Abstract
Malignant lymphomas are the third most common canine neoplasm and the most common tumour of the hematopoetic system in dogs. The biological behaviour of canine malignant lymphomas is difficult to predict on the basis of histomorphology alone. Recent classifications are complicated and have only minor prognostic value. Therefore, new methods are necessary to develop a simple, prognostic classification. In this regard, the role of proliferate activity has received special attention.
The aim of this study was to investigate the prognostic importance of age, sex, clinical stage, histomorphology, immunophenotype and different proliferation markers, especially AgNORs, for the evaluation of canine malignant lymphomas in treated and untreated dogs and to develop a prognostic useful classification.
Peripheral lymph node biopsies from 122 dogs with malignant lymphoma were evaluated microscopically and immunohistochemically. The extirpated lymph nodes from 87 dogs were studied prior to chemotherapy with a standardised protocol (vincristine, cyclophosphamide, prednisolone, doxorubicin and L-asparaginase). The other 35 dogs did not receive chemotherapy.
Age, sex, weight, tumor stage, histomorphological grade (Kiel classification, Working Formulation), immunophenotype (using markers for CD 3 and CD 79a) and cell proliferation (Ki-67, PCNA, mitotic index and AgNORs) were evaluated as indicators of prognosis. The AgNORs were assessed qualitatively and quantitatively using also image analysis. All markers were used on routinely formalin fixed, paraffin embedded tissues.
Survival measurements, e.g. tumour response, disease free survival time and survival time for treated dogs and survival time only for untreated dogs were compared to the prognostic criteria. Correlations were determined by univariate (74 dogs) and multivariate (53 dogs) logistic regression analysis based on the Cox proportional hazard model regression method with a forward step selection for treated dogs only. Untreated dogs were used as a comparison and important prognostic criteria for the total survival time were determined by univariate logistic regression analysis based on the Cox proportional hazard regression method with a forward step selection. All parameters with P-values less than 0.3 in the univariate analysis were included in the multivariate analysis for treated dogs only. A P-value of 0.05 was considered significant.
Age, sex, weight, tumour stage and histomorphological grade (Kiel classification, Working Formulation) had no prognostic significance for the survival time of untreated dogs and for the survival time and disease free survival time of treated dogs with malignant lymphoma. The mitotic index was not significant for the prognosis of survival time of untreated dogs, where as the proliferation marker Ki-67 (P = 0.001) and PCNA (P = 0.001) were only significant for the prognosis of survival for single groups of untreated dogs with malignant lymphoma. The maximal distance between two AgNORs (MAXDIST, P = 0.025) and the maximal AgNOR area (MAXNOR, P = 0.026) had a similar importance. Mean AgNOR number (NORNBC, P = 0.001), mean AgNOR area (MEANAR, P = 0.001), AgNOR location (LOCAT, P = 0.048), number of peripherally located AgNORs (BORDER, P = 0.027) and AgNOR type (P = 0.004) were valuable prognostic markers for the survival time of untreated dogs with malignant lymphoma. However, these results are based on the univariate survival analysis only.
In the multivariate analysis immunophenotype (P = 0.027), mean AgNOR area (MEANAR, P = 0.001), maximal distance between two AgNORs (MAXDIST, P = 0.009), ratio AgNOR area/nucleus (AR_RAT, P = 0.008) and total AgNOR area (S_AREA, P = 0.003) were of prognostic significance for the total survival time of treated dogs with malignant lymphoma. However, the total AgNOR area (S_AREA, P = 0.003) was only significant for the prognosis of survival for single groups of treated dogs with malignant lymphoma. Mitotic index (P = 0.028), PCNA (P = 0.009), Ki-67 (P = 0,087), mean AgNOR number (NORNBC, P < 0,001), maximal AgNOR area (MAXNOR, P = 0.001), AgNOR location (LOCAT, P = 0.019), and AgNOR type (P = 0.008) were only significant for the prognosis of survival for single groups of treated dogs with malignant lymphoma in the univariate analysis. Mean AgNOR number (NORNBC, P = 0.001), mean AgNOR area (MEANAR, P = 0.001), total AgNOR area (S_AREA, P = 0.001) and maximal AgNOR area (MAXNOR, P = 0.01) were significant in the multivariate analysis for the prognosis of disease free survival of treated dogs with malignant lymphoma. The mean AgNOR number (NORNBC, P = 0.001) was the most valuable prognostic marker for the disease free survival time. Mitotic index (P = 0.02) and AgNOR type (P = 0.012) were only significant in the univariate analysis for the prognosis of disease free survival for single groups of treated dogs with malignant lymphoma.
Immunophenotype and AgNORs were the most important prognostic markers for the treatment of canine malignant lymphomas. Longer survival time of treated dogs with malignant lymphoma correlated with B-cell type lymphomas, a larger mean AgNOR area, a shorter distance between two AgNORs, and a smaller ratio AgNOR area/nucleus. Longer disease free survival time correlated with a smaller mean AgNOR number per nucleus.
The immunophenotype, the mean AgNOR number (NORNBC) and the mean AgNOR area (MEANAR) were the most simple to evaluate and prognostically most important of all investigated parameters. The use of these three parameters only allows to separate high risk dogs with a low change of survival from dogs with possible survival times of up to three years, if treated with chemotherapy. The evaluation of additional AgNOR parameters would increase the accuracy of the prognostic classification, however the necessary computer equipment might present a too high of a cost factor for many small animal veterinary practices. In addition to a quantitative AgNOR analysis, AgNOR pattern distribution is a useful and simple diagnostic tool in predicting survival time. The here proposed classification based on AgNOR types had only in the univariate analysis prognostic value for the survival time of untreated dogs and for the disease free survival time and survival time of single groups of treated dogs with malignant lymphoma. In contrast the maximal distance between two AgNORs (MAXDIST) and the ratio AgNOR area/nucleus (AR_RAT) were important prognostic factors for the disease free survival time of treated dogs with malignant lymphoma. AgNOR pattern distribution should be further investigated with special reference to these two parameters to develop a classification for the prognostic evaluation of canine malignant lymphomas. At this point of time the here proposed AgNOR type classification combined with the evaluation of the immunophenotype and the mean AgNOR number is the fastest, most accurate and easiest to afford method for the prognostic evaluation of canine malignant lymphomas.
The value of prognostic markers becomes clear with the use of chemotherapy due to longer survival times of treated dogs compared to untreated dogs. Survival times of approximately 2 months only for untreated dogs with malignant lymphoma, early euthanasia after diagnosis and a small total number of cases prevented a multivariate survival analysis and permitted only indicative results with the univariate survival analysis.
The results of this work proof the possibility and importance of a prognostic evaluation of canine malignant lymphomas. An important issue is the use of all markers for a prognostic analysis on routinely formalin fixed, paraffin embedded tissues. Such material may be collected and processed or submitted for further evaluation in any small animal veterinary practice without significant additional costs and time efforts. |
