ORIGINAL ARTICLE

Soluble Mesothelin, Megakaryocyte Potentiating Factor, and Osteopontin as Markers of Patient Response and Outcome in Mesothelioma Kevin Hollevoet, PhD,* Kristiaan Nackaerts, MD, PhD,† Robert Gosselin, MD,‡ Walter De Wever, MD, PhD,§ Lionel Bosque´e, MD,储 Paul De Vuyst, MD, PhD,¶ Paul Germonpre´, MD, PhD,# Eliane Kellen, MD, PhD,** Catherine Legrand, PhD,†† Yoshiro Kishi, PhD,‡‡ Joris R. Delanghe, MD, PhD,§§ and Jan P. van Meerbeeck, MD, PhD*

Introduction: Soluble mesothelin (SM), megakaryocyte potentiating factor (MPF), and osteopontin (OPN) are blood biomarkers of mesothelioma. This study evaluates their use as markers of response to therapy and outcome. Methods: Sixty-two patients with malignant pleural mesothelioma were included in an observational multicenter study. Blood samples and matched computed tomography scans were collected at diagnosis and, when possible, during and after therapy. For each patient, the best overall radiological response was compared with the changes in serum SM, MPF, and plasma OPN levels across corresponding time points. Results: In five patients, blood sampling was done shortly before and after extrapleural pneumonectomy. SM and MPF levels markedly decreased after surgery, whereas OPN levels showed a median increase. Fifty-seven patients were surveilled during (and after) chemotherapy, of whom 27 (47%) had stable disease, 14 (25%) partial response, and 16 (28%) progressive disease. In patients with stable disease, SM and MPF levels did not change significantly across the corresponding time points, whereas OPN levels significantly decreased. In those with partial response, SM and MPF levels *Department of Respiratory Medicine, Ghent University Hospital, Ghent; †Department of Respiratory Medicine, University Hospital Gasthuisberg, Leuven; ‡Department of Radiology, Ghent University Hospital, Ghent; §Department of Radiology, University Hospital Gasthuisberg, Leuven; 储Department of Respiratory Medicine, CHU Sart Tilman, Lie`ge; ¶Department of Respiratory Medicine, Erasme Hospital ULB, Brussels; #Department of Respiratory Medicine, Antwerp University Hospital, Antwerp; **Center for Cancer Prevention, University Hospital Gasthuisberg, Leuven; ††Institut de Statistique, Biostatistique et Sciences Actuarielles, Universite´ Catholique de Louvain, Louvain-la-Neuve, Belgium, ‡‡Department of Research and Development, Ina Institute, Medical & Biological Laboratories, Co. Ltd., Nagano, Japan; and §§Department of Clinical Chemistry, Ghent University Hospital, Ghent, Belgium. Disclosure: J.v.M. received a research grant of Cis Bio International (France). Y.K. is an employee of Medical & Biological Laboratories (Japan). Funding sources had no role in collection and analysis of the data, writing, and approval of the manuscript. Address for correspondence: Kevin Hollevoet, PhD, Department of Respiratory Medicine, Ghent University Hospital, 7K12IE, De Pintelaan 185, Ghent 9000, Belgium. E-mail: [email protected] Copyright © 2011 by the International Association for the Study of Lung Cancer ISSN: 1556-0864/11/0611-1930

1930

significantly decreased, whereas OPN levels showed no significant change. In patients with progressive disease, all three biomarker levels significantly increased. Patient responses correlated with a 15% change in all three biomarkers, although SM and MPF appeared more accurate than OPN. Low baseline OPN levels were independently associated with favorable progression-free survival and overall survival. Neither SM nor MPF showed prognostic value. Conclusions: SM and MPF levels were more closely associated with disease course than OPN and might prove useful in monitoring patient response in mesothelioma. Baseline OPN levels were an independent negative predictor of survival. These promising results require further validation. Key Words: Mesothelioma, Biomarker, Mesothelin, Osteopontin, Megakaryocyte potentiating factor, Monitoring, Response, Prognosis. (J Thorac Oncol. 2011;6: 1930–1937)

M

alignant mesothelioma is a fatal asbestos-related malignancy, predominantly arising from the surface serosal cells of the pleura. The natural history of mesothelioma results in a median survival of 7 to 9 months.1 When treated with standard of care chemotherapy, an antifolate (pemetrexed or raltitrexed) and a platinum agent, median survival is approximately 1 year.2,3 Highly selected patients with early-stage epithelioid disease, treated with extrapleural pneumonectomy (EPP), either alone or in combination with chemotherapy and/or radiation therapy, have a median survival of up to 2 years.4 Several novel anticancer drugs are currently evaluated in clinical trials, with tumor shrinkage being the standard end point to evaluate their efficacy. The ability to reproducibly measure tumor response consequently plays a pivotal role in cancer therapy research.1,5 In patients with mesothelioma, this, however, is troublesome. The RECIST, the standard method for radiological response assessment,6 is not adequate for this malignancy, as mesothelioma often presents as a rind around the lungs, rather than as a spherical mass.5 An adjustment of these criteria, i.e., “the modified RECIST,”7 has only in part addressed this concern and still displays a low interobserver agreement.8 An additional limitation of radiological

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

assessment, inherent to mesothelioma, is the presence of pleural fluid and benign asbestos-related lesions. A blood biomarker, which accurately reflects disease course, would be a useful and relatively cheap adjunct to monitor patient response. In mesothelioma, soluble mesothelin (SM), megakaryocyte potentiating factor (MPF), and osteopontin (OPN) are currently the most promising blood biomarkers,9 –11 and the primary candidates for such purpose. Both SM and MPF originate from the mesothelin gene, which is overexpressed in epithelioid mesothelioma cells.12 The Mesomark™ enzyme-linked immunosorbent assay (ELISA), commonly used to measure SM in serum, was approved in 2007 by the US Food and Drug Administration to aid in the monitoring of patients with epithelioid or biphasic mesothelioma.13 Despite this approval and the recent studies which demonstrate the potential of SM as a marker of response,14 –16 this biomarker is not commonly used in clinical practice. For OPN, a recent study found no association between plasma levels and response in mesothelioma, whereas others do report a correlation with disease recurrence in lung and ovarian cancer.17,18 MPF has not yet been extensively studied as a marker of response. SM, MPF, and OPN could also improve therapeutic decision making as prognostic markers. For SM, different studies previously found that baseline serum levels are an independent negative predictor of overall survival,16,19 –21 whereas others failed to do so.22,23 For MPF, no data are currently available. For OPN, tumor expression and blood levels are associated with poor outcome or an aggressive phenotype in a variety of malignancies, including mesothelioma,20,24 but extensive validation is absent. The association of SM, MPF, and OPN with progression-free survival in mesothelioma has not been previously assessed. This prospective multicenter study aimed (a) to examine whether changes in SM, MPF, and OPN levels reflect the response to therapy in patients with mesothelioma and (b) to assess whether baseline biomarker levels have a prognostic value for progression-free survival and overall survival.

METHODS Participants Between January 2007 and December 2010, patients with malignant pleural mesothelioma, scheduled for further treatment, were included at diagnosis in an observational multicenter study. Diagnosis was based on pleural biopsies and validated by the pathologists of the Belgian Mesothelioma Panel. Age, sex, serum C-reactive protein (CRP) levels, performance status (PS), tumor stage, and histology were recorded at inclusion. Blood sampling was done at inclusion, and, when possible, at each visit during and after therapy, until either progressive disease (PD) or death occurred. Patients’ renal function was monitored with serum creatinine measurements. All chest computed tomography (CT) scans made between inclusion and last blood sampling were encoded and collected. Tumor staging was done according to the guidelines of the International Mesothelioma Interest Group, which are approved by the International Union Against Cancer.25 Serum and plasma were stored in aliquots

SM, MPF, and OPN

at ⫺80°C. This study was approved by the ethics committee of all participating hospitals. Written informed consent was obtained from all participants before inclusion. The baseline SM and MPF levels of these participants have recently been reported in a study, which compared the diagnostic accuracy of both biomarkers.11

Biomarker Assays SM (nmol/L) and MPF levels (ng/ml) were measured in serum using the Mesomark™ (Cis bio International, Gif sur Yvette, France) and Human MPF ELISA kit (Medical & Biological Laboratories, Nagano, Japan),26 respectively. OPN levels (ng/ml) were measured in plasma with the Human Osteopontin ELISA™ (Immunobiological Laboratories, Hamburg, Germany). Assays were run according to manufacturer’s instructions, blinded to patient data.

Radiological and Biomarker Response Two experienced thoracic radiologists independently assessed the radiological responses with the modified RECIST.7,8 Conflicts were solved by consensus, blinded to the biomarker results. Depending on the number of available CTs, one or more time point responses were recorded per patient, each delimiting a certain time interval during followup. Each CT had a matched blood sample, thus allowing to compare the radiological response with the relative and absolute changes in SM, MPF, and OPN levels across the corresponding time points. To ensure a relevant association, the time between CT and blood sample had to be less than 3 weeks. For each patient, the best overall radiological response since the start of treatment was recorded.27

Statistical Analysis Continuous variables were reported with their median and 25th to 75th percentile values. Correlation between baseline biomarker levels was assessed with the nonparametric Spearman rank test. Patients who had pre- and postsurgery samples available were discussed separately in a descriptive manner, as their small number did not allow for robust statistical analysis. In the group of patients who received (neoadjuvant) chemotherapy, the relationship between patient response and the corresponding biomarker changes was evaluated with a number of exploratory analyses. First, patients were stratified according to their best overall radiological response: stable disease (SD), partial response (PR), or PD. For each of these responses, the Wilcoxon signed-rank test assessed whether the differences in median biomarker levels across the corresponding time points differed significantly from 0. Second, patients were stratified according to their best overall response and the relative biomarker change: a decrease ⱖ15%, a change less than 15% (stable), or an increase ⱖ15%. Fisher’s exact test subsequently evaluated whether a significant association was present between these two variables. The 15% threshold was chosen to avoid interference with the ELISA variance, which can be up to 12%.11,28 Third, the sensitivity and specificity of a relative biomarker change for a specific response were evaluated. To differentiate patients with PR from those with SD and PD, the threshold was set at a decrease ⱖ15%; to

Copyright © 2011 by the International Association for the Study of Lung Cancer

1931

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

Hollevoet et al.

distinguish patients with SD from those with PR and PD, the threshold was a change less than 15%; and to discriminate patients with PD from SD and PR, an increase ⱖ15% was applied. Finally, as an illustration, all available time point responses during follow-up (in addition to the best overall responses) were plotted against the corresponding relative SM, MPF, and OPN changes in waterfall plots. Progression-free survival and overall survival analyses were performed with univariate and multivariate Cox proportional hazards regression. Progression-free survival was defined as the number of days from date of inclusion until PD. Patients still alive and without PD were censored on the date of the last available CT. Of note, this follow-up exceeded the period of blood sampling. Overall survival was defined as the number of days from date of inclusion until death. Patients who were alive at February 1, 2011, were censored on that date. For the continuous biomarker levels, the threshold which best differentiated those with poor from those with favorable prognosis was searched with an algorithm of maximization of hazards ratio (HR)29 and evaluated with Kaplan-Meier and log-rank statistics. All hypothesis tests were performed two sided at the 5% significance level. Statistical analyses were done

TABLE 1. Baseline Demographics of the 62 Patients with Malignant Pleural Mesothelioma Therapy Regimen Chemotherapy (n ⫽ 48)

Covariates Stage PS Histology

Multimodality (n ⫽ 14)

Stage PS Histology

n I II–IV 0 1–3 Epithelioid Biphasic II III 0 1–2 Epithelioid Sarcomatoid Biphasic

15 33 17 31 47 1 9 5 6 8 12 1 1

PS, performance status.

with SPSS (version 17, SPSS Inc., Chicago, IL) and SAS (version 9.2, SAS Institute Inc., Cary, NC).

RESULTS Participants In total, 62 patients with malignant pleural mesothelioma were included (53 males and 9 females). Patients had a median age of 64 years (59 –72 years) and a median CRP level of 2.11 mg/dl (0.50 – 8.00 mg/dl). Forty-eight patients were treated with combination chemotherapy only (pemetrexed and a platinum agent), whereas the 14 others received a multimodality therapy: neoadjuvant chemotherapy (pemetrexed and a platinum agent) and EPP, with or without postoperative radiation therapy. In 5 of these 14 patients (four epithelioid and one sarcomatoid subtype), a pre- and post-EPP blood sample was available, whereas the nine others were only surveilled during neoadjuvant chemotherapy. In 43 of the 62 patients, follow-up was only available during (a part of) the therapy, whereas in the 19 others, control visits were also included. Median follow-up period with blood sampling and CTs was 3.5 months (2.2– 6.3 months), during which patients received a median of four chemo cycles (2– 6 cycles). Forty-three of the 62 patients received cisplatin as platinum agent, whereas the 19 others had carboplatin. Based on the serum creatinine levels, none of the patients presented nephrotoxicity. Additional characteristics are displayed in Table 1.

Biomarker Levels at a Baseline At diagnosis, patients had a median SM level of 1.98 nmol/L (1.25– 4.72 nmol/L), a median MPF level of 15.26 ng/ml (8.83–36.07 ng/ml), and a median OPN level of 868.32 ng/ml (652.19 –1209.20 ng/ml). Baseline SM and MPF were highly correlated (r ⫽ 0.90, p ⬍ 0.001), whereas neither of them significantly correlated with OPN (rSM ⫽ 0.23, pSM ⫽ 0.07; rMPF ⫽ 0.23, pMPF ⫽ 0.07). All three biomarkers had significantly lower levels in patients with stage I disease, compared with those with stages II to IV (p ⬍ 0.05). Biomarker levels did not differ according to PS (0 versus 1–3) (data not shown).

32

2000

80 64

1932

8 4 2

OP PN levels (ng/m mL)

MPF levels (ng/m mL)

FIGURE 1. Biomarker levels before and after extrapleural pneumonectomy (EPP). In all five patients with mesothelioma, soluble mesothelin (SM) and megakaryocyte potentiating factor (MPF) levels decreased after surgery. Osteopontin (OPN) levels increased in four of the five individuals. Four patients had an epithelioid histology, whereas the patient with the lowest pre-EPP biomarker levels had a sarcomatoid subtype.

SM levels (nmo S o/L)

16 32

16

8

1500

1000

500

1 0.5

Pre-EPP Post-EPP

4

Pre-EPP Post-EPP

0

Pre-EPP Post-EPP

Copyright © 2011 by the International Association for the Study of Lung Cancer

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

Biomarker Levels After EPP

postsurgery biomarker measurements were available in one patient with an epithelioid subtype and showed that OPN levels, after the initial increase, substantially decreased 8.6 months after EPP (Figure 2). Furthermore, after the initial decrease in all three biomarkers, levels steadily increased and peaked approximately 2 years after EPP, when disease recurrence was radiologically established.

SM and MPF levels decreased in all five patients after EPP, resulting in a median fall of 78% (49 – 83%) and 76% (40 –78%), respectively (Figure 1). In contrast, OPN levels increased in four patients, resulting in a median rise of 20% (15–22%). In the one patient with a sarcomatoid histology, biomarker changes were only modest: ⫺21% for SM, ⫺11% for MPF, and ⫺13% for OPN. Median time between the EPP and the pre- and postsurgery sample was 21 days (13–22 days) and 36 days (33–50 days), respectively. Longitudinal

Radiological Response and Biomarker Change Of the 57 patients who were surveilled during (and after) chemotherapy, 27 (47%) had SD as best overall radiological response, 14 (25%) PR, and 16 (28%) PD. In the patients with SD, SM and MPF levels did not significantly change across the corresponding time points, whereas OPN levels significantly decreased (Table 2). In those with PR, SM and MPF levels significantly decreased, whereas OPN levels did not change significantly. In the patients with PD, all three biomarker levels significantly increased, although the relative increase in OPN levels was modest, compared with SM and MPF. Fisher’s exact test demonstrated a significant association between the best overall responses and a 15% change in, respectively, SM, MPF, and OPN levels (pSM ⬍ 0.001, pMPF ⬍ 0.001, pOPN ⬍ 0.05). Relative changes in SM and MPF levels, however, displayed a higher accuracy for SD, PR, and PD, than OPN (Table 3). When considering all available follow-up data, a total of 125 time point responses were recorded, including 70 SD (56%), 26 PR (23%), and 29 PD (21%). Twenty-three patients had one time point response, 15 patients had two, six patients had three, 11 patients had four, and two patients had five. The median duration of a time interval was 2.1 months (1.4 –3.5 months). The resulting waterfall plots again illustrated the lower accuracy of OPN compared with SM and MPF (Figure 3). Time intervals with PD and PR virtually all displayed an increase and decrease, respectively, in SM and MPF levels, although not all changes reached the 15% threshold. In time intervals with SD, the distribution of the relative biomarker changes was more dispersed, but the majority had stable or decreased biomarker levels, as illustrated in the online available waterfall plots of SM, MPF, and OPN

Relative biom marker ch hange (%)) after EPP

50

25

SM

0

SM, MPF, and OPN

MPF OPN

-25

-50 50

-75 75

Relapse -100 100 23

20

14

10

.4

.3

.3

.7

6 8.

2 5.

7 1. PP E epr

Time after EPP (months)

FIGURE 2. Longitudinal follow-up after extrapleural pneumonectomy (EPP) of one patient with biomarker measurements. In this patient with epithelioid mesothelioma, soluble mesothelin (SM), megakaryocyte potentiating factor (MPF) and, somewhat later, osteopontin (OPN) levels decreased after EPP. Biomarker levels gradually increased again and peaked when disease recurrence was radiologically established.

TABLE 2. Relative and Absolute Biomarker Changes According to Best Overall Radiological Response of 57 Patients with Mesothelioma Median (P25-P75) Biomarker Difference Across Corresponding Time Points

SM

MPF

OPN

Relative (%) Absolute (nmol/L) Wilcoxon rank test Relative (%) Absolute (ng/ml) Wilcoxon rank test Relative (%) Absolute (ng/ml) Wilcoxon rank test

Stable Disease (n ⴝ 27)

Partial Response (n ⴝ 14)

Progressive Disease (n ⴝ 16)

⫺0.81 (⫺7.18 to 7.29) ⫺0.01 (⫺0.20 to 0.11) p ⫽ 0.60 ⫺5.75 (⫺24.91 to 8.80) ⫺0.52 (⫺3.40 to 0.60) p ⫽ 0.11 ⫺13.80 (⫺30.53 to 7.38) ⫺97.87 (⫺360.04 to 58.04) p ⬍ 0.05

⫺34.19 (⫺63.40 to ⫺11.92) ⫺0.63 (⫺3.60 to ⫺0.31) p ⬍ 0.01 ⫺52.79 (⫺75.98 to ⫺26.47) ⫺10.56 (⫺35.91 to ⫺4.72) p ⬍ 0.01 ⫺6.28 (⫺28.91 to 7.61) ⫺26.77 (⫺344.85 to 152.07) p ⫽ 0.22

53.92 (17.25 to 94.34) 1.55 (0.29 to 2.67) p ⬍ 0.001 57.88 (21.58 to 147.50) 8.14 (2.92 to 24.49) p ⬍ 0.001 7.46 (⫺2.24 to 36.06) 84.79 (⫺23.86 to 486.87) p ⬍ 0.05

SM, soluble mesothelin; MPF, megakaryocyte potentiating factor; OPN, osteopontin.

Copyright © 2011 by the International Association for the Study of Lung Cancer

1933

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

Hollevoet et al.

TABLE 3. Accuracy of the Relative Biomarker Changes for the Best Overall Radiological Response of the 57 Patients with Mesothelioma Sensitivity Best Overall Response Stable disease (n ⫽ 27) Partial responseb (n ⫽ 14) Progressive diseasec (n ⫽ 16) a

Specificity

SM

MPF

OPN

SM

MPF

OPN

63% (17/27) 64% (9/14) 81% (13/16)

43% (13/27) 86% (12/14) 81% (13/16)

37% (10/27) 43% (6/14) 38% (6/16)

77% (23/30) 88% (38/43) 85% (35/41)

90% (27/30) 77% (33/43) 88% (36/41)

53% (16/30) 67% (29/43) 83% (34/41)

0 -25 50 -75

-100

Relative changes in MPF levels (%)

B Relative cha anges in SM le evels (%)

FIGURE 3. Waterfall plot of the relative biomarker changes across time intervals with partial response (A) and with progressive disease (B). Soluble mesothelin (SM), megakaryocyte potentiating factor (MPF), and osteopontin (OPN) levels typically decreased across the 26 time intervals with partial response (A) and increased across the 29 time intervals with progressive disease (B). Dotted lines at ⫺15% and 15% represent the thresholds for a significant change.

7000 700 700 150 150 100 50 0 -15

(Supplemental Figure 1, Supplemental Digital Content 1, http://links.lww.com/JTO/A103).

Survival Analysis The progression-free survival and overall survival analyses were performed in the 48 patients who were treated with combination chemotherapy only (Table 1). During the course of the study, PD was recorded in 31 patients (65%), whereas 30 patients (63%) died. Patients had an estimated median progression-free survival of 6.8 months (95% confidence interval [CI] ⫽ 3.5–10.1 months). In the univariate analysis, baseline OPN and CRP levels were significantly and inversely associated with progression-free survival. In the multivariate analysis, both retained significance (Table 4). Neither SM and MPF nor PS and tumor stage were informative for patient outcome. For OPN, the search algorithm29 identified 862.78 ng/ml as the

1934

Relative cha anges in OPN levels (%)

35

35

0 25 -50 -75 -100

800

150 150 125

75

25 0

Relative changes in OPN levels (%)

Relative changes in SM le evels (%)

A

Relative cha anges in MPF llevels (%)

Threshold is set at a achange ⬍15%, bdecrease ⱖ15%, or cincrease ⱖ15% in biomarker levels. SM, soluble mesothelin; MPF, megakaryocyte potentiating factor; OPN, osteopontin.

35

0 -25 -50 -75 -100

250 200

150 100 50 0 -30

optimal threshold to differentiate those with poor from those with favorable outcome. Patients with an OPN level below this threshold had an estimated median progression-free survival of 14.3 months (95% CI ⫽ 8.5–20.1 months), whereas the others had a median progression-free survival of only 5.6 months (95% CI ⫽ 4.4 – 6.8 months) (p ⬍ 0.01) (Figure 4A). When this threshold was included in the multivariate model, instead of the continuous OPN levels, it was an independent predictor of progression-free survival (HR ⫽ 2.912, 95% CI ⫽ 1.286 – 6.596). Patients had an estimated median overall survival of 15.4 months (95% CI ⫽ 12.6 –18.3 months). In the univariate analysis, advanced mesothelioma tumor stage (I versus II–IV) and baseline OPN and CRP levels were significant negative predictors of overall survival. In the multivariate analysis, only OPN and CRP retained signif-

Copyright © 2011 by the International Association for the Study of Lung Cancer

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

SM, MPF, and OPN

TABLE 4. Survival Analyses in 48 Patients with Mesothelioma Who Received Chemotherapy Only Univariate Analysis Outcome Progression-free survival

Overall survival

Multivariate Analysis

Covariates

Categories

HR (95% CI)

p

HR (95% CI)

p

SM (nmol/L) MPF (ng/ml) OPN (ng/ml) CRP (mg/dl) Mesothelioma stage PS SM (nmol/L) MPF (ng/ml) OPN (ng/ml) CRP (mg/dl) Mesothelioma stage PS

Continuous Continuous Continuous Continuous ⬎I ⬎0 Continuous Continuous Continuous Continuous ⬎I ⬎0

1.001 (0.963–1.039) 1.001 (0.996–1.006) 1.0009 (1.0001–1.0017) 1.019 (1.003–1.035) 1.941 (0.817–4.610) 1.068 (0.495–2.304) 0.993 (0.958–1.030) 1.001 (0.996–1.005) 1.0009 (1.0002–1.0016) 1.022 (1.006–1.037) 2.809 (1.129–6.986) 1.039 (0.486–2.224)

0.98 0.71 ⬍0.05 ⬍0.05 0.13 0.87 0.72 0.78 ⬍0.05 ⬍0.01 ⬍0.05 0.92

— — 1.0009 (1.0001–1.0018) 1.018 (1.002–1.035) — — — — 1.0008 (1.0001–1.0015) 1.019 (1.003–1.035) 2.002 (0.766–5.231) —

— — ⬍0.05 ⬍0.05 — — — — ⬍0.05 ⬍0.05 0.16 —

HR, hazard ratio; CI, confidence interval; CRP, C-reactive protein; PS, performance status; SM, Soluble mesothelin; MPF, megakaryocyte potentiating factor; OPN, osteopontin.

FIGURE 4. Survival curves stratified according to baseline osteopontin (OPN) levels. OPN levels are an indicator of favorable (A) progression-free survival and (B) overall survival.

icance (Table 4). SM, MPF, and PS were not informative for patient outcome. For OPN, the search algorithm29 identified 1141.11 ng/ml as the optimal threshold. Patients with levels below this threshold had an estimated median overall survival of 18.9 months (95% CI ⫽ 14.6 –23.1 months), whereas in the others, median overall survival was limited to 10.1 months (95% CI ⫽ 5.5–14.7 months) (p ⬍ 0.05) (Figure 4B). When this threshold was included in the multivariate model, instead of the continuous OPN levels, it was an independent predictor of overall survival (HR ⫽ 2.216, 95% CI ⫽ 1.050 – 4.678).

DISCUSSION This multicenter observational study aimed to gain insight in the role of SM, MPF, and OPN as markers of patient response and outcome in mesothelioma.

In agreement with previous findings, SM and MPF levels markedly decreased shortly after EPP, illustrating their close association with tumor bulk.9,15,30,31 Although less substantial, such decrease was also observed in a patient with a sarcomatoid histology. As this subtype does not overexpress mesothelin,12 this possibly indicated the presence of some epithelioid cell nests in the tumor. In contrast, OPN levels showed a median increase shortly after EPP. The longitudinal follow-up of one patient, however, revealed that OPN levels did decrease in the long run. This interesting observation was in accordance with a recent study in patients with nonsmall cell lung cancer.18 In their series, OPN levels decreased after surgery, but levels shortly after surgery were much higher than those 6 weeks later.18 According to the authors, this highlighted the role of OPN in wound healing and tissue remodeling.18 As such, the association between OPN and

Copyright © 2011 by the International Association for the Study of Lung Cancer

1935

Hollevoet et al.

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

tumor debulking appears to be blurred by its inflammatory response to the surgery. In 57 patients surveilled during (and after) chemotherapy, the best overall radiological responses (SD, PR, or PD) were significantly associated with relative changes in both SM, MPF, and OPN levels. These findings were in agreement with recent studies on SM14 –16 and demonstrated that MPF has a very similar behavior as marker of response. This was somewhat expected, as we recently reported that both biomarkers are highly correlated, have an equivalent diagnostic accuracy, and a similar longitudinal behavior in asbestosexposed controls.11,32 For OPN, our results were in agreement with previous findings in lung and ovarian cancer17,18 but in contrast with a recent study in 21 patients with mesothelioma, which found no association with patient response.15 These conflicting results might be due to the smaller sample size in the latter study. Compared with SM and MPF, however, OPN was less closely associated with the radiological responses. This was illustrated by the less marked changes in OPN levels in patients with PR or PD, the inferior sensitivity and specificity for patient responses, and the less straightforward waterfall plots. Our findings also indicated that, even for SM and MPF, the accuracy of biomarker changes for radiological response was not absolute. Although this was in part caused by considering a 15% threshold in relative biomarker changes, some patients with SD, for example, displayed a marked increase in biomarker levels. This could indicate that an increase in biomarker levels precedes the radiological presentation of progression. In ovarian cancer, CA-125, for example, has shown to increase several months before disease progression becomes radiologically evident.33 Another plausible cause is radiological misclassification, considering the current difficulties with the modified RECIST. Altogether, the more accurate association with tumor debulking and radiological response indicated that SM and MPF are more suitable for routine monitoring, when compared with OPN. If confirmed in larger series, SM or MPF could be used as an adjunct to radiological monitoring in patients with epithelioid or biphasic mesothelioma. The choice of threshold, which differentiates stable biomarker levels from a significant change, will hereby be important. Two recent studies demonstrated that a change in SM levels above 10% can already adequately reflect patient response,14,15 whereas others proposed a 25% threshold.16 We opted for a 15% threshold in relative biomarker changes. Although this led to a decrease in sensitivity, it increased specificity, hereby especially avoiding interference with the variance of the applied ELISA, which can be up to 12%.11,28 Further study is required to establish the optimal threshold for routine use. Of note, several issues can lead to false-positive changes during monitoring. First, low biomarker levels are more subject to assay variance, and a minor absolute change could easily lead to a high relative change. Second, we recently found that SM and MPF levels in controls, without evidence of malignant disease, significantly increase over time, in part due to aging and assay variance.32 Third, SM and MPF are both inversely associated with renal function,32,34

1936

and chemotherapeutics such as cisplatin may cause nephrotoxicity during monitoring,35 although this was not the case in our series. This study has some limitations, which require consideration. Because of the observational and exploratory nature of the study, blood sampling and CTs were not mandatory. As a consequence, the follow-up and time between blood sampling were not uniform across patients. In addition, only a limited number of patients were followed up after completion of the therapy. Finally, although the number of surveilled patients in this study was one of the largest so far, it still is relatively small. Further validation of our findings is consequently required to elucidate the clinical utility of these biomarkers in response evaluation. In agreement with the literature, CRP levels and tumor stage were associated with overall survival in mesothelioma. Low baseline OPN levels were independently associated with favorable progression-free survival and overall survival, which adds further proof to its prognostic significance in mesothelioma.20,24 In addition, OPN expression and blood levels are positively correlated with tumor stage, progression, invasion, and metastasis.36 Recent in vitro studies gave an insight in the underlying biology and indicated that OPN plays an important role in the multidrug resistance and enhancement of adhesion, proliferation, and migration activities of mesothelioma cells.37,38 For SM, different studies previously reported that baseline levels were an independent negative predictor of overall survival,16,19 –21 whereas other studies failed to do so.22,23 In our series, neither SM nor MPF predicted progression-free survival and overall survival. These survival analyses, however, were based on a relatively low number of events and lack power. As a consequence, our negative findings for SM and MPF should be interpreted with caution and further examined. In conclusion, SM and MPF levels were more closely associated with disease course than OPN and might prove useful in monitoring patient response in mesothelioma. Low baseline OPN levels were independently associated with improved progression-free survival and overall survival. Further validation of these biomarkers in larger prospective trials is necessary to determine their clinical utility.

ACKNOWLEDGMENTS Supported by the Foundation Against Cancer, a foundation of public interest, and the research grant Emmanuel van der Schueren of the Flemish League Against Cancer. Cis bio International (France) provided a research grant. Human MPF ELISA kits were kindly provided by Medical & Biological Laboratories (Japan). The authors thank all study participants and the investigators involved in the study at Ghent University Hospital, Ghent (F. Vande Walle, T. Verstraete, V. Stove), University Hospital Gasthuisberg, Leuven (L. Peeters), Erasme Hospital ULB, Brussels (G. Amand), CHU Sart Tilman, Lie`ge (S. Maccan and K. Mobarak), AZ Nikolaas, Sint-Niklaas (K. Deschepper and M. Van Hooste), AZ Sint-Maarten, Mechelen (M. Lambrechts), CHR de la Citadelle, Lie`ge (M. Crine and F. Bustin), AZ Imelda, Bonheiden (T. Lauwerier), AZ Virga

Copyright © 2011 by the International Association for the Study of Lung Cancer

Journal of Thoracic Oncology • Volume 6, Number 11, November 2011

Jesse, Hasselt (L. Spaas), and Medical & Biological Laboratories, Nagano (Y. Fujii and R. Uehara). REFERENCES 1. van Meerbeeck JP, Scherpereel A, Surmont VF, et al. Malignant pleural mesothelioma: the standard of care and challenges for future management. Crit Rev Oncol Hematol 2010;78:92–111. 2. Vogelzang NJ, Rusthoven JJ, Symanowski J, et al. Phase III study of pemetrexed in combination with cisplatin versus cisplatin alone in patients with malignant pleural mesothelioma. J Clin Oncol 2003;21: 2636 –2644. 3. van Meerbeeck JP, Gaafar R, Manegold C, et al. Randomized phase III study of cisplatin with or without raltitrexed in patients with malignant pleural mesothelioma: an intergroup study of the European Organisation for Research and Treatment of Cancer Lung Cancer Group and the National Cancer Institute of Canada. J Clin Oncol 2005;23:6881– 6889. 4. Weder W, Opitz I, Stahel R. Multimodality strategies in malignant pleural mesothelioma. Semin Thorac Cardiovasc Surg 2009;21:172– 176. 5. van Klaveren RJ, Aerts J, de Bruin H, et al. Inadequacy of the RECIST criteria for response evaluation in patients with malignant pleural mesothelioma. Lung Cancer 2004;43:63– 69. 6. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000;92:205–216. 7. Byrne MJ, Nowak AK. Modified RECIST criteria for assessment of response in malignant pleural mesothelioma. Ann Oncol 2004;15:257– 260. 8. Tsao AS, Garland L, Redman M, et al. A practical guide of the Southwest Oncology Group to measure malignant pleural mesothelioma tumors by RECIST and modified RECIST criteria. J Thorac Oncol 2011;6:598 – 601. 9. Robinson BW, Creaney J, Lake R, et al. Mesothelin-family proteins and diagnosis of mesothelioma. Lancet 2003;362:1612–1616. 10. Pass HI, Lott D, Lonardo F, et al. Asbestos exposure, pleural mesothelioma, and serum osteopontin levels. N Engl J Med 2005;353:1564 – 1573. 11. Hollevoet K, Nackaerts K, Thimpont J, et al. Diagnostic performance of soluble mesothelin and megakaryocyte potentiating factor in mesothelioma. Am J Respir Crit Care Med 2010;181:620 – 625. 12. Hassan R, Ho M. Mesothelin targeted cancer immunotherapy. Eur J Cancer 2008;44:46 –53. 13. Li Z, Verch T, Allard JW. MESOMARK威 in vitro diagnostic test for mesothelioma. Expert Opin Med Diagn 2007;1:137–142. 14. Grigoriu BD, Chahine B, Vachani A, et al. Kinetics of soluble mesothelin in patients with malignant pleural mesothelioma during treatment. Am J Respir Crit Care Med 2009;179:950 –954. 15. Wheatley-Price P, Yang B, Patsios D, et al. Soluble mesothelin-related peptide and osteopontin as markers of response in malignant mesothelioma. J Clin Oncol 2010;28:3316 –3322. 16. Creaney J, Francis RJ, Dick IM, et al. Serum soluble mesothelin concentrations in malignant pleural mesothelioma: relationship to tumor volume, clinical stage and changes in tumor burden. Clin Cancer Res 2011;17:1181–1189. 17. Schorge JO, Drake RD, Lee H, et al. Osteopontin as an adjunct to CA125 in detecting recurrent ovarian cancer. Clin Cancer Res 2004;10: 3474 –3478. 18. Blasberg JD, Pass HI, Goparaju CM, et al. Reduction of elevated plasma

19. 20. 21.

22. 23. 24. 25. 26. 27. 28. 29. 30.

31. 32. 33. 34. 35.

36. 37. 38.

SM, MPF, and OPN

osteopontin levels with resection of non-small-cell lung cancer. J Clin Oncol 2010;28:936 –941. Cristaudo A, Foddis R, Vivaldi A, et al. Clinical significance of serum mesothelin in patients with mesothelioma and lung cancer. Clin Cancer Res 2007;13:5076 –5081. Grigoriu BD, Scherpereel A, Devos P, et al. Utility of osteopontin and serum mesothelin in malignant pleural mesothelioma diagnosis and prognosis assessment. Clin Cancer Res 2007;13:2928 –2935. Schneider J, Hoffmann H, Dienemann H, et al. Diagnostic and prognostic value of soluble mesothelin-related proteins in patients with malignant pleural mesothelioma in comparison with benign asbestosis and lung cancer. J Thorac Oncol 2008;3:1317–1324. Pass HI, Wali A, Tang N, et al. Soluble mesothelin-related peptide level elevation in mesothelioma serum and pleural effusions. Ann Thorac Surg 2008;85:265–272. Creaney J, Yeoman D, Naumoff LK, et al. Soluble mesothelin in effusions: a useful tool for the diagnosis of malignant mesothelioma. Thorax 2007;62:569 –576. Cappia S, Righi L, Mirabelli D, et al. Prognostic role of osteopontin expression in malignant pleural mesothelioma. Am J Clin Pathol 2008; 130:58 – 64. Wittekind C, Greene FL, Hutter RVP, et al. TNM Atlas: illustrated guide to the TNM/pTNM classification of malignant tumours. Berlin: Springer, 2007:393. Iwahori K, Osaki T, Serada S, et al. Megakaryocyte potentiating factor as a tumor marker of malignant pleural mesothelioma: evaluation in comparison with mesothelin. Lung Cancer 2008;62:45–54. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228 –247. Beyer HL, Geschwindt RD, Glover CL, et al. MESOMARK: a potential test for malignant pleural mesothelioma. Clin Chem 2007;53:666 – 672. Mandrekar JN, Mandrekar SJ, Cha SS. Cutpoint determination methods in survival analysis using SAS. SAS Users Group International Proceedings. Seattle, Washington, 2003:261–228. Shiomi K, Hagiwara Y, Sonoue K, et al. Sensitive and specific new enzyme-linked immunosorbent assay for N-ERC/mesothelin increases its potential as a useful serum tumor marker for mesothelioma. Clin Cancer Res 2008;14:1431–1437. Onda M, Nagata S, Ho M, et al. Megakaryocyte potentiation factor cleaved from mesothelin precursor is a useful tumor marker in the serum of patients with mesothelioma. Clin Cancer Res 2006;12:4225– 4231. Hollevoet K, Van Cleemput J, Thimpont J, et al. Serial measurements of mesothelioma serum biomarkers in asbestos-exposed individuals: a prospective longitudinal cohort study. J Thorac Oncol 2011;6:889 – 895. Duffy MJ, Bonfrer JM, Kulpa J, et al. CA125 in ovarian cancer: European Group on Tumor Markers guidelines for clinical use. Int J Gynecol Cancer 2005;15:679 – 691. Hollevoet K, Bernard D, De Geeter F, et al. Glomerular filtration rate is a confounder for determination of soluble mesothelin in serum. Clin Chem 2009;55:1431–1433. Launay-Vacher V, Rey JB, Isnard-Bagnis C, et al. Prevention of cisplatin nephrotoxicity: state of the art and recommendations from the European Society of Clinical Pharmacy Special Interest Group on Cancer Care. Cancer Chemother Pharmacol 2008;61:903–909. El-Tanani MK. Role of osteopontin in cellular signaling and metastatic phenotype. Front Biosci 2008;13:4276 – 4284. Tajima K, Ohashi R, Sekido Y, et al. Osteopontin-mediated enhanced hyaluronan binding induces multidrug resistance in mesothelioma cells. Oncogene 2010;29:1941–1951. Ohashi R, Tajima K, Takahashi F, et al. Osteopontin modulates malignant pleural mesothelioma cell functions in vitro. Anticancer Res 2009; 29:2205–2214.

Copyright © 2011 by the International Association for the Study of Lung Cancer

1937