Wie unabhängig ist die Proliferationsaktivität kaniner Tumorzellen? Assoziation des mittels digitaler Bildanalytik erhobenen Ki67-Labelling Index ausgewählter Hundetumore mit individuellen Patienten- sowie Tumorparametern

Is the proliferative activity of canine tumour cells independent? Association of the Ki67-labelling index with individual patient characteristics and tumour parameters using digital image analysis

The proliferative activity of tumour cells is an essential component in the histopathological assessment of tumour malignancy in general or the tumour grade of neoplasms in both human and veterinary medicine. As an independent parameter or as part of grading schemes the proliferative activity is correlated with the disease outcome and prognosis of the patient. There is an increased trend towards an individualised tumour therapy, not only in human medicine. For that, highest precision in the assessment of the proliferative activity of tumour cells is necessary. However, individual patient characteristics have not been considered for diagnostics so far. In human and partly also in veterinary medicine, the proliferative activity showed age- or sex-dependencies in healthy tissues, ex vivo cultures and different tumour types. Therefore, the question arises as to why the influence of these individual characteristics have not been taken into account in diagnostics so far. In the case that individual patient- or tumour-related parameters have a decisive influence on the proliferative activity of tumour cells, tumour diagnostics could possibly be optimised and made more precise by correcting the proliferation for age, sex, or other parameters.
Based on the known age and sex dependencies of the proliferative activity in individual human tumour types, we hypothesised an association between patient-related parameters and tumour cell proliferation for canine tumours. Therefore, the aim of this exploratory study was to identify comparable associations between patient-related characteristics and tumour cell proliferation using a selection of common canine tumour types. In addition to patient age and sex, the evaluated patient parameters were extended by the castration status as well as the body weight/size, and tumour characteristics such as size and location were added. The tumour types chosen for this study should represent common canine neoplasms with homogenous morphologies. Based on that, the following tumour types were selected: cutaneous mast cell tumours (MCT), soft tissue sarcomas (STS), adenomas of the hepatoid perianal glands (PGA), plasmacytomas (PLA), trichoblastomas (TRI), interstitial or Leydig cell tumours (ICT), and Sertoli cell tumours (SCT). In this retrospective study a total of 495 tumours from the histological archive of the Institute of Veterinary Pathology at Freie Universität Berlin from 2010 to 2018 met the inclusion criteria. For the identification of proliferating tumour cells the proliferation marker Ki67 was chosen which is frequently used in research and diagnostics. The immunohistochemically stained sections were scanned and whole slide sections were digitally analysed. This enabled reliable and reproducible data of the proliferative activity.
As a core result of this study it was impossible to identify universal associations or correlations between the proliferative activity of tumour cells and the investigated patient- and tumourrelated parameters. Instead, the results turned out to be sporadic and rather tumour type specific. Thus, without considering gender, an association between Ki67-LI and patient age could only be observed for PGA with increasing Ki67-LI with increasing age. Not surprisingly, after stratification of the study population by sex, this increase in Ki67-LI was present only in males, whereas Ki67-LI in females decreased with increasing age. Since PGA occur more frequently in unneutered males and are often associated with hormone-producing testicular tumours, the differences in proliferative activity between the sexes seem rather predictable. In MCT, stratification of the study population by sex led to increasing Ki67-LI with increasing age only in females. However, for this type of tumour, more advanced malignant progression in older animals appears to be a likely cause for this observation. Additionally to age, effects of sex and castration on proliferative activity were investigated. Significant differences in proliferative activity were observed between castrated and intact males with PLA and females with MCT. For MCT, few studies exist claiming that castration plays a role in development and proliferation for that type of tumour, however, the complex hormonal relationships remain unclear so far. Further patient-related parameters investigated in this study were the body weight or size of the dogs but no associations with the Ki67-LI were observed here. Initially, subdivision of the study population by genetic background was sought but proved not to be possible due to small group sizes. This may have prevented the detection of potential associations of Ki67-LI with breed of the patient. Another focus of the study was on the association of tumour characteristics with the Ki67-LI. Here, a significant increase of the Ki67-LI with increasing tumour size was identified in MCT. This observation corresponds to characteristic of this and other malignant tumour types also from humans to undergo malignant progression with accumulation of mutations that can drive cell division. Conversely, Ki67-LI decreased in PGA with increasing tumour size. As a possible cause for this observation increasing hypoxia with increasing tumour size was hypothesized. Finally, possible effects of tumour location on the Ki67-LI were also investigated, which, however, were observed exclusively in TRI. Apart from the observations mentioned, no further associations between patient- or tumour-related parameters and Ki67-LI were found in the selected tumour types.
Based on the results of this explorative study proliferative activity is indeed sporadically associated with patient or tumour characteristics in some tumour types. However, no general conclusions can be drawn from the results, especially not on possible positive effects by taking these characteristics into account when using tumour gradings. Subsequent studies with a prospective design are necessary to evaluate positive effects by adaptation or correction of malignancy assessment in routine diagnostics. Nevertheless, this study has shown that proliferative activity of tumour cells may indeed be associated with patient or tumour characteristics in individual cases. With the background of growing personalised precision medicine, the influence of these variables should not be neglected.