Prevalence and Correlation of Human Papilloma Virus and its Types with Prognostic Markers in Patients with Invasive Ductal Carcinoma of the Breast in Kuwait *

OBJECTIVES
This study aimed to document the association of human papilloma virus (HPV) and its types in breast carcinoma tissues in Kuwaiti women, and correlate this with known prognostic markers.


METHODS
The clinicopathological data of archived tissue from 144 cases of invasive ductal breast carcinoma were studied (age, histological grade, size of tumour, lymph node metastases, oestrogen/progesterone receptors and human epidermal growth factor receptor 2 status). HPV frequency was documented using immunohistochemistry (IHC) and chromogenic in-situ hybridisation (CISH). HPV types were documented by CISH using HPV probes. CISH and IHC techniques were compared and HPV correlated with prognostic parameters.


RESULTS
The HPV prevalence as determined by CISH and IHC was 51 (35.4%) and 24 (16.7%) cases, respectively. The sensitivity of HPV by IHC was 37.3% and specificity was 94.6%. The sensitivity and specificity of HPV-CISH compared to HPVIHC was statistically significant (P <0.001). HPV-CISH was seen in 51 cases. A combination of HPV 6 and 11, and 16 and 18 was seen in 2 (3.9%) cases, and a combination of HPV 6, 11, 31 and 33 was seen in 7 (13.7%) cases. All three HPV probes: 6 and 11, 16 and 18, as well as 31 and 33 were present in 2 (3.9%) cases. The prevalence of HPVCISH in the Kuwaiti and non-Kuwaiti populations was 27 (52.9%) and 19 (37.2%), respectively. No correlation was observed with the prognostic parameters.


CONCLUSION
The frequency of HPV in breast carcinoma cases in Kuwait was 35.4% (CISH). Of those, 52.9% were Kuwaitis in whom both low- and high-risk HPV types were detected.


B
reast cancer is the most common form of cancer among women in Kuwait.In 2001, the breast cancer age-standardised incidence rate (ASR) was 36 among Kuwaiti and 39.5 among non-Kuwaiti female residents. 1Since then, its incidence has increased rapidly.Al-Shaibani et al. noticed that in Kuwait the majority of breast cancer patients were 40-49 years of age, and increased breast carcinoma risk was correlated with the menopause, recent hormone replacement therapy and family history. 2ell-known breast cancer risk factors associated with the development of breast cancer are age, familial history, hormonal intake, early menarche, late menopause, multiparity, first pregnancy at >30 years, obesity and personal history.Nevertheless, in 50-80% of cases, known risk factors have not been identified which has generated an interest in identifying new factors related to this cancer. 2,3he three most studied viruses found to cause breast cancer in humans are mouse mammary tumour virus (MMTV), the Epstein-Barr virus (EBV) and the human papilloma virus (HPV), with MMTV and EBV occurring in up to 37% and 50% of breast carcinoma cases, respectively. 4HPV is accepted as carcinogenic in human cervical, anogenital, and head and neck cancers.Molecular and epidemiological studies have shown that a persistent infection with high-risk types of HPV is the most important risk factor for both cervical cancer and its precursors. 5Recently, the possibility of HPV being aetiologically-related to breast cancer has been debated.However, the causal role of HPV infection in the development of breast carcinoma remains controversial. 6eports on the distribution of HPV infection in breast cancer are not only limited but also highly controversial. 7[10][11] However, a moderate frequency of 20-48% HPV infection and as high as 60-85% of HPV infection in breast carcinomas has been reported. 74][15][16][17][18][19][20][21][22][23] The prevalence of HPV-positive breast cancer in these studies was reported to vary from 4% in Mexican to 86% in American women. 13In all the studies, high-risk HPV was found in tumour tissue only and not in the surrounding normal tissue, with the exception of the study from Turkey where the virus was also detected in normal tissue but at a lower expression than in the cancerous tissue. 20ost of the studies have used standard and nested polymerase chain reaction (PCR) techniques.The disadvantage of this technique is its inability to confirm whether or not a positive reaction is in the mammary epithelial cells. 24Very few studies have used in-situ molecular methods for the demonstration of HPV, and in three of them a comparison between solution PCR and in-situ hybridisation methods for the same set of specimens was made and identified oncogenic HPV in mammary epithelium of 12% of the collective total versus 22% using standard or nested PCR. 18,13- 15,22,24In-situ methods have the potential to establish -Very few studies are available which correlate the presence of HPV with prognostic parameters in cases of breast carcinoma.
-The frequency of HPV-positivity in breast carcinoma could help guide the HPV vaccination policy.
more accurately the frequency of HPV in breast cancer tissue because of their ability to localise the signals in individual cells which can be categorised as benign or malignant.
The association of breast carcinoma with HPV has yet not been studied in Kuwait.The aim of this study was to document the association of HPV and its types (6, 11, 16, 18, 31 and 33) with breast carcinoma cases in Kuwait, both in Kuwaitis and non-Kuwaitis.The study also aimed to identify the correlation of HPV and its types with known prognostic and predictive markers, namely age, histological grade, tumour size, lymph node metastases, oestrogen (ER) and progesterone (PR) receptor expression and human epidermal growth factor receptor 2 (HER2/neu) amplification status.The secondary aim of this study was to compare HPV detection with immunohistochemistry (IHC) versus chromogenic in-situ hybridisation (CISH).

Methods
Archived formalin-fixed paraffin-embedded tissue from 144 cases of invasive ductal breast carcinoma, from the Department of Pathology at Hussain Makki Al-Juma Center for Specialized Surgery (HMJCSS) in Kuwait, were studied over a 3-year period (2009 to 2011).The study was performed according to the guidelines of the local ethics committee which conforms to the Helsinki Declaration.Cases were selected in which the following clinicopathological data were available: age of the patient at the time of diagnosis, histological grade and size of tumour, lymph node status, ER, PR and HER2/neu status.
Histology slides were reviewed and histological grading of the tumour was done using the modified Scarff-Bloom-Richardson grading system. 25The size of the tumour, the number of lymph nodes involved, and the ER, PR and HER2/neu status were reviewed for uniformity.
Paraffin-embedded tissue sections of the 144 breast cancer cases were analysed for HPV status by IHC and CISH.For IHC, the prefixed unstained tissue sections were used to demonstrate HPV by the standard IHC methodology using antibody from Dako (Carpinteria, California, USA), dilution of antibody (1:100) and the detection system, Dako EnVision, FLEX/HRP, High pH (Dako Global R & D, Glostrup, Denmark).Positive and negative controls were run with each batch.HPV staining was considered positive when there was distinct nuclear staining.
For CISH, a CISH implementation kit AP-NBT/ BCIP (Zytovision, Bremerhaven, Germany) was used.In CISH, all the cases were first screened for HPV status using the Zytovision biotin labelled HPV screening probe that allows the simultaneous screening of 7 HPV types (6, 11, 16, 18, 31, 33 and  35) without determination of specific HPV types.Positive and negative controls were run with each batch.HPV staining was considered positive when there was distinct nuclear staining.
For HPV typing, all positive cases were tested for the specific HPV types, namely the low-risk: 6 and 11; intermediate-risk: 31 and 33, and high-risk: 16 and 18, types using the specific biotin labelled 6 and 11, 31 and 33, and 16 and 18 Zytofast HPV DNA probes by the CISH technique.
The HPV types detected were correlated with the prognostic markers, namely age, histological grade, tumour size, lymph node metastases, ER, PR and HER2/neu.Data were analysed using Statistical Package for Social Sciences (SPSS), Version 17 (IBM, Corp., Chicago, Illinois, USA).Univariate analysis was used to compare the clinical and pathological features with the HPV types.Fisher's exact test was used to find the association between the two techniques of CISH and immunostaining.For all analyses, P <0.05 was considered statistically significant.
HPV was detected by CISH in 51 (35.4%) cases and by IHC in 24 (16.7%).The correlation of the HPV status in breast carcinoma cases by CISH and IHC.A good correlation was seen in 19 (13.2%) and 88 (61.1%) cases, being positive and negative, respectively, for HPV by both methods.The sensitivity and specificity of HPV-IHC was 37.3% and 95.6%.The sensitivity and specificity of HPV-CISH was 79.2% and 73.3%.This association was found to be significant by Fisher's exact test (P <0.001).
HPV typing performed on 51 cases found positive by CISH were HPV 6 and 11 positive in 23 (45.1%) cases, HPV 16 and 18 positive in 6 (11.8%) cases, and HPV 31 and 33 positive in 11 (21.6%)cases.A combination of HPV 6 and 11, and 16 and 18 was seen in 2 (3.9%); and a combination of HPV 6 and 11, and 31 and 33 was seen in 7 (13.7%)cases.The distribution by nationality of HPV types in breast carcinoma cases determined positive for HPV-CISH is shown in Table 1.
The distribution of HPV types and the age groups of the cases with breast carcinoma are shown in Table 2.The low-risk HPV types 6 and 11 were seen in all decades of life.The prevalence was higher in patients greater than 30 years of age.The high-risk HPV types 16 and 18, and intermediaterisk HPV 31 and 33 were detected in the fifth, sixth and seventh decades of life.
The correlation of HPV types determined by CISH according to age, histological grade, tumour size, lymph node metastases, ER, PR and HER2/ neu status is shown in Table 3. Univariate analysis between the HPV status determined by CISH was done with the histological grade of the tumour, size of the tumour, lymph node metastases, ER, PR and HER2/neu.No statistically significant differences were observed.Linear-by-linear association did not show significant correlation with the prognostic parameters although a trend was observed in that HPV types 16 and 18 were associated with lymphnodal metastases; however, the numbers were too limited to come to a definitive conclusion.

Discussion
This study documented HPV and the various types (6, 11, 16, 18, 31 and 33) in Kuwaiti and non-Kuwaiti women.We found HPV determined by CISH in 51 (35.41%) cases of invasive ductal carcinoma of the breast.
The reported prevalence of HPV in breast tissue between 1992 and 2012 varies between 4% in Mexico to 86% in the USA. 26HPV prevalence as determined by the PCR method was seen in 61.1% cases in Syria. 23

Simoes et al. conducted a comprehensive systematic review of studies
HPV = human papilloma virus; CISH = chromogenic in situ hybridisation; N/A = not applicable; [ ] = percentage within HPV types.
addressing worldwide HPV prevalence rates. 27In this meta-analysis of 29 studies comprising of 2,211 samples, the overall prevalence of HPV in breast carcinoma was 23% with great variation worldwide, ranging from 13.4-42.8%. 27This large worldwide variability could be related to methodology.Most studies utilise the PCR technique, which is more sensitive but less specific as compared to in-situ hybridisation techniques, which show the virus location. 24he HPV types found in our study, as determined by CISH, were HPV 6 and 11 (45.1%);HPV 16 and 18 (11.8%);and HPV 31 and 33 (21.6%).More than one HPV type was seen in 11 cases (21.6%).A combination of HPV 6 and 11 and HPV 16 and 18 was seen in two cases (3.9%) while a combination of HPV 6 and 11 and HPV 31 and 33 was seen in 7 cases (13.7%).In two cases (3.9%), all HPV types examined were identified [Table 1].In the Kuwaiti population, the distribution of HPV types 6 and 11; 16 and 18, and 31 and 33 were 40.7%, 7.4% and 18.5%, respectively, while in non-Kuwaitis it was 42.1%, 21.1% and 26.2%, respectively [Table 1].A combination of HPV 6 and 11, and 16 and 18 was seen in 7.4% of the Kuwaiti women while a combination of HPV 6 and 11, and 31 and 33 was found in 22.3%.All three HPV types were seen in one woman.In the non-Kuwaitis, only one (5.3%)case had a combination of HPV 6, 11, 31, 33  [Table 1].The high-and intermediate-risk HPV types were found in 16 of the 27 (59.3%)Kuwaiti women and 11 of the 19 (47.9%) non-Kuwaitis   1].This is the first time that the association of HPV types with breast carcinoma has been documented in Kuwaitis and non-Kuwaitis in Kuwait.We have no explanation for these differences, and more research work needs to be undertaken.Di Lonardo et al. were the first to report HPV 16 DNA using PCR in 29.4% of 40 breast cancer specimens. 14Using PCR, HPV types 11, 16 and 18 have been reported with increasing frequency from women living in USA and Brazil, while HPV type 18 was present in the majority of Australian women. 12,16,17HPV 33 was seen in 34.1% of breast carcinoma cases in Japan. 28Studies from the Middle East reveal that HPV types 18, 33 and 35 were present in cancerous and normal breast tissue in Turkish women. 20In Syria, a prevalence of HPV 16 (9%), HPV 33 (56%) and HPV 35 (37%) was observed. 23In 24 (34.78%)women, more than one HPV type was detected. 23Only two studies have been reported where in-situ PCR (IS-PCR), which involves amplification to detect HPV in breast tissue, was used. 13,24ery few studies have tried to correlate the presence of HPV in breast carcinoma cases with other prognostic markers. 12,29,30In our study, 51 HPV DNA-positive cases, as determined by CISH, were correlated with the age of the patients, histological grades, tumour sizes, lymph nodal metastases, ER, PR and HER2/neu expression.We did not find any significant correlation as our numbers were limited; thus, more work needs to be done in this regard.Kan et al. using PCR for detecting HPV, did not find any significant correlation between grade of tumour, mortality of patients, ER, PR, ERB-B2, p53 expression and the presence of p53 mutations. 12antu de Leon, using PCR for HPV determination, found that the greater the tumour size the greater the probability of finding viral DNA in the tumour, but not in tumours larger than 4 cm where the viral DNA appeared to be lost. 29However, there was no correlation with the histological grade, ER, PR or clinical stage. 29Mou et al., in a retrospective study from China of 62 breast cancer patients tested for HPV status by nested PCR, did not find any significant correlation between the presence of HPV and patient age, tumour size, metastases, or ER and PR status. 30here are several limitations to this study apart from the limited sample size.While the nucleic acid is degraded in archival paraffin-embedded tissue, CISH should still be considered a relevant technique when used on such material.Although IHC is not as sensitive a technique as PCR, we did not perform PCR in this study as in-situ methods have the advantage of detecting HPV in individual cells, which can be categorised as benign or malignant under conventional light microscopy.

Conclusion
The frequency of HPV in breast carcinoma cases in Kuwait was 35.4% when determined by the CISH method.The CISH method was useful in detecting HPV in malignant cells and should be considered as part of a standard investigation.Of the 51 HPVpositive cases, 27 (52.9%)were Kuwaitis.Both low-risk and high-risk HPV types, along with a combination of HPV types, were detected in the Kuwaiti population.No correlation was observed with the prognostic parameters.
d e c l a r at i o nThis research was supported by Kuwait University (Research Grant MG01/10).

Advances in Knowledge -Reports
on the distribution of human papilloma virus (HPV) infection in breast cancer patients are limited and highly controversial.-Thismanuscriptdescribes the prevalence of HPV in breast carcinoma cases in Kuwait and the distribution of the low risk(HPV 6 and  11)high risk (HPV 16 and 18) and intermediate risk(HPV 31 and 33)types in Kuwaiti and non-Kuwaiti women.Previous to this study, HPV and its subtypes in breast carcinoma in the Middle East had only been described in Turkey and Syria.

Table 1 :
Distribution by nationality of human papilloma virus (HPV) types in breast cancer cases testing positive for HPV by chromogenic in-situ hybridisation

Table 3 :
Distribution of human papilloma virus types in breast cancer cases with prognostic parameters (n = 51)

Table 2 :
Distribution of age according to human papilloma virus (HPV) types in breast cancer cases (n = 51)