Immune profile of primary and recurrent epithelial ovarian cancer cases indicates immune suppression, a major cause of progression and relapse of ovarian cancer

The burden of ovarian cancer is increasing globally. A 47% rise in ovarian cancer incidence is predicted by the year 2040. Second highest ovarian cancer incidences are reported in India whereas the mortality rate is highest in the world [1]. Relative frequency of high grade serous epithelial ovarian cancer in India is 87.6% which is also highest in the world [2]. Early diagnosis of the disease is rare due to its asymptomatic nature of growth and delayed onset of symptoms. The 5-year survival rate is approximately 25% for patients with high-grade serous epithelial ovarian cancer [3]. The major complication is the recurrence of the disease and acquired resistance towards standard platinum-based chemotherapeutic agents [4]. The outcome of the disease is heterogeneous even after considering the most common factors such as stage, grade, and response to therapy. This disparity is thought to be driven by both the tumor and the host [5]. Ovarian cancer is considered an immunogenic tumor. The tumor microenvironment is infiltrated by various immune and non-immune cells which is considered to be associated with the progression and metastasis of the disease [6, 7]. However, the role of natural killer (NK) cells in the prognosis of ovarian cancer is not yet clear. Based on the density of CD56 expression NK cells are dissected into CD3-CD56 + ^Bright and CD3-CD56 + ^Dim NK cell subsets. CD3-CD56 +^BrightNK cells are efficient cytokine produces whereas CD3-CD56 + ^Dim NK cells are primarily responsible for cytotoxic activity [8]. Tumor associated NK cells with low level of activating receptors were associated with a poor overall survival of ovarian cancer patients. However, total CD56 + NK cells were associated with better progression free survival and overall survival of these patients [9]. Thus, specific phenotype of NK cells may determine the functional outcome. To combat tumor escape and relapse; NK cells can recognize the loss of HLA molecules for activation, induce inflammation and kill target tumor cells in an antigen-independent manner [10]. NK cells have various mechanisms to kill their target cells such as polarization and release of cytotoxic granules after activation signal; another mechanism is the use of death receptor-induced killing of target cells [11]. NK cells express a repertoire of germline-encoded activating and inhibitory surface receptors. Major families of these receptors are Killer immunoglobulin-like receptors (KIR), Natural cytotoxicity receptors (NCR), and CD94/NKG2 heterodimers that belong to the C-type lectin superfamily. Engagement of these receptors with their cognate ligand and fine balance of threshold of the inhibitory and activating signal determines the functional status of NK cells [12]. It is widely accepted that aberrant expression of these receptors may affect the disease outcome. Furthermore, T cells are the major adaptive immune cells perform response to pathogen, allergen and tumors [13]. T cells have been extensively studied in ovarian cancer [14]. Baseline infiltrations of T cells have played a key prognostic factor for different tumors. The presence of intratumoral CD8 + and CD4 + T cells is independently associated with a good prognosis of the disease. However, the antitumor effect of these immune cells is affected by their number, relative frequency, and functional capabilities [15]. In addition to CD3 + CD56-T cells, CD3 + CD56 + Natural killer T (NKT-like) cells are another important subset of highly differentiated, CD1d independent and MHC unrestricted T cells subset considered as a link between innate and adaptive immune systems [16]. However, the function of NKT- like cells is severely compromised in ovarian cancer patients, this impairment in NKT-like cells function is driven by soluble MHC class-1 related sequences (MICs) by down-regulating the prominent activating receptors NKG2D expressed over the surface of NKT-like cells [17]. Moreover, various lymphocyte subsets work co-operatively, which also decides the patient’s survival [18]. Although the normal expression of NK cell receptors is required for the optimum functioning of these cells, the function of these receptors is largely influenced by the level of cognate ligands as well. A large group of molecules (ligands) expressed over the surface of multiple tumor types are recognized by the activating NK cell receptors [19]. NCR receptors are important for the activation of NK cells and these receptors recognize ligands over malignant cells. NKp30 has two ligands B7H6 and BAG6 restricted over malignant cells, while the ligand for NKp46 is not known [20]. NKG2D is a dominant activating cell surface receptor recognizing the ligands MICA, MICB, and ULBPs, considered as a potential therapeutic target [21]. Ligands for activating receptor DNAM-1, Nectin-2/CD112, and PVR/CD155 are widely distributed on hematopoietic, epithelial, and endothelial cells as well as on several tumors [22]. Understanding the interaction between tumors and the immune system has helped in the initiation of different immune-mediated therapies to improve patient management. Moreover, NK cell based therapy leads to stabilization of disease with mild side effects [23]. Ovarian cancer incidence rate, highest mortality rate, availability of treatment options indicate the importance of bridging studies from India to rationalize alternate treatment with correct prognosis. Hence, we proposed to study the differential immune profile of women with primary and recurrent ovarian cancer. The study focused on the frequency, the phenotype of NK, NKT-like, and T cells, their cognate ligands as well as cytokine profiles, in clinical samples such as blood and tissue to highlight differential immune profiles and their prognostic significance.

This is the first study to the best of our knowledge, to report expression of NK cell receptors, their surface ligands, soluble ligands and cytokine profile in Indian women with primary or recurrent ovarian cancer. Immune escape is one of the most challenging questions in cancer biology and how it affects the development of solid tumors remains unanswered. Antitumor immunity was found to be compromised in solid tumors [26, 27]. We have found immune changes in pEOC patients in terms of increased frequency of circulating CD56^DimNK cells which may indicate proliferation and activation of innate immune cells. A high level of circulatory NK cells in colorectal cancer was associated with better survival and can be used as an independent prognosticator [28]. However, other circulatory subsets such as CD56^BrightNK, NKT-like, and T cells were comparable with HC in both groups of EOC. The modulation of receptor expression could be an immune escape mechanism in invasive pEOC and rEOC. We have found that alteration of these receptors was more drastic on circulatory immune cells. Two novel mechanisms for down regulation of receptors on circulatory immune cells are the release of soluble ligands as “decoy molecules” which suppresses the expression of target receptors and the release of platelet derived TGF-β which down-regulates the expression of DNAM-1 and CD96 [29]. Present analysis on these soluble ligands revealed, high soluble MICA level in both groups of cases. Increased soluble MICA may lead to consistent decrease of NKG2D + immune cells. Modulation of surface receptor expression has been reported in several other malignancies [30, 31]. Down-regulation of activating receptors NKp30, NKp46, and NKG2D is an immune evasion mechanism which led to low cytolytic activity [27]. In our data, both subsets of NK cells with activating phenotype were reduced which might have lead to immune suppression. Moreover, we also noticed the change in the expression pattern of inhibitory receptors on NK cells in both groups of EOC patients. Circulatory immune cell subsets such as NKp46 + CD56^BrightNK, NKp30 + CD56^DimNK, NKp46 + CD56^DimNK, NKp44 + NKT-like, NKG2A + NKT-like, and NKG2A + T cells were reduced in both groups of EOC patients. This probably indicates immune dysregulation and common receptor expression pattern at primary and recurrent stage of the disease. However, certain subsets such as NKp30 + CD56^BrightNK, NKp44 + T and NKp46 + T cells were reduced specifically in pEOC. In contrast, the frequencies of NKp44 + CD56^DimNK cells were increased, whereas CD161 + CD56^BrightNK, NKG2A + CD56^DimNK, CD161 + NKT-like, and NKp30 + T cells were reduced in rEOC, depicting the immunological differences among these two groups of patients. Clinically relevant anti-metastatic role of NK cells was reported in several solid tumors [32, 33]. Thus, reduction in circulatory subsets with activating phenotype could have facilitated the breach of local tumor microenvironment by neoplastic cells, reach the circulation and colonize at distant sites. Furthermore, differential reduction in specific subsets, especially in recurrent cancer may be used to rationalize alternative therapeutic strategies and special care of relapsed cases. Again, these differential immune profiles might be a factor of time. As the disease progresses, the immune signature of pEOC undergoes some changes in circulation that might facilitate the disease relapse. Further, immune infiltration has been neglected for a long time but the tumor infiltrated immune cells can also predict the outcome of ovarian cancer [34]. The tumor infiltrated NK and T cell subsets positive for DNAM-1 are reduced in both groups of patients. Reduction in DNAM-1 receptor especially on NK cells in TME might have led to decrease in NK cell’s ability of synapse formation, a key step to execute the effecter function. Tumor-infiltrating T cells demonstrated a beneficial effect on ovarian cancer patients, especially as a predictive biomarker for the prognosis of ovarian cancer patients suggesting that these cells also play a major role in the outcome of the disease in EOC [23]. Profile of both groups of tumors was equally immunogenic in terms of ligand expression by tumor cells. The expression of ligands by tumor cells can modulate the phenotype of circulatory immune cells [35]. Further, we evaluated the serum cytokine level of both groups; exposure to these cytokines can modulate the phenotype and function of immune cells [36]. IL-6 and oncostatin M have been found to directly stimulate enhanced invasion of cancer cells, stimulate the promotion of cell cycle, enhance resistance to chemotherapy, and cause epithelial-to-mesenchymal transition (EMT) [37]. IL-10 was also associated with ovarian cancer cell migration and the worst disease-free survival of ovarian cancer patients [38]. Furthermore, TNF-α and IL-6 induces the generation of reactive oxygen species, nitrogen species and promote DNA damage which accelerates the initiation of tumorigenesis [39]. Thus, elevated level of IL-6, IL-10 and TNF-α seen in these cases might have led to enhanced invasion, resistance to chemotherapy and disease prognosis. Besides, elevated level of IL-5 and IL-15 in rEOC, it is very interesting to find the elevated level of IL-2 in their sera as IL-2 has promising therapeutic potential, a phase II clinical trial has shown its therapeutic benefits in platinum-resistant ovarian cancer patients [40]. IL-15 also acts as a super agonist that enhances NK cell function against ovarian cancer [41]. However, these beneficial effects of IL-2 and IL-15 are not seen in present study. So, to evaluate the influence of increased cytokine levels on the phenotype of different immune cells, spearman’s correlation analysis was carried out between serum cytokine level and receptor expression profile. Among the elevated cytokines, IL-6 was the only cytokine that was positively correlated with NKp46 + CD56^BrightNK cells. Similar cytokine profiles in serum of both patient groups were rarely correlated with phenotype of immune cells, while the correlation was more common in healthy control which indicates the disrupted axis of cytokine and immune cell interactions in both the group of cases in spite of different state of the disease. Although, study has its limitation as in our data, we did not have ligand and receptor expression profile from healthy tissue. Another limitation for the study is the small sample size in this multiparametric study. However, the study has its own strength, possibly highlighting for the first time a broad immune profile comprising expression of the surface receptors, ligands, soluble ligands and cytokine profiles in both pEOC and rEOC patients in an Indian scenario.

The present study highlights the immunological similarity as well as the differences of immune subsets in pEOC and rEOC. Circulatory NKG2D positive NK, NKT-like, and T cells were significantly reduced in both groups of EOC which is the prominent activating marker. A high level of soluble MICA might have acted as a “decoy” molecule. This could be a probable mechanism of decrease in NKG2D positive subsets in both groups of EOC. Furthermore, the study highlights the elevated level of serum cytokines in EOC patients that may help to understand the cytokines associated with ovarian cancer progression. Profiling of tumor infiltrated immune cells revealed the reduced level of DNAM-1 positive NK and T cells in both groups of EOC. A reduced level of DNAM-1 expression on tumor infiltrated NK cells may led to a possible decrease in NK cell’s ability to synapse formation, required to execute the effecter function. Significantly reduced expression of NKG2D, high level of MICA as well as IL-6, IL10 and TNF-α indicates immune suppression of ovarian cancer patients.. It also emphasized that restoration of cytokines level, NKG2D and DNAM-1on tumor infiltrated immune cells may be targeted to develop specific therapeutic approaches for high-grade serous epithelial ovarian cancer.