How Enhancing Immunity to Low-Risk HPV Could Cure Recurrent Respiratory Papillomatosis

Abstract

Recurrent respiratory papillomatosis (RRP) is currently treated with repeat surgical resection of papillomatous disease that does not address the fundamental underlying issue of chronic infection with low-risk human papillomavirus. Here, we review the biology and immunology of low-risk human papillomavirus (HPV) infections. Antiviral or antiangiogenic adjuvant treatments similarly address the papillomatous disease itself but do not activate HPV immunity. It is likely that only through immune-mediated clearance of low-risk HPV infection can patients with RRP be cured. In some patients, this occurs spontaneously. In others with more aggressive disease, adjuvant immunotherapy to activate immunity may be needed. Based on current understanding of antiviral immune responses, the only rational strategy to clear HPV-infected epithelial cells is through activation of the T-lymphocyte arm of the adaptive immune response. Translation of immunotherapies that are Food and Drug Administration-approved or under clinical study for cancer, such as immune checkpoint blockade or engineered therapeutic vaccines, may provide a path toward tolerable and efficacious adjuvant immunotherapy for RRP.

Level of Evidence:

NA

INTRODUCTION

Chronic infection with human papillomavirus (HPV) can cause a number of neoplastic disorders. In contrast with a few high-risk HPV subtypes that can cause malignant transformation of infected epithelial cells, greater than 200 low-risk HPV subtypes exist and can cause benign neoplasms of the skin and mucosa.¹ Chronic infection with two common subtypes of low-risk HPV—HPV 6 or 11—are the cause of anogenital warts (> 600,000 cases annually in the United States) and recurrent respiratory papillomatosis (RRP) (~3,000 cases annually).^2,3 RRP manifests as recurrent papillomatous growths in the upper aerodigestive tract that cause significant voice and airway dysfunction. In some patients, papillomatous lesions form in the airways of the lungs, leading to post-obstructive pneumonia that can be fatal. Many individuals (as high as 50% in some populations) are exposed to HPV 6 or 11, but only a few develop chronic infection and papillomatous lesions.^4,5 The underlying genetic and/or environmental risk factors that make an individual susceptible to chronic infection with a low-risk HPV are poorly understood.⁶

Diagnosis of RRP is usually straightforward, based on the presence of recurrent papillomas in the upper aerodigestive tract. Historically, treatment has centered around the repeat surgical debulking of papillomatous disease to maintain a patent airway and a functional voice. RRP is one of the few chronic infections with manifestations primarily treated with repeat surgery. Numerous adjuvant treatments have been used to treat RRP, including antivirals and antiangiogenics. The difficult nature of establishing in vitro or in vivo models of chronic low-risk HPV infection that causes papillomatous disease has made preclinical research slow and challenging.^7,8 As a result, most published research in RRP is clinical research. Because large prospective clinical trials studying new RRP treatment are difficult to fund and accrue, most published clinical research studies are the result of small retrospective, uncontrolled, single-cohort studies using Food and Drug Administration (FDA)-approved drugs with other indications. Many of these studies have contradictory results,^9,10 leading to a lack of widespread acceptance and use of any single adjuvant treatment.

Here, via review of published literature using keywords such as “recurrent respiratory papillomatosis,” “adjuvant treatment,” and “low-risk HPV 6 or 11 immunity,” we review the biology, replication, and persistence of HPV infection, as well as what is understood about the anti-HPV immune responses and how these are dysfunctional in patients with chronic low-risk HPV infections. The objective of this work is to provide a contemporary review of adjuvant therapies for RRP currently utilized in practice or clinical studies, with a focus on how immunity can possibly be harnessed to effectively reduce RRP disease burden. Previously studied antiviral and antiangiogenic treatment for neoplasms caused by low-risk HPV only treat the papillomatous manifestations of the real underlying issue—the chronic low-risk HPV infection. Significant recent advances in the understanding of how reversal of immune dysfunction and enhancement of anti-HPV immunity can be used to treat HPV-associated malignancies may be translatable to disorders caused by low-risk HPV. It is likely that only through immune-mediated clearance of the HPV infection can neoplastic disorders caused by chronic HPV infection, such as RRP, be cured.

HPV INFECTION

In normal stratified squamous epithelial mucosa, basal cells replicate, giving rise to daughter cells that progress through the cell cycle and differentiate as they move into more superficial layers. HPV infection of basal cells that give rise to infected daughter cells establishes a chronic HPV infection.¹¹ Recent evidence suggests that HPV actually enters basal epithelial cells via interaction with epithelial cell surface heparin sulfate proteoglycans and that cells must be in the M phase of mitosis to be susceptible to cellular entry.^12,13 It appears that viral gene amplification occurs once per cell cycle in basal cells. The expression of early (E) genes E6 and E7 can induce uncontrolled cell cycle entry and allow the viral genome to amplify in suprabasal squamous epithelial cell layers.^14,15

Functions of Low-Risk HPV Gene Products

HPVs encode eight genes: E1, E2, E4, E5, E6, E7, L1, and L2.¹⁶ Early (E) genes are transcribed early in the viral replication cycle, utilizing the host cell’s gene expression machinery. Late (“L”) genes are latent proteins are transcribed later in the replication cycle and encode capsid proteins. The E1, E2, E4, L1, and L2 genes are considered core genes, are required for viral replication and virion production, and are highly conserved between different HPV types. The E5, E6, and E7 genes are considered accessory genes and function to modify host cells to better tolerate viral infection and support viral genome replication. Generally, through their respective associations with p53 and retinoblastoma, E6 and E7 drive uncontrolled cell cycle progression. Here, critical differences between E6 and E7 from high-risk and low-risk HPV are observed. E6 and E7 from high-risk HPV bind their respective targets with greater affinity compared to E6 and E7 from low-risk HPV.^14,15,17,18 Binding of p53 by E6 from low-risk HPVs is not enough to induce its degradation.¹⁹ This may be the principle reason that chronic infection with high-risk HPV is sufficient to induce malignant transformation, whereas chronic infection with low-risk HPV is not. Another fundamental difference between high-risk and low-risk HPV is integration. E6 and E7 from high-risk HPV frequently integrates into the host genome, often losing some of the regulatory or apoptotic functions of other early genes (such as E2) that do not integrate, leading to significantly greater expression levels.^20,21 Low-risk HPVs do not integrate, remaining episomal within infected cells.²² Epithelial cells have multiple intracellular innate immune sensors that can detect cytoplasmic DNA; the identification of intrinsic cellular defects that allow episomal low-risk HPV to persist in the cytosol of infected cells remains an area of active study.²³

Although occurring in less than 2% of patients with pulmonary manifestations of RRP, malignant transformation of a previous papillomatous lesion can occur.²⁴ Such rare events are difficult to study; however, case series from carcinoma ex-papilloma events document frequent correlation with HPV 11 (and not HPV 6) infection, viral integration events into the host DNA, and the presence of mutated TP53, suggesting that these events may be mechanistically linked to the transition from papilloma to cancer.^25,26 Whether other mutations events in the host DNA are needed for this transformation requires further study.

SUMMARY ON ANTIVIRAL, ANTIANGIOGENIC, AND OTHER ADJUVANT TREATMENTS FOR RRP

Only one of eight proteins encoded by HPV is an enzyme; thus, HPV relies primarily on the host epithelial cell’s machinery (such as human DNA polymerase) for replication. This is the primary reason that there are no HPV-specific antiviral drugs or small molecule inhibitors: there are few enzymes to inhibit.²⁷ The active metabolite of cidofovir—cidofovir diphosphate—is highly selective for viral DNA polymerase, with only weak activity against human DNA polymerase. Accordingly, cidofovir potently inhibits the replication of viruses that encode a viral DNA polymerase. For reasons that are poorly understood, cidofovir does inhibit the replication of HPVs to some degree, even though they rely on the host cell’s human DNA polymerase for replication.²⁸ For decades, based on initial reports of clinical activity of locally injected cidofovir,⁹ clinicians have been using cidofovir for the adjuvant treatment of RRP. Retrospective studies reporting the presence of absence of clinical benefit with cidofovir have conflicting results,^29–31 and ultimately a Cochrane review (summarizing one randomized controlled study) concluded that no clinical benefit exists.¹⁰ Use of adjuvant intralesional cidofovir remains controversial given the paucity of controlled clinical data yet continued widespread use. Continued use of cidofovir is supported by some retrospective studies suggesting a clinical benefit and evidence that there is little chance of harm with repeat administration.³² Until more definitive and convincing clinical evidence emerges for other forms of adjuvant treatment for RRP, use of intralesional cidofovir will likely continue in select practices.

Given the high vascularity observed in papillomatous lesions, antiangiogenics, particularly the monoclonal antibody (mAb) bevacizumab that targets vascular endothelial growth factor, have been studied as an adjuvant treatment for RRP. The use of locally injected or systemic bevacizumab can induce dramatic regression of lary-ngotracheal papillomatous disease, as reported in several clinical case series.^33–36 Renal toxicity with long-term use of bevacizumab can occur,³⁷ leading to clinical questions about the optimal dosing and duration of treatment of bevacizumab for patients with RRP.³⁸ Another concern is rebound of papillomatous disease upon cessation of bevacizumab treatment given that VEGF inhibition alone does not induce anti-HPV immunity that could lead to elimination of HPV-infected epithelial cells. Prospective clinical trials powered to determine dosing and scheduling of bevacizumab (or the biosimilar) are needed to conclusively establish the safety and clinical activity of this treatment approach. Current adjuvant therapies commonly utilized for the treatment of aggressive RRP are summarized in Figure 1.

Fig. 1.

Schematic of currently utilized adjuvant treatments for aggressive RRP. Patients with aggressive RRP are commonly treated with cidofovir, which blocks viral DNA replication inside infected cells (top panels), bevacizumab that binds vascular endothelial growth factor (VEGF), and blood vessel formation and stability (middle panels), or immunotherapy designed to induce anti-HPV immunity. Based on current understanding, only the induction of anti-HPV immunity can lead to clearance of virally infected basal epithelial cells and cure of RRP.HPV = human papillomavirus; HSPG = heparin sulfate proteoglycans; RRP = recurrent respiratory papillomatosis; VEGF = vascular endothelial growth factor.

Other adjuvant treatments for RRP have been studies clinically. Cox-2 was found to be overexpressed in RRP lesions,³⁹ and a case report documented disease stabilization in a patient with aggressive RRP with combination celecoxib (cyclooxygenase-2 [COX-2] inhibitor) erlotinib (epidermal growth factor receptor [EGFR] inhibitor),⁴⁰ and thus a prospective clinical trial was executed. Preliminary results of this trial indicated that COX-2 inhibition with celecoxib does not provide RRP growth control (NCT00592319). Indole-3-carbinol (I-3-C), similarly with preclinical data demonstrating in vitro and in vivo RRP growth control,⁴¹ was evaluated in a prospective study of adults and children with RRP.⁴² This clinical study demonstrated stabilization of disease with an increased intersurgery interval in a subset of adult and pediatric patients; however, these results have not been followed up in larger studies. Whether adjuvant treatment with celecoxib or I-3-C benefits patients with aggressive RRP remains unclear.

HOW IMMUNITY CONTROLS HPV IN ALMOST EVERYONE

Antibodies bind extracellular targets (e.g., viruses), leading to their elimination before cellular infection. Antibodies cannot get inside cells. The T-lymphocyte (T cell) arm of the adaptive immune system is responsible for detecting and eliminating cells already infected with an intracellular virus. Viral proteins produced inside infected epithelial cells are broken down (like all intracellular proteins), and peptides 8 to 10 amino acids in length are loaded onto major histocompatibility complex (MHC) class I molecules and moved to the surface of the cell.⁴³ Here, T cells “sample” cells in passing, and when a T cell expresses a T cell receptor that recognizes an MHC class I:peptide complex for which it is specific, the T cell kills the infected cell. The specific peptide that induced the response from the T cell is called a T cell antigen. The more proteins a virus produces, the more of a chance that one or more peptides from those proteins can serve as a T cell antigen. This has implications for low-risk HPVs that do not integrate into the host genome, remain episomal, and express most or all of their genes. Compared to high-risk HPVs that integrate E6 and E7 and tend to lose several or all other genes, low-risk HPVs produce a greater diversity of proteins, which may portend a greater chance of producing one or more T cell antigens that would lead to the detection and elimination of infected cells. Anecdotally, this may be one explanation for why fewer patients develop chronic low-risk HPV infections compared to the number of patients that develop chronic high-risk HPV infections.⁴⁴

Because RRP is a result of chronic infection, it can be thought of as an immune disorder. For some reason, the immune system of a patient with RRP does not clear the low-risk HPV infection. The reasons why most individuals clear infection and why others develop chronic infection and possibly RRP remain unclear. Specific immunodeficiencies resulting in increased risk of developing RRP have not been identified,⁴⁵ and patients with RRP in general are not immunocompromised and are not at risk of the types of recurrent infections observed in individuals with genetic defects in antibody or T cell responses. Definitive genetic or epigenetic risk factors have not been identified.^6,46 Further, RRP rarely occurs in related individuals. Whether this could be a problem with clearing HPV infections in general or a problem specific to low-risk HPV is unknown (in patients with certain HLA class II haplotypes)⁴⁷; however, patients with RRP are not at higher risk of developing high-risk HPV infections.⁴⁸

HOW IMMUNITY IS DYSREGULATED IN PATIENTS WITH RRP

Although patients with existing RRP infections appear to have developed insufficient antibody responses to prevent the development of a chronic infection,^4,49,50 it is unclear why some patients fail to initially mount an antibody response sufficient to prevent infection. What is clear, however, is that boosting antibody levels before HPV infection with preventative vaccines, such as Gardasil (Merck, Kenilworth, NJ), dramatically decreases risk of chronic infection. There is no question that this works; prevention of HPV infection with Gardasil has already resulted in dramatically decreased incidence of RRP in Australia, where vaccination is government-mandated.⁵¹ Yet, a patient with RRP already has the HPV infection, and only T cells can eliminate infected cells to cure the underlying problem: chronic HPV infection. Suppression of effective T cell priming and function is what allows chronic HPV infection to persist.

Defective Innate Immunity and T Cell Priming

The ability of innate immune cells, primarily myeloid cells, to cross-present HPV antigen is a critical first step of T cell priming and activation. Myeloid cells in papillomas from patients with RRP are less able to produce type I interferon, which is critical for dendritic cell activation, and the T cell activating cytokines interleukin (IL)-12 and IL-18.^52,53 Natural killer cells also appear to be suppressed within papillomatous lesions from RRP patients due at least in part to lack of positive costimulatory cell surface receptors.⁵⁴

Suppression of Effective T Cell Activity

Once HPV-specific T cell clones develop and traffic into papillomatous lesions, the microenvironment can promote or inhibit their function. CD4+ T cells within papillomas tend to be polarized into a T helper type 2 (Th2) state; have altered intracellular signaling; and produce cytokines such as IL-4, L-10, and IL-13, which suppress the effector activity of CD8+ T cells.⁵⁵ The underlying cause of this Th2 polarization is unclear. Similar Th2 type cytokines may also polarize myeloid cells such as macrophages into functional states that support the development of regulatory T cells (Tregs).⁵⁶ Tregs potently suppress the effector functions of CD8+ T cells and are frequently found in RRP lesions.^57,58

Another mechanism by which HPV-infected epithelial cells could escape T cell detection and elimination is through the expression of immune checkpoints, the most popular of which are components of the programmed death (PD) pathway. When T cells are primed and activated (against for example an HPV antigen) they express the PD-1 inhibitory receptor. Activated T cells also produce interferon, which potently upregulates expression of the ligand for PD-1, PD-L1, on epithelial and stromal cells. This negative feedback loop prevents uncontrolled T cell activation and results in suppression of T cell activity. Therapeutic PD-1 or PD-L1 monoclonal antibodies can disrupt this immune checkpoint axis and activate existing T cells. Immune checkpoint blockade does not appear to induce new T cells—it activates what already exists that is being suppressed by PD-1/L1 signaling. Indeed, T cells express PD-1 and PD-L1 expression is present on both papilloma cells and other infiltrating immune cells in the majority of archived RRP samples.^57–59

Epithelial Cell-Specific Defects in Antigen Presentation

Intracellular processing and presentation of T cell antigens on MHC molecules is a complex process. Key to this process are the transporter associated with antigen-processing (TAP) proteins. TAP-1 expression is down-regulated in papilloma cells from patients with RRP, and the level of TAP-1 expression appears to inversely correlate with RRP disease recurrence.⁶⁰ Similarly, expression of HLA class I itself is reduced in papilloma cells in patients with RRP,^45,60 due at least in part to the activity of E7.⁶¹ Without sufficient expression of functional TAP or HLA molecules, antigen cannot be processed or presented, and HPV-infected cells will evade T cell detection.

IMMUNOTHERAPEUTIC STRATEGIES FOR PATIENTS WITH RRP

Through understanding how anti-HPV immunity is dysregulated in patients with RRP, new immunotherapy treatments are being studied. The need for enhanced immunity to clear the chronic HPV infection was recognized early, and initial attempts at immunotherapy with systemic interferon were met with some clinical success but unacceptable toxicity profiles.⁶² Subsequently, new immunotherapeutic approaches pushed into clinical development for cancer (especially HPV-associated cancer) afford new opportunities for the treatment of RRP.

Immune Checkpoint Blockade

Given frequent expression of PD-L1 on papilloma cells and PD-1-expressing T cells within RRP lesions, the safety and clinical activity of PD-L1 blockade with avelumab in adult patients with aggressive RRP has been studied.⁶³ The majority of patients demonstrated significant reduction in papilloma disease burden without surgical debulking, and measurable amplification of HPV-specific T cell responses was detected in several patients. At least one additional ongoing clinical trial of immune checkpoint blockade in RRP (with the PD-1 mAb pembrolizumab) has preliminarily reported similar results.⁶⁴ However, the majority of patients were not cured with immune checkpoint blockade alone. Given that immune checkpoint blockade only activates existing T cells, this and other data suggested that inadequate T cell priming may be an issue in RRP. In other words, immunotherapy designed to activate new HPV-specific T cell responses may be needed to eliminate HPV-infected cells and cure RRP.

Preventative and Therapeutic HPV Vaccines

Understanding this framework of how virally infected cells are detected and eliminated by the T cell arm of the adaptive immune system sets up one of the current clinical and scientific debates in this field of study. Preventative vaccines such as Gardasil induce a B lymphocyte (humoral) response that results in the production of antibodies.^65–67 HPV-specific antibodies can bind HPV virions before they infect epithelial cells, leading to their elimination by other arms of the immune system. Antibodies cannot, however, result in the detection and elimination of cells already infected with HPV. If the goal of the immunotherapy is to eliminate the cells already infected with HPV to eliminate the chronic infection, only T cells can accomplish this task. There is no evidence that preventative vaccines can educate or activate T cells. This biologic understanding of how different arms of the adaptive immune system prevent or control existing infection is in conflict with clinical data suggesting that adjuvant administration of the Gardasil preventative vaccine prevents RRP disease recurrence or significantly lengthens the time between surgeries. Several single cohort, retrospective studies suggest a clinical benefit with increased intersurgery interval after adjuvant Gardasil administration.^65,68,69 Of two multicohort studies that reported intersurgery interval in patients with and without adjuvant Gardasil, one reported clinical benefit and the other did not.^70,71 Gardasil treatment does not increase the intersurgery interval in patients with low-risk HPV-associated anogenital condyloma.⁷² The argument can be made that Gardasil is already commercially available; treatment is low risk; and that multiple HPV serotypes are covered to which a patient may not have yet been exposed. Therefore, patients with HPV-driven disorders should be treated regardless. This is clinically rational and may contribute to the most recent meta-analyses on the topic, which supports the use of Gardasil as an adjuvant treatment in RRP despite the lack of convincing, prospective, controlled clinical data⁷³ and contrary conclusions being made in meta-analyses of data in other disease types driven by HPV.⁷⁴

The design and clinical study of therapeutic vaccines for the treatment of established HPV infections is biologically rational, based on current understanding of immune responses against virally infected cells. Such therapeutic vaccines come in many forms but share the goal of delivering HPV DNA to the innate immune cells (usually dendritic cells or macrophages) for presentation of antigens to T cells and subsequent T cell priming and activation. De novo primed and activated HPV-specific T cells would then traffic to HPV-driven lesions and clear HPV infected cells. The approach of using a therapeutic vaccine to induce strong T cell responses and clinical benefit has already proven successful in the treatment of patients with premalignant cervical and vulvar intraepithelial neoplasia with a therapeutic vaccine encoding HPV 16 genes.^75,76

Reports of clinical benefit with therapeutic vaccines for patients with RRP exist. In two trials, one using the measles, mumps and rubella vaccine and the other using a modified vaccinia Ankara vector encoding the bovine papillomavirus E2 gene, intralesional injection of vaccine with or without surgical debulking of disease resulted in significant prolongation of the intersurgery interval or cure of disease in subsets of patients.^77,78 Neither report definitively demonstrated the induction of HPV-specific T cell responses, suggesting that part or all of the clinical benefit induced with these treatments was due to a local, nonspecific, antiviral inflammatory responses involving type I interferon (though such alterations were not studied). More recently, report of clinical benefit in three patients treated with a DNA vaccine encoding HPV 6 E6 and E7 included demonstration of induction of HPV 6-specific T cell responses.⁷⁹ These data have paved the path forward for multiple planned phase I and II clinical trials designed to study the safety and efficacy of novel therapeutic vaccines encoding HPV 6 and/or 11 genes.

A summary of adjuvant treatment for RRP currently in use clinically or in clinical development, including treatment potential strengths and weaknesses, is provided in Table I.

TABLE I.

Summary of Adjuvant Treatment for Patients With Recurrent Respiratory Papillomatosis Currently in Clinical Practice or Development.

Treatment	Pathway	Currently Used in Practice?	Currently in Prospective Clinical Trials?	Strengths	Weaknesses
Cidofovir	Viral DNA replication	Yes	No	Safe for repeat administration	Unclear clinical benefit, no induction of HPV immunity
Celecoxib	COX-2 inhibitor	No	No	Targets mitogenic signaling	No clinical benefit, no induction of HPV immunity
I-3-C	Viral DNA replication	Yes	No	Likely safe for long-term use	Unclear clinical benefit, no induction of HPV immunity
Bevacizumab	Angiogenesis and vessel stability	Yes	No	Can induce rapid regression of lesions	Nephrotoxicity when used systemically, no induction of HPV immunity
Gardasil	Induces humoral immunity	Yes	No	May increase intersurgery interval	Does not induce T cell immunity
Immune checkpoint blockade	Induces cellular immunity	Yes	Yes	May unleash existing anti-HPV immunity	Risk of immune-related adverse events
Therapeutic vaccines	Induces cellular immunity	No	Yes	May induce new anti-HPV immunity	Unknown

HPV = human papillomavirus; I-3-C = indole-3-carbinol; COX-2 = cyclooxygenase-2; T cell = T-lymphocyte.

CONCLUSION

Although considered a “benign” disorder, RRP causes profound voice and airway disturbance. Maintenance surgery to debulk papillomas has been the standard of care for decades but itself can lead to profound postoperative voice and airway dysfunction and does not address the fundamental problem: chronic infection of upper aerodigestive tract epithelial cells with low-risk HPV. Based on current understanding of antiviral immunity, clearance of chronic low-risk HPV infection by the T cell arm of the adaptive immune system is the single rational therapeutic goal. Future adjuvant therapies should focus on the induction of anti-HPV T cell immune responses. Translation of immunotherapies currently FDA-approved or under clinical investigation for cancer, such as immune checkpoint blockade and engineered therapeutic vaccines, provides a hopeful path forward toward the development of such adjuvant therapies.