The Microcirculation in the “Target Node“ as Outcome Prognosticator: Facts and Implications

In the recent paper of Kim et al. [1], the authors attempt for first time to examine the relationship between pharmacokinetic parameters, obtained by dynamic contrast-enhanced (DCE)-MRI, of a metastatic target node and treatment outcome in patients with neck cancer. The paper makes 3 important contributions to the DCE neck imaging: 1) adding to the evidence gained by Cao et al. [2], Kim et al. derived (based on a two-compartment pharmacokinetic model) quantitative perfusion-associated parameters 2) similarly to the work of Bisdas et al. [3] microcirculation parameters (other to blood flow, blood volume, and permeability) such as Ktrans (transfer constant), ve (extravascular extracellular space volume fraction) and τi (intracellular water lifetime) are introduced in the characterization of neck cancer; 3) for first time Kim et al. examine exclusively the pre-treatment microcirculation parameters of nodal disease in neck cancer, trying to evaluate their predictive value. But let’s take a closer look to these 3 important aspects of the paper.

The quantification of the perfusion parameters in neck cancer is valuable as the quantitative results may facilitate an objective disease monitoring in the same institution and, under certain circumstances, an interchangeability across different institutions. Nowadays, theoretical models deliver quantitative information (of course under certain inevitable assumptions concerning the relationship between MR signal and contrast agent concentration) which are obviously superior to heuristic (semi-quantitative) DCE parameters, such as peak enhancement, maximum upslope, time-to-peak enhancement, and washout slope. In the future, DCE-MRI should be besides CT a major player in this field and combined with diffusion-weighted sequences and spectroscopy may face equally the PET/CT.

Kim et al. focused on the nodal disease, which is a rather unattended aspect in the DCE imaging of neck cancer. The authors found significantly elevated baseline Ktrans in responders, which presumably has led to a better distribution of the chemotherapeutics than in the non-responders who had lower Ktrans values. This seems logical but on the other hand we should bear in mind that high Ktrans may imply severe neoangiogenesis, which, in turn, implies a more aggressive tumor with possibly higher microvascular density. Lower Ktrans may also be the result of necrotic areas thus, poor oxygenation and poor response to radiotherapy. Apparently, the interpretation of the microcirculatory parameters should not be one-sided and in a concomitant chemoradiation setting is definitely difficult to separate the effects of each therapy and draw easily logical conclusions. The authors could not demonstrate except of any significant association between ve, τi and response to therapy. This is not necessarily a drawback of the method but may reflect the heterogeneity of the volume of the target node as well as the different induction chemotherapy regimens across the patients. In a point of view what actually play the most important role are not the baseline microcirculatory parameter values themselves but how they shift during the therapy and after its completion. Obviously, the nodal and tumor response to therapy may have different time points, which are crucial for the further treatment planning. As expected, patients with small nodes in the present study [1] were complete responders after the preoperative chemoradiation, however, some of them had distant metastasis after 6 months. In other words, the 6-month follow-up is more reliable time point for deciding the predictive value of the DCE-imaging parameters. Furthermore, Kim et al. by demonstrating the feasibility of their method posed a very important question: how shall we analyse the nodal disease by means of DCE-MRI? Shall we calculate the microcirculation parameters on a single target node, on a node-to-node basis, or shall we average them?

The results in the presented paper are initial and in a small patient population, thus, far from drawing definite thresholds, cut-off values and significant predictive parameters. Ideally, the work of Kim et al. should trigger DCE-MRI studies that would: 1) compare the microcirculation parameters of histologically confirmed metastatic and reactive lymph nodes, 2) investigate the alteration of microcirculation parameters during the course of chemoradiation, defining the optimal time points for monitoring, and 3) compare the microcirculation parameters of tumoral and nodal residual disease/recurrence and chemoradiated neck tissue. Only under these premises, we would be able to use DCE-MR imaging as a diagnostic clinical tool and, thus, estimate the real predictive value of the microcirculation parameters.


1. Kim S, Loevner LA, Quon H, Kilger A, Sherman E, Weinstein G, Chalian A, Poptani H. Prediction of Response to Chemoradiation Therapy in Squamous Cell Carcinomas of the Head and Neck Using Dynamic Contrast-Enhanced MR Imaging. AJNR Am J Neuroradiol. 2009 Oct 1. [Epub ahead of print]
2. Cao Y, Popovtzer A, Li D, Chepeha DB, Moyer JS, Prince ME, Worden F, Teknos T, Bradford C, Mukherji SK, Eisbruch A. Early prediction of outcome in advanced head-and-neck cancer based on tumor blood volume alterations during therapy: a prospective study. Int J Radiat Oncol Biol Phys. 2008;72:1287-90.
3. Bisdas S, Baghi M, Wagenblast J, Vogl TJ, Thng CH, Koh TS. Gadolinium-enhanced echo-planar T2-weighted MRI of tumors in the extracranial head and neck: feasibility study and preliminary results using a distributed-parameter tracer kinetic analysis. J Magn Reson Imaging. 2008;27:963-9.

The Microcirculation in the “Target Node“ as Outcome Prognosticator: Facts and Implications
Sotirios Bisdas • Neuroradiology, University of Tuebingen