Ki-67 Expression in Penile Squamous Cell Carcinoma
Ki-67 Expression in Penile Squamous Cell Carcinoma
The study was approved by East London and The City Research Ethics Committee. We retrospectively retrieved 148 formalin-fixed paraffin-embedded PSCCs from the Cellular Pathology department of St George's Hospital between 2001– and 2007. Ninety-seven samples were of the usual type, 17 basaloid, 15 pure verrucous, 7 mixed verrucous/usual type, 7 mixed verrucous/warty, 2 warty and 3 warty/usual type. They were graded as well (grade 1), moderately (grade 2) and poorly differentiated (grade 3) carcinomas. All cases were reviewed again by an expert uropathologist (CC) including subtyping, grading and staging by standard methodologies. Thirty-five tumours were grade 1, 59 grade 2 and 54 grade 3. Forty-nine samples were stage 1, 70 stage 2, 16 stage 3 and 7 stage 4. There was no tumour stage available for six patients, who underwent penile circumcision or excision biopsy. Lymph node status was as follows: NX (not assessed) 13, N0 –93, N1 –23, N2 – 17 and N3 – 2 patients.
For tissue microarrays 3×1 mm tissue cores were taken from each cancer sample. Tissue sections were cut at 4 μm thickness. A heat-induced antigen retrieval method with citrate buffer pH 6.0 was applied. Immunochemistry was performed using the avidin–biotin technique. Primary antibody, clone MIB-1 was purchased from Dako (Glostrup, Denmark) and used in a concentration of 1:100 for 40 min. The positive control was normal tonsil tissue. The staining pattern was nuclear.
Sections were scored semi-quantitatively by a consultant genitourinary pathologist (DMB) and given an estimated visual score of 0–100%, representing the percentage of positively stained neoplastic nuclei. The core with highest score was chosen for each patient for analysis. Statistical analysis was performed using StatsDirect software, version 2.60.6000. The association between the antibody/lymph node status and the clinical data was evaluated by χ test or Fisher's exact probability test. Comparisons between histological types of SCCs were restricted to usual type, verrucous and basaloid as warty group of tumours was too heterogeneous. For the cut-off point for antibody positivity, median value was chosen (cut-off >40%). In all analyses, p<0.05 was considered to be statistically significant.
One hundred and two PSCC cases were previously genotyped for HPV using INNO-LiPA line probe assay (Innogenetics NV, Ghent, Belgium) and described elsewhere. The association between the high-risk HPV and Ki67 status was evaluated by χ test.
Follow-up data was available for 134 patients. The median follow-up was 24 months. Thirty-one patients had cancer recurrence, 26 died from PSCC and nine from other causes. To identify the best predictors, a Cox proportional hazard model was applied. Variables used for the univariate and multivariate analysis were: Ki-67 immunoexpression (as a continuous variable, %), age at diagnosis, histological grade and stage of the tumour, and tumour nodal status. Backwards stepwise regression using the likelihood ratio test was used to identify variables for the final multivariate model. A p<0.05 was regarded as statistically significant for all analysis. The Ki-67 protein was assessed as a dichotomous variable with a cut-off at 40% (same as for immunohistochemistry). Graphical representations of cumulative risk for death were calculated using Kaplan–Meier techniques. For comparing survival curves we used the log rank test.
Patients and Methods
Patients
The study was approved by East London and The City Research Ethics Committee. We retrospectively retrieved 148 formalin-fixed paraffin-embedded PSCCs from the Cellular Pathology department of St George's Hospital between 2001– and 2007. Ninety-seven samples were of the usual type, 17 basaloid, 15 pure verrucous, 7 mixed verrucous/usual type, 7 mixed verrucous/warty, 2 warty and 3 warty/usual type. They were graded as well (grade 1), moderately (grade 2) and poorly differentiated (grade 3) carcinomas. All cases were reviewed again by an expert uropathologist (CC) including subtyping, grading and staging by standard methodologies. Thirty-five tumours were grade 1, 59 grade 2 and 54 grade 3. Forty-nine samples were stage 1, 70 stage 2, 16 stage 3 and 7 stage 4. There was no tumour stage available for six patients, who underwent penile circumcision or excision biopsy. Lymph node status was as follows: NX (not assessed) 13, N0 –93, N1 –23, N2 – 17 and N3 – 2 patients.
Immunohistochemistry
For tissue microarrays 3×1 mm tissue cores were taken from each cancer sample. Tissue sections were cut at 4 μm thickness. A heat-induced antigen retrieval method with citrate buffer pH 6.0 was applied. Immunochemistry was performed using the avidin–biotin technique. Primary antibody, clone MIB-1 was purchased from Dako (Glostrup, Denmark) and used in a concentration of 1:100 for 40 min. The positive control was normal tonsil tissue. The staining pattern was nuclear.
Sections were scored semi-quantitatively by a consultant genitourinary pathologist (DMB) and given an estimated visual score of 0–100%, representing the percentage of positively stained neoplastic nuclei. The core with highest score was chosen for each patient for analysis. Statistical analysis was performed using StatsDirect software, version 2.60.6000. The association between the antibody/lymph node status and the clinical data was evaluated by χ test or Fisher's exact probability test. Comparisons between histological types of SCCs were restricted to usual type, verrucous and basaloid as warty group of tumours was too heterogeneous. For the cut-off point for antibody positivity, median value was chosen (cut-off >40%). In all analyses, p<0.05 was considered to be statistically significant.
HPV Analysis
One hundred and two PSCC cases were previously genotyped for HPV using INNO-LiPA line probe assay (Innogenetics NV, Ghent, Belgium) and described elsewhere. The association between the high-risk HPV and Ki67 status was evaluated by χ test.
Survival Analysis
Follow-up data was available for 134 patients. The median follow-up was 24 months. Thirty-one patients had cancer recurrence, 26 died from PSCC and nine from other causes. To identify the best predictors, a Cox proportional hazard model was applied. Variables used for the univariate and multivariate analysis were: Ki-67 immunoexpression (as a continuous variable, %), age at diagnosis, histological grade and stage of the tumour, and tumour nodal status. Backwards stepwise regression using the likelihood ratio test was used to identify variables for the final multivariate model. A p<0.05 was regarded as statistically significant for all analysis. The Ki-67 protein was assessed as a dichotomous variable with a cut-off at 40% (same as for immunohistochemistry). Graphical representations of cumulative risk for death were calculated using Kaplan–Meier techniques. For comparing survival curves we used the log rank test.
