p53 Expression and Cell Proliferation as Prognostic Factors in Laryngeal Squamous Cell Carcinoma By Pasi Hirvikoski, Eero Kumpulainen, Jukka Virtaniemi, Risto Johansson, Hannu Haapasalo, Sinikka Marin, Pirio Halonen, Heikki Helin, Hannu Raitiola, Juhani Pukander, Pirkko Kellokumpu-Lehtinen, and Veli-Matti Kosma Purpose: To investigate the prognostic significance of p53 expression and proliferation markers in primary laryngeal squamous cell carcinoma. Patients and Methods: Primary tumors for analyses were obtained from 103 patients, with complete followup data. All patients were treated between the years 1975 and 1990. The expression of p53 was analyzed with monoclonal D07 antibody and proliferative activity with Ki-67 (MIB-1) and PCNA (monoclonal 19A2) antibodies. Volume corrected mitotic (M/V) index and histological grade were determined in hematoxylin and eosin-stained slides. Results: Sixty-eight percent of the tumors overexpressedp53. During a median follow-up of 62 months, 41 (40%) of patients relapsed. In univariate analysis site of the primary tumor, stage, p53 expression, histologic grade, and M/V index were significant predictors of disease-free survival. In multivariate analysis, only M/V index was a statistically significant predictor of disease-

free survival. Overall survival was significantly better for those overexpressing p53 (10-year cumulative survival rate 68% v44%, P = .004). In multivariate analysis, M/ V index (P = .02), p53 (P = .02), and stage (P = .007) were statistically significant predictors of overall survival. When this analysis includes stratification according to the type of treatment received, M/V index (P = .007), stage (P = .0002), and p53 (P = .006) were even more significant predictors of overall survival. No association between p53 status and proliferative activity was found. Conclusion: Overexpression of p53 is associated with favorable disease-free and overall survival in laryngeal squamous cell carcinoma. It may also have an independent prognostic value in laryngeal cancer. M/V index, p53 overexpression, and stage predict with significant accuracy the 10-year overall survival. J Clin Oncol 15:3111-3120. © 1997 by American Society of Clinical Oncology.

EAD AND NECK squamous cell carcinoma is the Ht sixth most common cancer worldwide.' In the United States, more than 12,000 new laryngeal cancers were diagnosed in 1990.2 Epidemiologic data have strongly linked smoking and alcohol consumption to the development of laryngeal cancer.3 Commonly recognized prognostic factors in laryngeal cancer include exact site of the primary tumor (supraglottis, glottis, and subglottis), stage, histologic grade of the tumor, and sex. 4 These factors influence the therapeutic approach offered to an individual patient. Markers capable of distinguishing patients with a good prognosis from those more likely to relapse are needed. It is now widely accepted that p53 gene alterations are the most common changes occurring during malignant progression of different types of cancer.5 Most mutated forms of p53 proteins with novel oncogenic properties result in a prolonged half-life that is detectable by immunohistochemistry. Loss of normal p53 activity by mutation is a common event in laryngeal neoplasms. 6 However, no correlation to prognosis has been found in earlier studies. 7 9 A positive association between p53 expression and cell proliferation has been shown in several neoplasms. 101 '2 However, contradictionary results have also been presented. 3 "' 4 Nylander et al'5 found no correlation between p53 deregulation and tumor cell proliferation in squamous cell carcinoma of the head and neck. It has been shown that proliferative activity in a number of malignant tumors correlates with progression 8 Proliferative activity and prognosis of the disease. 12.' 6 -1 can be assessed by different methods. In addition to

flow cytometric S-phase fraction and counting mitoses, immunohistochemical markers that are commonly used to evaluate the proliferative activity of tumors include MIB-1 (an antibody against the Ki-67 antigen in formalin-fixed, paraffin-embedded tissues) and proliferating cell nuclear antigen (PCNA). All of these commercially available antibodies are reliable but only when thoroughly tested. 91 2 1 Previous studies6 9 have yielded conflicting results on the prognostic significance of p53 in head and neck cancer. Small patient materials, different immunohistochemical antibodies used, and a short follow-up period may partly explain this controversy. Our aim was to evaluate the prognostic significance of p53 expression and proliferation markers in 103 laryngeal cancer patients, followed up for a median of 62 months. Second, we wanted to

From the Departments of Pathology and ForensicMedicine, Oncology, Otorhinolaryngology, and Computer Centre, University of Kuopio and Kuopio University Hospital, Kuopio; and Departments of Pathology, Otorhinolaryngology, and Oncology, University of Tampere Medical School, Tampere, Finland. Submitted September 24, 1996; accepted June 18, 1997. Supported by Anna and Matti Roininen Memorial Foundation, EVOfunding of the Kuopio University Hospital,and Savon Syparahasto. Address reprint requests to Pasi Hirvikoski, MD, Department of Pathology, University of Kuopio, P.O.B. 1627, FIN-70211 Kuopio, Finland; Email [email protected]. © 1997 by American Society of Clinical Oncology.

0732-183X/97/1509-0021$3.00/0

Journal of Clinical Oncology, Vol 15, No 9 (September), 1997: pp 3111-3120

Downloaded from ascopubs.org by 37.44.207.114 on January 15, 2017 from 037.044.207.114 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3111

3112

HIRVIKOSKI ET AL

clarify the relationship between the expression of p53 and markers of proliferative activity assessed by M/V index and immunohistochemical analysis of MIB-1.

PATIENTS AND METHODS

Histologic Methods

Patients This retrospective study consists of 103 patients diagnosed and treated for laryngeal cancer at Kuopio (n = 85) and Tampere (n = 18) University Hospitals between 1975 and 1990. These patients were selected from the series of 270 laryngeal carcinoma patients identified from the Finnish Cancer Registry, 22 based on the availability of sufficient material from the primary tumor for analyses. We compared this subset of patients with p53 status with the whole population (Table 1). This comparison demonstrated that the age and sex of the patients, site of the primary tumor, and pretherapeutic performance status as classified by the Karnofsky scale 2 3 were almost identical between the two populations. However, the patients with more advanced tumors were more likely to be included into our material. On the basis of clinical characteristics of the patients and analysis of the known prognostic factors, the material is regarded representative of the whole cohort. All patients were followed-up until death or until December 1995. The patients with a previous history of malignancy were excluded from the study. Tumors were classified according to the TNM staging system 2 4 into four stages (I to IV). The TNM classification was based on the previous standard

Table 1. Clinicopathologic Features and Comparison of Patients With and Without p53 Data

Feature

Patients in the Study (%)

Total Population

(n = 103)

IN = 270)

Stage of primary tumor

Tis

0

11 III IV

21 (21%) 19 (18%) 31 (30%) 32 (31%)

89 69 70 34 P