• Copenhagen, Denmark
Recent publications
The majority of neoantigens arise from unique mutations that are not shared between individual patients, making neoantigen-directed immunotherapy a fully personalized treatment approach. Novel technical advances in next-generation sequencing of tumor samples and artificial intelligence (AI) allow fast and systematic prediction of tumor neoantigens. This study investigates feasibility, safety, immunity, and anti-tumor potential of the personalized peptide-based neoantigen vaccine, EVX-01, including the novel CD8⁺ T-cell inducing adjuvant, CAF®09b, in patients with metastatic melanoma (NTC03715985). The AI platform PIONEERTM was used for identification of tumor-derived neoantigens to be included in a peptide-based personalized therapeutic cancer vaccine. EVX-01 immunotherapy consisted of 6 administrations with 5–10 PIONEERTM-predicted neoantigens as synthetic peptides combined with the novel liposome-based Cationic Adjuvant Formulation 09b (CAF®09b) to strengthen T-cell responses. EVX-01 was combined with immune checkpoint inhibitors to augment the activity of EVX-01-induced immune responses. The primary endpoint was safety, exploratory endpoints included feasibility, immunologic and objective responses. This interim analysis reports the results from the first dose-level cohort of five patients. We documented a short vaccine manufacturing time of 48–55 days which enabled the initiation of EVX-01 treatment within 60 days from baseline biopsy. No severe adverse events were observed. EVX-01 elicited long-lasting EVX-01-specific T-cell responses in all patients. Competitive manufacturing time was demonstrated. EVX-01 was shown to be safe and able to elicit immune responses targeting tumor neoantigens with encouraging early indications of a clinical and meaningful antitumor efficacy, warranting further study.
Purpose Visceral venous aneurysms are rare, especially in the inferior mesenteric vein (IMV). We report a giant IMV aneurysm secondary to an iatrogenic arteriovenous fistula (AVF). Case Report A woman presented with an incidental finding of a 7 cm large IMV aneurysm and an inferior mesenteric arteriovenous shunt. The patient underwent successful endovascular occlusion of the shunt to avoid aneurysm rupture and portal hypertension. Conclusion Embolization is a possible treatment strategy for mesenteric venous aneurysms with an AVF. Clinical Impact We describe an unusual mesenteric AV-shunt from a surgical crush injury that caused a giant venous mesenteric aneurysm and offer technical aspects on minimally invasive endovascular treatment.
Introduction Systemic lupus erythematosus (SLE) is a chronic autoimmune disease. SLE is treated with immunosuppressants with suboptimal efficacy and high risk of serious side effects. Patients with SLE have increased risk of mortality, organ damage and debilitating treatment-resistant fatigue. Autonomic nervous system dysfunction (AD) is present in approximately half of the patients and may promote autoimmunity by weakening the vagally mediated anti-inflammatory reflex. Recent studies suggest that transcutaneous vagus nerve stimulation (tVNS) has few side effects and beneficial effects on fatigue, pain, disease activity and organ function. This study investigates whether adjuvant tVNS improves measures of fatigue (primary end point), AD, clinical disease activity, inflammation, pain, organ function and quality of life. Hence, this study will contribute to the understanding of AD as a potentially important precursor of fatigue, disease activity, progression and complications in SLE, and how tVNS mechanistically may attenuate this. As adjuvant tVNS use may reduce the need for traditional immunosuppressive therapy, this trial may prompt a shift in the treatment of SLE and potentially other autoimmune disorders. Methods and analysis Eighty-four patients with SLE with fatigue and AD will be randomised 1:1 to active or sham tVNS in this double-blinded parallel-group study. In period 1 (1 week), participants will receive a 4 min tVNS 4 times daily and report on fatigue daily. After a 2-week pause, period 2 (8 weeks) will entail tVNS twice daily and participants will report on fatigue, pain and disease activity weekly. Secondary end points will be assessed before and after each period and after 1 week in period 2. Ethics and dissemination The study is approved by the Danish Medical Research Ethical Committees (case no: 2120231) and results will be published in international peer-reviewed journals. Trial registration number: NCT05315739.
Recently, small extracellular vesicles (sEVs) secreted in vivo from chronic lymphocytic leukemia (CLL) preclinical murine models were characterized. Leukemia microenvironment sEV (LME-sEVs) selectively target CD8⁺ T-cells, inducing exhaustion and hampering anti-tumor immune response. Additionally, a sEV-related gene expression correlated with patient treatment-free survival, overall survival and clinical parameters.
Domain shift is a common problem in machine learning and medical imaging. Currently one of the most popular domain adaptation approaches is the domain-invariant mapping method using generative adversarial networks (GANs). These methods deploy some variation of a GAN to learn target domain distributions which work on pixel level. However, they often produce too complicated or unnecessary transformations. This paper is based on the hypothesis that most domain shifts in medical images are variations of global intensity changes which can be captured by transforming histograms along with individual pixel intensities. We propose a histogram-based GAN methodology for domain adaptation that outperforms standard pixel-based GAN methods in classifying chest x-rays from various heterogeneous target domains.KeywordsUnsupervised domain adaptationHistogram layerLung disease classification
The coronavirus 2019 (COVID‐19) infection pandemic has affected the care of patients with heart failure (HF). Several consensus documents describe the appropriate diagnostic algorithm and treatment approach for patients with HF and associated COVID‐19 infection. However, few questions about the mechanisms by which COVID can exacerbate HF in patients with high‐risk (Stage B) or symptomatic HF (Stage C) remain unanswered. Therefore, the type of HF occurring during infection is poorly investigated. The diagnostic differentiation and management should be focused on the identification of the HF phenotype, underlying causes, and subsequent tailored therapy. In this framework, the relationship existing between COVID and onset of acute decompensated HF, isolated right HF, and cardiogenic shock is questioned, and the specific management is mainly based on local hospital organization rather than a standardized model. Similarly, some specific populations such as advanced HF, heart transplant, patients with left ventricular assist device (LVAD), or valve disease remain under investigated. In this systematic review, we examine recent advances regarding the relationships between HF and COVID‐19 pandemic with respect to epidemiology, pathogenetic mechanisms, and differential diagnosis. Also, according to the recent HF guidelines definition, we highlight different clinical profile identification, pointing out the main concerns in understudied HF populations.
Serum free light chain (FLC) concentration is greatly affected by kidney function. Using a large prospective population-based cohort, we aimed to establish a reference interval for FLCs in persons with chronic kidney disease (CKD). A total of 75422 participants of the iStopMM study were screened with serum FLC, serum protein electrophoresis and immunofixation. Estimated glomerular filtration rate (eGFR) was calculated from serum creatinine. Central 99% reference intervals were determined, and 95% confidence intervals calculated. Included were 6461 (12%) participants with measured FLCs, eGFR < 60 mL/min/1.73 m ² , not receiving renal replacement therapy, and without evidence of monoclonality. Using current reference intervals, 60% and 21% had kappa and lambda FLC values outside the normal range. The FLC ratio was outside standard reference interval (0.26–1.65) in 9% of participants and outside current kidney reference interval (0.37–3.10) in 0.7%. New reference intervals for FLC and FLC ratio were established. New reference intervals for the FLC ratio were 0.46–2.62, 0.48–3.38, and 0.54–3.30 for eGFR 45–59, 30–44, and < 30 mL/min/1.73 m ² groups, respectively. The crude prevalence of LC-MGUS in CKD patients was 0.5%. We conclude that current reference intervals for FLC and FLC ratio are inaccurate in CKD patients and propose new eGFR based reference intervals to be implemented.
Background: Angiotensin-converting enzyme inhibitors attenuate left ventricular (LV) enlargement after acute myocardial infarction (AMI). Preclinical data suggest similar benefits with combined angiotensin receptor neprilysin inhibition, but human data are conflicting. The PARADISE-MI Echo Study (Prospective ARNI Versus ACE Inhibitor Trial to Determine Superiority in Reducing Heart Failure Events After Myocardial Infarction) tested the effect of sacubitril/valsartan compared with ramipril on LV function and adverse remodeling after high risk-AMI. Methods: In a prespecified substudy, 544 PARADISE-MI participants were enrolled in the Echo Study to undergo protocol echocardiography at randomization and after 8 months. Patients were randomized within 0.5 to 7 days of presentation with their index AMI to receive a target dose of sacubitril/valsartan 200 mg or ramipril 5 mg twice daily. Echocardiographic measures were performed at a core laboratory by investigators blinded to treatment assignment. The effect of treatment on change in echo measures was assessed with ANCOVA with adjustment for baseline value and enrollment region. The primary end points were change in LV ejection fraction (LVEF) and left atrial volume (LAV), and prespecified secondary end points included changes in LV end-diastolic and end-systolic volumes. Results: Mean age was 64±12 years; 26% were women; mean LVEF was 42±12%; and LAV was 49±17 mL. Of 544 enrolled patients, 457 (84%) had a follow-up echo at 8 months (228 taking sacubitril/valsartan, 229 taking ramipril). There was no significant difference in change in LVEF (P=0.79) or LAV (P =0.62) by treatment group. Patients randomized to sacubitril/valsartan demonstrated less increase in LV end-diastolic volume (P=0.025) and greater decline in LV mass index (P=0.037), increase in tissue Doppler e'lat (P=0.005), decrease in E/e'lat (P=0.045), and decrease in tricuspid regurgitation peak velocity (P=0.024) than patients randomized to ramipril. These differences remained significant after adjustment for differences in baseline characteristics. Baseline LVEF, LV end-diastolic volume, LV end-systolic volume, LV mass index, LAV, and Doppler-based diastolic indices were associated with risk of cardiovascular death or incident heart failure. Conclusions: Treatment with sacubitril/valsartan compared with ramipril after AMI did not result in changes in LVEF or LAV at 8 months. Patients randomized to sacubitril/valsartan had less LV enlargement and greater improvement in filling pressure. Measures of LV size, systolic function, and diastolic properties were predictive of cardiovascular death and incident heart failure after AMI in this contemporary, well-treated cohort. Registration: URL: https://clinicaltrials.gov; Unique identifier: NCT02924727.
Objectives To investigate thresholds for lymph node yield (LNY), lymph node density (LND), and pN in patients with oral squamous cell carcinoma in relation to previous findings in the literature. Study Design Retrospective register-based study. Setting Copenhagen Oral Cavity Squamous Cell Carcinoma database. Methods Appropriate thresholds for LNY, LND, and pN were determined by areas under the curve and subsequently subjected to multivariate analysis. Five-year overall survival and 3-year recurrence-free survival were determined by Kaplan-Meier survival curves. Results In total, 413 patients diagnosed with oral squamous cell carcinoma were included. In the pN0 cohort, no superior/prognostic LNY cutoff values were detected. In the pN+ cohort, areas under the curve determined thresholds of LNY, LND, and pN to be 21 nodes, 5%, and 3 metastases, respectively. The 5-year overall survival was 52% for patients with LNY ≥21 vs 38% for patients with LNY <21 (hazard ratio [HR], 1.49; 95% CI, 1.05-2.11; P < .05), 60% for patients with LND ≤5% vs 38% for patients with LND >6% (HR, 1.63; 95% CI, 1.03-2.57; P < .05), and 43% for patients with pN <3 vs 26% for patients with pN ≥3 (HR, 1.40; 95% CI, 1.04-2.15; P < .05). Conclusions Increased nodal yield, decreased LND, and decreasing number of pN were associated with significantly improved survival outcomes. LNY might serve as a prognosticator of survival as well as a surgical quality indicator. LND may have implications as a tool in cancer staging and treatment planning.
While intracranial aneurysms rarely develop after neurosurgical procedures, delayed pseudoaneurysm formation after foramen magnum decompression (FMD) has never been reported. A 52-year-old woman presented with an atypical subarachnoid hemorrhage in the posterior fossa 12 years after a FMD for symptomatic Chiari malformation type I was performed. A pseudoaneurysm on a dural-pial anastomosis was identified as the bleeding source and successfully occluded by endovascular means with full clinical recovery of the patient. Injury to the distal posterior inferior cerebellar artery related to surgery and postoperative infection likely caused formation of a dural-pial anastomosis. Additionally, hemodynamic stress or dissection may have contributed to delayed pseudoaneurysm formation and rupture.
Background Glioblastoma is one of the most aggressive cancers, and hypoxia plays an essential role in its tumor- microenvironment. Tumor-associated microglia and macrophages (TAMs) have been reported to constitute up to 30 % of the cells, a fraction that is even higher in hypoxic areas. Single-cell mRNA sequencing of glioblastoma tumors has revealed vast heterogeneity, but the spatial aspects are not entirely defined yet. The aim of this study was to uncover differences between the hypoxic and normoxic tumor-microenvironment of human glioblastoma by high-end multiplexing with imaging mass cytometry. Material and Methods A tissue microarray (TMA) with normoxic and hypoxic areas from 4 IDH-wildtype glioblastomas was prepared based on the hypoxia marker hypoxia-inducing factor 1 alpha (HIF1 alpha). The TMA was stained with 18 metal-tagged antibodies covering TAMs, lymphocytes, immune checkpoints, vessels, tumor cells and proliferation. The Hyperion-CYTOF technology was used to ablate the samples and the images were analyzed by MCD viewer, Visiopharm software, and customized R scripts. Results Single-cell analysis of 160 fields covering around 45,000 cells in the glioblastoma microenvironment revealed multiple cellular phenotypes. It was revealed that proliferating TAMs (IBA1+, Ki67+) were more frequent in hypoxia, whereas proliferating vessels (CD34+, Ki67+) were more frequent in normoxia. Additionally, proliferating stem-like tumor cells (OLIG-2+, Ki67+) were more frequent in normoxia regions. Conclusion Our study revealed multiple cellular phenotypes in the glioblastoma microenvironment. The TAMs, endothelial and tumor cell phenotypes revealed may play a critical role in glioblastoma biology however this needs to be elucidated in future studies.
Objectives To assess the resection quality of transurethral bladder tumour resection (TURBT) and the association to surgeon experience depending on the presence of detrusor muscle. Methods A retrospective study on 640 TURBT procedures performed at Zealand University Hospital, Denmark, from 1 January 2015 − 31 December 2016. Data included patient characteristics, procedure type, surgeon category, supervisor presence, surgical report data, pathological data, complications data and recurrence data. Analysis was performed using simple and multiple logistic regression on the association between surgeon experience and the presence of detrusor muscle in resected tissue from TURBT. Results Supervised junior residents had significant lower detrusor muscle presence (73%) compared with consultants (83%) (OR = 0.4, 95% CI = 0.21–0.83). Limitations were the retrospective design and the diversity of included TURBT. Conclusions It was found that surgical experience predicts detrusor muscle presence and supervised junior residents performing TURBT on patients resulted in less detrusor muscle than consultants.
Background Glioblastoma is the most aggressive cancer originating in the brain with an average survival of 15 months. One of the characteristics of glioblastoma is the high level of intra-tumour heterogeneity (ITH), but the composition and complexity at the single-cell level is poorly understood. Here, we aimed to assess the effects and consequences of immune checkpoint inhibitor (ICI) on the cellular and molecular heterogeneity of glioblastoma tumours using at the single cell level. Material and Methods In collaboration with the phase I trials unit at Rigshospitalet, we performed paired molecular analysis of glioma cells from primary and relapse surgery after ICI treatment. Samples were analysed using single-cell RNA sequencing (scRNA-seq) as well as bulk RNA sequencing and whole exome DNA sequencing. Results In an effort to trace cellular lineages we developed and refined methods to a identify copy number changes using scRNA-seq. To this end, we identified clonal and subclonal tumour cell populations in each sample. We found high levels of ITH prior to treatment, both with respect to the glioblastoma subtype enrichment and the cell type-specific gene expression. Using expression-based cell-type classification, we found defined recurrent cell-type populations present at both surgery time points. The immune checkpoint treatment had consequences on the cellular phenotypes and proportions of tumour cells, suggesting a level of plasticity in the neoplastic cells. Moreover, we identified examples of clonal dynamics and sweeps following ICI treatment, pointing to potential treatment response and resistance in these population. Conclusion In summary, we pursued single cell-focused analysis of ICI treated glioblastoma patients to study the cellular and molecular heterogeneity within and between glioblastoma patients, which pointed to recurrent patterns of cellular responses following ICI treatment.
Background Glioblastoma is an aggressive brain tumour with a median survival of less than 15 months despite aggressive standard treatment consisting of neurosurgery, radiotherapy plus concomitant and adjuvant chemotherapy with temozolomide. There is no standard treatment at recurrence, and all have limited efficacy. In this study we will enroll glioblastoma patients from the Danish clinical trial “Protarget” to get a better understanding of resistance mechanisms and vulnerabilities in glioblastoma to give personalized medicine to glioblastoma patients at recurrent disease. We hypothesise that two connected factors are responsible for the failure to successfully target glioblastoma: i) intra-tumour heterogeneity and within this ii) cancer stem cells. Material and Methods We will use tumour tissue from glioblastoma patients included in the Danish phase 2, prospective, non-randomized clinical trial “Protarget” to establish short-term cultured neurospheres to preserve the heterogeneity of the tumour including cancer stem cells. We will then perform single cell RNA sequencing before and after drug screening. This setup will allow for a new approach to identify drug vulnerabilities at the single cell level. If successful in identifying drugs that leads to a clinical response, according to the "ProTarget" guidelines, we will consider scaling up the effort, to allow broader evaluation of efficacy. Our plan is to enroll 10 patients as a start. Patients will be selected based on ProTarget inclusion criteria and molecular profiling. We will further select patients with the following characteristics:1. IDH wildtype 2. High tumour purity (>50%). 3. High degree of intra-tumour heterogeneity. Results Enrollment for "Protarget" began in 2020 and our cohort will be part of this study, which is why enrollment has already started, but the first patient has not been enrolled yet. Clinical trial registry number for "Protarget" is NCT04341181. Conclusion This new and flexible approach to identify drug vulnerabilities at the single cell level in glioblastoma in order to find targeted therapies is a promising tool for future treatment of glioblastoma patients.
Post–mortem computed tomography (PMCT) is a routine tool in many forensic pathology departments as it is fast and non-destructive and allows less gruesome visualization than photographs, and the images are indefinitely storable. Several studies investigated congruence between PMCT and autopsy for skull fracture but registered only the presence or absence of fracture systems. The objective of this study was to determine location-specific sensitivity and specificity of PMCT for individual fracture lines in blunt force head trauma. Accurate 3D models based on PMCT data with all fracture lines visible are important for future studies on fractures, applying finite element analysis (FEA). We retrospectively sampled adult cases from 2013 to 2019 with skull fracture mentioned in the autopsy report. PMCT was on a Siemens 64-slice scanner and autopsy according to international guidelines. The location and direction of all fracture lines at autopsy and at de novo interpretation of scans were registered and compared. Ninety-nine cases with 4809 individual findings were included. Age ranged from 18 to 100 years. The overall sensitivity was 0.58, and specificity was 0.91. For individual locations, sensitivity ranged from 0.24 to 0.85, and specificity ranged from 0.73 to 1.00. Intra-observer agreement was 0.74, and inter-observer agreement ranged from 0.43 to 0.58. In conclusion, PMCT is suited for detection of fracture systems, but not for detection of all individual fracture lines. Our results differed from the existing literature due to the methodological choices of registering individual fracture lines. Future studies utilising FEA must supplement PMCT with autopsy data.
Objectives To determine whether electrocardiogram (ECG) markers are associated with incident non-Alzheimer's dementia (non-AD) and whether these markers also improve risk prediction for non-AD. Materials and methods We retrospectively included 170,605 primary care patients aged 60 years or older referred for an ECG by their general practitioner and followed them for a median of 7.6 years. Using Cox regression, we reported hazard ratios (HRs) for electrocardiogram markers. Subsequently, we evaluated if addition of these electrocardiogram markers to a clinical model improved risk prediction for non-AD using change in area under the receiver-operator characteristics curve (AUC). Results The 5-year cumulative incidence of non-AD was 3.4 %. Increased heart rate (HR=1.06 pr. 10 bpm [95% confidence interval: 1.04–1.08], p<0.001), shorter QRS duration (HR=1.07 pr. 10 ms [1.05–1.09], p<0.001), elevated J-amplitude (HR=1.16 pr. mm [1.08–1.24], p<0.001), decreased T-peak amplitude (HR=1.02 pr. mm [1.01–1.04], p=0.002), and increased QTc (HR=1.08 pr. 20 ms [1.05–1.10], p<0.001) were associated with an increased rate of non-AD. Atrial fibrillation on the ECG (HR=1.18 [1.08–1.28], p<0.001) Sokolow-Lyon index > 35 mm (HR=1.31 [1.18–1.46], p<0.001) and borderline (HR=1.18 [1.11–1.26], p<0.001) or abnormal (HR=1.40 [1.27–1.55], p<0.001) QRS-T angle were also associated with an increased rate of non-AD. Upon addition of ECG markers to the Cox model, 5-year and 10-year C-statistic (AUC) improved significantly (delta-AUC, 0.36 [0.18–0.50] and 0.20 [0.03–0.35] %-points, respectively). Conclusions ECG markers typical of an elevated cardiovascular risk profile were associated with non-AD and improved both 5-year and 10-year risk predictions for non-AD.
Title Self-reported limitations in physical function are common 6 months after out-of-hospital cardiac arrest. Background Out-of-hospital cardiac arrest (OHCA) survivors generally report good health-related quality of life, but physical aspects of health seem more affected than other domains. Limitations in physical function after surviving OHCA have received little attention. Aims To describe physical function 6 months after OHCA and compare it with a group of ST elevation myocardial infarction (STEMI) controls, matched for country, age, sex and time of the cardiac event. A second aim was to explore variables potentially associated with self-reported limitations in physical function in OHCA survivors. Methods A cross-sectional sub-study of the Targeted Temperature Management at 33 °C versus 36 °C (TTM) trial with a follow-up 6 months post-event. Physical function was the main outcome assessed with the self-reported Physical Functioning-10 items scale (PF-10). PF-10 is presented as T-scores (0–100), where 50 represents the norm mean. Scores <47 at a group level, or <45 at an individual level indicate limitations in physical function. Results 287 OHCA survivors and 119 STEMI controls participated. Self-reported physical function by PF-10 was significantly lower for OHCA survivors compared to STEMI controls (mean 46.0, SD 11.2 vs. 48.8, SD 9.0, p = 0.025). 38% of OHCA survivors compared to 26% of STEMI controls reported limitations in physical function at an individual level (p = 0.022). The most predictive variables for self-reported limitations in physical function in OHCA survivors were older age, female sex, cognitive impairment, and symptoms of anxiety and depression after 6 months. Conclusion Self-reported limitations in physical function are more common in OHCA survivors compared to STEMI controls. Trial registration ClinicalTrials.gov Identifier: NCT01946932.
Background Bilateral carpal tunnel syndrome (CTS) is a common extracardiac manifestation of amyloidosis and usually predates overt cardiac amyloidosis (CA) by several years. Screening studies on patients undergoing CTS surgery have shown a low yield of CA (2.0%), but high prevalence of amyloid in the carpal ligament. The proportion of patients with amyloid in the carpal ligament who later develop CA is unknown. Objectives The authors sought to investigate the prevalence of undiagnosed CA 5 to 15 years after surgery for bilateral CTS. Methods Using national registries, the authors identified subjects aged 60 to 85 years with prior CTS surgery, where the first procedure on the second wrist was performed 5 to 15 years earlier. Invitations to participate in the study were sent by mail. Per international recommendations, the initial cardiac evaluation included echocardiography, 99mtechnetium-pyrophosphate scintigraphy, and assessment of monoclonal proteins in serum and urine. Results A total of 250 subjects (35.7% of those invited) participated in the study. The median age was 70.4 years, and 50% were female. CA was diagnosed in 12 patients (4.8%; 95% CI: 2.5%-8.2%), and all cases were wild-type transthyretin amyloidosis (ATTRwt). The prevalence of ATTRwt in men was 8.8% (95% CI: 4.5%-15.2%; n = 11), and 21.2% (95% CI: 11.1%-34.7%) in male subjects ≥70 years with a BMI <30 kg/m². All but 2 patients diagnosed with ATTRwt were in the lowest disease severity score (Mayo score). Conclusions Screening for CA in patients with prior surgery for bilateral CTS finds approximately 5% with early-stage transthyretin CA. The clinical yield was higher (>1 in 5) when focusing on nonobese men ≥70 years, showing potential for systematic screening.
Institution pages aggregate content on ResearchGate related to an institution. The members listed on this page have self-identified as being affiliated with this institution. Publications listed on this page were identified by our algorithms as relating to this institution. This page was not created or approved by the institution. If you represent an institution and have questions about these pages or wish to report inaccurate content, you can contact us here.
1,673 members
Mikael Palner
  • Neurobiology Research Unit
Omid Rezahosseini
  • Department of Infectious Diseases 8632
Daria N Podlekareva
  • Centre of Excellence for Health, Immunity and Infections (CHIP)
Jesper Frank Christensen
  • Department of Oncology
Blegdamsvej 9, DK-2100, Copenhagen, Denmark
Head of institution
Christian Gluud