The main criterion of effective anticancer treatment is survival after specialized treatment. Surgery is considered as the most effective method in kidney cancer patients. Prediction of survival in patients with renal cell carcinoma is the subject of active researches because it is very important for doctors and especially patients. The necessity of plain prognosis methodology survival of patients is obvious.
The aim of the presented study was to create our own predictive scoring system of cancer-specific survival among patients with renal cancer, who underwent surgery.
The results of this study are based on analyzing of 49 parameters in 343 patients with renal cell carcinoma that were characterized by patients conditions, results of clinical investigations parameters that characterize the immune system functioning.
For creation of a predictive model and selection of the most important factors it has been used Pearson coefficient of linear correlation (r). In cases when |r|≥0,5 we established strong correlation, if |r|=0,3…0,5 – moderate correlation, if |r|=0,1…0,3 – feeble correlation and if |r|≤0,1 – very feeble correlation.
12 parameters have been estimated, which, according to the statistical analysis, inherent in the closest correlation with indicators of cancer-specific survival. Any immune parameter was correlated with kidney cancer patient’s survival.
Simultaneously, morphological and clinical parameter such as tumour size, presence/ absence of capsule, hystologic parameters, tumour classification accordingly to TNM classification, presence/absence of necrosis, Karnofsky scale standardized index correlated with their disease-free duration of life, could be considered as prognostic factors of survival in renal cell carcinoma patients.
All statistically significant parameters were standardized accordingly to their meaning that allows assigning an individual score each of them. At the first stage it a calculation of the absolute value of each of the above parameters was carried out in order to obtain homogeneously positive values. The second phase was conducted by sorting linear correlation coefficients from the smallest to the largest value of considering all options indicators.
At the final stage of calculations prognostic points according to the magnitude of the effect of r were carried, as recommended by Cohen J., 1988. By implementation of Breslow test it has been evaluated that 3-years cancer-specific survival in renal cancer patients directly correlates with tumour stage at the moment of tumour diagnostics (xi2 = 48,4).
With the application of techniques ranging parameters according to Cohen J., 1988 & Field 2009, our own prognostic scale has been created that allows urologists to predict the probability of survival of patients with renal cancer in each case. Developed predictive system makes it possible to stratify renal cell carcinoma patients into subgroups for favorable (0–8 points), doubtful (9–14 points) and unfavorable prognosis (15–24 points).
At present numerous predictive models for kidney cancer patients are being developed which include evaluation of micro-ribonucleic acids (micro-RNAs) and another genes expression. It looks that increasing of expression micro-RNA-100 could be considered as independent predictive factor of cancer-specific survival in kidney cancer cases. Expression of miR-486 also could have predictive significance for survival in patients with kidney cancer who were underwent radical nephrectomies.
Thus, the developed prognostic scale could help urologists to predict survival in patients with kidney cancer. Proposed scoring system may be used in both urologic clinics and ambulances.
Full text: PDF (Ukr) 1.35M