- Email: [email protected]
The prognostic value of histological typing in nasopharyngeal carcinoma
Oral Oncology 48 (2012) 429–433
Contents lists available at SciVerse ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
The prognostic value of histological typing in nasopharyngeal carcinoma Florence Cheung a,⇑, Oscar Chan b, Wai Tong Ng b, Lucy Chan b, Anne Lee b, Siu Wah Pang a a b
Department of Clinical Pathology, Pamela Youde Nethersole Eastern Hospital (PYNEH), Chai Wan, Hong Kong Department of Clinical Oncology, Pamela Youde Nethersole Eastern Hospital (PYNEH), Chai Wan, Hong Kong
a r t i c l e
i n f o
Article history: Received 25 July 2011 Received in revised form 21 November 2011 Accepted 23 November 2011 Available online 16 December 2011 Keywords: Nasopharyngeal carcinoma Non-keratinizing Differentiation Histological typing Prognostic value Excision repair cross complementation group 1 protein expression Clinical outcome
s u m m a r y We analyzed the relation of histological typing in late stage nasopharyngeal carcinoma (NPC) with clinical outcome and excision repair cross complementation group 1 protein (ERCC1) expression. The biopsy specimens of 259 patients with NPC were reviewed by two pathologists for classification according to 2005 WHO subtypes. The patients were of stage III to IVB and treated with radiotherapy (RT) alone or concurrent–adjuvant chemoradiotherapy (CRT). Expression of ERCC1 protein detected by immunohistochemistry on paraffin sections was correlated with the histological subtypes. There were 10 cases (3.9%) of differentiated non-keratinizing carcinoma compared with 249 cases of conventional undifferentiated carcinoma. The former exhibited more advanced squamous differentiation with 3 cases belonging to the papillary variant. The degree of ERCC1 expression was generally high compared with the median of the cohort. Clinically, the differentiated group fared poorly compared with the undifferentiated group with respect to loco-regional failure-free rate, distant failure-free rate, disease-free survival and overall survival (p 6 0.05). Treatment modality of the 10 patients (5 RT, 5 CRT) was similar to the whole cohort. Contrary to general acceptance that differentiation of non-keratinizing NPC had little bearing on prognosis, we demonstrated that in endemic area differentiation in fact conferred a worse prognosis in stage III to IVB patients. There was positive correlation of differentiation with ERCC1 expression. We advocate precise histological typing of NPC in pathology report for prognostic purpose and outcome correlation. Ó 2011 Elsevier Ltd. All rights reserved.
Introduction
Materials and methods
Nasopharyngeal carcinoma (NPC) is a common cancer in southern China and Southeast Asia. Its aetiology in the endemic regions has been clearly shown to be related to Epstein–Barr virus (EBV).1 Hong Kong, being situated in the southern tip of Guangdong, has relatively high incidence of NPC. According to local statistics,2 NPC agestandardized incidence rate has been slowly decreasing over the past 16 years (1991–2007). It fell from 25 to 15 per 100,000 population for male, and 10 to 5 per 100,000 for female. In 2007, NPC still ranked 7th among Hong Kong’s 10 most common cancers. The crude mortality rate stood at 4.9 per 100,000 population. Although this cancer is highly radiosensitive, locoregional recurrence, distant metastasis and morbidity of treatment are problems we face in patient management. Concurrent chemoradiotherapy, entailing significant treatment complications, is the standard treatment nowadays for advanced locoregional disease.3 In order to identify specific tumour prognostic factors in deciding on the optimal treatment modality, we analyzed the histological patterns of archival biopsy material from a selected group of NPC patients. Data on the expression of excision repair cross complementation group 1 protein (ERCC1) were available from a previous study we conducted.4 These results were then correlated with the clinical outcome.
A total of 259 NPC patients were chosen from the period October 1998 to March 2005. They were managed in the clinical oncology department of PYNEH, a regional tertiary hospital in Hong Kong. Only those with adequate archival tissue material for immunohistochemical studies were included. The initial study which had been published4 aimed to test if ERCC1 expression had any predictive and prognostic role in NPC patients treated with radiotherapy with or without concurrent platinum chemotherapy. Since only stage III/ IV patients would be treated with concurrent chemoradiotherapy in that period, the initial study selected only these late stage patients. The current follow-up study was performed on the same group of patients to assess relationship of tumour differentiation and ERCC1 expression with clinical outcome. The patients were treated by radiotherapy (RT, 46%) alone or concurrent-adjuvant chemoradiotherapy (CRT, 54%). The median follow-up time was 6 years. Details of the treatment schemes had been previously described.4
⇑ Corresponding author. Tel.: +852 25955254; fax: +852 25159152. E-mail address: [email protected] (F. Cheung). 1368-8375/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2011.11.017
Histological typing The haematoxylin and eosin stained sections of biopsy material obtained for first diagnosis were retrieved and reviewed by two senior pathologists blind to the corresponding clinical information. The objective was to obtain a consensus typing according to the
430
F. Cheung et al. / Oral Oncology 48 (2012) 429–433
2005 World Health Organization classification5 based on microscopic morphology into: (1) Keratinizing squamous cell carcinoma (8071/3). (2) Non-keratinizing carcinoma (8072/3): differentiated and undifferentiated subtypes. Keratinizing squamous carcinoma is characterized by obvious squamous differentiation with intercellular bridges and keratinization of tumour cells over most areas. Undifferentiated subtype, by far the most common, forms syncytial sheets of large tumour cells with indistinct cell borders, vesicular nuclei and large central nucleoli (Fig. 1a). Differentiated subtype differs by showing more squamous features. There is cellular stratification, pavementing and well-defined cell border. Vague intercellular bridges or occasional keratinized cells are sometimes seen. The nuclei are slightly darker with indistinct nucleoli. Co-existence of undifferentiated and differentiated components does occur. In the present study, the differentiated component had to constitute more than 60% of the tumour tissue to be qualified as differentiated subtype.
castra Laboratories Ltd.) and the peptide nucleic acid (PNA) ISH Detection Kit (DAKO, Glostrup, Denmark). A fluorescein conjugated PNA probe directed against glyceraldehyde 3-phosphate dehydrogenase and a fluorescein labelled random PNA probe supplied with the kit were used as positive and negative control, respectively. The detection was performed according to the manufacturer’s instruction. In brief, 4 lm thick paraffin sections were cut and mounted onto TESPA (3-aminopropyltriethoxysilane) coated slides. The sections were deparaffinized, rehydrated and air-dried. The dewaxed slides were digested with 5 mmol/L proteinase K for 30 min at room temperature. The digestion was stopped by immersing the sections in 95% ethanol and air-dried. The sections were hybridized with fluorescein-conjugated DNA probe for EBV EBER mRNA overnight at 37 °C. The fluorescein-conjugated DNA:RNA hybrids were detected with alkaline phosphatase conjugated anti-fluorescein antibody (Dako, Glostrup, Denmark). The colour reaction was performed with nitroblue tetrazolium salt and 5-bromo-4-chloro-3-indolyl phosphate solution in the dark for 30 min. The sections were counterstained with 0.1% malachite green.
Immunohistochemistry of ERCC1 and microscopic interpretation In-situ hybridization (ISH) for EBV encoded early RNA (EBER) ISH for EBER had been done in some of the tumours during initial diagnosis. In the present review, it was carried out in the target group of differentiated carcinoma for confirmation of its relation to EBV.6 The test was performed using the fluorescein-conjugated DNA probe for EBV EBER mRNA (EBV Probe; Novo-
Paraffin embedded biopsy material from the 259 patients were retrieved and immunohistochemical studies were performed on 4 lm thick sections with mouse monoclonal antibody specific against the full length human ERCC1 protein (clone 8F1, Neomarkers, Fremont, CA). The exact methodology has been explained in our previous report4 and will not be repeated here. ERCC1
Figure 1 Microscopic morphology of nasopharyngeal non-keratinizing carcinoma is shown here with original magnification in brackets. (a) Undifferentiated subtype (200) showing syncytial sheets of tumour cells with indistinct cell border infiltrated by lymphocytes. (b) Differentiated subtype (200) with well defined tumour border indicated by black arrows and distinct cell border indicated by white arrow. (c) Tumour of patient 6 demonstrates focal keratinization indicated by white arrows. (d) Papillary variant (40). Inset (200) highlights markedly pleomorphic tumour cells around vascular core.
431
F. Cheung et al. / Oral Oncology 48 (2012) 429–433
expression of each section was assessed based on staining intensity and the proportion of tumour cells stained. The staining intensity of positively stained nuclei was graded on a scale from 0 to 3. The percentage of positively stained cells was estimated for each section and a proportion score assigned (0 if 0–9%, 50 if 10–49%, and 100 if =50%). The multiplication product of the staining intensity and the proportion score gave the final ERCC1 score. ERCC1 score higher than the median of the cohort was considered ‘‘high’’, equal or below was considered ‘‘low’’. Statistical analysis The locoregional failure-free rate (LRFFR), distant failure-free rate (DFFR), failure- free survival (FFS) and overall survival (OS) were estimated by the use of the Kaplan–Meier method. The differences between the major histological subgroups with respect to the above four parameters were compared by the use of the logrank test. Chi square test was used to compare their ERCC1 scores. All statistical tests were two-sided and p < 0.05 was considered statistically significant.
Table 1 Clinical data, histological features, ERCC1 score, treatment modality and clinical outcome of the differentiated group. Patient no.
Sex/age
Morphology
ERCC1 score
Treatment
Outcome
1 2 3 4 5 6 7 8 9 10
M/46 M/66 M/44 M/53 M/43 M/47 M/45 F/68 M/47 F/41
Papillary Papillary Squamous Squamous Papillary Squamous +K Squamous Squamous Papillary +UD Squamous +UD
300 100 300 300 300 200 300 300 300 300
CRT RT CRT RT RT CRT RT RT CRT CRT
DOD DOD Alive DOD DOD DOD DOD Alive Alive Alive
+K, with minor keratinized foci; +UD, with minor undifferentiated component; CRT, chemoradiotherapy; RT, radiotherapy; DOD, died of disease.
(p = 0.05). Multivariate analysis taking into consideration other clinical variables had not been done in this retrospective study. This was due to the small size of the differentiated group (3.9%) compared with the undifferentiated group (96.1%).
Results Among the 259 NPC cases, 249 cases were subtyped as non-keratinizing undifferentiated carcinoma while 10 cases were subtyped as non-keratinizing differentiated carcinoma. No cases belonged to the keratinizing squamous carcinoma group. Within the differentiated group, initial diagnosis of 3 cases was ‘‘undifferentiated carcinoma’’ while the other 7 cases were correctly categorized under ‘‘differentiated carcinoma’’. ISH for EBER was positive in the 10 differentiated carcinomas, confirming their relation to EBV. Histology and immunohistochemistry Within the differentiated group, 7 cases demonstrated squamous morphology (Fig. 1b) with appositional cell growth (pavementing), well-defined tumour border, stratification, distinct cell border, hyperchromatic nuclei and indistinct nucleoli. Occasional keratinization of tumour cells with eosinophilic cytoplasm was seen in patient 6 (Fig. 1c). The remaining 3 cases showed papillary configuration (Fig. 1d) mimicking squamous papilloma with dysplasia. High magnification revealed highly dysplastic tumour cells with complete loss of cell polarity (Fig. 1d inset) thus betraying their carcinoma nature. Minor component of undifferentiated morphology was seen in patients 9 and 10. ERCC1 expression for the whole cohort had a median score of 200. The majority of the undifferentiated group had low scores (69 high, 180 low). In comparison, ERCC1 scores for the differentiated group (Table 1) were mostly ‘‘high’’ (8 high, 2 low) with p < 0.01. Hence there was a positive correlation of squamous differentiation with ERCC1 expression. Clinical correlation Clinical data showed male predominance (Table 1) within the differentiated group (4:1) similar to the cohort of 3.3:1. There was no significant age, disease stage, or treatment difference. They were treated by RT (5 cases) or CRT (5 cases) in a proportion similar to the cohort. The locoregional failure-free rate, distant failure-free rate, failure-free survival and overall survival were presented in Fig. 2. There was statistically significant difference in the distant failure-free rate and overall survival rate between the undifferentiated and differentiated groups (p < 0.05) with the former being a much better performer. The locoregional failure-free rate and disease free survival rate showed the same diversion pattern with difference between the 2 groups approaching statistical significance
Discussion It has been generally accepted by oncologists and pathologists that histological classification of nasopharyngeal non-keratinizing carcinoma offers no clinical or prognostic significance.5,7 Therefore further subclassification is sometimes omitted in the pathology reporting of nasopharyngeal non-keratinizing carcinoma. There have been studies revealing worse prognosis in keratinizing squamous carcinoma of the nasopharynx compared with the non-keratinizing category in non-endemic and endemic areas.8,9 But scanty detailed clinico-pathological correlation has been carried out for the 2 subtypes within the non-keratinizing category in the modern era of sophisticated treatment modalities. We undertook this study on a cohort of 259 stage III to IVB patients and identified a small proportion of differentiated carcinoma (3.9%) compared with the large majority of undifferentiated carcinoma. Histological typing was done based on the major component (>60%) in the biopsy tissue, thus taking into consideration the possibility of combined subtypes. There was no significant age, disease stage, or treatment difference between the undifferentiated and the differentiated groups. However, based on the locoregional failure-free rate, distant failure-free rate, failure-free survival and overall survival, we found that the differentiated group fared worse with respect to locoregional disease control, distant metastasis and patient survival in spite of similar treatment modalities. Low ERCC1 expression has been reported to be associated with a better response to cisplatin-based chemotherapy in cancers like non-small-cell lung cancer.10 In our previous study on NPC patients, we could not demonstrate such association.4 However we were able to show that high expression of ERCC1 predicted a 2-fold increase in the risk of locoregional failure. This could be explained by resistance to radiotherapy in the high-score tumours brought about by more effective DNA repair after irradiation damage. A similar study by Sun et al. on NPC patients treated with cisplatin-based concurrent chemoradiotherapy did show that patients with ERCC1-negative tumours had longer disease-free survival and overall survival than ERCC1-positive patients.11 In the present study, we found high expression of ERCC1 in the differentiated group compared with the undifferentiated group. There was definite positive correlation between tumour differentiation, ERCC1 score and poor clinical outcome. Similar observation of poor overall survival associated with ERCC1 high expression and keratinizing or differentiated morphology has been made from a smaller
432
F. Cheung et al. / Oral Oncology 48 (2012) 429–433
Figure 2 Kaplan–Meier estimates of probability of (a) locoregional failure-free rate, (b) distant failure-free rate, (c) failure-free survival, and (d) overall survival of patients with undifferentiated (Undiff) and differentiated (Diff) carcinoma over a 5 years (yrs) period.
series in a non-endemic region.12 However, the proportion of keratinizing (17%) and differentiated (15%) subtypes in that study was significantly different from ours. This could be explained by different carcinogenetic pathways of nasopharyngeal cancer in endemic and non-endemic areas. Further validation of our findings by studies in endemic areas would be valuable. In conclusion, we found that in late stage NPC patients, differentiated subtype had a worse prognosis compared with the majority of undifferentiated subtype. This could be related to ERCC1 expression, pending further supportive evidence. We advocate precise histological typing of NPC in pathology report for prognostic prediction and clinical outcome correlation. Conflict of interest statement None declared.
Acknowledgements The study was approved by the ethics committee of PYNEH in Hong Kong. The authors thank the Lee Family Foundation for financial support to the study. Abstract of this study has been presented in the 5th International Symposium on Nasopharyngeal Carcinoma, Malaysia, 2011.
References 1. Pathmanathan R, Prasad U, Sadler R, Flynn K, Raab-Traub N. Clonal proliferations of cells infected with Epstein–Barr virus in preinvasive lesions related to nasopharyngeal carcinoma. N Engl J Med 1995;333(11):693–8. 2. Hong Kong Cancer Registry, Hospital Authority. Nasopharyngeal cancer. In: Hong Kong Cancer Stat 2007; 2007.
F. Cheung et al. / Oral Oncology 48 (2012) 429–433 3. Baujat B, Audry H, Bourhis J, Chan AT, Onat H, Chua DT, et al. Chemotherapy as an adjunct to radiotherapy in locally advanced nasopharyngeal carcinoma. Cochrane Database Syst Rev 2006;18(4):CD004329. 4. Chan SH, Cheung FM, Ng WT, Choi CW, Cheung KN, Yiu KH, et al. Can the analysis of ERCC1 expression contribute to individualized therapy in nasopharyngeal carcinoma? Int J Radiat Oncol Biol Phys 2011;79(5):1414–20. 5. Chan JKC, Bray F, Mc Carron P, Foo W, Lee AWM, Yip T, et al. Nasopharyngeal carcinoma. In: Barnes EL, Eveson JW, Reichart P, Sdransky D, editors. Pathology and genetics of head and neck tumours. Kleihues P, Sobin LH, editors. World Health Organization Classification of Tumours. Lyon, France: IARC Press; 2005. p. 85–97. 6. Wu TC, Mann RB, Epstein JI, MacMahon E, Lee WA, Charache P, et al. Abundant expression of EBER1 small nuclear RNA in nasopharyngeal carcinoma. A morphologically distinctive target for detection of Epstein–Barr virus in formalinfixed paraffin-embedded carcinoma specimens. Am J Pathol 1991;138(6):1461–9. 7. Saw D, Ho JH, Fong M, Chan CL, Tse CH, Lau WH. Prognosis and histology in stage I nasopharyngeal carcinoma (NPC). Int J Radiat Oncol Biol Phys 1985;11(5):893–8.
433
8. Reddy SP, Raslan WF, Gooneratne S, Kathuria S, Marks JE. Prognostic significance of keratinization in nasopharyngeal carcinoma. Am J Otolaryngol 1995;16(2):103–8. 9. Shanmugaratnam K, Chan SH, de-Thé G, Goh JE, Khor TH, Simons MJ, et al. Histopathology of nasopharyngeal carcinoma: correlations with epidemiology, survival rates and other biological characteristics. Cancer 1979;44(3):1029–44. 10. Olaussen KA, Dunant A, Fouret P, Brambilla E, André F, Haddad V, et al. DNA repair by ERCC1 in non-small-cell lung cancer and cisplatin-based adjuvant chemotherapy. N Engl J Med 2006;355(10):983–91. 11. Sun JM, Ahn MJ, Park MJ, Lee HY, Ahn JS, Lee S, et al. Expression of excision repair cross-complementation group 1 as predictive marker for nasopharyngeal cancer treated with concurrent chemoradiotherapy. Int J Radiat Oncol Biol Phys 2011;80(3):655–60. 12. Lee HW, Hwang YH, Han JH, Choi JH, Kang SY, Jeong SH, et al. High expression of excision repair cross-complementation group 1 protein predicts poor outcome in patients with nasopharyngeal cancer. Oral Oncol 2010;46(3):209–13.
Contents lists available at SciVerse ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
The prognostic value of histological typing in nasopharyngeal carcinoma Florence Cheung a,⇑, Oscar Chan b, Wai Tong Ng b, Lucy Chan b, Anne Lee b, Siu Wah Pang a a b
Department of Clinical Pathology, Pamela Youde Nethersole Eastern Hospital (PYNEH), Chai Wan, Hong Kong Department of Clinical Oncology, Pamela Youde Nethersole Eastern Hospital (PYNEH), Chai Wan, Hong Kong
a r t i c l e
i n f o
Article history: Received 25 July 2011 Received in revised form 21 November 2011 Accepted 23 November 2011 Available online 16 December 2011 Keywords: Nasopharyngeal carcinoma Non-keratinizing Differentiation Histological typing Prognostic value Excision repair cross complementation group 1 protein expression Clinical outcome
s u m m a r y We analyzed the relation of histological typing in late stage nasopharyngeal carcinoma (NPC) with clinical outcome and excision repair cross complementation group 1 protein (ERCC1) expression. The biopsy specimens of 259 patients with NPC were reviewed by two pathologists for classification according to 2005 WHO subtypes. The patients were of stage III to IVB and treated with radiotherapy (RT) alone or concurrent–adjuvant chemoradiotherapy (CRT). Expression of ERCC1 protein detected by immunohistochemistry on paraffin sections was correlated with the histological subtypes. There were 10 cases (3.9%) of differentiated non-keratinizing carcinoma compared with 249 cases of conventional undifferentiated carcinoma. The former exhibited more advanced squamous differentiation with 3 cases belonging to the papillary variant. The degree of ERCC1 expression was generally high compared with the median of the cohort. Clinically, the differentiated group fared poorly compared with the undifferentiated group with respect to loco-regional failure-free rate, distant failure-free rate, disease-free survival and overall survival (p 6 0.05). Treatment modality of the 10 patients (5 RT, 5 CRT) was similar to the whole cohort. Contrary to general acceptance that differentiation of non-keratinizing NPC had little bearing on prognosis, we demonstrated that in endemic area differentiation in fact conferred a worse prognosis in stage III to IVB patients. There was positive correlation of differentiation with ERCC1 expression. We advocate precise histological typing of NPC in pathology report for prognostic purpose and outcome correlation. Ó 2011 Elsevier Ltd. All rights reserved.
Introduction
Materials and methods
Nasopharyngeal carcinoma (NPC) is a common cancer in southern China and Southeast Asia. Its aetiology in the endemic regions has been clearly shown to be related to Epstein–Barr virus (EBV).1 Hong Kong, being situated in the southern tip of Guangdong, has relatively high incidence of NPC. According to local statistics,2 NPC agestandardized incidence rate has been slowly decreasing over the past 16 years (1991–2007). It fell from 25 to 15 per 100,000 population for male, and 10 to 5 per 100,000 for female. In 2007, NPC still ranked 7th among Hong Kong’s 10 most common cancers. The crude mortality rate stood at 4.9 per 100,000 population. Although this cancer is highly radiosensitive, locoregional recurrence, distant metastasis and morbidity of treatment are problems we face in patient management. Concurrent chemoradiotherapy, entailing significant treatment complications, is the standard treatment nowadays for advanced locoregional disease.3 In order to identify specific tumour prognostic factors in deciding on the optimal treatment modality, we analyzed the histological patterns of archival biopsy material from a selected group of NPC patients. Data on the expression of excision repair cross complementation group 1 protein (ERCC1) were available from a previous study we conducted.4 These results were then correlated with the clinical outcome.
A total of 259 NPC patients were chosen from the period October 1998 to March 2005. They were managed in the clinical oncology department of PYNEH, a regional tertiary hospital in Hong Kong. Only those with adequate archival tissue material for immunohistochemical studies were included. The initial study which had been published4 aimed to test if ERCC1 expression had any predictive and prognostic role in NPC patients treated with radiotherapy with or without concurrent platinum chemotherapy. Since only stage III/ IV patients would be treated with concurrent chemoradiotherapy in that period, the initial study selected only these late stage patients. The current follow-up study was performed on the same group of patients to assess relationship of tumour differentiation and ERCC1 expression with clinical outcome. The patients were treated by radiotherapy (RT, 46%) alone or concurrent-adjuvant chemoradiotherapy (CRT, 54%). The median follow-up time was 6 years. Details of the treatment schemes had been previously described.4
⇑ Corresponding author. Tel.: +852 25955254; fax: +852 25159152. E-mail address: [email protected] (F. Cheung). 1368-8375/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2011.11.017
Histological typing The haematoxylin and eosin stained sections of biopsy material obtained for first diagnosis were retrieved and reviewed by two senior pathologists blind to the corresponding clinical information. The objective was to obtain a consensus typing according to the
430
F. Cheung et al. / Oral Oncology 48 (2012) 429–433
2005 World Health Organization classification5 based on microscopic morphology into: (1) Keratinizing squamous cell carcinoma (8071/3). (2) Non-keratinizing carcinoma (8072/3): differentiated and undifferentiated subtypes. Keratinizing squamous carcinoma is characterized by obvious squamous differentiation with intercellular bridges and keratinization of tumour cells over most areas. Undifferentiated subtype, by far the most common, forms syncytial sheets of large tumour cells with indistinct cell borders, vesicular nuclei and large central nucleoli (Fig. 1a). Differentiated subtype differs by showing more squamous features. There is cellular stratification, pavementing and well-defined cell border. Vague intercellular bridges or occasional keratinized cells are sometimes seen. The nuclei are slightly darker with indistinct nucleoli. Co-existence of undifferentiated and differentiated components does occur. In the present study, the differentiated component had to constitute more than 60% of the tumour tissue to be qualified as differentiated subtype.
castra Laboratories Ltd.) and the peptide nucleic acid (PNA) ISH Detection Kit (DAKO, Glostrup, Denmark). A fluorescein conjugated PNA probe directed against glyceraldehyde 3-phosphate dehydrogenase and a fluorescein labelled random PNA probe supplied with the kit were used as positive and negative control, respectively. The detection was performed according to the manufacturer’s instruction. In brief, 4 lm thick paraffin sections were cut and mounted onto TESPA (3-aminopropyltriethoxysilane) coated slides. The sections were deparaffinized, rehydrated and air-dried. The dewaxed slides were digested with 5 mmol/L proteinase K for 30 min at room temperature. The digestion was stopped by immersing the sections in 95% ethanol and air-dried. The sections were hybridized with fluorescein-conjugated DNA probe for EBV EBER mRNA overnight at 37 °C. The fluorescein-conjugated DNA:RNA hybrids were detected with alkaline phosphatase conjugated anti-fluorescein antibody (Dako, Glostrup, Denmark). The colour reaction was performed with nitroblue tetrazolium salt and 5-bromo-4-chloro-3-indolyl phosphate solution in the dark for 30 min. The sections were counterstained with 0.1% malachite green.
Immunohistochemistry of ERCC1 and microscopic interpretation In-situ hybridization (ISH) for EBV encoded early RNA (EBER) ISH for EBER had been done in some of the tumours during initial diagnosis. In the present review, it was carried out in the target group of differentiated carcinoma for confirmation of its relation to EBV.6 The test was performed using the fluorescein-conjugated DNA probe for EBV EBER mRNA (EBV Probe; Novo-
Paraffin embedded biopsy material from the 259 patients were retrieved and immunohistochemical studies were performed on 4 lm thick sections with mouse monoclonal antibody specific against the full length human ERCC1 protein (clone 8F1, Neomarkers, Fremont, CA). The exact methodology has been explained in our previous report4 and will not be repeated here. ERCC1
Figure 1 Microscopic morphology of nasopharyngeal non-keratinizing carcinoma is shown here with original magnification in brackets. (a) Undifferentiated subtype (200) showing syncytial sheets of tumour cells with indistinct cell border infiltrated by lymphocytes. (b) Differentiated subtype (200) with well defined tumour border indicated by black arrows and distinct cell border indicated by white arrow. (c) Tumour of patient 6 demonstrates focal keratinization indicated by white arrows. (d) Papillary variant (40). Inset (200) highlights markedly pleomorphic tumour cells around vascular core.
431
F. Cheung et al. / Oral Oncology 48 (2012) 429–433
expression of each section was assessed based on staining intensity and the proportion of tumour cells stained. The staining intensity of positively stained nuclei was graded on a scale from 0 to 3. The percentage of positively stained cells was estimated for each section and a proportion score assigned (0 if 0–9%, 50 if 10–49%, and 100 if =50%). The multiplication product of the staining intensity and the proportion score gave the final ERCC1 score. ERCC1 score higher than the median of the cohort was considered ‘‘high’’, equal or below was considered ‘‘low’’. Statistical analysis The locoregional failure-free rate (LRFFR), distant failure-free rate (DFFR), failure- free survival (FFS) and overall survival (OS) were estimated by the use of the Kaplan–Meier method. The differences between the major histological subgroups with respect to the above four parameters were compared by the use of the logrank test. Chi square test was used to compare their ERCC1 scores. All statistical tests were two-sided and p < 0.05 was considered statistically significant.
Table 1 Clinical data, histological features, ERCC1 score, treatment modality and clinical outcome of the differentiated group. Patient no.
Sex/age
Morphology
ERCC1 score
Treatment
Outcome
1 2 3 4 5 6 7 8 9 10
M/46 M/66 M/44 M/53 M/43 M/47 M/45 F/68 M/47 F/41
Papillary Papillary Squamous Squamous Papillary Squamous +K Squamous Squamous Papillary +UD Squamous +UD
300 100 300 300 300 200 300 300 300 300
CRT RT CRT RT RT CRT RT RT CRT CRT
DOD DOD Alive DOD DOD DOD DOD Alive Alive Alive
+K, with minor keratinized foci; +UD, with minor undifferentiated component; CRT, chemoradiotherapy; RT, radiotherapy; DOD, died of disease.
(p = 0.05). Multivariate analysis taking into consideration other clinical variables had not been done in this retrospective study. This was due to the small size of the differentiated group (3.9%) compared with the undifferentiated group (96.1%).
Results Among the 259 NPC cases, 249 cases were subtyped as non-keratinizing undifferentiated carcinoma while 10 cases were subtyped as non-keratinizing differentiated carcinoma. No cases belonged to the keratinizing squamous carcinoma group. Within the differentiated group, initial diagnosis of 3 cases was ‘‘undifferentiated carcinoma’’ while the other 7 cases were correctly categorized under ‘‘differentiated carcinoma’’. ISH for EBER was positive in the 10 differentiated carcinomas, confirming their relation to EBV. Histology and immunohistochemistry Within the differentiated group, 7 cases demonstrated squamous morphology (Fig. 1b) with appositional cell growth (pavementing), well-defined tumour border, stratification, distinct cell border, hyperchromatic nuclei and indistinct nucleoli. Occasional keratinization of tumour cells with eosinophilic cytoplasm was seen in patient 6 (Fig. 1c). The remaining 3 cases showed papillary configuration (Fig. 1d) mimicking squamous papilloma with dysplasia. High magnification revealed highly dysplastic tumour cells with complete loss of cell polarity (Fig. 1d inset) thus betraying their carcinoma nature. Minor component of undifferentiated morphology was seen in patients 9 and 10. ERCC1 expression for the whole cohort had a median score of 200. The majority of the undifferentiated group had low scores (69 high, 180 low). In comparison, ERCC1 scores for the differentiated group (Table 1) were mostly ‘‘high’’ (8 high, 2 low) with p < 0.01. Hence there was a positive correlation of squamous differentiation with ERCC1 expression. Clinical correlation Clinical data showed male predominance (Table 1) within the differentiated group (4:1) similar to the cohort of 3.3:1. There was no significant age, disease stage, or treatment difference. They were treated by RT (5 cases) or CRT (5 cases) in a proportion similar to the cohort. The locoregional failure-free rate, distant failure-free rate, failure-free survival and overall survival were presented in Fig. 2. There was statistically significant difference in the distant failure-free rate and overall survival rate between the undifferentiated and differentiated groups (p < 0.05) with the former being a much better performer. The locoregional failure-free rate and disease free survival rate showed the same diversion pattern with difference between the 2 groups approaching statistical significance
Discussion It has been generally accepted by oncologists and pathologists that histological classification of nasopharyngeal non-keratinizing carcinoma offers no clinical or prognostic significance.5,7 Therefore further subclassification is sometimes omitted in the pathology reporting of nasopharyngeal non-keratinizing carcinoma. There have been studies revealing worse prognosis in keratinizing squamous carcinoma of the nasopharynx compared with the non-keratinizing category in non-endemic and endemic areas.8,9 But scanty detailed clinico-pathological correlation has been carried out for the 2 subtypes within the non-keratinizing category in the modern era of sophisticated treatment modalities. We undertook this study on a cohort of 259 stage III to IVB patients and identified a small proportion of differentiated carcinoma (3.9%) compared with the large majority of undifferentiated carcinoma. Histological typing was done based on the major component (>60%) in the biopsy tissue, thus taking into consideration the possibility of combined subtypes. There was no significant age, disease stage, or treatment difference between the undifferentiated and the differentiated groups. However, based on the locoregional failure-free rate, distant failure-free rate, failure-free survival and overall survival, we found that the differentiated group fared worse with respect to locoregional disease control, distant metastasis and patient survival in spite of similar treatment modalities. Low ERCC1 expression has been reported to be associated with a better response to cisplatin-based chemotherapy in cancers like non-small-cell lung cancer.10 In our previous study on NPC patients, we could not demonstrate such association.4 However we were able to show that high expression of ERCC1 predicted a 2-fold increase in the risk of locoregional failure. This could be explained by resistance to radiotherapy in the high-score tumours brought about by more effective DNA repair after irradiation damage. A similar study by Sun et al. on NPC patients treated with cisplatin-based concurrent chemoradiotherapy did show that patients with ERCC1-negative tumours had longer disease-free survival and overall survival than ERCC1-positive patients.11 In the present study, we found high expression of ERCC1 in the differentiated group compared with the undifferentiated group. There was definite positive correlation between tumour differentiation, ERCC1 score and poor clinical outcome. Similar observation of poor overall survival associated with ERCC1 high expression and keratinizing or differentiated morphology has been made from a smaller
432
F. Cheung et al. / Oral Oncology 48 (2012) 429–433
Figure 2 Kaplan–Meier estimates of probability of (a) locoregional failure-free rate, (b) distant failure-free rate, (c) failure-free survival, and (d) overall survival of patients with undifferentiated (Undiff) and differentiated (Diff) carcinoma over a 5 years (yrs) period.
series in a non-endemic region.12 However, the proportion of keratinizing (17%) and differentiated (15%) subtypes in that study was significantly different from ours. This could be explained by different carcinogenetic pathways of nasopharyngeal cancer in endemic and non-endemic areas. Further validation of our findings by studies in endemic areas would be valuable. In conclusion, we found that in late stage NPC patients, differentiated subtype had a worse prognosis compared with the majority of undifferentiated subtype. This could be related to ERCC1 expression, pending further supportive evidence. We advocate precise histological typing of NPC in pathology report for prognostic prediction and clinical outcome correlation. Conflict of interest statement None declared.
Acknowledgements The study was approved by the ethics committee of PYNEH in Hong Kong. The authors thank the Lee Family Foundation for financial support to the study. Abstract of this study has been presented in the 5th International Symposium on Nasopharyngeal Carcinoma, Malaysia, 2011.
References 1. Pathmanathan R, Prasad U, Sadler R, Flynn K, Raab-Traub N. Clonal proliferations of cells infected with Epstein–Barr virus in preinvasive lesions related to nasopharyngeal carcinoma. N Engl J Med 1995;333(11):693–8. 2. Hong Kong Cancer Registry, Hospital Authority. Nasopharyngeal cancer. In: Hong Kong Cancer Stat 2007; 2007.
F. Cheung et al. / Oral Oncology 48 (2012) 429–433 3. Baujat B, Audry H, Bourhis J, Chan AT, Onat H, Chua DT, et al. Chemotherapy as an adjunct to radiotherapy in locally advanced nasopharyngeal carcinoma. Cochrane Database Syst Rev 2006;18(4):CD004329. 4. Chan SH, Cheung FM, Ng WT, Choi CW, Cheung KN, Yiu KH, et al. Can the analysis of ERCC1 expression contribute to individualized therapy in nasopharyngeal carcinoma? Int J Radiat Oncol Biol Phys 2011;79(5):1414–20. 5. Chan JKC, Bray F, Mc Carron P, Foo W, Lee AWM, Yip T, et al. Nasopharyngeal carcinoma. In: Barnes EL, Eveson JW, Reichart P, Sdransky D, editors. Pathology and genetics of head and neck tumours. Kleihues P, Sobin LH, editors. World Health Organization Classification of Tumours. Lyon, France: IARC Press; 2005. p. 85–97. 6. Wu TC, Mann RB, Epstein JI, MacMahon E, Lee WA, Charache P, et al. Abundant expression of EBER1 small nuclear RNA in nasopharyngeal carcinoma. A morphologically distinctive target for detection of Epstein–Barr virus in formalinfixed paraffin-embedded carcinoma specimens. Am J Pathol 1991;138(6):1461–9. 7. Saw D, Ho JH, Fong M, Chan CL, Tse CH, Lau WH. Prognosis and histology in stage I nasopharyngeal carcinoma (NPC). Int J Radiat Oncol Biol Phys 1985;11(5):893–8.
433
8. Reddy SP, Raslan WF, Gooneratne S, Kathuria S, Marks JE. Prognostic significance of keratinization in nasopharyngeal carcinoma. Am J Otolaryngol 1995;16(2):103–8. 9. Shanmugaratnam K, Chan SH, de-Thé G, Goh JE, Khor TH, Simons MJ, et al. Histopathology of nasopharyngeal carcinoma: correlations with epidemiology, survival rates and other biological characteristics. Cancer 1979;44(3):1029–44. 10. Olaussen KA, Dunant A, Fouret P, Brambilla E, André F, Haddad V, et al. DNA repair by ERCC1 in non-small-cell lung cancer and cisplatin-based adjuvant chemotherapy. N Engl J Med 2006;355(10):983–91. 11. Sun JM, Ahn MJ, Park MJ, Lee HY, Ahn JS, Lee S, et al. Expression of excision repair cross-complementation group 1 as predictive marker for nasopharyngeal cancer treated with concurrent chemoradiotherapy. Int J Radiat Oncol Biol Phys 2011;80(3):655–60. 12. Lee HW, Hwang YH, Han JH, Choi JH, Kang SY, Jeong SH, et al. High expression of excision repair cross-complementation group 1 protein predicts poor outcome in patients with nasopharyngeal cancer. Oral Oncol 2010;46(3):209–13.