Nasopharyngeal Carcinoma: Epidemiology, Etiology, Screening, and Staging
Epidemiology
Nasopharyngeal carcinoma (NPC), unlike many other head and neck cancers, is an uncommon neoplasm in most parts of the world. The age-adjusted incidence for both sexes is less than 1 per 100,000 population per year in many countries.1 On a global scale there are only 80,000 new patients per year, constituting 0.7% of all cancers, making it the 23rd-commonest new cancer in the world.2 It has a very distinctive geographic distribution: a high incidence is seen in the southern part of China, especially among the inhabitants of Guangdong province, including Hong Kong. The reported incidence for males is 17.8 per 100,000 and for females is 6.7 per 100,000. A range of intermediate rates is observed in populations in Southeast Asia, and in natives of the Arctic region such as northern Canada,3 North Africa, and the Middle East.4 One report on the incidence of this tumor in the Eskimo population in Greenland from 1955 to 1976 gave 12.3 and 8.5 per 100,000 per year for males and females, respectively.5 The incidence in northern African countries was 5.4 and 1.9 per 100,000 population, respectively, and these are roughly 10 times higher than the incidence in Europe.6
This geographic variation in the incidence of NPC is also seen within China: low rates of 1.1 per 100,000 population are seen in northern China, such as Harbin city. The distribution of the disease among different ethnic groups is also not uniform. In the southern province of Guangdong, the incidence of NPC is three times higher among Cantonese speakers than in Hakka, Hokkien, or Chiu Chau dialect groups.7 In Malaysia, the incidence of NPC is again higher among the Cantonese Chinese than among the Hokkien and the Teochiu.8 This may be associated with social and racial intermingling. In the Vietnamese city of Hanoi, the incidence of NPC is twice that in Ho Chi Minh City as there are more Chinese descendants in Hanoi.9 In the United States, a study of 1,645 NPC patients also showed the highest incidence in Chinese, followed by Filipinos then the whites and blacks.10 The survival was also the highest among the Chinese.11
The incidence of NPC remains high among Chinese who have migrated to North America, but is lower among Chinese born in North America than in those born in southern China.12,13 A similar higher incidence was seen among Chinese who have immigrated to Australia14 and the southern part of England.15 Migrant studies also showed that the incidence of NPC is higher in the off-spring of north Africans who have migrated to Israel than among native Israelis.16
These findings suggest that geographic, ethnic, and environmental influences together with other factors contribute to the etiology of nasopharyngeal carcinoma.17
In nearly all reports the incidence of nasopharyngeal carcinoma is 2 to 3 times higher in males than in females.1 As for the distribution of the disease within age groups, in low-risk regions, NPC incidence increases with increasing age.18 In high-risk regions, the peak incidence is around ages 50 to 59 years and declines thereafter19,20; there is also a minor peak among young adults,21–23 consistent with exposure to carcinogenic agents in early life.
Etiology
Salt-Preserved Fish
Preservation of fish and other food material with salt makes them partially putrefied, resulting in the accumulation of significant levels of nitrosamines, which are known carcinogens.24,25 Salted fish is a traditional weaning food for infants, especially among the Cantonese population. A case–control study has shown that weekly consumption of salted fish before 10 years of age is associated with a 3-fold increased risk of developing NPC.26 The consumption of salted fish in childhood rather than adulthood,27 and the duration and frequency of intake, were independently associated with increased risk of NPC.28 However, even in low-risk regions such as the northern provinces of China29 and the United States of America,30 preserved food is a common dietary component. Thus consumption of salted fish per se is unlikely to be the cause of nasopharyngeal carcinoma. The consumption of fresh fruit and vegetables is associated with a low risk of nasopharyngeal carcinoma,31 especially in childhood.32 This might be related to the antioxidant effects of fresh fruit and vegetables rather than to the specific food items.33
Epstein-Barr Virus
The Epstein-Barr virus (EBV) is a double-stranded DNA virus that belongs to the herpesvirus family. The virus is ubiquitous in the human population; it may cause infections such as infectious mononucleosis and has also been found to be associated with certain lymphomas and NPC. It infects and persists latently in the global population; most of the infection is subclinical and transmission is common in crowded conditions. In Hong Kong, most children have been infected by the age of 10 years.34 The B lymphocytes are the target of EBV infection and it is known to be associated with Burkitt′s lymphoma.35 The association of EBV with NPC was postulated as patients suffering from NPC also have higher EBV antibody titers than controls.36 In particular, the IgA class of antibodies to viral capsid antigen and early antigen are elevated in NPC patients.37,38
There were reports showing that these antibodies titers were elevated for some time before the presentation of NPC39,40 and that they are correlated with tumor staging41 and prognosis.42
Recently, through PCR, cell-free EBV DNA was detected in the plasma of patients suffering from nasopharyngeal carcinoma.43 The number of copies of this EBV DNA can be quantified44 and this has been shown to be related to stage of the disease45 and survival.46
However, EBV is unlikely to be the sole causative factor of NPC as it is ubiquitous in human populations. It is more likely that under specific circumstances the virus plays an oncogenic role together with other cofactors in leading to the development of NPC.
Genetic Factors
The high incidence of NPC among southeastern Chinese and their descendants suggests the presence of a genetic factor.47 This familial aggregation is also seen in intermediate-incidence48 and low-incidence populations.49 The genetic evidence is further supported by the fact that NPC is 4 to 10 times more common in first-degree relatives of patients suffering from nasopharyngeal carcinoma than in controls.50,51 This familial clustering might be the result of shared genetic susceptibility and related environmental factors; probably multiple genetic and environmental factors in combination are responsible, rather than a single gene.52
The human leukocyte antigen (HLA) genes were identified to be the genes responsible for the development of NPC. These genes encode proteins that present foreign antigens such as viral particles to the immune system for lysis. Individuals with HLA alleles that have reduced ability to present the viral particle, in this case the Epstein-Barr virus, for lysis have an increased risk of developing NPC.53 In a meta-analysis of the studies on southern Chinese, the finding of patients with HLA alleles HLA-A2, B14, and B46 is associated with an increased risk of developing NPC.54
Meta-analysis of comparative genomic hybridization studies revealed several genomic spots where chromosomal losses and gains were identified in NPC.55 These losses on chromosomes 3p, 9p, 11q, 13q, and 14q suggested that tumor suppressor genes at these loci might be involved in NPC development.56,57
Screening
In regions where NPC is endemic, EBV serology has been used for population screening aimed at detection of the disease in its early stage. In a prospective study performed in the early 1980s, 1,136 individuals were detected to have elevated IgA antibodies against the viral capsid antigen of EBV. They were examined regularly for 4 years. During this period, 35 NPC patients were detected, most of them (91.5%) were diagnosed in the early stages.58 Comparable results were reported from a similar study conducted in Guangdong province, China.59 The predictive value of EBV serology was also reported in a study from Taiwan that included 9,699 patients. Their EBV serological status was cross-checked with the cancer and death registry for a 15-year period. The results showed that the longer the follow-up period, the greater was the difference in cumulative incidence of nasopharyngeal carcinoma between those with elevated serology and those with negative serology.40 In a prospective study of 42,048 patients, this raised level of antibody titer could be detected for up to 10 years before diagnosis of the tumor. The mean duration of this preclinical serological elevation window was 37 ± 28 months.60
Although this elevation of antibody titer of EBV has also been reported in low-risk regions,61 its value in general population screening aiming to detect early stage disease for more effective treatment is not likely to be cost-effective, even in high-incidence regions. It might be applicable in the screening of family members such as first-degree relatives in a high-risk region. In a prospective study of 1,199 asymptomatic family members of NPC patients, screening identified 16 patients.62 The sensitivity and specificity of EBV serology were 88.9% and 87.0%, respectively, and these patients have a 10 to 12% higher survival rate than those patients diagnosed without screening.63
When the NPC cell dies, the associated EBV is released and its quantity in the plasma can be detected by real-time polymerase chain reaction (PCR). This plasma cell-free EBV DNA is detectable in 96% of patients,44 especially in the undifferentiated carcinoma cells. It correlates with the stage of the disease and has prognostic significance.45 It is also of value in the diagnosis of recurrences and can be detected before the appearance of the recurrent tumor.64 Studies on EBV DNA detection were mainly performed on patients suffering from NPC; its role as a screening tool in the general population has not been determined.65
Staging
As with other malignancies, a clinical staging system for nasopharyngeal carcinoma is essential for planning treatment and evaluating the results of therapy. A simple staging system for NPC was first described in 1952.66 With this system, each stage covered too wide a range of tumor involvement and did not accurately reflect the clinical condition. Over the years a number of staging systems have been used for NPC; the Union for International Cancer Control/Union International Contre le Cancer (UICC) and the American Joint Committee on Cancer Staging (AJCC) systems are preferred in Europe and America, respectively, while Ho′s system is frequently used in Asia.67,68 The nodal classification in Ho′s system has demonstrated its prognostic significance, but its stratification of the T stages into five stages differs from other staging systems for malignant disease. The lack of a universally accepted staging system reflects, to some extent, the inadequacies of the various existing staging classifications.69
The development of a revised staging system since the 2000s has taken into consideration the experiences gained from various centers around the world. It also takes into account several prognostic factors, such as the extension of primary tumor to the paranasopharyngeal space,70 skull base erosion, the involvement of cranial nerves,71 and also the position and size of the cervical nodes.72
Both the UICC and the AJCC systems assess tumor extent in the nasopharynx by considering the number of tumor-affected sites within the nasopharynx, while the Ho system classifies all tumors confined to the nasopharynx as T1 disease. The UICC and AJCC systems have been unified since 1992 and the clinical staging is the same in their recent published manuals.73,74 In the staging system of 2009, T1 stage included all tumors that were confined to the nasopharynx or extended locally such as anteriorly to the nasal cavity, or inferiorly to the oropharynx. This is because the nasopharynx is an irregular structure and the margins of the walls are not precise, and the exact limit of the tumor is also difficult to define. There is also the additional problem of submucosal extension, which is difficult to determine even with endoscopy.75 The whole nasopharynx and its vicinity are included in the radiation field; so as long as the tumor is superficial and confined to the mucosa, it is classified as T1 disease. On the other hand, lateral tumor extension to involve the paranasopharyngeal space is important and this can be accurately determined by cross-sectional imaging. Thus, T2 stage included tumor that had extended to the paranasopharyngeal space. The T3 stage covered tumors that had extended to the skull base or other paranasal sinuses. T4 stage covered tumors that had extended into the infratemporal fossa, orbit, hypopharynx, and cranium, or had affected the cranial nerves.
The disparity between the UICC/AJCC and Ho staging systems is greater in the N stage. The UICC/AJCC currently recognizes the size of the lymph node as an important factor. In N staging for other head and neck cancers, staging between N1 and N2 depends on whether the node is smaller or larger than 3 cm, and the difference between N2 and N3 is the nodal size of 6 cm. Another criticism common to the various N-staging systems is that the retropharyngeal nodes, which are the first-echelon nodes, are not taken into account by any staging systems. These nodes, although difficult to examine clinically, can now be assessed by CT or MRI.76
These factors were addressed in the nodal staging section of the UICC/AJCC nasopharyngeal cancer staging system (Table 23.1). Here only a measurement of 6 cm is considered as a factor in size. Laterality and level of involvement such as the retropharyngeal region and the supraclavicular fossa are other important factors employed in the determination of N-staging. Under the current system, N1 refers to unilateral nodal involvement less than 6 cm in diameter and not reaching the supraclavicular fossa. The presence of bilateral retropharyngeal nodes, as long as they are less than 6 cm in diameter, remains N1. Bilateral nodal disease in the neck that has not reached N3 designation is classified as N2, irrespective of the size, number, and anatomical location of the nodes. Stage N3 disease refers to lymph nodes larger than 6 cm (N3a), or nodes that have extended to the supraclavicular fossa (N3b). There is agreement on M-staging, where M1 represents distant metastases, including any lymph node involvement below the level of the clavicle. The current unified staging system has enabled patients to be staged more precisely and simply; it also gives a better predictor of prognosis.77–79