Page 89 - GTM-2-4

Global Translational Medicine Prognosis of relapse in squamous cell lung cancer

measured parameters is observed exclusively in patients The sensitivity of the combined model for a TV of 0.195
who experience tumor recurrence within a year. In was 100%, the NPV was 100%, and the overall accuracy
patients without relapse, a slight increase in the level of was 96.5%. The AUC, at 0.928, characterizes the model as
one or another measured indicator occurred in some highly effective. The findings imply that if the TV surpasses
cases, but it did not exceed the TV. Other studies have 0.195, then in 85.7% of cases, the patient has a genuinely
reported an increase in CYFRA 21-1 concentrations in high probability of tumor recurrence, while a TV < 0.195
the serum of patients with NSCLC who experienced will correctly predict the recurrence in 100% of patients.
tumor recurrence within 1 year . Moreover, according to It is essential to highlight that the idea of combining
[17]
another study, among five serum tumor markers (CYFRA individual criteria into a unified complex to improve their
21-1, CEA, neuron-specific enolase [NSE], carbohydrate prognostic value is not new. This approach is frequently
antigen [CA] 125, and CA 19-9), only CYFRA 21-1 employed in the laboratory diagnosis of tumors and, in
emerged as the most sensitive marker for predicting certain cases, serves as an important improving tool [23,29-33] .
response to chemotherapy. An increase in its level after an For example, in predicting relapse for patients with lung
initial decrease correlated with a high likelihood of tumor AC post-resection, CYFRA 21-1 and human epididymal
relapse . However, the basis for this conclusion relied protein, HE4 exhibited a sensitivity of 80% and a specificity
[24]
solely on the results of ROC analysis without calculating of 57.1% for each marker. However, when both markers
the TV and indicators of prognostic accuracy. The study were collectively considered, the test specificity increased
presented only the area under the curve as an argument, to 69.7% . In this study, the combined determination
[23]
which can indicate the quality of the model without the of SCC antigen levels, along with the percentage of
practical applicability. lymphocytes with the CXCR2 receptor and monocytes
The utilization of SCC antigen levels in blood serum with the CD44v6 receptor in the blood cell populations,
has been proposed for monitoring treatment efficiency allowed us to achieve a sensitivity of 100% and a specificity
in patients with NSCLC . Elevated concentrations of of 95.6%. In practical terms, if the value of the combined
[25]
SCC antigen post-treatment have been demonstrated model falls below the TV, 95.6% of cases will indeed be free
to be associated with lower overall survival in patients. from relapse.
Furthermore, an increased postoperative SCC antigen It is crucial to acknowledge that the present study was
level has been recommended as a prognostic marker for conducted with a cohort of 57 patients. To validate the
recurrence in various cancers, including oral SCC , proposed model in the future, it is imperative to expand
[26]
cervical SCC , and head-and-neck SCC . the number of subjects and test its validity on samples.
[28]
[27]
In our study, we carried out post-operative monitoring
of SCC antigen levels, along with the percentage of 5. Conclusion
lymphocytes with the CXCR2 receptor and monocytes In this study, we observed that the levels of SCC antigen,
with the CD44v6 receptor in the blood cell populations. the percentage of lymphocytes with the CXCR2 receptor,
This monitoring revealed characteristic dynamics of and the percentage of monocytes with the CD44v6
change at different times after the radical removal of receptor in blood cell populations are highly informative
tumors in Stages I–II of NSCLC. Elevated levels before the for determining the likelihood of post-operative relapse
operation sharply decreased afterward, and a subsequent during the 3 weeks – 6 months after tumor resection.
significant increase in marker concentrations typically The accuracies are 89.5%, 86.0%, and 82.5%, respectively.
indicated the development of relapse. Surprisingly, the By employing a combined model, we could elevate the
prognostic value of these markers, even when considered precision of predicting the probability of relapse based
individually, significantly exceeded the accuracy of on the measurements of the aforementioned indicators
predicting the likelihood of tumor recurrence based on the during the 3 weeks – 6 months post-surgery to 96.5% (with
determination of CYFRA 21-1 and CEA. a sensitivity of 100% and a specificity of 95.6%, using a TV
In the final phase of the study, we achieved a further of 0.195).
enhancement in the prognostic accuracy by incorporating Acknowledgments
all the aforementioned indicators into regression analysis.
As a result, their combination was selected for the regression We appreciate Professor Victor T. Malkevich, the head
model, wherein, devoid of correlation with each other, of the Laboratory of Thoracic Oncopathology at N.N.
they demonstrated the highest predictive capability. This Alexandrov National Cancer Center of Belarus, for his
collective predictive ability significantly exceeded that of kind and fruitful cooperation and discussion of the study
each individual indicator included separately in the model. design.

Volume 2 Issue 4 (2023) 10 https://doi.org/10.36922/gtm.2209