Infectivity of Murine Papillomavirus in the surgical byproducts of treated tail warts

Abstract

Objective:

Human papillomavirus (HPV) is a highly stable DNA virus that causes disease in human organ systems, including the larynx and oropharynx. The treatment of HPV-associated diseases with scalpels, lasers, and other surgical instruments has the potential to release infectious particles, placing healthcare workers at risk. The objectives of this study are to create a reproducible in vivo animal model of papillomavirus infectivity and to compare the infectivity of byproducts of surgically-treated mouse papillomavirus (MusPV) warts.

Study Design:

Animal study

Methods:

Nude laboratory mice (Mus musculus) with established MusPV tail warts were treated with scalpel excision, KTP laser ablation, and coblator treatment. Uninfected nude mice were challenged with surgical byproducts, including ablated and heated tissue, and surgical smoke products. The incidence and time course of the appearance of warts was recorded.

Results:

There was rapid transmission of virus in mice challenged with scalpel-treated warts, with 50% penetrance of infection at Day 13, and 100% at Day 32. For KTP treated warts, there was the slower development of infection (50% by Day 35) but 100% penetrance by Day 52. Coblator-treated tissue reached 50% penetrance at day 59 and a maximum of 73% penetrance. Smoke plume captured during treatment with the KTP laser and coblator was highly infectious, as was the material captured in a laser filter.

Conclusion:

MusPV remains infectious in all modes of surgically-treated tissue and the smoke plume is capable of transmitting infection. Healthcare workers should use appropriate precautions to lower their risk of infection when treating papillomavirus-associated diseases.

Introduction

Human papillomavirus (HPV) causes a wide range of human diseases, including laryngeal papilloma and genital warts¹, oropharyngeal, cervical, and anal carcinoma², and cutaneous warts³. The worldwide burden of HPV-associated diseases is high, and the mainstay of treatment for these diseases is surgical management, as there exists no known effective medical treatment for HPV apart from preventative vaccines. In the head and neck specifically, the low-risk HPV types 6 and 11 are the cause of Recurrent Respiratory Papillomatosis (RRP)⁴, and high-risk HPV is the cause of the rising incidence of oropharyngeal carcinoma⁵. Current treatments for removal of such papillomatous or neoplastic tissue include robotic excision with electrocautery⁶, sharp excision with bladed instruments⁷, coblation⁸, or laser excision or ablation⁹. In all methods, it could be hypothesized that actively infectious viral particles are contained in the treated or heated tissue, released into the immediate surgical environment, or in the surgical plume generated by these devices.

The potential danger posed by aerosolized HPV particles has been recognized for many years¹⁰. Especially for carbon dioxide (CO2) laser excision, experimental and anecdotal evidence indicates that there is a real risk of infection. Bovine papillomavirus can be found intact in the smoke plume generated by a carbon dioxide laser and is capable of infection¹¹. At least one case study draws a correlation between a surgeon’s exposure to HPV-carrying smoke plume and developing an infection¹². The intact tissue itself that has been treated by surgical instruments could also pose a potential risk for operating room staff, physicians, and other patients if not adequately disposed of or cleaned.

For laryngeal papilloma CO2 is still used for surgical management, but formerly held a much more dominant market position.¹³ Newer technologies are now widely available and increasing in popularity, including laryngeal microdebrider¹⁴, coblation⁸, and potassium titanylphosphate (KTP) laser¹⁵. The KTP laser in particular offers specific surgical advantages in the treatment of papilloma⁹, being capable of either inducing ischemia, denaturization or complete ablation of target tissue, it also produces minimal off-target damage to surrounding epithelial tissue¹⁶. Previous literature has focused on the CO₂ laser and the risk of infectious virus in the surgical smoke but these newer modalities have not been studied.

The exact nature of how infectious surgical byproducts are in vivo has been difficult to answer due to the difficulty of transfecting papillomaviruses in cell culture and the challenges inherent in working with animal models such as cows. Since papillomaviruses require stratified squamous epithelium for their replication and persistence¹, an animal model could be very useful in answering these questions. Recently discovered, the murine papillomavirus (MusPV)¹⁷, and its host, the common laboratory animal Mus musculus, is potentially a model organism for in vivo papillomavirus studies. MusPV shares many homologues of human genes and physiology with HPV, including obligatory replication in squamous epithelium, but is much easier with which to work than rabbit, canine, or bovine¹⁸ models of papilloma. Therefore, in this study we employed the MusPV model to answer questions related to the risk of infectious papillomavirus particles in surgically treated warts and smoke plumes.

Material and Methods

Mice

Female athymic nude mice (6–8 weeks old) were purchased from Charles River (Frederick, MD) and maintained at the Johns Hopkins University animal facility. All procedures were performed in accordance with the Johns Hopkins Animal Care and Use Committee guidelines for proper care and use of laboratory animals. All procedures were performed under an approved protocol.

MusPV Challenge

A laboratory stock of mouse papilloma (MusPV) was used for infecting the tails of 10 athymic nude mice to serve as the surgical subjects of future experimentation. MusPV tail challenge was carried out as follows: MusPV stock was mixed with 1x PBS to a final 50:50 mixture. Tails of nude mice were treated with a Dremel sander to remove the first two layers of epidermis. 20 μL of the MusPV:PBS mix was pipetted onto the wounds of each mouse and an endocervical brush was run back and forth across the wounds with approximately 10 brush strokes. 20 μL of additional MusPV:PBS mix was pipetted across the wound for a total of 40 μL MusPV:PBS per tail challenge.

Surgical Treatment and Tissue Collection

Warts in the source mice were given four months to grow to a florid state, and prior to surgical manipulation mice were anesthetized with up to 120 μL of ketamine mixture. Two mice served as controls, where their warts were simply removed en bloc with a scalpel. To mimic laryngeal microbrider, two mice underwent piecemeal sharp excision and rapid mincing of the wart tissue in a saline rinse. Three mice underwent KTP laser (Aura XP, Boston Scientific) surgery, with individual warts treated with two different settings - blanching (30 W/15 ms pulses/2 pulses per second) and ablative (40 W/15 ms pulses/2 pulses per second)¹⁹. Three different mice underwent coblation (Procise Laryngeal Wand, Smith & Nephew), with individual warts treated at ‘ablate’ setting of 5 and ‘coag’ setting of 3. With both the KTP laser and coblation groups, individual warts were either blanched or nearly ablated to mimic different treatment techniques, and the desiccated tissue that sloughed off or had obviously undergone profound treatment effects was saved separately for each surgical method. Tissue was stored at −20°C until ready for processing.

Tissue processing for all tissue samples regardless of treatment type was done as follows: 250–500 μL of 0.8 M NaCl in DPBS was added to the tissue each time before placing the cryotube into the bead beater with ≤ 106 μm beads (Sigma) for a total of 3 cycles. The first bead beating session lasted for 3 minutes, the second session for 2 minutes, and the third session for 1 minute. The samples were spun in a microfuge and decanted by pipette. The decanted fluid consisting of 0.8 M NaCl in DPBS and viral particles were pooled into a 2 mL Eppendorf tube for each sample. The resultant stock was used for subsequent challenge in the naïve nude mice which comprised the experimental groups (Table 1). Multiple samples from uniquely treated warts were created for each treatment technique due to expected variability in treatment effect. Challenge of naïve mice tails was carried out as described above.

Table 1.

Treatment groups for tail challenge experiments.

Group	Treatment Method	Number of Mice Challenged	Calculated Viral Load at Challenge (viral particle units)
1	Cold instrument	3	1.73 × 1012
2	Cold instrument	3	3.86 × 1011
3	Cold instrument	3	8.43 × 1010
4	KTP laser blanched	3	2.54 × 1011
5	KTP laser blanched	3	4.18 × 1011
6	KTP laser ablated	3	1.26 × 1012
7	Coblation blanched	2	5.79 × 109
8	Coblation blanched	2	2.26 × 1011
9	Coblation blanched	2	6.75 × 1011
10	Coblation blanched	2	6.53 × 1010
11	Coblation ablated	3	1.01 × 1010
12	KTP laser plume	3	8.16 × 105
13	Coblation plume	2	1.47 × 105
14	Filter wash	2	2.27 × 105
15	MusPV stock control	3	4.00 × 1012

Surgical Smoke Collection

For the coblation and KTP laser methods, a bubbler apparatus was connected to the wall vacuum system with an inline filter (Surgimedics, San Antonio, TX, USA) between the bubbler apparatus and the wall. The bubbler apparatus consisted of a 250 mL Erlenmyer flask, rubber stopper, 5 mL plastic pipette (with the ends broken off), approximately 100 mL of Roswell Park Memorial Institute (RPMI) media, and tubing²⁰. During coblation, steam vapor was produced; during the KTP treatment, a smoke plume was produced. Both of these “plumes” were vacuumed into the bubbler apparatus through suction tubbing placed adjacent to the working tip of the instrument. The RPMI solution and proximal tubing was changed between each coblation and KTP treatment yielding two distinct “plume” solutions - however, the same distal tubing and filter membrane that connected the device to the wall suction was used for the duration of treatment for both groups. After all experiments were performed for each group, the solutions for KTP laser and coblation ‘plume’ samples were separately stored at −20°C until each was used to challenge naïve nude mice tails. The filter membrane that captured material distal to the bubbler from both the KTP laser and coblator was removed under sterile conditions in a laminar flow hood. The filter was placed in a 50 mL conical and 2.6 mL of Dulbecco’s Modified Eagle Medium (DMEM) was added with 13 μL of 130U penicillin and 13 μL of 50 μg streptomycin. The filter and DMEM was incubated at room temperature for 90 minutes. After incubation, the filter was spun at 2000 rpm for 5 minutes and the resulting supernatent saved. The resulting solution was saved at −20°C until ready for use. Challenge of naïve nude mice tails was carried out as described above.

PCR

DNA was extracted from approximately 200 μL of each sample using the ZYMO DNA Prep Kit (ZYMO Research, USA). Primers for MusPV E6 were custom ordered from IDT (USA). Forward primer 5’ – AGA GTG CAT GGC TGG CAA GA – 3’, reverse primer 3’ – CAT GTG GCG CAC CAA GTG AA – 5’, and probe 5’ - /56-FAM/TGG CAA GCC GCA CGC TTT GGC ATC A/36-/TAMSp/ − 3’ were used to amplify each sample. PCR was run in triplicates of 10 μL volume. Each reaction consisted of 5 μL iQ Supermix (BioRad, USA), 0.5 μL each of 10 μM forward and reverse primers, 0.5 μL 5 μM probe, and 1 μL nuclease free water. All reactions were run on the BioRad CFX96 qPCR machine (BioRad, USA). A standard curve of known MusPV plasmid concentration was used for calculating Starting Quantity (SQ) of MusPV virus in viral particle units (vpu). SQ was then used to calculate the viral load for each sample at inoculation.

Statistical Analysis

The development of warts was recorded on a Kaplan-Meier survival curve, with curves compared using the log-rank test, where alpha < 0.05, set as significance for rejection of the null hypothesis (no difference between the curves).

Results

After tail challenge with processed tissue samples, mice were serially examined for the development of MusPV tail warts. Sample warts are shown in Figure 1. In the cold instrument / bladed group, there was rapid growth of tail warts, which was an expected result due to the lack of manipulation of the wart tissue with heat or laser energy. This control group had 50% penetrance of tail warts at day 13 and all mice developed tail warts by post-challenge day 32 (Figure 2a). In the KTP laser tissue group, there was 50% penetrance of tail warts at day 35 with all mice having warts by day 52. In the coblator tissue group, there was 50% penetrance of tail warts at day 59. In contrast to the other modalities, not all mice in the coblator tissue group developed warts by the termination of the experiment; at 60 days post-challenge the penetrance of warts was 73%. Both coblated tissue and cold instrument tissue mice groups had their first appearance of warts at day 13, while the KTP laser tissue group had warts visible starting at day 34. When evaluated by the log-rank test, none of the curves differed to a statistically significant degree: KTP laser tissue vs. cold instrument, alpha = 0.074, Coblator tissue vs. cold instrument, alpha = 0.18, KTP laser tissue vs. coblator tissue, alpha =0.11.

Figure 1.

Sample tail warts in an athymic nude mouse.

Figure 2.

A) Kaplan-Meier survival curves for nude mice challenged with surgical tissue byproducts from cold instrument / bladed (n=9), KTP laser (n=9), and coblator (n=11) treated MusPV warts. B) Kaplan-Meier survival curves for nude mice challenged with material collected from the KTP laser plume (n=3), coblator plume (n=2), and filter wash (n=2) from the smoke evacuator.

When media that captured the surgical smoke plume from the KTP laser or the coblator was used to inoculate mice, there was also the rapid development of tail warts (Figure 2b). All KTP laser plume mice had visible warts by day 18, and coblation plume group had 50% penetrance of warts by 60 days post-challenge. When the filter wash that contained the smoke plume from both the KTP laser and coblator was used to inoculate mice, all mice had visible warts by day 38.

As the prior experiments indicated, there was clearly intact and infectious virus even in both the heavily treated residual tissue and smoke plumes. We were therefore further interested in the relationship between the viral DNA that could be retrieved from the surgical by-products and the temporal development or overall penetrance of wart development. The viral DNA measured in the surgical by-products is found in Table 1, broken down into individual tissues treated and converted into viral inoculum stock. As expected, the viral count was dramatically lower in the smoke plume samples compared to tissue samples. In general, surgical samples showed a 1 – 3 log reduction in viral count compared to untreated control.

As a result of the same volume of viral stock solution being used for the tail challenge, any differences in growth could be due to differences in the viral load between groups. In order to determine if the time course or penetrance of wart development was correlated with viral inoculum at any point along the continuum of wart development, comparisons between the proportion of uninfected mice and viral load were made at various time points (Figures 3a–d). Since there was a difference in processing technique between the treated tissue versus the material captured in the vacuum bubbler, analysis of the viral inoculum and the development of warts was confined to similarly processed tissue. Proportions of mice infected at Days 15, 30, 45, and 60 showed that the correlation between the development of warts and viral dose was generally positive, with the strongest dose / response relationship at Day 45 (R² = 0.65). When few mice have warts (Day 15, R² = 0.0015) or most mice have warts (R² = 0.116) the relationship was not strong since there was little divergence in outcomes at those time points.

Figure 3.

Correlation between viral inoculum and temporal development of warts demonstrates a positive relationship between dose and wart development, particular at Day 30 (B) and Day 45 (C). No relationship seen at Day 15 (A) or Day 60 (D) due to the high proportion of uninfected or infected mice respectively. Line of best fit is logarithmic.

Discussion

This series of experiments demonstrates that treatment of a murine papilloma with modern surgical equipment results in viral particles that are infectious. Using the KTP laser or coblator, even different settings and treatment durations (blanching vs. ablative) still resulted in infectious particles in the desiccated, charred, and otherwise heavily treated tissue. In addition, the surgical smoke plume created by the coblator and the KTP laser both contained infectious particles.

Regardless of the treatment method, all experimental groups eventually contained infected mice. However, it is notable that only the coblator group, in both the surgically treated tissue and in the surgical plume, contained some mice that were not infected by the termination of the experiment. In general, this reduced infectivity correlated with a lower viral inoculum in the coblator-treated tissue, as there was a positive correlation between viral dose and development of warts. The cold instrument-treated tissue, analogous to the microdebrider or tissue shaver, led to the most rapid development of warts, followed by the KTP laser, and then the coblator. It is interesting to note that treatment with the smoke captured from the surgical plume from both the KTP laser plume and the coblator also led to warts despite a ~7 log reduction in viral load compared to control tissue. This result demonstrates that even comparatively small amounts of virus are still infectious. In addition, the laser filter distal to the entire bubbler apparatus contained infectious material, meaning that aerosolized virus was easily able to pass through or over our bubbler capture system to the distal filter. As we did not sample a second filter distal to the first filter, we cannot comment whether the filter was efficacious in completely removing virus from the air flow. This could be a direction of future research and testing of laser safety devices.

Our results agree with many prior experiments with CO2 laser where viral DNA was retrieved from surgical smoke, and in some cases capable of transmitting infection^10,11,18,21. The value of an animal model to prove direct transmission of infection establishes definitively that intact virus present in the surgical product and surgical smoke is active and retains its infectious capability, in a way that simple retrieval and measurement of viral DNA cannot. While we are able to quantify the amount of viral DNA in our surgically-treated tissue and smoke plume, it is an assumption that this correlates directly to the amount of intact viral particles. It is likely that much of the virus measured is denatured or rendered uninfectious by the high heat and energy applied the by surgical instruments. This effect may account for difference seen between the viral counts from tissue compared to smoke, yet roughly equivalent infectivity in the in vivo animal model. Finally, we make the important note that establishment of viral infection in our model involves the direct inoculation of virus onto denuded skin. As mucosal trauma seems to be essential to the transmission of papillomavirus¹, our model requires the debridement of superficial epidermis and variable trauma could be another explanation for the variable development of warts.

Our study has important implications for surgical safety and the importance of safety masks, air filtration systems, and smoke evacuators²². Earlier case studies have described the transmission of HPV with the CO2 laser²³. This study demonstrates this concern is now clearly also applicable to KTP laser, coblator and surgical resection techniques. Of particular importance, plumes are infectious, regardless of whether the viral load in the plume is high or low. Precautions should be adopted by any associated operating room and cleaning crews, especially if the handling instruments or tissue immediately following papilloma treatment.

Papillomaviruses are species-specific, and while mouse papillomavirus is genetically homologous to the human counterpart and similar in its structure and lifecycle, we must acknowledge the species difference that places natural limitations on the predictive power of our model. Our relatively small sample size per group may contribute to some of the model’s inaccuracy – the costs in housing a quantity of mice for a sufficient amount of time to ensure the persistence of MuPV infection places limitations on the duration of the experiment. However, in this experiment we demonstrate that the easy availability of nude mice and the ease of establishing infection in this model is attractive for future experiments that seek to study the effect of lasers on papillomaviruses, evaluate the efficacy of laser safety devices, and the development of novel surgical devices to treat papillomavirus infections.

Conclusion

In an experimental model of murine papillomavirus infections treated with surgical devices, MusPV remains infectious, both the surgically-treated tissue and the smoke plume is capable of transmitting infection. Healthcare workers should use appropriate precautions to lower their risk of infection when treating papillomavirus-associated diseases.