According to a 2012 estimate, approximately 22,280 new cases of ovarian cancer (OC) are diagnosed and 15,500 women die of the disease in the United States annually [1]. OC is often detected when the disease is already widespread in the abdomen, with approximately 40–50% of all patients being diagnosed with stage III or IV disease [2]. The standard approach to OC treatment is debulking surgery, followed by combined platinum and taxane chemotherapy [3, 4]. Although significant progress in OC treatment has been achieved, approximately 55% of all patients develop recurrence within 2 years, and more than 70% of patients show recurrence within 5 years [2]. Patients with recurrent disease often receive additional second- and third-line chemotherapy regimens. In cases of resistance to platinum-based therapy, second-line single-agent chemotherapy with non-platinum drugs is offered, with a short-lived response rate of approximately 10–25%, regardless of drug types. Combination therapies have been suggested to offer no clinical benefits in these patients [5]. Most patients therefore receive only palliative care, indicating an urgent need for alternative approaches that could improve the survival rates of patients with recurrent ovarian cancer (ROC).

Dendritic cells (DCs) are critical antigen-presenting cells that are characterized by their efficient presentation of internalized antigens with major histocompatibility complexes (MHCs), required to orchestrate T-cell responses [6]. The first DC vaccination study was reported in 1996 [7], and a few clinical trials of DC-based immunotherapy for OC have been conducted [8–10]. However, these trials included patients with all stages of OC and had small sample sizes, thus making it difficult to accurately assess the efficacy of DC immunotherapy in ROC.

In 2009, the cancer antigen prioritization project of the National Cancer Institute ranked Wilms tumor 1 (WT1) as the first antigen, followed by mucin 1, cell-surface associated (MUC1) [11]. The oncogenic WT1 is expressed in various types of hematological and solid malignancies, including OC. The expression frequency of WT1 in OC ranges from 62% to 78% in immunohistochemistry (IHC) studies [12, 13].

MUC1 is a heavily glycosylated membrane glycoprotein with 5 potential O-glycosylation sites in each of its 20 amino acid-long tandem repeats, which comprise most of the extracellular domain [14]. Numerous studies have shown that MUC1 is widely expressed on carcinomas, including those of the breast, colon, rectum, stomach, and lung, as well as OC [15]. Wang et al. [16] reported that MUC1 expression was detected by IHC in 90% of OC cases. Additionally, cancer antigen (CA)-125 has also been detected in tissue sections of human epithelial OC, but not in normal ovarian tissue. Høgdall et al. [17] confirmed that CA125 expression was observed in 70% of all examined OC tissue samples.

Although WT1, MUC1, and CA125 are considered tumor antigens and potential targets for cancer immunotherapy, no reports are available on these antigens in ROC. Therefore, we retrospectively investigated the safety, immunological responses, and clinical effects of DC vaccines targeting synthesized peptides in ROC patients.

Although some clinical trials of DC immunotherapy for OC have been conducted, most of these were designed with small sample sizes (approximately 11–22 patients) and included patients with different stages of OC [8–10]. Differences in biological behavior and individual disease burden were some of the issues encountered when comparing the efficacy of DC-based immunotherapy in heterogeneous patient populations, including patients with recurrent, unresectable, and metastatic disease. The present study was therefore designed to include only ROC patients who had received prior chemotherapy (Additional file 1: Table S2) in order to accurately assess the therapeutic effects and evaluate alternative approaches. Despite its retrospective design and small sample size, the present study yielded 3 major findings in ROC patients receiving the DC vaccine. First, DC-based immunotherapy was well tolerated in all patients, with no serious complications observed. Second, DC administration may be adequate for generating immune responses. Finally, according to multivariate analysis, albumin levels of ≥4.0 g/dL and LDH levels of <200 IU/L before vaccination were significant independent factors for an improved MST.

In this study, the frequency of WT1-specific CTLs increased in 12 (70.6%) of the 17 patients who were examined using the WT1-specific CTL tetramer assay. We found that an increase in the frequency of WT1-specific CTLs was not significantly associated with MST. In addition, an increase in the frequency of WT1-specific CTLs did not affect CD8⁺ T cells. Thus, an increase in antigen-specific CTLs did not appear to directly contribute to the MST. Similarly, although vaccine therapies developed to date have demonstrated immunological responses, only minor clinical benefits have been observed [8, 9, 21].

ROC patients with initially platinum-sensitive disease usually receive platinum-based regimens, even as second-line chemotherapy. Their response rates range from 31% to 66%, with a MST of 17–29 months [22, 23]. In contrast, the MST of platinum-resistant patients is only 10.6 months [5]. In this study, the MST from diagnosis was 30.4 and that from the first vaccination was 14.5 months. DC-based immunotherapy was elected by the patients themselves without our recommendations, leading to inconsistencies in the date of DC-based immunotherapy initiation. Therefore, although our results suggest that DC immunotherapy might prolong survival time, we could not adequately address the differences in survival benefits between our study and previous ones because of the differences in starting points for survival time calculation, patient population, anticancer drugs, and therapy administration schedules. A future prospective study with a larger sample size is thus needed to confirm and validate our findings. The clinical design of such a study should include 4 independent prospective trials to compare chemotherapy platinum-sensitive +/-DC-immunotherapy and platinum-resistant +/-DC-immunotherapy in a controlled ROC patient population. Since DC-based immunotherapy has been reported to show late efficacy [24], the appropriate endpoint should be overall survival, regardless of the study sample size.

The multivariate Cox proportional hazard analysis identified albumin and LDH levels as significant independent factors. Albumin levels are associated with cachexia and ascites, leading to malnutrition in ROC patients. The nutritional status of a cancer patient, which can be assessed by the serum albumin level, is known to be associated with survival. Asher et al. [25] clearly demonstrated that a low serum albumin level was associated with poor survival and that the albumin level could be an independent prognostic predictor of survival in OC patients. In our previous study, low albumin levels were responsible for poor prognosis in patients receiving DC immunotherapy for biliary tract cancers [19], and the same phenomenon was observed in this study on ROC patients. In both studies, an albumin level of ≥4 g/dL seemed to indicate suitability for DC immunotherapy, and the nutritional status was found to be a promising factor for promoting therapeutic effects.

LDH is a major enzyme in glycolysis that reversibly catalyzes the conversion of pyruvate to lactic acid. Boran et al. [26] showed that high serum LDH levels are linked to a poor prognosis in patients with OC. There could be several reasons for the ominous prognostic significance of serum LDH in malignancies. For example, an acidic extracellular pH has been shown to activate gelatinase activity and cathepsin D production, which helps increase the invasiveness of cancer cells [27, 28]. Additionally, the lactate-mediated activation of macrophage-associated angiogenesis might also facilitate metastasis [29]. Finally, a low pH protects the mitochondria from oxidative stress and could account for the increased resistance of cancer cells to hypoxia-induced apoptosis [30]. Further investigation is required to determine whether the abovementioned or any additional mechanisms underlie the strong and independent association of LDH with the prognosis of ROC patients.

Inflammation plays a critical role in the pathogenesis and progression of cancer. Recently, the derived NLR has been shown to influence the clinical outcomes of various cancer types, including ROC [20]. The number of neutrophils might reflect the levels of circulating angiogenesis-regulating chemokines, growth factors, and proteases, which are major contributors to tumor-related angiogenesis [31]. In addition, lymphocytes are involved in the production of cytokines that inhibit the proliferation and metastasis of tumor cells [32]. Therefore, the NLR could act as a marker of the balance between the host inflammatory and immune responses. In our study, patients with a high NLR had a poor prognosis, suggesting that the NLR might be another predictor of the MST in ROC patients.

A number of cancer-associated gene products evoke immune recognition, but host reactions rarely impede disease progression. A strong therapeutic effect has not been confirmed in ROC patients. The weak immunogenicity of nascent tumors contributes to such a failure by the host defense. Our next option to treat ROC patients with DC immunotherapy is the concomitant use of either antagonists of immune-repressor molecules or agonists of immune-activating receptors such as checkpoint blockade receptors comprising cytotoxic T-lymphocyte associated antigen 4 (CTLA-4) (anti-CTLA-4 antibody, ipilimumab; Bristol-Myers-Squibb, NY, USA) and programmed death-1 (PD-1) (anti-PD-L1 monoclonal antibody therapy, BMS-936559). These drugs serve to accelerate antitumor immune responses and improve the therapeutic effect potential in OC [33, 34]. The combination of immunomodulatory properties with DC-based immunotherapy targeting synthesized peptides might help prolong the survival of ROC patients.