HPV L1-Capsid Protein Detection and Progression of Anal Squamous Neoplasia

Abstract

The progression of cervical intraepithelial lesions to invasive cancer is associated with corresponding reductions in human papillomavirus (HPV) L1-capsid antigen (L1) expression. We sought to determine whether a similar loss of L1 occurs during anal carcinogenesis using immunohistochemistry on paraffin-embedded sections as well as INNO-LiPA HPV Genotyping (Innogenetics, Gent, Belgium) technology to determine HPV infection status. We analyzed 31 squamous cell carcinomas (SCCs), 26 SCCs in situ (SCC-IS), and 11 normal anal mucosae from 36 patients. High-risk HPV subtypes were detected in all patients. L1 nuclear staining was identified in 38% of SCC-IS; however, there was no detection in normal anal mucosae, SCC, or recurrent SCC. Of those SCC-IS associated with a concomitant invasive SCC, only 15% demonstrated nuclear L1 expression as compared to 62% of isolated SCC-IS (P = .02). Nuclear expression of L1 is lost in the progression of anal SCC-IS to SCC and may serve as a possible prognostic marker of enhanced malignant potential.

The human papillomavirus (HPV) consists of an icosahedral capsid, composed of L1 and L2 proteins, which enclose a circular double-stranded DNA genome of approximately 8 kb.¹ The virus is capable of infecting mucosal epithelial cells of the anogenital tract and has been implicated as the causative agent for the overwhelming majority of malignancies affecting this anatomic region, including squamous cell carcinomas (SCCs) of the uterine cervix, vulva, anus, and penis.² Specifically HPV infection has been reported to be associated with 75% to 90% of anal SCCs.² Similar to cervical SCC and cervical intraepithelial neoplasia (CIN), HPV-dependent anal carcinogenesis is characterized by a step-wise progression through identifiable precursor lesions termed anal intraepithelial neoplasia, (AIN).³

Given that approximately 20 million Americans are seropositive for HPV and anogential tract cancers will eventually develop in only a small fraction, it is clear that infection alone is insufficient for carcinogenesis.^4,5 Similarly, the development of AIN or CIN does not necessarily result in SCC; some lesions may never progress, while others may actually regress.³ Infection with certain high-risk HPV subtypes however, HPV-16 and/or HPV-18 for example, increases the risk of developing cancer. As a result, the US Food and Drug Administration recently approved 2 vaccinations (human papillomavirus quadrivalent [types 6, 11, 16, and 18] vaccine, recombinant [Gardasil, Merck, Whitehouse Station, NJ] and human papillomavirus bivalent (types 16 and 18) vaccine, recombinant [Cervarix, GlaxoSmithKline, Brentford, England) for the prevention of HPV-induced cervical squamous neoplasia, both of which target HPV-16 and HPV-18 L1 capsid proteins.

The HPV capsomere primarily consists of L1 proteins, which confer viral entry ability via interaction with heparan sulfate proteoglycan and subsequent internalization by clathrin-dependent endocytosis.⁶ In addition to serving as a key therapeutic target, expression of the L1 capsid protein has also been shown to be prognostic. Immunohistochemical evaluation of HPV L1 positivity on Papanicolaou (Pap) smears demonstrated disease progression in 9% of L1+ patients and 26% of L1– patients with CIN 1 and CIN 2 lesions.⁷ In addition, regression was observed in 69% of Pap smears with mild or moderate dysplasia and L1 positivity, while L1– mild or moderate dysplastic Pap smears progressed in 76% of patients.⁸ In patients with cervical lesions, HPV L1 positivity is more highly associated with low-grade squamous intraepithelial lesions and low-risk HPV subtypes than with high-grade squamous intraepithelial lesion^9,10 and high-risk HPV subtypes.¹⁰

Although HPV has been extensively studied in female genital tract cancers, relatively little is known about its role in the pathogenesis of anal SCC. In addition to characterizing the HPV infection status of our patient population, we sought to characterize the pattern of HPV L1 capsid protein expression across the spectrum of anal squamous cell neoplasia.

Materials and Methods

Following institutional review board approval, the records of all patients treated at the H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL, with the diagnosis of anal SCC were reviewed. Cases with a pathologic diagnosis of SCC of the anus and sufficient paraffin-embedded tissue for analysis were identified. Pertinent clinical data were obtained retrospectively by using our institutional electronic medical record system. Tissue samples were rereviewed by one of us (D.C.) to ensure accuracy in diagnosis. Of note, the term SCC in situ (SCC-IS), which will be used for the purposes of this study, is equivalent to high-grade squamous intraepithelial lesion and high-grade AIN.

Immunohistochemical Analysis

Unstained slides were recut from original formalin-fixed, paraffin-embedded (FFPE) tissue blocks. The slides were dewaxed by heating at 55°C for 30 minutes, followed by three 5-minute washes with xylene. Tissues were rehydrated by a series of 5-minute washes using ethanol (100%, 95%, and 80%) and distilled water. Antigen retrieval was performed by heating the samples at 95°C for 30 minutes in 10 mmol/L sodium citrate (pH 6.0). Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 20 minutes. After blocking with universal blocking serum (Ventana Medical Systems, Tucson, AZ) for 30 minutes, samples were incubated with a monoclonal anti-HPV L1 antibody (Cell Marque, Rocklin, CA) at a 1:5 dilution overnight at 4°C. The sections were then incubated with biotin-labeled secondary antibody and streptavidin-horseradish peroxidase for 30 minutes each (Ventana Medical Systems). The samples were developed with 3,3’-diaminobenzidine tetrahydrochloride substrate (Ventana Medical Systems) and counterstained with hematoxylin (product No. 760-2021, Ventana Medical Systems). The slides were dehydrated and coverslipped. Negative control samples were included by omitting HPV L1 antibody during the primary antibody incubation.

HPV Genotyping

Genomic DNA was extracted from FFPE tissue by using the QIAamp DNA FFPE Tissue Kit (catalog No.56404, Qiagen, Germantown, MD) Broad-spectrum HPV DNA amplification was performed using biotinylated primers (SPF₁₀) that define a short fragment (65 base pairs) from the HPV L1 region. Polymerase chain reactions were run on an MJ PTC-200 DNA engine thermocycler (MJ Research, Waltham, MA). HPV type-specific detection was conducted by using the INNO-LiPA HPV Genotyping Extra (Innogenetics, Gent, Belgium) based on the principle of reverse hybridization. We amplified 24 HPV subtypes Table 1. The amplicons were then added to each provided probe-specific nitrocellulose test strip. Each strip contained a conjugate control, DNA control, HPV control 1, and HPV control 2. Genotyping was carried out on the AutoBlot 3000H 20 Strip (MedTec, Chapel Hill, NC). Strips were placed on an adhesive LiPA-Scan Reading template and analyzed using the LiRAS for LiPA HPVE, version 2.01 software (Innogenetics).

Table 2.

HPV Subtypes Amplified

High Risk

HPV-16

HPV-18

HPV-26

HPV-31

HPV-33

HPV-35

HPV-39

HPV-45

HPV-52

HPV-53

HPV-56

HPV-58

HPV-59

HPV-66

HPV-68

HPV-73

HPV-82

Low Risk

HPV-6

HPV-11

HPV-40

HPV-43

HPV-44

HPV-54

HPV-70

HPV, human papillomavirus.

Statistical Analysis

Statistical comparisons were performed by using the Fisher exact test. Significance was accepted with 95% confidence.

Results

From 2000 to 2009, we identified 63 cases treated at the H. Lee Moffitt Cancer Center with the diagnosis of SCC or SCC-IS of the anus and with available paraffin-embedded tissues obtained during surgical resection or endoscopic biopsy. Of the cases, 36 had adequate tissue preparations for immunohistochemical analysis and, thus, constituted the study population. The mean age for the 36 patients (17 men and 19 women) was 48 years (range, 26–81 years). Of the patients, 17 had previously received chemoradiation, and 8 patients were immunocompromised secondary to HIV (5), prior organ transplantation (2), and congenital immune deficiency (1).

From these 36 cases, 68 paraffin-embedded tissue slides were evaluated: 11 slides of histologically normal mucosa, 26 of SCC-IS, and 31 of SCC, including 7 local recurrences and 6 metastases. Normal anal mucosa was obtained from 6 patients with SCC-IS, 1 patient with SCC, and 4 with SCC-IS and SCC.

Nuclear Staining for L1 Capsid Protein

HPV L1 nuclear positivity was identified in 38% (10/26) of SCC-IS cases Image 1A. However, there was no L1 nuclear positivity detected in normal mucosa or SCC Image 1B and Image 1C, including distant metastases. Of the SCC-IS cases associated with a concomitant invasive SCC, only 15% (2/13) demonstrated nuclear L1 expression Image 2 compared with 62% (8/13) of isolated SCC-IS (P = .002) Image 3. L1 nuclear staining was located in the superficial squamous epithelium Image 1D. There was no correlation between previous chemoradiation treatment and L1 nuclear positivity (P = .61).

Image 1.

Human papillomavirus L1 capsid immunohistochemical analysis. A, Normal anal mucosa (×40). B, Squamous cell carcinoma (SCC) in situ (×40). C, SCC (×40). D, Staining in the superficial layer of the epithelium (×10).

Image 2.

Isolated squamous cell carcinoma in situ demonstrating nuclear L1 positivity at ×10 (A) and ×40 (B).

Image 3.

Squamous cell carcinoma (SCC) in situ associated with a component of SCC (not shown) without demonstrable human papillomavirus L1 positivity (×20).

Cytoplasmic Staining for L1 Capsid Protein

HPV L1 cytoplasmic positivity was identified in 9% (1/11) of normal mucosa, 4% (1/26) of SCC-IS, and 23% (7/31) of SCC cases Image 4. L1 cytoplasmic staining was located in the superficial squamous epithelium, although at slightly greater depth than observed with nuclear staining (Image 1D). There was no correlation between previous chemoradiation treatment and L1 cytoplasmic positivity (P, not significant).

Image 4.

Squamous cell carcinoma demonstrating cytoplasmic human papillomavirus L1 positivity (×40).

HPV Genotyping

HPV genotyping was performed on tissues from 24 patients, including normal mucosa (5), SCC-IS (12), and SCC (16) Table 2. HPV was detected in each tissue examined, including 4 histologically normal mucosal sections. All patients were infected with at least 1 high-risk HPV subtype. HPV subtype 16 was identified in 73% (24/33) of the tissue sections examined and was the only viral subtype detected in 52% (17/33) of the tissue sections. HPV-18 was the dominant oncogenic subtype detected in 2 (8%) of 24 cases, while there was only 1 case in which neither HPV-16 nor HPV-18 was detected. Multiple HPV subtypes were identified in 12 (50%) of the 24 cases studied. No correlation between polyviral subtype infection and histologic diagnosis (normal mucosa, SCC-IS, or SCC) or immunocompetence was identified.

Table 2.

HPV Genotyping Results for Tissue Samples From 24 Patients^*

HPV Identified	Patients (n = 24)	Multiple Infections (n = 12)	Single Infection (n = 21)	Infection and SCC (n = 16)	Infection and SCC-IS (n =12)	Infection and NM (n = 5)
High risk
HPV-16	20 (83)	6 (50)	19 (90)	15 (94)	6 (50)	4 (80)
HPV-18	3 (13)	3 (25)	0 (0)	0 (0)	2 (17)	1 (20)
HPV-31	7 (29)	7 (58)	0 (0)	4 (25)	4 (33)	0 (0)
HPV-33	7 (29)	7 (58)	0 (0)	4 (25)	4 (33)	0 (0)
HPV-39	3 (13)	3 (25)	0 (0)	0 (0)	2 (17)	1 (20)
HPV-52	7 (29)	7 (58)	0 (0)	4 (25)	1 (8)	0 (0)
HPV-53	7 (29)	7 (58)	0 (0)	4 (25)	4 (33)	0 (0)
HPV-58	7 (29)	7 (58)	0 (0)	4 (25)	4 (33)	0 (0)
HPV-68	2 (8)	2 (17)	0 (0)	0 (0)	1 (8)	1 (20)
HPV-73	2 (8)	2 (17)	0 (0)	0 (0)	1 (8)	1 (20)
Low risk
HPV-6	2 (8)	2 (17)	2 (10)	0 (0)	3 (25)	0 (0)
HPV-11	4 (17)	4 (33)	0 (0)	2 (13)	1 (8)	1 (20)
HPV-40	7 (29)	7 (58)	0 (0)	4 (25)	4 (33)	0 (0)
HPV-54	5 (21)	5 (42)	0 (0)	3 (19)	2 (17)	0 (0)

HPV, human papillomavirus; IS, in situ; NM, normal mucosa; SCC, squamous cell carcinoma.

^*Data are given as number (percentage).

Discussion

We identified that the immunohistochemical evaluation of the HPV L1 capsid protein may harbor prognostic potential for patients with anal SCC-IS. Specifically, the lack of L1 expression may identify patients at higher risk for progression to SCC, while expression of the L1 protein may identify lower risk lesions. Although invasive disease will develop in only about 15% of patients with SCC-IS, this population is unidentifiable a priori, which mandates treatment, including excision and/or local ablative therapies, for all patients.¹¹ In other words, 85% of patients with SCC-IS may be exposed to procedural and/or drug-related morbidity without therapeutic benefit. The evaluation of HPV L1 expression may, therefore, serve as a stratification tool for patients with anal SCC-IS, facilitating identification of patients in need of more aggressive treatment while avoiding interventions in patients who may require only close surveillance.

Of the 24 cases evaluated for the presence of HPV DNA, all were found to be infected with 1 or more HPV subtypes and with at least 1 high-risk subtype. The predominant oncogenic subtype was HPV-16, with a few patients harboring HPV-18. Other high-risk HPV subtypes involved in polyinfections included HPV-31, HPV-33, HPV-52, HPV-53, and HPV-58. These findings are consistent with previously reported studies demonstrating the relationship between high-risk HPV infection and anal SCC.¹² Of the 14 patients with insufficient DNA for HPV genotyping, 5 had positive L1 capsid staining, demonstrating the presence of HPV in at least 80% of the cases evaluated in this study, which is in agreement with the reported associations with HPV and anal cancer.² It is important to note that this results suggests that the lack of L1 capsid positivity is likely not due to the absence of HPV infection. Nonetheless, it remains possible that a small proportion of cases may be HPV−.

The prognostic potential that we have demonstrated for HPV L1 immunohistochemical analysis is in concordance with current knowledge regarding the HPV viral lifecycle. Early infection has been characterized by a productive phase, during which active viral replication and shedding occur from the superficial epithelium.¹³ During this time, L1 encapsulates viral genomes in the nucleus, forming progeny virions capable of reinitiating infection. The productive phase may be followed by a transformation phase, during which the HPV genome is integrated into host DNA.¹⁴ Integration of the viral genome is generally a precursor event for progression to carcinoma, although not an absolute necessity. We found nuclear L1 positivity solely in SCC-IS, an early or precursor lesion to invasive SCC, and observed staining only within the superficial epithelium. Furthermore, we did not identify a single case of nuclear positivity in invasive SCC, suggesting that L1 expression is lost with viral integration. Of note, loss of L1 capsid protein expression has previously been demonstrated to occur in cervical carcinoma in conjunction with viral integration.¹³

Epidemiologic data have demonstrated significantly higher seropositivity rates for HPV infection among patients with anal SCC-IS vs patients with invasive anal SCC.² Possible explanations include loss of expression of the major immunogenic elements of HPV (L1 capsid protein) with viral integration, as has been shown in the development of cervical cancer,¹⁴ higher seropositivity rates or secondary to immune suppression leading to decreased or undetectable antibody titers and thereby progression to carcinoma. In support of the latter hypothesis is the prevalence of anal SCC approaching epidemic proportions in the HIV population.¹⁵ Our data, however, suggest that HPV infection may act similarly in cervical and anal mucosa, with the loss of L1 capsid protein expression during the progression to invasive cancer.

In contrast with nuclear L1 capsid staining, we found no correlation with disease progression and L1 cytoplasmic positivity. We observed what appears to be random L1 cytoplasmic positivity among normal mucosa, SCC-IS, and SCC. Occasional cytoplasmic L1 staining has previously been observed in cervical carcinoma without any known significance.¹⁰ These results suggest that cytoplasmic staining could possibly be artifactual. Our immunohistochemical analysis used a monoclonal antibody generated against a nonconformational internal linear epitope of the L1 capsid protein, which may have some cross-reactivity with host endogenous peptide(s). As an alternative, positive cytoplasmic L1 staining may be related to recognition of degradation peptides that persist in the cytosol for reasons that remain unclear. Curiously, we observed cytoplasmic staining at slightly greater depths in the superficial epithelium than was observed with nuclear staining.

HPV infection and subsequent integration into the host genome are among the necessary steps in the pathogenesis of anal SCC. Similar to cervical SCC, anal carcinoma progresses through identifiable precursor lesions, SCC-IS, offering patients a window for preemptive treatment. However, the clinical dilemma remains the accurate identification of patients who may benefit most from therapeutic intervention. The results of this study suggest that immunohistochemical evaluation of HPV L1 expression may be used to stratify patients, particularly patients with SCC-IS, for treatment selection. Larger scale prospective studies addressing this issue are clearly warranted.

Upon completion of this activity you will be able to:

• describe the pathogenic importance of the human papillomavirus (HPV) L1 capsid protein and identify the HPV subtypes most commonly associated with anal squamous cell carcinoma.

• define the percentage of patients with anal squamous cell carcinoma in situ who will progress to develop invasive cancer.

• recognize the absence of accurate prognostic markers for the development of invasive disease and describe the potential prognostic utility of HPV L1 capsid immunohistochemistry for anal squamous cell carcinoma.