| dc.description.abstracteng | Lung cancer is one of the most frequently occurring malignancies at the present time and still the leading cause of cancer deaths worldwide. Checkpoint pathway inhibitors are a family of cancer immunotherapy that has been shown to improve disease progression-free survival as well as overall survival in patients with advanced NSCLC.
In this retrospective study, we analyzed data of 66 consecutive patients with metastatic NSCLC who received pembrolizumab as a first-line monotherapy. We attempted to validate and improve the reliability of a previously described score of treatment response prediction. The score entails various laboratory values (LDH, CRP and granulocyte/lymphocyte ratio) as well as the change in tumor size derived by CT imaging. Further, we investigated other clinical, radiological and pathological characteristics and correlated them to patients’ response to the treatment with pembrolizumab.
Aiming to examine the possibility of the prediction of patients’ response, we applied the pre-diction score to our cohort of patients after three cycles of the treatment, and as a result, we classified them according to the score into four predicted response groups: stable disease, objective response, pseudoprogressive and real progressive disease. We examined the reliability of this prediction comparing the score results to patients’ real response, which was defined based on the RECIST 1.1 criteria. The prediction score showed a prediction sensitivity of almost 88% with a specificity of about 64%. Sensitivity was defined as the ability of the predictor to sort out patients who are going to benefit from the treatment, i.e., non RPD (= OR, PsPr and SD), from patients with RPD. Furthermore, we validated the score by testing the possibility to predict patients’ survival probability, the predicted one-year survival showed a statistical significance (p= 0.00089) in survival probability between the predicted groups. The one-year survival probability of predicted non-RPD was 64% compared to 69% in reality.
As there is no indisputable test to predict patient response to treatment with immune check point inhibitors (ICIs), a score could be a helpful tool in monitoring patients under treatment with ICIs, especially, as during treatment with ICIs a so-called pseudoprogression is hard to distinguish from a real progressive disease. A prospective validation of the concluded results and the score of prediction in a large cohort of patients would give a more reliable and applicable results. | de |