Fax: (011) 852 2636-4346
Pretherapy quantitative measurement of circulating Epstein–Barr virus DNA is predictive of posttherapy distant failure in patients with early-stage nasopharyngeal carcinoma of undifferentiated type
Version of Record online: 28 MAY 2003
Copyright © 2003 American Cancer Society
Volume 98, Issue 2, pages 288–291, 15 July 2003
How to Cite
Leung, S.-f., Chan, A. T. C., Zee, B., Ma, B., Chan, L. Y. S., Johnson, P. J. and Dennis Lo, Y. M. (2003), Pretherapy quantitative measurement of circulating Epstein–Barr virus DNA is predictive of posttherapy distant failure in patients with early-stage nasopharyngeal carcinoma of undifferentiated type. Cancer, 98: 288–291. doi: 10.1002/cncr.11496
- Issue online: 2 JUL 2003
- Version of Record online: 28 MAY 2003
- Manuscript Accepted: 1 APR 2003
- Manuscript Revised: 31 MAR 2003
- Manuscript Received: 7 FEB 2003
- Hong Kong Research Grants Council (Central Allocation Group Research Projects)
- Institute of Molecular Oncology of the Chinese University of Hong Kong
- Kadoorie Charitable Foundation
- Epstein–Barr virus (EBV) DNA;
- nasopharyngeal carcinoma (NPC);
Patients with International Union Against Cancer (UICC) Stage I–II nasopharyngeal carcinoma (NPC) appear to have a relatively favorable prognosis and generally are excluded from trials of combined modality treatment. More recently, plasma/serum cell-free Epstein-Barr virus (EBV) DNA has been shown to be measurable in the majority of NPC patients at the time of diagnosis, and appears to have prognostic significance. However, within Stage I-II disease, in which failure events are infrequent, the prognostic impact of the pretreatment EBV DNA level has not been addressed to our knowledge. This issue has management implications because different therapeutic strategies currently are employed for patients with good-risk and those with poor-risk NPC.
A cohort of 90 patients with UICC Stage I-II NPC (World Health Organization Grade 2/3 histology) had their pretherapy plasma/serum EBV DNA levels determined by a quantitative polymerase chain reaction assay and correlated with the probability of posttherapy failure. All patients received radiation therapy only, except for three patients who also received concurrent chemotherapy. Kaplan–Meier plots of the probability of locoregional failure, distant failure, and cancer-specific survival were compared with reference to clinical stage and EBV DNA levels.
With a median follow-up time of 45 months, 12 patients and 7 patients, respectively, had developed locoregional and distant failures, including 2 patients with both local and distant failures. Patients with distant failure had significantly higher pretherapy EBV DNA levels than those without failure (a median of 13,219 copies/mL [interquatile-range, 274,635 copies/mL] vs. a median of 423 copies/mL [interquatile-range, 2753 copies/mL]). The probability of distant failure was significantly higher in patients with high (> 4000 copies/mL plasma) compared with low EBV DNA levels (P = 0.0001, log-rank test) and for Stage IIB disease compared with Stage I and Stage IIA disease combined (P = 0.0149, log-rank test), but was not significantly different between patients with Stage II and those with Stage I disease. The risks of locoregional failure were not significantly different between patients with high and those with low EBV DNA levels, and also was not significantly different between clinical substages. Approximately 35% of patients with Stage IIB disease were in the at-risk group for distant failure, as identified by high EBV DNA levels.
Within a group of patients with UICC Stage I-II NPC, the pretherapy plasma EBV DNA level was found to identify a poor-risk group with a probability of distant failure similar to that of patients with advanced stage disease. This group of patients may warrant management considerations currently applicable only to cases of Stage III-IV disease. The prognostic significance of designating Stage IIB disease as per the 1997 UICC staging was confirmed, although the pretherapy EBV DNA level appears to be a more powerful prognostic discriminator in patients with early-stage NPC. Cancer 2003;98:288–91. © 2003 American Cancer Society.
The current standard therapy for nasopharyngeal carcinoma (NPC) is combined chemotherapy and radiation therapy for Stage III and IV disease,1–4 and radiation alone for early-stage disease. Although tumor stage generally is accepted as the most important prognostic factor guiding therapy for NPC, more recent investigations suggest that the level of Epstein–Barr virus DNA (EBV DNA) in the plasma/serum5 or in the peripheral blood cells6 of NPC patients is a useful prognostic factor. In cases of Stage III and IV disease, in which metastatic events are relatively frequent, it has been shown that prognostic subsets could be segregated by plasma EBV DNA levels.5 Failure is infrequent in patients with Stage I-II disease, and to our knowledge the issue of whether treatment failure can be predicted by pretherapy plasma DNA levels in these patients with a relatively good prognosis has not been addressed.
MATERIALS AND METHODS
In this study, we postulated that among patients with International Union Against Cancer (UICC) Stage I–II NPC, there is a subset of patients considered to be at good risk with a nearly 5% recurrence rate,7, 8 and a poor-risk subset with a close to 30% recurrence rate, which is similar to patients with Stage III–IV disease.3, 4, 9, 10 To achieve an 80% power with which to detect such a difference, using a 2-sided 5% level test, approximately 90 patients were required to take part in the study.
Patients and Oncologic Treatment
The pretherapy plasma or serum EBV DNA level was determined for a cohort of 90 patients with Stage I and II (1997 UICC staging system11) NPC. Consecutive consenting patients were recruited for tumor marker determination at an oncology clinic before the commencement of oncologic therapy. All patients had World Health Organization (WHO) Grade 2 or Grade 3 NPC. There were 23 patients with Stage I disease 16 patients with Stage IIA disease, and 51 patients with Stage IIB disease. All patients were staged using a uniform protocol comprised of computed tomography scan of the nasopharynx, base of the skull, and upper neck; nasopharyngoscopy; clinical examination; chest X-ray; and serum alkaline phosphatase level. Positron emission tomography (PET) and other imaging for distant metastases were not performed unless the patient had symptoms suspicious for distant metastases or abnormal baseline investigation findings. No patient in the current study cohort had suspicious symptoms, but two patients had slightly elevated pretherapy serum alkaline phosphatase levels that led to bone scintigraphy and abdominal ultrasound that confirmed the absence of distant metastatic disease. Treatment was comprised of curative-intent radiation therapy to the nasopharynx and neck. The nasopharynx received at least 66 Grays (Gy) by external beam radiation therapy. Patients with extension of disease in the parapharynx (classified as T2b) were given a boost dose of 20 Gy of external radiation. Patients with T1 and T2a disease who were found to have suspicious endoscopic findings in the nasopharynx at the time of completion of external radiation therapy were given a boost dose by intranasopharyngeal brachytherapy delivering 24 Gy in 3 fractions to points 1 cm from the sources (19 patients). Three patients also received concurrent chemotherapy with weekly cisplatin at a dose of 40 mg/m2 within a clinical trial setting in which there were cervical lymph nodes measuring > 4 cm in greatest dimension. The study was approved by the institution's clinical ethics committee and written informed consent was obtained from all subjects.
Plasma EBV DNA Assay
Plasma or serum EBV DNA was measured by a quantitative polymerase chain reaction (PCR) system as described previously.5 Plasma samples were subject to DNA extraction using a QIAamp Blood Kit (Qiagen, Hilden, Germany) using the “blood and body fluid protocol” as recommended by the manufacturer. A total of 400–800 μL of the plasma samples were used for DNA extraction per column. The exact amount was documented for the calculation of the target DNA concentration. A final elution volume of 50 μL was used to elute the DNA from the extraction column. Circulating EBV DNA concentrations were measured using a real-time quantitative PCR system that amplified a DNA segment in the BamHI-W fragment region of the EBV genome. The principles of real-time quantitative PCR and reaction set-up procedures were as described previously.5 Data were collected using an ABI Prism 7700 Sequence Detector and analyzed using the Sequence Detection System software (version 1.6.3) developed by PE Biosystems (Foster City, CA). Results were expressed as copies of EBV genomes per milliliter of plasma. All plasma DNA samples also were subjected to real-time PCR analysis for the β-globin gene, which gave a positive signal on all tested samples, thus demonstrating the quality of the extracted DNA. Multiple negative water blanks were included in every analysis. Results were expressed as copies of DNA per milliliter of plasma.
Clinical Outcome Assessment
Patients were followed at 2–3-month intervals during the first 3 years after therapy. Analysis was performed when the last-recruited patient in the cohort had been followed for 30 months. The occurrence of locoregional and distant failures and cancer-related death was correlated with tumor stage and the pretherapy plasma EBV DNA levels.
At a median follow-up time of 45 months, 12 patients and 7 patients, respectively, had developed locoregional and distant failures, including 2 patients with both local and distant failures. Three patients had died of NPC. Patients with distant failure had significantly higher pretherapy EBV DNA levels than those without distant failure (median of 13,219 copies/mL [interquatile range, 274,635 copies/mL] vs. median of 423 copies/mL [interquatile range, 2753 copies/mL]). Using a cutoff level of 4000 copies/mL for EBV DNA, the risk of distant failure was significantly higher in patients with high EBV DNA levels compared with those with low EBV DNA levels (P = 0.0001, log-rank test) (Fig. 1). The cutoff level of 4000 copies/mL was chosen because it had been shown to predict prognosis in an earlier study of NPC of all stages, using the same measurement system.12 Distant failure also was more probable in patients with Stage IIB disease when compared with those with Stage I and Stage IIA disease combined (P = 0.0149, log-rank test), but was not significantly different between patients with Stage II and Stage I disease (P = 0.0967, log-rank test). There was no significant difference in the probability of locoregional failure between patients with high and low EBV DNA levels (P = 0.719, log-rank test), between those with Stage I and those with Stage II disease (P = 0.1167, log-rank test), or between those with Stage I/IIA disease and patients with Stage IIB disease (P = 0.1495, log-tank test). Repeat analysis was performed on the data set after exclusion of the three patients who had received concurrent chemotherapy and the two patients who had slightly elevated pretherapy alkaline phosphatase levels (as described earlier), and the same statistical conclusions were obtained. The distribution of EBV DNA levels within each clinical substage, in relation to distant failure, is shown in Figure 2. The proportions of patients with high EBV DNA levels within each clinical substage were 1 of 23 patients with Stage I disease, 2 of 16 patients with Stage IIA disease, and 18 of 51 patients with Stage IIB disease. It was noted that all distant failures had occurred in patients with Stage IIB disease. Testing EBV DNA by clinical stage interaction was not found to be significant (P = 0.402), indicating that there is not enough evidence to suggest that the association between EBV DNA and distant failure differs among clinical stages. Patients with high EBV DNA had an actuarial distant failure rate of 31% (95% confidence interval [95% CI], 9–52%) at 45 months. For those patients with both Stage IIB disease and high EBV DNA levels (comprising 35% of the patients with Stage IIB disease), the actuarial distant failure rate was as high as 37% (95% CI, 12–63%) at 45 months of follow-up.
The prognosis of patients with Stage I-II NPC generally is favorable, and in view of the infrequent metastatic events, this group of patients for the most part was excluded from clinical trials of combined modality therapy. The results of the current study demonstrate that within Stage IIB disease, a poor risk group with a distant failure rate comparable to that of advanced stage disease could be identified by their pretherapy EBV DNA levels. Furthermore, the pretherapy EBV DNA level appears to be a more significant prognostic factor than clinical staging in this regard. This finding has two clinical management implications. First, imaging screening for distant metastases could be considered selectively for the group of patients with a high EBV DNA level. Such distant metastatic screening was not performed in those patients in the current study with early-stage disease in view of the lack of evidence of their cost-effectiveness. In fact, data from the pre-PET era suggest that distant metastatic screening in patients with NPC of a nonadvanced stage is not cost-effective.13 Although the value of PET in staging has been demonstrated in several cancer models, to our knowledge its role in the pretherapy staging of NPC has yet to be defined. Second, consideration needs to be given to the recruitment of these patients into clinical trials of treatment intensification with combined modality therapy, such as concurrent1, 4 and/or neoadjuvant2, 3 chemotherapy. In fact, the actuarial risk of distant failure for Stage IIB patients of 37% in the high EBV DNA level group was comparable to that of patients with Stage III-IV disease, in whom a distant failure rate in the range of 20–40% usually was reported.3, 4, 9, 10 These observations also hint at the potential usefulness of a combined clinical-molecular marker staging system for NPC. The data from the current study suggest that the prognosis of Stage II patients with high circulating EBV DNA levels is similar to that for patients Stage III-IV disease, whereas the remaining patients constitute a good-risk group with a minimal risk of distant failure similar to that for patients with Stage I disease. Although the number of cancer-related deaths were few in the current analysis, it is probable that the prognostic impact of the pretherapy EBV DNA level would be reflected in cancer mortality on longer follow-up because the majority if not all distant metastases are expected to translate into death.
The association between the pretherapy EBV DNA level and distant failure but not locoregional failure also is in concordance with the observations in the study by Lin et al., who assessed the prognostic effect of peripheral blood cell EBV DNA in a group of patients with all stages of NPC.6 To explain the disparity in prognostic effect on distant failure and local failure, we hypothesized that quantitative EBV DNA is a marker of tumor burden and that recurrence within the radiation port (i.e., locoregional failure) is partly the result of radioresistance whereas recurrence outside the radiation port (i.e., distant failure) is the result of occult distant metastases already present prior to therapy that essentially were untreated by local radiation therapy. In patients with high EBV DNA levels, it is the occult distant metastases that contributed to the majority of the pretherapy EBV DNA.
The authors would like to acknowledge the contributions to this work from Frankie Mo, B.Sc., Maria Lai, B.Sc., and Wai-yee Lee, B.Sc. from the Department of Clinical Oncology at the Chinese University of Hong Kong.
- 2International Nasopharynx Cancer Study Group. Preliminary results of a randomized trial comparing neoadjuvant chemotherapy (cisplatin, epirubicin, bleomycin) plus radiotherapy vs. radiotherapy alone in stage IV (≥ N2, M0) undifferentiated nasopharyngeal carcinoma: a positive effect on progression-free survival. Int J Radiat Oncol Biol Phys. 1996; 35: 463–469.
- 11FlemingID, CooperJS, HensonDE, editors. American Joint Committee on Cancer manual for staging of cancer, 5th edition. Philadelphia: Lippincott-Raven, 1997.