Detection of Recurrence through microRNA-371a-3p Serum Levels in a Follow-up of Stage I Testicular Germ Cell Tumors in the DRKS-00019223 Study

Abstract

Purpose: Surveillance of clinical stage I (CSI) testicular germ cell tumors (GCT) is hampered by low sensitivity and specificity of current biomarkers for detecting relapses. This study evaluated if serum levels of microRNA371a-3p (M371 test) can: (i) Accurately detect relapses, (ii) detect relapses earlier than conventional technology, and (iii) if elevated postoperative M371 levels may predict relapse.

Experimental design: In a multicentric setting, 258 patients with testicular CSI GCT were prospectively followed by surveillance for a median time of 18 months with serial measurements of serum M371 levels, in addition to standard diagnostic techniques. Diagnostic characteristics of M371 for detecting relapses were calculated using ROC curve analysis.

Results: Thirty-nine patients recurred (15.1%), all with elevated M371 levels; eight without relapse had elevations, too. The test revealed the following characteristics: area under the ROC curve of 0.993, sensitivity 100%, specificity 96.3%, positive predictive value 83%, negative predictive value 100%. Earlier relapse detection with the test was found in 28%, with non-significant median time gain to diagnosis. Postoperative M371 levels did not predict future relapse.

Conclusions: The sensitivity and specificity of the M371 test for detecting relapses in CSI GCTs are much superior to those of conventional diagnostics. However, post-orchiectomy M371 levels are not predictive of relapse, and there is no significant earlier relapse detection with the test. In all, there is clear evidence for the utility of the M371 test for relapse detection suggesting it may soon be ready for implementation into routine follow-up schedules for patients with testicular GCT.

Figure 1.

Study profile. Outline of the selection process of patients with CSI GCT upon enrollment.

Figure 2.

Performance characteristics of M371 test. ROC curve graphically displays the excellent sensitivity and specificity of the M371 test to detect relapses in patients with CSI GCT.

Figure 3.

Relapse detection with M371 during follow-up. A, Individual M371 measurements of relapsing and non-relapsing patients during follow-up (n = 258). Each dot represents the highest M371 measurement of each single patient at the corresponding time point of follow-up. Red dots denote relapsing patients, green dots non-relapsing. The dashed black line marks the cutoff. The y-axis is depicted in a logarithmic scale. Clearly, all relapsing patients are located beyond the cut-off line (elevated M371 levels), and none of the relapsing cases lies below the line, but very few non-relapses lie beyond the cut-off line (false-positives). B, Relapse-free survival curves showing time-points of relapse detection with M371 test and with conventional methods. The graph shows that time-points of relapse detection with the M371 test and with standard methodology are not different from each other (P = 0.956; Log rank test). Median time to relapse detection is 6 months with both methods as indicated by dashed lines. The hypothesis of earlier relapse detection with the M371 test is not substantiated by these data. Green curve: relapse detection with M371 test, red curve: relapse detection with imaging and classical tumor markers.

Figure 4.

Median M371 levels in patients with and without recurrence at different time points. The figure shows that M371 levels measured immediately after orchiectomy are not different among the subgroups of relapsing (n = 5, right side of image) and non-relapsing cases (n = 59, left side of image), respectively. The hypothesis that elevated M371 levels after orchiectomy may predict relapse is to be rejected by these data. As expected, last M371 measurements are significantly higher in patients with relapse than in those without. These data refer to a subsample of n = 64 cases of the entire study population. Blue boxes: median preoperative M371 levels; green boxes: levels measured immediately after orchiectomy; violet boxes: levels measured at last follow-up visit. The y-axis is depicted in a logarithmic scale. Error bars denote 95% CIs.