We hypothesized that the use of G-CSF in nSCLC patients could decrease OS because a possible increase in the neutrophil count would reduce the survival rate by increasing the NLR and the dNLR. However, the results of this study showed higher OS in the group of patients who received G-CSF supportive therapy prior to the treatment with aPD1-i in comparison to that of the group that did not receive G-CSF (HR = 0.63, 95%CI: 0.32-1.25, p (log-rank) = 0.186). This difference in OS was not statistically significant, but it was observed in all the subgroups (Figure 2).
OS prognostic factors for aPD1-i drugs in nSCLC and other cancer can be classified into factors dependent on the patient, tumor, drugs, and environment; the patient-dependent prognostic factors include NLR and dNLR. Patients with NLR ≥5 or dNLR ≥3 have a lower OS (20),(21). Thus, patients with an elevated neutrophil count may have a poor response to aPD1-i drugs. In contrast, patients with NLR <5 or dNLR <3 have better OS. In our population, the prognostic value of NLR and dNLR can be observed in the OS (Table 2). The median OS was higher and statistically significant in the case of patients with a dNLR <3. For patients with a NLR <5, the OS was higher without reaching statistical significance.
In the group of patients who received G-CSF, the median OS was similar in the subgroups with NLR ≥5 and NLR <5 (Table 3). A similar result was achieved in the subgroup of patients with dNLR ≥3, as its median OS was like that in the subgroup with dNLR <3. These results show an improvement in the prognosis of nSCLC when G-CSF is used in a group of patients with a priori poor prognosis, i.e., patients with NLR ≥5 or dNLR ≥3.
The improvement in prognosis with the use of G-CSF was observed in all the subgroups in terms of histology, aPD1-i drug performance status, and time elapsed since the initiation of platinum therapy until aPD1-i drug administration. However, conclusions cannot be drawn because of the difference in the number of cases in each subgroup.
The results of this study are difficult to interpret because previous studies have reported that the use of G-CSF increases the neutrophil count, NLR and dNLR, therefore, causes a worse prognosis (19), (20), (21). The role of neutrophils in antitumor immunity remains unclear. Data indicate that the tumor microenvironment induces a deterioration in antitumor immunity through modulation of PD-L1 expression in tumor-infiltrating neutrophils. It has also been shown in vitro that granulocyte-macrophage colony-stimulating factor (GM-CSF) can increase the expression of PD-L1 in neutrophils of the tumor microenvironment thus suppressing the proliferation and activation of T cells (23).
On the other hand, there are data that show that with the use of G-CSF, the significant increase in absolute numbers of lymphocytes, monocytes, CD3+, CD4+, and CD8+ T cells, and the ratios of CD4/CD8 and dendritic cells in peripheral blood grafts is higher than in bone marrow grafts (24) which causes a faster immune reconstitution in the receptors of blood stem cell transplants; they have higher lymphocyte subset count, which results in fewer infections (25). These data indicate a beneficial effect of G-CSF administration on the antitumor immunity of lymphocytes.
The population included in this study was previously treated for lung cancer and received G-CSF prior to treatment with the aPD1-i drug. The regimens used in the first-line therapy for nSCLC are classified as intermediate risk for febrile neutropenia (incidence 10-20%); therefore, prophylactic use of G-CSF would be considered in the first chemotherapy cycle in case of the existence of associated risk factors for neutropenia and after chemotherapy cycle if dose-limiting neutropenic event or febrile neutropenia (22). In our study, 24.1% of the patients received G-CSF for febrile neutropenia, a value slightly higher than the 20% expected incidence for febrile neutropenia. With the results of incidence of febrile neutropenia and the OS for patients who received G-CSF, obtained in this study, use since the first cycle of chemotherapy in the first line of nSCLC could be considered.
Currently, first-line treatment for nSCLC includes the combination of aPD1-i plus chemotherapy. In this situation, G-CSF could also be used from the first cycle of treatment as prophylaxis for febrile neutropenia.
Therefore, in this study, we elucidate the effectiveness of G-CSF during chemotherapy prior to treatment with aPD1-i and associated with combination regimens of aPD1-i plus chemotherapy for first-line treatment of nSCLC. The use of G-CSF as primary prophylaxis of febrile neutropenia, in these situations, could be considered as high-risk neutropenia chemotherapy regimens. This approach should be validated in larger studies using real world data or prospective studies should be conducted based on this hypothesis.