Time

Prog #

Wilburn E. Reddick¹, John O. Glass¹, Catherine A. Billups¹, Fred A. Hoffer¹, Jesse J. Jenkins¹, Juan Quintana², Milena Villarroel², Lori Luchtman-Jones³, Najat C. Daw¹

¹St. Jude Children's Research Hospital, Memphis, Tennessee, USA; ²Luis Calvo Mackenna Hospital, Santiago, Chile; ³Washington University Medical School, St. Louis, Missouri, USA

The current study prospectively evaluated the use of a two-compartment pharmacokinetic model of DCE-MRI to non-invasively evaluate tumor response to neoadjuvant chemotherapy in children treated on a single phase II trial at three institutions (2 US and 1 Chile). DCE-MRI was performed at baseline, week 9, and week 12 of preoperative chemotherapy prior to definitive surgery. Histological assessment of tumor response was determined from en bloc resections. Decreased Ktrans and ve at week 9 relative to presentation was significantly associated with good histologic response to preoperative chemotherapy. Patients with higher Ktrans and ve at presentation were more likely to have good histological response. DCE-MRI is a non-invasive method that can be feasibly used to assess tumor response to neoadjuvant chemotherapy in OS even between institutions in an international setting.

Jason A. Koutcher¹, Ya Wang¹, Howard T. Thaler¹, Paul Meyers¹, Leonard H. Wexler¹, Jonathan P. Dyke², David M. Panicek¹, Lawrence H. Schwartz¹, Maayan E. Korenblit¹, Wei Huang¹

¹Memorial Sloan Kettering Cancer Center, New York, New York, USA; ²Weill-Cornell Medical College, New York, New York, USA

Bone sarcomas are among the most common pediatric tumors.  The most useful prognostic marker has been percent necrosis at time of surgery post chemotherapy – patients with > 90% necrosis have a better prognosis.  Using DCE-MRI, we have found the combination of measuring ve (interstitial volume fraction) and pathological necrosis is a highly significant predictor of long term survival in a group of 25 patients followed for up to 9 years (p<0.002) and is superior to percent necrosis alone which has only borderline statistical significance.  Only one of the 25 patients was misclassified by this method.

Fernando Arias-Mendoza¹, Geoffrey S. Payne², Kristen Zakian³, Marion Stubbs⁴, Jaime G. Cruz-Lobo¹, A J. Schwarz², Amita Dave³, Franklyn Howe⁵, N R. Maisey², D Cunningham², H Poptani⁶, M R. Smith⁷, O A. O'Connor¹, R Pettengell⁵, M O. Leach², J A. Koutcher³, J R. Griffiths⁴, A Heerschap Heerschap⁸, J D. Glickson⁶

¹Columbia University, New York, New York, USA; ²Royal Marsden Hospital, London, UK; ³Memorial Sloan Kettering Cancer Center, New York, New York, USA; ⁴Cambridge University, Cambridge, UK; ⁵St. George's Hospital, London, UK; ⁶University of Pennsylvania, Philadelphia, Pennsylvania, USA; ⁷Fox Chase Cancer Center, Philadelphia, Pennsylvania, USA; ⁸Radboud University Nijmegen Medical Center, Nijmegen, Netherlands

Our ongoing international multi-institutional research program has demonstrated that the sum of the integrated ³¹P-MR resonances of phosphoethanolamine plus phosphocholine [Etn-P + Cho-P] normalized by the total nucleotide triphosphates (NTP) acquired prior to treatment initialization can predict treatment failure and drug-free survival in various forms of lymphomas. To make our results diagnosis- and treatment- specific, patients with diffuse large B-cell lymphomas (DLBCL) that were treated with CHOP or R-CHOP were studied. As in our previous reports, we show here that the pre-treatment [Etn-P + Cho-P]/NTP ratio predicts long-term failure to treatment and drug-free survival in these set of patients.

Sungheon Kim¹, Sanjeev Chawla¹, Laurie A. Loevner¹, Harry Quon¹, Eric J. Sherman¹, Gregory S. Weinstein¹, Harish Poptani¹

¹University of Pennsylvania, Philadelphia, USA

In this study ¹H MRS was performed in the metastatic cervical lymph nodes of squamous cell carcinomas of the head and neck (HNSCC) to assess the potential of choline as a biomarker for treatment response. Single-voxel ¹H MR spectroscopy was performed in 32 patients who were treatment naïve. A negative correlation was observed between tCho and residual tumor volume at the end of treatment suggesting that higher pre-treatment tCho values indicate better treatment response. These studies indicate that pre-treatment choline levels can be used as a marker for prediction of treatment response in HNSCC.

Hyeon-Man Baek¹, Jeon-Hor Chen, ², Shadfar Bahri, Rita S. Mehta, Orhan Nalcioglu, Min-Ying Su

¹ University of California-Irvine, Irvine, California , USA; ²China Medical University Hospital, Taichung 404, Taiwan

The changes of choline and tumor size in 35 patients receiving neoadjuvant chemotherapy were analyzed. Depending on the final pathological findings, patients were separated into pathologic complete response (pCR) vs. residual disease (non-pCR) groups. In F/U-1 after 2cycles AC, the median change of Cho and tumor size was -62% and -17%, respectively in the pCR group, and was -36% and -15% in the non-pCR group. In F/U-2, all pCR patients had non-detectable Cho. The changes of Cho were greater than the size changes, and higher in pCR than in non-pCR group. Therefore, its role for response prediction warrants further investigation.

Laura Martincich¹, Ilaria Bertotto¹, Alfonso Fausto², Filippo Montemurro¹, Daniele Regge¹, Francesco Sardanelli², ³

¹Institute for Cancer Research and Treatment, Candiolo, Italy; ²Policlinico S.Donato, S. Donato Milanese, Italy; ³University of Milan, Milan, Italy

In vivo effects of NCT on tumor vascularity are not exhaustively studied by MRI. The evaluation of breast vascularity by DCE-MRI may have a role in the evaluation of tumor response, identifying patients who better benefit from therapy.

Hanneke W.M. van Laarhoven¹, Jacques A. van Asten¹, Ingrid Desar¹, Gian Paolo Rizzardi², Claudio Bordignon², Cornelis Punt¹, Carla van Herpen¹, Arend Heerschap¹

¹Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands; ²Molmed, Milan, Italy

Targeted delivery to the tumor of picogram doses of TNF-alpha can be achieved by coupling TNF-alpha with CNGRC (NGR-TNF), a peptide that specifically targets tumor neovasculature. The aim of this study was to asses the effect of NGR-TNF on solid tumors and healthy liver tissue with dynamic contrast enhanced MRI (DCE-MRI) during a phase I trial with NGR-TNF. Two hours after treatment we observed a significant decrease in kep and Ktrans in solid tumors compared to baseline, but not in healthy liver tissue. This suggests a tumor-specific anti-vascular effect of NGR-TNF.

Anwar R. Padhani¹, Dow-Mu Koh², NJ Jane Taylor¹, James J. Stirling¹, David J. Collins², Gordon J. Rustin¹, Adrian Harris³, R Sinha, Jane Boxall¹, J Smythe³, N Fisher⁴, Hillori Connors⁵, Ian Judson, Paul Nathan¹

¹Mount Vernon Cancer Centre, Northwood, UK; ²Institute of Cancer Research, Sutton, UK; ³John Radcliffe Hospital, Oxford, UK; ⁴National Blood Transfusion Service, Oxford, UK; ⁵Oxigene Inc., Waltham, Massachusetts, USA

In this study we document reversible tumour vascular shutdown following the first administration of the vascular disruptive agent (CA4P) and show failure of recovery of tumour vasculature when CA4P is given with the anti-VEGF antibody bevacizumab. Marked heterogeneity in individual lesion responses were seen but cohort analysis convincingly shows failure of tumour vasculature to recover when CA4P is given with bevacizumab. This is the first mechanistic demonstration in humans of increased anti-vascular action of CA4P and bevacizumab using DCE-MRI.

Bob Lei Hou¹, Sasan Karimi², Lauren E. Abrey², Andrei I. Holodny², Philip H. Gutin²

¹MSKCC, New York City, USA; ²MSKCC, New York City, New York, USA

Anti-angiogenic therapy using bevacizumab (Avastin) affects abnormal neovascularity of high grade primary brain tumors. We used DCE and DSE perfusion techniques on 9 patients with recurrent high grade gliomas before and after the first cycle of bevacizumab.  Mean bolus wash-in slopes, Ktrans, fBV, and rCBV ratios for the 9 patients decreased respectively from 2.07+/-1.77 to 0.84+/-0.85, 1.86 +/-0.80 to 1.26+/-0.21, 1.83 +/-0.85 to 1.36+/-0.71, and 1.92+/-0.99 to 1.20+/-0.28.  Ktrans ratio (p=0.043) and bolus wash-in slopes (p=0.015) before and after treatment were significantly different suggesting the bolus wash-in slope from DCE perfusion is the most sensitive parameter for tumor vessel changes.

Dow-Mu Koh¹, ², Anwar Padhani³, Matthew Blackledge⁴, Ben Wilton⁴, David J. Collins⁴, Toni Wallace², N J. Taylor³, James J. Stirling³, Rajesh Sinha², Hillori Connors⁵, Martin O. Leach⁴, Ian Judson⁴, Paul Nathan⁶

¹Institute of Cancer  Research, Sutton, UK; ²Royal Marsden Hospital, Sutton, UK; ³Mount Vernon Hospital, UK; ⁴Institute of Cancer Research, Sutton, UK; ⁵Oxigene Inc, USA; ⁶Mount Veron Hospital, UK

The aims of this study were to determine the ADC measurement variability in a two centre clinical trial, and to track ADC changes in response to the administration of the vascular disrupting drug Combrestatin A4 phosphate (CA4P) and the anti-VEGF antibody bevacizumab. We found that ADC showed excellent reproducibility. However, mean ADC values proved to be insensitive to treatment related effects in our study and bio-informatics based approaches to data analysis should be developed.