EURURO Vol. 71 No. 5

778 844

according to the manufacturer’s protocol. Beadchips were scanned using
a HiScan system (Illumina, San Diego, CA, USA) and analysis of array data
was performed using GenomeStudio software (version 2011.1, Illumina).
Samples were normalized by the cubic spline algorithm, and probes with
all signals lower than two times the mean background level were
excluded, leaving 21 675 probes for further analysis.
2.5. Multivariate modeling and univariate analysis
Principal component analysis (PCA), an unsupervised projection method,
was used to create an overview of the variation in data and to detect
clusters and trends among metastasis samples and expressed genes
Data were mean-centered and scaled to unit variance before
analysis. Models were validated via sevenfold cross-validation. Multi-
variate statistical analyses were performed in SIMCA version 14.0 (MKS
Umetrics AB, Umea˚ , Sweden).
Univariate analysis was applied to compare subgroups identified by
PCA with respect to differences in gene expression and clinical
characteristics. Groups were compared using the Mann-Whitney U-test
for continuous variables and the x2 for categorical variables. Univariate
statistical analyses were performed using SPSS 23.0 software (SPSS,
Chicago, IL, USA).
2.6. Functional enrichment analysis
Functional enrichment analysis was generated via Ingenuity Pathway
Analysis (IPA; ). IPA core analysis was used to
identify altered canonical pathways. The significance of associations
between lists of differently expressed genes and canonical pathways were
assessed using (1) the ratio of differentially expressed genes (molecules)
that map to a specific pathway, [14_TD$DIFF]given [15_TD$DIFF]in relation to the total number of
molecules included in the canonical pathway and (2) Fisher’s exact test to
determine the probability that the relationship between the molecules in
the data set and the canonical pathway is explained by chance.
Upstream analysis was used to identify regulators with a probability
of being responsible for the changes in gene expression observed, by
calculating an overlap p value with Fisher’s exact test and an activation
z-score. Details of the IPA core analysis can be obtained at
. Both upregulated and downregulated identifiers were submitted as
parameters for the analysis. IPA core analysis default settings were used,
but limited to the human knowledge base.
2.7. Real-time RT-PCR
Samples of 200 ng of total RNA were reversed transcribed using a
Superscript VILO cDNA synthesis kit (Invitrogen, Stockholm, Sweden) in a
total volume of 10 ml. Subsequent qRT-PCR analysis was performed using
TaqMan assays for quantification ofHLA-A, TAP1, and PSMB9 mRNA levels
(Hs01058806_g1, Hs00388675_m1, and Hs00160610_m1; Life Technol-
ogies, Stockholm, Sweden) on an ABI Prism 7900HT sequence detection
system according to the manufacturers’ protocols. Each sample was run in
duplicate and adjusted for the corresponding RPL13A mRNA level
(Hs01578912_m1, Life Technologies) using the ddCt method. Statistical
differences in mRNA levels between groups were identified using the
Kruskal-Wallis test followed by theMann-Whitney U-test. Paired samples
were compared using the Wilcoxon signed-rank test.
2.8. Immunohistochemistry
Tissue sections were deparaffinized in xylene and rehydrated in a
graded ethanol series. Immunohistochemistry was performed using an
Table 1 – Clinical characteristics for patients with prostate cancer or other malignancies who underwent surgery for metastatic spinal cord
compression
Prostate cancer Other malignancies
Castration-resistant Not treated Not [11_TD$DIFF]treated
Patients (n) 54 11 14
Age at diagnosis (yr) 69 (63-74) 76 (64-82) 67 (61-79)
Age at metastasis surgery (yr) 72 (67-79) 76 (64-82) 67 (61-79)
Serum PSA at diagnosis (ng/ml) 72 (36-530) 690 (82-2500) –
Serum PSA at metastasis surgery (ng/ml) 290 (85-780) 690 (82-2500) –
Gleason score at diagnosis –
6 4 (7.4) 1 (9.1)
7 19 (35) 1 (9.1)
8–10 20 (37) 1 (9.1)
Not available 11 (20) 8 (73)
Bicalutamide before surgery – –
Yes 24 (44)
No 30 (56)
Chemotherapy before surgery – –
Yes 9 (17)
No 45 (83)
Radiation before surgery – –
Yes 8 (15)
No 46 (85)
Follow-up after metastasis surgery (mo) 5.9 (2.0-15) 37 (24-72) 7.0 (3.6-16)
Data are presented as median (25th–75th percentile) for continuous variables and as number (precentage) for categorical variables. PSA = prostate-specific
antigen.
a Castration-resistant patients had disease progression after long-term androgen deprivation therapy including surgical ablation, luteinizing hormone–releasing
hormone/GNRH agonist therapy, and therapy with anti-androgens (bicalutamide).
b Other malignancies included kidney (n = 3), colorectal (n = 2), lung ( n = 1), liver (n = 1), unknown adenocarcinoma (n = 3), myeloma (n = 2), lymphoma (n = 1),
and sarcoma (n = 1).
c Chemotherapy included taxotere in six cases, estramustine in two cases, and taxotere, carboplatin, and etoposide in one case.
d Radiation towards operation site.
E U R O P E A N U R O L O G Y 7 1 ( 2 0 1 7 ) 7 7 6 – 7 8 7 778