Systemic bevacizumab for recurrent respiratory papillomatosis. A case series

Abstract

Recurrent respiratory papillomatosis (RRP) is a human papilloma virus (HPV) 6/11 related, predominantly histologically benign neoplasm of the upper and lower airway affecting both younger and older patients. Though potentially declining due to HPV vaccination RRP is still challenging in management and a significant cause of affected quality of life. Systemic bevacizumab has shown efficacy for aggressive disease in other case series in juvenile-onset RRP (JORRP) as in adult-onset RRP (AORRP). We present five consecutive patients with AORRP treated with systemic bevacizumab for aggressive laryngeal and tracheal papillomatosis. Adding to the findings of previous reports we show that systemic bevacizumab treatment could have positive influence on aggressive AORRP without pulmonary spread with manageable side effects.

Level of Evidence

IV - case series

Keywords:

• Respiratory papillomatosis
• bevacizumab
• systemic
• case series

Introduction

Recurrent respiratory papillomatosis (RRP) is a human papilloma virus (HPV) 6/11 related, predominantly histologically benign neoplasm of the upper and lower airway affecting both younger (juvenile onset - JORRP) and older patients (adult onset - AORRP). Incidence of 1,8–4/100.000 [1], highest in AORRP; though potentially declining due to HPV vaccination [2], patients often require multiple surgeries to maintain airway patency and voice quality either in local or general anesthesia. Even so, a percentage of patients show an even more aggressive course of the disease defined as more than four surgeries per year and/or tracheobronchopulmonal spreading and they pose a challenge to surgical management [3].

Though malignant transformation in papilloma from HPV 6/11 is low [4], it seems to be a little higher in this group of patients with aggressive disease and especially in patients with pulmonary spread [5] and treatment to control the disease is needed. The epithelium of the papilloma strongly expresses vascular endothelial growth factor (VEGF-A) and its receptors (VEGF-R) and is supposed to be part of the pathogenesis [6].

Bevacizumab is a VEGF antibody widely used in advanced and metastatic cancer with a theoretical effect on RRP due to its angiogenic inhibition. While there is a risk of severe complications, such as bleeding, thromboembolism and thrombocytopenia, these are rare in otherwise healthy individuals, and most patients experience side effects limited to hypertension and proteinuria in previous published series [5, 7,8]. Local bevacizumab has also been administered with effect and safety as an adjuvant to surgery in small series [9,10].

Long-term side effects of treatment with bevacizumab are not clear. In a study with long-term follow-up, examples of malignant transformation have been noted during treatment, but it is difficult to conclude that it is because of treatment as aggressive course of the disease itself predisposes to this transformation [5]. Still, the use of bevacizumab in treating recurrent respiratory papilloma is off-label and patients should be given extensive information about potential side effects and should be reserved for severe cases.

Case series

In our tertiary referral center with extensive experience in treating juvenile and adult-onset recurrent respiratory papilloma (JORRP/AORRP), we have treated 133 patients with RRP over the last 13 years. Our preferred surgical methods have shifted from ablation with CO₂ laser to microdebrider and now mostly ablation with photoangiolytic lasers (KTP laser and Blue Laser).

In this case series, we present five patients with complicated RRP treated with systemic bevacizumab. All patients referred for this treatment were adults with AORRP, previous multiple surgeries at least more than 4 per year and/or aggressive illness and distal spreading. An experienced team of three laryngologists were responsible for referring patients and observing and documenting treatment results. All patients were discussed individually with the tumor board and all patients referred were eventually treated. A team of two designated oncologists verified indication, administered treatment, and handled side effects.

Intervention was a standardized bevacizumab 10 mg/kg, reduced to 5 mg/kg with serious side effects. Initiation therapy was a 3-week interval for 4 to 6 series. Increase of intervals was subsequently done between 6 and 10 weeks: 10 weeks (P1), 8 weeks (P4), 6 weeks (P5).

Monitoring the treatment effect was done by the laryngologist by laryngeal and tracheobronchial endoscopy. Computed Tomography (CT) scan was not used as no patients showed signs of pulmonary spread. Surgery was performed before treatment introduction to minimize the amount of visual papillomatosis and upon recurrence after discontinuation of treatment. The endpoints were the resulting increase in interval between surgeries or visual and symptomatic regression as evaluated by the otolaryngologists. Due to the heterogeneity of the small cohort, we emphasized the results in partial response (PR) – regression but still multiple visible papilloma and in need of repeated surgery, good response (GR) – hardly visible papilloma and no need of repeated surgery and complete response (CR) – in the one patient that showed no recurrence of visible papilloma even though treatment with bevacizumab was stopped.

HPV analysis in tissue biopsies was performed in the 4 of 5 patients with the VisionArray® HPV Chip 1.0 that detects 41 subtypes of HPV by polymerase chain reaction (PCR), in one patient analysis was done with in situ hybridization (ISH) at time of diagnosis but not repeated.

This publication was approved by the local data board and written consent obtained from all patients.

See synopsis in Table 1 for demographics, Table 2 for treatment overview and Figure 1 for outcomes.

Figure 1. Outcome for patient 1–5. Above horizontal line, each vertical line marks a surgery up to 2 years before initiating treatment with bevacizumab. Below horizontal line, each vertical line mark initiation of a bevacizumab treatment series

Figure 2. images of patient nr 2. First image is of the larynx at time of presentation. Second image the difficult exposure of the larynx during attempted surgery. Third picture of the larynx a year after presentation and 6 months after last bevacizumab treatment.

Table 1. Demografics.

Patient case	RRP onset age	Lokalization	HPV subtype	Follow up months (years)
1	37	Larynx, upper trachea	6	138 (11,5)
2	35	Larynx	negative	74 (6,2)
3	25	Larynx, upper trachea	11	484 (40,3)
4	20	Larynx	11	340 (28,3)
5	52	Larynx	11	39 (3,2)

RRP: recurrent respiratory papillomatosis, HPV: human papilloma virus.

Table 2. Treatment overview.

Dosing of bevacizumab	N series (total treatments)	Best response (Clinical evaluation)	Side effects
10 mg/kg	4 (30 a.c.)	Partial response	Worsening in known hypertension
10 mg/kg	1 (5)	Complete response	None^a
10-5 mg/kg	1 (4)	Partial response	Multiple^b
10 mg/kg	3 (23 a.c.)	Good response	None
10 mg/kg	1 (11 a.c.)	Godd response	None

^aTreatment discontinued probably due to pulmonary tuberculosis, see text, ^bsee text, a.c.: and counting.

Patient no 1

Forty-nine-year-old Caucasian male (2022), papilloma HPV type 11 (analysis with PCR) and mild dysplasia involving glottic and supraglottic area and trachea, debut at age 37. He showed no pulmonal spreading on bronchoscopy. Tracheostomy in early stage, to date 31 surgeries with trans-laryngeal and trans-tracheostomy debridement with microdebrider, CO₂ and KTP laser ablation. He had two times four series and one time five series of bevacizumab 10 mg/kg in 3-week interval over a period of four years (2015–2019) with only few relapses in need for surgery. Overall good effect of treatment with visual regression of papilloma, though still present especially around the tracheostoma (partial response). In 2020, bevacizumab was reintroduced and continues to date, with increasing treatment intervals, now 10-week interval. Successfully decannulated in July 2022. He has received Gardasil-4 vaccination in 2013.

Patient no 2

Forty-one-year-old woman of south-east Asian origin (2022) with rapid debut at age 35 of glottic and supraglottic papilloma, biopsy consistent with papilloma without dysplasia or necrotizing granulomas but HPV negative on PCR analysis. The papillomatosis involved the vocal cords arytenoid region, and epiglottis. Surgically very difficult to access and she was quickly in need of tracheostomy. She had three attempts for surgical debridement with KTP-laser ablation before referral to systemic bevacizumab and received five series 10 mg/kg in 3-week interval (2016). Due to low hemoglobin and declining general condition treatment was terminated, but there was still no recurrence of papilloma (complete response) after five years of follow-up (see Figure 2). She was decannulated 3 months after the last series but due to severe glottic agglutination after treatment, seven attempts have been made to address the webbing; patient is forming keloid also in other locations. The declining general condition was found to be concurrent with pulmonary tuberculosis, diagnosis probably delayed because of suspected side effects to bevacizumab treatment. Revision of the pathology report did find classic papilloma morphology and not necrotizing granulomas. She has not received HPV vaccination.

Patient no 3

Sixty-five-year-old Caucasian male (2022) with debut in youth (age 25) with a total of 63 surgeries of laryngeal and tracheal papilloma with mostly CO₂ laser ablation. Biopsy with HPV type 6 (ISH) and mild dysplasia in the past. Due to disease specific and surgical side effects, he suffers from laryngeal agglutination treated with tracheostomy, Montgomery® tracheal stent and speaking valve. He had four series of bevacizumab 10 mg/kg in 3-week intervals reduced to 5 mg/kg due to side effects as stomatitis, thrombocytopenia, proteinuria, hypertension, and bleeding per rectum. The initial effect declined after reduction in dosage (partial response), treatment was terminated, and he continues to have half-yearly debridement of the remaining papilloma in trachea, mostly around the tracheostoma with changing of the stent as before bevacizumab treatment. Recently vaccinated with Gardasil-9, the potential outcome is not yet evaluated.

Patient no 4

Unprecedented case to us of 104 surgeries for aggressive glottic papilloma in a 48-year-old Caucasian woman (2022). Debut at age 20, HPV 11 (PCR analysis) positive papilloma showing occasional mild dysplasia. 33 surgeries up to our registration (2009), after which surgeries in periods occured almost monthly with microdebrider and later Blue Laser for voice and breathing problems. CT scan and bronchoscopy with biopsy revealed non-symptomatic lymph node sarcoidosis but no distal spreading of papilloma has been shown. After surgery no 101, she received bevacizumab 10 mg/kg for 6 series in 3-week interval (2020) under which no surgery was required but papilloma did not vanish entirely (partial response). Three months after pausing due to local agreement papilloma reoccurred with almost equal aggressivity and bevacizumab was reintroduced after two surgeries with 1-month interval. After six series in 3-week interval, she had surgery for voice improving and continued bevacizumab treatment with increasing intervals up to 6-week interval, the papilloma hardly visible (good response). HPV vaccination with Silgard in 2013 appeared to have no impact, sublesional bevacizumab administered three times had absolutely no effect and recurrences increased.

Patient no 5

Fifty-five-year-old Caucasian man (2022), debut at age 52, with initially mild disease of HPV 11 (PCR) positive glottic papillomatosis, but quickly decreasing intervals between surgeries and very few weeks of tolerable voice between interventions. He was operated on with microdebrider and Blue Laser. He never experienced breathing difficulty. The patient had a total of nine surgeries over 2 years, three of which included three perioperative submucosal administrations of bevacizumab, before starting systemic bevacizumab 10 mg/kg in 3-week interval increasing to 6-week interval 10 months prior to this date, not yet requiring additional surgery and papilloma visually regressing (good response). HPV-vaccination with Gardasil-9 shortly after debut with no apparent benefit.

Discussion

Due to the lack of an efficient lasting therapy for treating RRP, management has until now been primarily surgical. However, a fraction of patients has a more aggressive disease with rapidly relapsing disease and spreading to the entire respiratory tract. In these cases, adjuvant treatment is highly desired. Currently, bevacizumab locally or systemically is the treatment of choice in many centers.

Since the first case report from Germany in 2009 several other centers have until now published their case series of treatment with systemic bevacizumab in both the adult and juvenile population, also systematic reviews agree on the efficacy and safety of the treatment [5,7,8,11–15].

In comparison with our study, other series treat patients with pulmonary spreading and juvenile onset and hence more severe disease. Included patients in our series are equally severe cases, though without pulmonal involvement. Fortunately, there has not been JORRP with need of systemic treatment in our center.

Overall, treatment with systemic bevacizumab seems to be efficient and safe to control aggressive disease, though not curing it, as all referenced cases and series find partial or good response with visible changes to the papilloma structures and appearances with limited side effects [5,7,8,11–14]. In our series, we do have one partial responder, but dosage was reduced due to side effects. The other four patients had no or very mild side effects.

Ten mg/kg seems to be the most frequently used dosage though effect has been shown with lower dosage even in an adult [8].

Beside dosage, consensus about treatment intervals and maintenance therapy differ throughout the series and even in the same series, but most continue treatment with increased intervals [5,7,13]. When we treated our first patient, experience was limited both nationally and internationally and in collaboration with the department of clinical oncology we agreed to treatment series resembling the oncological series of 3-week interval and evaluation after four series with discontinuation of the treatment. Based on the international experience, we now agree on six series as standard initiation therapy and individually find the interval of the maintenance treatment when we achieve good response. This change over time, however, makes comparison to other case series difficult. There is no uniform treatment protocol yet and treatment choice depends on individual disease activity which is seen in our series as well as in others. Others have found improvement in intersurgical intervals up to several years [13], but our previous tradition of discontinuing treatment and waiting for recurrence will account for our limited statistical improvement. We must wait for more long-term follow-up studies to see the optimal treatment intervals, which may show itself to be very individual. Surgical interventions will still be necessary despite bevacizumab therapy. And it is obvious that recurrence after discontinuing treatment is a likely outcome [5].

One patient in our series showed complete response still after discontinuing treatment and we did not find HPV in the biopsies which make her case very different from most other cases. As tuberculosis is known to trigger immune response, it might have had an impact on the disease course.

The standard array in our department could not detect HPV, even though detecting 41 different subtypes of HPV. It remains unclear whether her South-East Asian origin has an influence on HPV type. A VEGF-A or VEGF-R analysis could be performed in this case to maybe recommend bevacizumab treatment also for HPV negative papilloma in the future if other centers find similar results. Due to the severity of disease in patients considered for systemic bevacizumab, it is not surprising that several cases report malignant transformation. The general risk of malignant transformation in RRP is low, but variably reported between 3–7% [4] though seems higher in these series [5,7]. It is more often seen in distal and especially pulmonal spreading. This might explain why we see no malignant transformation in our cohort which primarily have glottic or subglottic disease.

Submucous administration of bevacizumab as adjuvant therapy to surgery has also shown promising results in small cohorts as shown by Best and Zeitels et al. who treated 43 patients with no side effects and a randomized cohort of 20 patients/40 vocal cords with significant decline in need of surgery on the treated vocal cord [9,10]. But its use remains indicated in cases where the bulks are surgically removable and sublesional administration is possible. We have offered patients sublesional treatment according to this experience if relapses were frequent but not concluded as aggressive.

All patients in the cohort received vaccination with the HPV vaccine. All after debut of the HPV-related disease and as a potential adjuvant for disease control according to support of the Rosenberg et al. metaanalysis [16]. As the vaccine cannot be an adjuvant standing alone in their cases, it might have a place in a combination of treatments as it is not known why some patients experience this aggressive course of disease and the possible immune modulation of a vaccine may result beneficial. In our cohort, it seemed to make little difference.

The obvious limitations of this case series are the lack of control group, and the short follow-up. The former is difficult to find due to the low number of cases, consideration for treatment only with unmanageable illness and off-label treatment with potential serious side effects.

With the limited number of publications in this field, we still find it useful to publish yet another case series.

As controlled trials with systemic bevacizumab are probably impossible to conduct, consensus on inclusion and treatment modalities are important and very welcome [17]. A consensus on treatment intervals and dosage will be the next step [18]. Also, in future, we need more reports on long-term treatment with bevacizumab to see if more side effects or complications or even resistance occur.

Conclusion

Adding to the findings of previous reports, we show that systemic bevacizumab treatment could have positive influence on aggressive AORRP without pulmonary spread with manageable side effects. We hope to contribute to consensus in the area and to the recommendations difficult to raise in level of evidence.

Informed consent

We have obtained written consent to this publication from all patients in the cohort.

Disclosure statement

No potential conflict of interest was reported by the author(s).