|Home | About | Journals | Submit | Contact Us | Français|
To investigate the effectiveness of a polydisulfide-based biodegradable macromolecular contrast agent, (Gd-DTPA)-cystamine copolymers (GDCC), in assessing the efficacy of indocyanine green enhanced photothermal cancer therapy using dynamic contrast enhanced MRI (DCE-MRI).
Breast cancer xenografts in mice were injected with indocyanine green and irradiated with laser. The efficacy was assessed using DCE-MRI with GDCC of 40 KDa (GDCC-40) at 4 hours and 7 days after the treatment. The uptake of GDCC-40 by the tumors was fit to a two-compartment model to obtain tumor vascular parameters, including fractional plasma volume (fPV), endothelium transfer coefficient (KPS), and permeability surface area product (PS).
GDCC-40 resulted in similar tumor vascular parameters at three doses with larger standard deviations at lower doses. The values of fPV, KPS and PS of the treated tumors were smaller (p < 0.05) than those of untreated tumors at 4 hours after the treatment and recovered to pretreatment values (p > 0.05) at 7 days after the treatment.
DCE-MRI with GDCC-40 is effective for assessing tumor early response to dye-enhanced photothermal therapy and detecting tumor relapse after the treatment. GDCC-40 has a potential to non-invasively monitor anticancer therapies with DCE-MRI.
Laser thermal ablation is an effective cancer therapy used in clinical practice. First introduced in 1983, laser ablation (1) has been used for the treatment of tumors throughout the body, including head and neck (2), liver (3,4), breast (5,6), and etc. Near infrared laser has relative deep tissue penetration and is commonly used for laser ablation (7, 8). Organic dyes that absorb infrared laser are often used to enhance the therapeutic efficacy of laser tumor ablation. Indocyanine green is a clinically approved water soluble dye (9) and has strong absorption at 800 nm in plasma (10). It has shown the effectiveness of causing tumor cell destruction and enhancing laser ablation of tumors in preclinical studies (11–13). The challenge for dye-enhanced photothermal therapy is to completely ablate and eradicate tumor tissue (13). MRI can provide accurate target localization, instrument visualization, online temperature monitoring, and assessment of therapeutic efficacy (6). Contrast enhanced MRI is able to provide accurate evaluation of completeness of tumor ablation and to detect the residual tumor (14, 15). Dynamic contrast enhanced MRI (DCE-MRI) is effective for rapid assessment and prediction of tumor response to anticancer therapies, including laser ablation, based on the changes of tumor vascular parameters, including fractional plasma volume (fPV), endothelium transfer coefficient (KPS) and permeability surface area product (PS), before any morphological changes can be observed (16).
Paramagnetic gadolinium(III) chelates, including Gd-DTPA, Gd-DOTA and their derivatives, are MRI contrast agents approved for clinical uses. However, these agents are small molecular chelates and often over-estimate the tumor vascular properties with DCE-MRI in evaluating tumor response to therapies (17). Macromolecular gadolinium(III) chelates (MW > 20 KDa) are reported to provide more accurate determination of parameters of tumor vascularity because they have limited diffusion through normal vasculature and are able to discriminate leaky microvessels from normal vasculature (18–20). Unfortunately, macromolecular contrast agents can not proceed into clinical development because they excrete slowly from the body and result in long-term tissue accumulation of toxic Gd(III) ions (21, 22). To alleviate this problem, a novel class of polydisulfide-based macromolecular Gd(III) complexes has been recently developed as biodegradable macromolecular MRI contrast agents (21,23–26,29). As shown in animal models, these agents initially retain the properties of macromolecular contrast agents. They can be readily degraded into small chelates and rapidly excreted from the body with minimal tissue Gd(III) accumulation comparable to small molecular weight contrast agents (21,24–26,29). These agents have demonstrated advantageous features over currently available clinical low molecular contrast agents and other reported macromolecular MRI contrast agents in terms of effective contrast enhancement and rapid elimination after the MRI examinations. The biodegradable macromolecular contrast agents are promising for further clinical development as macromolecular contrast agents.
Accurate and timely evaluation of tumor response is critical in assessing therapeutic efficacy for further optimizing cancer therapies and improving patient survival. The biodegradable macromolecular MRI contrast agents have a promise to be used for image-guided laser ablation and accurate assessment of tumor response to the therapy. In this study, we investigated the effectiveness of a biodegradable macromolecular contrast agent, Gd-DTPA) cystamine copolymers (GDCC), in assessing tumor response to indocyanine green enhanced photothermal therapy with DCE-MRI in a mouse tumor model bearing MDA-MB-231 human breast carcinoma xenografts. The dose effect of GDCC was also evaluated to identify the minimally effective dose.
Female athymic nude mice (24–32 grams, Frederick, MD, National Cancer Institute) were cared for under the guidelines of a protocol approved by the University of Utah Institutional Animal Care and Use Committee. The MDA-MB-231 human breast cancer cell line was cultured in the complete medium (Leibovitz’s L-15 medium with 2 mM L-glutamine and 10% fetal bovine serum) at 37°C in a humidified atmosphere of 5% CO2. 5×106 cells in a mixture of 50 μL complete medium and 50 μL Matrigel (Becton-Dikinson, Franklin Lakes, NJ) were inoculated subcutaneously on the hips (both left and right) of the mouse. When the tumor size reached about 300 mm3, they were subjected to laser ablation treatment. The tumor size was monitored regularly using digital caliber and calculated using the ellipsoid volume formula: tumor volume = π/6×A×B×C, where A and B are the equatorial diameters, and C is the polar diameter. Relative tumor volume was calculated as the ratio of the tumor volume at day 7 or day 12 to the tumor volume at Day 0 (relative tumor volume = tumor volumeday 7,12/tumor volumeday 0).
The details of experiment conditions are included in Table 1. Six mice with 2 tumors each were used in the experiment. One tumor in each mouse was treated and the other one was used as an untreated control. When the tumor volume reached approximately 300 mm3, aqueous solution of indocyanine green (Sigma, St. Louis, MO, 1.5 wt-%, 100 μL) was prepared and injected intratumorally 4 hours before the treatment. The fiber tip was maintained at a 5 mm distance from the skin overlying the tumor and the light dose was evenly distributed to the whole tumor. To modify the radial temperature distribution and decrease skin temperature, cold water was dripped from the top of the tumor for dye enhanced photothermal therapy. The tumor bearing mice were sacrificed 2 weeks after the treatment.
The mice were anesthetized by the intraperitoneal administration of a mixture of ketamine (90 mg/kg) and xylazine (10 mg/kg). Three mice with an untreated control tumor and a treated tumor each after indocyanine green enhanced photothermal therapy were scanned on a Siemens Trio 3T MRI scanner. The system body coil was used for RF excitation and a human wrist coil was used for RF reception. DCE-MRI with GDCC-40 (0.05 mmol-Gd/kg) was performed at 4 hours and 7 days after the treatment to evaluate tumor response and therapeutic efficacy. To test the dose effect on vascular parameters, DCE-MRI with GDCC-40 at the doses of 0.01 and 0.025 mmol-Gd/kg was performed in tumor bearing mice.
Two-dimensional fast low angle shot pulse sequence (2D FLASH) images were repetitively acquired for 16 minutes for DCE-MRI with the following parameters: TR/TE = 104/4.46 ms, α= 30°, 0.5 × 0.5 × 1.5 mm, average = 1, acquisition time = 11 seconds, and a total of 10 slices per acquisition. GDCC-40 saline solution was injected intravenously via a tail vein cannulation 45 seconds after the 2D FLASH started in order to acquire 4 acquisition for the calculation of baseline signal intensity (SI). 2D spin echo (SE) images were acquired before and at the end of DCE-MRI using the following parameters: TR/TE = 400/10 ms, α = 90°, 0.4 × 0.4 × 2 mm, average 2, acquisition time = 61 seconds.
The 2D SE images were reconstructed and analyzed using Osirix (http://homepage.mac.com/rossetantoine/osirix). A package of programs based on MATLAB (The MathWorks, Inc., Natick, MA, USA) was developed to process dynamic 2D FLASH data in the digital imaging and communications in medicine (DICOM) format. A two-compartment model consisting of the tumor plasma compartment and the extravascular and extracellular space (EES) was used for analyzing DCE-MRI data. Tumor vascular parameters, including fractional plasma volume (fPV), endothelium transfer coefficient (KPS) and permeability surface area product (PS), were similarly calculated by fitting SIs to the two-compartment model using the methods as described in the literature (27). It is assumed that ΔSI is proportional to the change of the contrast agent concentration, which is a reasonable approximation at low contrast agent concentration (27). Since a high temporal resolution of 11 second per acquisition was used in this study, the maximum ΔSIblood was used directly to calculate the fPV with the two-compartment model. 2D fPV, KPS, and PS maps of tumor tissue were constructed by pixel by pixel calculation using the MATLAB program with the two-compartment model. Regions of interest for the whole tumor were manually drawn on the parameter (fPV, KPS, and PS) maps to calculate the average values using the MATLAB program.
The control and treated tumors were collected and fixed in buffered formalin, and embedded in paraffin after the mice were sacrificed two weeks after the treatment. Hematoxylin & eosin stained tumor slices (5 μm) were examined by a pathologist (LE) who was blinded to the treatment of the tumors. An estimation of tumor necrosis was made for each tumor by the following grading scheme: scores 1, 2 and 3 represent 0 – 33 %, 34 – 65 %, and 66 – 100 % tumor necrosis, respectively.
Statistical analysis was performed using a student t-test (GraphPad Prism; GraphPad Software, San Diego, CA). P values were two-tailed with a confidence interval of 95%. The difference was considered significant when p < 0.05.
Indocyanine green has a strong absorption at 810 nm. The presence of indocyanine green in solid tumor can significantly increase tumor absorption of laser photon and improve the efficacy of laser ablation. The tumor volume was in the range of 184 to 710 mm3 with an average size of 290 mm3 for the control group. The tumor volume of the treatment group was in the range of range from 189 to 763 mm3 with an average size of 316 mm3. Tumor sections along the path of laser irradiation had reduced volumes after the treatment, indicating the efficacy of indocyanine green enhanced photothermal therapy. The relative tumor volumes at 7 days and 12 days after the treatment are listed in Table 1. The growth of the treated tumors was arrested by the dye enhanced photothermal therapy and the size of the treated tumors was significantly smaller compared to the control tumors 7 days (p = 0.006) and 12 days (p = 0.003) after the treatment. The control tumors had minimal central necrosis with a score of 1 at the end of experiment. The treated tumors had an average necrosis score of 2.
The dose effect of the biodegradable macromolecular contrast agent GDCC-40 on DCE-MRI was evaluated in untreated tumors at reduced doses. Table 2 lists the signal enhancement in SE images of the tumor tissue at 16 minutes post-injection with a spin-echo sequence and the vascular parameters determined by the DCE-MRI at different doses. At 16 minutes post-injection, tumor enhancement decreased with reduced doses. Approximately 62% ± 19% tumor enhancement was observed for GDCC-40 at the dose of 0.05 mmol-Gd/kg, 26% ± 5% at 0.025 mmol-Gd/kg and 14% ± 3% at 0.01 mmol-Gd/kg. The 14% signal enhancement at the lowest dose (0.01 mmol-Gd/kg) still resulted in visible enhancement in tumor tissue at 16 minutes post-injection. Although lower doses resulted in lower signal noise ratio for image analysis, the vascular parameters determined at the tested doses were not significantly different from each other (p > 0.05). However, DCE-MRI with lower doses resulted in larger deviations of the vascular parameters.
Figure 1 shows the representative T1-weighted 2D spin-echo axial images of a mouse before and 16 minutes after the injection of GDCC-40 at a dose of 0.05 mmol-Gd/kg at four hours after the treatment. Strong enhancement was observed in the untreated control tumor. In contrast, the treated tumor showed heterogeneous enhancement. Little enhancement was observed for the tumor tissue close to the laser irradiation due to the vascular damage caused by the dye-enhanced photothermal therapy. Relatively strong enhancement was observed in deep tumor tissue away from the laser irradiation, but the enhancement was less than that in the untreated tumor.
Figure 2 shows the representative time courses of the signal intensity in the tumor rim of the untreated tumor, the SI in the muscle around the tumor tissue, and their ratios of DCE-MRI after the injection of GDCC-40 at 0.05 mmol-Gd/kg. The SI reached plateau in muscle approximately 1.5 minutes after the injection and the maximum in the tumor rim at 12 minutes for GDCC-40. The enhancement ratio increased during the period of experiment and was more than 3 at 15 minutes.
Table 3 lists the tumor vascular parameters (fPV, KPS, and PS) of the treated tumors and control tumors estimated using a two-compartment model (27) from the DCE-MRI at 4 hours and 7 days after the treatment. Dye enhanced photothermal therapy resulted in significant reduction of the tumor vascular parameters. The values of fPV, KPS, and PS of the treated tumors at 4 hours after the treatment were approximately 43%, 34%, and 11 % of those of the untreated tumors (p = 0.01, 0.01 and 0.02 for fPV, KPS, and PS, respectively). The vasculature recovery of the treated tumors were observed at 7 days after the treatment based on the increase in fPV, KPS, and PS values. All three parameters of the treated tumors 7 days after the treatment were not significantly different from those of the untreated ones (p values were 0.97, 0.23, 0.42 for fPV, KPS, and PS, respectively).
Figure 3 shows the representative fPV and KPS maps of both control and treated tumors constructed by pixel-to-pixel analysis of the DCE-MRI data. The treated tumor showed more heterogeneous distribution of fPV and KPS values than the untreated control tumor. Significant decrease of fPV and KPS in the tumor tissue along the path of laser irradiation was shown in the maps of the treated tumor as compared to those of the untreated tumors at both 4 hours and 7 days after the treatment. The tumor tissue away from laser irradiation showed relatively high fPV and KPS at 4 hours after treatment, which increased 7 days later.
This study is the first attempt of using the biodegradable macromolecular MRI contrast agent GDCC to monitor the efficacy of anticancer therapies based on the vascular parameters estimated by DCE-MRI. GDCC with a narrow molecular weight distribution and a molecular weight of approximately 40 KDa was selected in the study because it has a similar hydrodynamic volume as albumin-(Gd-DTPA) (29,30,31), a prototype macromolecular contrast agent that has been extensively investigated for tumor characterization. The vascular parameters estimated using GDCC-40 in untreated tumor tissues were comparable to those with albumin-(Gd-DTPA). For example, the average fPV values of the untreated tumors estimated using GDCC-40 was in the range of 0.069 to 0.080, close to those estimated using albumin-(Gd-DTPA) (0.034 to 0.065) (27, 29), while fPV estimated with a low molecular weight agent Gd(DTPA-BMA) could be as high as 0.16 (29). The KPS values of untreated tumors estimated by GDCC-40 was in the range of 7 to 8.5 ml/min/100 cc, similar to the reported KPS estimated with albumin-(Gd-DTPA) (1.4 to 13.5 ml/min/100 cc), while Gd(DTPA-BMA) gave much higher KPS (29 ml/min/100 cc) (27, 29).
GDCC-40 has shown similar effectiveness in characterization of tumor vascular parameters as the extensively studied macromolecular agent albumin-(Gd-DTPA) (29). The biodegradable macromolecular contrast agent is advantageous over albumin-(Gd-DTPA) because it can degrade in low molecular weight complexes that can be rapidly excreted after the DCE-MRI studies. Our previous studies showed that GDCC had minimal long-term tissue accumulation comparable to low molecular weight contrast agents (25). In comparison, albumin-(Gd-DTPA) excretes slowly from the body, resulting in high tissue accumulation. This study also showed that the dose of GDCC could be further reduced to as low as 0.01 mmol-Gd/kg in DCE-MRI studies. Application of the biodegradable contrast agent at a low dose has a potential to reduce the potential Gd-related toxic side effects, e.g. nephrogenic systemic fibrosis (32).
Dye-enhanced photothermal therapy damaged the tumor vasculature and resulted in tumor growth arrest. Dynamic contrast enhanced MRI with GDCC-40 revealed that the tumor vascular parameters significantly decreased at 4 hours after the dye-enhanced photothermal therapy as compared those of untreated tumors. The results indicate that DCE-MRI with GDCC-40 is able to provide accurate assessment of early tumor response to the therapy. Both fPV and KPS maps of the treated tumors showed heterogeneous vascular parameters cross the tumor tissue as compared to the untreated tumors. The deep tumor tissue had relatively high fPV and KPS values, indicating low tumor response as shown in the maps, which was possibly due to non-homogeneous distribution of indocyanine green and/or uneven laser energy distribution due to the laser attenuation along the irradiation path. Indocyanine green was injected intratumorally and its diffusion in the tumor tissue could be affected by tumor heterogeneity. Heterogeneous tumor response resulted in tumor relapse 7 days after the treatment as shown by DCE-MRI with GDCC-40. At 7 days after treatment, the treated tumors had similar average fPV and KPS values as the control tumors (p > 0.05). The fPV and KPS maps showed increased fPV and KPS values in deep tumor tissues where low early response to the treatment was observed at 4 hours after the treatment. These observations also correlated well to the increased tumor volume at 7 days after the treatment. DCE-MRI with the biodegradable macromolecular contrast agent can predict the early therapeutic response and as well as relapse before any tumor size change is detected.
It is also interesting to note that DCE-MRI with the biodegradable macromolecular contrast agent may be able to differentiate the enhancement of inflammation from that of tumor angiogenesis. Significant enhancement was observed at the tumor surface irradiated by laser in both precontrast and postcontrast images as compared to the rest treated tumor tissue at 4 hours after the treatment (Figure 1). The strong enhancement at tumor surface may be the result of inflammatory damage caused by heating from the un-attenuated laser. However, the inflammatory region did not have high fPV and KPS values as shown in the fPV and KPS maps constructed from the DCE-MRI data, Figure 3a and c. The fPV and KPS values in the region were lower than the deep tissue of the treated tumors. This indicates that DCE-MRI with GDCC-40 is able to distinguish tumor from inflammatory tissues based on vascular parameters.
In conclusion, the biodegradable macromolecular MRI contrast agent GDCC-40 is effective to characterize tumor vascular properties in DCE-MRI at a dose as low as 0.01 mmol-Gd/kg. DCE-MRI with GDCC-40 can assess early tumor response to indocyanine green enhanced photothermal therapy and the efficacy of therapy and detect relapse of treated tumor. The biodegradable macromolecular contrast agent is promising for non-invasive assessment of tumor response to anti-cancer therapies and the efficacy of the therapies.
Grant sponsors: National Institutes of Health (NIH) Grant R01 EB00489 and a Research contract from the Cao Group Inc.
The authors also thank Melody Johnson for technical assistance for MRI data acquisition, Dr. Yong-En Sun for animal handling, and Thanh Nguyen for her assistance with laser characterization.