|Home | About | Journals | Submit | Contact Us | Français|
Mesothelin, a tumor differentiation antigen highly expressed in mesothelioma and ovarian cancer, is the receptor for CA-125 (MUC 16) and this interaction may play a role in tumor metastasis. MORAb-009 is a chimeric anti-mesothelin monoclonal antibody.
Twenty-four patients with mesothelin expressing cancers were treated on a phase I study of MORAb-009 administered as an intravenous infusion (12.5 mg/m2 –400 mg/m2) weekly × 4 doses with 2 weeks off before the next cycle. This report summarizes the effect of MORAb-009 on serum CA-125 kinetics in the eight patients with mesothelioma who had CA-125 levels measured before and at different time-points following therapy.
MORAb-009 treatment led to a marked increase in serum CA-125 levels in all patients including those without elevated CA-125 levels before therapy. The increase in CA-125 levels was not due to disease progression since CA-125 levels decreased rapidly after stopping MORAb-009 therapy. No patients had signs of peritoneal or pleural inflammation as the possible cause of CA-125 rise. In addition, the elevated CA-125 levels were not due to MORAb-009 interfering with the laboratory assay used to measure CA-125.
The increase in serum CA-125 produced by treatment with MORAb-009 is most likely due to MORAb-009 inhibiting the binding of tumor shed CA-125 to mesothelin present on mesothelial cells lining the pleural and peritoneal cavities. Inhibiting the mesothelin-CA-125 interaction could be a useful strategy to prevent tumor metastasis in mesotheliomas and ovarian cancer.
Mesothelin is a tumor differentiation antigen whose expression in normal human tissues is limited to mesothelial cells lining the pleura, pericardium and peritoneum [1,2]. Mesothelin is highly expressed in many human cancers, including virtually all epithelial mesotheliomas and pancreatic adenocarcinomas, and approximately 70% of ovarian cancers and 50% of lung adenocarcinomas [3–7]. The mesothelin gene encodes a precursor protein of 71 kDa that is processed to a 31 kDa shed protein called megakaryocyte potentiating factor and a 40 kDa fragment, mesothelin, that is attached to the cell membrane by a glycosyl-phosphatidylinositol anchor [2,8]. This expression pattern makes mesothelin an attractive target for cancer therapy and several agents targeting mesothelin are currently in clinical trials . In addition, some cell bound mesothelin is shed into the serum and elevated levels are present in many patients with mesothelioma and ovarian cancer [10,11].
The normal biologic function of mesothelin is unknown. Mutant mice in which both copies of the mesothelin gene were inactivated showed no detectable abnormalities as compared to wild-type littermates . The mesothelin gene is differentially regulated by members of the Wnt signal transduction pathway and in C57MG mouse mammary epithelial cells, mesothelin was up-regulated by Wnt-1 . It was originally suggested that mesothelin might have a role in adhesion, because 3T3 cells transfected with mesothelin were more difficult to remove from tissue culture plates than non-transfected cells . Recent studies have supported the hypothesis that mesothelin plays a role in cell adhesion by showing that it is the receptor for CA-125 (MUC 16), and this interaction between mesothelin and CA-125 leads to heterotypic adhesion [14,15]. CA-125, the ligand for mesothelin, is a cell surface glycoprotein that is present on normal mesothelial cells lining the body cavities [16,17]. Increased cell surface expression of CA-125 is seen in tumors such as ovarian cancer and mesothelioma as well as some other cancers [16,18–20]. It is also shed into the circulation and serum CA-125 is a commonly used test for monitoring disease progression in ovarian cancer and is also elevated in mesothelioma and some benign conditions [21–23]. The gene encoding the peptide moiety of CA-125 has been cloned and termed MUC16, because it shares characteristics associated with mucin proteins [24,25]. The finding of heterotypic adhesion through mesothelin-CA-125 high affinity interaction, suggests that mesothelin and/or CA-125 present on tumor cells can lead to intra-cavitary tumor metastasis by binding to their respective ligands on the mesothelial cells lining the pleura or peritoneum [14,15].
MORAb-009 is a high affinity chimeric (mouse/human) monoclonal IgG1/κ which was obtained by attaching the heavy and light chain variable regions of a mouse anti-mesothelin single chain Fv to human IgG1 and κ constant regions . The mouse Fv was obtained by panning a phage display library made from splenic mRNA of a mouse immunized with mesothelin cDNA on mesothelin protein . Laboratory studies show that MORAb-009 kills mesothelin-expressing cell lines via antibody dependent cellular cytotoxicity and, in addition, it inhibits the binding of mesothelin to CA-125 . Based on these studies a three-institution phase I clinical trial of MORAb-009 was conducted and recently completed in patients with mesothelin expressing cancers.1 This report describes the effect of MORAb-009 on raising the serum CA-125 level in all eight patients with mesothelioma treated at our site. In addition, the possible mechanism for the elevation of CA-125 and the implications of our findings for therapy of mesothelioma and ovarian cancer are discussed.
Twenty-four patients with mesothelin expressing tumors were treated on this phase I study of MORAb-009 conducted at the National Cancer Institute (NCI), Johns Hopkins and Fox Chase Cancer Center. All patients signed informed consent prior to being enrolled on this study. This report describes the effect of MORAb-009 on serum CA-125 levels in the eight patients with mesothelioma treated at the NCI. These included four patients each with pleural and peritoneal mesothelioma. Serum CA-125 levels were measured before and at different time-points after treatment. Since serum CA-125 levels were not followed serially in patients treated at the other institutions that data is not available.
Key inclusion criteria for participation in this study included patients with mesothelin expressing mesothelioma, pancreatic cancer, ovarian cancer and lung adenocarcinoma who had failed standard therapies for their cancer.1 Other eligibility criteria included Eastern Cooperative Oncology Group performance status 0–2, adequate organ function and no co-morbid conditions. Patients were treated with MORAb-009 given as a weekly intravenous infusion × 4 weeks with 2 weeks off (considered as one cycle). Patients with stable or responding disease could receive additional cycles of therapy given in the same schedule. Tumor assessment for response was performed prior to therapy and after each cycle. Since this was a phase I dose escalation trial, patients received doses of MORAb-009 ranging from 12.5 mg/m2 to 400 mg/m2.
Blood samples for serum CA-125 measurement were obtained from patients within 4 weeks prior to infusion of MORAb-009 and at various time-points after treatment. Serum CA-125 levels were measured by the Department of Laboratory Medicine of the NIH Clinical Center, using the AxSYM CA-125 assay (Abbott Laboratories, Abbott Park, IL USA). It is a fully automated microparticle enzyme immunoassay that uses the OC125 monoclonal antibody. Normal reference range for CA-125 for this assay is 0–35 U/ml, with a coefficient of variation of less than 10%.
Because MORAb-009 present in patient serum could potentially interfere with the AxSYM CA-125 assay, the effect of MORAb-009 on the assay was determined. Serum samples were obtained from two patients with ovarian cancer, one with CA-125 within normal limits (sample A) and the other with elevated CA-125 levels (sample B). These samples were supplemented with either MORAb-009 100 μg/ml, rituximab 100 μg/ml or phosphate buffered saline at room temperature for 1 hour and then CA-125 levels were measured as usual with the AxSYM CA-125 assay.
Table I describes the characteristics of the eight patients, six males and two females, with malignant mesothelioma treated at NCI. All patients had serum CA-125 measurement before and at different time-points after therapy. Out of these eight patients four had pleural and four had peritoneal mesothelioma of the epithelial subtype; this tumor type has high cell surface mesothelin expression. These patients had all failed prior chemotherapy before going on this study. The patients were treated at various dose levels of MORAb-009 as per the phase I dose escalation design. One patient each was treated at 12.5, 25, 50 and 100 mg/m2, and two patients each were treated at 200 and 400 mg/m2.1 Out of the six patients who were evaluable for tumor response four had stable disease and two had progressive disease, using the Response Evaluation Criteria in Solid Tumors. Patients 1007 and 1008 were not evaluable for tumor response. Patient 1007 received only one dose of MORAb-009 and was taken off-study because of transient grade 4 elevation of liver transaminases, which was judged to be unrelated to MORAb-009 since the patient had a prior history of transient grade 2–3 liver transaminase elevations. Patient 1008 was taken off-study since she developed serum sickness after the second dose of MORAb-009 and was not evaluable for tumor response. Details regarding the clinical trial including dose escalation, adverse events, tumor response and MORAb-009 pharmacokinetics in the 24 patients treated on-study will be presented separately.
Although measurement of serum CA-125 levels was not part of the protocol, we routinely follow CA-125 levels in patients treated in our Mesothelioma Clinic at NCI, because several studies have previously shown that it may be a useful tumor marker to follow response to therapy in patients with mesothelioma [22, 23]. In all patients we noted a very interesting effect of MORAb-009 on serum CA-125 levels, which are summarized in Table II.
Elevated serum CA-125 levels prior to treatment with MORAb-009 were present in 6/8 patients, but not patients 1002 and 1007. There was a wide variation in serum levels among patients, which is most likely a reflection of tumor burden. Following treatment with MORAb-009 all patients had an increase in their serum CA-125 levels. For example, the baseline CA-125 level in patient 1001 was 1,887 U/ml and by day 35 had increased to 2,917 U/ml. In patient 1002, the baseline CA-125 levels were within the normal range prior to treatment and increased to 47 U/ml on day 35. This increase in CA-125 levels following treatment with MORAb-009 was seen in patients with both peritoneal and pleural mesothelioma.
As illustrated by patients 1002 and 1007, even patients with CA-125 levels within the normal range prior to treatment had an increase in CA-125 levels following treatment with MORAb-009. Furthermore, our results demonstrate that the marked increase in CA-125 levels can occur even with a single-dose of MORAb-009. In patient 1006 with pleural mesothelioma, the CA-125 levels increased from 128 U/ml to 262 U/ml after one infusion of MORAb-009 and to 644 after four infusions. Similarly, in patient 1007 the pre-treatment CA-125 level was 8 U/ml and increased more than 13-fold to 110 U/ml after just one infusion of MORAb-009.
Figure 1 illustrates the kinetics of CA-125 rise and fall following treatment with MORAb-009. In patient 1003 with peritoneal mesothelioma (Fig. 1a), the serum CA-125 levels increased more than 15-fold on day 22 to 752 U/ml (i.e. after patient had received 3 weekly infusions of MORAb-009) and gradually fell to five times the baseline levels on day 49 to 495 U/ml and were close to the pre-treatment values on day 176. This patient has continued to have stable disease on imaging studies on day 176. Similarly in patient 1005 (Fig. 1b), who had pleural mesothelioma with an elevated baseline CA-125 value of 399 U/ml and received two cycles of MORAb-009, the CA-125 levels continued to increase while the patient was receiving therapy but decreased to pre-treatment levels once treatment was completed. The peak CA-125 level of 882 U/ml occurred on day 83, about 2 weeks after the last MORAb-009 dose on day 70 and by day 120 the CA-125 levels had decreased by about half to 466 U/ml. These results illustrate that the increase in serum CA-125 levels was not due to disease progression but rather a consequence of MORAb-009 therapy.
To rule out the possibility that the rise in serum CA-125 was due to MORAb-009 interfering with the clinical assay, we did the CA-125 assay in the presence of MORAb-009. Serum samples from patients with ovarian cancer with serum CA-125 levels within normal limits (sample A, Fig. 2a) or with elevated serum CA1–25 level (sample B, Fig. 2b) were spiked with 100 μg/ml of MORAb-009, and then CA-125 levels measured. As a control, these serum samples were also spiked with the isotype matched antibody rituximab using the same concentration. The concentration of antibody added to the serum samples was similar to the peak MORAb-009 levels observed in patients treated at the highest dose level of MORAb-009 (unpublished data). As shown in Figure 2, the addition of MORAb-009 (or rituximab) did not interfere with the measurement of serum CA-125 levels demonstrating that the elevation in serum CA-125 levels observed in our patients was not a consequence of MORAb-009 present in the serum interfering with the clinical assay used to measure CA-125.
Our results show that patients with mesothelioma treated with MORAb-009, a chimeric antibody reacting with mesothelin, have a very significant elevation in their CA-125 levels. This finding suggests that MORAb-009 is causing an elevation of CA-125, possibly by preventing it from binding to mesothelin present on mesothelial cells lining the pleura and peritoneum. Inhibition of the binding of cell associated mesothelin and CA-125 by MORAb-009 could prevent intraperitoneal/intrapleural spread of ovarian cancer and mesothelioma.
The results of our clinical trial show that MORAb-009 therapy led to a marked increase in serum CA-125 levels in all patients with mesothelioma treated at our institution. This increase in serum CA-125 levels was noted in patients with pleural as well as peritoneal mesothelioma, including patients who had CA-125 levels within the normal range prior to treatment. The marked increase in serum CA-125 levels was also observed with a single dose of MORAb-009. Elevation of CA-125 levels was seen in patients treated at different doses of MORAb-009, ranging from 12.5 mg/m2 to 400 mg/m2, demonstrating that even low doses of MORAb-009 are sufficient to increase serum CA-125.
This increase in serum CA-125 was not due to tumor progression but rather a consequence of MORAb-009 therapy since the CA-125 levels decreased to baseline values once the treatment was stopped. It is unlikely that the rise in serum CA-125 levels is due to peritoneal or pleural inflammation, because none of these patients had clinical signs of pleuritis or peritonitis. In addition, serum C-reactive protein, a marker of inflammation, was within normal limits in patient 1003 at the time when the serum CA-125 levels were markedly elevated, making peritoneal inflammation an unlikely cause for the rise in serum CA-125. We also show that MORAb-009 does not interfere with the laboratory assay used to measure CA-125 demonstrating that the rise in serum CA-125 levels is not due to MORAb-009 in patient serum interfering with this assay. Elevations of serum CA-125 levels have been reported in patients treated with mouse monoclonal antibodies who develop a human anti-mouse antibody (HAMA) response . This false increase in CA-125 levels is due to HAMA bridging the mouse capture and reporter antibodies used in these assays, resulting in positive interference. In our case, the patients were treated with a chimeric antibody, which is unlikely to interfere with the clinical assay even if the patients developed a human anti-chimeric antibody (HACA) response, because human antibodies are not used in the CA-125 assay. Furthermore, the CA-125 levels would not decrease rapidly as seen in our patients because any HACA formed would be predicted to still be present at significant levels at the time of follow-up for the repeat CA-125 measurement.
Although it would be of interest to measure changes in serum mesothelin concentration in patients treated with MORAb-009, our laboratory studies have shown that MORAb-009 interferes with the clinical assay (MESOMARK® Assay, Fujirebio Diagnostics, Inc., Malvern, PA USA) used to measure serum mesothelin (Hassan R, unpublished data), because of MORAb-009 binding to the same epitope on mesothelin as one of the antibodies used in the MESOMARK® Assay. Therefore serum mesothelin will not be a useful marker in monitoring patients being treated with MORAb-009.
Pre-clinical studies have demonstrated that mesothelin is the receptor for the tumor antigen CA-125 and that it binds to it with high affinity and specificity [14, 15]. These studies suggest that tumors expressing CA-125 may bind to mesothelin present on the mesothelial cells lining body cavities, leading to heterotypic cell adhesion and tumor metastasis. Tumors such as ovarian cancer and mesothelioma, that highly express mesothelin as well as CA-125, are both characterized by extensive intra-cavitary metastasis. Based on the pre-clinical studies, showing inhibition of mesothelin-CA-125 interaction by MORAb-009, and results of this clinical trial, it is likely that the rise in serum CA-125 seen in patients while on MORAb-009 therapy is due to MORAb-009 inhibiting mesothelin-CA-125 binding. The CA-125 made by the tumor cells is released into the peritoneal/pleural cavity and binds to mesothelin present on mesothelial cells lining the body cavities, and is also shed into the serum. MORAb-009 most likely binds to mesothelin on the mesothelial cells and this interaction could increase the pool of soluble CA-125 in the pleural/peritoneal cavity, which can then enter the systemic circulation leading to elevated serum levels as was seen in our patients. As the serum MORAb-009 concentration decreases, tumor secreted CA-125 can now bind to mesothelin lining the pleural/peritoneal cavity, leading to fall in serum CA-125 levels. This is illustrated by the model in Figure 3.
Our findings are important for two reasons. Firstly, our results suggest that serum CA-125 levels are not useful to assess tumor progression in mesothelioma patients treated with MORAb-009 and probably will also not be useful to follow ovarian cancer patients. However, this would require doing a formal analysis of CA-125 levels in patients with ovarian cancer treated with MORAb-009 since increased CA-125 values could be a consequence of MORAb-009 therapy rather than due to ovarian cancer progression. Secondly, these results provide insight into mesothelin-CA-125 interaction that can be exploited for cancer therapy. The potential interactions between mesothelin and CA-125 and their role in cancer metastasis are depicted in Figure 4. As per this model, binding of CA-125 present on tumor cells to mesothelin on the mesothelial cells lining the pleural or peritoneal cavity could lead to heterotypic adhesion and tumor spread. In addition, interaction between CA-125 that is normally present on mesothelial cells to tumor associated mesothelin can also lead to heterotypic adhesion and further dissemination of tumor along pleural or peritoneal cavities. Besides the heterotypic adhesion described above, interaction between mesothelin and CA-125 present on tumor cells may lead to tumor aggregation and potentiate intra-cavitary metastasis.
Inhibition of mesothelin-CA-125 interaction could be a potentially useful target for cancer therapy for tumors such as ovarian cancer and mesothelioma that highly express CA-125 and mesothelin. Clinical trials of agents such as MORAb-009 are needed to determine if inhibition of mesothelin-CA-125 binding results in clinical benefit.
This research was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Center for Cancer Research. The clinical trial was conducted under a Clinical Research And Development Agreement between the National Cancer Institute and Morphotek Inc.
1Laheru DA, Cohen SJ, Phillips DK, et al. A phase I study of MORAb-009, a monoclonal antibody against mesothelin, in mesothelioma, pancreatic and ovarian cancer. Abstract #3578, Presented at the 44th ASCO Annual Meeting, Developmental Therapeutics: Molecular Therapeutics Session, Chicago, IL, May 30-June 3, 2008
Conflict of interest
Charles Schweizer and Susan C. Weil are employed by Morphotek Inc. The other authors indicated no potential conflicts of interest.