Abstract
In a recent article published in Acta Radiologica, Hagtvedt and colleagues reported on reliability of diffusion-weighted magnetic resonance imaging (DW-MRI) and 18F-FDG/PET-CT for assessing early treatment response in 27 patients with lymphomatous lymph-nodes of the neck (1). By comparing changes in quantitative parameters measured before and after the onset of the first cycle of chemotherapy, they found, in responder patients, a significant reduction of SUVmax but unchanged apparent diffusion coefficients (ADCs). Hence, they concluded that the ADC is an ineffective marker of early response and that DW-MRI cannot replace PET-CT for evaluating early treatment response in lymphoma. Although they stated that the different acquisition time of DW-MRI and PET-CT after the first administration of treatment was a limitation of their study (mean time, 2 days versus 19 days, respectively), some considerations should be highlighted in order to explain the “misleading” inadequate outcome of DW-MRI reported by the authors.
From 1 to 4 days after first administration of chemotherapy, they found a mean ADC change of 5.3% that ranged from a decrease of 0.28 × 10−3 mm2s−1 to an increase of 0.20 × 10−3 mm2s−1 being significantly greater in non-Hodgkin’s lymphoma (NHL) (16.2%, 8 subjects with DLBCL) compared to Hodgkin’s lymphoma (HL) (0.8%, 19 subjects). Although some studies demonstrated that ADC increases were noted as early as 3–7 days after the first dose of chemotherapy in different organs (2), studies on lymphoma that showed an increase of ADC as early as the first week after administration of chemotherapy have evaluated only NHL (3,4). Conversely, for HL, the ADC changes induced by chemotherapy may be observed later compared to NHL, because in HL the nodal tissue is represented by Reed-Sternberg cells admixed within an abundant reactive cell infiltrate (5). The early inflammatory changes immediately induced at the onset of chemotherapy in this matrix can lead to a transient decrease of ADC in the first days after chemotherapy, a condition that occurred in the study by Hagtvedt (1). Moreover, another factor that could explain a transient decrease or steady state of ADC in responder patients, soon after initiation of therapy, is cellular swelling that has been noted to occur in the early phases of apoptosis in response to anticancer treatment (2). Thus, similarly to PET-CT, DW-MRI may be influenced by early post-therapeutic changes, such as tissue edema and inflammation, that can occur immediately after first administration of chemotherapy, especially in HL (2,5).
In addition, simple manual ADC measurements of central tendency of lymphoma with evaluation of mean or median changes in ADCs obtained on a single slice, as the authors made, have a limited ability in detecting treatment-related changes, particularly if there are areas of both increase and decrease in ADCs for tissue heterogeneity, as can occur in HL. Thus, tumors that are heterogeneous in their spatial ADC distribution for their histological composition or for variable treatment response can be better evaluated with histogram analysis, in which the ROI is identified on multiple scans and a plot of the number of voxels at each ADC value is depicted as a histogram (2) (Fig. 1). Hence, changes in histogram shape or descriptors can be more precisely correlated with the overall tissue response to treatment (2).
A 34-year-old woman with nodular sclerosis Hodgkin’s lymphoma of the anterior mediastinum (arrows). The ADC-MAP of DW-MRI obtained at diagnosis (left image) demonstrates a soft tissue mass with heterogeneous low signal intensity that reflects high cellular density and cellular atypia. The areas of high signal intensity correspond to the fibrotic counterpart of the tumour typically detected in nodular sclerosis HL. The ADC-MAP obtained 20 days later the first administration of the first ABVD cycle (right image) shows that the signal intensity is diffusely increased, suggesting early response to treatment. The ADC-MAP histogram analysis of the entire lesion at diagnosis (left image) reveals a mean ADC of 1.2 × 10−3 mm2 s−1 (standard deviation, 0.42 × 10−3 mm2 s−1) with a significant number of voxels with low ADC values (min, 0.49 × 10−3 mm2 s−1; max, 2.7 × 10−3 mm2 s−1). After the first cycle, the histogram analysis of the entire lesion (right image) shows a significant increase in ADC (mean, 2.60 × 10−3 mm2 s−1) with a significant number of voxels with high ADC values (min, 0.51 × 10−3 mm2 s−1; max, 3.7 × 10−3 mm2 s−1). The change in histogram shape demonstrates early response to treatment. (x axis, ADC value; y axis, number of voxels x 10).
In conclusion, this study demonstrates that, similarly to PET-CT, DW-MRI may not be useful in the first days after onset of chemotherapy to predict treatment response, especially in HL. Nevertheless, because the assessment of early treatment response in lymphoma is crucial to stratify risk and predict prognosis in order to aid treatment modification for improving long-term outcome, studies that evaluate changes in ADCs after the end of the first cycle are needed to evaluate the reliability of DW-MRI in assessing early treatment response.