Manuel Wiesenfarth

Combined Clinical Parameters and Multiparametric Magnetic Resonance Imaging for Advanced Risk Modeling of Prostate Cancer—Patient-tailored Risk Stratification Can Reduce Unnecessary Biopsies

BACKGROUND Multiparametric magnetic resonance imaging (mpMRI) is gaining widespread acceptance in prostate cancer (PC) diagnosis and improves significant PC (sPC; Gleason score≥3+4) detection. Decision making based on European Randomised Study of Screening for PC (ERSPC) risk-calculator (RC) parameters may overcome prostate-specific antigen (PSA) limitations. OBJECTIVE We added pre-biopsy mpMRI to ERSPC-RC parameters and developed risk models (RMs) to predict individual sPC risk for biopsy-naïve men and men after previous biopsy. DESIGN, SETTING, AND PARTICIPANTS We retrospectively analyzed clinical parameters of 1159 men who underwent mpMRI prior to MRI/transrectal ultrasound fusion biopsy between 2012 and 2015. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS Multivariate regression analyses were used to determine significant sPC predictors for RM development. The prediction performance was compared with ERSPC-RCs, RCs refitted on our cohort, Prostate Imaging Reporting and Data System (PI-RADS) v1.0, and ERSPC-RC plus PI-RADSv1.0 using receiver-operating characteristics (ROCs). Discrimination and calibration of the RM, as well as net decision and reduction curve analyses were evaluated based on resampling methods. RESULTS AND LIMITATIONS PSA, prostate volume, digital-rectal examination, and PI-RADS were significant sPC predictors and included in the RMs together with age. The ROC area under the curve of the RM for biopsy-naïve men was comparable with ERSPC-RC3 plus PI-RADSv1.0 (0.83 vs 0.84) but larger compared with ERSPC-RC3 (0.81), refitted RC3 (0.80), and PI-RADS (0.76). For postbiopsy men, the novel RMs discrimination (0.81) was higher, compared with PI-RADS (0.78), ERSPC-RC4 (0.66), refitted RC4 (0.76), and ERSPC-RC4 plus PI-RADSv1.0 (0.78). Both RM benefits exceeded those of ERSPC-RCs and PI-RADS in the decision regarding which patient to receive biopsy and enabled the highest reduction rate of unnecessary biopsies. Limitations include a monocentric design and a lack of PI-RADSv2.0. CONCLUSIONS The novel RMs, incorporating clinical parameters and PI-RADS, performed significantly better compared with RMs without PI-RADS and provided measurable benefit in making the decision to biopsy men at a suspicion of PC. For biopsy-naïve patients, both our RM and ERSPC-RC3 plus PI-RADSv1.0 exceeded the prediction performance compared with clinical parameters alone. PATIENT SUMMARY Combined risk models including clinical and imaging parameters predict clinically relevant prostate cancer significantly better than clinical risk calculators and multiparametric magnetic resonance imaging alone. The risk models demonstrate a benefit in making a decision about which patient needs a biopsy and concurrently help avoid unnecessary biopsies.

Sensitivity and specificity of antibodies against HPV16 E6 and other early proteins for the detection of HPV16-driven oropharyngeal squamous cell carcinoma

To determine the sensitivity and specificity of HPV16 serology as diagnostic marker for HPV16‐driven oropharyngeal squamous cell carcinoma (OPSCC), 214 HNSCC patients from Germany and Italy with fresh‐frozen tumor tissues and sera collected before treatment were included in this study. Hundred and twenty cancer cases were from the oropharynx and 94 were from head and neck cancer regions outside the oropharynx (45 oral cavity, 12 hypopharynx and 35 larynx). Serum antibodies to early (E1, E2, E6 and E7) and late (L1) HPV16 proteins were analyzed by multiplex serology and were compared to tumor HPV RNA status as the gold standard. A tumor was defined as HPV‐driven in the presence of HPV16 DNA and HPV16 transformation‐specific RNA transcript patterns (E6*I, E1∧E4 and E1C). Of 120 OPSCC, 66 (55%) were HPV16‐driven. HPV16 E6 seropositivity was the best predictor of HPV16‐driven OPSCC (diagnostic accuracy 97% [95%CI 92–99%], Cohens kappa 0.93 [95%CI 0.8–1.0]). Of the 66 HPV‐driven OPSCC, 63 were HPV16 E6 seropositive, compared to only one (1.8%) among the 54 non‐HPV‐driven OPSCC, resulting in a sensitivity of 96% (95%CI 88–98) and a specificity of 98% (95%CI 90–100). Of 94 HNSCC outside the oropharynx, six (6%) were HPV16‐driven. In these patients, HPV16 E6 seropositivity had lower sensitivity (50%, 95%CI 19–81), but was highly specific (100%, 95%CI 96–100). In conclusion, HPV16 E6 seropositivity appears to be a highly reliable diagnostic marker for HPV16‐driven OPSCC with very high sensitivity and specificity, but might be less sensitive for HPV16‐driven HNSCC outside the oropharynx.

Integrative genome-scale analysis identifies epigenetic mechanisms of transcriptional deregulation in unfavorable neuroblastomas

The broad clinical spectrum of neuroblastoma ranges from spontaneous regression to rapid progression despite intensive multimodal therapy. This diversity is not fully explained by known genetic aberrations, suggesting the possibility of epigenetic involvement in pathogenesis. In pursuit of this hypothesis, we took an integrative approach to analyze the methylomes, transcriptomes, and copy number variations in 105 cases of neuroblastoma, complemented by primary tumor- and cell line-derived global histone modification analyses and epigenetic drug treatment in vitro We found that DNA methylation patterns identify divergent patient subgroups with respect to survival and clinicobiologic variables, including amplified MYCN Transcriptome integration and histone modification-based definition of enhancer elements revealed intragenic enhancer methylation as a mechanism for high-risk-associated transcriptional deregulation. Furthermore, in high-risk neuroblastomas, we obtained evidence for cooperation between PRC2 activity and DNA methylation in blocking tumor-suppressive differentiation programs. Notably, these programs could be re-activated by combination treatments, which targeted both PRC2 and DNA methylation. Overall, our results illuminate how epigenetic deregulation contributes to neuroblastoma pathogenesis, with novel implications for its diagnosis and therapy. Cancer Res; 76(18); 5523-37. ©2016 AACR.

Direct Simultaneous Inference in Additive Models and Its Application to Model Undernutrition

This article proposes a simple and fast approach to build simultaneous confidence bands and perform specification tests for smooth curves in additive models. The method allows for handling of spatially heterogeneous functions and its derivatives as well as heteroscedasticity in the data. It is applied to study the determinants of chronic undernutrition of Kenyan children, with a particular focus on the highly nonlinear age pattern in undernutrition. Model estimation using the mixed model representation of penalized splines in combination with simultaneous probability calculations based on the volume-of-tube formula enable the simultaneous inference directly, that is, without resampling methods. Finite sample properties of simultaneous confidence bands and specification tests are investigated in simulations. To facilitate and enhance its application, the method has been implemented in the R package AdaptFitOS.

Prediction of malignancy by a radiomic signature from contrast agent‐free diffusion MRI in suspicious breast lesions found on screening mammography.

To assess radiomics as a tool to determine how well lesions found suspicious on breast cancer screening X‐ray mammography can be categorized into malignant and benign with unenhanced magnetic resonance (MR) mammography with diffusion‐weighted imaging and T2‐weighted sequences.

The Finite Sample Performance of Semi- and Nonparametric Estimators for Treatment Effects and Policy Evaluation

The finite sample performance of a comprehensive set of semi- and non-parametric estimators for treatment evaluation is investigated. The simulation design is based on Swiss labor market data and considers estimators based on parametric, semiparametric, and nonparametric propensity scores, as well as approaches directly controlling for covariates. Among the methods included are pair, radius, kernel, and genetic matching, inverse probability weighting, regression, doubly robust estimation, entropy balancing, and empirical likelihood estimation. The simulation designs vary w.r.t. sample size, selection into treatment, effect heterogeneity, and (non-)omission of a subset of the all in all 3 continuous and 11 binary confounders. Several nonparametric estimators outperform commonly used treatment estimators based on parametric propensity scores in terms of root mean squared error (RMSE), even though average RMSEs based on the 16 simulation designs considered are not statistically significantly different across the estimators investigated.

Antibodies against human papillomaviruses as diagnostic and prognostic biomarker in patients with neck squamous cell carcinoma from unknown primary tumor: HPV antibodies as biomarker in NSCCUP patients

Treatment of patients with neck lymph node metastasis of squamous cell carcinoma (SCC) from unknown primary tumor (NSCCUP) is challenging due to the risk of missing occult tumors or inducing toxicity to unaffected sites. Human papillomavirus (HPV) is a promising biomarker given its causal link to oropharyngeal SCC and superior survival of patients with HPV‐driven oropharyngeal SCC and NSCCUP. Identification of HPV‐driven NSCCUP could focus diagnostic work‐up and treatment on the oropharynx. For the first time, we assessed HPV antibodies and their prognostic value in NSCCUP patients. Antibodies against E6 and E7 (HPV16/18/31/33/35), E1 and E2 (HPV16/18) were assessed in 46 NSCCUP patients in sera collected at diagnosis, and in follow‐up sera from five patients. In 28 patients, HPV tumor status was determined using molecular markers (HPV DNA, mRNA and cellular p16INK4a). Thirteen (28%) NSCCUP patients were HPV‐seropositive for HPV16, 18, 31, or 33. Of eleven patients with HPV‐driven NSCCUP, ten were HPV‐seropositive, while all 17 patients with non‐HPV‐driven NSCCUP were HPV‐seronegative, resulting in 91% sensitivity (95% CI: 59–100%) and 100% specificity (95% CI: 80–100%). HPV antibody levels decreased after curative treatment. Recurrence was associated with increasing levels in an individual case. HPV‐seropositive patients had a better overall and progression‐free survival with hazard ratios of 0.09 (95% CI: 0.01–0.42) and 0.03 (95% CI: 0.002–0.18), respectively. For the first time, seropositivity to HPV proteins is described in NSCCUP patients, and high sensitivity and specificity for HPV‐driven NSCCUP are demonstrated. HPV seropositivity appears to be a reliable diagnostic and prognostic biomarker for patients with HPV‐driven NSCCUP.

RAS-pathway mutation patterns define epigenetic subclasses in juvenile myelomonocytic leukemia

Juvenile myelomonocytic leukemia (JMML) is an aggressive myeloproliferative disorder of early childhood characterized by mutations activating RAS signaling. Established clinical and genetic markers fail to fully recapitulate the clinical and biological heterogeneity of this disease. Here we report DNA methylome analysis and mutation profiling of 167 JMML samples. We identify three JMML subgroups with unique molecular and clinical characteristics. The high methylation group (HM) is characterized by somatic PTPN11 mutations and poor clinical outcome. The low methylation group is enriched for somatic NRAS and CBL mutations, as well as for Noonan patients, and has a good prognosis. The intermediate methylation group (IM) shows enrichment for monosomy 7 and somatic KRAS mutations. Hypermethylation is associated with repressed chromatin, genes regulated by RAS signaling, frequent co-occurrence of RAS pathway mutations and upregulation of DNMT1 and DNMT3B, suggesting a link between activation of the DNA methylation machinery and mutational patterns in JMML.Juvenile myelomonocytic leukemia (JMML) is an aggressive disease with limited options for treatment. Here, the authors analyse the DNA methylome and mutational profile of JMML to define three subgroups with unique molecular and clinical characteristics.

Exploiting the potential of unlabeled endoscopic video data with self-supervised learning

PurposeSurgical data science is a new research field that aims to observe all aspects of the patient treatment process in order to provide the right assistance at the right time. Due to the breakthrough successes of deep learning-based solutions for automatic image annotation, the availability of reference annotations for algorithm training is becoming a major bottleneck in the field. The purpose of this paper was to investigate the concept of self-supervised learning to address this issue.MethodsOur approach is guided by the hypothesis that unlabeled video data can be used to learn a representation of the target domain that boosts the performance of state-of-the-art machine learning algorithms when used for pre-training. Core of the method is an auxiliary task based on raw endoscopic video data of the target domain that is used to initialize the convolutional neural network (CNN) for the target task. In this paper, we propose the re-colorization of medical images with a conditional generative adversarial network (cGAN)-based architecture as auxiliary task. A variant of the method involves a second pre-training step based on labeled data for the target task from a related domain. We validate both variants using medical instrument segmentation as target task.ResultsThe proposed approach can be used to radically reduce the manual annotation effort involved in training CNNs. Compared to the baseline approach of generating annotated data from scratch, our method decreases exploratively the number of labeled images by up to 75% without sacrificing performance. Our method also outperforms alternative methods for CNN pre-training, such as pre-training on publicly available non-medical (COCO) or medical data (MICCAI EndoVis2017 challenge) using the target task (in this instance: segmentation).ConclusionAs it makes efficient use of available (non-)public and (un-)labeled data, the approach has the potential to become a valuable tool for CNN (pre-)training.

Radiomic Machine Learning for Characterization of Prostate Lesions with MRI: Comparison to ADC Values

Purpose To compare biparametric contrast-free radiomic machine learning (RML), mean apparent diffusion coefficient (ADC), and radiologist assessment for characterization of prostate lesions detected during prospective MRI interpretation. Materials and Methods This single-institution study included 316 men (mean age ± standard deviation, 64.0 years ± 7.8) with an indication for MRI-transrectal US fusion biopsy between May 2015 and September 2016 (training cohort, 183 patients; test cohort, 133 patients). Lesions identified by prospective clinical readings were manually segmented for mean ADC and radiomics analysis. Global and zone-specific random forest RML and mean ADC models for classification of clinically significant prostate cancer (Gleason grade group ≥ 2) were developed on the training set and the fixed models tested on an independent test set. Clinical readings, mean ADC, and radiomics were compared by using the McNemar test and receiver operating characteristic (ROC) analysis. Results In the test set, radiologist interpretation had a per-lesion sensitivity of 88% (53 of 60) and specificity of 50% (79 of 159). Quantitative measurement of the mean ADC (cut-off 732 mm2/sec) significantly reduced false-positive (FP) lesions from 80 to 60 (specificity 62% [99 of 159]) and false-negative (FN) lesions from seven to six (sensitivity 90% [54 of 60]) (P = .048). Radiologist interpretation had a per-patient sensitivity of 89% (40 of 45) and specificity of 43% (38 of 88). Quantitative measurement of the mean ADC reduced the number of patients with FP lesions from 50 to 43 (specificity 51% [45 of 88]) and the number of patients with FN lesions from five to three (sensitivity 93% [42 of 45]) (P = .496). Comparison of the area under the ROC curve (AUC) for the mean ADC (AUCglobal = 0.84; AUCzone-specific ≤ 0.87) vs the RML (AUCglobal = 0.88, P = .176; AUCzone-specific ≤ 0.89, P ≥ .493) showed no significantly different performance. Conclusion Quantitative measurement of the mean apparent diffusion coefficient (ADC) improved differentiation of benign versus malignant prostate lesions, compared with clinical assessment. Radiomic machine learning had comparable but not better performance than mean ADC assessment.