Recurrent respiratory papillomatosis (RRP) is a devastating albeit rare disease in which papillomas of the airway cause hoarseness and airway obstruction.
The disease is caused by human papillomavirus 6 or 11, the same subtypes responsible for the development of genital warts and low-risk cervical intraepithelial neoplasia of the genitourinary tract.
The age of disease onset and clinical course are highly variable.
Surgical therapy for RRP requires a team approach with otolaryngologists, anesthesia providers, and operating room personnel working together in a facility properly equipped to manage difficult airways.
In addition to surgical debridement accomplished with the use of microdebriders, pulsed-dye and carbon dioxide lasers, and microlaryngoscopy “cold steel” techniques, multiple medical modalities have been used without consistently effective results.
Parental support and education are invaluable adjuncts to the safe care of children with RRP.
The recent introduction of a quadrivalent human papillomavirus vaccine offers hope for prevention of transmission of the virus to neonates and may significantly reduce the future incidence of RRP and oropharyngeal cancers.
Recurrent respiratory papillomatosis (RRP) is the most common benign neoplasm of the larynx in children. Despite its benign histology, RRP has potentially fatal consequences and is often difficult to treat because of its tendency to recur and spread throughout the respiratory tract. Long neglected from an epidemiologic standpoint, initiatives to better understand this disease process have been launched in the United States through coordination between the Centers for Disease Control and Prevention (CDC) and pediatric otolaryngologists worldwide.
This chapter discusses the etiology, immunology, epidemiology, and transmission of RRP. Clinical features that include pertinent aspects of the history, physical examination, airway endoscopy, and other considerations are highlighted. Management principles for surgical and nonsurgical treatments and the indications for the use of adjuvant therapies are discussed along with their results. Ongoing research initiatives and promising strategies to improve our understanding of this frustrating disease are also reviewed, including the potential for a quadrivalent human papilloma virus (HPV) vaccine to reduce the incidence.
RRP is a disease of viral etiology, caused by HPV types 6 and 11, and it is characterized by the proliferation of benign squamous papillomas within the aerodigestive tract. Although it is a benign disease, RRP has potentially fatal consequences because of its involvement of the airway and because of the risk, albeit low, of malignant conversion. In addition to the emotional burden to the patients and their families associated with the need for repeated surgery, the economic cost of this relatively rare chronic disease is high, having been estimated at $150 million annually.
RRP is both the most common benign neoplasm of the larynx among children and the second most frequent cause of childhood hoarseness. The disease is often difficult to treat because of its tendency to recur and spread throughout the aerodigestive tract. Although it most often involves the larynx, RRP may involve the entire aerodigestive tract. The course of the disease is variable; some patients experience spontaneous remission, whereas others may suffer from aggressive papillomatous growth and respiratory compromise that requires multiple surgical procedures through many years.
RRP can have its clinical onset during either childhood or adulthood; it affects people of any age, with the youngest patient identified at 1 day of life and the oldest at 84 years. Two distinct forms of RRP are generally recognized: a juvenile, or aggressive, form and an adult, or less aggressive, form. The aggressive form, although most prevalent in children, can also occur in adults. Children whose RRP was diagnosed at younger ages (<3 years) have been found to be 3.6 times more likely to have more than four surgical procedures per year and almost two times more likely to have two or more anatomic sites affected than children whose RRP was diagnosed at older ages (>3 years). Similarly, children with disease progression are generally diagnosed at younger ages than those who remain stable or become disease free. In most pediatric series, the delay in diagnosis from the time of onset of symptoms averages about 1 year. In 75% of children with RRP, diagnosis was made before the fifth birthday.
The true incidence and prevalence of RRP are uncertain. Numerous studies have been performed to elucidate the true incidence of RRP. It is estimated that between 80 and 1500 new cases of childhood-onset RRP occur in the United States each year. Campisi created a national database to incorporate all children (<14 years) with RRP in Canada treated by pediatric otolaryngologists; this study estimated the national incidence of juvenile-onset RRP (JORRP) from 1994 to 2007 to be 0.24 per 100,000 with a prevalence of 1.11 per 100,000. In a Danish study that incorporated 50% of the population of that country, the overall incidence of RRP was 3.84 cases per 100,000; the rate among children in that study was 3.62 per 100,000, whereas the adult-onset cases occurred at a rate of 3.94 per 100,000. Interestingly, a pilot study of a large database of publicly and privately insured patients in the United States consistently showed that RRP incidence was higher in publicly insured patients compared with those with private insurance (3.21 vs. 1.98 per 100,000, respectively). Whereas the incidence among children in the United States is estimated at 4.3 per 100,000 children, the incidence among adults is 1.8 per 100,000. These figures are comparable with those found in the Danish survey. The National Registry of children with RRP, composed of the clinical practices at 22 pediatric otolaryngology sites, calculates a mean number of procedures of 19.7 per child with an average of 4.4 procedures per year. Based on the incidence data, this translates into more than 10,000 surgical procedures annually for children with RRP in the United States.
Virology of Human Papillomavirus
The HPVs are members of a large family of papillomaviruses that infect vertebrates ranging from birds to humans, and they cause epithelial neoplasms that can be benign or malignant. These viruses are designated by their natural host species (e.g., bovine papillomavirus, murine papillomavirus, and HPV). Each papillomavirus is specific for its host species, and this specificity is thought to be absolute. Within each species, similar types of papillomavirus exhibit specificity for epithelial tissues of different sites, such as the oral mucosa, genital mucosa, or skin. In humans, this tissue specificity is less absolute, and some HPV types exhibit more of a preference for certain tissues. HPV, belonging to the Papovaviridae family, is a small deoxyribonucleic acid (DNA)–containing, nonenveloped, icosahedral (20-sided), capsid virus with a double-stranded circular DNA that is 7900 base-pairs long. HPV is epitheliotropic and infects epithelial cells. The HPVs have been grouped on the basis of shared genetic code homology; viruses that share less than 90% identity in specific regions of the virus are defined as separate types. On this basis, the HPVs are numbered to distinguish them, and nearly 100 different HPV types have been identified. Grouping HPV types based on their DNA homology has allowed us to identify closely related types. Functionally, these groupings correlate with their tissue preference as well as similar pathophysiology.
Until the 1990s, HPV had been suspected but not confirmed as the causative agent in RRP. This uncertainty resulted from an inability to culture the virus in vitro and from the failure to demonstrate viral particles consistently in papilloma lesions using electron microscopy or HPV antibodies. With the advent of molecular probes, HPV DNA has been identified in virtually every papilloma lesion studied. The most common types identified in the airway are HPV 6 and HPV 11—the same types responsible for more than 90% of genital condylomas. Specific viral subtypes may be correlated with disease severity and clinical course. Children infected with HPV 11 appear to have more aggressive papilloma growth, which results in a more obstructive airway course early in the disease and a greater likelihood of undergoing tracheotomy to maintain a safe airway.
In addition to the HPV group that includes HPVs 6 and 11, two other major groups of HPV are associated with mucosal lesions in the aerodigestive and genital tracts. HPVs 6 and 11, responsible for the majority of RRPs, are members of a group believed to have a low malignant potential compared with some other groups. In contrast, the group that contains HPV 16 and HPV 18 is associated with malignancies in the genital and aerodigestive tracts. The group that contains HPV 31 and HPV 33 exhibits malignant potential that lies somewhere in between.
HPV is thought to infect stem cells within the basal layer of mucosa. After infection of the stem cells, the viral DNA can either be actively expressed or it can exist as a latent infection in mucosa that remains clinically and histologically normal. To produce viral proteins or to replicate the virus, the viral DNA must somehow reactivate the host replication genes. The viral genome consists of three regions: an upstream regulatory region and the two regions named according to the phase of infection in which they are expressed—early (E) and late (L) regions. The E-region genes are involved in the replication of the viral genome, interaction with the host cell intermediate filaments, and transforming activities; they are potential oncogenes, depending on the HPV type. The L-region genes encode the viral structural proteins. The induction of cellular proliferation is a fundamental property of HPV, although the mechanism of action remains unclear. We are slowly gathering information regarding the interaction of viral gene products with cellular proteins. For example, several of the viral E-region gene products have been shown to bind and inactivate certain cellular tumor-suppressor proteins. Conversely, HPV has been shown to activate the epidermal growth factor (EGF) receptor pathway known to be associated with proliferation of epithelial cells. Thus there are likely several mechanisms by which HPV induces cellular proliferation in the aerodigestive mucosa.
Histologically, this mucosal proliferation results in multiple “fronds” or fingerlike projections with a central fibrovascular core covered by stratified squamous epithelium ( Fig. 25-1 ). When papillomas are microscopic, they can assume a superficial-spreading configuration that has a velvety appearance. When they exhibit a more macroscopic or exophytic growth pattern, they appear grossly as “cauliflower” or “grapelike” projections ( Figs. 25-2 and 25-3 ). Papilloma lesions may be sessile or pedunculated and often occur in irregular exophytic clusters. Typically, the lesions are pinkish to white in color. Iatrogenic implantation of papilloma may be preventable by avoiding injury to nondiseased squamous or ciliated epithelium adjacent to areas of frank papilloma. Ciliated epithelium undergoes squamous metaplasia when exposed to repeated trauma and is replaced with nonciliated epithelium that creates an iatrogenic squamociliary junction. This may also explain the observation that RRP flourishes in the presence of uncontrolled gastroesophageal reflux. Most RRPs do not exhibit dysplasia, abnormal mitoses, or hyperkeratosis. Without exception, RRP exhibits delayed maturation of the epithelium and results in a significantly thickened basal cell layer and nucleated cells in the superficial layers. This is thought to be in part because of the interaction of HPV gene products with the EGF receptor pathway. The sum result of these cellular effects is that although HPV-infected cells do not rapidly divide, a disproportionate increase occurs in the number of dividing basal cells. Thus expansion of the RRP tissue mass may occur very rapidly because of the large number of dividing cells.
During viral latency, very little viral ribonucleic acid (RNA) is expressed. Even so, HPV DNA can be detected in normal-appearing mucosa in RRP patients who have been in remission for years, and unknown stimuli can result in reactivation and clinical recurrence following years of remission. Thus activation of viral expression can occur any time after the establishment of latent infection. Gene products of early region genes—E6, E7, and possibly E5—are required for papilloma induction, but the details of the mechanism of HPV activation are unknown. To “cure” RRP, it is necessary to modulate the host response to the virus and, ideally, to eliminate latent infection; to date, this has not been achieved.
It is likely that the host immune system plays an important role in the pathogenesis of HPV-induced lesions. Both the humoral and the cellular immune responses may be compromised in children with RRP, and the patient’s immunocompetence may be associated with the clinical course of the disease. The role of cytokines—such as interleukins 2, 4, and 10—and expression of the major histocompatibility complex antigens in the dysfunction of the cell-mediated immune response in children with RRP have been demonstrated.
An association between cervical HPV infection in the mother and the incidence of RRP has been well established. However, the precise mode of transmission is still not clear. The universality of HPV in the lower genital tract rivals that of any other sexually transmitted disease in humans. It is estimated that at least 1 million cases of genital papillomas occur per year in the United States. These are most commonly manifested as condylomata acuminata that involve the cervix, vulva, or other anogenital sites in women or the penis of male sexual partners of affected women. Colposcopic (subclinical) changes are seen in about 4% of women, whereas DNA-positive biopsies without a visible lesion are seen in 10% of women. HPV antibody positivity (without DNA or a clinical lesion) is estimated in 60% of women (81 million). HPV has been estimated to be present in the genital tract of as many as 25% of all women of child-bearing age worldwide. One study reported that the incidence of HPV infections in sexually active young college women is highest, with a cumulative incidence of 43% during a 36-month period. Clinically apparent HPV infection has been noted in 1.5% to 5% of pregnant women in the United States. More than 30% of American women are currently infected with HPV with 7.5 million 14- to 24-year-olds currently infected and 25% of women younger than 60 years of age infected at any given time. Up to 90% of lesions are undetectable clinically at 2 years. The highest prevalence is in women 20 to 24 years of age, and more than 50% of women will initially acquire HPV within 4 years of their first sexual intercourse. As in RRP, HPV 6 and HPV 11 are the most common subtypes identified in cervical condylomas.
Vertical transmission that occurs during delivery through an infected birth canal is presumed to be the major mode of transmitting the infection in children, whereas in utero and transplacental transfer of HPV, sexual abuse, and direct contact are thought to play a minor role. The support for vertical transmission lies in the fact that overt maternal condylomas are seen in more than 50% of mothers who give birth to children with RRP. The same subtypes (HPVs 6 and 11) are involved, and cesarean delivery of children seems to be preventive to some extent.
Patients with childhood-onset RRP are more likely to be first born and vaginally delivered than are control patients of a similar age. Kashima and others hypothesized that primigravid mothers are more likely to have a long second stage of labor and that the prolonged exposure to the virus leads to a higher risk of infection in the first-born child. They also suggested that newly acquired genital HPV lesions are more likely to shed virus than long-standing lesions, which accounts for the higher incidence of papilloma disease observed among the offspring of young mothers of low socioeconomic status—the same group that is most likely to acquire sexually transmitted diseases such as HPV.
Despite the close association between maternal condylomas and the development of RRP, only a small portion of children exposed to genital condylomas at birth actually go on to development of clinical RRP. Although HPV could be recovered from the nasopharyngeal secretions of 30% of infants exposed to HPV in the birth canal, the number of infants expected to manifest evidence of RRP is only a small fraction of this population. Clearly, other factors such as patient immunity, timing, length and volume of virus exposure, and local trauma must be important determinants in the development of RRP. Whereas cesarean section delivery would seem to reduce the risk of transmission of the disease, this procedure is associated with higher rates of morbidity and mortality for the mother and a much higher economic cost than elective vaginal delivery. Furthermore, reports of neonatal papillomatosis following cesarean delivery suggest that in at least some cases, transmission may occur in utero. However, with such a high rate of subclinical maternal HPV infection and such a low rate of actual new cases of childhood RRP, elective cesarean delivery as a means of preventing RRP is currently not practical or recommended. The risk of a child contracting the disease from a mother who has an active genital condyloma lesion during vaginal delivery is only approximately 1 in 231 to 400. The characteristics that distinguish this one child from the other 230 to 399 remain elusive. In summary, a better understanding of the risk factors associated with RRP is needed before the efficacy of cesarean delivery or other preventive measures can be fully assessed.
The potential exists for a breakthrough in prevention of RRP with the emergence of widespread use of the quadrivalent HPV vaccine (Gardasil; Merck) in the developed world. The vaccine has been used for the prevention of cervical cancer, adenocarcinoma in situ, and intraepithelial neoplasia grades 1 through 3; vulvar and vaginal intraepithelial neoplasias grades 2 and 3; and genital warts associated with HPVs 6, 11, 16, and 18. A second-generation vaccine that contains five additional HPV subtypes is currently under investigation.
The ability of the quadrivalent vaccine to prevent cervical and genital disease associated with HPVs 6, 11, 16, and 18 was established in the phase 3 FUTURE I and II trials, and their immunogenicity in the target group of vaccine recipients was established in immunogenicity bridging trials. FUTURE I and II studies were conducted in women aged 15 to 26 years, known to be at highest risk for HPV acquisition. In FUTURE I, the quadrivalent vaccine was 100% effective in preventing cervical intraepithelial neoplasia or worse, genital warts, and vulvovaginal neoplasia. In FUTURE II, the vaccine was 100% effective in preventing cervical intraepithelial neoplasia associated with HPV 16 or 18. Both FUTURE I and FUTURE II were conducted in women in the age range at highest risk for HPV acquisition. However, it appears that the vaccine will be most effective if administered to individuals who have not yet become sexually active. Immunogenicity bridging studies established that immunogenicity among younger girls was equal to, if not superior to, the response among 16- to 23-year-old women, which suggests that the quadrivalent HPV vaccine is immunogenic in this population and is thus likely to be effective in preventing disease. Similarly, a separate study in 9- to 15-year-olds established that in this younger population, immunogenicity lasts at least 18 months.
Accordingly, the CDC’s Advisory Committee on Immunization Practices has recommended that all boys and girls 11 to 12 years of age (prior to the age of sexual debut) be vaccinated. Vaccination has also been recommended in boys and girls under age 25 who missed out on earlier opportunities to receive the vaccine. The vaccine’s efficacy as a therapeutic option is markedly more limited, and although the antibody response may be boosted by administration, surveillance for future lesions should be continued. Based upon the pivotal clinical studies, the quadrivalent vaccine is predicted to reduce the incidence, morbidity, and mortality of cervicovaginal HPV disease.
A bivalent vaccine has also been licensed that provides protection against HPVs 16 and 18 but not 6 and 11. Early phase 2 data for this vaccine suggest that it is 100% effective in preventing incident and persistent cervical HPVs 16 and 18 infections in the according-to-protocol sample and that it was found to be 93% effective in preventing disease related to HPV 16 or 18 in the intention-to-treat analysis; efficacy against disease was not presented for the according-to-protocol cohort. This vaccine’s efficacy against HPVs 16 and 18 suggests that, like the quadrivalent vaccine, it may reduce the incidence of HPV-associated head and neck cancers. However, because the bivalent vaccine does not protect against HPVs 6 and 11, it will not likely affect the vertical transmission of HPVs 6 or 11 from mother to child.
Even with a successful HPV vaccination program, Papanicolaou screening and postlicensure safety monitoring will still be needed. The quadrivalent vaccine may still benefit women with existing HPV infection with one virus type by boosting the antibody response and increased protection from infection by other HPV vaccine types.
The often overlooked benefits of nearly universal vaccination against HPV may be a concomitant decrease in the incidence of oropharyngeal squamous cell carcinoma and RRP in all age groups. HPV has also been recognized to play a significant role in the development of oropharyngeal squamous cell carcinoma. HPV 16–positive oropharyngeal squamous cell carcinoma commonly occurs in younger patients who do not have a history of smoking or alcohol abuse. The tonsil is the most common oropharyngeal subsite. Because tens of thousands of new diagnoses of oropharyngeal squamous cell carcinoma are made each year, a preventative vaccination could have enormous epidemiologic impact in the next several decades. Thus, widespread use of the quadrivalent HPV vaccine promises to dramatically reduce the morbidity and mortality of cervical cancer and to drastically reduce the incidence of genital warts. If the vaccine is as effective in preventing HPV infection of the oral cavity as it is of the cervix and genital tract, vaccination could be expected to reduce the incidence of HPV-associated oropharyngeal cancers by as much as 30%. Thus vaccination against HPV offers additional benefits above and beyond prevention of cervical cancers, some of which are unique to the quadrivalent vaccine.
Universal vaccination with the quadrivalent vaccine also holds promise to eliminate the maternal and paternal reservoir of HPV and to lead to a near eradication of RRP caused by HPVs 6 and 11. With widespread vaccination, a reduction in the incidence of cervicovaginal HPV would be projected to result in many fewer cases of RRP that result from vertical transmission. Having only recently been introduced, the long-term results of widespread vaccination have yet to be realized. Several obstacles have limited universal vaccination, including moral objections and cost. Critics have objected that vaccination programs may promote unprotected and dangerous sexual activity, concerns that have not been supported by the scientific literature regarding AIDS awareness and contraception. Public health officials may have justifiable concerns about the merits of HPV vaccination based on the financial and logistic burdens that may be imposed on families and schools, and they also may legitimately worry about adverse events that might occur with large-scale vaccination programs that involve children. The cost to fully immunize all 11- and 12-year-old girls and boys is estimated to exceed $1.7 billion per year, although this expense will be offset by a predicted decrease of $6.5 billion in health care costs annually. Such costs will have to be covered by the Vaccine for Children program for the uninsured, Medicaid, and State Children’s Health Insurance Program families and by most private insurers. Given the rarity of RRP, multiinstitutional studies will need to be continued for many years to observe any decrease in the incidence of RRP secondary to widespread vaccination.
Inasmuch as the vocal fold is usually the first and predominant site of papilloma lesions, hoarseness is the principal presenting symptom in RRP. The child’s voice may be described as hoarse or weak from the time of birth. Particularly in very young children, changes in voice may go unnoticed. Stridor is often the second clinical symptom to develop, beginning as an inspiratory noise and becoming biphasic with progression of the disease. Less commonly, chronic cough, recurrent pneumonia, failure to thrive, dyspnea, dysphagia, or acute life-threatening events may be the presenting symptoms. The duration of symptoms before diagnosis varies. Not uncommonly, a mistaken diagnosis of asthma, croup, allergies, vocal nodules, or bronchitis is entertained before a definitive diagnosis is made. However, the initial presentation in infants, whose airway dimensions are small, may be with acute respiratory distress during an otherwise routine upper respiratory tract infection. The natural history of RRP is highly variable. After presentation, the disease may undergo spontaneous remission or it may persist in a stable state that requires only periodic surgical treatment. At the other extreme, RRP may become extremely aggressive and may require frequent surgical treatment (every few days to weeks) and may prompt early institution of medical adjuvant therapy. A waxing-waning clinical course of remissions and exacerbations is common for RRP.
Because of the rarity of RRP and the slowly progressive nature of the disease, some cases may go unrecognized until respiratory distress results from papillomas that obstruct the airway. The result is a relatively high need for tracheotomy to be performed in these children. Shapiro and others noted that RRP tracheotomy patients presented at a younger age and with more widespread disease, which often involved the distal airway before tracheotomy. In their experience with 13 patients, they did not believe that the tracheotomy itself led to spread of disease outside the larynx. In the CDC’s National RRP Registry, children with tracheotomy were initially diagnosed with RRP at a younger age (2.7 years) than those without a tracheotomy (3.9 years). Others have suggested that tracheotomy may activate or contribute to the spread of disease lower in the respiratory tract. Cole and colleagues reported that tracheal papillomas developed in half of their tracheotomy patients and that despite attempts to avoid this procedure, 21% of their patients still required a long-term tracheotomy. Prolonged tracheotomy and the presence of subglottic papillomas at the time of tracheotomy have been associated with an increased risk of distal tracheal spread. Most authors agree that tracheotomy is a procedure to be avoided unless absolutely necessary. Interestingly, Boston and colleagues noted successful laryngotracheal reconstruction in a series of children with subglottic stenosis and RRP. When a tracheotomy is unavoidable, decannulation should be considered as soon as the disease is managed effectively with endoscopic techniques. Children with bronchopulmonary dysplasia who require prolonged endotracheal intubation may also be at increased risk for development of RRP. Through interruption of the continuous respiratory mucosal surface, an endotracheal tube may have the same risk for distal spread via mechanical dissemination-implantation of RRP as tracheotomy. Several authors have noted an association between RRP caused by HPV 11 (as opposed to HPV 6) and distal spread of papilloma.
Extralaryngeal spread of respiratory papillomas has been identified in approximately 30% of children and in 16% of adults with RRP. The most frequent sites of extralaryngeal spread were, in decreasing order of frequency, the oral cavity, trachea, and bronchi ( Figs. 25-4 and 25-5 ). Pulmonary papilloma lesions begin as asymptomatic noncalcified peripheral nodules. These lesions eventually enlarge to undergo central cavitation and central liquefactive necrosis with air-fluid levels on a computed tomography scan ( Fig. 25-6 ). These patients present clinically with recurrent bronchiectasis, pneumonia, and declining pulmonary status. The clinical course of the pulmonary spread of RRP is insidious and may progress through the years, but it eventually manifests in respiratory failure because of destruction of lung parenchyma. For this reason, the finding of pulmonary lesions in a patient with RRP is a grave development with no currently available treatment modality that has shown more than anecdotal promise. Furthermore, pulmonary dissemination is anecdotally associated with a higher risk of malignant transformation of RRP.
Malignant transformation of RRP into squamous cell carcinoma has been documented in several case reports. A total of 26 patients were identified as having progressed to squamous cell carcinoma in the task force survey. Dedo and Yu reported malignant transformation in 4 of 244 RRP patients (1.6%) treated over two decades. When death occurs in a patient with RRP, it is usually as a complication of frequent surgical procedures or is caused by respiratory failure because of distal disease progression. RRP that presents in the neonatal period is thought to be a negative prognostic factor, with a greater likelihood for mortality and need for tracheotomy.
Persistent or progressive stridor and dysphonia, with the possible development of respiratory distress, are the most consistent signs and symptoms of RRP in children (see the Practice Pathway Flow Chart, Fig. 25-7 ). In the absence of severe respiratory distress, a careful history should be obtained. Information regarding the time of onset of symptoms; possible airway trauma, including a history of previous intubation; and characteristics of the cry are obviously important. Hoarseness, although a common and often benign clinical complaint in young children, always indicates some abnormality of structure or function. Because of the precision of laryngeal mechanics, hoarseness may result from a remarkably small lesion and may thus be an early sign in the course of a disease process. On the other hand, if the lesion’s origin is remote from the vocal cords, hoarseness may present as a late sign. Although histologically the same lesion, a papilloma that produces hoarseness in one patient may produce stridor and obstruction in another, depending on the size and location of the lesion. The quality of the voice change may give only limited clues to its etiology, whereas other characteristics such as age of onset, rate of progression, associated infection, history of trauma or surgery, and the presence of respiratory or cardiac distress may be of much greater significance. A low-pitched, coarse, fluttering voice suggests a subglottic lesion, whereas aphonia, a high-pitched, cracking voice, or a breathy voice suggests a glottic lesion. Associated high-pitched stridor also suggests a glottic or subglottic lesion. Although stridor that has been present since birth is more often associated with laryngomalacia, subglottic stenosis, vocal cord paralysis, or a vascular ring, it should be recognized that papillomatosis may also present in neonates. Associated symptoms such as feeding difficulties, allergic symptoms, vocal abuse, and the presence of hereditary congenital anomalies may help distinguish RRP from alternative diagnoses, including vocal fold nodules, vocal fold paralysis, subglottic cysts, subglottic hemangioma, and subglottic stenosis. In the absence of any history that suggests these lesions, review of the perinatal period may reveal a history of maternal or paternal condylomas. If the onset of stridor and dysphonia is gradual and progressive through weeks or months, neoplastic growth that compromises the airway must be considered and investigated.
Certainly not every child with a hoarse voice or cry merits investigation beyond an assessment of symptoms. However, in the presence of hoarseness with respiratory distress, tachypnea, decreased air entry, tachycardia, cyanosis, dysphagia, chronic cough, failure to thrive, recurrent pneumonia, or dysphagia, the larynx must be visualized, and a firm diagnosis of the cause of hoarseness must be made. Any child with slowly progressive hoarseness merits investigation, and the clinician should not wait until total aphonia or airway problems occur.
Children who present with symptoms consistent with RRP must undergo a thorough and organized physical examination. The child’s respiratory rate and degree of distress must first be assessed. The physician should observe the child for tachypnea or onset of fatigue that may indicate impending respiratory collapse. The child should be observed for flaring of the nasal ala and the use of accessory neck or chest muscles. Increasing cyanosis and air hunger may cause the child to sit with the neck hyperextended in an attempt to improve airflow. If a child is gravely ill, additional examination should not be undertaken outside the operating room (OR), the emergency department, or the pediatric intensive care unit, where resuscitation equipment for intubation of the airway, endoscopic evaluation, and possible tracheotomy is readily available. In the stable, well-oxygenated child, additional examination can proceed. The most important part of the examination is auscultation with the aid of a stethoscope. The physician should listen over the nose, open mouth, neck, and chest to help localize the probable site of the respiratory obstruction. We prefer to pull the bell off the stethoscope and listen over these areas with the open tube. The respiratory cycle, which is normally composed of a shorter inspiratory phase and a longer expiratory phase, should then be observed. Stridor of a laryngeal origin is most often musical and may begin as inspiratory, but it will progress to biphasic stridor as airway obstruction worsens. Infants with stridor should be placed in various positions to elicit any changes in the stridor. A child with RRP would not be expected to demonstrate much change in the stridor with position change, in contrast to infants with laryngomalacia, a vascular ring, or a mediastinal mass. Pulse oximetry can add objective information on the child’s respiratory status. In the stable patient in whom asthma is a likely diagnosis, pulmonary function testing may also be helpful.
The preoperative diagnosis of RRP is best made with a flexible fiberoptic nasopharyngoscope. Careful, sequential inspection of the pharynx, hypopharynx, larynx, and subglottis provides the critical information necessary to make the diagnosis of RRP. It also allows estimation of lumen size and vocal cord mobility and determines the urgency of operative intervention. Advances in instrumentation of flexible nasopharyngoscopes have resulted in instruments as small as 1.8 mm in diameter, which when combined with the newest low-light, distal-chip endoscopic cameras allow passage in even the smallest newborns. Even the smallest diameter scopes provide excellent images that can be viewed on a video monitor and recorded for later review. Topical decongestant and local anesthesia can be applied by spray, dropper, or pledget. Oxymetazoline is the decongestant of choice because of its lack of cardiac side effects. Topical tetracaine may be used to enhance patient cooperation, but the dosage must be critically monitored in small infants to avoid cardiotoxicity.
Most clinicians find that visualization with the flexible nasopharyngoscope is far superior to that obtained with indirect mirror laryngoscopy in young children. Patient cooperation, however, is required even with good topical anesthesia. In infants, this is not a major issue, because they can easily be restrained in an upright, sitting position in an adult’s lap for evaluation. Likewise, most children older than 6 or 7 years of age can be “talked into” cooperating for the examination. It is children in this intermediate age group, between 1 and 6 years of age, who may be the most difficult to examine, taxing the patience and skill of even the most experienced clinicians. Although dynamic evaluation can be appreciated when children are spontaneously breathing, endoscopy in the OR under anesthesia is warranted in any child suspected to have RRP who cannot be fully examined in the outpatient setting.
At present, there is no “cure” for RRP, and no single modality has consistently been shown to be effective in its eradication. The current standard of care is surgical therapy with a goal of complete removal of papillomas and preservation of normal structures. In patients in whom anterior or posterior commissure disease or highly aggressive papillomas are present, the goal may be subtotal removal with clearing of the airway. It is advisable to debulk as much disease as possible while preserving normal morphology and anatomy and preventing the complications of subglottic and glottic stenosis, web formation, and resulting airway stenosis.
Until recently, the carbon dioxide (CO 2 ) laser has been favored over cold instruments in the treatment of RRP that involves the larynx, pharynx, upper trachea, and nasal and oral cavities. When coupled to an operating microscope, the laser vaporizes the lesions with precision and causes minimal bleeding. When used with a no-touch technique, it minimizes damage to the vocal cords and limits scarring. Dedo reported on a series of 244 patients with RRP treated with the CO 2 laser used every 2 months and achieved “remission” in 37%, “clearance” in 6%, and “cure” in 17%. The CO 2 laser has an emission wavelength of 10,600 nm and converts light to thermal energy. The CO 2 laser delivers energy that is absorbed by intracellular water, effectively vaporizing the cells. It provides a controlled destruction of tissues with vaporization of water, and it cauterizes tissue surfaces. The smoke plume contains water vapor and vaporized tissue material. Its drawbacks relate to safety of the OR personnel, patient, and surgeon. The laser may glance off metal used in the suspension of the larynx and injure eyes or skin that it hits. A misfire may also hit areas on the patient that are not protected by a wet towel to absorb the laser energy. In addition, the laser smoke, or “plume,” has been found to contain active viral DNA—a potential source of infection. Smoke evacuators are necessary for the safety of exposed individuals. Most importantly, because the laser generates heat, ignition of the endotracheal tube may occur with an inadvertent laser strike. In the oxygen-rich environment provided by anesthetic gases, this can lead to explosion or fire in the airway. The surgeon who is not aware of injury to deeper tissue layers with injudicious laser usage may cause unacceptable scarring and subsequent abnormal vocal fold function. Inappropriate and aggressive use of the laser may also cause injury to tissues that are not affected and can create an environment suitable for implantation of viral particles. Use of the CO 2 laser can also result in delayed local tissue damage, which may be related to the total number of laser surgeries and the severity of RRP disease.
Although the CO 2 laser allows for precision of surgery and excellent hemostasis, multiple procedures are often necessary. Frequent interval laser laryngoscopies are recommended in an attempt to avoid tracheotomy and to permit the child to develop good phonation with preservation of normal vocal cord anatomy. The latest generation of laser microspot micromanipulators enables the surgeon to use a spot size of 250 mm at a 400-mm focal length and 160 mm at a 250-mm focal length.
The use of the CO 2 laser on the true vocal folds must be judicious given the potential for significant postoperative scar tissue formation from unrecognized heat transfer. To minimize the risk of scar formation in the true vocal folds, cold-steel excision can be used successfully following the principles of phonomicrosurgery, submucosal dissection, and microinstrumentation. This approach may have treatment advantages over CO 2 laser surgery, especially in the adult RRP patient. In their initial series, Zeitels and Sataloff report recurrence of papillomas in none of the six adults at 2-year follow-up after resection for primary disease. Of those who presented with recurrent papillomatosis, 6 of 16 (38%) had continued recurrence after their microflap procedure.
The potassium-titanyl-phosphate (KTP) laser has also garnered attention for its use in papillomatosis. The KTP laser delivers light at a wavelength of 532 nm. At this wavelength, oxyhemoglobin absorbs the energy contained within the laser. The laser can be used in the office setting, because fibers as narrow as 0.3 mm can be used through the working channel of a flexible bronchoscope. This laser has also been used to manage vascular ectasia of the glottis, because it is able to initiate vascular coagulation while not destroying the superficial layer of the lamina propria. In a review by Zeitels, KTP laser use resulted in disease regression of greater than 75% in 18 of 29 patients. Proponents of the KTP laser favor its use because of its lower cost and the ability to use narrow-gauge fibers, which limits some of the mechanical problems present with use of other lasers. The pulsed-dye laser (PDL) uses light and a lasing medium that can be varied based on the target chromophore, the portion of the molecule responsible for its color, and it is tuned to a specific wavelength at which maximum absorption of energy can occur. The chromophore for blood is 577 nm, and the PDL works at a wavelength of 585 nm. The benefits of a PDL include the ability to induce microvascular coagulation while preserving overlying epithelium, which is ultimately beneficial in preservation of the voice, while appropriately managing the neoplastic process. Patients with sessile lesions, disease that involves the ventricle, and areas with significant scarring may benefit the most from use of the PDL. Use of the PDL has been popularized because it can be used in outpatient and office settings. Patient comfort with the office-based procedure has been documented in adults: 87% of patients preferred it to similar OR-based upper aerodigestive tract PDL procedures. The benefits of performing these procedures in the office include a potential cost savings of $5,000 per case, once all hospital, anesthesia, and OR costs have been taken into account. The majority of pediatric patients, however, still need to be treated under general anesthesia given cooperation issues and airway maintenance concerns. When coupled with the use of bevacizumab, this modality may offer superior results to the use of ablative techniques alone.
A number of investigators are now replacing their use of the CO 2 laser with the endoscopic microdebrider as a means of quickly debulking laryngeal disease. In a small, randomized study, Pasquale and colleagues observed improved voice quality, less OR time, less mucosal injury, and a cost benefit in direct comparison of the microdebrider and the CO 2 laser. El-Bitar and Zalzal and Patel and others have observed similar improved outcomes with the use of the endoscopic microdebrider. A web-based survey of members of the American Society of Pediatric Otolaryngology found the majority of respondents now favoring the use of “shaver” technology.
Because no therapeutic regimen reliably eradicates the HPV, when there is a question about whether papilloma in an area needs to be removed, it is prudent to accept some residual papilloma rather than risk damage to normal tissue and produce excessive scarring. Even with the removal of all clinically evident papilloma, latent virus remains in adjacent tissue. Therefore the goals of therapy in extensive disease should be to reduce the tumor burden, decrease the spread of disease, create a safe and patent airway, optimize voice quality, and increase the time interval between surgical procedures. Staged papilloma removal for disease in the anterior commissure is appropriate to prevent the apposition of two raw mucosal surfaces to minimize the risk of web formation ( Fig. 25-8 ).