BG. Comparative analysis of human papillomavirus detection by polymerase chain reaction and Virapap/Viratype kits. Am J Clin Pathol. 1990;94:554-560. 22. Kiviat NB, Koutsky LA, Critchlow CW, et al. Comparison of Southern transfer hybridization and dot filter hybridization for detection of cervical human papillomavirus infection with types 6,11,16, 18,
31, 33 and 35. Am J Clin Pathol. 1990;94:561-565. 23. Barrett TJ, Silbar JD. Genital warts: a venereal disease. JAMA. 1954;154:333-334. 24. Oriel JD. Maternal history of genital warts. Br J Vener Dis. 1971;47:1-13. 25. Barrasso R, DeBrux J, Croissant O, Orth G. High prevalence of
papillomavirus-associated penile intraepithelial neoplasia
in sexual
partners of women with cervical intraepithelial neoplasia. New Engl J Med. 1987;317:916-923. 26. Kiviat NB, Koutsky LA, Paavonen JA, et al. Prevalence of genital
papillomavirus infection in women attending a college student health clinic or a sexually transmitted disease clinic. J Infect Dis. 1989;159:293\x=req-\ 302. 27. Syrjanen K, Vayrynen M, Castren O, et al. Sexual behavior of women with human papillomavirus (HPV) lesions by the uterine cervix. Br J Vener Dis. 1984;60:243-248. 28. Cohen BA, Honig P, Androphy E. Anogenital warts in children: clinical and virologic evaluation for sexual abuse. Arch Dermatol. 1990;126:1575-1580. 29. Padel AF, Venning VA, Evans MF, Quantrill AM,
Fleming KA. Hu-
papillomaviruses in anogenital warts in children: typing by in situ hybridization. Br Med J. 1990;300:1491-1494. 30. DeJong AR, Weiss JC, Brent RL. Condyloma acuminata in chilman
dren. AJDC. 1982;136:704-706. 31. Shelton TB, Jerkins GR, Noe HN. Condylomata acuminata in the pediatric patient. J Urol. 1986;135:548-549. 32. Gilles EE, Hitchcock BS, Elvik SL. Human papilloma virus in children. Ninth National Conference on Child Abuse and Neglect; September 14-17, 1991; Denver, Colo. Abstract B.1, 55. 33. Rock B, Naghashfar E, Barnett N, Buscema J, Woodruff SD, Shah K. Genital tract papillomavirus infection in children. Arch Dermatol.
1986;122:1129-1132. 34. Gibson PE, Gardner SD, Best SJ. Human papillomavirus types in anogenital warts of children. J Med Virol. 1990;30:142-145. 35. Martinez J, Smith R, Farmer M, et al. High prevalence of genital tract papillomavirus infection in female adolescents. Pediatrics. 1988;
82:604-608. 36. Rosenfeld WD, Vermund SH, Wentz ST, Burk RD. High prevalence rate of human papillomavirus infection and association with abnormal Papanicolaou smears in sexually active adolescents. AJDC.
1989;143:1443-1447. 37. Hillard PJA, Biro FM, Wildey LS, Bradley JM, Burket RL, Rauh JL. Cervical dysplasia and human papillomavirus: evaluation in an adolescent dysplasia clinic. Adolesc Pediatr Gynecol. 1989;2:32-36. 38. Piver MS, Baker TR. Cervical cancer in the adolescent patient. Pediatr Ann. 1986;15:536-541. 39. Moscicki A-B, Palefsky J, Gonzales J, Schoolnik GK. Human papillomavirus infection in sexually active females: prevalence and risk factors. Pediatr Res. 1990;28:507-513.
Gutman, MD; Karen St. Claire, MD; Marcia E. Herman-Giddens, PA, MPH; William W. Johnston, MD; William C. Phelps, PhD
\s=b\ Objective.\p=m-\The
objective of this study was to compare the prevalence of intravaginal human papillomavirus\x=req-\ associated disease in two groups of girls to develop information regarding the means of transmission of anal-genital human papillomavirus disease. Design.\p=m-\Apair of parallel studies of prevalences of human papillomavirus infections in two populations of prospectively enrolled girls. Patients.\p=m-\Index patients consisted of 15 consecutive girls aged 11 years or younger who were confirmed to have been sexually abused, had signs or symptoms of vaginal disease, and required generalized anesthesia for evaluation. Selection of nonabused control patients was based on negative findings from screening evaluations and physical examinations. Main Outcome Measures.\p=m-\Prevalencesof cervical\x=req-\ vaginal human papillomavirus infections in the two populations were compared. Vaginal wash samples from index and control patients were assayed for human papillomavirus 1, 2, 4, 6, 11, and 16 by reverse-blot and Southern substantial body of epidemiologie data indicates that human papillomavirus (HPV)-related diseases of the adult genital tract are transmitted primarily through sexual contact.1 In contrast, there have been conflicting reports about the means of transmission of anal-genital warts to postneonatal children. The major area of uncer¬ tainty is whether the primary mode of transmission of genital warts to children is similar to that in adults, ie, sexual contact. This question has been studied primarily through the assessment of children who have been iden¬ tified as having genital warts and were evaluated for pos¬ sible sexual abuse. However, the methods of evaluating
A
Accepted for publication
March 25, 1992. From the Departments of Pediatrics (Drs Gutman and St. Claire and Ms Herman-Giddens) and Pathology (Dr Johnston), Duke University Medical Center, Durham, NC, and the Burroughs-Wellcome Research Laboratories, Research Triangle Park, NC (Drs Gutman and Phelps). Presented in part at the 1991 Papillomavirus Workshop, Seattle, Wash, July 23,1991, and at the joint plenary session for the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy and the Infectious Diseases Society of America, Chicago, Ill, September 30, 1991. Reprint requests to Box 3971, Duke University Medical Center, Durham, NC 27710 (Dr Gutman).
Downloaded From: by a Midwestern University User on 01/22/2019
transfer hybridization methods. Papanicolaou smears were examined from index patients. Results.\p=m-\Vaginalwash samples from five (33%) of 15 index patients were positive for human papillomavirus 6, 11, or 16, compared with none of 17 controls. The presence or absence of external anal-genital warts was not correlated with results from the assay of intravaginal samples. Blinded readings of vaginal exfoliative cytologic findings of the index patients showed koilocytosis, atypia, or inflammatory reactions in four of five human papillomavirus\x=req-\ positive girls, and normal cytologic findings in one human
papillomavirus-positive girl. Conclusion.\p=m-\Thesefindings support
other studies that indicate that sexual contact is a major route in the transmission of anal-genital human papillomavirus-related disease in children. Evaluation of intravaginal specimens was
required to identify human papillomavirus-infected girls since the results of the wash samples were not correlated with the presence or absence of external anal-genital warts. (AJDC. 1992;146:694-699)
for sexual abuse have varied greatly, and, thus, reported rates of confirmed child sexual abuse in children with anal-genital warts have ranged from 0% to 91%.2j3 Evalu¬ ation of abused girls for intravaginal HPV-related infec¬ tion has not been recommended, and no data on the prevalence of intravaginal HPV infection in children have been published. The intravaginal microbiologie findings of nonabused young girls have not been well characterized because of the social and technical limitations in acquiring samples. Prevalence studies in defined populations represent a means of investigating the relationship between sexual abuse and anal-genital HPV infection in children. How¬ ever, because of the limitations on sample collection, studies to identify intravaginal HPV infection in either sexually abused girls or nonabused girls have not been
reported.
The present study documents the prevalence of a selected number of HPV types from vaginal wash samples of two groups of young girls whose status regarding sex¬ ual abuse was fully characterized. The first group com¬ prised girls who had been victims of invasive child sexual abuse, while girls in the second group had not been sex¬ ually abused. This study differs in two respects from oth¬ ers that have examined the means of transmission of anal-genital HPV disease in children. In this study, two
of girls were compared, and the study focused on the diagnosis of intravaginal HPV disease as manifested by HPV, DNA, and cytologie findings, rather than on external genital warts.
populations
SUBJECTS AND METHODS Selection of Index Patients
The Duke Child Protection Team examines approximately 250 girls per year for suspected child sexual abuse, and abuse is con¬ firmed in about 40% of examined girls. The index population comprised 15 girls aged 11 years or younger who were Tanner stage 1 or 2 in sexual development and who had been referred to the Duke Child Protection Team for evaluation of suspected child sexual abuse. The index population included all girls from a 41/4-year period beginning in 1986 who fulfilled the following criteria: (1) The Child Protection Team completed the evaluation for suspected child sexual abuse. (2) The girls had a history of symptomatic vaginal disease (ie, discharge, bleeding, or pain) that required a medical examination. (3) Generalized anesthesia was used because the girls were unable to tolerate the examina¬ tion while awake. (4) The girls were confirmed to have been abused by the Child Protection Team. Examinations using anes¬ thesia were indicated when there were persistent medical signs or symptoms and the child could not tolerate the examination while awake. Index patients were referred for an examination with anesthesia for evaluation of acute injury and/or recurrent or chronic symptomatic vaginal disease manifested as bleeding, discharge, pain, or lesions (Table 1).
Table 1.—Genital Signs and Symptoms of Index Patients (n 15) =
No. of Patients
Presenting complaints
Vaginal discharge Chronic Acute
or
recurrent
Vaginal bleeding
Labial abrasions Perivaginal lesions
Perivaginal condyloma acuminatum Perivaginal molluscum contagiosum
Perianal lesions
Additional findings during anesthesia administration Scars, tears, notches, or distortions of hymen
Hymenal opening transverse diameter 3=8
mm
Changing external examination results with serial evaluations Abnormal external genital examination results
7 2 5 3 3 1 1
8
10 2
15
Table Characteristic
Selection of Control Patients Control patients were recruited from the pediatrie clinics at the Duke Eye Center, Durham, NC. They were girls aged 11 years or younger and Tanner stage 1 in sexual development who had been scheduled to receive ophthalmic examinations or proce¬ dures while receiving generalized anesthesia. Control patients underwent a screening assessment for sexual abuse because child sexual abuse is a common problem and nonabuse of this population could not be assumed." The median age of the index population was 59 months, with a range of 23 to 131 months. The median age of the control pop¬ ulation was 34 months, with a range of 3 to 68 months (Table 2). The control population was significantly younger than the index population (P=.006; two-sample Wilcoxon's rank sum test). Eight (53%) of the index children were black and seven (47%) were white. Of 17 controls, six (35%) were black and 11 (65%) were white. The groups did not differ significantly in the percentage of black children (P= .47; Fisher's Exact Test).
Evaluation for Child Sexual Abuse Methods used in evaluating the index patients for sexual abuse have been described7 and follow currently recommended proce¬ dures, including parent interviews and child interviews and ex¬ aminations.8 Five groups of data were obtained: (1) a review of the behavioral history of the child, (2) a review of the medical history of the child, (3) multiple diagnostic interviews with the child, (4) a forensic medical examination, and (5) assessment for other sexually transmitted diseases. In the present study, "dis¬ closure" refers to the detailed verbalization by the child or other witness of a specific sexually abusive act. Perpetrators were "identified" if the child or a caretaker had provided specific de¬ scription of the person as an assailant. Child sexual abuse was confirmed on the basis of any of the following: (1) the child or a witness disclosed the abuse; (2) gen¬ ital gonorrhea or acquired syphilis was diagnosed; (3) the gen¬ ital examination results were abnormal, with findings clearly in¬ dicative of abuse as outlined at the 1985 National Child Sexual Abuse Summit meeting and subsequently updated9"12; (4) serial genital examinations showed changing abnormalities that were indicative of active sexual abuse; or (5) there was acknowledg¬ ment by the perpetrator. The following methods were used in evaluating control patients for concerns of sexual abuse. At an eye center clinic visit, the study and procedure were described to the parents and child by a member of the Child Protection Team and informed consent was obtained. An intake questionnaire was administered by one of us (K.S.C.) that included questions about each of the follow¬ ing medical areas: chronic illness and previous medical inter¬ ventions (five questions asked); medications (two questions); genitourinary signs and symptoms (nine questions); gastrointes¬ tinal signs and symptoms (three questions); vaginal signs and symptoms (five questions); and behavioral disturbances (10 questions). Open questions were asked regarding any concerns the parents may have had concerning physical or sexual abuse. The physical examination of the external genitalia was per¬ formed at the time of administering anesthesia. The evaluation did not include a separate diagnostic interview with the child.
2.—Study Populations
Index Patients (n = 15)
(range)
Controls (n = 17)
59 (23-131) 34 (3-68) .006 (Wilcoxon's rank sum test) age, mo Race, No. (%) of patients .47 (Fisher's Exact Test) Black 8(53) 6(35) 11 (65) White 7(47) Assays positive for HPV6, 11, or 16, 5 (33)* .015 (Fisher's Exact Test) No. (%) of patients *HPV indicates human papillomavirus infection. The HPV types identified included HPV6 in two patients and HPV11, HPV 6/11, and HPV16 in one patient each.
Median
Downloaded From: by a Midwestern University User on 01/22/2019
The
study protocol required that parents be informed of find¬
ings that raised concerns about possible child sexual abuse and
the parents were aware that a report of suspected sexual abuse would be made should findings strongly indicate abuse. The study was approved by the Duke Institutional Review Board and written parental consent was obtained.
Physical
Examination and Collection of Specimens
The external genital examination for both groups of children included inspection of the labia and introitai area for redness, discharge, lesions, or injuries; identification of hymenal type and contour; and measurement of the horizontal diameter of the hy¬ menal opening.13 Samples obtained included vaginal swabs and washes. Intravaginal samples from index patients were obtained during direct visualization of the vagina and cervix, and cervicalvaginal Papanicolaou smears were obtained. Vaginal wash ma¬ terial was collected through a flexible, small-bore plastic cathe¬ ter that was fitted to the end of a 10-mL syringe. The catheter was inserted through the hymenal opening, and 10 mL of Dulbecco's phosphate-buffered saline was washed in and out of the vagina two or three times. Direct intravaginal visualization was not at¬ tempted in the control children and therefore Papanicolaou smears were not obtained.
Virologie Diagnosis Plasmid-cloned isolates of HPV1, 2, and 4 have been previ¬ ously described.14 The HPV6 probe consists of the Fnu4Hl (b. 509) fragment to the Mstll (b. 4520).15 The HPV16 consists of the full genome from a pUC18 clone derived from the proto¬ type.16 The HPVll/pUC18 was cloned from a patient specimen
(L.T.G, unpublished data, 1991).
The HPV DNA fragments were isolated electrophoretically and labeled with 32P using random oligonucleotide primers for use as
probes.
HPV DNA Analysis Cellular DNA was isolated from the pellet of vaginal wash material by overnight digestion, phenol extraction, and alcohol precipitation. The extracted DNA was digested with RNAse, ex¬ tracted with phenol, and alcohol precipitated a second time. The concentration of DNA was determined with a spectrophotometer. The range of DNA extracted from samples from index patients was 630 ng to 2.25 µg and that from control patients was 315 to 630 ng. A minimum of 300 ng was required to complete the study protocol. Reverse Southern
Hybridization
Reverse Southern hybridization for HPV1, 2, 4, 6, 11, and 16 has been previously described.17-18 Briefly, 200 ng of each of HPV types 1, 2, 4, 6,11, and 16 was immobilized on a nylon filter. The probe consisted of 50 to 100 ng of patient DNA that was labeled with 32P using random nucleotide primers. Hybridization was performed at 65°C. The sensitivity of this assay was less than or equal to 1 pg of HPV DNA in 1 µg of cellular DNA.
Southern Transfer
Hybridization
Two hundred nanograms of cellular DNA from each index and control patient was digested with the restriction endonuclease enzymes, BamHI or Psf I, and analyzed with Southern transfer
hybridization procedures.19 Samples were separated by electro¬ phoresis, transferred to nylon filter, and probed with 32P-labeled HPV6/11 and HPV16. Hybridization produced a readily visible signal with less than 0.2 pg of HPV DNA per 200 ng of cellular DNA.
Cervical-Vaginal Cytologie Findings Cervical-vaginal swabs were collected from index patients for
cytologie examination. The preparations were interpreted by a single observer (W.W.J.), who was blinded to the results of the HPV assays. Descriptive observations were made and findings were categorized according to the Bethesda System.20 Koilocytes are mature squamous cells characterized by a large perinuclear cavity writh sharp borders. The cellular cytoplasm is dense and the nucleus shows varying degrees of degeneration. These cells are commonly present in HPV-infected lesions. RESULTS
Control and Index Patients Fifteen index patients met the criteria for entry into the study. Thirteen girls (87%) gave a full disclosure of sexual abuse and two (13%) gave a partial disclosure of sexual abuse. Thirteen girls had abnormal examination findings of the introitai area or hymen that were consistent with healing or healed trauma from sexual abuse (Table 1). The 15 index patients came from 15 separate households and no index patient was related to any other patient. Seventeen control patients underwent a behavioral, medical, and physical screening evaluation for child sex¬ ual abuse and had negative findings. A vaginal wash sample was obtained from each of these girls for evalua¬ tion for HPV. Two potential control patients were omit¬ ted from the study because of concerns of possible abuse that were identified during the screening evaluation. Both children disclosed their abuse and identified the perpe¬ trators.
Perpetrators (33%) of the index children had been assaulted by single perpetrators, four (27%) had been assaulted by multiple perpetrators, three (20%) disclosed abuse but did not identify a specific perpetrator(s), and three (20%) identified one perpetrator with at least one additional perpetrator suspected. All identified perpetrators were Five
either related to the child or had access to the child because of a relationship with another member of the
family.
The types of abuse that were disclosed by the index pa¬ are listed in Table 3. Eleven children (73%) de¬ scribed penetrating penile-vaginal or penile-anal abuse, and two (13%) described digital abuse. Two children (13%) were unable to give details of their assault. tients
Table
3.—Type of Sexual Abuse Disclosed (n 15)* =
Site of Contactt I-1
Assailant
No. of Index
Child
Children
Penis
Vagina
9
Penis
Anus
4
Penis
Mouth
2
Finger
Vagina Vagina multiple types
2
Mouth
1
*Five children disclosed of abuse. tThe site of contact was unknown for two children.
Downloaded From: by a Midwestern University User on 01/22/2019
HPV DNA
Assay Vaginal samples from five (33%) of the 15 index patients were positive for HPV using Southern transfer hybridization. Four (80%) of these five were HPV types 6 wash
11, and one was HPV type 16 (Table 2). The results of the reverse-blot assays corroborated the findings of the Southern transfer hybridization assays (Figs 1 and 2). La¬ beled DNA from index patients did not hybridize with HPV1, 2, or 4 on reverse blots. The prevalences of HPV or
Fig 1.—Restriction endonuclease digestion results from a cervicalvaginal wash specimen. BamHI and Pst / demonstrated fragments
characteristic of HPV6a.
DNA from the two study populations, 0 of 17 controls vs five of 15 index cases, differed significantly (P= .015) us¬ ing Fisher's Exact Test. The 95% confidence interval for HPV positivity in the index population was 9% to 57%.
External Genital Warts The relationship between the presence or absence of external genital lesions and intravaginal infection is shown in Table 4. Four of the index patients had clinically evident external anal-genital warts, one in the perianal area and three in the perivaginal area. Two girls also had an intravaginal specimen that was HPV positive and two did not. Three (60%) of the five girls who had HPVpositive intravaginal specimens did not have identifiable external genital warts. Biopsy material of external lesions from one patient was examined for HPV DNA, and both wash and biopsy results showed HPV6a.
Cytologie Findings Of the five girls with HPV-positive wash specimens, only one had cytologie findings that included koilocytosis (Fig 3). Cytologie findings from a second HPV-positive child showed a severe acute inflammatory reaction in which the parabasal cells were surrounded by a dense aggregation of polymorphonuclear leukocytes. The nuclei of parabasal cells showed atypia with darkening of the nuclear chromatin, thickening of the nuclear envelope, and increased nuclear to cytoplasmic ratio. For the third child, cytologie findings were restricted to chronic inflam¬ matory reactions showing lymphocytic and monocytic cellular predominance and relatively fewer parabasal cells. Cytologie findings for the fourth child included a severe acute inflammatory reaction with a predominance Downloaded From: by a Midwestern University User on 01/22/2019
Fig 2.—Restriction endonuclease digestion results from a cervicalvaginal wash specimen. BamHI and Pst / demonstrated fragments characteristic of HPV11.
Table 4.—Relationship Between Presence or Absence of External Genital Lesions and Positive or Negative Assay Results From Intravaginal Wash Samples For Index Patients With Human Papillomavirus (HPV) Infection
Vaginal Wash Assay Results, No. of Patients I
Clinical Evaluations External
I
HPV DNA
HPV DNA
Positive
Negative
Total
anogenital warts
present
Perianal
Perivaginal External anogenital warts
absent Total No. of index
patients
0 2
1 1 13
3
8
11
10
15
5
of polymorphonuclear leukocytes and parabasal cells that showed mild perinuclear clearing but not overt koilocytosis. The fifth child had normal cytologie findings, with¬ out evidence of inflammatory reaction, atypia, or koilo-
cytosis.
Cytologie examinations from the 10 index girls who had
negative HPV DNA vaginal washes showed neither cel¬ lular reactions suggestive of HPV infection nor acute or
chronic
inflammatory reactions.
COMMENT The index patients represent a subpopulation of sexu¬ ally abused girls who were particularly symptomatic from the trauma that they had experienced and whose abuse
other
studies,
more than 60% of adults in established re¬ with partners who have genital warts were concordant for the presence of lesions.24-25 Similar data for child victims of sexual abuse are not available because, in these cases, perpetrators identified by children are rarely charged with the offense. This effectively prohibits any evaluation for sexually transmitted diseases. Other evidence that HPV infection is sexually trans¬ mitted in adults includes data showing a relationship be¬ tween prevalence of HPV infection and numbers of sex¬ ual partners. In adults, these two factors increase proportionally.26-27 Similarly for children, a recent study has shown a significant association between the presence of multiple, other sexually associated organisms in chil¬ dren who have genital warts and who have been abused
lationships
by multiple perpetrators.3
Fig 3.—Papanicolaou
smear
from human
papillomavirus-positive
child, demonstrating koilocytic findings in squamous cell. The nu¬ cleus is marginated and condensed, the cell is enlarged, anda halo surrounds the nucleus
(original magnification 320).
probably unusually severe. At least four (27%) of the 15 patients had been abused by multiple perpetrators. All had signs or symptoms of in tra vaginal disease. Most of the types of disclosed abuse that they experienced were invasive. The 33% rate of recovery of HPV DNA from pa¬ tient samples is similar to the recovery rate of HPV from a sexually at-risk adult population that was studied using was
dot-blot hybridization method.21 Using the Southern transfer hybridization method, the recovery rate of HPV DNA from index cases was slightly higher than the 15% rate of recovery from another at-risk adult population.22 Results from the HPV assays of the index children therefore closely resemble results from adults who are highly sexually active. Even so, the 33% prevalence rate of HPV from the index population is ex¬ pected to be an underrepresentation of the true preva¬ lence of this infection in children who have been severely abused because sample sizes were small, assays for only six HPV types were performed, and only a single sample was taken from each child. In adults, the primary means of transmission of HPVrelated genital disease is through sexual contact. Evidence of sexual transmission in adults comes from studies of couples; 90% of new sexual contacts of persons with gen¬ ital warts result in the development of genital lesions.23 In a
Downloaded From: by a Midwestern University User on 01/22/2019
Medical literature on children with anal-genital papil¬ lomavirus disease contains conflicting information about sexual contact as the means of transmission of this infec¬ tion to children. Unfortunately, many of these reports and studies were done prior to the development of currently established methods for evaluating child sexual abuse. Most of these earlier reports did not use a sexual abuse evaluation process, which must include a diagnostic interview of the child and parent by a skilled interviewer and a forensic genital examination by a qualified exam¬ iner. Rather, most of the earlier reports relied on the child to volunteer the disclosure and categorized nondisclosing children who had not received a thorough evaluation as nonabused. Several of the more recent studies have also omitted any description of the evaluation methods and findings that they used to diagnose sexual abuse, or they have catego¬ rized children as nonabused when the child's status was inadequately documented or unknown.2'2829 Conse¬ quently, reported associations of sexual abuse with HPV transmission to children are likely to be inaccurately low. Despite the limitations in their diagnosis of sexual abuse, however, these earlier reports document that at least 50% of children with genital warts disclosed that they had been sexually abused.30-31 The only published study using cur¬ rent diagnostic methods for child sexual abuse in girls having genital warts showed that 10 (91%) of 11 had been sexually abused. In that study disclosure of abuse was provided in 71% of cases, and in a similar study a high disclosure rate was also recorded.3,32 Sexual abuse was confirmed in four of the five children aged 3 years or younger.3 This strongly suggests that even in children who are preverbal or only marginally verbal the presence of genital tract lesions due to HPV is frequently associated with a confirmed diagnosis of sexual abuse. The current study used the Southern transfer hybrid¬ ization assay for several reasons. First, previous studies using this method have demonstrated that most genital papillomas of children are caused by HPV6 or 11 and only infrequently by HPV2,16, or 18.33-34 Other types are rarely or never identified in genital lesions of children. Conse¬ quently, our assay was designed to examine specimens for a limited number of HPV types. Second, because nonamplified DNA was measured, quantitative informa¬ tion was provided, and identification of the type was precise. Third, the method is the criterion standard. In the present study, racial distribution of the index and control children did not differ significantly. The mean age of the control population, however, was significantly
younger than that of the index population. The control patients were therefore closer in time to possible neona¬ tal transmission of HPV, and they were also more likely to have difficulty in evaluations for sexual abuse due to their preverbal or poorly verbal developmental stages. These factors resulted in a conservative bias for the study. Of particular concern for the HPV-positive children is that the long-term consequences of early exposure to this infection are unknown. Of note are reports from several medical centers about the increased incidence of genitaltract papillomavirus infections in female adolescents. Be¬ nign, dysplastic, and malignant HPV-associated lesions are also now being reported in this age range.35"39 Because the length of time between HPV infection and progression to dysplastic and malignant stages may be quite long, some of these female adolescents may have acquired their infection prior to the legal age of consent for intercourse and through abusive sexual contact as a child. The present data indicated that our index population of abused girls resembles populations of sexually active adults in the prevalence of intravaginal HPV DNA, in the presence of anal-genital lesions, and in cytologie findings. These children have acquired their internal genital HPV disease at very immature ages. Data regarding their prognosis and appropriate treatment are needed. Study findings also support other data that children presenting with anal-genital papillomavirus disease are at significant risk of having been abused and should be evaluated for possible sexual abuse and other sexually transmitted in¬ fections. These studies were supported in part by a grant from the Mellon Foundation. The authors gratefully acknowledge the active participation of Robert Machemer, MD, and Edward Buckley, MD, and the faculty and personnel of the Duke Eye Center, Durham, NC. Gerald Hajian, PhD, Burroughs-Wellcome Co, Research Triangle Park, NC, pro¬ vided expert assistance with data analysis and statistical evaluation of the results, and David Barry, MD, provided editorial review. Julia Barnes, Burroughs-Wellcome Co, provided technical advice. Nancy Berson, MSW, Carmenza Salgado, PA, and Jeanne Niemeyer, MSW, Duke University Medical Center, conducted skillful interviews and examinations of index patients, and Gale Hill, PhD, Duke Univer¬
sity, provided microbiologie support.
References 1. Koutsky LA, Galloway DA, Holmes KK. Epidemiology of genital human papillomavirus infection. Epidemiol Rev. 1988;10:122-163. 2. Obalek S, Jablonska S, Favre M, Walczak L, Orth C. Condylomata acuminata in children: frequent association with human papillomaviruses responsible for cutaneous warts. J Am Acad Dermatol.
1990;23:205-213.
3. Herman-Giddens ME, Gutman LT, Berson NL, Duke Child Protection Team. Association of coexisting vaginal infections and multiple abusers in female children with genital warts. Sex Transm Dis.
1988;15:63-67.
4. Russell DEH. The incidence and
of intrafamilial and extrafamilial sexual abuse of female children. Child Abuse Negl.
prevalence
1983;7:133-146. 5. Moore KA, Nord CW, Peterson JL. Nonvoluntary sexual activity among adolescents. Fam Plann Perspect. 1989;21:110-114. 6. Siegel JM, Sorenson SB, Golding JM, Burnam MA, Stein JA. The
prevalence of childhood sexual assault: the Los Angeles Epidemiologic Catchment Area Project. Am J Epidemiol. 1987;126:1141-1153. 7. Gutman LT, St Clair KK, Weedy C, et al. Human immunodeficiency virus transmission by child sexual abuse. AJDC. 1991;145:137-141. 8. Krugman RD, Bays JA, Chadwick DL, Levitt CJ, McHugh MT, Whitworth JM. Guidelines for the evaluation of sexual abuse of children. Pediatrics. 1991;87:254-260. 9. Tipton AC. Child sexual abuse: physical examination techniques and interpretation of findings. Adolesc Pediatr Gynecol. 1989;2:10-25. 10. White ST, Ingram DL, Lyna PR. Vaginal introital diameter in the evaluation of sexual abuse. Child Abuse Negl. 1989;13:217-224.
Downloaded From: by a Midwestern University User on 01/22/2019
11. Paradise JE. The medical evaluation of the sexually abused child. Pediatr Clin North Am. 1990;37:839-862. 12. Muram D. Child sexual abuse-genital tract findings in prepubertal girls, I: the unaided medical examination. Am J Obstet Gynecol.
1989;60:328-333. 13. McCann J, Voris J, Simon M, Wells R. Comparison of genital examination techniques in prepubertal girls. Pediatrics. 1990;85:182-187. 14. Heilman C, Law M-F, Israel MA, Howley PM. Cloning of human
papilloma virus genomic DNAs and analysis of homologous polynucleotide sequences. J Virol. 1980;36:395-407. 15. Schwarz E, Durst M, Demankowski C, et al. DNA sequence and genome organization of genital human papillomavirus type 6b. EMBO
J. 1983;2:2341-2348. 16. Durst M, Gissman L,
Ikenberg H, Zur Hausen H. A papillomaviDNA from a cervical carcinoma and its prevalence in cancer biopsy samples from different geographic regions. Proc Natl Acad Sci U S A. rus
1983;80:3812-3815. 17. deVilliers E-M, Schneider A, Gross G, zur Hausen H. Analysis of
benign and malignant urogenital tumors for human papillomavirus infection by labelling cellular DNA. Med Microbiol Immunol. 1986; 174:281-286. 18. Webb
DH, Rogers RE, Fife KH. A one-step method for detecting and typing human papillomavirus DNA in cervical scrape specimens from women with cervical dysplasia. J Infect Dis. 1987;156:912-919. 19. Sambrook J, Fritsch EF, Maniatis T, eds. Molecular Cloning. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1989. 20. Solomon D. The 1988 Bethesda system for reporting cervical/ vaginal cytologic diagnoses. Anal Quant Cytol Histol. 1989;11:291\x=req-\ 297. 21. Burmer GC, Parker JD, Bates J, East
K, Kulander BG. Comparative analysis of human papillomavirus detection by polymerase chain reaction and Virapap/Viratype kits. Am J Clin Pathol. 1990;94:554-560. 22. Kiviat NB, Koutsky LA, Critchlow CW, et al. Comparison of Southern transfer hybridization and dot filter hybridization for detection of cervical human papillomavirus infection with types 6,11,16, 18,
31, 33 and 35. Am J Clin Pathol. 1990;94:561-565. 23. Barrett TJ, Silbar JD. Genital warts: a venereal disease. JAMA. 1954;154:333-334. 24. Oriel JD. Maternal history of genital warts. Br J Vener Dis. 1971;47:1-13. 25. Barrasso R, DeBrux J, Croissant O, Orth G. High prevalence of
papillomavirus-associated penile intraepithelial neoplasia
in sexual
partners of women with cervical intraepithelial neoplasia. New Engl J Med. 1987;317:916-923. 26. Kiviat NB, Koutsky LA, Paavonen JA, et al. Prevalence of genital
papillomavirus infection in women attending a college student health clinic or a sexually transmitted disease clinic. J Infect Dis. 1989;159:293\x=req-\ 302. 27. Syrjanen K, Vayrynen M, Castren O, et al. Sexual behavior of women with human papillomavirus (HPV) lesions by the uterine cervix. Br J Vener Dis. 1984;60:243-248. 28. Cohen BA, Honig P, Androphy E. Anogenital warts in children: clinical and virologic evaluation for sexual abuse. Arch Dermatol. 1990;126:1575-1580. 29. Padel AF, Venning VA, Evans MF, Quantrill AM,
Fleming KA. Hu-
papillomaviruses in anogenital warts in children: typing by in situ hybridization. Br Med J. 1990;300:1491-1494. 30. DeJong AR, Weiss JC, Brent RL. Condyloma acuminata in chilman
dren. AJDC. 1982;136:704-706. 31. Shelton TB, Jerkins GR, Noe HN. Condylomata acuminata in the pediatric patient. J Urol. 1986;135:548-549. 32. Gilles EE, Hitchcock BS, Elvik SL. Human papilloma virus in children. Ninth National Conference on Child Abuse and Neglect; September 14-17, 1991; Denver, Colo. Abstract B.1, 55. 33. Rock B, Naghashfar E, Barnett N, Buscema J, Woodruff SD, Shah K. Genital tract papillomavirus infection in children. Arch Dermatol.
1986;122:1129-1132. 34. Gibson PE, Gardner SD, Best SJ. Human papillomavirus types in anogenital warts of children. J Med Virol. 1990;30:142-145. 35. Martinez J, Smith R, Farmer M, et al. High prevalence of genital tract papillomavirus infection in female adolescents. Pediatrics. 1988;
82:604-608. 36. Rosenfeld WD, Vermund SH, Wentz ST, Burk RD. High prevalence rate of human papillomavirus infection and association with abnormal Papanicolaou smears in sexually active adolescents. AJDC.
1989;143:1443-1447. 37. Hillard PJA, Biro FM, Wildey LS, Bradley JM, Burket RL, Rauh JL. Cervical dysplasia and human papillomavirus: evaluation in an adolescent dysplasia clinic. Adolesc Pediatr Gynecol. 1989;2:32-36. 38. Piver MS, Baker TR. Cervical cancer in the adolescent patient. Pediatr Ann. 1986;15:536-541. 39. Moscicki A-B, Palefsky J, Gonzales J, Schoolnik GK. Human papillomavirus infection in sexually active females: prevalence and risk factors. Pediatr Res. 1990;28:507-513.
