Prognostic value of disseminated tumor cells in unresectable pancreatic ductal adenocarcinoma: a prospective observational study

Abstract

Background: Although pancreatic ductal adenocarcinoma (PDAC) rarely metastasizes to the skeleton, disseminated tumor cells have been detected in bone marrow samples from patients with this disease. The prognostic value of such findings is currently unclear. Thus, the current study aimed to clarify the prognostic information associated with disseminated tumor cell detection in samples from patients with PDAC.

Methods: Bone marrow aspirates were obtained from 48 patients with locally advanced (n = 11) or metastatic (n = 37) PDAC, before and after 2 months of chemotherapy. Disseminated tumor cells were detected with an mRNA panel and quantitative reverse transcription PCR. We used the highest levels measured in healthy bone marrow (n = 30) as a threshold to define the positive detection of disseminated tumor cells. Progression-free and overall survival were analyzed with Kaplan-Meier and Cox proportional hazards regression analyses.

Results: Disseminated tumor cells were detected in 15/48 (31%) bone marrow samples obtained before starting chemotherapy and in 8/25 (32%) samples obtained during chemotherapy. Patients with disseminated tumor cells detected before therapy had significantly shorter progression-free (p = 0.03; HR = 2.0) and overall survival (p = 0.03; HR = 2.0), compared to those without disseminated tumor cells in the bone marrow. When restricting disseminated tumor cell detection to keratins KRT7 and KRT8, the prognostic information was substantially stronger (p = 1 × 10^-6; HR = 22, and p = 2 × 10^-5; HR = 7.7, respectively). The multivariable Cox regression analysis demonstrated that disseminated tumor cell detection prior to treatment had independent prognostic value. In contrast, disseminated tumor cells detected during treatment did not have prognostic value.

Conclusions: Disseminated tumor cells detected before commencing chemotherapy had prognostic value in patients with inoperable PDAC.

Keywords: Bone marrow; DTC; Disseminated tumor cells; PDAC; Pancreatic ductal adenocarcinoma; Prognosis; Survival.

Fig. 1

Relative concentrations of DTC markers in bone marrow samples. The relative concentrations of 9 DTC marker mRNAs in bone marrow samples are compared between healthy controls (yellow) and patients with PDAC (blue).The highest level of each mRNA observed in control samples is indicated with a horizontal black line segment. Measurements below detection limits are not shown (5 measurements)

Fig. 2

Correlations between relative concentrations of DTC markers. The scatterplots on the left side of the matrix show the median-normalized relative concentrations of DTC markers, plotted on logarithmic axes. Pearson correlation coefficients are shown on the right side of the matrix; the colors and intensities indicate the direction and strengths of the correlations (green: R = 1, white: R = 0, blue: R =  − 1). Statistical significance is indicated with asterisks (***p < 0.001; *p < 0.05). The plots placed on the diagonal show the data densities of the mRNAs indicated in the top bar

Fig. 3

Kaplan–Meier survival estimates stratified by positive/negative DTC detection in bone marrow. A, C, E Progression-free survival and B, D, F overall survival estimates are shown for patients that showed positive (blue) or negative (orange) DTC detection in bone marrow samples acquired A-D before chemotherapy (BM1) or E and F during chemotherapy (BM2). DTC detection was based on all epithelial markers (A, B, E, F) or restricted to KRT7 and KRT8 mRNAs (C, D). P-values were calculated with log-rank tests; the numbers at risk are shown in the panels below each survival curve