Juvenile-onset recurrent respiratory papillomatosis, also referred to as laryngeal papillomatosis, is a rare disease associated with low mortality but significant morbidity. Lesions occur most often in the larynx, vocal cords, and oral and nasal mucosa, particularly at junctions of the squamous and ciliary epithelium. 1 Incidence estimates range from two to 43 cases per 1,000,000 children and are dependent on the populations examined and methods used. 2–7 Respiratory papillomatosis is the most common benign neoplasm of the larynx in children 3,8–10 and causes childhood hoarseness, a change in voice, chronic cough, choking, recurrent respiratory infections, and failure to thrive. Derkay 3 reported malignant progression of respiratory papillomatosis in less than 5% of cases and extralaryngeal spread in approximately 30% of cases. Treatment primarily consists of surgical removal, but recurrence is common. A tracheotomy may be performed to avoid life-threatening airway obstruction, and has been reported in 14% of juvenile cases. 3 Case reports of respiratory papillomatosis–associated deaths are rare. 11
The infectious nature of respiratory papillomatosis was first demonstrated in 1923 when Ullman 12 developed warts after inoculating himself with the laryngeal tissue from an infected 6-year-old boy. Subsequent case reports suggested that human papillomavirus (HPV) infection could be transmitted from mother to infant, resulting in respiratory papillomatosis. 13,14 In addition, investigators have reported that over 50% of laryngeal papillomatosis cases have a maternal history of genital warts, 15,16 thus suggesting an etiological association between genital warts and respiratory papillomatosis. The identification of HPV types 6 and 11 from both genital warts and respiratory papillomatosis 17–19 provided the definitive evidence for the etiological link between genital warts and respiratory papillomatosis.
The Danish national registries are a valuable resource that allowed for an evaluation of factors associated with the development of respiratory papillomatosis among births over a 20-year period. All women with a history of genital warts during pregnancy were identified from the national registries. Respiratory papillomatosis cases were identified by a comprehensive medical record review from ear, nose, and throat departments. Our data allowed for the estimation of the absolute risk of respiratory papillomatosis in women with genital warts in pregnancy. We also evaluated the effect of young maternal age, nulliparity, and vaginal delivery, which were previously reported risk factors of disease. 20–22 Additional potential risk factors available in the registry data were also evaluated, including increased delivery times and procedures and complications during delivery.
MATERIALS AND METHODS
Between 1974 and 1993, the Danish Medical Birth Registry recorded 1,206,213 births in Denmark. We identified 3033 women who had genital warts during pregnancy by linking the data in the Danish National Hospital Registry with those in the Danish Medical Birth Registry. A representative sample (n = 7902) of all births was selected by a simple random sample from the Danish Medical Birth Registry database to serve as a comparison group for the women with genital warts. We identified 57 cases of respiratory papillomatosis by review of medical records from ear, nose, and throat departments. A retrospective cohort design was used to identify risk factors for respiratory papillomatosis among Danish births.
The data were extracted from the Danish National Registries, which have been described elsewhere. 23,24 Briefly, the Danish Medical Birth Registry 23 contains data on all live births and stillborn fetuses by women living in Denmark and all infant deaths, defined as less than or equal to 1 year of age. From this database, we obtained sociodemographic information on the mothers and newborns, previous pregnancy and delivery histories of the mothers, and details of the delivery, including gestational age, birth weight, birth length, and delivery duration. The procedures performed and complications encountered during the delivery process were also included. The Danish National Hospital Registry 24 includes detailed information on all hospital admissions. The registry can accommodate up to 20 discharge diagnoses per admission, which are coded according to the International Classification of Diseases. From this registry, we obtained all diagnoses of condylomata acuminata (International Classification of Diseases, 8th revision code 99.9) among pregnant women. The two national registries described above were linked using each mother's ten-digit Central Population Registry number, a unique identifier that has been provided to all Danish citizens since 1968. The result was the identification of the earliest diagnoses of genital warts during pregnancy among women who gave birth between 1974 and 1993.
To identify respiratory papillomatosis cases, we contacted all Danish hospitals with an ear, nose, and throat department (n = 22). Each hospital provided a list of patients who were admitted between January 1974 and December 1999 with a discharge diagnosis of benign neoplasm of the larynx, trachea, bronchus, or lung. Discharge diagnoses were based on the International Classification of Diseases, 8th revision before 1994 (International Classification of Diseases, 8th revision codes 508, 212, 231), and the International Classification of Diseases, 10th revision after 1994 (International Classification of Diseases, 10th revision codes J38, D14, D38). International Classification of Diseases, 9th revision codes were not implemented in Denmark. Medical records of all patients identified from the lists and born between 1974 and 1993 were reviewed.
Diagnoses of respiratory papillomatosis were distinguished from other benign conditions, such as hemangiomas, fibromas, chondromas, myxomas, neurofibromas, polyps, and singer's nodes, by the histopathologic report, which was included in the medical record. All cases identified in the medical record review with at least one histopathologically confirmed respiratory papillomatosis lesion were included in our study. We excluded cases with an adult onset, defined as 17 years of age or greater. The methods used to identify cases by medical record review were identical to the methods used by Lindeberg and Elbrond, 5 who reported the incidence of respiratory papillomatosis in a subpopulation in Denmark from 1965 to 1984. We independently identified all 17 cases described in this earlier study that matched the criteria for our study, indicating that our surveillance was excellent.
Approvals from the Danish Data Surveillance Authority, the Scientific–Ethics Committees in Denmark, the Johns Hopkins University, Bloomberg School of Public Health, and the Office for Human Research Protections of the United States Department of Health and Human Services were obtained to access the national registries and access the medical records of respiratory papillomatosis cases.
Relative risks and 95% confidence intervals (CIs) were calculated to compare the risk of disease among “exposed” children with a maternal history of genital warts in pregnancy with the risk of disease among two “unexposed” groups: 1) the random sample of children without a maternal history of genital warts and 2) all live births without a maternal history of genital warts. A population-attributable risk 25 was also calculated to quantify the proportion of respiratory papillomatosis attributable to a maternal history of genital warts in pregnancy.
Additional analyses to identify cofactors for disease development were performed using time-to-event procedures to account for different lengths of follow-up after birth. Data from all respiratory papillomatosis cases (n = 57), and all births with a maternal diagnosis of genital warts (n = 3033) were included. Unadjusted and adjusted relative hazards (similar interpretation to the relative risk) were computed using the Cox proportional hazards model. 26 The outcome used in the Cox models was the time to respiratory papillomatosis diagnosis, which allowed for variable lengths of follow-up. Those subjects who ended follow-up without a diagnosis of respiratory papillomatosis were considered censored 26 at December 31, 1999. Several groups of characteristics were evaluated, including maternal demographics and delivery history, newborn characteristics, procedures performed during pregnancy, and complications occurring during pregnancy.
SAS 8.0 (SAS Inpstitute Inc., Cary, NC) was used for all analyses.
Of the 1.2 million live births in Denmark during the study period, we identified 57 respiratory papillomatosis cases, corresponding to an overall risk of 4.7 per 100,000 births and an overall rate of 3.5 per 1,000,000 person-years. There were no statistically significant differences (P < .05) of selected characteristics between the 3033 newborns identified from the Danish national registries with a maternal history of genital warts and the 7902 randomly selected sample of births without such a history in pregnancy (Table 1). In particular, the proportion of births with a cesarean delivery was similar in the two groups (P = .51), suggesting that the presence of genital warts was not an indication for cesarean delivery in this population.
Of the 57 respiratory papillomatosis cases, 21 were identified among the 3033 births with a maternal history of genital warts, and 36 were from the remaining births without a maternal history of genital warts (Table 2). No cases were identified from the 7902 randomly sampled births. The corresponding risk of respiratory papillomatosis was 6.9 cases per 1000 births among births with a maternal history of genital warts. Among births without a maternal history of genital warts in pregnancy, the corresponding risk was 0.0 cases per 1000 births in the random sample and 0.03 cases per 1000 among all live births without a maternal history of genital warts (Table 2). The risk of disease among births with a maternal history of genital warts in pregnancy was 231.4 (95% CI 135.3, 395.9) times higher than that among births without a maternal history of genital warts. Among births during the 20-year period, 37% of respiratory papillomatosis cases were attributable to a maternal history of genital warts during pregnancy (Table 2).
Maternal genital warts in pregnancy were the overwhelming risk factor for respiratory papillomatosis. Thus, subsequent analyses evaluated cofactors of disease between all cases and all births with a maternal history of genital warts that did not result in respiratory papillomatosis. Univariate results are presented for maternal and infant characteristics in Table 3 and for procedures and complications in Table 4. The results of the final multivariable model are listed in Table 5. The only statistically significant (P < .05) findings in the univariate examinations (Tables 3 and 4) were a decreased risk of respiratory papillomatosis among female infants (relative hazard 0.6; 95% CI 0.3, 1.0) and an increased risk if not living with the newborn's father (relative hazard 3.5; 95% CI 1.4, 8.7) (Table 3). We did not identify any complications or procedures performed during pregnancy that were associated with a statistically significant (P < .05) increased risk of respiratory papillomatosis in our univariate models. Of note, elective cesarean, acute cesarean, and instrumentally assisted deliveries did not affect the risk of respiratory papillomatosis relative to the remaining vaginal deliveries. Several factors were associated with a slightly increased, although statistically insignificant (P ≥ .05) risk, including younger maternal age, deliveries with duration of 10 hours or more, cardiotopography, induced stimulation of labor (medical or rupture of membranes), breech presentation, shoulder dystocia, and a narrow maternal pelvis.
In the multivariable model, deliveries longer than 10 hours in duration were associated with an increased risk of respiratory papillomatosis (relative hazard 2.0; 95% CI 1.1, 3.5). The duration of the second stage of labor, which may be a better surrogate for contact of the fetus with infective tissue, was not available in the registry data.
We also observed an increased risk for not living with the newborn's father (relative hazard 3.6; 95% CI 1.4, 9.1). Cesarean delivery, maternal age, parity, and gender were statistically insignificant (P ≥ .05).
The valuable resource of the Danish national registries allowed for the longitudinal evaluation of maternal and infant characteristics associated with respiratory papillomatosis. Fifty-seven respiratory papillomatosis cases occurring during a follow-up were identified by medical record review, corresponding to an overall rate of 3.5 per 1,000,000 person-years, which is similar to the rate reported by Lindeberg and Elbrond. 5 A maternal history of genital warts in pregnancy was identified as the principal risk factor for the development of respiratory papillomatosis. Seven in every 1000 births with a maternal history of genital warts resulted in disease, corresponding to one case for every 144 births with maternal genital warts. Among births without a history of genital warts, the risks were three cases per 100,000 births or one case for every 33,421 births. Thus, maternal genital warts in pregnancy conferred a 231 times higher risk of disease relative to births without a maternal history of genital warts in pregnancy. Examinations of the oral cavity of children and newborns using nasopharyngeal swabs have detected HPV deoxyribonucleic acid (DNA) in some studies, 27,28 suggesting that HPV can be transmitted from mother to infant, but other studies have been negative. 29 A higher risk of respiratory papillomatosis among women with clinical HPV disease confirms the mechanism of mother-to-infant transmission of infection.
The small group of children (0.25% of all births) exposed to maternal genital warts accounted for 37% of disease. The remaining 36 cases of respiratory papillomatosis cases presumably resulted from HPV transmission from a mother with subclinical HPV infections. We did not have the data to estimate the risk of respiratory papillomatosis from subclinical maternal infections, but it is likely to be lower because of less virus shedding in subclinical infections than in clinical infections.
We identified longer delivery times (at least 10 hours) and not living with the newborn's father as additional cofactors among women with genital warts for the development of respiratory papillomatosis in the newborns. Mothers with more than 10 hours between onset of labor and delivery were twice as likely to transmit HPV resulting in respiratory papillomatosis. The increased risk for longer delivery times was likely a result of increased exposure to infection during delivery, providing evidence for intrapartum transmission of infection. Not living with the newborn's father was associated with an increased risk of disease in both univariate and multivariable models. One explanation is that this factor may be correlated with unstable relationships, resulting in more recent exposures to sexually transmitted diseases such as HPV. Confirmation of these cofactors in other populations and settings is necessary.
Of 57 cases in our study, seven (12.5%) were delivered by cesarean, of which only three were elective cesarean deliveries; this percentage was only slightly lower than the overall cesarean delivery rate among noncases (15.7%) and was not statistically significant. Three previous studies by Shah and colleagues 20–22 have examined method of delivery and other maternal cofactors for HPV transmission resulting in respiratory papillomatosis. Eight of 273 (2.9%) respiratory papillomatosis cases were delivered by cesarean in the three studies, which is lower than expected based on national statistics. Maternal young age (less than 20 years) and nulliparity were also more common among cases. 20,21 We observed a slightly increased risk for younger maternal age, but not for parity or cesarean delivery. The discrepancies between studies may in part be explained by differences in both study design and the populations examined. First, our study used a cohort design, which reduces the influence of recall bias, whereas previous investigations used a case–control design using either a selected control group or national statistics for comparison. Furthermore, previous studies had no data on maternal genital warts. Finally, Denmark has substantially lower cesarean delivery rates than the United States, where the previous studies were based. Thus, the power to detect a reduced cesarean delivery rate among respiratory papillomatosis cases in Denmark was limited, particularly with such a small case group.
Fonnest et al 30 have examined the incidence of neonatal herpes in this population, a disease with pathogenesis similar to that of respiratory papillomatosis. They reported 30 neonatal herpes cases among 862,298 births, corresponding to a risk of 3.5 per 100,000 births, a rate similar to the risk of respiratory papillomatosis found in our study (4.7 per 100,000 births). Cesarean delivery prevents transmission of herpes, although the protective effect was not observed when the amniotic membranes ruptured more than 4–6 hours before delivery. 31 Primary maternal herpes infections 32 and the use of fetal scalp electrodes 33 increased the risk for neonatal herpes. We were able to evaluate the role of prelabor rupture of membranes and cesarean delivery for the risk of respiratory papillomatosis, although we did not have data on fetal scalp electrodes or primary versus recurrent HPV infections.
A limitation of this study is the potential misclassification of genital warts. Friis et al 34 examined women hospitalized for condylomata acuminata and noted that those with mild cases of disease would more likely seek treatment in an outpatient setting and would therefore be missed in studies of hospitalized cases. The misclassification may be somewhat reduced in our study because pregnant women were closely monitored throughout pregnancy and were hospitalized during labor and delivery. The result of any misclassification that did occur would be to dilute the observed association of genital warts and the risk of respiratory papillomatosis. The likelihood of the misclassification of respiratory papillomatosis cases is unknown. We had no information, for example, on mild cases that did not receive medical attention. However, in Denmark, all children with clinically relevant lesions in the aerodigestive tract would eventually be referred to an ear, nose, and throat specialist. It is also conceivable, although unlikely, that severe cases would be missed if they resulted in death before receiving medical care. Third, a common limitation of registry-based studies is the lack of investigator control of examined variables. We did not have data on genital wart treatments, use of fetal scalp electrodes, or duration of the second stage of labor, all of which may have an important role in the transmission process. Finally, as discussed, the small sample of cases makes it difficult to examine certain exposures, particularly if these exposures are rare, as in Denmark, which has a low rate of elective cesarean delivery. Nevertheless, larger case groups may be difficult to obtain without sacrificing the high-quality data from national registries.
Through the use of well-designed and maintained national registries, we evaluated cofactors for respiratory papillomatosis during pregnancy that had not previously been examined using a longitudinal, population-based design. These exposures were collected and recorded during the hospital admission, thus eliminating the possibility of recall bias. An additional benefit of registry-based data was the independent collection of exposures and outcomes, avoiding some types of information and selection biases. Identification of respiratory papillomatosis cases was not influenced by the method of delivery or the presence or absence of a maternal history of genital warts in pregnancy.
In summary, we identified a maternal history of genital warts in pregnancy as the primary risk factor for respiratory papillomatosis. Increased contact between the fetus and mother, as measured by delivery time, was associated with an increased risk of respiratory papillomatosis, supporting an intrapartum transmission of HPV. However, considering the lack of a protective effect of cesarean delivery as well as a lack of association with complications, such as prelabor rupture of membranes, or procedures, such as instrumental delivery, in utero or postnatal transmission may be more common than previously suspected. Some studies have found HPV DNA in the amniotic fluid and fetal membranes, 28,35,36 which may have resulted from ascending infection from the maternal genital tract through unperceived tears in the fetal membranes. Additional studies are needed with a larger series of cases to clearly describe the effect of cesarean delivery. Other important factors that may affect the risk of respiratory papillomatosis also deserve further investigation, including location and size of the genital warts lesion in pregnant women, genital warts treatments used, and HPV subtype.
1. Kashima H, Mounts P, Leventhal B, Hruban RH. Sites of predilection in recurrent respiratory papillomatosis. Ann Otol Rhinol Laryngol 1993;102:580–3.
2. Strong MS, Vaughan CW, Healy GB. Recurrent respiratory papillomatis. In: Healy GB, McGill TJ, eds. Laryngotracheal problems in the pediatric patient. Springfield, Illinois: Thomas, 1980:88–98.
3. Derkay CS. Task force on recurrent respiratory papillomas. A preliminary report. Arch Otolaryngol Head Neck Surg 1995;121:1386–91.
4. Armstrong LR, Preston EJ, Reichert M, Phillips DL, Nisenbaum R, Todd NW, et al. Incidence and prevalence of recurrent respiratory papillomatosis among children in Atlanta and Seattle. Clin Infect Dis 2000;31:107–9.
5. Lindeberg H, Elbrond O. Laryngeal papillomas: The epidemiology in a Danish subpopulation 1965–1984. Clin Otolaryngol 1990;15:125–31.
6. Christensen PH, Jorgensen K, Grontved A. Juvenile papillomatosis of the larynx. Acta Otolaryngol Suppl 1984; 412:37–9.
7. Bomholt A. Juvenile laryngeal papillomatosis. An epidemiological study from the Copenhagen region. Acta Otolaryngol 1988;105:367–71.
8. Narozny W, Mikaszewski B, Stankiewicz C. Benign neoplasms of the larynx. Auris Nasus Larynx 1995;22:38–42.
9. Jones SR, Myers EN, Barnes L. Benign neoplasms of the larynx. Otolaryngol Clin North Am 1984;17:151–78.
10. Yoshida T, Kuratomi K, Mitsumasu T. Benign neoplasms of the larynx. A 10-year review of 38 patients. Auris Nasus Larynx 1983;10Suppl:S61–71.
11. Balazic J, Masera A, Poljak M. Sudden death caused by laryngeal papillomatosis. Acta Otolaryngol Suppl 1997; 527:111–3.
12. Ullman EV. On the aetiology of the laryngeal papilloma. Acta Otolaryngol 1923;5:317–34.
13. Hajek E. Contribution to the etiology of laryngeal papilloma in children. J Laryngol Otol 1956;70:166.
14. Kaufman RS. Verrucas and juvenile laryngeal papillomas. Arch Otolaryngol 1969;89:748–9.
15. Quick CA, Watts SL, Krzyzek RA, Faras AJ. Relationship between condylomata and laryngeal papillomata. Clinical and molecular virological evidence. Ann Otol Rhinol Laryngol 1980;89:467–71.
16. Cook TA, Brunschwig JP, Butel JS, Cohn AM, Goepfert H, Rawls WE. Laryngeal papilloma: Etiologic and therapeutic considerations. Ann Otol Rhinol Laryngol 1973;82:649–55.
17. Mounts P, Shah KV, Kashima H. Viral etiology of juvenile- and adult-onset squamous papilloma of the larynx. Proc Natl Acad Sci U S A 1982;79:5425–9.
18. Gissmann L, Wolnik L, Ikenberg H, Koldovsky U, Schnurch HG, zur Hausen H. Human papillomavirus types 6 and 11 DNA sequences in genital and laryngeal papillomas and in some cervical cancers. Proc Natl Acad Sci U S A 1983;80:560–3.
19. Gissmann L, deVilliers EM, zur Hausen H. Analysis of human genital warts (condylomata acuminata) and other genital tumors for human papillomavirus type 6 DNA. Int J Cancer 1982;29:143–6.
20. Shah KV, Stern WF, Shah FK, Bishai D, Kashima HK. Risk factors for juvenile onset recurrent respiratory papillomatosis. Pediatr Infect Dis J 1998;17:372–6.
21. Kashima HK, Shah F, Lyles A, Glackin R, Muhammad N, Turner L, et al. A comparison of risk factors in juvenile-onset and adult-onset recurrent respiratory papillomatosis. Laryngoscope 1992;102:9–13.
22. Shah K, Kashima H, Polk BF, Shah F, Abbey H, Abramson A. Rarity of cesarean delivery in cases of juvenile-onset respiratory papillomatosis. Obstet Gynecol 1986;68:795–9.
23. Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med Bull 1998;45:320–3.
24. Andersen TF, Madsen M, Jorgensen J, Mellemkjoer L, Olsen JH. The Danish National Hospital Register. A valuable source of data for modern health sciences. Dan Med Bull 1999;46:263–8.
25. Szklo M, Nieto FJ. Epidemiology beyond the basics. Gaithersburg, Maryland: Aspen Publishers Inc., 2000.
26. Cox DR, Oates D. Analysis of survival data. London: Chapman and Hall, 1984.
27. Puranen M, Yliskoski M, Saarikoski S, Syrjanen K, Syrjanen S. Vertical transmission of human papillomavirus from infected mothers to their newborn babies and persistence of the virus in childhood. Am J Obstet Gynecol 1996;174:694–9.
28. Sedlacek TV, Lindheim S, Eder C, Hasty L, Woodland M, Ludomirsky A, et al. Mechanism for human papillomavirus transmission at birth. Am J Obstet Gynecol 1989; 161:55–9.
29. Watts DH, Koutsky LA, Holmes KK, Goldman D, Kuypers J, Kiviat NB, et al. Low risk of perinatal transmission of human papillomavirus: Results from a prospective cohort study. Am J Obstet Gynecol 1998;178:365–73.
30. Fonnest G, de la Fuente Fonnest I, Weber T. Neonatal herpes in Denmark 1977–1991. Acta Obstet Gynecol Scand 1997;76:355–8.
31. Gibbs RS, Amstey MS, Lezotte DC. Role of cesarean delivery in preventing neonatal herpes virus infection. JAMA 1993;270:94–5.
32. Brown ZA, Selke S, Zeh J, Kopelman J, Maslow A, Ashley RL, et al. The acquisition of herpes simplex virus during pregnancy. N Engl J Med 1997;337:509–15.
33. Amann ST, Fagnant RJ, Chartrand SA, Monif GR. Herpes simplex infection associated with short-term use of a fetal scalp electrode. A case report. J Reprod Med 1992;37:372–4.
34. Friis S, Kjaer SK, Frisch M, Mellemkjaer L, Olsen JH. Cervical intraepithelial neoplasia, anogenital cancer, and other cancer types in women after hospitalization for condylomata acuminata. J Infect Dis 1997;175:743–8.
35. Wang X, Zhu Q, Rao H. Maternal-fetal transmission of human papillomavirus. Chin Med J (Engl) 1998;111:726–7.
36. Armbruster-Moraes E, Ioshimoto LM, Leao E, Zugaib M. Presence of human papillomavirus DNA in amniotic fluids of pregnant women with cervical lesions. Gynecol Oncol 1994;54:152–8.