Send to

Choose Destination
See comment in PubMed Commons below
Radiother Oncol. 1992 Feb;23(2):79-93.

Prognostic factors in nasopharyngeal carcinoma investigated by computer tomography--an analysis of 659 patients.

Author information

Department of Clinical Oncology, Prince of Wales Hospital, Shatin, N.T., Hong Kong.


A total of 659 freshly diagnosed nasopharyngeal carcinoma (NPC) (1984-1987), were investigated by computed tomography (CT), treated with locoregional radiotherapy to radical dose, and given neoadjuvant chemotherapy (CHEMO) with 2-3 courses of cisplatinum and 5-fluorouracil for bulky (greater than or equal to 4 cm) cervical nodal metastasis and booster radiotherapy (PPB) for parapharyngeal disease. All except 15 patients were fully evaluable with complete data entry till death or to the last follow-up (minimum 2 years). The data have been analysed extensively to identify variables of potential prognostic significance. The assessed factors include patients' sex and age, nasal involvement (NAS), oropharyngeal involvement (ORO), parapharyngeal involvement (PAR), muscle involvement (MU), skull base involvement (BS), cranial nerves (II-VIII) palsy (CN1), cranial nerves (IX-XII) palsy (CN2), intracranial extension (IC), laryngopharyngeal extension (HYP), confinement to nasopharynx (NP), Ho's N-stage (Nho), maximal nodal size (Nmax), nodal mobility (Nf- fixed, Npf- doubt in mobility, Nm- mobile), nodal laterality (unilateral, contralateral, bilateral), nodal multiplicity (single, multiple), and presentation with distant metastasis (M1). These factors have been assessed as to their interdependence and correlation with the clinical course (study endpoints) using both monovariate analyses and Cox's Regression model. Significant association among Ho's T2 and T3 features was identified. Advanced Ho's N-stage correlated significantly with bulky nodes, multiple nodes, fixed nodes, and, contralateral and bilateral nodes. Poor prognostic factors found to be significant by both monovariate analyses and Cox's Regression model included the M1, Nho (advanced), CN1, BS, and CN2 for the actuarial survival (ASR) for all patients (659), the Nho (advanced), CN1, CN2, and BS for the ASR for the non-metastatic patients (628), the absence of NP and the male sex for the local failure rate (628), the Nho (advanced), CN2, and BS for the distant metastasis rate (628), and the Nho (advanced), CN1, and BS for the disease-free survival (DFS) (628). In addition, old age, male sex, and the presence of parapharyngeal disease were probably significant in predicting poor survival (ASR); CN1 was probably significant in predicting more local failures, and, the parapharyngeal disease and the intracranial extension for more distant metastases. The Ho's N-staging is superior to the other N-stage classifications, because once the Ho's N-stage has been determined, other nodal characteristics including nodal size, multiplicity, laterality, and fixity, are prognostically insignificant.

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Support Center