Clinical utility of FDG-PET/CT for post-surgery surveillance of malignant pleural mesothelioma – Comparison with contrast-enhanced CT

Objectives To assess the diagnostic accuracy of fluorodeoxyglucose (FDG)-positron emission tomography/computed tomography (PET/CT) findings for recurrent malignant pleural mesothelioma (MPM) after a radical surgery procedure and their impact on clinical management in comparison with contrast-enhanced CT. Results Treatment failure was confirmed in 40 patients. The patient-based area under the receiver-operating characteristic (ROC) curves (AUC)/sensitivity/specificity/accuracy were 0.915/90.0%/80.0%/88.0% for FDG-PET/CT, and 0.805/75.0%/90.0%/78.0% for contrast-enhanced CT, respectively. AUC and sensitivity values were significantly different between the modalities (both p=0.041). Patient-based AUC values for diagnosing locoregional recurrence (ipsilateral hemithoracic recurrence) and distant metastasis, including peritoneal dissemination and lung, bone, muscle, and liver metastasis, were also significantly different (p=0.023 and p=0.035, respectively). The findings of FDG-PET/CT resulted in a change of management for 14 of the 50 patients (28%) by initiating new treatment. Of six patients judged as not having recurrence by contrast-enhanced CT but truly having recurrence based on FDG-PET/CT findings, 4 patients received new treatment due toFDG-PET/CT. Methods Fifty patients who underwent radical surgery for MPM received FDG-PET/CT and contrast-enhanced neck/chest/abdomen/pelvis CT examinations for surveillance or suspected recurrence within a 2-week period. Diagnostic ability was determined on a patient and lesion-site basis by 2 experienced examiners, and the modalities were compared using ROC analysis and McNemar test results. Lesion status was determined on the basis of histopathology, radiological imaging and clinical follow-up for longer than 6 months. Conclusion FDG-PET/CT findings were shown to be more accurate for assessing MPM recurrence and more often led to therapy change than contrast-enhanced CT.


INTRODUCTION
Advances in both surgical and nonsurgical therapeutic strategies for cancer continue, cancer recurrence and distant metastasis following the initial treatment remain as major issues for patients with malignant pleural mesothelioma (MPM). Early and accurate detection of recurrence in these patients has been shown to have an important influence on therapy choices, with selection of an appropriate subsequent treatment strategy expected to significantly impact survival [1].
Contrast-enhanced computed tomography (CT) is a widely used modality for assessing treatment failure in MPM patients. Unfortunately, those results are not always specific enough to distinguish between a recurrent tumor and benign post-therapeutic changes. Recent studies have examined use of fluorodeoxyglucose (FDG)-positron emission tomography/computed tomography (PET/CT) for detecting recurrent [2][3][4]. This imaging modality detects increased utilization of glucose by malignant cells, thus revealing their high glucose uptake, making diagnosis of cancer recurrence and distant metastasis in the preclinical stage possible earlier than with more conventional techniques. As a result, FDG-PET/CT results are considered to have a higher level of accuracy for diagnosis and restaging of MPM than contrast-enhanced CT. In the present study, we assessed FDG-PET/CT regarding its clinical usefulness for diagnosis of recurrent and metastatic MPM, and compared the findings with those obtained using contrast-enhanced CT. Table 1 shows patient demographics and clinicopathologic variables. Forty (80%) of the 50 enrolled patients had confirmed treatment failure. Tumor recurrence confirmation after 2 imaging examinations scans was determined by biopsy (n=10), cytology of ascites (n=2), or average clinical and imaging follow-up (n=28) findings over a period of at least 6 months. Four representative cases are presented in Figures 1-4. Using the FDG-PET/CT results, the value for patient-based AUC was 0.915, and values for sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy for detecting MPM recurrence were, 90.0% (36/40), 80.0% (8/10), 94.7% (36/38), 66.7% (8/12), and 88.0% (44/50), respectively, while those for contrast-enhanced CT were 0.805, 75.0% (30/40), 90.0% (9/10), 96.8% (30/31), 47.4% (9/19), and 78.0% (39/50), respectively(Table 2) ( Figure 5). AUC and sensitivity values with FDG-PET/CT were significantly higher as compared to those with contrast-enhanced CT (p=0.041 and p=0.041, respectively).

Patient-based diagnostic analysis
Based on FDG-PET/CT findings, there were 4 false-negative and 2 false-positive cases. As for the false-negative cases, there was 1 each of tiny contralateral pleural dissemination, tiny peritoneal dissemination, tiny ipsilateral pleural dissemination and lymph node metastasis, and tiny ipsilateral pleural dissemination and peritoneal dissemination, while the false-positive cases included 1 for which FDG-PET/ CT showed increased FDG uptake (SUVmax 6.04) in the right cardiophrenic angle that was misjudged as ipsilateral pleural dissemination (locoregional recurrence) and 1 with a surgery-proven peritoneal desmoid with moderate FDG uptake (SUVmax 3.89) that was misjudged as peritoneal dissemination in FDG-PET/CT findings.
The values for AUC, sensitivity, and accuracy of FDG-PET/CT for diagnosing ipsilateral hemithoracic recurrence (locoregional recurrence) were significantly higher as compared to those for contrast-enhanced CT (p=0.023, p=0.026, and p=0.0077, respectively). Additionally, the accuracy of FDG-PET/CT for diagnosing lymph node involvement was significantly higher (p=0.041). As for sensitivity, that of FDG-PET/ CT for diagnosing bone and muscle metastasis tended to be higher than that of contrast-enhanced CT, though the difference was not significant.
Patient-based diagnostic performance for thoratic recurrence, lymph node involvement, and distant metastasis showed similar tendency. The only AUC for diagnosing thoratic recurrence in a EPP group reached the significant difference between FDG-PET/CT and contrastenhanced CT (p=0.031)."

Impact on patient management
FDG-PET/CT findings impacted the management of 14 (28.0%) of the 50 patients, including initiation of chemotherapy in 8, radiotherapy in 3, resection in 2, and radiofrequency ablation in 1 as a new treatment modality. Six patients judged to not have recurrence based on contrast-enhanced CT findings were shown to have recurrence in FDG-PET/CT findings, of whom 2 had locoregional recurrence (ipsilateral hemithoracic recurrence), 1 bone metastasis, 1 muscle metastasis, 1 nodal involvement, and 1 lung and bone metastasis, and locoregional recurrence (ipsilateral hemithoracic recurrence). Two of those patients underwent sequential chemotherapy treatments, thus FDG-PET/CT was considered to have had an influence on additional therapeutic management (chemotherapy) in 4 cases.

DISCUSSION AND CONCLUSION
Using FDG-PET/CT findings, the patient-based value for AUC was 0.915, while those values for (C) The PET/CT value for AUC in cases with nodal involvement tended to be greater than that for contrast-enhanced CT, though the difference was not significant (p=0.1). (D) The PET/CT value for AUC (0.957) in cases with distant metastasis was significantly greater than that (0.852) for contrast-enhanced CT (p=0.035).   Previous studies evaluated FDG-PET/CT diagnostic performance for detection of MPM recurrence following a radical surgery procedure, including extrapleural pneumonectomy (EPP) and pleurectomy/decortication (P/D) [2][3][4]. Those authors reported patient-based sensitivity, specificity, PPV, NPV, and accuracy values of FDG-PET/CT for detecting MPM recurrence following radical surgery ranging from 94-98%, 75-100%, 86-100%, 83-95%, and 94-98%, respectively, similar to those found in the present cohort. In an investigation that included 57 patients, of whom 35 underwent surgery, Niccoli-Asabella et al. [4] noted those patient-based values for detecting MPM recurrence after various treatments as 98.1%, 100%, 100%, 83.3%, and 98.2%, respectively, for FDG-PET/CT, and 96.2%, 20.0%, 92.5%, 33.3%, and 89.4%, respectively, for contrast-enhanced CT. Again, the present findings were similar and demonstrate that FDG-PET/CT is more accurate for diagnosis of MPM recurrence in patients who have undergone radical surgery as compared to contrast-enhanced CT. As for our cohort, FDG-PET/CT findings had an impact on management, as they indicated the need for new treatment in 14 (35%) of 40 patients with confirmed failure. Also, Gerbaudo et al. [3] reported that FDG-PET/CT findings were helpful to select 12 (29%) of their 42 patients for an additional previously unplanned treatment because of evidence of initial treatment failure.
To our knowledge this is the first study to compare the diagnostic value of FDG-PET/CT and contrastenhanced CT between EPP and P/D groups separately. A similar tendency was observed in between in EPP and P/D groups and the only AUC for diagnosing thoratic recurrence in a EPP group reached the significant difference between FDG-PET/CT and contrast-enhanced CT. Although FDG-PET/CT may be more valuable to detect thoratic recurrence after EPP than after PD, this should be confirmed in a prospective and multicenter study with large number of patients." The present study has some limitations, including the relatively few patients, who were enrolled from a single institution. There also may have been bias regarding patient selection because of the retrospective design and strict criteria for patient enrollment. The gold standard for analysis of this type is histological confirmation of results. Nevertheless, clinical follow-up examination results are considered valid for diagnostic accuracy and therapy response evaluations, and it would have been unethical to utilize an invasive procedure to investigate all PET/CT-detected lesions. Finally, a consensus method for determining results, as employed in this study, is not preferred, with the ideal being independent determinations with formal interobserver variability.
In conclusion, the present results showed FDG-PET/CT superior for assessment of MPM recurrence. Additionally, those findings led to a change in subsequent appropriate therapy more often than findings obtained with contrast-enhanced CT.

Patients
From May 2006 to December 2017, 149 patients underwent a radical surgery procedure at our institution for histopathologically proven MPM, including EPP and P/D. Thereafter, 131 were regularly followed on an outpatient basis and received contrast-enhanced CT examinations for surveillance, which included 110 who also received whole-body FDG-PET/CT examinations. Following surgery, most of the patients received a contrast-enhanced CT examination every 3 months and FDG-PET/CT examination every 6 or 12 months for surveillance. For the present study, 50 patients (40 males, 10 females; average 62.5 years old, range 37-73 years old) who had received both whole-body FDG-PET/CT and contrast-enhanced neck/chest/abdomen/pelvis CT examinations for surveillance or suspected recurrence within 2 weeks were enrolled. Of those, 31 underwent contrast-enhanced CT first, followed by FDG-PET/CT, while 19 underwent FDG-PET/CT and then contrastenhanced CT. The mean interval between surgery and the initial imaging scan (FDG-PET/CT or contrast-enhanced CT) was 416.3±115.6 days (range 169-612 days). Tumor recurrence confirmation after the 2 imaging examinations was determined from histologic or cytologic analysis findings, or after at least 6 months of clinical and imaging follow-up examinations, such as contrast-enhanced CT, FDG-PET/CT, and/or magnetic resonance imaging (MRI).

FDG-PET/CT and contrast-enhanced CT examinations
The FDG-PET/CT examinations were performed using a PET/CT scanner (Gemini GXL16 or Gemini TF64; Philips Medical Systems, Eindhoven, The Netherlands) equipped with a gadolinium oxyorthosilicate detector. Details of the FDG-PET/CT procedures have been described [5]. Briefly, patients were instructed to fast for 5 hours before the scan, then blood glucose was measured immediately prior to injection of FDG at 4.0 MBq/kg body weight for the GXL16 and 3.0 MBq/kg for the TF64. All in the present cohort had a blood glucose level lower than 160 mg/dL. Approximately 60 minutes after the injection, static emission images were obtained. Helical CT scan images from the top of the head to the bottom of the feet were obtained for attenuation correction and anatomic localization, using the following parameters: tube voltage 120 kV, effective tube current auto-mA up to 120 mA (GXL16) or 100 mA (TF64), gantry rotation speed 0.5 seconds, detector configuration 16×1.5 mm (GXL16) or 64×0.625 mm (TF64), slice thickness 2 mm, and transverse field of view 600 mm. Immediately upon completion of the CT examination, PET images from the head to the mid-thigh were acquired for 90 seconds per bed position and from the mid-thigh to the toes for 30 seconds per bed position using a variable sampling method. Next, images at 12-14 bed positions for 90 seconds each and 6-7 bed positions for 30 seconds each were obtained in 3-dimensional mode. Thus, 22 to 24-26 minutes of emission scanning was required for each patient. During PET scanning, the patients were allowed to breathe normally.
Pre-contrast and contrast-enhanced CT images of the neck, chest, abdomen, and pelvis were obtained using a 128-detector row CT (SOMATOM Definition AS) at 120 kV, with an effective mA of 220 (CAREDose4D), beam pitch of 0.6, collimation of 1.2×32 mm, and B31 + medium smooth + image reconstruction. The detailed procedures for contrast-enhanced CT have been described [5]. Briefly, the level of blood creatinine was checked prior to the examination and did not exceed 1.5 mg/dL in any of the patients. Using a power injector, iodinated contrast material (Iopamiron Inj, Syringe, Bayer Schering Pharma, Berlin, Germany) containing 300 mg of iodine per ml at a dose of 600 mg of iodine per kg of body weight was intravenously administered, then scanning was started at 120 seconds after the injection.

Image analysis
FDG-PET/CT images were interpreted on a retrospective basis and consensus determinations made by 2 board-certified nuclear medicine physician and radiologists with 10 and 5 years of experience, respectively, in FDG-PET/CT, and no knowledge of the other imaging results or clinical data for the present cohort. A recurrent or metastatic lesion was diagnosed when abnormal focal FDG uptake shown by PET corresponded to an abnormal mass revealed by CT. Lymph nodes with increased glucose uptake were considered positive for metastatic spread, even in cases when the short-axis diameter was less than 1 cm [6], whereas cases with no detectable tracer uptake were considered negative for metastatic spread even when the short-axis diameter was 1 cm or greater.
Semiquantitative analysis of abnormal radiotracer uptake for each lesion was also performed using SUVmax values, calculated as follows: SUV = volume of interest radioactivity concentration (Bq/mL)/[injected dose (Bq)/ patient weight (g)]. SUVmax was defined as the greatest SUV value for pixels with the highest count within the VOI, with the highest for each used as the representative value of each scan. GI-PET (AZE Co., Ltd., Tokyo, Japan), a commercially available software package able to harmonize SUV values across different PET/CT systems [7], was employed.
Contrast-enhanced CT images were retrospectively evaluated and consensus determinations made by 2 board-certified radiologist and nuclear medicine physicians, with 17 and 10 years of experience, respectively, with CT, and no knowledge of the other imaging results or clinical data used in the present study. Images of coronal, axial, and sagittal sections were viewed and analyzed, with appropriate winding applied. A lymph node with a short-axis diameter greater than 1 cm was defined as malignant. A central unenhanced area suggesting central necrosis was also considered to be a sign of malignancy, whereas the presence of peripheral low attenuation suggesting a fatty hilum within a lymph node was considered to indicate benign, regardless of node size [6].
The diagnostic ability of the scanning modalities was determined for each individual patient as well as 8 different lesion sites, including ipsilateral hemithorax (locoregional recurrence), contralateral hemithorax, lymph node, peritoneum, lung, bone, muscle, and liver, with a 5-point grading system (1, definitely absent; 2, probably absent; 3, indeterminate; 4, probably present; 5, definitely present) utilized [8]. Each set of data was reviewed in random order after an interval of at least 4 weeks to avoid decision threshold bias. www.oncotarget.com

Statistical analysis
Patient-and lesion site-based analyses of the PET/ CT results obtained by consensus in comparison with contrast-enhanced CT findings were performed. PET/ CT and contrast-enhanced CT results for each region were also compared with the true lesion status (gold standard), then classified as true positive, false positive, true negative, or false negative. Using standard statistical formulae, we then determined sensitivity, specificity, PPV, NPV, and accuracy, with the 95% confidence interval (95% CI) value calculated for each parameter. For sensitivity and specificity, a score of 4 of 5 was considered to indicate positive, because of higher diagnostic accuracy. A McNemar test was used for differences between the 2 imaging modalities, with a p value less than 0.05 considered to indicate statistical significance. All analyses were performed using the SAS software package, version 9.3 (SAS Institute, Cary, NC).