Human papillomavirus and non-small cell lung cancer
Abstract
Lung cancer is the most common cause of cancer-related deaths in the world, causing more than one million deaths worldwide each year. Human papillomavirus (HPV) are small non-enveloped DNA viruses that infect squamous epithelial cells. Relevant studies have reported lung cancer-related HPV infection rates that fluctuate between 10% and 80%, depending on the various research methods and geographical factors. Various scholars gathered statistics from global research reports and found that 22.4% of the patients with lung cancer presented with an HPV infection, which suggested that HPV infection may relate to the tumorigenesis of non-small cell lung cancer. This article will review the history and discovery of HPV, the correlation between HPV and lung cancer development, and carcinogenesis caused by HPV regulatory genes, such as p53, p21, p16INK4a, and genes related to hypermethylation and genome instability in lung cancer patients with HPV infection. In addition, because studies have highlighted the difference in clinical prognosis for HPV-positive and HPV-negative patients, articles demonstrating the correlation between HPV infection and prognosis for lung cancer patients will be reviewed.
Introduction
Lung cancer is the leading cause of cancer-related mortality worldwide, accounting for 3.1% of the annual total number of deaths worldwide, and 17.6% of cancer-related deaths.1 The causes of cancer can be extremely complex. However, epidemiological studies have found that the incidence of lung cancer is related to smoking tobacco (including second-hand smoke), air pollution, work-related factors (such as the inhalation of asbestos fibers), radioactive substances (such as radon), hereditary factors, chronic diseases, gender factors, viral infections, and various risk factors.
Possible viruses that lead to lung cancer include human papillomaviruses (HPVs),2 the John Cunningham virus (JCV),3 the simian vacuolating virus 40 (SV40), the BK virus, and the cytomegalovirus (CMV).4 These viruses can influence the cell life cycle, impede cell apoptosis, and disrupt cell division. In recent years, international scholars have conducted numerous studies on lung cancer and HPV infections.2, 5-9 Relevant studies have reported lung cancer-related HPV infection rates that fluctuate between 10% and 80%, depending on the various research methods (such as polymerase chain reaction [PCR]; in-situ hybridization [ISH]; or nested ISH [NISH]), and geographical factors. Recently, Finnish scholars gathered statistics from global research reports and found that of 7381 lung cancer diagnoses from various regions, 22.4% (1653) of the patients also presented a HPV infection.10 A recent literature review highlighted the relationship between HPV and lung cancer.10-12 In this article, we explored HPV and lung cancer, with a focus on molecular events.
Human papillomaviruses (HPVs) and cancer
HPV is a small DNA virus (diameter of 55 nm) comprising six early regulatory genes, namely E1, E2, E4, E5, E6, and E7, and two late genes, namely L1 (major viral capsid protein), and L2 (minor viral capsid protein). Currently, more than 100 different HPV types are known. A HPV genome is now defined as a new HPV type when it shows a more than 10% dissimilarity in the combined nucleotide sequences of the E6, E7, and L1 genes, when compared with those of any previously known type; a subtype is defined when it shows a 2–10% dissimilarity, and a variant when less than 2%.13-16 Viruses are among the causes of cancer.17 The main viruses associated with human cancer are as follows: hepatitis B virus (HBV), which causes liver cancer by inducing chronic viral infection;18 human T-lymphotropic virus (HTLV), which can lead to adult T-cell leukemia (ATL);19 Kaposi's sarcoma-associated herpes virus (KSHV), which can cause Kaposi's sarcoma (KS) and body cavity lymphomas (BCLs); and Epstein-Barr virus (EBV), which can lead to the development of Burkitt's lymphoma, Hodgkin's lymphoma, and nasopharyngeal carcinomas (NPCs).20 Over 5% of the cases of cancer worldwide are affected by continuous HPV infections. Among the over 100 types of HPV, approximately 40 cause genital tract infections, and 15 expose women with the virus to a high risk of cervical cancer; HPV 16, 18, 31, 33 and 45 are considered high risk; 6 and 11 are considered low risk. It is known that HPV is the primary cause of cervical, skin, anal, and penile cancer.21
Correlation between HPV and lung cancer development
Research into the correlation between lung cancer and HPV has been proposed in a number of countries. Relevant studies have reported that 79% of patients diagnosed with squamous cell lung cancer were also infected with HPV.22 According to Thomas et al.,8 78.3% of patients diagnosed with lung cancer involving squamous cell carcinomas and adenocarcinomas were also infected with HPV, and HPV could be identified in lung adenocarcinoma cells and nearby squamous cell carcinomas. Among individuals with lung adenocarcinoma in the Western world, the prevalence of pulmonary HPV infection ranges from 0% to 36%.10, 22, 23 In Asia, the prevalence of pulmonary HPV infection ranges from 9% to 78%.23 These results suggest that the correlation between HPV infection and lung cancer is influenced by ethnic and geographical differences. Thus, the involvement of HPV infection in lung cancer tumor development in Taiwan, especially among non-smoking women, is a valuable subject of study.
Cheng et al. examined the difference in the HPV infection rates among people with lung cancer and people without cancer. They analyzed two high-risk (HPV 16/18) and two low-risk (HPV 6/11) HPV infections, including 141 lung cancer patients and 60 non-cancer participants (control group), employing a nested polymerase chain reaction (nested PCR). The results showed infection rates of 35.5%, 41.1%, 28.4%, and 10.0% for the four HPV types tested on the participants with lung cancer. The HPV 6, 16, and 18 infection rates for participants with and without lung cancer differed significantly, indicating that infections of HPV 16, 18, and 6 may be related to the development of lung cancer. Including the variables of smoking behavior and gender, the HPV 16/18 infection rates for non-smoking women with lung cancer were astonishingly high, at 60% and 73%, respectively, significantly higher than that for men with lung cancer (HPV 16: 24%, HPV 18: 26%). However, male smokers had the highest HPV 6 infection rate at 38.1%, and female non-smokers the lowest rate at 11.1%. This indicates that, considering gender and smoking behaviors, various types of HPV may affect the development of lung cancer. HPV 16/18 infections may be associated with the development of lung cancer in non-smoking women in Taiwan.2 Taiwanese women have a history of cervical intraepithelial lesions and/or HPV 16/18 infection.24 In a case-control study on the association between Taiwanese cervical neoplasia and HPV infection, it has been shown that HPV DNA was detected in 91% of high-grade cases, 54% of low-grade cases, and 9% of controls. HPV 16 was the predominant type among HPV-positive high-grade cases, as reported in Western countries.25 In addition, it is well known that HPV 6/11 are frequently associated with upper aerodigestive and respiratory diseases. Cheng et al. showed that a higher prevalence of HPV 6 was detected in lung tumors of smoking male patients with early tumor stage than those with advanced stages, but not in non-smoking male and female patients.26 Furthermore, Aguayo et al. found that HPV was detected in 20 (29%) out of 69 lung carcinomas by PCR and Southern blot, and was more frequently detected in squamous cell carcinoma (SQC) than in adenocarcinomas (46 vs. 9%, P = 0.001). Interestingly, the HPV-16 E2/E6 ratio, estimated from real-time PCR analysis, was much lower than the unity, suggesting that at least a partial HPV-16 genome was integrated in all but one HPV-16-positive SQC. Although the viral load was in general very low and HPV E6 expression was either weak or non-existent, further studies seem warranted to examine aetiological involvement of high-risk HPV in lung carcinogenesis.21 Therefore, the molecular mechanism for HPV infected lung tumorigenesis would be different between smoking and non-smoking lung cancer patients.27
Although a number of reports revealed that no HPV infection or only a low frequency of HPV infection was found in lung tumors,28, 29 some data from Japan and Northern European countries, such as Finland and Norway, showed that significantly higher frequencies of HPV infection (69–78.3%) were determined in lung carcinomas.30, 31 The association between HPV infection and lung cancer is, therefore, reasonably suggested to be geography- and race-dependent.
Possible HPV infection transmission paths for lung cancer patients
Recent research has found that HPV infections are not only related to cervical cancer, but are also correlated to many other human cancers, such as head and neck, laryngeal, esophageal, and colorectal cancer. For patients with head and neck cancer, the HPV infection rate in non-tumor tissues ranges between 18.5 and 35.9%, and can reach 34.5% in tumor tissue (416/1205). Additionally, head and neck cancer infection rates for high-risk type HPV 16 and 18 are 40.0% and 11.9%, respectively. Among all head and neck cancers, patients with oral cancer exhibit the highest HPV infection rate at up to 59%, followed by those with pharyngeal and laryngeal cancer at 43.0% and 33.0%, respectively. The oral HPV infection rate for the average person ranges between 1% and 60%.32 Consequently, HPV in lung tumor tissue may develop through an infection in the mouth or respiratory tract. Olatunbosun et al.33 found that HPV DNA could be detected in the sperm of 53% (24/45) of patients with a current or previous HPV infection. Additionally, 8% of the sperm samples from participants in the control group, who were not infected with HPV, contained HPV. Research has shown that age is another factor that influences HPV infection in oral mucosa cells, particularly for people under 20 years of age. The infection rate is 8.7% for children under seven years of age, 0% for children between seven and 13 years, and 5.2% for children between 13 and 20 years. However, these HPV infections were not affected by gender or ethnicity. Additionally, in the participant group comprising adolescents between 13 and 20 years of age, the HPV infection rate was not correlated to smoking or sexual behavior. This indicates that for adolescents, HPV infections may primarily result from vertical transmissions between mother and infant during birth.34
Using nested PCR, Chiou et al. measured the prevalence of HPV 16/18 in the blood of lung cancer patients. They reported a HPV 16 rate of 47.7%, which exceeded the 12.6% reported for the non-cancer control group (P < 0.0001). The HPV 18 infection rate for the two groups was 30.9% and 5.2% (P < 0.0001), respectively. Women with lung cancer had a higher prevalence rate (57.6%), compared to men with lung cancer (41.1%, P = 0.048). The risk-adjusted rates showed that the number of participants identified with HPV 16 DNA in their blood in the lung cancer group was 6.5 times that of the group without cancer (P < 0.0001). The number of participants with HPV 18 DNA in the lung cancer group was 9.2 times that of the group without cancer (P < 0.0001). Additionally, the number of participants with both HPV 16 and 18 DNA in the lung cancer group was 75.7 times that of the group without cancer (P < 0.0001). These results imply that HPV DNA in blood may be a risk factor for lung cancer patients.35
In Norway, Hennig et al. analyzed how HPV DNA infected lung tumor tissue in 75 participants with lung cancer (bronchogenic carcinoma). They reported a 49% infection rate (37/75). In the 37 lung cancer patients with a HPV infection, 34 had a history of cervical diseases, with a 74% cervical HPV infection rate (25/34). They inferred that the HPV found in the patients' lung tumor tissue may have originated from cervical infections, and been transmitted to the lungs through the blood.23 In Japan, Iwamasa et al. investigated high HPV infection rates. They reported that HPV DNA was observed in lung tumor tissue, and that 80% of the patients had a history of cervical diseases.6 Therefore, the HPV found in the lung tumor tissues in these two countries probably originated from cervical infections and was transmitted to the patients' lungs via the blood. However, Cheng et al. indicated that most patients did not have a history of cervical disease, although they reported HPV 16 and 18 infection rates in lung tumor tissues of up to 35.5% and 41.1%, respectively.
The cell surface receptor for HPV is still unidentified. Several reports have indicated that α6β4, α6β1 integrin and heparan sulfate proteoglycans may be the receptor for HPV.36-41 Based on our knowledge, α6β4, α6β1 integrin, and heparan sulfate proteoglycans have been expressed in T lymphocytes, B lymphocytes, and macrophages.42-45 A previous study showed that HPV could infect peripheral blood lymphocytes (PBLs), especially lymphocytes, monocytes and macrophages.46 Thus, we suggest that HPV infection may be mediated through blood circulation.
Carcinogenesis caused by HPV regulatory genes
Serial research regarding HPV infection and the molecular mechanism of lung cancer
p53
Werness et al.47 found that HPV 16 and HPV 18 E6 are capable of binding to p53 protein, and Scheffner et al.48 verified that E6 protein binds to and degrades p53 through the ubiquitin regulatory pathway. E6 degrades p53 mainly through E6-associated protein (E6-AP). The mechanism is as follows: Once E6 binds to E6-AP, p53 binds to the p53-binding domain in the E6 protein sequence. Then p53 is degraded in the ubiquitin regulatory pathway.47, 48 E6-AP has a molecular mass of approximately 100 kd, and its function is similar to that of E3 protein in the ubiquitin regulatory pathway.49 For E6 to degrade p53, both E6 and E6-AP must be present simultaneously. If mutated E6-AP cannot bind to E6 or mutated p53 cannot bind to E6, the p53 will not degrade. Therefore, the E6/E6-AP composite, instead of simply E6-AP, functions as an E3 ligase during the degradation of p53.50, 51 In addition, the polymorphism of the p53 gene codon 72 influences the effectiveness of p53 degradation by E6. The polymorphism of the p53 gene codon 72 mainly causes amino acid sequences to transform from Arg→Pro. Arg and Pro have three genotypes, that is, Arg/Arg, Arg/Pro, and Pro/Pro. Arg/Arg has a higher susceptibility to E6, followed by Arg/Pro, and Pro/Pro has the lowest susceptibility to E6.52, 53
The inactivation of p53 by HPV 16/18 E6 is crucial to the cancerization of cervical cancer. Cheng et al. examined the relationship between HPV 16/18 E6 and p53 protein expression in 122 lung cancer tissue samples. They found that the expression of HPV 16/18 E6 was negatively correlated to the expression of p53 protein. The results of real-time reverse transcription PCR analysis showed that E6-positive cancer cells possessed less p21WAF1/CIP1 and mdm2 mRNA, than E6-negative cells. Additionally, both HPV-positive and HPV-negative lung adenocarcinoma cell lines have been successfully developed for research. The expression of E6 protein in cells infected with HPV 16 was confirmed using the Western blot technique. The amount of p53 in high expressions of E6 protein was lower than that in low E6 expressions, and the p21WAF1/CIP1 and mdm2 mRNA in E6-positive cancer cells was lower than that in E6-negative cells. Furthermore, RNA interference and cell comparison showed that, in E6 knockdown cells, expressions of p53, p21,WAF1/CIP1 and mdm2 mRNA were restored. The results indicated that the expression of HPV 16/18 E6 in HPV-DNA-positive lung cancer is partially caused by the regulatory expression of p21WAF1/CIP1 and mdm2 mRNA through p53 inactivation.46
p21
Chang et al. and Chao et al. indicated that the inactivation of human DEAD-box (DDX) RNA helicases by the nucleoproteins HBx and HCV viruses may occur through a p53-independent pathway, and that the suppression of p21 transcription results in cell proliferation.54, 55 Wu et al. found that in the first and second stages of lung cancer, p21 expression was positively correlated to DDX expression, but negatively correlated to E6 expression, by using real-time PCR and immunohistochemistry (IHC) to analyze DDX and p21 expressions in 138 lung cancer tissue samples. However, an inferior relapse-free survival rate was observed for early-stage lung cancer with a lower p21 expression.56
P16INK4A
Located on chromosome 9p21, P16INK4A suppresses CDK4 and cyclin D1 effects and suppresses cell proliferation after binding to cyclin-dependent kinase 4 (CDK4).57, 58 Relevant studies have shown that hypermethylation of the P16INK4A promoter can typically be detected in non-small-cell lung carcinomas or cancer.59, 60 Belinsky et al. suggested that P16INK4A could be used as an early indicator of lung cancer. However, whether a correlation exists between P16INK4A hypermethylation and smoking is still disputed.61 Because only 10% of women diagnosed with lung cancer in Taiwan are tobacco smokers, Wu et al. investigated hypermethylation of the P16INK4A promoter. For 162 participants with lung cancer, the proportion of male smokers, male non-smokers, and female non-smokers, was 59.7%, 36.6%, and 60.3%, respectively, demonstrating the influence that sex and smoking behavior has on hypermethylation of the P16INK4A promoter. In addition, for female non-smokers with hypermethylation of the P16INK4A promoter, 70% also had HPV, whereas 33% of the female non-smokers who did not have HPV, did exhibit hypermethylation of the P16INK4A promoter. Furthermore, a significant difference (P = 0.017) between the rates of HPV infection and hypermethylation of the P16INK4A promoter was only found for non-smoking women, but this phenomenon was not discovered for smoking or non-smoking men. Additionally, a negative correlation between p16INK4A immunostaining and hypermethylation of the P16INK4A promoter was only observed in the lung cancer tissue of female non-smokers. Therefore, we contend that the cancerization mechanism of HPV-related lung cancer is at least partially caused by hypermethylation of the P16INK4A promoter, which inactivates P16INK4A.62
Hypermethylation
Relevant studies have identified excessive expressions of DNA methyltransferases, such as DNTM1, DNMT3a and DNMT3b, as common events in human lung cancer.63, 64 Lin et al. found a significant correlation (P < 0.0001) between hypermethylation of the P16INK4A promoter and DNMT protein expression in lung cancer patients with HPV infection. Specifically, the expression of DNMT3b was significantly correlated to hypermethylation of the P16INK4A promoter (P < 0.023) and HPV infection (P < 0.001). However, the significant correlation between hypermethylation of the P16INK4A promoter and DNMT3b expression was only observed in women diagnosed with lung cancer (P = 0.035). This indicates that the hypermethylation of the P16INK4A promoter caused by DNMT3b expression strongly demonstrated the cancerization mechanism for female non-smokers diagnosed with lung cancer and infected with HPV.65
Genome instability
Scholars have investigated fragile histidine triad protein (FHIT) located on chromosome 3p14.2 to understand the correlation between genome instability and HPV. The results show that the loss of heterozygosity for FHIT alleles is frequently observed in lung cancer and cervical cancer cells.66 Tseng et al.67 found that FHIT protein inactivation commonly occurs in pre-stage tracheal cancer and first-stage non-small-cell lung cancer (NSCLC), and that the loss of FHIT protein expression posed a latent contribution to lung cancer cancerization. Wilke et al.68 found that HPV could anchor or embed into a fragile FRA3B site near the FHIT gene, resulting in a loss of heterozygosity for FHIT alleles. Wang et al. investigated 150 lung cancer tissue samples and the results suggested that the loss of heterozygosity for FHIT alleles is commonly observed in men, smokers, and people diagnosed with squamous cell cancer or carcinoma. However, the likelihood of a loss of FHIT allele heterozygosity among women with lung cancer is significantly higher for those with HPV 16 (46%), compared to those without HPV 16 (16%). These results indicate that high incidence rates for losses in FHIT allele heterozygosity contribute to cancerization for female lung cancer patients infected with HPV 16.69
Epidermal grow factor receptor (EGFR)
The potential relationship between HPV infection and sensitizing epidermal growth factor receptor (EGFR) mutation expression in lung cancer is still limited. Na et al. found that in 108 patients with squamous cell carcinoma of the head and neck, none of the patients with EGFR mutations (16%) were HPV positive (9%).70 Márquez-Medina et al. first suggested a direct relationship between HPV infection and sensitizing EGFR mutation expression by retrospectively analyzing 40 NSCLC samples. However, the study did not provide definitive conclusions because of the limited sample size.71 Kato et al. recently confirmed a significantly higher incidence of HPV DNA in lung cancers harboring sensitizing EGFR mutations. The study examined the association between the HPV presence and mutations in exons 19 and 21 of the EGFR gene in Japanese lung cancer patients. Thirteen (31%) out of 42 cases had EGFR mutations. HPV DNA was found in 7/42 (17%) lung tumors. The presence of HPV DNA was significantly related to EGFR mutations (P = 0.021), especially in adenocarcinomas of the lung (P = 0.014). HPV-positive lung tumors accounted for 38% and 7% of those with and without EGFR mutations, respectively. These results suggest that EGFR mutations are associated with the presence of HPV in Japanese patients with lung cancer.72
Correlation between HPV infection and prognosis for lung cancer patients
According to numerous studies, several prognostic factors can be used to predict the survival rate of patients diagnosed with NSCLC. The relevant prognostic factors are as follows: (i) clinical factors, such as age;73 (ii) laboratory findings, such as pre-surgery carcinoembryonic antigen (CEA);64 (iii) radiological findings, such as tumor size;74 (iv) surgical assessments, such as mediastinal lymph node evaluations;75 (v) pathological findings, such as pathological staging results;76 and (vi) biological predictive factors, including indices of cell proliferation (K-ras oncogene activation and change),77 modifications in cell loops (Rb77 and p2178), apoptosis (p5379 and bcl-280), and angiogenesis, such as the vascular endothelial growth factor receptor (VGFR).81 These factors contribute to evaluations of patient survival rates.
Whether the prognosis of cancer patients is affected by HPV infection remains disputed by academia. Research regarding oral cancer has indicated that patients with HPV 16 have higher survival rates.82 Tonsillar cancer patients with HPV DNA expressions in tumor tissues have a three-year survival rate of 65.3%, which is significantly higher than that for patients without HPV (31.5%). Their five-year survival rate is also higher (53.5% vs. 31.5%).83 Bezerra et al. found no correlation between HPV infection and prognosis for penile cancer patients.84 Lo et al.85 indicated that the prognosis for cervical cancer patients with HPV 18 was worse than that for patients without, although HPV 16 infection was not correlated to the prognosis. A possible reason for these inconsistent results is that HPV does not necessarily anchor and produce oncoproteins in host DNA, thereby minimally influencing cells. In addition, relevant studies have shown that patients with HPV-positive head and neck squamous cell cancer have higher survival rates than that of HPV-negative patients.86, 87 Ragin and Taioli88 conducted a meta-analysis on the correlation between overall survival rates, disease-free survival rates, and HPV infection for patients with squamous cell neck cancer. Their results showed lower death and relapse rates for HPV-positive head and neck squamous cell cancer patients, than for HPV-negative patients.
Existing literature has not adequately indicated whether HPV infection influences the prognosis for lung cancer patients. Iwamasa et al.6 reported that a superior prognosis is expected for lung cancer patients with HPV in the lung tumor tissue, whose tissue biopsies show higher Langerhans cell expression than for patients without HPV. Additionally, their results showed that patients infected with HPV show superior differentiation of lung tumor tissue. This may be why HPV-positive lung cancer patients have superior prognoses. Miyagi et al reported identical results.89 Hsu et al.90 examined the expressions of oncoproteins related to HPV 16/18 E6 in samples obtained from 148 men and 69 women diagnosed with first-stage NSCLC post-surgery. A significant difference (P = 0.055) was found between the 17 patients with expressions of HPV 16/18 E6 oncoproteins and the remaining 154 patients without. The patients with HPV had a comparatively superior survival rate.
Nevertheless, the higher survival rate for HPV-positive cancer patients has not been clearly explained. The assumptions include increased radiation response, viral antigen immunity monitoring, and a lack of field cancerization for non-smoking patients.91 In addition, HPV-positive tumors contain wild-type p53, which can suppress E6 oncoprotein activity. By contrast, specific p53 mutations always exist in HPV-negative tumors and these mutations are caused by smoking.91 In summary, the E6-related proteins in HPV-positive tumors that cause p53 degradation demonstrate an imbalance with HPV-negative tumors with p53 mutations.92, 93 Therefore, HPV-positive tumors may have a complete apoptosis process that is responsive to radiological and chemical therapies.94 The findings show that first-stage NSCLC patients without HPV 16 and 18 have inferior prognoses, thus, subsequent supplementary chemical therapies should be considered.
Future implications
Numerous molecular carcinogenesis mechanisms by which HPV infection causes cell cancerization have been identified, and studies have highlighted the difference in clinical prognoses for HPV-positive and HPV-negative patients. Consequently, if customized therapies and appropriate molecular targeted medication, such as telomerase inhibitors or TRAIL-sorafenib, can be provided to patients according to the status of HPV infection, concurrent therapies can enhance the effectiveness of clinical treatment for lung cancer patients with HPV.
Disclosure
No authors report any conflict of interest.
