Annals of Surgical Oncology

, Volume 19, Issue 9, pp 3116–3122

Bioluminescence Imaging Serves as a Dynamic Marker for Guiding and Assessing Thermal Treatment of Cancer in a Preclinical Model

  • Joyce T. Au
  • Lorena Gonzalez
  • Chun-Hao Chen
  • Inna Serganova
  • Yuman Fong
Translational Research and Biomarkers

DOI: 10.1245/s10434-012-2313-7

Cite this article as:
Au, J.T., Gonzalez, L., Chen, CH. et al. Ann Surg Oncol (2012) 19: 3116. doi:10.1245/s10434-012-2313-7



Bioluminescence has been harnessed as a dynamic imaging technique in research. This is a proof of principle study examining feasibility of using bioluminescent proteins as a marker to guide therapeutic ablation.


Mesothelioma cancer cells (MSTO-Td) were transfected with a retroviral vector bearing firefly luciferase gene, plated in serial dilutions, and imaged to compare bioluminescence signal to cell number, determining threshold of bioluminescence detection. MSTO-Td cells were subjected to thermal treatment in a heated chamber; the bioluminescence signal and number of remaining live cancer cells were determined. Mice with MSTO-Td xenografts underwent electrocautery tumor ablation guided by bioluminescence imaging; bioluminescence signal and tumor size were monitored for 3 weeks.


MSTO-Td cells emitted a bright, clear, bioluminescence signal that amplified with the cell number (P < .001) and was detectable with as few as 10 cells in cell culture. Bioluminescence decreased in a dose-dependent fashion upon thermal treatment as temperature increased from 37 to 70 °C (P < .001) and as treatment duration increased from 5 to 20 min (P < .001). This correlated with the number of remaining live MSTO-Td cells (Pearson coefficient = 0.865; P < .001). In mice, the bioluminescence signal correlated with tumor size posttreatment and effectively guided the ablation procedure to completion, achieving 0 % tumor recurrence.


Bioluminescence imaging is a sensitive, real-time imaging approach; bioluminescence reporters such as firefly luciferase can assess and guide thermal treatment of cancer. This encourages research into bioluminescence imaging as a molecular technique with potential to target tumors via biomarkers and optimize thermal treatment procedures in a clinical setting.

Radiofrequency ablation, microwave ablation, and laser ablation are common thermal techniques used to treat focal solid malignancies, including colorectal metastases in the liver, prostate tumors, and renal cell carcinomas.13 As the temperature rises, cells progress from susceptibility to damage, to irreversible cellular damage, to cell death by coagulative necrosis.1 Ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI) are current imaging modalities that monitor the ablative procedure and are based on physical properties. We hereby propose that the thermal treatment of cancer may also be guided and assessed by real-time bioluminescence imaging.

Bioluminescence imaging is a molecular imaging technique based on the chemical reaction between a luciferase enzyme, such as the firefly luciferase, and otherwise inactive substrates, which results in the emission of light. Besides being a pretty sight in nature in fireflies and jellyfish, bioluminescence can be harnessed for practical purposes. Since the cloning, gene transfer, and expression of the firefly luciferase gene in 1985, bioluminescence has become an almost ubiquitous imaging reporter in research laboratories.4

Several inherent advantages predicate the utility of bioluminescence imaging of cells and animals in research. A high signal-to-noise ratio of the emitted light confers a quality of exceptional sensitivity for luciferase activity. Furthermore, it is well suited for mammalian tissues where there is very low background autobioluminescence. In addition, the requirement of ATP in the chemical reaction focuses the study on only the live cells, without any confusion with surrounding necrotic tissue and debris. The procedure is also easy to perform and low cost. Bioluminescence imaging has thus been used extensively to study a variety of dynamic biological processes from signal transduction pathways, to angiogenesis, to optimization of gene therapy.5,6 Our objective in this study is to investigate the utility of bioluminescence imaging in monitoring thermal treatment of cancer.


MSTO-Td Cells

Mesothelioma-211 cancer cells were transfected with the retroviral vector bearing the firefly luciferase (FLuc) as previously described (MSTO-Td).7 Briefly, a vector with SFG-tdRFP-cmvFLuc had the firefly luciferase gene under control of the constitutive CMV promoter. After transfection, cells were sorted by fluorescence-activated cell sorting (FACS) using tdRFP red fluorescence protein and cultured in RPMI media for experiments.

Bioluminescence Imaging

D-luciferin was diluted in phosphate-buffered solution to a final concentration of 15 mg/ml and added directly to cells at a volume of 100 μl or injected intraperitoneally into mice at a volume of 200 μl or 150 μg/mg mouse body weight. Bioluminescence was detected with a cooled charge-coupled device (CCD) camera (Xenogen IVIS 200 system, Hopkinton, MA). Images were analyzed with Living Image 2.60 software (Caliper, Hopkinton, MA), and total photon emission in the region of interest was measured in photons/s.

Cell Number and Threshold of Bioluminescence Signal

MSTO-Td cancer cells were counted, serially diluted by a factor of 10, and plated at a cell number from 1 × 106 to 1 × 100 cells in 100 μl of media per well in 96-well plates. Separate plates were used for each cell number. Bioluminescence imaging was performed. Control wells contained media and but no cells.

Thermal Treatment In Vitro

MSTO-Td cells were plated at 1 × 105 cells per well in 6-well plates and incubated overnight for adherence. The next day, the media was aspirated and replaced with fresh media that was preheated in a separate chamber to 60 °C. The plate was then immediately placed in the heated chamber for 5, 10, 15, or 20 min. This was repeated with temperatures of 65 and 70 °C. Bioluminescence imaging was performed immediately after thermal treatment, and 24 h later, a trypan blue assay was performed to count the remaining live cells. Control wells contained MSTO-Td cells incubated at the standard 37 °C temperature without any thermal treatment.

Tumor Ablation in Mice

All animal work was conducted with approval from the Institutional Animal Care and Usage Committee at Memorial Sloan-Kettering Cancer Center. Athymic mice were injected subcutaneously with 1 × 107 MSTO-Td cells to generate xenograft flank tumors. After 1 week, when tumors reached 7 to 10 mm in diameter, mice underwent tumor ablation under general anesthesia with isoflurane at 2 to 3 %. After skin incision and minimal dissection for exposure, the tumor was grasped between forceps and slowly ablated by Bovie electrocautery (Bovie Medical Corporation, Clearwater, FL) at 2 mA on the blend setting for a minimum of 10 s. Bioluminescence imaging was then performed. Six mice underwent tumor ablation for as long as needed until the bioluminescence signal on repeat imaging was completely abolished; this constituted the “extinguished signal group.” Another six mice underwent ablation treatment for 10 s, were verified on imaging to have residual bioluminescence signal, and no further ablation was done; this constituted the “residual signal group.” Three mice had sham procedure with skin incision but no ablation; this was the “sham group” that served as a control. Wounds were closed with clips and buprenorphine administered for analgesia. Postoperative monitoring was carried out with bioluminescence imaging and tumor measurement by calipers for 3 weeks posttreatment.

Statistical Analysis

All statistics were carried out on IBM SPSS 19 software (IBM Corporation, Somers, NY). Comparison of bioluminescence to different cell numbers and different temperatures and duration of thermal treatment was done with one-way ANOVA analysis. Correlation of bioluminescence to remaining live cells and tumor size after ablation was done with bivariate correlation analysis. Comparison of one treatment group of mice to another was done with independent samples t test.


Determining the Threshold of Bioluminescence Signal

No bioluminescence signal was observed in control wells, whereas the bioluminescence signal was clear and bright in all wells with MSTO-Td cells. The strength of bioluminescence rose in proportion to the cell number (P < .001, Pearson correlation = 0.693) (Fig. 1a), and even those wells with only 10 cells demonstrated bioluminescence (Fig. 1b).
Fig. 1

Bioluminescence is well correlated to the cell number and highly sensitive. a As the number of MSTO-Td cancer cells increase, the bioluminescent signal increases. b A bioluminescent signal is detectable with as few as 10 cancer cells

Thermal Treatment In Vitro

Bioluminescence significantly decreased with increasing temperature (P < .001) and duration (P < .001) of thermal treatment (Figs. 2 and 3), with a >6000-fold difference between nontreated cells and cells treated at 70 °C for 20 min. Furthermore, the bioluminescence signal was significantly correlated to the number of remaining live cancer cells after treatment (P < .001, Pearson 0.865) (Fig. 4a). When grouped by level of bioluminescence from <1 × 106 photons/s, to 1 × 106 to 1 × 107 photons/s, to >1 × 107 photons/s, there were no remaining live cancer cells at <1 × 106 photons/s (Fig. 4b).
Fig. 2

Bioluminescence of MSTO-Td is responsive to thermal treatment and diminishes with increasing temperature and duration of thermal treatment

Fig. 3

Bioluminescence imaging can be used to monitor the thermal treatment of MSTO-Td cancer cells, which demonstrate less bioluminescence with rising temperature and duration of treatment

Fig. 4

a Bioluminescence imaging is indicative of the viability of cancer cells after thermal treatment. b As a lower bioluminescent signal is detected, the mean number of remaining live cells found drops. At a signal of <1 × 106 photons/s, there are no live cancer cells left

Tumor Ablation in Mice

After thermal tumor ablation by electrocautery, the absence of bioluminescence persisted in the extinguished signal group of mice, while bioluminescence grew in the residual signal group and the sham group (Fig. 5). At 3 weeks, bioluminescence in the extinguished signal group was significantly lower at 2.64 × 104 photons/s than in the residual signal group, which had a bioluminescence signal of 7.51 × 109 photons/s (P = .032), and in the sham group, which had a bioluminescence signal of 8.67 × 109 photons/s (P = .002). Thus, fully treated mice in the extinguished signal group had a >300,000-fold reduction in bioluminescence signal at 3 weeks after ablation compared with untreated mice. Also, there was no tumor recurrence in mice in the extinguished signal group, while tumor size rose in the residual signal group and the sham group, with a significant correlation of bioluminescence to tumor size (P = .001, Pearson 0.448) (Fig. 6). In one mouse, the bioluminescence signal had spread from the flank to the abdomen, and metastatic lesions were indeed found upon necropsy. Having the bioluminescence signal extinguished by the end of the ablation procedure had a significant effect on tumor response to treatment over time (P = .010).
Fig. 5

Bioluminescence imaging can be used to monitor the effects of thermal ablation of tumor in mice. At 3 weeks follow-up, mice from the sham group and the residual signal group exhibit an increase in bioluminescence in the area of their tumors, while mice from the extinguished signal group still do not exhibit any bioluminescence

Fig. 6

Bioluminescence imaging can be used to monitor thermal ablation to prevent tumor recurrence. a At 3 weeks after treatment, bioluminescence is high in the MSTO-Td tumors of the mice from the sham group (blue) and the residual signal group (yellow) but low from the extinguished signal group (green). b Tumors continue to grow in the sham group (blue) and stay about the same in the residual signal group (yellow), but are absent in the extinguished signal group (green) indicating complete response at 3 weeks follow-up


We began this study by first establishing that bioluminescence intensity was an excellent marker of tumor burden. The strength of the bioluminescence signal was proportional to the number of cancer cells, and the low detection threshold of only 10 cells in vitro showcased how robust the bioluminescence signal was. This impressive threshold has also been noted in the literature to be as low as 4 to 40 cells in vitro and 400 to 1,000 cells in vivo after tumor inoculation, making bioluminescence an incredibly sensitive modality of imaging.813 Presence and strength of bioluminescence has also been correlated to tumor weight, caliper measurement of tumor volume, number of cancer cells in the tissue by FACS analysis, MRI signal, and positron emission tomography signal of the tumor.9,14 This body of evidence validates bioluminescence imaging as a tool for assessing tumor burden.

Bioluminescence imaging is also effective in detecting metastasis. Although unexpected, we discovered bioluminescence signal not only in the flank where the cancer cells were inoculated, but also in the abdomen where the cancer had metastasized as found grossly on necropsy. Previously, our lab had shown that bioluminescence imaging effectively identified both macroscopic and microscopic lymph node metastasis in an animal model of melanoma.15 Other groups have shown that bioluminescence imaging may be used to detect breast cancer metastasis in the bone, adrenal gland, and lungs, and though early bone marrow metastases are difficult to detect on conventional radiological imaging, they can be detected by bioluminescence imaging in a lesion as small as 0.5 mm3.13,16 This ability to identify and localize microscopic lesions makes bioluminescence imaging valuable in the context of dealing with malignancies.

Our study demonstrated that bioluminescence imaging is an effective, quantitative, dynamic imaging approach to assess response to thermal treatment of tumors. The strength of the bioluminescence signal progressively diminished as the intensity and duration of treatment augmented and was correlated to the number of remaining live cancer cells after treatment in vitro. Continuing to in vivo experiments, the strength of the bioluminescence signal was correlated to tumor size during postablation monitoring. We here demonstrate that thermal ablation can immediately change the bioluminescence signal and that such change correlates with effectiveness of treatment. These are proof of concept that bioluminescence imaging can be a sensitive and useful monitoring modality for ablative therapy.

The use of bioluminescence imaging to monitor cancer size and localization may be applied to other oncologic treatment strategies. For instance, bioluminescent imaging has allowed for evaluation of mammary tumors in response to chemotherapy and prostate tumors in response to hormone therapy.17,18 Such use of bioluminescence imaging to reveal the success or failure of treatment by tumor presence, size, and location is common in longitudinal cancer research studies to investigate the effects of various novel, therapeutic agents from small molecule inhibitors to cytotoxic T cells, oncolytic viruses, and small interfering RNA among others.7,9,1922

A novel use of bioluminescence imaging in our study was to guide the thermal ablation procedure itself. In all animals where tumor ablation was continued until the bioluminescence signal was extinguished, there was no tumor recurrence. This remarkable 100 % complete response rate was attributed to the efficacy of bioluminescence imaging in ensuring the totality of ablation during the procedure, thus avoiding disease recurrence and the need for a second procedure. The bioluminescence signal also directly reflected the viability of the tumor in real-time, giving instantaneous feedback on the presence and location of any microscopic residual tumor to be ablated. It did so in a manner specific to only live cancer cells and made them conspicuous without being obscured by necrotic coagulated tissue and debris. Given that microbubbles of gas, inflammation, and other effects of thermal ablation sometimes obscure the margins of ablation on conventional imaging modalities, the specificity for live cells in bioluminescence imaging is a crucial advantage for optimizing treatment.

Bioluminescence imaging is continually being developed with many future prospects. The flaw of a two-dimensional image may soon be overcome as studies work toward adding a third dimension to the image in bioluminescence tomography.23,24 Our lab has used bioluminescence imaging to show hypoxia-driven gene expression in an orthotopic liver tumor model, and others have shown that luciferase can be attached to the therapeutic agent to show how it homes toward the tumor.19,21,25 Thus, bioluminescence imaging can be a versatile imaging adjunct to a specific biological process, agent, or biomarker to target a tumor and its activity. Future studies will need to combine bioluminescence reporters with tumor-specific antibodies or tumor-specific metabolites to allow clinical application of this technology.2631

Whether it is the liver or the prostate, the principles of imaging in thermal ablation therapy are to localize the tumor, guide the ablation process, and assess the effects of ablation on follow-up.1,32 This study has shown the ability of bioluminescence imaging to perform these functions based on its real-time capability as well as its sensitivity and specificity for viable tumor. This encourages research into bioluminescence imaging as a molecular technique with the potential to target tumors via biomarkers and to optimize thermal treatment procedures in a clinical setting.


The authors would like to thank Susanne Carpenter, Joshua Carson, Dana Haddad, Arjun Mittra, Valerie Longo, and Pat Zanzonico for their advice and support of this work.

Copyright information

© Society of Surgical Oncology 2012

Authors and Affiliations

  • Joyce T. Au
    • 1
  • Lorena Gonzalez
    • 1
  • Chun-Hao Chen
    • 1
  • Inna Serganova
    • 2
  • Yuman Fong
    • 1
  1. 1.Department of SurgeryMemorial Sloan-Kettering Cancer CenterNew YorkUSA
  2. 2.Molecular Pharmacology and Chemistry ProgramMemorial Sloan-Kettering Cancer CenterNew YorkUSA

Personalised recommendations