Introduction

Accurate chemotherapy monitoring of cancer provides clinically useful information for treatment response and its prognostication. In Hodgkin lymphoma (HL), 18-fluorodeoxyglucose (FDG) Positron Emission Tomography (PET) with CT (PET/CT) is most commonly used for diagnosing, staging and interim monitoring using quantitative parameter such as standard uptake value (SUV)¹. However, PET/CT uses ionizing radiation, and radiotracers is sometimes contraindicated in few patients because of associated complications. Non-invasive monitoring can be achieved using Intravoxel incoherent motion (IVIM) and Diffusion kurtosis imaging (DKI) which provides microstructure and macrovascular information of tumor without any contrast injection^2,3. Objective of this study was to investigate the role of IVIM and DKI in interim response assessment of HL in comparison with clinical standard, PET imaging.

Methods

Patients and Interim assessment: Three patients with Hodgkin lymphoma (n=3; age: 42.33±13.58 years; M:F=1:2) were recruited and underwent biopsy procedure after the ethical approval of Institutional Ethics Board and written informed consent. All patients were treated with standard ABVD (Adriamycin, Bleomycin-sulfate, Vinblastine-sulfate, and Dacarbazine) regime. Interim assessments were done after four cycles of chemotherapies i.e. within 8 weeks.
18F-FDG PET/CT acquisition and analysis: PET/CT examinations were performed with a PET/CT scanner (Biograph Mct; Siemens Healthcare,Henkestr,Erlangen,Germany) and Discovery 710 (General electric company, Fairfield, Connecticut, USA). Dosage of 6–11 MBq/kg (0.16–0.18 mCi/kg; minimum, 3 mCi) FDG intravenously injected. After the 45–60-minute uptake period, the patients were taken for the PET/CT study. SUVmax (maximum Standard Uptake Value) and SUVmean were calculated using equation below:
$$SUV=\frac{r}{a'/w}$$
where, r=radioactivity activity concentration (kBq/ml), a′=decay-corrected amount of injected radiolabeled FDG (kBq), and w=weight of the patient (g).
MR acquisition and Analysis: All patients were scanned in 1.5T MRI (Ingenia; Philips Healthcare, Best, the Netherlands) at AIIMS, New Delhi, India, with a STIR (Thoracic:TR=1.503s and TE=0.09s;Abdomen: TR=1.503s and TE=0.06s), including IVIM-DKI with 9b-values= 0,35,50,100,175,300,500,1500,2000 s/mm2 using phased-array surface coil (Thoracic: TR=12.44s, TE=0.081s; Abdomen: TR=12.44s, TE=0.081s) for thoracic and abdominal area.
All parameters were estimated using Non-linear least square optimization based IVIM-DKI model with in-house toolbox using MATLAB. ADC was calculated voxelwise using monoexponential model.
IVIM parameters such as D, D*, and f were evaluated using two methods i) BE model² and ii) BE with Total variation penalty function (TV)^4,5. BE model is defined below²:
$$S⁄S_0=fexp(-bD^* )+(1-f)exp(-bD)$$
where S and S₀ are diffusion signals with and without diffusion gradient b in s/mm². In BE+TV model, image gradient was calculated with parameter values updated iteratively (TV parameters, alpha and beta set to 0.005 and 0.99)⁵.
DKI parameters such as Dk and k were analyzed using DKI model³:
$$S⁄S_0=exp(-bD_k+b^2D_k^2 K/6)$$
ROI localization: Image DWI at b=2000 s/mm² (hyperintense), ADC map (hypointense) and PET-SUV maps (hyperintense) were used to localize tumor as shown in Figure1.

Results

Diffusion, Kurtosis and Perfusion-related parameters in Baseline and Interim monitoring of HL: Figure2, circled area shows tumor region appeared hypointense in ADC, Dk D (BE), D (BE+TV) and k map at baseline for a representative patient. After treatment, intensity in tumor region reduced and corresponding changes can be seen in D (BE+TV), Dk, k map distinctly.
Figure3 shows D* parameter obtained from BE and BE+TV model, appeared hypointense at baseline as compared to interim assessment whereas f showed opposite trend. D* (BE) at both times showed noisy maps.
Quantitative assessment of IVIM and DKI parameters with PET parameters: Figure4 shows ADC, Dk, D (BE+TV), f (BE) and f (BE+TV) mean values reduced at interim measurement as compared to baseline, same changes were also seen in PET parameters such as SUVmax and SUVmean. Whereas, D* (BE) and D* (BE+TV) with k showed increased mean values at interim assessment. D (BE) showed no changes w.r.t treatment response.

Discussion and Conclusions

IVIM analysis consists of diffusion and perfusion parameter explaining the structural and dynamic information of tissue in the tumor region whereas DKI parameters provides information on tissue heterogeneity. To obtain spatial homogeneity in parameter maps, total variation penalty function was included into BE model to remove any spurious values in maps^4,5.
Previous literature has shown D, D* and f parameters to be good biomarker for treatment response^6-8. In this study, quantitative and qualitative assessment of IVIM and DKI parameters were performed in comparison with PET parameters for patients with HL. Diffusion related parameters such as ADC, Dk, D (BE+TV) showed decreasing trend in chemotherapy interim assessment which is in accordance with PET indicating a positive treatment response⁹. Kurtosis showed opposite trend with increase during response interim assessment, indicating increase in tissue heterogeneity. Perfusion-related parameters such as f showed reduction and opposite trend was observed with D* for response assessment which is in accordance with literature^6-8. However, due to small sample size, no statistical analysis could be performed. Still with a limited dataset this preliminary study showed good evidence to support further exploration of IVIM & DKI in chemotherapy response assessment for Hodgkin Lymphoma and to evaluate if it can be used in place of PET imaging.

Acknowledgements

This study was supported by IIT Delhi and AIIMS Delhi. AVM was supported by research fellowship fund from Ministry of Human Resource Development, Government of India.