Markers for the non-invasive diagnosis of mesothelioma: a systematic review

Abstract

Background: Numerous markers have been evaluated to facilitate the non-invasive diagnostic work-up of mesothelioma. The purpose of this study was to conduct a structured review of the diagnostic performance of non-invasive marker tests for the detection of mesothelioma in patients with suspected mesothelioma.

Methods: Studies on the diagnostic accuracy of serum and cytological markers published till 31 December 2009, available in either PUBMED or Embase, to detect or exclude the presence of mesothelioma were extracted. Study quality was assessed with use of the Quadas criteria.

Results: In total, 82 articles were included in this systemic review. Overall, quality of the incorporated studies to address our objective was poor. The most frequently studied immunohistochemical markers for cytological analysis were EMA, Ber-Ep4, CEA, and calretinin. The most frequently investigated serum marker was soluble mesothelin-related protein (SMRP). The markers CEA, Ber-EP4, and calretinin were most valuable in discriminating mesothelioma from other malignant diseases. Markers EMA and SMRP were most valuable in discriminating mesothelioma from non-malignant diseases. No marker performed well in discriminating between mesothelioma and all other diseases.

Conclusion: Currently, there is only limited evidence to properly assess the value of non-invasive marker tests in the diagnosis of mesothelioma. Studies were of limited value to address our objective and results showed considerable unexplained study heterogeneity.

Figure 1

Flowchart of the selection of the relevant articles. Serum and effusion markers include tests to detect serum and effusion marker levels; immunohistochemical markers include marker tests used for cytological analysis of effusion samples; genetic markers include polymerase chain reaction tests to detect specific gene expressions or fluorescence in situ hybridisation (FISH) tests to detect gene deletions with the use of specialised gene probes.

Figure 2

Summary of quality of the included studies according to the Quadas criteria (see Supplementary Appendix).

Figure 3

Sensitivity against 1-specificity in ROC space to discriminate mesothelioma from other malignant diseases. The height of the blocks is proportional to the reciprocal of the number of mesothelioma patients (mesothelioma yes subjects) and the width of the blocks is proportional to the reciprocal of the number of patients with other malignant diseases (mesothelioma no subjects).

Figure 4

Sensitivity against 1-specificity in ROC space to discriminate mesothelioma from non-malignancy. The height of the blocks is proportional to the reciprocal of the number of mesothelioma patients (mesothelioma yes subjects) and the width of the blocks is proportional to the reciprocal of the number of non-malignant patients (mesothelioma no subjects).