A prognostic hypoxia gene signature with low heterogeneity within the dominant tumour lesion in prostate cancer patients

Abstract

Background Gene signatures measured in a biopsy have been proposed as hypoxia biomarkers in prostate cancer. We assessed a previously developed signature, and aimed to determine its relationship to hypoxia and its heterogeneity within the dominant (index) lesion of prostate cancer. Methods The 32-gene signature was assessed from gene expression data of 141 biopsies from the index lesion of 94 patients treated with prostatectomy. A gene score calculated from the expression levels was applied in the analyses. Hypoxic fraction from pimonidazole immunostained whole-mount and biopsy sections was used as reference standard for hypoxia. Results The gene score was correlated with pimonidazole-defined hypoxic fraction in whole-mount sections, and the two parameters showed almost equal association with clinical markers of tumour aggressiveness. Based on the gene score, incorrect classification according to hypoxic fraction in whole-mount sections was seen in one third of the patients. The incorrect classifications were apparently not due to intra-tumour heterogeneity, since the score had low heterogeneity compared to pimonidazole-defined hypoxic fraction in biopsies. The score showed prognostic significance in uni-and multivariate analysis in independent cohorts. Conclusions Our signature from the index lesion reflects tumour hypoxia and predicts prognosis in prostate cancer, independent of intra-tumour heterogeneity in pimonidazole-defined hypoxia.

Integrated analysis of cervical squamous cell carcinoma cohorts from three continents reveals conserved subtypes of prognostic significance

“…For digital quantification scanned images of all sections at a high resolution of 0.46 um/pixel (20×), which was reduced to 0.92 um/pixel for analysis, were used. Digital score was calculated by quantifying the area fraction of stained CD8 cells in relation to the entire section in the digital assessment, as described previously for pimonidazole staining 118 . For a detailed description of this method together with representative staining images, please see Supplementary Methods.…”

ALDH1A1 drives prostate cancer metastases and radioresistance by interplay with AR- and RAR-dependent transcription

“…In contrast to ALDH1A1, ALDH1A3 expression in primary tumors does not significantly correlate with patients' outcomes ( Figure S3 B). Analysis of the gene expression dataset from an independent cohort of patients with primary intermediate or high-risk PCa (Oslo cohort 29 , n = 95) validated a negative correlation between ALDH1A1 and ALDH1A3 genes ( Figure S3 C). We also confirmed that ALDH1A3 negatively correlates with clinical parameters associated with cancer aggressiveness, such as pathological tumor stage, Gleason score and tumor size ( Figure S3 D).…”

“…In the current study, we also concerned the patient cohort that included 95 prostate cancer patients referred to robot-assisted laparoscopic radical prostatectomy (RALP) at Oslo University Hospital between October 2011 and May 2016 (Oslo cohort). All patients were enrolled in the FuncProst study (NCT01464216), and detailed clinical characteristics were described previously 29 . The following clinical parameters were taken for the correlation analysis: age at inclusion, cT1vsCT2vcCT3 (grouped clinical tumor stages), Glscorepat (Gleason score determined by pathologists after surgery), largest extent (largest extent histologically determined by the pathologists based on HE-stained whole-mount sections), N status pato PSA (lymph node status of patients; determined by pathological examination of lymph nodes when pelvic lymph node dissection (PLND) was performed.…”

“…The following clinical parameters were taken for the correlation analysis: age at inclusion, cT1vsCT2vcCT3 (grouped clinical tumor stages), Glscorepat (Gleason score determined by pathologists after surgery), largest extent (largest extent histologically determined by the pathologists based on HE-stained whole-mount sections), N status pato PSA (lymph node status of patients; determined by pathological examination of lymph nodes when pelvic lymph node dissection (PLND) was performed. In patients without PLND, negative MRI in combination with undetectable PSA at 6 weeks after surgery was regarded as nodal stage 0), PSA (PSA measured before surgery); pT2vs3vs4 (grouped pathological tumor stage); risk classification (D'amico risk classification: 1 = low, 2 = intermediate, 3 = high) 29 .…”

MiR-21 Is Induced by Hypoxia and Down-Regulates RHOB in Prostate Cancer

“…Indeed, a recent meta-analysis study identified miR-21 as one of several miRNAs that could predict responses to androgen-deprivation therapy [95]. Elsewhere, various research groups have developed hypoxic signature panels of gene markers to predict prostate cancer prognoses and treatment outcomes, although notably these did not include microRNAs [96][97][98][99]. Our research suggests that hypoxia-induced miRNAs should really be included in such panels, especially if blood-based markers are preferred.…”