Annals of Oncology 18: 29–35, 2007

original article doi:10.1093/annonc/mdl320

Published online 23 October 2006

Epidemiology of nasopharyngeal carcinoma in

the United States: improved survival of Chinese
patients within the keratinizing squamous cell
carcinoma histology
S.-H. I. Ou1,3*, J. A. Zell1,3, A. Ziogas2,3 & H. Anton-Culver2,3
1
Chao Family Comprehensive Cancer Center, Division of Hematology/Oncology; 2Genetic Epidemiology Research Institute; 3Division of Epidemiology,
Department of Medicine, School of Medicine, University of California Irvine, Irvine, USA

Received 2 June 2006; revised 19 July 2006; accepted 31 July 2006

Background: This study examined potential survival differences among nasopharyngeal carcinoma (NPC) patients
from various ethnicities in the United States.
Patients and methods: A total of 2436 newly diagnosed NPC patients from 1992 to 2002 were analyzed from the
population-based Surveillance, Epidemiology, and End Results (SEER) program. Five-year survival rate estimates and
Kaplan–Meier survival curves were calculated. Cox proportional hazard ratios (HRs) were used to identify independent
prognostic factors for survival.
Results: By multivariate analyses, early age of diagnosis, localized stage at presentation (versus distant, HR = 0.35;
P < 0.0001), radiation therapy (versus none; HR = 0.48; P < 0.0001), undifferentiated non-keratinizing carcinoma

original
article
(versus keratinizing squamous cell carcinoma; HR = 0.67; P < 0.0001), and Chinese ethnicity (versus Caucasian;
HR = 0.78; P = 0.0010) were associated with improved survival. Within keratinizing squamous cell carcinoma
histology, the survival advantage of Chinese patients remained even after adjustment for other prognostic factors.
Conclusions: The significant survival advantage of Chinese NPC patients within the keratinizing squamous cell
carcinoma histology contributed largely to Chinese ethnicity being an independent and favorable prognostic
factor for survival in NPC.
Key words: epidemiology, ethnicity, nasopharyngeal carcinoma, prognosis, SEER

background was further divided as being ‘differentiated’ and ‘undifferentiated’.

The differentiated subtype constituted the original WHO Type 2
Nasopharyngeal carcinoma (NPC) is endemic in Southeast Asia and the undifferentiated subtype constituted the original
but is relatively rare in the United States [1]. The high incidence WHO Type 3 [2, 4].
of NPC in Southeast Asia is likely due to increased susceptibility Epidemiologic studies in the United States have identified
for Epstein–Barr virus (EBV) infection to establish latency in WHO histologic type as an independent prognostic factor for
the nasopharynx [2]. In 1978, the World Health Organization survival in NPC [5, 6]. Additionally, Chinese- or Asian-
(WHO) classified NPC into three different histologic types: Americans have been observed to have improved survival when
keratinizing squamous cell carcinoma is classified as WHO Type compared with non-Chinese- or non-Asian-American patients
1, non-keratinizing carcinoma is classified as WHO Type 2 and [5–8], although one study from a single institution in the United
undifferentiated carcinoma is classified as WHO Type 3. WHO States did not find any survival advantage for Chinese NPC
Type 3 is the major histologic type in endemic areas whereas patients [9]. This superior survival of Chinese/Asian-American
WHO Type 1 usually comprises fewer than 5% of the endemic NPC patients is generally attributed to the much higher
population [3]. In 1991, WHO classification retained the prevalence of the favorable non-keratinizing carcinoma
keratinizing squamous cell carcinoma subtype (original WHO histology diagnosed among Chinese/Asian-American NPC
Type 1) and combined WHO Type 2 and 3 histologic subtypes patients [7]. Non-keratinizing carcinoma of the nasopharynx
into ‘non-keratinizing carcinoma’. Non-keratinizing carcinoma are more often controlled by ionizing radiation than
keratinizing histologies, and the 5-year survival rate is
*Correspondence to: Dr S.-H. I. Ou, Chao Family Comprehensive Cancer Center,
significantly better for non-keratinizing carcinoma than
University of California Irvine Medical Center, 101 The City Drive South, Building 56,
Room 241, RT 81, Orange, CA 92868-3298, USA. Tel: +1-714-456-8104; keratinizing squamous cell carcinoma of the nasopharynx (51%
Fax: +1-714-456-2242; E-mail: [email protected] versus 6%) [10]. In addition, non-keratinizing carcinomas are

ª 2006 European Society for Medical Oncology

original article Annals of Oncology

generally associated with EBV positivity and EBV positivity in ethical considerations
turn has been shown to be associated with improved survival This research study involved analysis of existing data from the
[11]. Most of the population-based studies in NPC are from aforementioned SEER database with no subject intervention. Information
endemic countries where the ethnic makeup of the population was recorded without identifiers linked to subjects. The University of
was fairly homogeneous and the undifferentiated non- California, Irvine, Institutional Review Board (IRB) approved this study
keratinizing type histology is the predominant histology. Thus, under the category ‘exempt’ status (IRB #2005-4265).
it remains unknown whether ethnicity is truly an independent
prognostic factor for survival in NPC in general and within each results
WHO histologic type. A recent study from Australia partially
addressed this question but was limited by the few numbers of age at diagnosis
WHO Type 1 NPC patients [12]. In the United States, every The mean age of diagnosis for the entire group was 52.7 years. The
WHO histology is diagnosed in each of the major racial groups mean age of diagnosis for African-Americans was 48.2 years, 57.4
[5, 6]. In this study, we tested the hypothesis that the observed years for Caucasians, 50.8 years for Chinese, 48.3 years for
survival benefit for Chinese/Asian-American race is attributed Hispanics, and 55.1 years for Other. The mean age of diagnosis
solely to the higher proportion of these patients presenting with for keratinizing squamous cell carcinoma patients was 55.8 years,
the favorable NPC non-keratinizing carcinoma histology by 53.8 years for differentiated non-keratinizing carcinoma patients,
examining NPC cases in the Surveillance, Epidemiology, and 47.9 years for undifferentiated non-keratinizing carcinoma
End Results (SEER) database from 1992 to 2002. patients, and 52.1 years for carcinoma NOS. For the whole
cohort, age group 40–49 years was the most common age group
patients and methods (23.8%) closely followed by age group 50–59 years (23.2%).

study cohort and diagnostic codes sex

We identified 2640 incident NPC cases from 1992 to 2002 using the 2003
In all, 29.1% of the patients were female and there was no
public data of the SEER-13 Program of the National Cancer Institute [13].
statistical difference in the male : female ratio among the four
SEER-13 comprises the states of Connecticut, Iowa, Hawaii, New Mexico, and
different WHO histologic groups in this study (P = 0.37) (Table
Utah and metropolitan areas of Atlanta, GA; Detroit, MI; Los Angeles County,
CA; San Francisco–Oakland, CA; San Jose–Monterey, CA; Seattle–Puget
1). The proportion of female NPC patients increased with
Sound, WA; and two supplemental registries (Alaska native and rural Georgia). increasing age (P = 0.0001).
Tumor site and histology were coded according to criteria specified by
the WHO in International Classification of Diseases for Oncology (ICD-O-3) WHO histology and ethnicity
[14]. All new NPC cases (SEER site code 20060) diagnosed from 1992 to In the United States, NPC patients presented most commonly
2002 were identified. Histologic types of the tumors were grouped with keratinizing squamous cell carcinoma (39.4%) followed by
according to the WHO classification scheme using the ICD-O-3 codes. undifferentiated non-keratinizing carcinoma (25.0%). In all,
Squamous cell carcinoma (ICD-O codes 8070 and 8071) formed the 21.6% of the NPC cases were not classified further and were
keratinizing squamous cell carcinoma group. Large- and small-cell grouped under the carcinoma NOS category for this study
non-keratinizing carcinomas (ICD-O codes 8072 and 8073) formed the purpose. The ethnicities of the patients within each WHO
differentiated non-keratinizing carcinoma group. Undifferentiated, histologic type are shown in Table 1. Chinese comprised the
anaplastic, lymphoepithelial carcinoma (ICD-O codes 8020, 8021, and 8082)
majority of the differentiated non-keratinizing carcinoma
formed the undifferentiated non-keratinizing carcinoma group. Patients
(55.9%), undifferentiated non-keratinizing carcinoma (58.0%),
with ‘carcinoma not otherwise specified’ (ICD-O code 8010) (N = 525)
and carcinoma NOS (61.3%) histologies. On the other hand,
were classified as carcinoma NOS. Patients with prior diagnosis of
Caucasian (55.4%), African-American (44.4%), and Hispanic
malignancies (N = 198) and ‘carcinoma in situ’ (N = 6) were excluded.
(41.3%) NPC patients predominantly presented with
A total of 2436 patients were analyzed in this study.
Staging was grouped into three broad categories: localized, regional, and
keratinizing squamous cell carcinoma.
distant disease according to the SEER summary staging classification [15].
Patient age was categorized in 20-year age group increments from age 0 to 39 stage at presentation
years and in 10-year age group increments from 40 years of age upwards. Stage at diagnosis was available for 86.7% of the cases. Regional
Race was coded as African-American, Caucasian, Chinese, Hispanic, and disease was the most common stage at presentation among the
Other (Native Americans/non-Chinese Asian). WHO histologic group (see Table 1). There were no significant
differences in stage at presentation among the four histologic
statistical analysis categories (P = 0.62).
Comparisons of demographic, clinical, and pathologic variables between
patients with various categories were carried out using Pearson chi-square
radiation treatment
statistic or Fisher’s exact test for nominal variables and Student’s t-test for
continuous variables. Analysis of variance with Tukey’s post hoc test was used Data on radiation treatment were available for 2400 out of 2436
for multiple comparisons of continuous variables. Survival curves were patients (98.5%). Of these 2400 patients, 89% received
constructed using the Kaplan–Meier method and compared with the log-rank radiation: 88.3% of keratinizing squamous cell carcinoma,
test. Multivariate Cox regression analyses were used to determine factors 93.2% of differentiated non-keratinizing carcinoma, 90.9% of
significantly associated with survival. All statistical analyses were conducted undifferentiated non-keratinizing carcinoma, and 85% of
using SAS version 9.1 statistical software (SAS Institute, Cary, NC). Statistical carcinoma NOS (P = 0.0011) (Table 1). In all, 90.2% of
significance was assumed for a two-tailed P value <0.05. localized, 93.6% of regional, and 84.7% of distant NPC received

30 | Ou et al. Volume 18 | No. 1 | January 2007

Annals of Oncology original article
Table 1. Clinicopathologic variables of all NPC patients (n = 2436)

Variable Keratinizing squamous Differentiated non-keratinizing Undifferentiated Carcinoma

cell carcinoma carcinoma non-keratinizing carcinoma NOS
No. % No. % No. % No. %
All cases 959 342 610 525
Age
0–19 6 0.63 2 0.6 40 6.6 13 2.5
20–39 123 12.8 59 17.3 114 18.7 98 18.7
40–49 184 19.2 84 24.6 175 28.7 136 25.9
50–59 251 26.1 63 18.4 133 21.8 119 22.7
60–69 218 22.7 85 24.9 100 16.4 88 16.8
70+ 177 18.5 49 14.3 48 7.9 71 13.5
Sex
Male 684 71.3 237 69.3 446 73.1 361 68.8
Female 275 28.7 105 30.7 164 26.9 164 31.2
Ethnic origin
Caucasian 445 46.4 94 27.5 146 23.9 118 22.5
Chinese 346 36.1 19 55.9 354 58.0 322 61.3
African-American 92 9.6 32 9.4 52 8.5 31 5.9
Hispanic 59 6.2 16 4.7 37 6.1 31 5.9
Other 17 1.8 9 2.6 21 3.4 23 4.4
Stage Out of 823 Out of 315 Out of 541 Out of 432
Localized 114 13.9 52 16.5 85 15.7 57 13.2
Regional 505 61.4 191 60.6 339 62.7 279 64.6
Distant 204 24.8 72 22.9 117 21.6 96 22.2
Radiation treatment
Yes 833 88.3 315 93.2 549 90.9 439 85.2
No 110 11.7 23 6.8 55 9.1 76 14.8

NOS, not otherwise specified; NPC, nasopharyngeal carcinoma.

radiation (P < 0.0001). In all, 86.3% of Caucasian, 85.7% of

African-American, 91.4% of Chinese, 91.3% of Hispanic, and
83.6% of Other received radiation treatment (P = 0.0014).
However, when analyzed within individual WHO histological
type, there was no statistical difference among the various
racial groups who received radiation treatment except within
carcinoma NOS: keratinizing squamous cell carcinoma (P =
0.12), differentiated non-keratinizing carcinoma (P = 0.23),
undifferentiated non-keratinizing carcinoma (P = 0.41), and
carcinoma NOS (P = 0.0007). There was no difference in
the proportion of patients annually who received radiation
over the 10-year study period (P = 0.36).
Figure 1. Overall survival of nasopharyngeal carcinoma patients stratified
by World Health Organization histologies. (A) Keratinizing squamous
overall survival analyses cell carcinoma, (B) differentiated non-keratinizing carcinoma, (C)
undifferentiated non-keratinizing carcinoma, and (D) carcinoma not
WHO histology. Undifferentiated non-keratinizing carcinoma
otherwise specified.
histology had the highest 5-year survival rate (68.1%) followed
by differentiated non-keratinizing carcinoma (57.6%),
carcinoma NOS (55.9%), and keratinizing squamous cell Chinese patients were observed to have the best 5-year survival
carcinoma (45.6%) (P <0.0001) (Figure 1). compared with other ethnic groups within keratinizing
squamous cell carcinoma: Chinese (59.0%), Caucasian (40.1%),
ethnicity. Overall, Chinese had the best 5-year survival rate African-American (35.7%), Hispanic (30.8%), and Other
(63.8%) followed by Caucasian (48.2%), African-American (20.9%) (P <0.0001) (Figure 3). There were no statistical
(45.4%), Hispanic (44.0%), and Other (41%) (P <0.0001) differences in overall survival (OS) across the ethnic groups
(Figure 2). within differentiated non-keratinizing carcinoma: Chinese
When univariate survival analyses were carried out for (64.1%), Caucasian (49.0%), African-American (44.3%),
individual WHO histologic type using ethnicity as a variable, Hispanic (68.1%), and Other (43.8%) (P = 0.09); or

Volume 18 | No. 1 | January 2007 doi:10.1093/annonc/mdl320 | 31

original article Annals of Oncology

undifferentiated non-keratinizing carcinoma: Chinese (70.2%), multivariate survival analyses

Caucasian (69.4%), African-American (74.4%), Hispanic After adjustment for sex, age at diagnosis, stage of presentation,
(50.9%), and Other (52.4%) (P = 0.22). Within carcinoma ethnicity, WHO histology type, and radiation treatment, the
NOS, Chinese (61.9%) had a significant better survival than variables in each category with the strongest survival
Caucasian (51.4%), Hispanic (50.6%), African-American characteristics were as follows: early age of diagnosis, localized
(24.7%), and Other (42.5%) (P = 0.0001). When comparing stage at presentation [versus distant; hazard ratio (HR = 0.35;
only among Chinese NPC patients by the four histologic types, P < 0.0001)], radiation treatment (versus none; HR = 0.48;
OS differences did not achieve statistical significance P < 0.0001), undifferentiated non-keratinizing carcinoma
(P = 0.063). Using Chinese patients with keratinizing squamous (versus keratinizing squamous cell carcinoma; HR = 0.67;
cell carcinoma as a standard, pairwise comparison of Chinese P < 0.0001), and Chinese race (versus Caucasian; HR = 0.78;
NPC patients revealed significant OS difference between P = 0.0010) (Table 2).
undifferentiated non-keratinizing carcinoma (P = 0.0090) or Multivariate Cox proportional hazards models were then
Chinese non-keratinizing carcinoma patients (P = 0.011). carried out for each WHO histologic type. Early age at diagnosis,
localized stage at presentation, and radiation treatment were
stage at presentation. Localized stage at presentation had the best associated with improved survival for each NPC histologic type.
5-year survival estimate when compared with other stages at After adjustment for sex, age at diagnosis, stage at presentation,
presentation (local = 75.7%; regional = 58.2%; distant = 34.9%; and radiation treatment, the Chinese race was the only other
P < 0.0001). significant prognostic factor identified for the keratinizing
squamous cell carcinoma histology when compared with
Caucasian race (HR = 0.67; P = 0.0001). The HR ratios of
radiation treatment. Patients who received radiation had
individual ethnic groups stratified by WHO histologies are
a much better 5-year survival than patients who did not receive
radiation (58.5% versus 31.2%, P < 0.0001).
Table 2. Survival among all NPC cases

sex. There was no difference in the 5-year survival rate between HR 95% HR CI P
female (55.7%) and male (54.9%) patients (P = 0.74).
Ethnic origin
Caucasian 1.000
Chinese 0.784 (0.678–0.907) 0.0010
African-American 1.212 (0.978–1.501) 0.0784
Hispanic 1.163 (0.897–1.508) 0.2532
Other 1.480 (1.063–2.061) 0.0201
Histology
Keratinizing squamous 1.000
cell carcinoma
Differentiated non-keratinizing 0.752 (0.616–0.918) 0.0051
carcinoma
Undifferentiated non-keratinizing 0.672 (0.564–0.801) <0.0001
carcinoma
Carcinoma NOS 0.920 (0.780–1.084) 0.3171
Stage at presentation
Figure 2. Overall survival of nasopharyngeal carcinoma patients Distant 1.000
stratified by ethnicity. (A) Chinese, (B) Caucasian, (C) African-American, Regional 0.697 (0.610–0.796) <0.0001
(D) Hispanic, and (E) Other. Local 0.348 (0.275–0.441) <0.0001
Age of diagnosis
70+ 1.000
60–69 0.605 (0.507–0.723) <0.0001
50–59 0.423 (0.351–0.510) <0.0001
40–49 0.340 (0.279–0.414) <0.0001
20–39 0.238 (0.188–0.301) <0.0001
0–19 0.136 (0.075–0.246) <0.0001
Sex
Male 1.000
Female 1.055 (0.921–1.209) 0.4396
Radiation treatment
No 1.000
Yes 0.479 (0.407–0.565) <0.0001

Figure 3. Overall survival of keratinizing squamous cell carcinoma patients Adjusted analysis using Cox proportional hazards model.
stratified by ethnicity. (A) Chinese, (B) Caucasian, (C) African-American, CI, confidence interval; HR, hazard ratio; NOS, not otherwise specified;
(D) Hispanic, and (E) Other. NPC, nasopharyngeal carcinoma.

32 | Ou et al. Volume 18 | No. 1 | January 2007

Annals of Oncology original article
Table 3. Race HR estimates with 95% CIs stratified by WHO histologic type for NPC patients diagnosed from 1992 to 2002 using Cox proportional
hazards model

Race Keratinizing squamous Differentiated non-keratinizing Undifferentiated non-keratinizing Carcinoma

cell carcinoma carcinoma carcinoma NOS
HR 95% CI P HR 95% CI P HR 95% CI P HR 95% CI P
Caucasian 1.00 1.00 1.00 1.00
Chinese 0.67 (0.54–0.83) 0.0002 0.92 (0.60–1.40) 0.686 1.32 (0.90–1.93) 0.155 0.84 (0.60–1.18) 0.314
African-American 1.04 (0.78–1.40) 0.779 1.58 (0.87–2.86) 0.132 1.61 (0.89–2.91) 0.113 1.41 (0.84–2.38) 0.198
Hispanic 1.12 (0.78–1.59) 0.539 0.86 (0.35–2.11) 0.735 2.32 (1.22–4.42) 0.010 1.19 (0.64–2.20) 0.586
Other 1.55 (0.87–2.78) 0.141 1.63 (0.61–4.32) 0.328 2.15 (1.10–4.22) 0.026 1.34 (0.71–2.56) 0.370

All models were adjusted for the following prognostic variables: sex, age at diagnosis, stage at presentation, and radiation treatment.
CI, confidence interval; HR, hazard ratio; NOS, not otherwise specified; NPC, nasopharyngeal carcinoma; WHO, World Health Organization.

presented in Table 3. Of note, for the undifferentiated non- been associated with cigarette smoking and heavy alcohol
keratinizing histology, Hispanic and Other had a slight but consumption [18]. Keratinizing squamous cell carcinoma of the
statistically significant poorer survival when compared with nasopharynx is similar in histology to the majority of the
Caucasian and Chinese NPC patients (Table 3). squamous cell carcinoma of the head and neck (SCCHN) region
and may represent distinct genetic alterations between Chinese
and non-Chinese patients.
Overexpression of the epidermal growth factor receptor
discussion
(EGFR) gene has been shown to be a poor prognostic factor in
The present study confirmed a survival advantage of the NPC patients [19, 20] and similarly in patients with SCCHN
undifferentiated non-keratinizing carcinoma over differentiated [21]. Allelic polymorphisms in the EGFR gene that control its
non-keratinizing carcinoma and keratinizing squamous cell transcription and thus expression level have been identified with
carcinoma and that Chinese NPC patients had better OS when Asian/Chinese having a higher proportion of the
compared with other racial groups in the United States as had polymorphism that tend to result in lower EGFR expression [22,
been reported in other studies [5–8]. This study confirmed that 23]. One such polymorphism is a dinucleotide CA repeat within
WHO histology is an independent prognostic factor as reported the intron 1 of the EGFR gene. The number of repeats range
by others [5, 6] and identified Chinese ethnicity as another from 14 to 21 within the general population and showed
independent and favorable prognostic factor for survival by remarkable interethnic variability with Asians/Chinese having
multivariate analyses. Therefore, this observed improved a significant lower proportion of the shorter repeats when
survival of Chinese NPC patients cannot simply be explained by compared with Caucasians and African-Americans. For
the higher proportion of favorable non-keratinizing carcinoma. example, the frequency of the allele containing 16 repeats was
This study is also the first to report statistically significant 43% for Caucasians, 42% for African-Americans, 25% for non-
improved survival of Chinese keratinizing squamous cell Chinese Asians, and only 6% for Chinese [22]. The level of
carcinoma NPC patients compared with keratinizing squamous
transcription of the EGFR gene can differ by five-fold depending
cell carcinoma NPC patients from other ethnicities even after
on the number of the CA repeats, with the longer the repeats, the
adjustment for other prognostic factors. This survival advantage
lower the EGFR gene transcription [24]. The significance of this
of Chinese NPC patients with keratinizing squamous cell
variation of the number of CA repeats within intron 1 of EGFR
carcinoma largely contributed to the Chinese race being an
gene has been shown to result in a statistically significant
independent and favorable prognostic factor in the multivariate
difference in the delay of pelvic recurrence in rectal cancer
analyses. Our study also largely confirmed a report by Corry
et al. [12] which demonstrated no statistically significant patients treated with chemoradiation [25] and disease
survival difference between Asian and non-Asian NPC patients progression in colon cancer patients treated with platinum-
with non-keratinizing carcinoma. based therapy, with the longer repeats having delayed recurrence
What may be the explanation for the improved survival of [26]. This CA repeat polymorphism can also occur among
Chinese NPC patients within the keratinizing squamous cell intraethnic SCCHN patient population [27]. Another
carcinoma histology? EBV plays a pivotal role in the potentially significant EGFR polymorphism is a single-
pathogenesis of non-keratinizing carcinoma, whereas nucleotide polymorphism at ÿ216 position (ÿ216G/T) of the
keratinizing squamous cell carcinoma histological type is the promoter region of EGFR gene which affects the binding affinity
least related to EBV infection among the three different of transcription factor Sp1 to these EGFR promoter. Sp1 binds
histologies [16]. The association of squamous cell carcinoma of with significantly higher affinity to the T allele than the G allele.
the nasopharynx with EBV showed geographical variation with The ÿ216T allele also led to significant higher EGFR expression
the highest proportion of squamous cell carcinoma in endemic both in vitro and in vivo. Of note, the ÿ216T haplotype had
areas to be EBV positive [17]. In non-endemic areas, a significant higher distribution in African-Americans (29.3%)
keratinizing squamous cell carcinoma of the nasopharynx has and Caucasians (34.1%) when compared with Asians (8.7%)

Volume 18 | No. 1 | January 2007 doi:10.1093/annonc/mdl320 | 33

original article Annals of Oncology

[23]. Therefore, molecular analysis of these EGFR gene 2. Lo KW, To KF, Huang DP. Focus on nasopharyngeal carcinoma. Cancer Cell
polymorphisms may help explain this observed survival 2004; 5: 423–428.
advantage of Chinese patients with keratinizing squamous cell 3. Shanmugaratnam K, Sobin LH. The World Health Organization histological
classification of tumours of the upper respiratory tract and ear. A commentary on
carcinoma of the nasopharynx.
the second edition. Cancer 1993; 71: 2689–2697.
There are several limitations in this study of the SEER 4. Barnes L, Eveson JW, Reichart P, Sidransky D. World Health Organization
database. Firstly the use of chemotherapy was not evaluated Classification of Tumours: Pathology and Genetics of Head and Neck Tumours.
because SEER data are generally captured from hospital records, Lyon, France: IARC Press 2005.
and chemotherapy was given outside the hospital setting. In 5. Levine PH, Connelly RR, Easton JM. Demographic patterns for nasopharyngeal
order to capture the use of chemotherapy, the SEER database carcinoma in the United States. Int J Cancer 1980; 26: 741–748.
has to be linked to the Medicare database. The vast majority of 6. Burt RD, Vaughan TL, Mcknight B. Descriptive epidemiology and survival analysis
NPC patients, however, were diagnosed at a much earlier age to of nasopharyngeal carcinoma in the United States. Int J Cancer 1992; 52:
qualify for Medicare (age >65) and thus will be excluded in the 549–556.
SEER–Medicare database. Other limitations of the current study 7. Marks JE, Philips JL, Menck HR. The National Cancer Data Base report on the
relationship of race and national origin to the histology of nasopharyngeal
include the inability of the SEER database to provide tobacco
carcinoma. Cancer 1998; 83: 582–588.
exposure, EBV status, concurrent comorbidities, and 8. Bhattacharyya N. The impact of race on survival in nasopharyngeal carcinoma:
socioeconomic status that could affect the survival of NPC a matched analysis. Am J Otolaryngol 2004; 25: 94–97.
patients. About 20% of the patients did not have their WHO 9. Su CK, Wang CC. Prognostic value of Chinese race in nasopharyngeal cancer. Int
histologies further classified (NOS) but were included in the J Radiat Oncol Biol Phys 2002; 54: 752–758.
analysis in this study. When the carcinoma NOS patients were 10. Reddy SP, Raslan WF, Gooneratne S et al. Prognostic significance of keratinization
excluded from the study, the conclusions of this report in nasopharyngeal carcinoma. Am J Otolaryngol 2005; 16: 103–108.
remained the same. 11. Shi W, Pataki I, MacMillan C et al. Molecular pathology parameters in human
Studies that reported survival differences among the three nasopharyngeal carcinoma. Cancer 2002; 94: 1997–2006.
WHO histology types did not analyze the interactions among 12. Corry J, Fisher R, Rischin D et al. Relapse patterns in WHO 2/3 nasopharyngeal
cancer: is there a difference between ethnic Asian versus non-Asian patients?
ethnicity, WHO histology, and survival [5–7] or did not have
Int J Radiat Oncol Biol Phys 2006; 64: 63–71.
significant numbers of Chinese patients in each of the histologic 13. Anonymous. Surveillance, Epidemiology, and End Results (SEER) Program
groups [8]. Most of the NPC studies from endemic countries are (www.seer.cancer.gov) Public-Use Data (1973–2002). National Cancer
with the undifferentiated non-keratinizing carcinoma histology, Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch,
and thus, comparing the survival among WHO histologies released April 2005, based on the November 2004 submission 2005.
within a homogeneous population in endemic countries yielded 14. Fritz A, Percy C, Jack A et al. International Classification of Disease for Oncology,
limited results. Chan et al. [28] have investigated the survival 3rd edition. Geneva, Switzerland: World Health Organization 2000.
15. National Cancer Institute. Summary Staging Guide, SEER Program. Bethesda,
difference between Chinese NPC patients with keratinizing
MD: National Institute of Health; 1981. NIH Publication No 81.
squamous cell carcinoma and non-keratinizing carcinoma
16. Niedobitek G, Hansmann ML, Herbst H et al. Epstein–Barr virus and carcinomas:
patients in an endemic area. Due to the small number of patients undifferentiated carcinomas but not squamous cell carcinoma of the nasopharynx
with keratinizing squamous cell carcinoma histology in the are regularly associated with the virus. J Pathol 1991; 165: 17–24.
study, no difference in disease-free survival and OS was 17. Nicholls JM, Agathanggelou A, Fung K et al. The association of squamous
observed. We reported that there was a statistically significant cell carcinomas of the nasopharynx with Epstein–Barr virus shows
OS advantage of Chinese patients with undifferentiated geographical variation reminiscent of Burkitt’s lymphoma. J Pathol 1997;
183: 164–168.
non-keratinizing carcinoma (P = 0.0090) or patients with
18. Vaughan TL, Shapiro JA, Burt RD et al. Nasopharyngeal cancer in a low-risk
non-keratinizing carcinoma as a whole over Chinese patients
population: defining risk factors by histological type. Cancer Epidemiol Biomark
with keratinizing squamous cell carcinoma (P = 0.011). This Prev 1996; 5: 587–593.
study contributes to the existing literature on NPC by 19. Ma BBY, Poon TCW, To KF et al. Prognostic significance of tumor angiogenesis,
identifying Chinese ethnicity as an independent prognostic Ki 67 p53 oncoprotein, epidermal growth factor receptor and HER2 receptor
factor in NPC and that this survival advantage is largely due to protein expression in undifferentiated nasopharyngeal carcinoma—a prospective
better survival within the keratinizing squamous cell carcinoma study. Head Neck 2003; 25: 864–872.
histology. The observed survival advantage for Chinese patients 20. Chua DTT, Nicholls JM, Sham JST et al. Prognostic value of epidermal growth
factor receptor expression in patients with advanced stage nasopharyngeal
with keratinizing squamous cell carcinoma may reflect
carcinoma treated with induction chemotherapy and radiotherapy. Int J Radiat
underlying gene and/or environmental exposures. This Oncol Biol Phys 2004; 59: 11–20.
difference warrants further investigations at the molecular level 21. Ang KK, Berkey BA, Tu X et al. Impact of epidermal growth factor receptor
since keratinizing squamous cell carcinoma is the predominant expression on survival and pattern of relapse in patients with advanced head and
histology in SCCHN. Thus future molecular and genetic neck carcinoma. Cancer Res 2002; 62: 7350–7356.
research may yield additional prognostic information on 22. Liu W, Innicenti F, Chen P et al. Interethnic difference in the allelic distribution of
response to chemotherapy and radiation therapy for NPC human epidermal growth factor receptor intron 1 polymorphism. Clin Cancer
Res 2003; 9: 1009–1012.
and SCCHN.
23. Liu W, Innocenti F, Wu MH et al. A functional common polymorphism in
a Sp1 recognition site of the epidermal growth factor receptor gene promoter.
Cancer Res 2005; 65: 46–53.
references
24. Gebhardt F, Burger H, Brandt B. Modulation of EGFR gene transcription by
1. Yu MC, Yuan J-M. Epidemiology of nasopharyngeal carcinoma. Semin Cancer a polymorphic repetitive sequence—a link between genetics and epigenetics.
Biol 2002; 12: 421–429. Int J Biol Markers 2000; 15: 105–110.

34 | Ou et al. Volume 18 | No. 1 | January 2007

Annals of Oncology original article
25. Zang W, Park DJ, Lu B et al. Epidermal growth factor receptor gene 27. Etienne-Grimaldi M-C, Pereira S, Formento J-L et al. Analysis of the
polymorphisms predict pelvic recurrence in patients with rectal cancer treated dinucleotide repeat polymorphism in the epidermal growth factor receptor
with chemoradiation. Clin Cancer Res 2005; 11: 600–605. (EGFR) gene in head and neck cancer patients. Ann Oncol 2005; 16:
26. Zhang W, Stoehlmacher J, Park DJ et al. Gene polymorphisms of epidermal 934–941.
growth factor receptor and its downstream effector, interleukin-8, predict 28. Chan ATC, Teo ML, Lee WY et al. The significance of keratinizing squamous
oxaliplatin efficacy in patients with advanced colorectal cancer. Clin Colorectal cell histology in Chinese patients with nasopharyngeal carcinoma. Clin
Cancer 2005; 5: 124–131. Oncol 1998; 10: 161–164.

Volume 18 | No. 1 | January 2007 doi:10.1093/annonc/mdl320 | 35