Stratification of ovarian tumor pathology by expression of programmed cell death-1 (PD-1) and PD-ligand- 1 (PD-L1) in ovarian cancer

Stratification of ovarian tumor pathology by expression of programmed cell death-1 (PD-1) and... Background: Ovarian cancer is the major cause of death among gynecologic cancers with 75% of patients diagnosed with advanced disease, and only 20% of these patients having a survival duration of five years. Treatments blocking immune checkpoint molecules, programmed cell death (PD-1) or its ligand PD-ligand- I (PD- L1) have produced a beneficial and prolonged effect in a subgroup of these patients. However, there is debate in the literature concerning the prognostic value of the expression of these molecules in tumors, with immunotherapy responsiveness, and survival. We evaluated the immune landscape of the ovarian tumor microenvironment of patients, by measuring the impact of the expression of tumor PD-1, PD-L1 and infiltrating lymphocytes on stage and grade of tumors and survival, in a cohort of 55 patients with gynecologic malignancies. Most patients under study were diagnosed with advanced disease ovarian cancer. Results: Our studies revealed that a low density of PD-1 and of PD-L1 expressing cells in tumor tissue were significantly associated with advanced disease (P = 0.028 and P = 0.033, respectively). Moreover, PD-L1 was expressed significantly more often in high grade tumors (41.5%) than in low grade tumors of patients (7.7%) (P = 0.040). The presence of CD3 or of FoxP3 infiltrating cells with PD-L1 in patient tumors did not impact the significance of the association of PD-L1 with high grade tumors (P = 0.040), and our analyses did not show an association between the presence of PD-1 or PD-L1 and survival. Conclusions: We conclude that a subgroup of advanced disease ovarian cancer patients with high grade tumors, expressing PD-L1, may be prime candidates for immunotherapy targeting PD-1 signaling. Keywords: Programmed cell death-1, Programmed cell death-1 ligand, High grade disease, Cancer immunotherapy, Ovarian cancer * Correspondence: mdrakes@luc.edu Cardinal Bernardin Cancer Center, Oncology Research Institute, Department of Medicine, Loyola University Chicago, Bldg. 112, Room 232, 2160 South First Avenue, Maywood, IL 60153, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 2 of 11 Background Pembrolizumab (Keytruda®) is FDA approved for melan- The early signs of ovarian cancer are asymptomatic and oma and NSCLC, and a blocking anti-PD-L1 antibody Ate- thus approximately 75 % of cases are detected in the ad- zolizumab (Tecentriq®) is also FDA approved for vanced metastatic stages. Conventional management strat- unresectable bladder cancer and for NSCLC. Blockade of egies for advanced disease include cytoreductive surgery this pathway is particularly useful in patients as it is applic- and chemotherapy. Most current treatments are not cura- able to a wide range of cancers, and because it induces tive for patients with advanced disease and hence survival anti-tumor immune responses capable of targeting mutated for this category of patients is low [1]. It is estimated that proteins [20]. Importantly, treatment targeting PD-1 signal- in 2017 there will be 22,440 new cases of ovarian cancer ing has fewer high grade toxicities than other immunother- in the Unites States, and that 14,080 patients will die due apies [13, 21]. to this disease [2]. Approximately 80 % of patients diag- Medical centers are currently utilizing these agents in nosed with late stage ovarian cancer die within five years. ongoing clinical trials for various cancers, including To provide more effective treatment options for patients, ovarian cancer [7, 22, 23]. Initial reports of some trials several clinical trials are ongoing using novel single and show promising objective response rates (ORR) for the combination regimens to improve survival. For cancer ther- treatment of ovarian cancer with anti-PD-1 antibody apy, there have been several distinct landmarks in the de- nivolumab (ORR of 15%, n = 20 patients), and pembroli- velopment of new therapies and FDA approved treatments zumab (ORR 11.5%, n = 49), or an anti-PD-L1 antibody over the last decade [3]. However, even with the current avelumab (ORR 10%, n = 124) [3, 24]. Those responding treatment options a considerable number of patients are often had durable responses, suggesting that if we could not yet receiving adequate therapy for the management of identify the subgroup that might typically respond, we advanced stage ovarian cancer and other malignancies. could advance the therapeutic options in this subgroup The development and optimization of the use of novel of ovarian cancer patients. therapies such as immunotherapy, requires an in-depth un- PD-1 is primarily expressed on CD4+ and CD8+ T cells derstanding of specific target molecules and cellular inter- and is associated with T cell exhaustion [11, 12, 14]. PD- actions in tumors. Early efforts in immunotherapy can be L1 is expressed on many cell types including tumor cells traced to 1891, in which administration of intra-tumoral in- and macrophages, including those with an immunosup- jections of bacteria led to a shrinkage of patients’ tumor [4, pressive phenotype [12, 25, 26]. Ligation of PD-L1 on 5]. Since then, significant progress has been made in the tumor cells with PD-1 on T cells, for example, abrogates T field [6]. One of the recent highlights in novel treatment cell proliferation, diminishes T cell activation and leads to options for cancer has been the targeting of immune check- a predominance of a T helper 2 (Th2) cytokine tumor point inhibitory molecules [7–9]. Immune checkpoints are microenvironment, with a pro-tumor propensity. Anti- critically important in health and disease. They represent body blocking of PD-1 or PD-L1 restores T cell prolifera- co-signaling pathways which are either costimulatory or tive and cytotoxic functions, and induces a T-helper 1 coinhibitory. In the body, linkage of coinhibitory receptor (Th1) phenotype, thereby re-invigorating T cells, with and ligand suppresses T cell receptor signaling, and limits resulting potent anti-tumor capacity [14, 27, 28]. immune responses. Whereas this function of checkpoint in- The immune mechanisms of disease improvement with hibitory molecules is beneficial during resolution of infec- administration of checkpoint inhibitory molecules are not tion, or in the development of self tolerance to prevent well understood. Clinically, there is also ongoing debate autoimmune conditions [10–12], ligation of checkpoint in- over which patients will benefit from this therapy, whether hibitory molecules can be a powerful and unwanted mech- patients who respond initially will continue to show anism of immunosuppression in cancer [13–15]. Since the complete responses (CR) or partial responses (PR), and successful introduction and FDA approved use of an anti- whether patients’ tumors need to express PD-1 and/ or body targeting checkpoint inhibitory molecule cytotoxic T PD-L1 in abundance, to predict beneficial responses to lymphocyte associated-4 (CTLA-4) (Ipilimumab; Yervoy®) checkpoint inhibitory molecule blocking therapy targeting in patients with unresectable or metastatic melanoma in these molecules. At the present time, there are more ques- 2011 [16], this agent is now in use in over 40 countries. At- tions than answers. tention has more recently focused on another checkpoint As a study of the immune microenvironment of ovarian inhibitory molecule programmed cell death-1 (PD-1) and cancer patient tumors offers insight into the baseline im- its ligand programmed cell death-1 ligand (PD-L1) [17–19]. mune landscape associated with patient survival and tumor Antibodies inhibiting PD-1 and PD-L1 have recently been pathology, and implicates broader scope for targeting these FDA approved for the treatment of cancer. For example, molecules in combination studies with conventional therapy the agent nivolumab (Opdivo®) is approved for unresectable and with other novel therapies, we undertook these present or metastatic melanoma, non-small cell lung cancer investigations. We primarily selected advanced disease ovar- (NSCLC), Hodgkin’s lymphoma and renal cell carcinoma. ian cancer patients for study, since this group typically have Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 3 of 11 poor outcome with standard therapy, and our future goal in single patient on each slide for detection of PD-L1, PD-1, translational medicine is to address the need for novel alter- CD3 and CD8 by immunohistochemistry (IHC). For stain- native treatment options in this patient sector. We evaluated ing of FoxP3 on T cells, patient tissue arrays were con- the expression and localization of PD-1 and PD-L1 in a co- structed from the paraffin embedded blocks and adhered hort of ovarian cancer formalin fixed paraffin embedded on a total of 2 glass slides with a core of tissue from each of (FFPE) tumor sections, and investigated whether the relative 27 or 28 patients, as well as control tissues. Positive control expression levels of these molecules can be relevant patient thymus tissue highly expressed the molecules/ markers prognostic indicators. We also studied the impact of tumor under study. Negative control tissue was sections of benign infiltrating lymphocytes (TILS) along with these checkpoint ovarian disease such as polycystic ovarian disease. Sections molecules, on patient status including tumor grade, disease on slides were de-paraffinized in xylene and then rehy- stage and survival post diagnosis. drated in a series of decreasing concentrations of alcohols. Antigen retrieval for PD-L1 and PD-1 was performed by boiling slides in a pressure cooker for 5 min in Universal Methods HIER retrieval agent (ab 208,572, Abcam, Cambridge, MA) Patients at a 1X concentration. Sections were washed in 0.1% tween Patients underwent surgery between 2003 and 2006 at in Dulbecco’s phosphate buffered saline (DPBS; 1X, Lonza, Loyola University Medical Center (LUMC) for ovarian and Walkersville, MD) and then blocked in 0.4% hydrogen per- other gynecologic associated cancers. Tissues were embed- oxide in DPBS, followed by blocking in 10% goat serum ded in paraffin blocks for patient diagnosis to characterize (S1000, Vector Laboratories, Burlingame, CA) for 1 h. stage and grade of cancer in tissue sections, and blocks Antigen retrieval for FoxP3, CD3 and CD8 was performed were stored in the Department of Pathology, LUMC. After by boiling sections in a pressure cooker for 5 min in Reveal receiving approval by the Institutional Review Board (IRB) Decloaker (RV1000G1, Biocare Medical, Concord, CA). After for the Protection of Human Subjects, we selected a cohort washing in DPBS, sections were blocked in 0.4% hydrogen of 55 patients for study, most of whom were diagnosed with peroxide in DPBS for 20 min, 10% goat serum or 10% horse advanced disease ovarian cancer (Table 1). Patient histories serum (S1000 or S2000 respectively, Vector Laboratories) for in the LUMC medical records were evaluated by two inves- 20 min, and then in Avidin/ Biotin blocking reagents (SP tigators and data collected for parameters including: age, 2001, Vector Laboratories) to further reduce non-specific date of birth, date of diagnosis, pre-treatment status before staining of primary antibody (FoxP3, CD3, or CD8). surgery, cancer stage, tumor grade, date of last encounter, and whether the patient was alive or deceased. Dates of Identification and assessment of antigens in patient death were retrieved from the patients’ medical records sections when this date was available, or found by a search on a Tissue sections were incubated overnight in 5% blocking website such as http://www.dobsearch.com/death-records/. serum with or without primary antibody at a pre-determined and optimized dilution. PD-1 (ab137132, Abcam) and PD- Antigen revealing L1 (ab205921, Abcam) were used at 1:500 dilution for IHC Formalin fixed paraffin embedded (FFPE) tissue sections staining. The next day sections were washed in 0.1% tween (4 μm) were adhered to glass slides using tissue from a in DPBS, and an amplifier polymer detection system specific Table 1 Categories of patient tumors Stage Tumor type Histological subtype No. of patients Low (I/ II) All papillary mucinous cystadenocarcinoma 1 Ovarian mixed adenocarcinoma, serous and endometroid 1 papillary serous adenocarcinoma 2 papillary serous carcinoma 4 serous carcinoma 1 High (III/ IV) Ovarian poorly differentiated serous carcinoma 1 Ovarian papillary serous carcinoma 24 Ovarian papillary serous adenocarcinoma 10 Fallopian tube papillary serous carcinoma 1 Peritoneum papillary serous carcinoma 4 Peritoneum papillary serous adenocarcinoma 1 Endometrium papillary serous carcinoma 1 Endometrium serous carcinoma 1 Omentum papillary serous carcinoma 2 Total 55 Tumor stage reflects the International Federation of Gynecology and Obstetrics (FIGO) classification Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 4 of 11 for rabbit anti-human primary antibodies (ab 20,901, Rabbit with Log Rank tests. The frequency of occurrence of each specific IHC polymer detection kit; HRP/ DAB) added ac- marker was graded on a scale from 0 to 4 (Table 2), and cording to the manufacturers’ guidance.Tissue was also Cochran Armitage tests used to determine the statistical stained overnight with primary antibodies for FoxP3 (236A/ significance of trends by patient characteristics including E7; ab 20,034, 1:1600 dilution, Abcam), CD8 (C8/144B; 1: age, cancer stage and tumor grade. Associations between 100 dilution, Cell Marque, Rocklin, CA 1:1000 dilution) and the presence of PD-1, PD-L1, CD3, CD8 and FoxP3 posi- CD3 (F7.2.38; 1:1000 dilution, Dako, Glostrup, Denmark). tive cells with patient age at the time of diagnosis, cancer Sections were washed in DPBS and a biotinylated secondary stage, or tumor grade were determined with chi-square or antibody for peroxidase (PK 6102, Vector Laboratories) Fisher’s exact tests as appropriate. Hazard ratios for overall added for 30 min, followed by an avidin-biotin peroxidase survival (O/S) were determined from univariable Cox pro- complex and enzyme reagent (ABC, Vector laboratories). All portional hazard regression models for each patient charac- sections were washed in DPBS and developed in Vector teristic and each marker. Analyses were performed using NovaRED (SK4800) or diaminobenzidine (DAB; SK4100, SAS 9.4 (SAS Institute, Cary, NC). Vector Laboratories). Sections were counterstained in hematoxylin and rehydrated in xylene, followed by alcohol, Results then mounted in Vectamount H-5000 (Vector Laboratories). Patient characteristics TumorsectionswereexaminedbypathologistsSMand The cohort consisted of 55 patients as follows: ovarian (45 MA to investigate the frequency of occurrence of markers, patients), fallopian tube (1 patient), peritoneum (5 pa- the degree of staining intensity and location of tumor cells or tients), endometrium (2 patients) and omentum (2 pa- lymphocytes expressing each molecule. A customized scor- tients) cancer tissue blocks. Patients were diagnosed as ing system was developed by the abovementioned patholo- stage I or stage II (stage 1 or 2, low/ early stage; 9 patients) gists to obtain a numerical score to represent the average and stage III or IV (stage 3 or 4, high stage or advanced; frequency of antigens as visualized over 7–10 high power 46 patients) disease (Table 1). One patient did not have fields (hpf) of IHC stained tissue sections (Table 2), where known cancer grade, and of the remaining, about three “0” was the lack of expression, and “4” represented the high- quarters of patients had tumor grade 3 (n = 41, 75.9%).The est frequency of expression of molecules in sections. In mean age of patients at the time of diagnosis was 61 years addition to the scoring pattern shown in Table 2, in statistical (standard deviation = 12), with a range of 26 to 85 years. analysis, combined PD-1 was assigned as a mathematical At the time of last follow up, 6 patients were alive, 9 pa- score which was derived by adding the observed pathology tients were lost to follow up (3 of whom were deceased at scores (0–2) for T-PD-1 and S-PD-1 in each patient section. an unknown date), and 40 other patients had recorded In some statistical analysis PD-1 and PD-L1 expression dates of death, 38 of whom had died as a cause of ovarian was classified as low (score of 1) or high frequency (score cancer. The median length of survival time for all patients of 2–4) to decipher correlations between the levels of ex- was 3.10 years (95% confidence interval (CI): 2.24–5.19). pression of these molecules and parameters studied. Localization of checkpoint inhibitory molecules in ovarian Statistical analysis cancer tissue Patient O/S was displayed visually in Kaplan Meier plots We studied the distribution of PD-1 and PD-L1 in ovarian and significance of differences by strata were determined cancer tissue sections by IHC staining. Sections were visu- alized and staining evaluated by methods as outlined in Table 2 Pathological interpretation of IHC stained tissue Table 2. Results showed that PD-1 in ovarian cancer tissue Marker Scoring criteria based on cells/ high power fields was localized primarily to cell membranes. PD-1 stained cells appeared to be primarily tumor infiltrating lympho- T-PD-1 0 = < 1; 1 = 1–10, 2= > 10–50 and 3= > 50 cytes (TILs), with varying degrees of intensity of staining S-PD-1 0 = < 1, 1 = 1–25; 2= > 25–50 and 3= > 50 and frequency of occurrence of these cells. Additionally, PD-L1 0 = < 1, 1= > 1–5, 2= > 5–10 and 3= > 10 this staining was strikingly evident in two separate com- CD3 0= <5, 1 =5–15, 2= > 15–25, 3= > 25–40 and 4= > 40 partments of the tumor microenvironment, thus we scored CD8 0 = < 1, 1 = 1–25, 2= > 25–50 and 3= > 50 this stain in two different categories. Intraepithelial TILs FoxP3 0 = < 1, 1 = 1–5, 2= > 5–15, 3= > 15–25 and 4= > 25 cells/ hpf completely enclosed by tumor epithelial cells and positive Based on the known frequency of each marker in various cancer tissues, each for PD-1 were designated as tumor PD-1 cells (T-PD-1). stained section was observed over 7–10 hpf and given an average numerical These cells were either clustered (Fig. 1a), or scattered in a value to represent the scoring frequency based on the categorization in Table less dense pattern in the tumor epithelium (Fig. 1b). Cells 2. PD-1 was assigned 2 different scores to reflect the observed localization of this molecule in tumor tissue. The percentage of PD-1 stained TILs completely which stained positive for PD-1 in the stromal compart- enclosed by tumor epithelial cells was termed tumor-PD-1 (T-PD-1). PD-1 ment were termed stromal PD-1 (S-PD-1). In the stroma, stained lymphocytes in the stromal compartment were classified as stromal-PD-1 (S-PD-1) there were aggregates of cells staining positive for PD-1 Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 5 of 11 Fig. 2 Membranous staining of PD-L1 positive cells in ovarian tumors. Diffuse membranous staining of almost all ovarian cancer cells (a) Fig. 1 Distinct IHC staining patterns of PD-1 positive cells in the versus focal strong membranous staining in a few scattered tumor cells stroma and epithelium of tissue. In ovarian cancer tissue sections, positive for PDL1 (b) tumor infiltrating lymphocytes (TILS) exhibiting strong membranous and cytoplasmic staining for PD-1 are apparent as clusters (a)or single scattered cells (b). TILS are also seen in aggregates (c)and as single Variable expression of tumor infiltrating lymphocytes in cells (d) in the reactive stroma associated with tumor ovarian cancer Immunocompetent TILS and FoxP3 T regulatory cells, are (Fig. 1c), or regions of fewer cells staining for this molecule strategically located in ovarian cancer tissue. We sought to (Fig. 1d). Table 3 shows a summary of scores for each PD-1 determine whether the levels of CD3, CD8 and/ or FoxP3 T classification as well as the frequency of PD-1 observations cells in the tumors of patients would alter the potential role in tissue in the cohort of 55 patients studied. A total of 48 of checkpoint inhibitory molecules as predictors of disease (87%) patients expressed PD-1 on cells, whilst 40 (73%) pa- pathology or of outcome. In FFPE tissue sections, of all T tients expressed PD-1 in both the epithelial (T-PD-1) and cells, CD3 positive T cells were found in highest frequency, as stromal (S-PD-1) compartments (data not shown). expected (Table 3), sometimes staining intensely. Some pa- In tissue sections, PD-L1 was localized to the cell mem- tients had an abundance of these cells, while other cases brane. In lesions of some patients there were solid tumor ag- showed a scattered arrangement (Fig. 3a and b respectively). gregates with diffuse membranous staining of these cells for Staining for CD8 T cells was widely observed in patient sec- this marker (Fig. 2a). PD-L1 was also identified in scattered tions as membranous reactivity, either with a dense distribu- tumor cells with focal strong membranous staining (Fig. 2b). tion, or with lower frequency (Fig. 3c and d respectively). Additionally, in some sections PD-L1 staining was observed in cells with the morphology of immune cells (not shown). Staining for this marker was positive in approximately one- third (18 of 55, 33%) of patients with ovarian cancer. Table 3 Occurrence of checkpoint inhibitory molecules and TILS in tumors Molecules in Frequency Total patients tumor sections positive T-PD-1 10 35 10 0 0 45 S-PD-1 12 39 4 0 0 43 PD-L1 37 14 3 1 0 18 CD3 2 23185 753 CD8 17 15 18 5 0 38 Fig. 3 High and low density of TILs in ovarian tumors. IHC staining FoxP3 7 12 15 19 2 48 of T cell subsets in patients’ FFPE tissue sections. CD3 exhibiting diffuse strong staining in clusters of tumor infiltrating lymphocytes Tumor sections of 55 ovarian and related cancers were IHC stained for molecules as indicated. Sections were visualized by microscopy and a (a) versus focal staining in scattered TILs in less dense areas (b). pathological score assigned from 0 to 4, based on the frequency of expression Distribution of CD8 T cells in clusters and as single cells is apparent of each molecule, where “0” indicates the absence and “4” the highest in c and d respectively frequency of expression Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 6 of 11 FoxP3 staining of cells was revealed as strong nuclear reactiv- The presence of PD-L1 is positively associated with high ity, either in clustered foci in some patient sections (Fig. 4a), grade tumors or scattered in the tumor stroma in other cases (Fig. 4b). We undertook detailed analysis to determine how the pres- FoxP3 positive cells were present in 48 of 55 patients. ence of checkpoint inhibitor molecules, PD-1 or PD-L1 correlated with patient history. We did not find any signifi- cant associations between the presence of PD-1, with pa- Clinical parameters as predictors of patient prognosis tient age at the time of diagnosis, stage of cancer or tumor We examined trends between age of patients at the time grade (Table 4). of diagnosis, tumor grade or tumor stage with overall sur- Only one patient with low tumor grade expressed PD- vival. Significant parameters for these associations are rep- L1, thus 17/18 (94%) patients who expressed PD-L1 in resented in Kaplan Meier plots (Additional file 1:Figure their ovarian tumors had a tumor grade of 3. Thus PD- S1). We found that patients diagnosed with ovarian cancer L1 was expressed significantly more often in patients when over 60 years of age had a higher hazard of death with high grade tumor than in those with low grade (HR: 2.63, 95% CI: 1.34–5.16) and were significantly more tumor (n = 54; 41.5% versus 7.7%, P = 0.040) (Table 4). likely to die as a cause of disease than patients diag- We did not observe any other significant associations nosed at a younger age (P = 0.005), (Additional file 2: with the presence of any other molecules studied as sin- Table S1). Additionally, patients diagnosed with ad- gle predictors, with age at the time of diagnosis, stage of vanced disease (stage III or IV) had a higher hazard of disease or tumor pathology. death (HR: 3.70, 95% CI: 1.28–10.76) and were signifi- Further analysis of data showed that the presence of cantly more likely to die than patients diagnosed in CD3 or of FoxP3 infiltrating cells together with PD-L1 the early stages of the disease (P = 0.016). Tumor grade in patient tumors did not impact the significance of the had no significant association with patient survival association of PD-L1 with high grade tumors (P = 0.040) (Additional file 2:Table S1). (Table 5). There was borderline significant association Similar analysis was performed to determine survival between the presence of S-PD-1 cells together with estimates as a function of the expression of each marker. FoxP3 positive cells in tumors with high grade It was found that while survival was similar in the first (P = 0.075). The presence of TILs and of cells expressing few years of follow-up for those with or without PD-L1, PD-1 or PD-L1 occurring together in ovarian tumors those with PD-L1 present in tissue sections appeared to had no impact on disease stage (Table 5), or on survival have a survival advantage with increased time after diag- (data not shown). nosis, even though this trend was not statistically signifi- cant (Additional file 3: Figure S2A). The presence of PD- 1 and of CD3 showed modest but insignificant trends to- Low frequency of expression of PD-1 and of PD-L1 ward improved survival (Additional file 3: Figure S2 B correlates with advanced ovarian cancer and C). CD8 or FoxP3 did not show a significant associ- We further investigated whether high or low frequency of ation with survival, even though for high expression of expression of PD-1 correlated with the stage or grade of FoxP3 there was a trend towards decreased survival cancer. Based on our scoring pattern for the occurrence of (data not shown). Table 4 Patient history and presence of PD-1 and PD-L1 combined PD-1 PD-L1 No. Patients n (%) p-value n (%) P-value Age at diagnosis < 60 27 24 (88.9) 0.99 9 (33.3) 0.93 ≥ 60 28 24 (85.7) 9 (32.1) Stage of cancer I-II 9 9 (100.0) 0.58 5 (55.6) 0.13 III-IV 46 39 (84.8) 13 (28.3) Tumor grade 1–2 13 12 (92.3) 0.99 1 (7.7) 0.040 3 41 35 (85.4) 17 (41.5) Fig. 4 Distribution of FoxP3 expressing T regulatory cells in ovarian tumors. Intense nuclear staining of FoxP3 positive lymphocytes in a The presence of PD-1 in the tumor epithelial or stromal compartment (combined PD-1), and PD-L1 was defined as the occurrence of these molecules focal arrangement (a). Other cases (b) showed nuclear staining of (scores 1, 2, 3, or 4) in tissue sections. Tumor grade was unknown for one fewer cells positive for this marker patient. Significant P-values are indicated in italics Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 7 of 11 Table 5 Association with immune markers, grade and stage No. Patients n (%) with stage III-IV P-value n (%) with grade 3 P-value Overall 55 46 (83.6) 41/54 (75.9) Presence with CD3 T-PD-1 + CD3 44 35 (79.5) 0.18 31 (72.1) 0.26 S-PD-1 + CD3 43 34 (79.1) 0.18 34 (81.0) 0.13 Combined PD-1 + CD3 47 38 (80.9) 0.33 34 (73.9) 0.66 PD-L1 + CD3 18 13 (72.2) 0.13 17 (94.4) 0.040 Presence with FoxP3 T-PD-1 + FoxP3 41 33 (80.5) 0.42 30 (75.0) 0.99 S-PD-1 + FoxP3 41 33 (80.5) 0.42 33 (82.5) 0.075 Combined PD-1 + FoxP3 44 36 (81.8) 0.67 33 (76.7) 0.99 PD-L1 + FoxP3 18 13 (72.2) 0.13 17 (94.4) 0.040 Evaluation of the association of the presence of PD-1 or PD-L1 together with CD3 and FoxP3 T cells on advanced disease and on high grade tumors. Significant P-values are indicated in italics PD-1 in the tumor epithelium (T-PD-1) or in the stroma or moderate staining for PD-L1 in either stromal or tumor (S-PD-1), we initially analyzed observations in these com- epithelial compartment, and an average of cells stained over partments separately. As expected, the frequency of PD-1 7 hpfs taken into account. Most patients (5/9; 56%) with expressing cells in tumor tissue was in general not as high early stage disease expressed PD-L1. A lower percentage of as those expressing CD3 or CD8 (Table 3). A higher per- patients with advanced disease expressed PD-L1 (13/46; centage of patients with early stage cancer were more 28%) (Table 7). Most patients (11/13; 85%) with advanced likely to have a higher frequency of expression (pathology disease who expressed PD-L1, expressed this molecule in score of2)ofT-PD-1 thaninpatients withadvanced tumors with a score of 1 (on a scale of 0–3). A higher per- disease, but this difference was not significant (P =0.13, centage of patients with early stage disease (22%) expressed data not shown). In the case of S-PD-1, a higher percent- highlevels of PD-L1(frequencyscore 2or3)incomparison age of patients with early stage cancer had higher levels with patients with advanced disease (4.3%) (Table 7). There (frequency score of 2, 22%) of S-PD-1 than for advanced were no significant trends in frequencies of CD3, CD8 or disease patients (frequency score of 2, 2%) (P =0.033) FoxP3 with age at the time of diagnosis, stage of disease, or (Additional file 4: Table S2). tumor grade (data not shown). A significantly higher percentage (33%) of early stage patients had a higher frequency (score 3 and 4) of com- bined PD-1 (PD-1 in the tumor epithelium and/or stroma) Discussion than those with advanced disease (17%) (P =0.028, Ovarian cancer is usually diagnosed in the advanced meta- Table 6). Thus, a high density of PD-1 was significantly as- static stages. Treatment of advanced stage disease with sociated with early stage disease diagnosis, and a low PD-1 conventional therapies is only sufficiently effective in a density was associated with advanced disease. limited number of patients, thus in about 80 % of these PD-L1 was expressed in 33% of the patient cohort. Scor- patients there is disease progression or disease recurrence ing for PD-L1 was performed by scoring regions of intense and death, within five years of diagnosis. In many cancers, Table 6 Patient tumor and combined frequency of PD-1 expression No. Patients 0 1 2 3 4 P-value Combined PD-1 Level, n (%) Stage of cancer I-II 9 0 (0.0) 0 (0.0) 6 (66.7) 1 (11.1) 2 (22.2) 0.028 III-IV 46 7 (15.2) 8 (17.4) 23 (50.0) 7 (15.2) 1 (2.2) Tumor grade 1–2 13 1 (7.7) 4 (30.8) 7 (53.8) 0 (0.0) 1 (7.7) 0.54 3 41 6 (14.6) 4 (9.8) 21 (51.2) 8 (19.5) 2 (4.9) Combined PD-1 was derived by adding the observed pathology scores for tumor PD-1 (T-PD-1) and S-PD-1 in each patient tumor section. Significant P-value in table is indicated in italics Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 8 of 11 Table 7 Patient tumor characteristics and frequency of PD-L1 expression No. Patients 0 1 2 3 P-value PD-L1 Level, n (%) Stage of cancer I-II 9 4 (44.4) 3 (33.3) 1 (11.1) 1 (11.1) 0.033 III-IV 46 33 (71.7) 11 (23.9) 2 (4.3) 0 (0.0) Tumor grade 1–2 13 12 (92.3) 1 (7.7) 0 (0.0) 0 (0.0) 0.033 3 41 24 (58.5) 13 (31.7) 3 (7.3) 1 (2.4) Significant P- values in table are indicated in italics investigators are focusing on the development of novel and TILS in the tumors of ovarian or related cancer pa- therapies as alternative and more robust options to exist- tients, diagnosed with advanced disease. Firstly, results ing therapies. Whereas conventional therapies primarily showed that patients’ age was an independent prognostic focus on the destruction of tumor cells, many novel ther- factor in survival, with patients over 60 years of age apies are designed to stimulate immune cells to elaborate more likely to die than those diagnosed when younger augmented anti-tumor immune responses. In this respect, than 60 years. This may be due to the fact that younger checkpoint immune inhibitory molecules have come full patients can tolerate more aggressive surgery and circle over the last decade for cancer immunotherapy. chemotherapy than older patients. Additionally, as ex- In a normal functioning immune system, T cell pected, disease stage was also an independent prognostic activating and inhibitory receptors balance immune factor in outcome, such that patients diagnosed with ad- tolerance, and the amplification of immune responses. vanced disease had a lower median survival than pa- In the body, immune checkpoints are designed to reduce tients diagnosed with early stage disease. These findings autoimmune responses, or to attenuate immune re- are in agreement with those of other investigators [32]. sponses which were elaborated after infections [10–12, To shed lightonthe relevanceofPD-1and PD-L1in 14, 27, 28]. In cancer, blocking of immune checkpoint ovarian cancer outcome, we studied the abundance of these molecules with antibodies is a novel and promising ther- molecules in the TME. PD-1 was compartmentalized in the apy, as it potentiates anti-tumor immune responses in stroma and in the tumor epithelium, and this molecule was patients [9, 13, 15, 19]. expressed in 87% of tumors. PD-L1 was only present in the The first of these checkpoint inhibitory molecules to tumors of 33% of patients. Patients who expressed PD-L1 be targeted for blocking in therapy, and is now FDA ap- had a trend towards survival, as did those expressing PD-1 proved for cancer therapy, is CTLA-4 [16, 17]. Therapy or CD3, even though these trends were not significant. In design is based on the following principle. Binding of our cohort we did not find a significant association with costimulatory molecules CD80 or CD86 on antigen pre- FoxP3 and survival. The presence of PD-L1 and FoxP3 to- senting cells to CD28 on T cells delivers a positive costi- gether in high grade tumors showed the same level of asso- mulatory signal contributing to T cell activation. On the ciation as the presence of PD-L1 alone. Some studies have contrary, linkage of CD80 or CD86 to CTLA-4 (a mol- reported that FoxP3 positive cells in ovarian tumors is ecule closely related to CD28) results in inhibition of im- negatively associated with outcome, however a meta- mune responses, and exhausted T cells, which are less analysis of 7 ovarian cancer studies with a total of 869 pa- able to proliferate or to secrete T helper 1 (Th1) cyto- tients, did not find FoxP3 TILS in ovarian cancer to be a kines [29] . Blocking of this inhibitory pathway with significant prognostic indicator [33]. anti-CTLA-4 antibodies, results in re-invigorated T cells In cancer there are conflicting reports concerning the with greater proliferative function, and durable anti- expression patterns of PD-1 in patients’ tumors and the tumor potency [30, 31]. By similar rationale, there are association with survival, with either positive or a nega- several other immune checkpoints in the immune tive association [34–36]. One recent report found that system, which are now being targeted in cancer clinical PD-1 positive TILS and /or PD-L1 positive tumor cells trials. In the case of blocking PD-1 signaling, there are had a positive association with survival of ovarian cancer currently about nine different antibodies in cancer patients [37]. clinical trials targeting the PD-1/ PD-L1 pathway [3, 17]. The expression of PD-L1 in tumors was shown to be This study was designed to better understand the ovar- positively associated with survival in NSCLC [38] and in ian cancer tumor microenvironment (TME) with rela- ovarian cancer [37]. On the contrary, others report a tion to the localization and frequency of PD-1, PD-L1 negative prognostic impact of PD-L1 expression in Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 9 of 11 ovarian cancer [39, 40] and breast cancer [41]. In a re- tumors. Patients then undergo courses of standard ther- view and meta-analysis of 17 studies using data of 2869 apy such as chemotherapy. Consistent with this practice, head and neck cancer (HNC) patients, authors found we found that a study of our patient cohort treatment that there was no significant association between the ex- plans revealed that only 3 of 55 (5.5%) patients had pression of PD-L1 on survival in HNC patients [42]. chemotherapy in the interval before surgery. Additionally, similar analysis of reports with NSCLC pa- A low frequency of PD-1 in tumors was associated tients also did not show a significant association with with advanced disease. This association between low PD-L1 expression and survival [43]. Taken together, this PD-1 density and advanced disease was only significant indicates that the prognostic impact of PD-1/ PD-L1 ex- when measuring S-PD-1 or combined PD-1, whereas pression in tumors is not yet established. Here, our stud- low T-PD-1 density alone was not associated with ad- ies did not find a significant association between survival vanced disease. Although beyond the scope of this study, and PD-1 or PD-L1 expression in ovarian cancer. this finding raises the possibility that T-PD-1 and S-PD- Differences in reports of the expression of these mole- 1 positive cells may perform unique immunosuppressive cules and associations with survival may be attributed to roles in the ovarian TME. several reasons. Firstly, it is possible that there may be PD-L1 expression was almost exclusively restricted to different survival outcomes due to the site of cancer. For high grade tumors, such that there was a positive and sig- example, Paulsen and colleagues [38] found that whereas nificant association between PD-L1 and high grade tumors. in a cohort of patients a high density of PD-1 and PD-L1 This finding may be of translational significance in selecting had a favorable impact on NSCLC, this association was patients for therapy blocking PD-1/ PD-L1 signaling, and not present when these molecules were studied at meta- we suggest that patients with high grade tumors, with pre- static sites such as lymph nodes of the same patients. existing PD-L1 expression might be excellent candidates This is highly likely because each cancer site has a differ- for therapy blocking this pathway. In support of this idea, a ent immune landscape, and levels of soluble molecules recent report shows that in an ongoing study of urothelial such as IFN-γ which is a strong regulator of PD-L1 ex- bladder cancer patients, treatment with durvalumab pression [44, 45]. (MED14736; an anti-PD-L1 antibody) resulted in improved Secondly, differences in reports may be due to varia- outcome in PD-L1 tumor positive patients. In pre- tions in staining protocols across individual laboratories. treatment tumor biopsies patients, 40 patients were PD-L1 Many investigators report difficulty in IHC staining for positive and 21 patients negative for PD-L1. In 42 evaluable PD-L1. In these present investigations, we initially used patients, the ORR was 31.0% (95% CI, 17.6 to 47.1), the tumor arrays to study the expression of PD-L1 and PD-1 ORR was 46.4% (95%CI, 27.5 to 66.1) in the PD-L1 positive in tumors. When optimizing our staining protocol, we patient subgroup, and 0% (95% CI, 0.0 to 23.2) in patients found that it was difficult to select cores which had a negative for PD-L1 [46]. good representation of tumor and stromal areas for ac- We suggest that due to the conflicting reports concerning curate visualization of the density of these molecules in the impact of PD-1 and PD-L1 on survival in cancer pa- tissue arrays. Therefore, in this study we used whole tients, a future larger study is needed investigating these tumor sections for the identification of these molecules, molecules in ovarian tissue, with standardized protocols as is done for patient diagnosis. For molecules such as and defined cut off points for positive staining and scoring PD-L1 especially, which is not widely expressed in criteria across centers, to minimize study variations. Even tumor tissues, reports in which IHC staining was per- so, the potency of patient responses to PD-1/ PD-L1 block- formed using tumor cores may give different findings ingantibodytherapy maybeinfluenced by thedensity of than reports from other laboratories in which staining other pre-existing or emerging checkpoint molecules in tu- was performed using whole tissue sections. mors, including T-cell immunoglobulin and mucin-domain Thirdly, the use of different primary antibody clones to containing-3 (TIM-3), lymphocyte-activation gene 3 (LAG- identify PD-1 or PD-L1 in tumor tissue in each laboratory, 3) and V-domain Ig suppressor of T cell activation (VISTA). may also lead to variability in staining of sections and in Other parameters such as the presence of myeloid derived interpretation. Finally, manual staining protocols in com- suppressor cells, levels of Th2 cytokines (for example IL- parison with automated staining may further contribute to 10) and of indoleamine 2,3-dioxygenase (IDO) can also variations in staining interpretation. limit anti-cancer immune responses to therapy blocking Finally, additional parameters which may alter the ex- PD-1/ PD-L1 [19, 47, 48]. Furthermore, genetic alterations pression of these molecules in tumors, is the administra- within the tumor (including DNA rearrangements, muta- tion of treatments such as chemotherapy to patients tions, deletions and insertions) alter tumor mutational prior to surgery. However, at our center, for patients di- loads, and it is reported that tumors with high mutational agnosed with ovarian and related gynecologic cancer, the loads have the greatest response to checkpoint inhibitory primary treatment is most often surgery for removal of blockade therapy [17, 49, 50]. Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 10 of 11 Finally, due to the multiplicity of factors regulating Authors’ contributions MLD Conceptualization, investigation, patient history data, data analysis, ORR, we believe that antibody therapy targeting the PD- writing- original draft, review and editing, and project administration. SM 1/ PD-L1 pathway in ovarian cancer will be of maximum Pathology grading and scoring, writing- review and editing. MA Pathology efficacy when used in combination with other treatment grading and scoring, patient history data, writing- review and editing. RP Patient surgery, writing- review and editing. YL Investigation, writing- review. regimens. Such treatments include standard therapy, im- AG Investigation, writing- review. CJ Statistical analysis. Writing- statistical munotherapy blocking other checkpoint inhibitory mol- writing, review and editing. TO Statistical analysis, writing- review. MSS ecules, dendritic cell vaccines, chimeric antigen receptor Conceptualization, investigation, data analysis, writing- review and editing, and project administration. PJS Investigation, writing- review and editing, and (CAR) T cell therapy, or targeted therapy, all of which project administration. All authors read and approved the final manuscript. can downregulate other immune suppressive mecha- nisms in patients, concomitantly. Ethics approval and consent to participate Investigators received approval by the Institutional Review Board (IRB) for the Protection of Human Subjects at Loyola University Medical Center for the use of the cohort of patients in study. Conclusions Treatments inhibiting PD-1 and PD-L1 are beneficial only Competing interests in some ovarian cancer patients. Our studies reveal that a The authors declare that they have no competing interests. low density of PD-1 and of PD-L1 expressing cells in tumor tissue are significantly associated with advanced disease, Publisher’sNote and that PD-L1 is expressed significantly more in high Springer Nature remains neutral with regard to jurisdictional claims in grade tumors than in low grade tumors. We conclude that published maps and institutional affiliations. a subgroup of advanced disease ovarian cancer patients with Author details high grade tumors, bearing PD-L1, may be the best candi- 1 Cardinal Bernardin Cancer Center, Oncology Research Institute, Department dates for immunotherapy targeting PD-1 and/ or PD-L1. of Medicine, Loyola University Chicago, Bldg. 112, Room 232, 2160 South First Avenue, Maywood, IL 60153, USA. Department of Pathology, Stritch School of Medicine, Loyola University Chicago, Maywood, IL, USA. Department of Obstetrics and Gynecology, Loyola University Chicago, Maywood, IL, USA. Department of Chemistry & Biochemistry, Harper Cancer Additional files Research Institute, University of Notre Dame, South Bend, IN, USA. Department of Public Health Sciences, Loyola University Chicago, Maywood, Additional file 1: Figure S1. Survival estimates by patient age and 6 IL, USA. Department of Mathematics & Statistics, and Institute of tumor grade. Kaplan Meier survival analysis to estimate overall survival in Environmental Sustainability, Loyola University Chicago, Chicago, IL, USA. patients as a function of age (A) or tumor stage (B). Patient survival was displayed visually in Kaplan Meier plots and significance of differences Received: 5 January 2018 Accepted: 9 May 2018 determined with Log Rank tests. (PPTX 427 kb) Additional file 2: Table S1. Univariable hazard ratios for Cox proportional hazards models (DOCX 25 kb) References Additional file 3: Figure S2. Survival estimates by expression of immune 1. Luvero D, Milani A, Ledermann JA. Treatment options in recurrent ovarian molecules. Kaplan Meier survival analysis to estimate overall survival in patients cancer: latest evidence and clinical potential. Ther Adv Med Oncol. 2014; expressing PD-L1 (A), PD-1 (B) and CD3 (C) in tumor sections. (PPTX 456 kb) 6(5):229–39. 2. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin. 2017; Additional file 4: Table S2. Patient tumor and frequency of S-PD-1 67(1):7–30. expression (DOCX 25 kb) 3. Iwai Y, Hamanishi J, Chamoto K, Honjo T. Cancer immunotherapies targeting the PD-1 signaling pathway. J Biomed Sci. 2017;24(1):1–11. 4. McCarthy EF. The toxins of William B. Coley and the treatment of bone and Abbreviations soft-tissue sarcomas. The Iowa orthopaedic journal. 2006;26:154–8. CR: complete response; CTLA-4: cytotoxic T lymphocyte associated-4; 5. Nauts HC, Swift WE, Coley BL. The treatment of malignant tumors by FFPE: formalin fixed paraffin embedded; HNC: head and neck cancer; bacterial toxins as developed by the late William B. Coley, M.D., reviewed in hpf: high power fields; IHC: immunohistochemistry; NSCLC: non-small cell the light of modern research. Cancer Res. 1946;6:205–16. lung cancer; O/S: overall survival; ORR: objective response rate; PD- 6. Chester C, Dorigo O, Berek JS, Kohrt H. Immunotherapeutic approaches to 1: programmed cell death-1; PD-L1: programmed cell death-1 ligand; ovarian cancer treatment. Journal for immunotherapy of cancer. 2015;3:1–10. PR: partial response; S-PD-1: stromal PD-1; Th1: T helper-1; TILS: tumor 7. Hamanishi J, Mandai M, Konishi I. Immune checkpoint inhibition in ovarian infiltrating lymphocytes; TME: tumor microenvironment; T-PD-1: Tumor PD-1 cancer. Int Immunol. 2016;28(7):339–48. 8. Hodi FS, Butler M, Oble DA, Seiden MV, Haluska FG, Kruse A, Macrae S, Nelson M, Canning C, Lowy I, et al. Immunologic and clinical effects of Funding antibody blockade of cytotoxic T lymphocyte-associated antigen 4 in This research was supported by a joint research award offered by the previously vaccinated cancer patients. Proc Natl Acad Sci U S A. 2008;105(8): Cardinal Bernardin Cancer Center, Loyola University (to M.L.D.), and the 3005–10. Harper Cancer Research Institute, University of Notre Dame (to M.S.S.), made 9. Pardoll DM. The blockade of immune checkpoints in cancer possible by funds from the Cronk Foundation. Additional support was from immunotherapy. Nat Rev Cancer. 2012;12(4):252–64. the Perritt Charitable Foundation (to P.J.S.). Funding agencies did not 10. Bayry J. Autoimmunity: CTLA-4: a key protein in autoimmunity. Nat Rev participate in the design of the study, collection, analysis or interpretation of Rheumatol. 2009;5(5):244–5. data, or in writing the manuscript. 11. Blank C, Mackensen A. Contribution of the PD-L1/PD-1 pathway to T-cell exhaustion: an update on implications for chronic infections and tumor Availability of data and materials evasion. Cancer Immunol Immunother. 2007;56(5):739–45. All data generated or analyzed during this study are included in this 12. Keir ME, Butte MJ, Freeman GJ, Sharpe AH. PD-1 and its ligands in tolerance published article (and its additional information files). and immunity. Annu Rev Immunol. 2008;26:677–704. Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 11 of 11 13. Brahmer J, Reckamp KL, Baas P, Crino L, Eberhardt WE, Poddubskaya E, 35. Schmidt LH, Kummel A, Gorlich D, Mohr M, Brockling S, Mikesch JH, Antonia S, Pluzanski A, Vokes EE, Holgado E, et al. Nivolumab versus Grunewald I, Marra A, Schultheis AM, Wardelmann E, et al. PD-1 and PD-L1 docetaxel in advanced squamous-cell non-small-cell lung Cancer. N Engl J expression in NSCLC indicate a favorable prognosis in defined subgroups. Med. 2015;373(2):123–35. PLoS One. 2015;10(8):1–15. 14. Flies DB, Sandler BJ, Sznol M, Chen L. Blockade of the B7-H1/PD-1 pathway 36. Thompson RH, Dong H, Lohse CM, Leibovich BC, Blute ML, Cheville JC, for cancer immunotherapy. Yale J biol Med. 2011;84(4):409–21. Kwon ED. PD-1 is expressed by tumor-infiltrating immune cells and is associated with poor outcome for patients with renal cell carcinoma. Clin 15. Larkin J, Chiarion-Sileni V, Gonzalez R, Grob JJ, Cowey CL, Lao CD, Cancer Res. 2007;13(6):1757–61. Schadendorf D, Dummer R, Smylie M, Rutkowski P, et al. Combined 37. Darb-Esfahani S, Kunze CA, Kulbe H, Sehouli J, Wienert S, Lindner J, Budczies Nivolumab and Ipilimumab or monotherapy in untreated melanoma. N J, Bockmayr M, Dietel M, Denkert C, et al. Prognostic impact of Engl J Med. 2015;373(1):23–34. programmed cell death-1 (PD-1) and PD-ligand 1 (PD-L1) expression in 16. Hodi FS, O'Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB, cancer cells and tumor-infiltrating lymphocytes in ovarian high grade Gonzalez R, Robert C, Schadendorf D, Hassel JC, et al. Improved survival serous carcinoma. Oncotarget. 2016;7(2):1486–99. with ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010; 38. Paulsen EE, Kilvaer TK, Khanehkenari MR, Al-Saad S, Hald SM, Andersen S, 363(8):711–23. Richardsen E, Ness N, Busund LT, Bremnes RM, et al. Assessing PDL-1 and 17. Gaillard SL, Secord AA, Monk B. The role of immune checkpoint inhibition in PD-1 in non-small cell lung Cancer: a novel Immunoscore approach. Clin the treatment of ovarian cancer. Gynecol Oncol Res Pract. 2016;3:1–14. Lung Cancer. 2017;18(2):220–33. e228 18. Okazaki T, Honjo T. PD-1 and PD-1 ligands: from discovery to clinical 39. Hamanishi J, Mandai M, Iwasaki M, Okazaki T, Tanaka Y, Yamaguchi K, application. Int Immunol. 2007;19(7):813–24. Higuchi T, Yagi H, Takakura K, Minato N, et al. Programmed cell death 1 19. Topalian SL, Drake CG, Pardoll DM. Immune checkpoint blockade: a ligand 1 and tumor-infiltrating CD8+ T lymphocytes are prognostic factors common denominator approach to cancer therapy. Cancer Cell. 2015;27(4): of human ovarian cancer. Proc Natl Acad Sci U S A. 2007;104(9):3360–5. 450–61. 40. Maine CJ, Aziz NH, Chatterjee J, Hayford C, Brewig N, Whilding L, George AJ, 20. Gubin MM, Zhang X, Schuster H, Caron E, Ward JP, Noguchi T, Ivanova Y, Ghaem-Maghami S. Programmed death ligand-1 over-expression correlates Hundal J, Arthur CD, Krebber WJ, et al. Checkpoint blockade cancer with malignancy and contributes to immune regulation in ovarian cancer. immunotherapy targets tumour-specific mutant antigens. Nature. 2014; Cancer immunology, immunotherapy : CII. 2014;63(3):215–24. 515(7528):577–81. 41. Muenst S, Soysal SD, Gao F, Obermann EC, Oertli D, Gillanders WE. The 21. Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, presence of programmed death 1 (PD-1)-positive tumor-infiltrating Powderly JD, Carvajal RD, Sosman JA, Atkins MB, et al. Safety, activity, and lymphocytes is associated with poor prognosis in human breast cancer. immune correlates of anti-PD-1 antibody in cancer. N Engl J Med. 2012; Breast Cancer Res Treat. 2013;139(3):667–76. 366(26):2443–54. 42. Li J, Wang P, Xu Y. Prognostic value of programmed cell death ligand 1 22. De Felice F, Marchetti C, Palaia I, Musio D, Muzii L, Tombolini V, Panici PB. expression in patients with head and neck cancer: a systematic review and Immunotherapy of ovarian Cancer: the role of checkpoint inhibitors. J meta-analysis. PLoS One. 2017;12(6):1–16. Immunol Res. 2015;2015:1–7. 43. Zhong A, Xing Y, Pan X, Shi M, Xu H. Prognostic value of programmed cell 23. Zhu X, Lang J. Programmed death-1 pathway blockade produces a death-ligand 1 expression in patients with non-small-cell lung cancer: evidence synergistic antitumor effect: combined application in ovarian cancer. J from an updated meta-analysis. OncoTargets and therapy. 2015;8:3595–601. Gynecol Oncol. 2017;28(5):1–19. 44. Benci JL, Xu B, Qiu Y, Wu TJ, Dada H, Twyman-Saint Victor C, Cucolo L, Lee 24. Hamanishi J, Mandai M, Ikeda T, Minami M, Kawaguchi A, Murayama T, DS, Pauken KE, Huang AC, et al. Tumor interferon signaling regulates a Kanai M, Mori Y, Matsumoto S, Chikuma S, et al. Safety and antitumor multigenic resistance program to immune checkpoint blockade. Cell. 2016; activity of anti-PD-1 antibody, Nivolumab, in patients with platinum- 167(6):1540–54. e1512 resistant ovarian Cancer. J Clin Oncol. 2015;33(34):4015–22. 45. Garcia-Diaz A,Shin DS, Moreno BH,SacoJ, Escuin-Ordinas H,Rodriguez GA, 25. Abiko K, Mandai M, Hamanishi J, Yoshioka Y, Matsumura N, Baba T, Zaretsky JM, Sun L, Hugo W, Wang X, et al. Interferon receptor signaling Yamaguchi K, Murakami R, Yamamoto A, Kharma B, et al. PD-L1 on tumor pathways regulating PD-L1 and PD-L2 expression. Cell Rep. 2017;19(6):1189–201. cells is induced in ascites and promotes peritoneal dissemination of ovarian 46. Massard C, Gordon MS, Sharma S, Rafii S, Wainberg ZA, Luke J, Curiel TJ, cancer through CTL dysfunction. Clin Cancer Res. 2013;19(6):1363–74. Colon-Otero G, Hamid O, Sanborn RE, et al. Safety and efficacy of 26. Ritprajak P, Azuma M. Intrinsic and extrinsic control of expression of the Durvalumab (MEDI4736), an anti-programmed cell death Ligand-1 immune immunoregulatory molecule PD-L1 in epithelial cells and squamous cell checkpoint inhibitor, in patients with advanced urothelial bladder Cancer. J carcinoma. Oral Oncol. 2015;51(3):221–8. Clin Oncol. 2016;34(26):3119–25. 27. Taube JM, Klein A, Brahmer JR, Xu H, Pan X, Kim JH, Chen L, Pardoll DM, 47. Koyama S, Akbay EA, Li YY, Herter-Sprie GS, Buczkowski KA, Richards WG, Topalian SL, Anders RA. Association of PD-1, PD-1 ligands, and other Gandhi L, Redig AJ, Rodig SJ, Asahina H, et al. Adaptive resistance to features of the tumor immune microenvironment with response to anti-PD- therapeutic PD-1 blockade is associated with upregulation of alternative 1 therapy. Clin Cancer Res. 2014;20(19):5064–74. immune checkpoints. Nat Commun. 2016;7:1–9. 28. Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJ, Robert L, Chmielowski B, 48. Yigit R, Massuger LF, Figdor CG, Torensma R. Ovarian cancer creates a Spasic M, Henry G, Ciobanu V, et al. PD-1 blockade induces responses by suppressive microenvironment to escape immune elimination. Gynecol inhibiting adaptive immune resistance. Nature. 2014;515(7528):568–71. Oncol. 2010;117(2):366–72. 29. Krummel MF, Allison JP. CD28 and CTLA-4 have opposing effects on the 49. Chalmers ZR, Connelly CF, Fabrizio D, Gay L, Ali SM, Ennis R, Schrock A, response of T cells to stimulation. J Exp Med. 1995;182(2):459–65. Campbell B, Shlien A, Chmielecki J, et al. Analysis of 100,000 human cancer 30. Buchbinder EI, Desai A. CTLA-4 and PD-1 pathways: similarities, differences, genomes reveals the landscape of tumor mutational burden. Genome Med. and implications of their inhibition. Am J Clin Oncol. 2016;39(1):98–106. 2017;9(1):1–14. 31. Peggs KS, Quezada SA, Chambers CA, Korman AJ, Allison JP. Blockade of CTLA- 50. Van Allen EM, Miao D, Schilling B, Shukla SA, Blank C, Zimmer L, Sucker A, 4 on both effector and regulatory T cell compartments contributes to the Hillen U, Foppen MHG, Goldinger SM, et al. Genomic correlates of response to antitumor activity of anti-CTLA-4 antibodies. J Exp Med. 2009;206(8):1717–25. CTLA-4 blockade in metastatic melanoma. Science. 2015;350(6257):207–11. 32. Deng F, Xu X, Lv M, Ren B, Wang Y, Guo W, Feng J, Chen X. Age is associated with prognosis in serous ovarian carcinoma. J Ovarian Research. 2017;10(36):1–9. 33. Shang B, Liu Y, Jiang SJ, Liu Y. Prognostic value of tumor-infiltrating FoxP3+ regulatory T cells in cancers: a systematic review and meta-analysis. Sci Rep. 2015;5:1–9. 34. Badoual C, Hans S, Merillon N, Van Ryswick C, Ravel P, Benhamouda N, Levionnois E, Nizard M, Si-Mohamed A, Besnier N, et al. PD-1-expressing tumor-infiltrating T cells are a favorable prognostic biomarker in HPV- associated head and neck cancer. Cancer Res. 2013;73(1):128–38. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Ovarian Research Springer Journals

Stratification of ovarian tumor pathology by expression of programmed cell death-1 (PD-1) and PD-ligand- 1 (PD-L1) in ovarian cancer

Free
11 pages

Loading next page...
 
/lp/springer_journal/stratification-of-ovarian-tumor-pathology-by-expression-of-programmed-Tb0r6aOThI
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Gynecology; Reproductive Medicine
eISSN
1757-2215
D.O.I.
10.1186/s13048-018-0414-z
Publisher site
See Article on Publisher Site

Abstract

Background: Ovarian cancer is the major cause of death among gynecologic cancers with 75% of patients diagnosed with advanced disease, and only 20% of these patients having a survival duration of five years. Treatments blocking immune checkpoint molecules, programmed cell death (PD-1) or its ligand PD-ligand- I (PD- L1) have produced a beneficial and prolonged effect in a subgroup of these patients. However, there is debate in the literature concerning the prognostic value of the expression of these molecules in tumors, with immunotherapy responsiveness, and survival. We evaluated the immune landscape of the ovarian tumor microenvironment of patients, by measuring the impact of the expression of tumor PD-1, PD-L1 and infiltrating lymphocytes on stage and grade of tumors and survival, in a cohort of 55 patients with gynecologic malignancies. Most patients under study were diagnosed with advanced disease ovarian cancer. Results: Our studies revealed that a low density of PD-1 and of PD-L1 expressing cells in tumor tissue were significantly associated with advanced disease (P = 0.028 and P = 0.033, respectively). Moreover, PD-L1 was expressed significantly more often in high grade tumors (41.5%) than in low grade tumors of patients (7.7%) (P = 0.040). The presence of CD3 or of FoxP3 infiltrating cells with PD-L1 in patient tumors did not impact the significance of the association of PD-L1 with high grade tumors (P = 0.040), and our analyses did not show an association between the presence of PD-1 or PD-L1 and survival. Conclusions: We conclude that a subgroup of advanced disease ovarian cancer patients with high grade tumors, expressing PD-L1, may be prime candidates for immunotherapy targeting PD-1 signaling. Keywords: Programmed cell death-1, Programmed cell death-1 ligand, High grade disease, Cancer immunotherapy, Ovarian cancer * Correspondence: mdrakes@luc.edu Cardinal Bernardin Cancer Center, Oncology Research Institute, Department of Medicine, Loyola University Chicago, Bldg. 112, Room 232, 2160 South First Avenue, Maywood, IL 60153, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 2 of 11 Background Pembrolizumab (Keytruda®) is FDA approved for melan- The early signs of ovarian cancer are asymptomatic and oma and NSCLC, and a blocking anti-PD-L1 antibody Ate- thus approximately 75 % of cases are detected in the ad- zolizumab (Tecentriq®) is also FDA approved for vanced metastatic stages. Conventional management strat- unresectable bladder cancer and for NSCLC. Blockade of egies for advanced disease include cytoreductive surgery this pathway is particularly useful in patients as it is applic- and chemotherapy. Most current treatments are not cura- able to a wide range of cancers, and because it induces tive for patients with advanced disease and hence survival anti-tumor immune responses capable of targeting mutated for this category of patients is low [1]. It is estimated that proteins [20]. Importantly, treatment targeting PD-1 signal- in 2017 there will be 22,440 new cases of ovarian cancer ing has fewer high grade toxicities than other immunother- in the Unites States, and that 14,080 patients will die due apies [13, 21]. to this disease [2]. Approximately 80 % of patients diag- Medical centers are currently utilizing these agents in nosed with late stage ovarian cancer die within five years. ongoing clinical trials for various cancers, including To provide more effective treatment options for patients, ovarian cancer [7, 22, 23]. Initial reports of some trials several clinical trials are ongoing using novel single and show promising objective response rates (ORR) for the combination regimens to improve survival. For cancer ther- treatment of ovarian cancer with anti-PD-1 antibody apy, there have been several distinct landmarks in the de- nivolumab (ORR of 15%, n = 20 patients), and pembroli- velopment of new therapies and FDA approved treatments zumab (ORR 11.5%, n = 49), or an anti-PD-L1 antibody over the last decade [3]. However, even with the current avelumab (ORR 10%, n = 124) [3, 24]. Those responding treatment options a considerable number of patients are often had durable responses, suggesting that if we could not yet receiving adequate therapy for the management of identify the subgroup that might typically respond, we advanced stage ovarian cancer and other malignancies. could advance the therapeutic options in this subgroup The development and optimization of the use of novel of ovarian cancer patients. therapies such as immunotherapy, requires an in-depth un- PD-1 is primarily expressed on CD4+ and CD8+ T cells derstanding of specific target molecules and cellular inter- and is associated with T cell exhaustion [11, 12, 14]. PD- actions in tumors. Early efforts in immunotherapy can be L1 is expressed on many cell types including tumor cells traced to 1891, in which administration of intra-tumoral in- and macrophages, including those with an immunosup- jections of bacteria led to a shrinkage of patients’ tumor [4, pressive phenotype [12, 25, 26]. Ligation of PD-L1 on 5]. Since then, significant progress has been made in the tumor cells with PD-1 on T cells, for example, abrogates T field [6]. One of the recent highlights in novel treatment cell proliferation, diminishes T cell activation and leads to options for cancer has been the targeting of immune check- a predominance of a T helper 2 (Th2) cytokine tumor point inhibitory molecules [7–9]. Immune checkpoints are microenvironment, with a pro-tumor propensity. Anti- critically important in health and disease. They represent body blocking of PD-1 or PD-L1 restores T cell prolifera- co-signaling pathways which are either costimulatory or tive and cytotoxic functions, and induces a T-helper 1 coinhibitory. In the body, linkage of coinhibitory receptor (Th1) phenotype, thereby re-invigorating T cells, with and ligand suppresses T cell receptor signaling, and limits resulting potent anti-tumor capacity [14, 27, 28]. immune responses. Whereas this function of checkpoint in- The immune mechanisms of disease improvement with hibitory molecules is beneficial during resolution of infec- administration of checkpoint inhibitory molecules are not tion, or in the development of self tolerance to prevent well understood. Clinically, there is also ongoing debate autoimmune conditions [10–12], ligation of checkpoint in- over which patients will benefit from this therapy, whether hibitory molecules can be a powerful and unwanted mech- patients who respond initially will continue to show anism of immunosuppression in cancer [13–15]. Since the complete responses (CR) or partial responses (PR), and successful introduction and FDA approved use of an anti- whether patients’ tumors need to express PD-1 and/ or body targeting checkpoint inhibitory molecule cytotoxic T PD-L1 in abundance, to predict beneficial responses to lymphocyte associated-4 (CTLA-4) (Ipilimumab; Yervoy®) checkpoint inhibitory molecule blocking therapy targeting in patients with unresectable or metastatic melanoma in these molecules. At the present time, there are more ques- 2011 [16], this agent is now in use in over 40 countries. At- tions than answers. tention has more recently focused on another checkpoint As a study of the immune microenvironment of ovarian inhibitory molecule programmed cell death-1 (PD-1) and cancer patient tumors offers insight into the baseline im- its ligand programmed cell death-1 ligand (PD-L1) [17–19]. mune landscape associated with patient survival and tumor Antibodies inhibiting PD-1 and PD-L1 have recently been pathology, and implicates broader scope for targeting these FDA approved for the treatment of cancer. For example, molecules in combination studies with conventional therapy the agent nivolumab (Opdivo®) is approved for unresectable and with other novel therapies, we undertook these present or metastatic melanoma, non-small cell lung cancer investigations. We primarily selected advanced disease ovar- (NSCLC), Hodgkin’s lymphoma and renal cell carcinoma. ian cancer patients for study, since this group typically have Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 3 of 11 poor outcome with standard therapy, and our future goal in single patient on each slide for detection of PD-L1, PD-1, translational medicine is to address the need for novel alter- CD3 and CD8 by immunohistochemistry (IHC). For stain- native treatment options in this patient sector. We evaluated ing of FoxP3 on T cells, patient tissue arrays were con- the expression and localization of PD-1 and PD-L1 in a co- structed from the paraffin embedded blocks and adhered hort of ovarian cancer formalin fixed paraffin embedded on a total of 2 glass slides with a core of tissue from each of (FFPE) tumor sections, and investigated whether the relative 27 or 28 patients, as well as control tissues. Positive control expression levels of these molecules can be relevant patient thymus tissue highly expressed the molecules/ markers prognostic indicators. We also studied the impact of tumor under study. Negative control tissue was sections of benign infiltrating lymphocytes (TILS) along with these checkpoint ovarian disease such as polycystic ovarian disease. Sections molecules, on patient status including tumor grade, disease on slides were de-paraffinized in xylene and then rehy- stage and survival post diagnosis. drated in a series of decreasing concentrations of alcohols. Antigen retrieval for PD-L1 and PD-1 was performed by boiling slides in a pressure cooker for 5 min in Universal Methods HIER retrieval agent (ab 208,572, Abcam, Cambridge, MA) Patients at a 1X concentration. Sections were washed in 0.1% tween Patients underwent surgery between 2003 and 2006 at in Dulbecco’s phosphate buffered saline (DPBS; 1X, Lonza, Loyola University Medical Center (LUMC) for ovarian and Walkersville, MD) and then blocked in 0.4% hydrogen per- other gynecologic associated cancers. Tissues were embed- oxide in DPBS, followed by blocking in 10% goat serum ded in paraffin blocks for patient diagnosis to characterize (S1000, Vector Laboratories, Burlingame, CA) for 1 h. stage and grade of cancer in tissue sections, and blocks Antigen retrieval for FoxP3, CD3 and CD8 was performed were stored in the Department of Pathology, LUMC. After by boiling sections in a pressure cooker for 5 min in Reveal receiving approval by the Institutional Review Board (IRB) Decloaker (RV1000G1, Biocare Medical, Concord, CA). After for the Protection of Human Subjects, we selected a cohort washing in DPBS, sections were blocked in 0.4% hydrogen of 55 patients for study, most of whom were diagnosed with peroxide in DPBS for 20 min, 10% goat serum or 10% horse advanced disease ovarian cancer (Table 1). Patient histories serum (S1000 or S2000 respectively, Vector Laboratories) for in the LUMC medical records were evaluated by two inves- 20 min, and then in Avidin/ Biotin blocking reagents (SP tigators and data collected for parameters including: age, 2001, Vector Laboratories) to further reduce non-specific date of birth, date of diagnosis, pre-treatment status before staining of primary antibody (FoxP3, CD3, or CD8). surgery, cancer stage, tumor grade, date of last encounter, and whether the patient was alive or deceased. Dates of Identification and assessment of antigens in patient death were retrieved from the patients’ medical records sections when this date was available, or found by a search on a Tissue sections were incubated overnight in 5% blocking website such as http://www.dobsearch.com/death-records/. serum with or without primary antibody at a pre-determined and optimized dilution. PD-1 (ab137132, Abcam) and PD- Antigen revealing L1 (ab205921, Abcam) were used at 1:500 dilution for IHC Formalin fixed paraffin embedded (FFPE) tissue sections staining. The next day sections were washed in 0.1% tween (4 μm) were adhered to glass slides using tissue from a in DPBS, and an amplifier polymer detection system specific Table 1 Categories of patient tumors Stage Tumor type Histological subtype No. of patients Low (I/ II) All papillary mucinous cystadenocarcinoma 1 Ovarian mixed adenocarcinoma, serous and endometroid 1 papillary serous adenocarcinoma 2 papillary serous carcinoma 4 serous carcinoma 1 High (III/ IV) Ovarian poorly differentiated serous carcinoma 1 Ovarian papillary serous carcinoma 24 Ovarian papillary serous adenocarcinoma 10 Fallopian tube papillary serous carcinoma 1 Peritoneum papillary serous carcinoma 4 Peritoneum papillary serous adenocarcinoma 1 Endometrium papillary serous carcinoma 1 Endometrium serous carcinoma 1 Omentum papillary serous carcinoma 2 Total 55 Tumor stage reflects the International Federation of Gynecology and Obstetrics (FIGO) classification Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 4 of 11 for rabbit anti-human primary antibodies (ab 20,901, Rabbit with Log Rank tests. The frequency of occurrence of each specific IHC polymer detection kit; HRP/ DAB) added ac- marker was graded on a scale from 0 to 4 (Table 2), and cording to the manufacturers’ guidance.Tissue was also Cochran Armitage tests used to determine the statistical stained overnight with primary antibodies for FoxP3 (236A/ significance of trends by patient characteristics including E7; ab 20,034, 1:1600 dilution, Abcam), CD8 (C8/144B; 1: age, cancer stage and tumor grade. Associations between 100 dilution, Cell Marque, Rocklin, CA 1:1000 dilution) and the presence of PD-1, PD-L1, CD3, CD8 and FoxP3 posi- CD3 (F7.2.38; 1:1000 dilution, Dako, Glostrup, Denmark). tive cells with patient age at the time of diagnosis, cancer Sections were washed in DPBS and a biotinylated secondary stage, or tumor grade were determined with chi-square or antibody for peroxidase (PK 6102, Vector Laboratories) Fisher’s exact tests as appropriate. Hazard ratios for overall added for 30 min, followed by an avidin-biotin peroxidase survival (O/S) were determined from univariable Cox pro- complex and enzyme reagent (ABC, Vector laboratories). All portional hazard regression models for each patient charac- sections were washed in DPBS and developed in Vector teristic and each marker. Analyses were performed using NovaRED (SK4800) or diaminobenzidine (DAB; SK4100, SAS 9.4 (SAS Institute, Cary, NC). Vector Laboratories). Sections were counterstained in hematoxylin and rehydrated in xylene, followed by alcohol, Results then mounted in Vectamount H-5000 (Vector Laboratories). Patient characteristics TumorsectionswereexaminedbypathologistsSMand The cohort consisted of 55 patients as follows: ovarian (45 MA to investigate the frequency of occurrence of markers, patients), fallopian tube (1 patient), peritoneum (5 pa- the degree of staining intensity and location of tumor cells or tients), endometrium (2 patients) and omentum (2 pa- lymphocytes expressing each molecule. A customized scor- tients) cancer tissue blocks. Patients were diagnosed as ing system was developed by the abovementioned patholo- stage I or stage II (stage 1 or 2, low/ early stage; 9 patients) gists to obtain a numerical score to represent the average and stage III or IV (stage 3 or 4, high stage or advanced; frequency of antigens as visualized over 7–10 high power 46 patients) disease (Table 1). One patient did not have fields (hpf) of IHC stained tissue sections (Table 2), where known cancer grade, and of the remaining, about three “0” was the lack of expression, and “4” represented the high- quarters of patients had tumor grade 3 (n = 41, 75.9%).The est frequency of expression of molecules in sections. In mean age of patients at the time of diagnosis was 61 years addition to the scoring pattern shown in Table 2, in statistical (standard deviation = 12), with a range of 26 to 85 years. analysis, combined PD-1 was assigned as a mathematical At the time of last follow up, 6 patients were alive, 9 pa- score which was derived by adding the observed pathology tients were lost to follow up (3 of whom were deceased at scores (0–2) for T-PD-1 and S-PD-1 in each patient section. an unknown date), and 40 other patients had recorded In some statistical analysis PD-1 and PD-L1 expression dates of death, 38 of whom had died as a cause of ovarian was classified as low (score of 1) or high frequency (score cancer. The median length of survival time for all patients of 2–4) to decipher correlations between the levels of ex- was 3.10 years (95% confidence interval (CI): 2.24–5.19). pression of these molecules and parameters studied. Localization of checkpoint inhibitory molecules in ovarian Statistical analysis cancer tissue Patient O/S was displayed visually in Kaplan Meier plots We studied the distribution of PD-1 and PD-L1 in ovarian and significance of differences by strata were determined cancer tissue sections by IHC staining. Sections were visu- alized and staining evaluated by methods as outlined in Table 2 Pathological interpretation of IHC stained tissue Table 2. Results showed that PD-1 in ovarian cancer tissue Marker Scoring criteria based on cells/ high power fields was localized primarily to cell membranes. PD-1 stained cells appeared to be primarily tumor infiltrating lympho- T-PD-1 0 = < 1; 1 = 1–10, 2= > 10–50 and 3= > 50 cytes (TILs), with varying degrees of intensity of staining S-PD-1 0 = < 1, 1 = 1–25; 2= > 25–50 and 3= > 50 and frequency of occurrence of these cells. Additionally, PD-L1 0 = < 1, 1= > 1–5, 2= > 5–10 and 3= > 10 this staining was strikingly evident in two separate com- CD3 0= <5, 1 =5–15, 2= > 15–25, 3= > 25–40 and 4= > 40 partments of the tumor microenvironment, thus we scored CD8 0 = < 1, 1 = 1–25, 2= > 25–50 and 3= > 50 this stain in two different categories. Intraepithelial TILs FoxP3 0 = < 1, 1 = 1–5, 2= > 5–15, 3= > 15–25 and 4= > 25 cells/ hpf completely enclosed by tumor epithelial cells and positive Based on the known frequency of each marker in various cancer tissues, each for PD-1 were designated as tumor PD-1 cells (T-PD-1). stained section was observed over 7–10 hpf and given an average numerical These cells were either clustered (Fig. 1a), or scattered in a value to represent the scoring frequency based on the categorization in Table less dense pattern in the tumor epithelium (Fig. 1b). Cells 2. PD-1 was assigned 2 different scores to reflect the observed localization of this molecule in tumor tissue. The percentage of PD-1 stained TILs completely which stained positive for PD-1 in the stromal compart- enclosed by tumor epithelial cells was termed tumor-PD-1 (T-PD-1). PD-1 ment were termed stromal PD-1 (S-PD-1). In the stroma, stained lymphocytes in the stromal compartment were classified as stromal-PD-1 (S-PD-1) there were aggregates of cells staining positive for PD-1 Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 5 of 11 Fig. 2 Membranous staining of PD-L1 positive cells in ovarian tumors. Diffuse membranous staining of almost all ovarian cancer cells (a) Fig. 1 Distinct IHC staining patterns of PD-1 positive cells in the versus focal strong membranous staining in a few scattered tumor cells stroma and epithelium of tissue. In ovarian cancer tissue sections, positive for PDL1 (b) tumor infiltrating lymphocytes (TILS) exhibiting strong membranous and cytoplasmic staining for PD-1 are apparent as clusters (a)or single scattered cells (b). TILS are also seen in aggregates (c)and as single Variable expression of tumor infiltrating lymphocytes in cells (d) in the reactive stroma associated with tumor ovarian cancer Immunocompetent TILS and FoxP3 T regulatory cells, are (Fig. 1c), or regions of fewer cells staining for this molecule strategically located in ovarian cancer tissue. We sought to (Fig. 1d). Table 3 shows a summary of scores for each PD-1 determine whether the levels of CD3, CD8 and/ or FoxP3 T classification as well as the frequency of PD-1 observations cells in the tumors of patients would alter the potential role in tissue in the cohort of 55 patients studied. A total of 48 of checkpoint inhibitory molecules as predictors of disease (87%) patients expressed PD-1 on cells, whilst 40 (73%) pa- pathology or of outcome. In FFPE tissue sections, of all T tients expressed PD-1 in both the epithelial (T-PD-1) and cells, CD3 positive T cells were found in highest frequency, as stromal (S-PD-1) compartments (data not shown). expected (Table 3), sometimes staining intensely. Some pa- In tissue sections, PD-L1 was localized to the cell mem- tients had an abundance of these cells, while other cases brane. In lesions of some patients there were solid tumor ag- showed a scattered arrangement (Fig. 3a and b respectively). gregates with diffuse membranous staining of these cells for Staining for CD8 T cells was widely observed in patient sec- this marker (Fig. 2a). PD-L1 was also identified in scattered tions as membranous reactivity, either with a dense distribu- tumor cells with focal strong membranous staining (Fig. 2b). tion, or with lower frequency (Fig. 3c and d respectively). Additionally, in some sections PD-L1 staining was observed in cells with the morphology of immune cells (not shown). Staining for this marker was positive in approximately one- third (18 of 55, 33%) of patients with ovarian cancer. Table 3 Occurrence of checkpoint inhibitory molecules and TILS in tumors Molecules in Frequency Total patients tumor sections positive T-PD-1 10 35 10 0 0 45 S-PD-1 12 39 4 0 0 43 PD-L1 37 14 3 1 0 18 CD3 2 23185 753 CD8 17 15 18 5 0 38 Fig. 3 High and low density of TILs in ovarian tumors. IHC staining FoxP3 7 12 15 19 2 48 of T cell subsets in patients’ FFPE tissue sections. CD3 exhibiting diffuse strong staining in clusters of tumor infiltrating lymphocytes Tumor sections of 55 ovarian and related cancers were IHC stained for molecules as indicated. Sections were visualized by microscopy and a (a) versus focal staining in scattered TILs in less dense areas (b). pathological score assigned from 0 to 4, based on the frequency of expression Distribution of CD8 T cells in clusters and as single cells is apparent of each molecule, where “0” indicates the absence and “4” the highest in c and d respectively frequency of expression Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 6 of 11 FoxP3 staining of cells was revealed as strong nuclear reactiv- The presence of PD-L1 is positively associated with high ity, either in clustered foci in some patient sections (Fig. 4a), grade tumors or scattered in the tumor stroma in other cases (Fig. 4b). We undertook detailed analysis to determine how the pres- FoxP3 positive cells were present in 48 of 55 patients. ence of checkpoint inhibitor molecules, PD-1 or PD-L1 correlated with patient history. We did not find any signifi- cant associations between the presence of PD-1, with pa- Clinical parameters as predictors of patient prognosis tient age at the time of diagnosis, stage of cancer or tumor We examined trends between age of patients at the time grade (Table 4). of diagnosis, tumor grade or tumor stage with overall sur- Only one patient with low tumor grade expressed PD- vival. Significant parameters for these associations are rep- L1, thus 17/18 (94%) patients who expressed PD-L1 in resented in Kaplan Meier plots (Additional file 1:Figure their ovarian tumors had a tumor grade of 3. Thus PD- S1). We found that patients diagnosed with ovarian cancer L1 was expressed significantly more often in patients when over 60 years of age had a higher hazard of death with high grade tumor than in those with low grade (HR: 2.63, 95% CI: 1.34–5.16) and were significantly more tumor (n = 54; 41.5% versus 7.7%, P = 0.040) (Table 4). likely to die as a cause of disease than patients diag- We did not observe any other significant associations nosed at a younger age (P = 0.005), (Additional file 2: with the presence of any other molecules studied as sin- Table S1). Additionally, patients diagnosed with ad- gle predictors, with age at the time of diagnosis, stage of vanced disease (stage III or IV) had a higher hazard of disease or tumor pathology. death (HR: 3.70, 95% CI: 1.28–10.76) and were signifi- Further analysis of data showed that the presence of cantly more likely to die than patients diagnosed in CD3 or of FoxP3 infiltrating cells together with PD-L1 the early stages of the disease (P = 0.016). Tumor grade in patient tumors did not impact the significance of the had no significant association with patient survival association of PD-L1 with high grade tumors (P = 0.040) (Additional file 2:Table S1). (Table 5). There was borderline significant association Similar analysis was performed to determine survival between the presence of S-PD-1 cells together with estimates as a function of the expression of each marker. FoxP3 positive cells in tumors with high grade It was found that while survival was similar in the first (P = 0.075). The presence of TILs and of cells expressing few years of follow-up for those with or without PD-L1, PD-1 or PD-L1 occurring together in ovarian tumors those with PD-L1 present in tissue sections appeared to had no impact on disease stage (Table 5), or on survival have a survival advantage with increased time after diag- (data not shown). nosis, even though this trend was not statistically signifi- cant (Additional file 3: Figure S2A). The presence of PD- 1 and of CD3 showed modest but insignificant trends to- Low frequency of expression of PD-1 and of PD-L1 ward improved survival (Additional file 3: Figure S2 B correlates with advanced ovarian cancer and C). CD8 or FoxP3 did not show a significant associ- We further investigated whether high or low frequency of ation with survival, even though for high expression of expression of PD-1 correlated with the stage or grade of FoxP3 there was a trend towards decreased survival cancer. Based on our scoring pattern for the occurrence of (data not shown). Table 4 Patient history and presence of PD-1 and PD-L1 combined PD-1 PD-L1 No. Patients n (%) p-value n (%) P-value Age at diagnosis < 60 27 24 (88.9) 0.99 9 (33.3) 0.93 ≥ 60 28 24 (85.7) 9 (32.1) Stage of cancer I-II 9 9 (100.0) 0.58 5 (55.6) 0.13 III-IV 46 39 (84.8) 13 (28.3) Tumor grade 1–2 13 12 (92.3) 0.99 1 (7.7) 0.040 3 41 35 (85.4) 17 (41.5) Fig. 4 Distribution of FoxP3 expressing T regulatory cells in ovarian tumors. Intense nuclear staining of FoxP3 positive lymphocytes in a The presence of PD-1 in the tumor epithelial or stromal compartment (combined PD-1), and PD-L1 was defined as the occurrence of these molecules focal arrangement (a). Other cases (b) showed nuclear staining of (scores 1, 2, 3, or 4) in tissue sections. Tumor grade was unknown for one fewer cells positive for this marker patient. Significant P-values are indicated in italics Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 7 of 11 Table 5 Association with immune markers, grade and stage No. Patients n (%) with stage III-IV P-value n (%) with grade 3 P-value Overall 55 46 (83.6) 41/54 (75.9) Presence with CD3 T-PD-1 + CD3 44 35 (79.5) 0.18 31 (72.1) 0.26 S-PD-1 + CD3 43 34 (79.1) 0.18 34 (81.0) 0.13 Combined PD-1 + CD3 47 38 (80.9) 0.33 34 (73.9) 0.66 PD-L1 + CD3 18 13 (72.2) 0.13 17 (94.4) 0.040 Presence with FoxP3 T-PD-1 + FoxP3 41 33 (80.5) 0.42 30 (75.0) 0.99 S-PD-1 + FoxP3 41 33 (80.5) 0.42 33 (82.5) 0.075 Combined PD-1 + FoxP3 44 36 (81.8) 0.67 33 (76.7) 0.99 PD-L1 + FoxP3 18 13 (72.2) 0.13 17 (94.4) 0.040 Evaluation of the association of the presence of PD-1 or PD-L1 together with CD3 and FoxP3 T cells on advanced disease and on high grade tumors. Significant P-values are indicated in italics PD-1 in the tumor epithelium (T-PD-1) or in the stroma or moderate staining for PD-L1 in either stromal or tumor (S-PD-1), we initially analyzed observations in these com- epithelial compartment, and an average of cells stained over partments separately. As expected, the frequency of PD-1 7 hpfs taken into account. Most patients (5/9; 56%) with expressing cells in tumor tissue was in general not as high early stage disease expressed PD-L1. A lower percentage of as those expressing CD3 or CD8 (Table 3). A higher per- patients with advanced disease expressed PD-L1 (13/46; centage of patients with early stage cancer were more 28%) (Table 7). Most patients (11/13; 85%) with advanced likely to have a higher frequency of expression (pathology disease who expressed PD-L1, expressed this molecule in score of2)ofT-PD-1 thaninpatients withadvanced tumors with a score of 1 (on a scale of 0–3). A higher per- disease, but this difference was not significant (P =0.13, centage of patients with early stage disease (22%) expressed data not shown). In the case of S-PD-1, a higher percent- highlevels of PD-L1(frequencyscore 2or3)incomparison age of patients with early stage cancer had higher levels with patients with advanced disease (4.3%) (Table 7). There (frequency score of 2, 22%) of S-PD-1 than for advanced were no significant trends in frequencies of CD3, CD8 or disease patients (frequency score of 2, 2%) (P =0.033) FoxP3 with age at the time of diagnosis, stage of disease, or (Additional file 4: Table S2). tumor grade (data not shown). A significantly higher percentage (33%) of early stage patients had a higher frequency (score 3 and 4) of com- bined PD-1 (PD-1 in the tumor epithelium and/or stroma) Discussion than those with advanced disease (17%) (P =0.028, Ovarian cancer is usually diagnosed in the advanced meta- Table 6). Thus, a high density of PD-1 was significantly as- static stages. Treatment of advanced stage disease with sociated with early stage disease diagnosis, and a low PD-1 conventional therapies is only sufficiently effective in a density was associated with advanced disease. limited number of patients, thus in about 80 % of these PD-L1 was expressed in 33% of the patient cohort. Scor- patients there is disease progression or disease recurrence ing for PD-L1 was performed by scoring regions of intense and death, within five years of diagnosis. In many cancers, Table 6 Patient tumor and combined frequency of PD-1 expression No. Patients 0 1 2 3 4 P-value Combined PD-1 Level, n (%) Stage of cancer I-II 9 0 (0.0) 0 (0.0) 6 (66.7) 1 (11.1) 2 (22.2) 0.028 III-IV 46 7 (15.2) 8 (17.4) 23 (50.0) 7 (15.2) 1 (2.2) Tumor grade 1–2 13 1 (7.7) 4 (30.8) 7 (53.8) 0 (0.0) 1 (7.7) 0.54 3 41 6 (14.6) 4 (9.8) 21 (51.2) 8 (19.5) 2 (4.9) Combined PD-1 was derived by adding the observed pathology scores for tumor PD-1 (T-PD-1) and S-PD-1 in each patient tumor section. Significant P-value in table is indicated in italics Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 8 of 11 Table 7 Patient tumor characteristics and frequency of PD-L1 expression No. Patients 0 1 2 3 P-value PD-L1 Level, n (%) Stage of cancer I-II 9 4 (44.4) 3 (33.3) 1 (11.1) 1 (11.1) 0.033 III-IV 46 33 (71.7) 11 (23.9) 2 (4.3) 0 (0.0) Tumor grade 1–2 13 12 (92.3) 1 (7.7) 0 (0.0) 0 (0.0) 0.033 3 41 24 (58.5) 13 (31.7) 3 (7.3) 1 (2.4) Significant P- values in table are indicated in italics investigators are focusing on the development of novel and TILS in the tumors of ovarian or related cancer pa- therapies as alternative and more robust options to exist- tients, diagnosed with advanced disease. Firstly, results ing therapies. Whereas conventional therapies primarily showed that patients’ age was an independent prognostic focus on the destruction of tumor cells, many novel ther- factor in survival, with patients over 60 years of age apies are designed to stimulate immune cells to elaborate more likely to die than those diagnosed when younger augmented anti-tumor immune responses. In this respect, than 60 years. This may be due to the fact that younger checkpoint immune inhibitory molecules have come full patients can tolerate more aggressive surgery and circle over the last decade for cancer immunotherapy. chemotherapy than older patients. Additionally, as ex- In a normal functioning immune system, T cell pected, disease stage was also an independent prognostic activating and inhibitory receptors balance immune factor in outcome, such that patients diagnosed with ad- tolerance, and the amplification of immune responses. vanced disease had a lower median survival than pa- In the body, immune checkpoints are designed to reduce tients diagnosed with early stage disease. These findings autoimmune responses, or to attenuate immune re- are in agreement with those of other investigators [32]. sponses which were elaborated after infections [10–12, To shed lightonthe relevanceofPD-1and PD-L1in 14, 27, 28]. In cancer, blocking of immune checkpoint ovarian cancer outcome, we studied the abundance of these molecules with antibodies is a novel and promising ther- molecules in the TME. PD-1 was compartmentalized in the apy, as it potentiates anti-tumor immune responses in stroma and in the tumor epithelium, and this molecule was patients [9, 13, 15, 19]. expressed in 87% of tumors. PD-L1 was only present in the The first of these checkpoint inhibitory molecules to tumors of 33% of patients. Patients who expressed PD-L1 be targeted for blocking in therapy, and is now FDA ap- had a trend towards survival, as did those expressing PD-1 proved for cancer therapy, is CTLA-4 [16, 17]. Therapy or CD3, even though these trends were not significant. In design is based on the following principle. Binding of our cohort we did not find a significant association with costimulatory molecules CD80 or CD86 on antigen pre- FoxP3 and survival. The presence of PD-L1 and FoxP3 to- senting cells to CD28 on T cells delivers a positive costi- gether in high grade tumors showed the same level of asso- mulatory signal contributing to T cell activation. On the ciation as the presence of PD-L1 alone. Some studies have contrary, linkage of CD80 or CD86 to CTLA-4 (a mol- reported that FoxP3 positive cells in ovarian tumors is ecule closely related to CD28) results in inhibition of im- negatively associated with outcome, however a meta- mune responses, and exhausted T cells, which are less analysis of 7 ovarian cancer studies with a total of 869 pa- able to proliferate or to secrete T helper 1 (Th1) cyto- tients, did not find FoxP3 TILS in ovarian cancer to be a kines [29] . Blocking of this inhibitory pathway with significant prognostic indicator [33]. anti-CTLA-4 antibodies, results in re-invigorated T cells In cancer there are conflicting reports concerning the with greater proliferative function, and durable anti- expression patterns of PD-1 in patients’ tumors and the tumor potency [30, 31]. By similar rationale, there are association with survival, with either positive or a nega- several other immune checkpoints in the immune tive association [34–36]. One recent report found that system, which are now being targeted in cancer clinical PD-1 positive TILS and /or PD-L1 positive tumor cells trials. In the case of blocking PD-1 signaling, there are had a positive association with survival of ovarian cancer currently about nine different antibodies in cancer patients [37]. clinical trials targeting the PD-1/ PD-L1 pathway [3, 17]. The expression of PD-L1 in tumors was shown to be This study was designed to better understand the ovar- positively associated with survival in NSCLC [38] and in ian cancer tumor microenvironment (TME) with rela- ovarian cancer [37]. On the contrary, others report a tion to the localization and frequency of PD-1, PD-L1 negative prognostic impact of PD-L1 expression in Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 9 of 11 ovarian cancer [39, 40] and breast cancer [41]. In a re- tumors. Patients then undergo courses of standard ther- view and meta-analysis of 17 studies using data of 2869 apy such as chemotherapy. Consistent with this practice, head and neck cancer (HNC) patients, authors found we found that a study of our patient cohort treatment that there was no significant association between the ex- plans revealed that only 3 of 55 (5.5%) patients had pression of PD-L1 on survival in HNC patients [42]. chemotherapy in the interval before surgery. Additionally, similar analysis of reports with NSCLC pa- A low frequency of PD-1 in tumors was associated tients also did not show a significant association with with advanced disease. This association between low PD-L1 expression and survival [43]. Taken together, this PD-1 density and advanced disease was only significant indicates that the prognostic impact of PD-1/ PD-L1 ex- when measuring S-PD-1 or combined PD-1, whereas pression in tumors is not yet established. Here, our stud- low T-PD-1 density alone was not associated with ad- ies did not find a significant association between survival vanced disease. Although beyond the scope of this study, and PD-1 or PD-L1 expression in ovarian cancer. this finding raises the possibility that T-PD-1 and S-PD- Differences in reports of the expression of these mole- 1 positive cells may perform unique immunosuppressive cules and associations with survival may be attributed to roles in the ovarian TME. several reasons. Firstly, it is possible that there may be PD-L1 expression was almost exclusively restricted to different survival outcomes due to the site of cancer. For high grade tumors, such that there was a positive and sig- example, Paulsen and colleagues [38] found that whereas nificant association between PD-L1 and high grade tumors. in a cohort of patients a high density of PD-1 and PD-L1 This finding may be of translational significance in selecting had a favorable impact on NSCLC, this association was patients for therapy blocking PD-1/ PD-L1 signaling, and not present when these molecules were studied at meta- we suggest that patients with high grade tumors, with pre- static sites such as lymph nodes of the same patients. existing PD-L1 expression might be excellent candidates This is highly likely because each cancer site has a differ- for therapy blocking this pathway. In support of this idea, a ent immune landscape, and levels of soluble molecules recent report shows that in an ongoing study of urothelial such as IFN-γ which is a strong regulator of PD-L1 ex- bladder cancer patients, treatment with durvalumab pression [44, 45]. (MED14736; an anti-PD-L1 antibody) resulted in improved Secondly, differences in reports may be due to varia- outcome in PD-L1 tumor positive patients. In pre- tions in staining protocols across individual laboratories. treatment tumor biopsies patients, 40 patients were PD-L1 Many investigators report difficulty in IHC staining for positive and 21 patients negative for PD-L1. In 42 evaluable PD-L1. In these present investigations, we initially used patients, the ORR was 31.0% (95% CI, 17.6 to 47.1), the tumor arrays to study the expression of PD-L1 and PD-1 ORR was 46.4% (95%CI, 27.5 to 66.1) in the PD-L1 positive in tumors. When optimizing our staining protocol, we patient subgroup, and 0% (95% CI, 0.0 to 23.2) in patients found that it was difficult to select cores which had a negative for PD-L1 [46]. good representation of tumor and stromal areas for ac- We suggest that due to the conflicting reports concerning curate visualization of the density of these molecules in the impact of PD-1 and PD-L1 on survival in cancer pa- tissue arrays. Therefore, in this study we used whole tients, a future larger study is needed investigating these tumor sections for the identification of these molecules, molecules in ovarian tissue, with standardized protocols as is done for patient diagnosis. For molecules such as and defined cut off points for positive staining and scoring PD-L1 especially, which is not widely expressed in criteria across centers, to minimize study variations. Even tumor tissues, reports in which IHC staining was per- so, the potency of patient responses to PD-1/ PD-L1 block- formed using tumor cores may give different findings ingantibodytherapy maybeinfluenced by thedensity of than reports from other laboratories in which staining other pre-existing or emerging checkpoint molecules in tu- was performed using whole tissue sections. mors, including T-cell immunoglobulin and mucin-domain Thirdly, the use of different primary antibody clones to containing-3 (TIM-3), lymphocyte-activation gene 3 (LAG- identify PD-1 or PD-L1 in tumor tissue in each laboratory, 3) and V-domain Ig suppressor of T cell activation (VISTA). may also lead to variability in staining of sections and in Other parameters such as the presence of myeloid derived interpretation. Finally, manual staining protocols in com- suppressor cells, levels of Th2 cytokines (for example IL- parison with automated staining may further contribute to 10) and of indoleamine 2,3-dioxygenase (IDO) can also variations in staining interpretation. limit anti-cancer immune responses to therapy blocking Finally, additional parameters which may alter the ex- PD-1/ PD-L1 [19, 47, 48]. Furthermore, genetic alterations pression of these molecules in tumors, is the administra- within the tumor (including DNA rearrangements, muta- tion of treatments such as chemotherapy to patients tions, deletions and insertions) alter tumor mutational prior to surgery. However, at our center, for patients di- loads, and it is reported that tumors with high mutational agnosed with ovarian and related gynecologic cancer, the loads have the greatest response to checkpoint inhibitory primary treatment is most often surgery for removal of blockade therapy [17, 49, 50]. Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 10 of 11 Finally, due to the multiplicity of factors regulating Authors’ contributions MLD Conceptualization, investigation, patient history data, data analysis, ORR, we believe that antibody therapy targeting the PD- writing- original draft, review and editing, and project administration. SM 1/ PD-L1 pathway in ovarian cancer will be of maximum Pathology grading and scoring, writing- review and editing. MA Pathology efficacy when used in combination with other treatment grading and scoring, patient history data, writing- review and editing. RP Patient surgery, writing- review and editing. YL Investigation, writing- review. regimens. Such treatments include standard therapy, im- AG Investigation, writing- review. CJ Statistical analysis. Writing- statistical munotherapy blocking other checkpoint inhibitory mol- writing, review and editing. TO Statistical analysis, writing- review. MSS ecules, dendritic cell vaccines, chimeric antigen receptor Conceptualization, investigation, data analysis, writing- review and editing, and project administration. PJS Investigation, writing- review and editing, and (CAR) T cell therapy, or targeted therapy, all of which project administration. All authors read and approved the final manuscript. can downregulate other immune suppressive mecha- nisms in patients, concomitantly. Ethics approval and consent to participate Investigators received approval by the Institutional Review Board (IRB) for the Protection of Human Subjects at Loyola University Medical Center for the use of the cohort of patients in study. Conclusions Treatments inhibiting PD-1 and PD-L1 are beneficial only Competing interests in some ovarian cancer patients. Our studies reveal that a The authors declare that they have no competing interests. low density of PD-1 and of PD-L1 expressing cells in tumor tissue are significantly associated with advanced disease, Publisher’sNote and that PD-L1 is expressed significantly more in high Springer Nature remains neutral with regard to jurisdictional claims in grade tumors than in low grade tumors. We conclude that published maps and institutional affiliations. a subgroup of advanced disease ovarian cancer patients with Author details high grade tumors, bearing PD-L1, may be the best candi- 1 Cardinal Bernardin Cancer Center, Oncology Research Institute, Department dates for immunotherapy targeting PD-1 and/ or PD-L1. of Medicine, Loyola University Chicago, Bldg. 112, Room 232, 2160 South First Avenue, Maywood, IL 60153, USA. Department of Pathology, Stritch School of Medicine, Loyola University Chicago, Maywood, IL, USA. Department of Obstetrics and Gynecology, Loyola University Chicago, Maywood, IL, USA. Department of Chemistry & Biochemistry, Harper Cancer Additional files Research Institute, University of Notre Dame, South Bend, IN, USA. Department of Public Health Sciences, Loyola University Chicago, Maywood, Additional file 1: Figure S1. Survival estimates by patient age and 6 IL, USA. Department of Mathematics & Statistics, and Institute of tumor grade. Kaplan Meier survival analysis to estimate overall survival in Environmental Sustainability, Loyola University Chicago, Chicago, IL, USA. patients as a function of age (A) or tumor stage (B). Patient survival was displayed visually in Kaplan Meier plots and significance of differences Received: 5 January 2018 Accepted: 9 May 2018 determined with Log Rank tests. (PPTX 427 kb) Additional file 2: Table S1. Univariable hazard ratios for Cox proportional hazards models (DOCX 25 kb) References Additional file 3: Figure S2. Survival estimates by expression of immune 1. Luvero D, Milani A, Ledermann JA. Treatment options in recurrent ovarian molecules. Kaplan Meier survival analysis to estimate overall survival in patients cancer: latest evidence and clinical potential. Ther Adv Med Oncol. 2014; expressing PD-L1 (A), PD-1 (B) and CD3 (C) in tumor sections. (PPTX 456 kb) 6(5):229–39. 2. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin. 2017; Additional file 4: Table S2. Patient tumor and frequency of S-PD-1 67(1):7–30. expression (DOCX 25 kb) 3. Iwai Y, Hamanishi J, Chamoto K, Honjo T. Cancer immunotherapies targeting the PD-1 signaling pathway. J Biomed Sci. 2017;24(1):1–11. 4. McCarthy EF. The toxins of William B. Coley and the treatment of bone and Abbreviations soft-tissue sarcomas. The Iowa orthopaedic journal. 2006;26:154–8. CR: complete response; CTLA-4: cytotoxic T lymphocyte associated-4; 5. Nauts HC, Swift WE, Coley BL. The treatment of malignant tumors by FFPE: formalin fixed paraffin embedded; HNC: head and neck cancer; bacterial toxins as developed by the late William B. Coley, M.D., reviewed in hpf: high power fields; IHC: immunohistochemistry; NSCLC: non-small cell the light of modern research. Cancer Res. 1946;6:205–16. lung cancer; O/S: overall survival; ORR: objective response rate; PD- 6. Chester C, Dorigo O, Berek JS, Kohrt H. Immunotherapeutic approaches to 1: programmed cell death-1; PD-L1: programmed cell death-1 ligand; ovarian cancer treatment. Journal for immunotherapy of cancer. 2015;3:1–10. PR: partial response; S-PD-1: stromal PD-1; Th1: T helper-1; TILS: tumor 7. Hamanishi J, Mandai M, Konishi I. Immune checkpoint inhibition in ovarian infiltrating lymphocytes; TME: tumor microenvironment; T-PD-1: Tumor PD-1 cancer. Int Immunol. 2016;28(7):339–48. 8. Hodi FS, Butler M, Oble DA, Seiden MV, Haluska FG, Kruse A, Macrae S, Nelson M, Canning C, Lowy I, et al. Immunologic and clinical effects of Funding antibody blockade of cytotoxic T lymphocyte-associated antigen 4 in This research was supported by a joint research award offered by the previously vaccinated cancer patients. Proc Natl Acad Sci U S A. 2008;105(8): Cardinal Bernardin Cancer Center, Loyola University (to M.L.D.), and the 3005–10. Harper Cancer Research Institute, University of Notre Dame (to M.S.S.), made 9. Pardoll DM. The blockade of immune checkpoints in cancer possible by funds from the Cronk Foundation. Additional support was from immunotherapy. Nat Rev Cancer. 2012;12(4):252–64. the Perritt Charitable Foundation (to P.J.S.). Funding agencies did not 10. Bayry J. Autoimmunity: CTLA-4: a key protein in autoimmunity. Nat Rev participate in the design of the study, collection, analysis or interpretation of Rheumatol. 2009;5(5):244–5. data, or in writing the manuscript. 11. Blank C, Mackensen A. Contribution of the PD-L1/PD-1 pathway to T-cell exhaustion: an update on implications for chronic infections and tumor Availability of data and materials evasion. Cancer Immunol Immunother. 2007;56(5):739–45. All data generated or analyzed during this study are included in this 12. Keir ME, Butte MJ, Freeman GJ, Sharpe AH. PD-1 and its ligands in tolerance published article (and its additional information files). and immunity. Annu Rev Immunol. 2008;26:677–704. Drakes et al. Journal of Ovarian Research (2018) 11:43 Page 11 of 11 13. Brahmer J, Reckamp KL, Baas P, Crino L, Eberhardt WE, Poddubskaya E, 35. Schmidt LH, Kummel A, Gorlich D, Mohr M, Brockling S, Mikesch JH, Antonia S, Pluzanski A, Vokes EE, Holgado E, et al. Nivolumab versus Grunewald I, Marra A, Schultheis AM, Wardelmann E, et al. PD-1 and PD-L1 docetaxel in advanced squamous-cell non-small-cell lung Cancer. N Engl J expression in NSCLC indicate a favorable prognosis in defined subgroups. Med. 2015;373(2):123–35. PLoS One. 2015;10(8):1–15. 14. Flies DB, Sandler BJ, Sznol M, Chen L. Blockade of the B7-H1/PD-1 pathway 36. Thompson RH, Dong H, Lohse CM, Leibovich BC, Blute ML, Cheville JC, for cancer immunotherapy. Yale J biol Med. 2011;84(4):409–21. Kwon ED. PD-1 is expressed by tumor-infiltrating immune cells and is associated with poor outcome for patients with renal cell carcinoma. Clin 15. Larkin J, Chiarion-Sileni V, Gonzalez R, Grob JJ, Cowey CL, Lao CD, Cancer Res. 2007;13(6):1757–61. Schadendorf D, Dummer R, Smylie M, Rutkowski P, et al. Combined 37. Darb-Esfahani S, Kunze CA, Kulbe H, Sehouli J, Wienert S, Lindner J, Budczies Nivolumab and Ipilimumab or monotherapy in untreated melanoma. N J, Bockmayr M, Dietel M, Denkert C, et al. Prognostic impact of Engl J Med. 2015;373(1):23–34. programmed cell death-1 (PD-1) and PD-ligand 1 (PD-L1) expression in 16. Hodi FS, O'Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB, cancer cells and tumor-infiltrating lymphocytes in ovarian high grade Gonzalez R, Robert C, Schadendorf D, Hassel JC, et al. Improved survival serous carcinoma. Oncotarget. 2016;7(2):1486–99. with ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010; 38. Paulsen EE, Kilvaer TK, Khanehkenari MR, Al-Saad S, Hald SM, Andersen S, 363(8):711–23. Richardsen E, Ness N, Busund LT, Bremnes RM, et al. Assessing PDL-1 and 17. Gaillard SL, Secord AA, Monk B. The role of immune checkpoint inhibition in PD-1 in non-small cell lung Cancer: a novel Immunoscore approach. Clin the treatment of ovarian cancer. Gynecol Oncol Res Pract. 2016;3:1–14. Lung Cancer. 2017;18(2):220–33. e228 18. Okazaki T, Honjo T. PD-1 and PD-1 ligands: from discovery to clinical 39. Hamanishi J, Mandai M, Iwasaki M, Okazaki T, Tanaka Y, Yamaguchi K, application. Int Immunol. 2007;19(7):813–24. Higuchi T, Yagi H, Takakura K, Minato N, et al. Programmed cell death 1 19. Topalian SL, Drake CG, Pardoll DM. Immune checkpoint blockade: a ligand 1 and tumor-infiltrating CD8+ T lymphocytes are prognostic factors common denominator approach to cancer therapy. Cancer Cell. 2015;27(4): of human ovarian cancer. Proc Natl Acad Sci U S A. 2007;104(9):3360–5. 450–61. 40. Maine CJ, Aziz NH, Chatterjee J, Hayford C, Brewig N, Whilding L, George AJ, 20. Gubin MM, Zhang X, Schuster H, Caron E, Ward JP, Noguchi T, Ivanova Y, Ghaem-Maghami S. Programmed death ligand-1 over-expression correlates Hundal J, Arthur CD, Krebber WJ, et al. Checkpoint blockade cancer with malignancy and contributes to immune regulation in ovarian cancer. immunotherapy targets tumour-specific mutant antigens. Nature. 2014; Cancer immunology, immunotherapy : CII. 2014;63(3):215–24. 515(7528):577–81. 41. Muenst S, Soysal SD, Gao F, Obermann EC, Oertli D, Gillanders WE. The 21. Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, presence of programmed death 1 (PD-1)-positive tumor-infiltrating Powderly JD, Carvajal RD, Sosman JA, Atkins MB, et al. Safety, activity, and lymphocytes is associated with poor prognosis in human breast cancer. immune correlates of anti-PD-1 antibody in cancer. N Engl J Med. 2012; Breast Cancer Res Treat. 2013;139(3):667–76. 366(26):2443–54. 42. Li J, Wang P, Xu Y. Prognostic value of programmed cell death ligand 1 22. De Felice F, Marchetti C, Palaia I, Musio D, Muzii L, Tombolini V, Panici PB. expression in patients with head and neck cancer: a systematic review and Immunotherapy of ovarian Cancer: the role of checkpoint inhibitors. J meta-analysis. PLoS One. 2017;12(6):1–16. Immunol Res. 2015;2015:1–7. 43. Zhong A, Xing Y, Pan X, Shi M, Xu H. Prognostic value of programmed cell 23. Zhu X, Lang J. Programmed death-1 pathway blockade produces a death-ligand 1 expression in patients with non-small-cell lung cancer: evidence synergistic antitumor effect: combined application in ovarian cancer. J from an updated meta-analysis. OncoTargets and therapy. 2015;8:3595–601. Gynecol Oncol. 2017;28(5):1–19. 44. Benci JL, Xu B, Qiu Y, Wu TJ, Dada H, Twyman-Saint Victor C, Cucolo L, Lee 24. Hamanishi J, Mandai M, Ikeda T, Minami M, Kawaguchi A, Murayama T, DS, Pauken KE, Huang AC, et al. Tumor interferon signaling regulates a Kanai M, Mori Y, Matsumoto S, Chikuma S, et al. Safety and antitumor multigenic resistance program to immune checkpoint blockade. Cell. 2016; activity of anti-PD-1 antibody, Nivolumab, in patients with platinum- 167(6):1540–54. e1512 resistant ovarian Cancer. J Clin Oncol. 2015;33(34):4015–22. 45. Garcia-Diaz A,Shin DS, Moreno BH,SacoJ, Escuin-Ordinas H,Rodriguez GA, 25. Abiko K, Mandai M, Hamanishi J, Yoshioka Y, Matsumura N, Baba T, Zaretsky JM, Sun L, Hugo W, Wang X, et al. Interferon receptor signaling Yamaguchi K, Murakami R, Yamamoto A, Kharma B, et al. PD-L1 on tumor pathways regulating PD-L1 and PD-L2 expression. Cell Rep. 2017;19(6):1189–201. cells is induced in ascites and promotes peritoneal dissemination of ovarian 46. Massard C, Gordon MS, Sharma S, Rafii S, Wainberg ZA, Luke J, Curiel TJ, cancer through CTL dysfunction. Clin Cancer Res. 2013;19(6):1363–74. Colon-Otero G, Hamid O, Sanborn RE, et al. Safety and efficacy of 26. Ritprajak P, Azuma M. Intrinsic and extrinsic control of expression of the Durvalumab (MEDI4736), an anti-programmed cell death Ligand-1 immune immunoregulatory molecule PD-L1 in epithelial cells and squamous cell checkpoint inhibitor, in patients with advanced urothelial bladder Cancer. J carcinoma. Oral Oncol. 2015;51(3):221–8. Clin Oncol. 2016;34(26):3119–25. 27. Taube JM, Klein A, Brahmer JR, Xu H, Pan X, Kim JH, Chen L, Pardoll DM, 47. Koyama S, Akbay EA, Li YY, Herter-Sprie GS, Buczkowski KA, Richards WG, Topalian SL, Anders RA. Association of PD-1, PD-1 ligands, and other Gandhi L, Redig AJ, Rodig SJ, Asahina H, et al. Adaptive resistance to features of the tumor immune microenvironment with response to anti-PD- therapeutic PD-1 blockade is associated with upregulation of alternative 1 therapy. Clin Cancer Res. 2014;20(19):5064–74. immune checkpoints. Nat Commun. 2016;7:1–9. 28. Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJ, Robert L, Chmielowski B, 48. Yigit R, Massuger LF, Figdor CG, Torensma R. Ovarian cancer creates a Spasic M, Henry G, Ciobanu V, et al. PD-1 blockade induces responses by suppressive microenvironment to escape immune elimination. Gynecol inhibiting adaptive immune resistance. Nature. 2014;515(7528):568–71. Oncol. 2010;117(2):366–72. 29. Krummel MF, Allison JP. CD28 and CTLA-4 have opposing effects on the 49. Chalmers ZR, Connelly CF, Fabrizio D, Gay L, Ali SM, Ennis R, Schrock A, response of T cells to stimulation. J Exp Med. 1995;182(2):459–65. Campbell B, Shlien A, Chmielecki J, et al. Analysis of 100,000 human cancer 30. Buchbinder EI, Desai A. CTLA-4 and PD-1 pathways: similarities, differences, genomes reveals the landscape of tumor mutational burden. Genome Med. and implications of their inhibition. Am J Clin Oncol. 2016;39(1):98–106. 2017;9(1):1–14. 31. Peggs KS, Quezada SA, Chambers CA, Korman AJ, Allison JP. Blockade of CTLA- 50. Van Allen EM, Miao D, Schilling B, Shukla SA, Blank C, Zimmer L, Sucker A, 4 on both effector and regulatory T cell compartments contributes to the Hillen U, Foppen MHG, Goldinger SM, et al. Genomic correlates of response to antitumor activity of anti-CTLA-4 antibodies. J Exp Med. 2009;206(8):1717–25. CTLA-4 blockade in metastatic melanoma. Science. 2015;350(6257):207–11. 32. Deng F, Xu X, Lv M, Ren B, Wang Y, Guo W, Feng J, Chen X. Age is associated with prognosis in serous ovarian carcinoma. J Ovarian Research. 2017;10(36):1–9. 33. Shang B, Liu Y, Jiang SJ, Liu Y. Prognostic value of tumor-infiltrating FoxP3+ regulatory T cells in cancers: a systematic review and meta-analysis. Sci Rep. 2015;5:1–9. 34. Badoual C, Hans S, Merillon N, Van Ryswick C, Ravel P, Benhamouda N, Levionnois E, Nizard M, Si-Mohamed A, Besnier N, et al. PD-1-expressing tumor-infiltrating T cells are a favorable prognostic biomarker in HPV- associated head and neck cancer. Cancer Res. 2013;73(1):128–38.

Journal

Journal of Ovarian ResearchSpringer Journals

Published: May 30, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off