The Effect of Previous Pregnancies in Patients With Gynecological Neoplasms Treated With PD-1/PD-L1 Inhibitors

In the last decade, the advent of immunotherapy has completely changed the therapeutic landscape for various diseases, leading to significant improvements in survival for patients with different conditions such as non-small cell lung cancer (NSCLC), melanoma, kidney cancer, and triple-negative breast cancer (TNBC). Immunotherapy is an innovative treatment modality that leverages the activation of the immune system, stimulating it to respond against cancer cells. Among the various immunotherapeutic approaches experimented with over the decades, the one that has achieved the best results is based on monoclonal antibodies capable of inhibiting certain immune checkpoints (ICI) to stimulate pathways of "self-tolerance". By inhibiting these inhibitory pathways, i.e., inhibiting the "physiological brake," the immune system is reactivated to stimulate it against cancer cells. Specifically, both CTLA-4 (cytotoxic T-lymphocyte- associated-4) and PD-1 (programmed cell death protein 1) are receptors generally expressed on T lymphocytes that interact respectively with CD80/CD86 and PD-L1/PD-L2. This interaction promotes immune evasion by cancer cells, favoring disease progression.

More recently, immunotherapy has also been explored in gynecological tumors. In first-line cervical cancer, the combination of chemotherapy and Pembrolizumab +/- Bevacizumab has shown benefits in the population with PD-L1 expression ≥ 1%, both in terms of overall survival (median not reached, HR for death 0.64; 95% CI 0.50 to 0.81, P<0.001) and in terms of progression-free survival compared to chemotherapy alone. Immunotherapy has also been studied in second-line treatments. Tewari and colleagues demonstrated the greater efficacy of Cemiplimab (anti-PD- 1) compared to chemotherapy chosen by the investigator in patients with cervical cancer progressing after a platinum-based first-line treatment. The study was positive in terms of overall survival (12.0 vs. 8.5 months; HR for death, 0.69; 95% CI, 0.56 to 0.84; two-sided P<0.001) but not in terms of progression-free survival (2.8 vs 2.9 months; HR for progression or death, 0.75; 95% CI, 0.63 to 0.89; two-sided P<0.001).

Immunotherapy has also been explored in endometrial cancer. In particular, in the recurrent/metastatic setting pretreated with at least one line of platinum-based chemotherapy, the phase 3 KEYNOTE-775 study randomized patients (stratified by microsatellite instability, MSI) to receive Pembrolizumab in combination with Lenvatinib versus investigator's choice chemotherapy. This study demonstrated benefits in both the overall MSS/pMMR population and in MSI-High/dMMR patients, in terms of overall survival (18.3 vs 11.4 months; HR for death 0.62; 95% CI, 0.51 to 0.75; P<0.001 and 17.4 vs. 12.0 months; HR for death, 0.68; 95% CI, 0.56 to 0.84; P<0.001) and in terms of progression-free survival (7.2 vs 3.8 months; HR for progression or death, 0.56; 95% CI, 0.47 to 0.66, P<0.001 and 6.6 vs. 3.8 months; HR for progression or death, 0.60; 95% CI, 0.50 to 0.72; P<0.001). Another immunotherapeutic drug used in the same setting but only in MSI-High/dMMR patients is Dostarlimab. The GARNET study showed a probability of disease progression-free at 6 months (95% CI 4.1-18 months) and overall survival not reached (95% CI 27.1-NR).

Despite the excellent results demonstrated in studies and effectively verified in clinical practice, not all patients can achieve lasting benefits from ICI therapy. For example, in NSCLC, which represents the largest group of patients treated with immunotherapy by numbers, 16% of patients are still alive at 5 years, and about half of the patients treated with immunotherapy did not obtain a real benefit from the treatment. In melanoma, the data, although more encouraging than in other pathologies, show a percentage of long-term survivors at 5 years with anti-PD-1 ICI of about 40%. Given the non-efficacy of immunotherapy in all patients, it is increasingly clinically relevant to find possible predictive factors for the response to this type of treatment. To date, the only predictive factor for response to anti-PD- 1/PD-L1 ICI treatment, albeit with many limitations and very different specificity even among different pathologies, remains the expression of PD-L1.

In the attempt to find factors capable of better guiding us in choosing patients to administer ICIs, it has been observed that the female sex is associated with less benefit compared to the male sex in patients with melanoma and NSCLC. Reviewing the available data in the literature, this reduced efficacy is confirmed only in mono-immunotherapy settings, while when chemo-immuno combinations are administered, women show more benefits than men.

Among the possible differences related to the response to immunotherapy based on sex, pregnancy may play a role. For a pregnancy to be successful, a proper balance between the "invasive" trophoblast and the "receptive" maternal decidua is necessary. Historically, it is believed that during pregnancy, there is a shift from a Th1-mediated response to a Th2-mediated response due to the presence of the placenta producing progesterone and cytokines such as IL-4, aimed at inducing a Th2-mediated response to ensure fetal tolerance according to the host-graft model. More recent studies have highlighted that the immunology in pregnancy cannot be reduced to the host-graft model, but during pregnancy, there are three stages: initially, a pro-inflammatory stage associated with implantation and placentation, then an immunosuppressive stage associated with fetal growth, and finally a new pro-inflammatory state aimed at initiating labor .

To maintain an immunological tolerance state towards paternal antigens during pregnancy, the role of Treg lymphocytes, a subgroup of T lymphocytes with immunosuppressive properties and an indispensable role in the immune response to cancer, is fundamental. Indeed, during normal pregnancies, a higher number of Tregs is observed. The expansion of fetal-specific Treg lymphocytes during gestation and their maintenance postpartum can explain the protective effect of the first pregnancy against subsequent immune-related complications. The necessity that fetal-specific Tregs are essential for a proper pregnancy is demonstrated in a study on murine models where the blockage of Treg differentiation is associated with fetal loss. Indeed, after birth, maternal Tregs with fetal specificity remain at elevated levels for the first 100 days post-pregnancy and then gradually decrease, still maintaining the ability to rapidly re-expand during a second pregnancy upon encountering the same paternal fetal antigen.

During pregnancy, women also show an increase in PD-1 expression on T lymphocytes compared to non-pregnant women, while reduced levels of PD-1 and PD-L1 seem to be associated with the proliferation of the Th17 phenotype with a Treg/Th17 imbalance and pregnancy complications. The immunosuppressive effect of the PD-1 pathway seems to have similar mechanisms in both pregnancy and cancer, although with opposite effects.

Another point of considerable interest in the context of immune system alterations with possible effects on the response to immunotherapy is represented by fetal microchimerism, a process that begins already during pregnancy. Microchimeric cells are rare cells found in an individual that originates from another individual and are genetically distinct from the host, while microchimerism refers to the circulation of fetal cells expressing paternal-fetal antigens that can circulate in the maternal blood for years. It has also been demonstrated that fetal microchimeric cells (FMCs) are not a random "souvenir" of pregnancy but are retained intentionally to favor potential future pregnancies. The progressive increase of fetal cells within different maternal tissues during pregnancy correlates with the systemic expansion of immunosuppressive CD4+ T lymphocytes. Fetal antigen-specific CD4+ T cells preferentially proliferate and differentiate into Tregs, as demonstrated by the increased expression of FOXP3. Therefore, it can be inferred that fetal microchimerism can cause an antigenic reminder that may be crucial for maintaining the immunological memory of pregnancy-induced Tregs. These microchimeric cells have shown protective anticancer functions in meningioma and glioblastoma. In particular, for glioblastoma patients, the presence of FMCs has shown a positive prognostic value, being associated with prolonged survival. Their expression has also been associated with protection against ovarian cancer, considering that women who have had pregnancies at older ages have higher levels of FMCs and are more protected from ovarian cancer compared to women who had pregnancies at younger ages. However, considering the different composition of the tumor microenvironment based on tumor histology, FMCs might play different roles depending on the tumor types. For example, Jørgensen et al. found that the presence of FMCs was associated with a reduced risk of breast cancer but an increased risk of developing colorectal cancer.

A possible explanation for the generation of pregnancy-induced lymphocytic memory is the exposure to paternal-fetal antigens. Initially, naive T cells are exposed to the male antigen through seminal fluid. A subsequent encounter with the antigens occurs during pregnancy through exposure to fetal antigens on the trophoblast and through fetal microchimerism. Finally, in the postpartum period, the maternal immune system remains exposed to fetal antigens through microchimerism.

Following the hypothesis that pregnancy leads to the activation, generation, and expansion of effector and central memory lymphocyte pools existing before pregnancy, and that pregnancy itself causes a persistent increase of these components postpartum, Kieffer et al. designed an observational study to evaluate lymphocyte subpopulations in pregnant women, post-pregnancy women (median of 18 months postpartum), and nulliparous women. This study showed that pregnancy has both short-term and long-term effects on CD4+ and CD8+ lymphocytes in peripheral blood. In the short term, a higher percentage of CD4+ effector memory and activated CD4+ memory lymphocytes is demonstrated in primigravida. As persistent effects, a higher proportion of CD4+ and CD8+ effector memory, CD4+ central memory, and activated CD4+ memory was seen in parous women. These data are consistent with other studies on murine and human models, supporting the hypothesis that memory T cell subtypes are generated during pregnancy and remain altered after childbirth. Both CD4+ central memory and effector memory lymphocytes also expressed CD69 after pregnancy, suggesting persistent activation due to antigen exposure, thus suggesting a possible correlation with microchimerism. Regarding the CD8+ lymphocyte population, changes were observed following pregnancy, specifically a higher population of CD8+ effector memory in peripheral blood in women who have had a pregnancy compared to women who have never had one.

Another element to consider is the existence of antigen sharing between placenta and tumor (e.g., MAGE-A3, MAGE- A4, NY-ESO-1) reported in various studies conducted on murine models and also in patients with solid tumors. Considering that the placenta shows growth and vascularization mechanisms, antigenic determinants, and immune escape mechanisms very similar to a tumor, a role for immunoplacental vaccination as an antitumor immunological therapy has been hypothesized in in vivo models. The fact that the placenta presents several antigens (such as those belonging to the cancer-testis family) in common with solid neoplasms suggests that multiparous women might have memory T lymphocytes capable of recognizing and activating in the presence of specific antigenic determinants, positively influencing the outcome of immunotherapy treatment. Conversely, the persistence in women with previous pregnancy of regulatory T cells specific for placental antigens also expressed by the tumor could suggest that pregnancy induces greater tolerance towards tumor antigens, negatively impacting the outcome of immune checkpoint inhibitors.

Finally, pregnancy could influence tumor development due to exposure to high levels of sex hormones. Although in some experimental models progesterone has demonstrated immunosuppressive activity, increased levels of circulating progesterone in the early stages of pregnancy have shown a protective effect on breast cancer development, especially in women under 50 years, while high levels of estradiol and testosterone during pregnancy are correlated with an increased risk of breast cancer .

In light of all the data reported so far, considering that the immune system is a dynamic but sensitive system to changes, it is of considerable interest to evaluate how long-term changes such as those caused by pregnancy can cause variations in the response to ICI and thus a different prognosis. The objective of this study is to evaluate whether having had one or more pregnancies and the distance from the start of treatment can influence the potential benefit related to treatment with ICI (immunotherapy, combination of chemotherapy with immunotherapy) in patients with gynecological neoplasms.