Toni K Choueiri

Dana-Farber Cancer Institute, Boston, Massachusetts, United States

Are you Toni K Choueiri?

Claim your profile

Publications (351)2607.66 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Adrenocortical carcinoma (ACC) is a rare tumor in which prognostic factors are still not well established. Programmed Death Ligand-1 (PD-L1) expression in ACC and its association with clinico-pathological features and survival outcomes are unknown. Formalin-fixed paraffin-embedded (FFPE) specimens were obtained from 28 patients with ACC. PD-L1 expression was evaluated by immunohistochemistry (IHC) in both tumor cell membrane and tumor infiltrating mononuclear cells (TIMC). PD-L1 positivity on tumor cells was defined as ≥5% tumor cell membrane staining. TIMC were evaluated by IHC using a CD45 monoclonal antibody. For PD-L1 expression in TIMC, a combined score based on the extent of infiltrates and percentage of positive cells was developed. Any score greater that zero was considered PD-L1 positive. Baseline clinico-pathological characteristics and follow up data were retrospectively collected. Comparisons between PD-L1 expression and clinico-pathological features were evaluated using unpaired t-test and Fisher's exact test. Kaplan-Meier method and log-rank test were used to assess association between PD-L1 expression and 5-year overall survival (OS). Among 28 patients with surgically treated ACC, 3 (10.7%) were considered PD-L1 positive on tumor cell membrane. On the other hand, PD-L1 expression in TIMC was performed in 27 specimens and PD-L1 positive staining was observed in 19 (70.4%) patients. PD-L1 positivity in either tumor cell membrane or TIMC was not significantly associated with higher stage at diagnosis, higher tumor grade, excessive hormone secretion, or OS. PD-L1 expression can exist in ACC in both tumor cell membrane and TIMC with no relationship to clinico-pathologic parameters or survival.
    12/2015; 3(1). DOI:10.1186/s40425-015-0047-3
  • Sumanta K. Pal · Toni K. Choueiri · Jose A. Karam · Daniel Y.C. Heng ·

  • [Show abstract] [Hide abstract]
    ABSTRACT: Radiological assessment of tumor response remains a challenge in patients treated with immune checkpoint inhibitors. In metastatic melanoma, for example, a spectrum of imaging response patterns to immunotherapies have been recognized and associated with clinical benefit. In metastatic renal cell carcinoma (mRCC), less than half of patients treated with immune checkpoint inhibitors achieve objective responses, but some of the responses have been durable. In this series, five different imaging patterns of response and progression are described mRCC patients treated with anti-PD-1/PD-L1 agents: 1) early and complete response, 2) pseudo-progression, 3) disease stability before ultimate response, 4) mixed response with new lesions, and 5) early progression/primary refractory disease. The implications of the different imaging patterns of response on prognosis are discussed and highlight the need for individualized patient assessment with the use of novel immune-targeted agents.
    11/2015; DOI:10.1158/2326-6066.CIR-15-0197

  • JAMA The Journal of the American Medical Association 11/2015; 314(19):2077-2079. DOI:10.1001/jama.2015.7273 · 35.29 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: Vascular endothelial growth factor (VEGF)-targeted therapies are standard treatment for metastatic renal cell carcinoma (mRCC); however, toxicities can lead to drug discontinuation, which can affect patient outcomes. This study was aimed at identifying risk factors for toxicity and constructing the first model to predict toxicity-related treatment discontinuation (TrTD) in mRCC patients treated with VEGF-targeted therapies. Methods: The baseline characteristics, treatment outcomes, and toxicity data were collected for 936 mRCC patients receiving first-line VEGF-targeted therapy from the International Metastatic Renal Cell Carcinoma Database Consortium. A competing risk regression model was used to identify risk factors for TrTD, and it accounted for other causes as competing risks. Results: Overall, 198 (23.8%) experienced TrTD. Sunitinib was the most common VEGF-targeted therapy (77%), and it was followed by sorafenib (18.4%). The median time on therapy was 7.1 months for all patients and 4.4 months for patients with TrTD. The most common toxicities leading to TrTD included fatigue, diarrhea, and mucositis. In a multivariate analysis, significant predictors for TrTD were a baseline age ≥60 years, a glomerular filtration rate (GFR) <30 mL/min/1.73 m(2) , a single metastatic site, and a sodium level <135 mmol/L. A risk group model was developed that used the number of patient risk factors to predict the risk of TrTD. Conclusions: In the largest series to date, age, GFR, number of metastatic sites, and baseline sodium level were found to be independent risk factors for TrTD in mRCC patients receiving VEGF-targeted therapy. Based on the number of risk factors present, a model for predicting TrTD was built to be used as a tool for toxicity monitoring in clinical practice. Cancer 2015. © 2015 American Cancer Society.
    Cancer 11/2015; DOI:10.1002/cncr.29773 · 4.89 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background Papillary renal-cell carcinoma, which accounts for 15 to 20% of renal-cell carcinomas, is a heterogeneous disease that consists of various types of renal cancer, including tumors with indolent, multifocal presentation and solitary tumors with an aggressive, highly lethal phenotype. Little is known about the genetic basis of sporadic papillary renal-cell carcinoma, and no effective forms of therapy for advanced disease exist. Methods We performed comprehensive molecular characterization of 161 primary papillary renal-cell carcinomas, using whole-exome sequencing, copy-number analysis, messenger RNA and microRNA sequencing, DNA-methylation analysis, and proteomic analysis. Results Type 1 and type 2 papillary renal-cell carcinomas were shown to be different types of renal cancer characterized by specific genetic alterations, with type 2 further classified into three individual subgroups on the basis of molecular differences associated with patient survival. Type 1 tumors were associated with MET alterations, whereas type 2 tumors were characterized by CDKN2A silencing, SETD2 mutations, TFE3 fusions, and increased expression of the NRF2-antioxidant response element (ARE) pathway. A CpG island methylator phenotype (CIMP) was observed in a distinct subgroup of type 2 papillary renal-cell carcinomas that was characterized by poor survival and mutation of the gene encoding fumarate hydratase (FH). Conclusions Type 1 and type 2 papillary renal-cell carcinomas were shown to be clinically and biologically distinct. Alterations in the MET pathway were associated with type 1, and activation of the NRF2-ARE pathway was associated with type 2; CDKN2A loss and CIMP in type 2 conveyed a poor prognosis. Furthermore, type 2 papillary renal-cell carcinoma consisted of at least three subtypes based on molecular and phenotypic features. (Funded by the National Institutes of Health.).
    New England Journal of Medicine 11/2015; DOI:10.1056/NEJMoa1505917 · 55.87 Impact Factor

  • International journal of radiation oncology, biology, physics 11/2015; 93(3):E199. DOI:10.1016/j.ijrobp.2015.07.1054 · 4.26 Impact Factor

  • International journal of radiation oncology, biology, physics 11/2015; 93(3):E199. DOI:10.1016/j.ijrobp.2015.07.1055 · 4.26 Impact Factor

  • 11/2015; 3(Suppl 2):P353. DOI:10.1186/2051-1426-3-S2-P353

  • 11/2015; 3(Suppl 2):P197. DOI:10.1186/2051-1426-3-S2-P197
  • [Show abstract] [Hide abstract]
    ABSTRACT: Importance: There is extensive evidence suggesting that black men with localized prostate cancer (PCa) have worse cancer-specific mortality compared with their non-Hispanic white counterparts. Objective: To evaluate racial disparities in the use, quality of care, and outcomes of radical prostatectomy (RP) in elderly men (≥65 years) with nonmetastatic PCa. Design, setting, and participants: This retrospective analysis of outcomes stratified according to race (black vs non-Hispanic white) included 2020 elderly black patients (7.6%) and 24 462 elderly non-Hispanic white patients (92.4%) with localized PCa who underwent RP within the first year of PCa diagnosis in the Surveillance, Epidemiology, and End Results (SEER)-Medicare database between 1992 and 2009. The study was performed in 2014. Main outcomes and measures: Process of care (ie, time to treatment, lymph node dissection), as well as outcome measures (ie, complications, emergency department visits, readmissions, PCa-specific and all-cause mortality, costs) were evaluated using Cox proportional hazards regression. Multivariable conditional logistic regression and quantile regression were used to study the association of racial disparities with process of care and outcome measures. Results: The proportion of black patients with localized prostate cancer who underwent RP within 90 days was 59.4% vs 69.5% of non-Hispanic white patients (P < .001). In quantile regression of the top 50% of patients, blacks had a 7-day treatment delay compared with non-Hispanic whites. (P < .001). Black patients were less likely to undergo lymph node dissection (odds ratio [OR], 0.76 [95% CI, 0.66-0.80]; P < .001) but had higher odds of postoperative visits to the emergency department (within 30 days: OR, 1.48 [95% CI, 1.18-1.86]); after 30 days or more (OR, 1.45 [95% CI, 1.19-1.76]) and readmissions (within 30 days: OR, 1.28 [95% CI, 1.02-1.61]); ≥30 days (OR, 1.27 [95% CI, 1.07-1.51]) compared with non-Hispanic whites. The surgical treatment of black patients was associated with a higher incremental annual cost (the top 50% of blacks spent $1185.50 (95% CI , $804.85-1 $1566.10; P < .001) more than the top 50% of non-Hispanic whites). There was no difference in PCa-specific mortality (P = .16) or all-cause mortality (P = .64) between black and non-Hispanic white men. Conclusions and relevance: Blacks treated with RP for localized PCa are more likely to experience adverse events and incur higher costs compared with non-Hispanic white men; however, this does not translate into a difference in PCa-specific or all-cause mortality.
    10/2015; DOI:10.1001/jamaoncol.2015.3384
  • Guillermo de Velasco · Lana Hamieh · Suzanne Mickey · Toni K Choueiri ·
    [Show abstract] [Hide abstract]
    ABSTRACT: The introduction of molecularly targeted therapies (TTs) has transformed the management of metastatic renal cell carcinoma (mRCC). Within a relatively short period of time, systemic treatment of mRCC has evolved from a disease treated only by cytokines to a disease where TT is the cornerstone of patient management. Since the approval of sorafenib, an inhibitor of vascular endothelial growth factor receptor (VEGFR), in December 2005, 7 drugs have been introduced that have provided a high level of clinical efficacy in patients with mRCC, with a median survival of ~30 months in an unselected patient population that generally fits trials eligibility. Despite such success, advancements in therapies have reached a plateau: different combinations of targeted agents have not demonstrated additional benefit mainly owing to toxicity concerns, and some novel agents have failed to show benefit over approved drugs in clinics. In this review, we aim to focus on optimizing selection of agents in mRCC after progression on first-line TT. We also review how new drugs may transform existing guidelines and break through the current plateau reached with approved agents.
    Urologic Oncology 10/2015; DOI:10.1016/j.urolonc.2015.08.007 · 2.77 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Purpose: We aimed to describe changes in treatment patterns for clinical T3 prostate cancer (PCa) from 1998 to 2012, specifically investigating what factors influence receipt of prostatectomy or radiation. Materials and methods: Using the Surveillance, Epidemiology, and End Results database, we studied 11,604 men with clinical T3N0M0 PCa from 1998 to 2012, with treatment categorized as radiation, radical prostatectomy (RP), or no curative therapy. We calculated rate of treatment type by year of diagnosis to investigate trends in treatment patterns, further stratifying by clinical T3a, defined as unilateral and bilateral extracapsular extension (n = 3,842), vs. T3b (defined as extension to seminal vesicles (n = 3,665). Finally, a multivariable logistic regression analysis measured association of demographic and clinical variables with type of treatment received for years 2010 to 2011. Results: Rates of prostatectomy increased significantly from 1998 to 2012 (12.5% vs. 44.4%), radiation decreased significantly (55.8% vs. 38.4%), and receipt of no treatment also decreased (31.7% vs. 17.2%, all P<0.001). These trends were similar for clinical T3a and T3b. Rates of prostatectomy surpassed radiation by 2008 in clinical T3a, reaching 49.8% vs. 37.1%, respectively, in 2012 (P = 0.002), and were statistically similar to radiation in 2012 for clinical T3b, reaching 41.6% vs. 42.1% (P = 0.92). Multivariable logistic regression analysis demonstrated that patients were less likely to receive prostatectomy than radiation if biopsy Gleason scores of 8 to 10 (adjusted odds ratio [AOR] = 0.41, 0.32-0.53), higher initial prostate-specific antigen (AOR = 0.97, 0.97-0.98), and older age (AOR = 0.92, 0.90-0.03, all P<0.01). The likelihood of RP was similar among cT3b vs. cT3a (AOR = 0.95, 0.71-1.26, P = 0.74). Conclusions: Since 1998, there has been a significant increase in the use of RP for clinical T3 PCa and a significant decrease in the use of radiation such that in 2012, the use of prostatectomy exceeded the use of radiation.
    Urologic Oncology 10/2015; DOI:10.1016/j.urolonc.2015.09.002 · 2.77 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Brain metastases are associated with a dismal prognosis. Whether brain metastases harbor distinct genetic alterations beyond those observed in primary tumors is unknown. We performed whole-exome sequencing of 86 matched brain metastases, primary tumors, and normal tissue. In all clonally related cancer samples, we observed branched evolution, where all metastatic and primary sites shared a common ancestor yet continued to evolve independently. In 53% of cases, we found potentially clinically informative alterations in the brain metastases not detected in the matched primary-tumor sample. In contrast, spatially and temporally separated brain metastasis sites were genetically homogenous. Distal extracranial and regional lymph node metastases were highly divergent from brain metastases. We detected alterations associated with sensitivity to PI3K/AKT/mTOR, CDK, and HER2/EGFR inhibitors in the brain metastases. Genomic analysis of brain metastases provides an opportunity to identify potentially clinically informative alterations not detected in clinically sampled primary tumors, regional lymph nodes, or extracranial metastases. SIGNIFICANCE: Decisions for individualized therapies in patients with brain metastasis are often made from primary-tumor biopsies. We demonstrate that clinically actionable alterations present in brain metastases are frequently not detected in primary biopsies, suggesting that sequencing of primary biopsies alone may miss a substantial number of opportunities for targeted therapy.
    Cancer Discovery 09/2015; DOI:10.1158/2159-8290.CD-15-0369 · 19.45 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background: Cabozantinib is an oral, small-molecule tyrosine kinase inhibitor that targets vascular endothelial growth factor receptor (VEGFR) as well as MET and AXL, each of which has been implicated in the pathobiology of metastatic renal-cell carcinoma or in the development of resistance to antiangiogenic drugs. This randomized, open-label, phase 3 trial evaluated the efficacy of cabozantinib, as compared with everolimus, in patients with renal-cell carcinoma that had progressed after VEGFR-targeted therapy. Methods: We randomly assigned 658 patients to receive cabozantinib at a dose of 60 mg daily or everolimus at a dose of 10 mg daily. The primary end point was progression-free survival. Secondary efficacy end points were overall survival and objective response rate. Results: Median progression-free survival was 7.4 months with cabozantinib and 3.8 months with everolimus. The rate of progression or death was 42% lower with cabozantinib than with everolimus (hazard ratio, 0.58; 95% confidence interval [CI] 0.45 to 0.75; P<0.001). The objective response rate was 21% with cabozantinib and 5% with everolimus (P<0.001). A planned interim analysis showed that overall survival was longer with cabozantinib than with everolimus (hazard ratio for death, 0.67; 95% CI, 0.51 to 0.89; P=0.005) but did not cross the significance boundary for the interim analysis. Adverse events were managed with dose reductions; doses were reduced in 60% of the patients who received cabozantinib and in 25% of those who received everolimus. Discontinuation of study treatment owing to adverse events occurred in 9% of the patients who received cabozantinib and in 10% of those who received everolimus. Conclusions: Progression-free survival was longer with cabozantinib than with everolimus among patients with renal-cell carcinoma that had progressed after VEGFR-targeted therapy. (Funded by Exelixis; METEOR number, NCT01865747.).
    New England Journal of Medicine 09/2015; 373(19). DOI:10.1056/NEJMoa1510016 · 55.87 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: Nivolumab, a programmed death 1 (PD-1) checkpoint inhibitor, was associated with encouraging overall survival in uncontrolled studies involving previously treated patients with advanced renal-cell carcinoma. This randomized, open-label, phase 3 study compared nivolumab with everolimus in patients with renal-cell carcinoma who had received previous treatment. Methods: A total of 821 patients with advanced clear-cell renal-cell carcinoma for which they had received previous treatment with one or two regimens of antiangiogenic therapy were randomly assigned (in a 1:1 ratio) to receive 3 mg of nivolumab per kilogram of body weight intravenously every 2 weeks or a 10-mg everolimus tablet orally once daily. The primary end point was overall survival. The secondary end points included the objective response rate and safety. Results: The median overall survival was 25.0 months (95% confidence interval [CI], 21.8 to not estimable) with nivolumab and 19.6 months (95% CI, 17.6 to 23.1) with everolimus. The hazard ratio for death with nivolumab versus everolimus was 0.73 (98.5% CI, 0.57 to 0.93; P=0.002), which met the prespecified criterion for superiority (P≤0.0148). The objective response rate was greater with nivolumab than with everolimus (25% vs. 5%; odds ratio, 5.98 [95% CI, 3.68 to 9.72]; P<0.001). The median progression-free survival was 4.6 months (95% CI, 3.7 to 5.4) with nivolumab and 4.4 months (95% CI, 3.7 to 5.5) with everolimus (hazard ratio, 0.88; 95% CI, 0.75 to 1.03; P=0.11). Grade 3 or 4 treatment-related adverse events occurred in 19% of the patients receiving nivolumab and in 37% of the patients receiving everolimus; the most common event with nivolumab was fatigue (in 2% of the patients), and the most common event with everolimus was anemia (in 8%). Conclusions: Among patients with previously treated advanced renal-cell carcinoma, overall survival was longer and fewer grade 3 or 4 adverse events occurred with nivolumab than with everolimus. (Funded by Bristol-Myers Squibb; CheckMate 025 number, NCT01668784.).
    New England Journal of Medicine 09/2015; 373(19). DOI:10.1056/NEJMoa1510665 · 55.87 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Despite the widespread use of kinase-targeted agents in clear cell renal cell carcinoma (CC-RCC), comprehensive kinase activity evaluation (kinomic profiling) of these tumors is lacking. Thus, kinomic profiling of CC-RCC may assist in devising a classification system associated with clinical outcomes, and help identify potential therapeutic targets. Fresh frozen CC-RCC tumor lysates from 41 clinically annotated patients who had localized disease at diagnosis were kinomically profiled using the PamStation®12 high-content phospho-peptide substrate microarray system (PamGene International). Twelve of these patients also had matched normal kidneys available that were also profiled. Unsupervised hierarchical clustering and supervised comparisons based on tumor vs. normal kidney and clinical outcome (tumor recurrence) were performed and coupled with advanced network modeling and upstream kinase prediction methods. Unsupervised clustering analysis of localized CC-RCC tumors identified 3 major kinomic groups associated with inflammation (A), translation initiation (B), and immune response and cell adhesions (C) processes. Potential driver kinases implicated include PFTAIRE (PFTK1), PKG1, and SRC, which were identified in groups A, B, and C, respectively. Of the 9 patients who had tumor recurrence, only one was found in Group B. Supervised analysis showed decreased kinase activity of CDK1 and RSK1-4 substrates in those which progressed compared to others. Twelve tumors with matching normal renal tissue implicated increased PIM's and MAPKAPK's in tumors compared to adjacent normal renal tissue. As such, comprehensive kinase profiling of CC-RCC tumors could provide a functional classification strategy for patients with localized disease and identify potential therapeutic targets.
    PLoS ONE 09/2015; 10(9):e0139267. DOI:10.1371/journal.pone.0139267 · 3.23 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Objective: To identify contemporary, clinically low-risk patients with ≥50% cores positive and compare the risk of upgrading at prostatectomy to other low- or intermediate-risk patients. Methods: We studied 14,902 patients with prostate cancer in the Surveillance, Epidemiology, and End Results database in 2010-2011 with prostatectomy. Patients were categorized by NCCN clinical risk-groups, separating low-risk patients by percent positive biopsy cores (PBC). We measured incidence of pathologic high-risk disease, defined as pT3a-T4 or Gleason 8-10, and multivariable logistic regression (MVA) was used to determine if patients with clinical low-risk disease and ≥50% PBC were similar to other low- or intermediate-risk patients. This analysis was repeated with favorable- and unfavorable-intermediate risk. Results: At prostatectomy, 9.2% of clinically low-risk and <50%PBC, 18.6% of clinically low-risk and ≥50%PBC, and 27.6% of clinically intermediate-risk patients had occult, high-risk disease (p<0.001). On MVA low-risk with ≥50%PBC were more likely than low-risk with <50%PBC to have pathologic high-risk disease (Adjusted odds ratio [AOR] 2.28, 95%CI 1.90-2.73, p<0.001), had similar risk to favorable-intermediate disease overall (AOR 1.09, 0.91-1.31, p=0.33), and had higher risk than favorable-intermediate among men over 60 (AOR 1.28, 1.00-1.64, p=0.04). Low-risk and ≥50%PBC had a mean tumor size similar to unfavorable-intermediate risk (21.3 vs. 21.0mm, p=0.82). Conclusion: Nearly one in five clinically low-risk prostate cancer patients with ≥50% positive biopsy cores harbor occult pT3a-T4 or Gleason 8-10, suggesting that national guidelines should not classify low-risk with ≥50% cores positive as "low-risk" and patients should be made aware of this excess risk if considering active surveillance.
    Urology 09/2015; DOI:10.1016/j.urology.2015.08.026 · 2.19 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: In clinical trials, the use of intermediate time to event endpoints (TEE) is increasingly common, yet their choice and definitions are not standardized. This limits the usefulness for comparing treatment effects between studies. The aim of the DATECAN Kidney project is to clarify and recommend definitions of TEE in renal cell cancer (RCC) through a formal consensus method for endpoint definitions. Material and methods: A formal modified Delphi method was used for establishing consensus. From a 2006-2009 literature review, the Steering Committee (SC) selected 9 TEE and 15 events in the non-metastatic (NM) and metastatic/advanced (MA) RCC disease settings. Events were scored on the range of 1 (totally disagree to include) to 9 (totally agree to include) in the definition of each endpoint. Rating Committee (RC) experts were contacted for the scoring rounds. From these results, final recommendations were established for selecting pertinent endpoints and the associated events. Results: Thirty-four experts scored 121 events for 9 endpoints. Consensus was reached for 31%, 43% and 85% events during the first, second and third rounds respectively. In the NM setting: Disease-Free Survival (contralateral renal cell cancer, appearance of metastases, local or regional recurrence, death from RCC or protocol treatment), Metastasis-Free Survival (appearance of metastases, regional recurrence, death from RCC); and Local Regional Free-Survival (local or regional recurrence, death from RCC). In the MA setting: Kidney Cancer Specific Survival (death from RCC or protocol treatment) and Progression-Free Survival (death from RCC, Local, regional, or metastatic progression). Conclusions: The consensus method revealed that intermediate endpoints have not been well defined, since all of the selected endpoints had at least one event definition for which no consensus was obtained. These clarified definitions of TEE should become standard practice in all RCC clinical trials, thus facilitating reporting and increasing precision in between trial comparisons.
    Annals of Oncology 09/2015; DOI:10.1093/annonc/mdv380 · 7.04 Impact Factor

Publication Stats

6k Citations
2,607.66 Total Impact Points


  • 2008-2015
    • Dana-Farber Cancer Institute
      • • Department of Medical Oncology
      • • Department of Imaging
      • • Lank Center for Genitourinary Oncology
      Boston, Massachusetts, United States
    • Harvard Medical School
      • Department of Radiation Oncology
      Boston, Massachusetts, United States
  • 2014
    • University of Massachusetts Boston
      Boston, Massachusetts, United States
  • 2010-2014
    • Brigham and Women's Hospital
      • Lank Center for Genitourinary Oncology
      Boston, Massachusetts, United States
    • Beverly Hospital, Boston MA
      BVY, Massachusetts, United States
  • 2008-2014
    • Harvard University
      Cambridge, Massachusetts, United States
  • 2013
    • University of Connecticut
      • Department of Statistics
      Storrs, Connecticut, United States
  • 2009-2010
    • Tom Baker Cancer Centre
      Calgary, Alberta, Canada
  • 2007
    • Wojskowy Instytut Medyczny
      Warszawa, Masovian Voivodeship, Poland
  • 2006-2007
    • Cleveland Clinic
      • Department of Solid Tumor Oncology
      Cleveland, Ohio, United States