Fig 1.
The patient screening process.
IP: irinotecan plus cisplatin; SSAs: somatostatin analogs.
Table 1.
Clinical characteristics and serum CgA levels of patients.
Fig 2.
Serum CgA levels in patients with GEP-NETs and healthy individuals.
There were higher CgA levels in patients with GEP-NETs compared with healthy individuals (Log10 scale). **P<0.01 (Mann–Whitney test).
Fig 3.
The ROC curve of serum CgA levels.
The sensitivity and specificity were 78.8% and 73.8% at the cutoff level of 46.2 ng/mL. The area under the curve (AUC) was 0.85 with 95% confidence interval of 0.79–0.91.
Fig 4.
Relationship between CgA level changes and clinical response.
(A) Of the twenty-eight patients that received IP chemotherapy as the first-line treatment, sixteen of twenty-three with non-progressive disease (PR and SD) showed decreased or stable CgA levels; three out of five with progressive disease (PD) exhibited increased CgA levels. (B) Of the eighteen patients who received somatostatin analogs as the first-line treatment, twelve of thirteen with non-progressive disease had decreased or stable CgA levels; two of the five patients having progressive disease showed increased CgA levels.
Fig 5.
Dynamic changes of CgA levels during treatment.
(A), Patient 1 was treated with IP regimen for three months (6 cycles) with partial response (PR) followed by 6 months (6 cycles) of somatostatin analogues with progressive disease (PD) and 3 cycles of IP regimen with PR. (B), Patient 2 was treated with 6 cycles of IP regimen with PR followed by 3 cycles of SSAs with PD and 3 cycles of IP regimen with PD. Patient 3 (C) and patient 4 (D) were treated with 6 cycles of IP regimen with SD followed by 6 cycles of SSAs. Patient 5 (E) and patient 6 (F) were treated with 6 cycles of IP regimen followed by 6 cycles of SSAs.
Fig 6.
Overall survival curve of patients based on CgA levels.
Patients with high CgA levels (≥46.2 ng/mL) had a significantly poorer survival than patients with low CgA levels (<46.2 ng/mL).