Green Area, Room 2 (Level 0)
The EORTC SURTIME trial explored a period of targeted therapy (sunitinib) prior to CN as an alternative approach to immediate CN. The sequence of CN and sunitinib did not affect the progression free rate (PFR) but a survival signal was seen for deferred CN. Previous retrospective studies suggest increased surgery related adverse events (AEs) after presurgical VEGF-targeted therapy. Here we report surgical safety from a randomized comparison of both sequences.
Patients with mRCC were randomized 1:1 to immediate CN followed by sunitinib versus 3 cycles sunitinib followed by CN and sunitinib. Inclusion required histologically confirmed clear-cell subtype, resectable asymptomatic primary tumour and <= 3 surgical risk factors (Culp et al., Cancer 2010). Due to poor accrual, the PFR at week 28 was reported (98 patients needed). We analysed in-hospital mortality, 30-day readmission/prolonged hospitalization and AEs related to surgery (intra- and 30-day postoperative AEs) according to CTCAE version 4. Postoperative AEs were additionally graded according to Clavien-Dindo.
The study closed after 5.7 years with 99 patients entered and a median follow-up of 3.3 years. In the immediate CN arm, 46 of 50 patients had CN. In the deferred CN arm, 40 of 48 had CN 24 hours after sunitinib. None of the primary tumours in the deferred arm became unresectable and only 2 patients had a sunitinib-related delay of CN > 2 weeks. AEs related to surgery (CTCAE v4 all grades and ≥ grade 3) occurred in 50.0% (95% CI 36.1-63.9) and 28.2% (17.3-42.5) of patients in the immediate arm and 45.0% (30.7-60.2) and 27.5% (16.1-42.8) of patients in the deferred arm, respectively, although the number of intraoperative AEs were higher in the immediate arm. Postoperative AEs Clavien-Dindo ≥ 3, 30-day readmission/ prolonged hospitalization and in-hospital mortality were 6.5%, 13% and 4.3% in the immediate arm and 2.5%, 7.5% and 2.5% in the deferred arm, respectively. There were no differences for surgery time, blood loss and hospitalization.
The deferred approach initiates therapy quickly and does not lead to inability to perform CN. CN after sunitinib appears safe with numerically less surgical AEs, readmissions and mortality than immediate CN.
During recent years, statin medication has shown beneficial effects as carcinogenesis inhibiting agent in laboratory studies, but population-based evidence is weak and plausible biases may affect studies concerning statin use and cancer. We studied the association between statin use and tumor extent of renal cell carcinoma (RCC) at diagnosis and the risk of RCC death while controlling for immortal time bias and selection bias.
Nationwide Finnish cohort of 13,873 participants diagnosed with RCC between 1995–2012 were analyzed retrospectively using register-based data. Information of RCC, statin use and comorbidities were compiled from The National Cancer Registry, the Social Insurance Institute of Finland and the Care Registry for Health Care. Hypertension, diabetes and obesity as either registered diagnoses or drug-purchases related to these conditions were adjusted in the analysis.
Tumor extent was dichotomized as localized/locally advanced vs. metastatic extent. Logistic regression model adjusted for age, gender and co-morbidities was used to analyze the risk association between statin medication and tumor extent. Cox regression with further adjustment for tumor extent and primary treatment was used to analyze the risk of RCC death. Statin use before and after RCC diagnosis was analyzed separately. To control immortal time bias post-diagnostic use was analyzed as time-dependent variable. Selection bias due to selective discontinuation of statin use in cancer patients with limited life-expectancy was controlled by keeping the statin users in the user-category from the first purchase. Dose-dependency of risk associations was estimated by calculating cumulative statin use for each participant and stratifying the participants accordingly.
5,179 patients died of RCC and 1,199 of them had statin medication. Median follow-up after RCC diagnosis was longer among statin users (7 vs 6 years, p=0.013). Pre-diagnostic statin use was not associated with tumor extent but with increased risk of RCC death (HR 1.03, 95% CI 1.03–1.03), driven by users ceasing statin use before the diagnosis (HR 1.68, 95% CI 1.48–1.90). Any post-diagnostic statin use was not associated with RCC survival, but this depended on the annual dosage of use; low-dose use (419 doses/year; HR 0.62, 95% CI 0.52–0.72). Limitation is possibility of residual confounding and missing information on smoking.
Retrospective studies include many possible sources of bias, which were controlled for using detailed nationwide data. Our study shows improved RCC survival in statin users compared non-users in long-term and high-dose use. Statin use may have benefit in RCC patients, but the risk association is not straightforward and high-quality retrospective studies are regarded before considering prospective studies.
Blue Area, Room 2 (Level 0)
The purpose of this study was to assess findings of intraoperative endoscopic ultrasound (EUS), ex vivo MRI and immunohistochemistry regarding the presence of a solitary aldosterone producing adenoma (APA) or multinodular hyperplasia (MNH). Also, the usability of endoscopic ultrasound was tested for future partial adrenalectomies.
Fifteen patients with primary aldosteronism caused by a solitary APA, based on preoperative CT scan and adrenal vein sampling, were treated by laparoscopic adrenalectomy. Intraoperative adrenal imaging was performed with EUS. Post-operative imaging of the specimen was performed using a 11.7T small animal magnetic resonance (MR) system. EUS images and MR images were compared to the preoperative CT scan and the histopathology report (including macroscopic slices, HE slides and immunohistochemistry for aldosterone synthesis). In diffusion-weighted MR images ADC values were calculated and compared between normal adrenal cortex and adrenal adenomas.
All fifteen APAs were clearly visualized and located with intraoperative endoscopic ultrasound. Besides these APAs, a thickened adrenal limb (suspicious for multinodularity) was seen in seven cases. Based on the histopathological report, five thickened limbs were presumably caused by MNH. One suspicion for multinodularity seemed to be an appendage of the APA and one suspicion was caused by optical illusion. Compared with the MRI report, multinodularity was missed by EUS in 8 glands. Using ex-vivo MRI, the radiologist classified 86,7% (13 of 15) of the adrenals as MNH, whereas on HE slides only, the pathologist initially classified 46,7% (7 of 15) of the adrenals as MNH. Combining ex-vivo MRI, macroscopic slices and HE slides the pathologist morphologically classified 12 of 15 adrenals as multinodular. Eight of 12 multinodular glands contained one single nodule positive for aldosterone production, which was always the largest nodule present. Three of 12 multinodular glands contained two nodules positive for aldosterone production. One multinodular gland contained six different aldosterone producing nodules. Mean ADC values of normal adrenal cortex were not different from mean ADC values of adrenal nodules (0,490 mm2/s vs 0,493 mm2/s).
Endoscopic ultrasound may have potential to detect the largest adrenal adenoma and can facilitate performing a partial adrenalectomy. Ex vivo MRI detects small adrenal gland micronodules that are frequently missed by routine histopathology. Some of these small nodules also produce aldosterone. Whether this leads to persistent PA after partial adrenalectomy remains to be investigated, for instance with genetic analysis. Therefore, radical adrenalectomy currently remains the standard treatment for patients with primary aldosteronism.
Blue Area, Room 5 (Level 0)
The host’s immune system plays a pivotal role in many tumour-types, including squamous cell carcinomas (SCCs). The complex tumour-microenvironment involves many immunological factors. These include 1) persistent infection with high-risk human papilloma virus (hrHPV), 2) tumour-antigen recognition via classical HLA class I by cytotoxic T-cells, and 3) tumour escape via checkpoint molecules (PD-L1), regulatory T-cells and M2-polarized macrophages. We aimed to identify immunological factors of penile SCC that were prognostic for lymph node metastases (LNM) and disease-specific survival (DSS).
Penile SCC patients (n = 213) treated in the Netherlands Cancer Institute between 2001 and 2009 of whom formalin-fixed, paraffin-embedded (FFPE) tumour material was available, were analysed. We used data of previously determined hrHPV (-/+), classical and non-classical human leukocyte antigen (HLA) class I (normal expression/partial downregulation/complete downregulation, and normal expression/upregulation, respectively), tumoural and stromal Programmed Death-Ligand 1 (PD-L1) expression (-/+), and tumour PD-L1-expression pattern (diffuse or predominantly at the tumour-stroma margin). Additional immunohistochemical measurements of cytotoxic and regulatory T-cells (CD8 and FoxP3 respectively), and macrophages (CD163) were included. Regression analyses were performed to investigate relationships of the immune parameters with LNM and DSS.
Results on hrHPV, HLA-expression and PD-L1-expression analyses were previously published, showing unfavourable associations for non-hrHPV induced tumours, tumours with downregulated classical HLA expression, and tumours with a diffuse PD-L1 expression pattern. In this combined and extended analysis, we confirmed our previous findings regarding hrHPV and PD-L1; patients with hrHPV- tumours had higher chance of disease-specific death in comparison to patients with hrHPV+ tumours; and that patients with diffuse PD-L1 expressing tumours had higher chance of disease-specific death and LNM in comparison to patients with negative/margin PD-L1 expressing tumours. Furthermore we showed that T-cells and macrophages had favourable (CD8) and unfavourable (FoxP3, CD163) associations in univariable and subgroup analyses. Both classical and non-classical HLA expression showed significant favourable (complete vs. partial downregulation of classical HLA class I) and unfavourable (expression of non-classical HLA class I) associations only in univariable analysis. In contrast to our previous study on HLA, both did not show associations with LNM or DSS in the multivariable model.
Established associations with LNM and decreased DSS for hrHPV- and diffusely PD-L1+ tumours remained significant when corrected for T-cells, macrophages and clinicopathological factors.
Squamous cell carcinomas (SCC) represent the majority of malignant penile tumours. About 50% of all SCCs are associated with human papillomaviruses (HPV), but the value of HPV in terms of prognosis is controversially discussed. Furthermore, the knowledge of molecular biological processesin tumourigenesis and metastasis with regard to HPV is limited. The aim of this international multicentre study is to examine prognostic markers for penile SCC.
We included 123 patients from Russia and Germany. The tumours have been classified according to the novel 2016 WHO classification. HPV genotyping and immunohistochemical staining of p16 was performed. Clinical data were correlated to the histological subtype and HPV status (high risk HPV negative or positive). RNA was isolated from FFPE tissues as well as paired normal tissue (skin from the penis). For each subtype, metastasised and non-metastasised tumours were selected. MicroRNA microarray analysis was performed using the Agilent platform. Selected miRNAs have been validated by qPCR.
In general, no difference in cancer specific (CSS) and progression free survival (PFS) was observed with regard to HPV. The CSS was associated with histological subtype, whereby the worst outcome was observed in HPV-related basaloid carcinomas and the best in non-HPV-related warty carcinomas. Considering the HPV status within the histological subtype, HPV-related patients exhibited a better CSS and PFS. MicroRNA-profiles of PeCa subtypes varied significantly. Within the subtypes, microRNAs distinguished between tumour and corresponding normal tissue as well as metastasised and non-metastasised tumours (usual type, 40 microRNAs; p<0.05; FC>1.5; basaloid carcinoma, 90 microRNAs; p<0.05; FC>1.5).
Our results indicate that the histological subtype combined with HPV status but not the HPV status alone determines the prognosis of the disease. The differentially expressed microRNAs could serve as molecular markers for prognosis and deliver further insights into the molecular pathogenesis of penile SCCs.
Red Area, Room 1 (Level 0)
Among molecular targeted therapies, tyrosine kinase inhibitors (TKIs) including sunitinib are most common first-line setting for the treatment of advanced renal cell carcinoma (RCC). However, TKIs therapies are not expected to have curative effects because they extend progression-free survival only slightly due to acquisition of resistance to these drugs. Recently, bromodomain family proteins such as BRD4 (Bromodomain containing 4), epigenetic regulator of acetylated histones in chromatin, have been identified as promising therapeutic targets, and its inhibitor JQ1 has shown inhibitory effects in several types of cancer. However, anti-cancer effects of JQ1 in RCC, especially sunitinib-resistant RCC is still unclear. The aim of this study is to elucidate anti-cancer effects of BRD4 inhibition by JQ1, and the mechanism underlying BRD4 inhibition in sunitinib-sensitive and -resistant RCCs.
BRD4 expressions in RCC cell lines including sunitinib-resistant 786-o RCC cell line (SU-R-786-o), which we previously established, were determined by qRT-PCR. To determine the anti-cancer efficiency of JQ1, Cell proliferation assays, cell cycle assays and apoptosis assays in vitro, and xenograft assays were performed. To elucidate molecular mechanisms regulated by JQ1 treatment, we performed RNA sequence analyses using JQ1 treated- or untreated- SU-R-786-o and parent 786-o cells. In addition, we performed statistical analyses based on The Cancer Genome Atlas (TCGA) database of RCC patients.
The expression level of BRD4 was significantly increased in RCC cell lines compared to normal kidney (P < 0.05). In addition, SU-R-786-o showed elevation of BRD4 expression in comparison with parent 786-o (P = 0.0014). TCGA database showed that high BRD4 expression group had poorer overall survival (OS) compared with medium and low BRD4 expression group (P = 0.0003), and BRD4 was an independent predictor for OS (P = 0.0063). JQ1 reduced cell growth in vivo and in vitro by inducing G0/G1 cell cycle arrest and apoptosis in SU-R-786-o cells more than in the parent cells (P < 0.005). Also, JQ1 suppressed MYC expression in SU-R-786-o cells (P < 0.0001). RNA sequence analysis revealed that several oncogenes were significantly downregulated in JQ1 treated- SU-R-786-o. Finally, among those oncogenes, we found that SCG5, SPOCD1, RGS19, and ARHGAP22 were novel independent predictors for OS in RCC by analyzing TCGA database (P = 0.0425, P = 0.0003, P = 0.0013, and P = 0.0179, respectively).
BRD4 inhibition by JQ1 could be a potential therapeutic approach in sunitinib-resistant RCC. We also found that MYC, known target of JQ1, as well as, novel independent predictors of RCC such as SCG5, SPOCD1, RGS19, and ARHGAP22 were suppressed by JQ1. The discovery of molecular targets regulated by BRD4 could provide important insights into potential mechanisms of sunitinib resistance in RCC.
Blue Area, Room 2 (Level 0)
Recent evidence suggests that patients with Prostate cancer (PC) and diabetes have an inferior oncologic outcome. The Androgen receptor (AR) is a well known major driver of PC growth and progression. The aim of the present study was to assess an association of type 2 diabetes with alterations in AR signaling in benign and malignant prostate tissue.
Tumor content of 73 prostate tissue samples of men with type 2 diabetes (42 cancer, 31 benign) and 78 samples of patients without diabetes (34 cancer, 44 benign) was quantified by an experienced pathologist. mRNA expression of AR, insulin receptor isoform A (IR-A) and B (IR-B), IGF-1 receptor (IGF1R), Cyp27A1 and Cyp7B1, PSA gene KLK3, PSMA gene FOLH1, and Ki-67 gene MKI67 were analyzed by RT-qPCR. Results were analysed in terms of cell signaling and correlated to the PC content of the respective sample.
There were substantial differences between patients with and without diabetes: AR expression was associated with tumor content exclusively in patients with diabetes (p=0.02). A clear correlation with downstream targets PSA and PSMA and increased cell proliferation was observed. Only in diabetes, AR expression was correlated to higher IR-A / IR-B ratio and lower IR-B / IGF1R ratio, thus, in favor of the mitogenic isoforms. Reduced Cyp27A1 and increased Cyp7B1 expressions in tumor suggest lower levels of protective estrogen receptor ligands in diabetes.
We report elevated androgen receptor signaling presumably due to altered insulin/IGF-1 receptors and decreased levels of protective estrogen receptor ligands in PC in men with diabetes. Our results reveal new insights why these patients have worse prognosis. These findings provide new hints towards therapy modification in men with PC and diabetes.
Clinical outcomes of patients with high-risk localized prostate cancer (PCa) remain very heterogeneous, because the available clinical classification systems do not take into account the tumor biology. The objective of this study is to identify potential drivers of metastatic progressive disease of primary high-risk PCa (HRPCa) and to investigate the underlying biology.
A retrospective matched case-control study of 2 clinically identical groups of HRPCa patients was performed. Both groups were treated with radical prostatectomy, but one group developed metastatic recurrence (n=19) and the other group did not (n=25) despite a median follow-up of 12 years. The primary index tumor was identified, followed by DNA and RNA extraction for somatic copy number aberration (CNA) and MicroArray based gene expression analysis. An integrative analysis was performed to identify potential driver genes of metastatic progression. Further in vitro analyses were performed with standard molecular biology methodology unless stated differently.
The integrative CNA and gene expression analyses identified that higher expression of antizyme inhibitor 1 (AZIN1) due to a focal amplification of 8q22.3 was associated with metastatic recurrence of HRPCa, which was validated in 4 independent PCa cohorts. AZIN1 is both one of the regulators of the polyamine homeostasis (plays a key role in many cellular processes) and Cyclin D1 stability (essential for the regulation of the cell cycle). Our in vitro analyses confirmed that modulation of AZIN1 expression determines both growth and migratory potential of PCa cells. RNA sequencing after knockdown of AZIN1 in PCa cells further revealed several transcriptional programs being activated/deactivated. This showed, among others, a significant upregulation of genes involved in extracellular matrix composition, including genes encoding for subunits of collagen IV, which is an integral part of the basement membrane. A correlation between AZIN1 and collagens’ gene expression levels was confirmed in primary PCa tumors as well. Ongoing experiments focus on identifying mechanisms by which AZIN1 regulates collagen IV expression.
An integrative CNA and gene expression analysis of primary high-risk PCa tumors identified AZIN1 as a potential driver of metastatic progression which was confirmed by our in vitro data. Knockdown of AZIN1 resulted in a significant upregulation of genes involved in extracellular matrix composition, which could explain the cells’ phenotype. Ongoing experiments focus on the mechanisms of collagen IV regulation by AZIN1, which will give us more insight in the underlying biology and could allow us to potentially inhibit metastatic progression of primary PCa.
Blue Area, Room 5 (Level 0)
Suicide has been shown to have a higher incidence in populations with cancer in several countries. However no national study has investigated this in the UK, and none specifically looking at disparities in urological malignancies. We aimed to identify all patients diagnosed with prostate, bladder or kidney cancer from 2001-2011 in England, and analyse incidence of suicide, suicide attempts, time to suicide and potential predisposing factors for suicidality.
We retrospectively analysed the records of patients diagnosed with cancer from the Hospital Episode Statistics database from 1st April 2001 to 31st January 2011. Office of National Statistics data was used to identify suicide as a cause of death. After data exclusion, from the initial 3,520,638 records, we identified 980,761 cases of the 10 commonest cancers, of which 328,372 had a diagnosis of urological malignancy at final data review.
We identified 1,222 suicide attempts, and 162 completed suicides. Suicide incidence was 49/100,000 in this cohort (48/100,000 in bladder, 36/100,000 in kidney, and 52/100,000 in prostate cancer), whereas population rates in England are 10/100,000. For non-urological cancers incidence was 30/100,000. Ratio of suicide/attempt was 1/7 for bladder and prostate, and 1/10 for kidney cancer, whereas in the general population it is thought to be 1/25. Median time to suicide was shortest for kidney cancer at 175 days, 1037 days for bladder, and 846 days for prostate. There was no disparity in suicide rates between genders, but older patients with prostate and bladder cancer were more likely to attempt suicide(p=<0.01), whereas more suicide attempts were seen in those under 50 for kidney cancer(p=0.01). There was no disparity in suicide attempts between Charlson comorbidity scores in prostate or bladder, but for kidney cancer there were more suicide attempts in those with fewer comorbidities(p=0.04). Analysing the index of multiple deprivation, there were more attempted suicides in those from lower socioeconomic groups.
Patients with urological cancers are five times more likely than the general population, and 63% more likely than other cancer patients to commit suicide. Also the ratio of attempt to suicide is lower than the general population, indicating more suicidal intent. The cause for the disparity with other cancers may be multifactorial including the possible social stigma associated with cancer of the urological tract. Formal identification of those at risk and early intervention should form the standard of care for patients diagnosed with urological cancers.
Red Area, Room 1 (Level 0)
Reliable biomarkers for renal cell carcinoma (RCC) have yet to be employed. Circulating cell-free DNA (cfDNA) is an emerging resource for the diagnosis and prognosis of various cancers. The study objective is to identify novel blood biomarkers for RCC.
Plasma cfDNA was extracted from RCC patients (n = 92) and healthy controls (n = 41). All patients with pathological features of clear cell RCC were eligible. Absolute copy number of plasma cfDNA was determined using quantitative real-time PCR of actin-beta as the target gene, and cfDNA fragments size was measured using a microfluidics platform. Diagnostic potential was assessed using receiver operating characteristic (ROC) and logistic regression analysis, and prognostic potential was evaluated using log rank test.
Mean absolute copy numbers of plasma cfDNA from RCC patients were significantly higher than those from healthy controls (5176 vs 2584 copies/ml, p < 0.001). Elevated absolute copy number of plasma cfDNA were confirmed even in cT1aN0M0 RCC patients compared to healthy controls (p < 0.001). Fragments size of plasma cfDNA from RCC patients were shorter than those from healthy controls (168.4 vs 171.4 bp, p = 0.052). To evaluate absolute copy number of plasma cfDNA as a diagnostic tool for RCC, ROC curve analysis revealed a sensitivity of 63.0% and a specificity of 78.1%. Multivariate analysis showed that age, gender and absolute copy number of plasma cfDNA were significantly associated with RCC diagnosis (p < 0.001, p = 0.013, p < 0.001, respectively). In terms of pathological status, RCC patients with Fuhrman nuclear grade 3 or 4 had significantly higher absolute copy number and shorter fragment sizes than those without it. Additionally, shorter fragment size of plasma cfDNA was associated with worse progression-free survival (p = 0.006). In patients belonging to the “short fragment size” group (≤ 166 bp) prior to surgical removal of the tumor, cfDNA fragment size became significantly longer in postoperative states.
Our study demonstrates that cfDNA characteristics such as absolute copy number and fragment size in RCC patients has promising diagnostic and prognostic potential. This is to our knowledge the first study showing that shorter fragment size of plasma cfDNA is significantly associated with poor prognosis.
Blue Area, Room 1 (Level 0)
There is a clear unmet clinical need for perioperative medical therapy in bladder cancer (BC). In phase 1-2 studies, treatment with recMAGE-A3 immunotherapy was associated with clinical responses in melanoma, lungcancer (LC) and BC. After evidence showed that the MAGE-A3 antigen was expressed in ~50% of BC patients, the Magnolia trial was initiated. Objectives were to evaluate disease free survival (DFS), the immune response and a gene signature predictive for response.
We aimed to screen 1300 patients who received no prior chemotherapy and randomize 273 patients with histologically confirmed BC (pT2–3 pN0–2 cM0 or pT4 pN0 cM0) with expression of MAGE-A3 antigen and clinically free of disease after cystectomy in 10 European countries. Patients were randomly assigned in a 2:1 ratio to receive either the MAGE-A3 cancer immunotherapeutic (M3CI) or placebo (P) in a double-blind way. Biological samples, i.e. tissue, blood, serum and urine were collected for transnational research. Upon GSK Biologicals' decision to stop further development of MAGE-A3 stand-alone immunotherapy following the negative outcomes in Phase 3 melanoma and LC trials and despite absence of safety issues, the trial’s recruitment was prematurely stopped.
A total of 529 patients gave consent; 448 tumor tissue samples were screened; 49.8% expressed MAGE-A3 and only 83 patients were randomized of whom 77 received in total 650 injections. Unexpected, only 37% of the patients with MAGE-A3 expression were randomized. Main reasons were: 1) patient was not recovered from surgery within 9 weeks where after the patient gave no consent anymore; 2) not fulfilling the eligibility criteria after pathological and imaging procedures postsurgery and 3) country specific competing (neo) adjuvant chemotherapy. For reasons 1 and 3, protocol amendments were developed. Apart from grade 1-2 flu-like symptoms and injection site sequel in the M3CI arm, safety was not different between arms. DFS was 27.5 months (95%CI:22.7–32.3) in the M3CI arm vs. 19.8 months (95%CI:15.7–23.9) in the P arm.
The Magnolia study represents a unique opportunity to investigate the concept of immunotherapy in BC. After premature termination, no clinically significant signal of efficacy could be shown. Lessons learned included the identification of multiple barriers to patient accrual including study complexity, multiple departments' involvement, poor communication with referral urologists, timing of informed consent, slow patient recovery, eligibility criteria and country specific competing (neo)adjuvant therapy that should be recognized in order to plan and conduct successful future clinical trials on perioperative therapy in BC in Europe. Biological samples will be used for further translational research in the targeted therapy perioperative setting.
This trial was financially supported by GSK Biologicals, Belgium.
Six cycles of platinum-based chemotherapy (CT) are conventionally targeted to treat locally advanced unresectable or metastatic urothelial carcinoma (UC). However, cisplatin is associated with significant cumulative toxicities, which render it challenging to deliver 6 cycles. Since this issue has not been investigated prospectively, we conducted a retrospective analysis.
The Retrospective International Study of Invasive/Advanced Cancer of the Urothelium (RISC) database was employed to conduct a retrospective analysis. The association of the number of cycles of platinum-based first-line CT with overall survival (OS) was investigated by a Cox regression utilizing multivariate analysis after controlling for previously recognized prognostic factors used in a nomogram (Eur Urol, 2017). The primary analysis was a comparison of <6 cycles vs. ≥6 cycles. Six-month landmark analysis was applied throughout, accounting for OS events. Additionally, we excluded patients (pts) receiving <3 or >9 cycles to reduce confounding due to early removal for toxicities, progression and patient decision and increased number of cycles due to response and pt-related factors.
Of 1020 pts available from RISC, 472 (cisplatin=338, carboplatin=134) were valuable for the landmark analysis with 281 events. A total of 157 pts received 3-5 cycles (median 4) and 315 received 6-9 cycles (median 6). There was no significant difference between 3-5 vs. 6-9 cycles of platinum-based chemotherapy (HR 1.02, 95%CI: 0.77-1.33, p=0.91). No significant interactions were observed with type of platinum (p=0.09) and “completed planned CT” (p=0.56). Comparison of 4 vs. 6 cycles (p=0.57) and <6 vs 6 vs 7-9 (p=0.9) also yielded no significant differences for association with OS. No differential association was observed with survival for 3-5 vs. 6-9 cycles when examining by nomogram-defined risk group tertiles. Limitations of a hypothesis-generating retrospective analysis apply.
Four cycles of platinum based first-line CT appear adequate to treat advanced UC. The omission of excessive cycles will avoid unnecessary toxicities and facilitate better transition to second-line and switch maintenance therapy.
Blue Area, Room 2 (Level 0)
Overdiagnosis and overtreatment of prostate cancer (PC) is a major challenge and use of PSA test is not without issues. Consequently, there is an urgent need for novel diagnostic and prognostic molecular biomarkers for PC. Tumors shed genomic DNA into the blood, known as circulating tumor DNA (ctDNA) and cancer-specific DNA methylation aberrations have shown particularly promising biomarker potential for PC. Here, we hypothesize that ctDNA represents a clinically relevant liquid tumour biopsy for PC and thus could be a source for discovery of new minimally-invasive diagnostic biomarkers. We aimed to identify hypermethylated PC-specific CpG sites that are suitable for liquid biopsy testing in order to improve diagnostic accuracy.
Genome-wide methylome patterns in 5031 male tissue samples from the Marmal-aid database were bioinformatically analysed. We looked for CpG sites that were hypermethylated in PC tissue and hypomethylated in other tissue. In order to avoid false positive signals caused by methylated cell free DNA from blood cells (mainly leukocytes), we filtered out all CpG sites that showed detectable levels of methylation in blood. Finally, we designed methylation specific PCR assays for the 11 top candidate markers. Sensitivity and specificity of these assays were tested on PC tissue (n=20) and BPH tissue (n=20) as well as leukocyte samples (n=40), before clinical validation using plasma samples (BPH, localized and metastatic PC).
We identified 11 top candidate DNA methylation markers with high sensitivity and specificity for PC in the Marmal-aid dataset. The 11 markers were significantly hypermethylated in PC tissue samples (n=184) when compared to non-malignant prostate tissue samples (n=81), other cancer tissue samples (n=2292) and normal blood cells (n=932). Experimental validation on tissue samples, reduced this list to six top candidate markers, which had 75-94 % sensitivity and 84 –100 % specificity for PC and 0% false-negative rates for detection of leukocyte DNA. The diagnostic potential of the six top candidate markers are currently being investigated using methylation-sensitive digital droplet PCR in plasma samples from large patient cohorts (n=150), including men with vs. without PC, recurrent vs. non-recurrent PC after radical prostatectomy and localized vs. metastatic PC.
We identified six novel diagnostic epigenetic biomarker candidates with high specificity and sensitivity for prostate cancer in tissue based analyses. Additionally, they displayed a promising potential for liquid tumor biopsy analyses, as they did not detect methylated DNA from leukocytes.
Blue Area, Room 3 (Level 0)