Circulating levels of cell adhesion molecule L1 as a prognostic marker in gastrointestinal stromal tumor patients

The study was approved by the Ethics Committee of the Chamber of Physicians in Hamburg, Germany. Written informed consent was obtained from all patients for using tissue and serum samples. For this study, 93 patients with GIST were chosen retrospectively. It was possible to evaluate a rather large number of patients suffering from this rare tumor because of the cooperation of a self-help organization for GIST patients called "Lebenshaus e.v.", Bad Nauheim, Germany, which helped to collect serial serum samples from patients across Germany, Austria and Switzerland. The blood samples were taken from patients at time of diagnosis and every 3 months thereafter. We selected patients on the basis of availability of specimens and did not stratify them due to rare occurrence and different treatment strategies. All data including sex, age, tumor size, mitotic count, tumor location, metastasis, relapse, and imatinib therapy were obtained from the clinical and pathological records. The clinicopathological data from externally treated patients participating in the self-help project were sent regularly by their attending physicians. As healthy controls, 151 blood bank donors were included in the study. Limitations in this study were that we had some missing items in the clinical and pathological data of our patients due to logistic difficulties in collecting data of patients from different hospitals. However, a sufficient number of GIST patients were included in the study.

For the detection of soluble L1, 96-well flexible microtiter plates (Costar 9019) were coated with 50 μl per well of 10 μg/ml of capturing antibody (anti-L1 mAb 5G3, BD Bioscience) overnight at 4°C. Wells were blocked with 3% w/v bovine serum albumin (BSA; Fraktion V, 98% purity, Sigma Aldrich, Munich, Germany) in PBS/T (phosphate buffered saline, pH 7.3, containing 0.05% v/v Tween) for 45 min at room temperature and then incubated for 1 h with human sera diluted 1:5 in PBS at room temperature. After five washes with PBS/T, bound protein was detected with biotin-conjugated monoclonal antibody UJ127 (Dianova, Hamburg, Germany)- which does not cross-react with the close homolog of L1 (chL1) - followed by streptavidin-conjugated peroxidase using TMB (3,3', 5,5"-tetramethylbenzidine) as substrate. The color reaction was stopped by the addition of 10 mM H₂SO₄ and analysed at 450 nm using an ELISA reader. Human L1-Fc protein served as an internal standard for the assay.

To ensure that the immunoassay was suitable for measuring clinical serum samples, reproducibility, linearity and crossreactivity were examined. The assay showed no crossreactivity with the closed homolog of L1 CHL1, excellent linearity with serial dilutions and a low coefficient of variation (CV) in intra- and inter-assay variability.

For analysis of L1cam protein, PBS washed GIST882 cells were lysed in RIPA buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM Na₄P₂O₇, 1 mM NaF, 1 mM EDTA, 2 mM Na₃VO₄, 1 mM PMSF, 1% NP-40 plus complete EDTA-free protease inhibitor mixture, pH 7.5) for 30 min at 4°C and then subjected to SDS-Page and Western blot analysis. Equal amount of protein of each sample was mixed with 2x Laemmli Buffer (4% SDS, 20% glycerol, 10% 2-mercaptoethanol, 0.004% bromphenol blue, 0.125 M Tris HCl, pH 6.8) and boiled at 95°C for 5 min. Protein extracts were separated by 4 - 12% SDS-PAGE (NuPage, Invitrogen, Karlsruhe, Germany) and transferred onto nitrocellulose membrane (Protran; Whatman Schleicher and Schuell Dassel, Germany). For immunoblotting, membranes were blocked with StartingBlock (TBS) Blocking Buffer (Pierce, Rockford, IL, USA) and incubated overnight at 4°C with primary 1:1000 biotinylated antibody L1-UJ127 (Dianova, Hamburg Germany). UJ127 binds to a membrane proximal Fn repeat. This allows detection of both the ADAM-cleaved 180 kDa shed fragment and the 80 kDa membrane-bound fragment presumably generated by plasmin cleavage. After washes with TBST (TBS with 0.1% Tween 20), membranes were incubated with HRP-conjugated streptavidin (Sigma Aldrich, Munich, Germany) (1:500) for 1 h at room temperature. After extensive washes, streptavidin-reactive bands were visualized using enhanced chemiluminescent substrate (ECL, Pierce) on x-ray films (Kodak Biomax-ML; Sigma-Aldrich, Munich, Germany).

SPSS^® for Windows (Version 11.5.1) (SPSS^® Inc., Chicago, IL, USA) was used for statistical analysis. Statistical significance was evaluated by the Mann-Whitney U and Kruskal-Wallis test. Cross-table statistics were performed using Fisher's test. Recurrence-free survival was plotted using the Kaplan-Meier method and analysed by the log-rank test. Significance statements refer to p-values of two-tailed tests less than 0.05.

A total of 93 patients suffering from GIST and 151 apparently healthy controls were included in our study. The median age of GIST patients was 61 years (range 28-81 years) and 47 years (range 30-70 years) for the healthy controls. Fifty-one (55%) of the GIST patients were male and 42 (45%) female, Ninety-eight (65%) of the healthy controls were male and 53 (35%) female. Patients` characteristics and clinicopathological data are listed in Table 1.

L1 could be detected in sera of GIST patients, representatively shown in Figure 1. In sera proteolytically cleaved soluble L1 fragments as well as full-length L1 were specifically recognised by the monoclonal antibody UJ127, which was also used in the ELISA, underlining the specificity of the assay. Lysate of the cell line GIST882 [26] was used to visualise membrane bound or associated L1 compared to soluble cleavage products. In patient sera, the shed extracellular domain of 180 kDa, generated by cleavage of ADAM10/17, is detectable but not the 80 kDa form generated by PC5 or Plasmin.

Soluble L1 (s-L1) concentrations in the sera of GIST patients were significantly higher than in healthy controls: median (25th and 75th percentiles) soluble L1 values were 1.4 (0.8 and 2.7 respectively) ng/ml in the GIST patients and 0.5 (0 and 1 respectively) ng/ml in the controls (p < 0.001; Mann-Whitney U test) (Figure 2). The mean (± SD) soluble L1 concentration was 2.1 ± 2.5 ng/ml for GIST patients and 0.7 ± 1.4 ng/ml for healthy controls. Detailed clinicopathological data are given in Table 1. In addition s-L1 levels were not elevated in other benign and malignant gastic tumors (n = 59, data not shown). There was no influence of sex, age, tumor size, mitotic count, tumor location, resection margin or imatinib treatment on soluble L1 concentrations.

Table 1 shows that GIST patients with recurrence had significantly higher median soluble L1 concentrations than those without. Soluble L1 concentration (25th and 75th percentiles) was 1.7 (0.9 and 3) ng/ml in GIST patients with recurrence and 1.1 (0.8 and 1.7) ng/ml in relapse-free patients (p = 0.049; Mann-Whitney U test).

The median follow-up time in the study was 37 months (range 0-273 months). For additional statistical analysis, patients with high (>2 ng/ml) and low (<2 ng/ml) soluble L1 concentrations were compared by survival curves plotted by the Kaplan-Meier method for recurrence-free survival (Figure 3). GIST patients with high soluble L1 levels showed a significantly worse outcome (38 months [95%-CI: 27; 50] of recurrence-free survival compared to 73 months [95%-CI: 56; 90] in patients with low soluble L1 levels (p = 0.016, log-rank test). Accordingly, the five-year recurrence-free survival rate among patients with low soluble L1 levels was 52%, while it was only 19% in patients with high soluble L1 levels. Absolute recurrence counts were significantly higher in patients with high soluble L1-levels (p = 0.025, Fischer's test, data not shown). We refrained from analysing overall survival because of low numbers (n = 4) of deaths in the patient group in this study.

Neither mitotic count nor tumor size had an influence on recurrence-free survival plotted by Kaplan-Meier analyses (p = 0.231 and p = 0.199, log rank test; data not shown). Also mitotic count and tumor size could not be significantly correlated with recurrence (p = 0.075 and p = 0.725, Fisher's test).