1. Biophysical Determinants of Photodynamic Therapy and Approaches to Improve Outcome Theresa M. Busch, Ph.D. Department of Radiation Oncology University of Pennsylvania, Philadelphia, PA

2. What is Photodynamic Therapy ?  PDT is a directed, light-based method of damaging malignant or otherwise abnormal tissues. Image from Wikipedia

3. How Does it Work? Energy transfer 3O23O2 Type 2 Reaction 1O21O2 Photosensitizer hv Photosensitizer 3 Oxidation of Organic Substrates

4. How Does it Work?  Mechanisms of PDT action Direct Cell Effects Direct 1 O 2 -mediated toxicity to tumor cells Indirect Effects Vascular damage During light treatment Delayed development within several hours after light treatment Stimulation of host immune responses.  Cell death may occur by apoptosis, necrosis, and/or autophagy

5. PDT Variables  Photosensitizer Drug type Dose Drug-light interval  Light Delivery Wavelength Fluence Fluence rate

6. What is it used for? FDA-Approved Indications (Oncology)  Obstructive esophageal cancer*  Obstructive endobronchial lung cancer*  Microinvasive endobronchial lung cancer  Actinic keratosis  Barrett’s esophagus/ high grade dysplasia *for palliative intent Clinical Trials  Pleural spread of nonsmall cell lung cancer  Mesothelioma  Intraperitoneal malignant tumors  Head and Neck- pre-malignant through advanced disease  Brain tumors  Skin cancer  Prostate cancer

7. Heterogeneity in PDT Photosensitizer distribution Tissue optical properties (light distribution) Microenvironment Tumor oxygenation Vascular network

8. Heterogeneity in Photosensitizer Uptake: A Lesson From the Intraperitoneal PDT Clinical Trial Hahn SM, et al. Clin Cancer Res 12:5464-70, 2006

9. How about light distribution?

10. Light absorption and scattering affects the fluence rate seen by the tissue. Normalized fluence rate Distance (mm) Tumor surface 3 mm depth 75 mW/cm 2 630 nm

11. The tumor microenvironment is highly heterogeneous…. Busch TM, et al. Clin. Cancer Res. 10: 4630–4638, 2004

12. …. and PDT exacerbates heterogeneity in hypoxia distribution Control RIF TumorDuring PDT 5 mg/kg Photofrin 135 J/cm 2, 75 mW/cm 2 Busch TM, et al. Cancer Res. 62:, 7273-7279, 2002

13. Heterogeneity Abounds So what to do?  Modify  Modify  Monitor  Monitor

14. Approach 1: Modify Light Delivery Rationale:  Lowering PDT fluence rate reduces the rate of photochemical oxygen consumption.  Better maintenance of tumor oxygenation during illumination.  Improves long-term tumor responses  Enhanced direct cell kill  Enhanced vascular shutdown in the treatment field

15. Hypoxia Assay EF3 and EF5 are nitroimidazole- based drugs that binds to hypoxic cells as an inverse function of oxygen tension. Detection is by a fluorochrome- conjugated monoclonal antibody. Fluorescent micrographs are digitally analyzed for binding. Section, Stain for EF3/5 Fluorescence microscopy

16. Labeling of Hypoxia during PDT RIF murine tumor EF3 at 52 mg/kg Treated animals receive Photofrin-PDT at 75 or 38 mW/cm 2, 135 J/cm 2 Hoechst 33342 at 1.5 min before tumor excision Cryosectioning, immunohistochemistry, fluorescence microscopy EF3 Hoechst Hoechst (perfusion) Anti-EF3 Anti-CD31 Hoechst (tissue label) PDT

17. Fluence rate effects on PDT-created hypoxia 38 mW/cm 2 75 mW/cm 2 EF3 Binding

18. Low fluence rate reduces intratumor heterogeneity in PDT-created hypoxia

19. Causes of depth-dependent hypoxia during PDT Light distribution? Normalized fluence rate Distance (mm) Tumor surface 3 mm depth

20. Causes of depth-dependent hypoxia during PDT Photosensitizer distribution? Photofrin Uptake (ng/mg) S D

21. Causes of depth-dependent hypoxia during PDT Does not appear to be a result of photochemical oxygen consumption. How about PDT-induced vascular effects?

22. Getting at heterogeneity in vascular response during PDT Diffuse Correlation Spectroscopy Measures the temporal correlation of fluctuations in the intensity of transmitted light (785 nm) to provide information on the motion of tissue scatters, e.g. red blood cells Data used to calculate relative blood flow, i.e. flow normalized to a pre-treatment baseline Monitoring throughout PDT is facilitated by a non-contact camera probe equipped with optical filters to block the 630 nm treatment light Separation distance between unique source- detector pairs determines the depth of tissue probed. Distance (mm) sources detectors

23. Substantial intratumor heterogeneity exists in PDT-created vascular effects PDT induces an initial increase in blood flow. PDT leads to significant depth-dependent intratumor heterogeneity in blood flow response during illumination.

24. Intratumor heterogeneity in vascular effects (controls)

25. Lower fluence rate reduces intratumor heterogeneity in relative blood flow during PDT Max rbfMax time (s)Min rbfMin time (s)CV (%)% of values 0.75-1.00 75 mW/cm 2 1.72 ± 0.13325 ± 570.47 ± 0.71195 ± 172 15 ± 313 ± 2 38 mW/cm 2 1.76 ± 0.19752 ± 175*0.31 ± 0.03*1647 ± 2499 ± 1*26 ± 5*

26. Low fluence rate reduces intratumor heterogeneity in cytotoxic response.

27. Low fluence rate improves long-term tumor response % of animals with tumors <400 mm 3

28. Lowering PDT fluence rate improves therapeutic outcome (summary)  Delivering a light dose more slowly provides  Less intra-tumor heterogeneity in PDT-created hypoxia during illumination  Less intra-tumor heterogeneity in vascular responses during illumination  Greater direct cell kill of tumor cells  Better long-term treatment response

29. Heterogeneity Abounds So what to do?  Modify  Modify  Monitor  Monitor

30. Monitoring: Rationale PDT can create significant hypoxia in even vascular-adjacent tumor cells. Vascular monitoring, including oxygenation and/or blood flow, may be indicative of tumor response.

31. Monitoring: Methods Diffuse optical spectroscopy Broadband reflectance spectroscopy with a noninvasive probe Measures tissue optical properties in the range of 600-800 nm Data used to calculate concentrations of oxyhemoglobin (HbO2) and deoxyhemoglobin (Hb) Tissue hemoglobin oxygen saturation (SO 2 or StO 2 ) = [HbO 2 ]/[HbO 2 + Hb] In mouse tissues SO 2 of 50% at pO 2 of 40 mmHg Diffuse correlation spectroscopy with a non-contact probe Measures temporal fluctuations in transmitted light (785 nm) to provide information on the motion of tissue scatters, e.g. red blood cells Data used to calculate relative blood flow, i.e. flow normalized to a pre-treatment baseline Monitoring throughout PDT is facilitated by a non-contact camera probe equipped with optical filters to block the 630 nm treatment light

32. PDT induces variable changes in tumor hemoglobin oxygen saturation 0 10 20 30 40 50 SO 2 (%) Before PDT After PDT 0 h 3 h

33. Pre- or post-PDT SO 2 is not associated with tumor response SO 2 before PDT(%) Time-to-400 mm 3 (days) SO 2 after PDT (%) Time-to-400 mm 3 (days)

34. The PDT-induced change in SO 2 in individual tumors is highly predictive of response Relative-SO 2 Time-to-400 mm 3 (days) Relative SO 2 = SO 2 after PDT SO 2 before PDT Wang H-W, et al. Cancer Res. 64(20):7553-7561, 2004

35. The PDT-induced change in blood flow is highly predictive of response Slope of decrease in blood flow Time to a tumor volume of 400 mm 3 (days) Yu G, et al. Clin Cancer Res. 11:3543-52, 2005

36. Monitoring (Summary) Pre-existing tumor SO 2 of similarly-sized tumors of the same line can be highly heterogeneous. PDT-induced changes in SO 2 and blood flow can vary from tumor-to-tumor, even for the same PDT treatment conditions. Individualized measurement of PDT effect on blood flow or blood oxygenation in a given tumor is predictive of long term response in that animal. Changes associated with better maintenance of tumor oxygen (smaller PDT-induced decreases in SO 2 or blood flow) lead to better tumor response. Diffuse optical spectroscopy, can be readily applied in the clinic and thereby may provide a means for the rapid, individualized assessment of PDT outcome.

37. Conclusions Both and clinical and preclinical studies indicate that tumors can be characterized by substantial heterogeneity in the essential components of PDT. MODIFICATION (e.g. light delivery or tumor microenvironment) can be used reduce physiologic, hemodynamic, and cytotoxic heterogeneity. MONITORING offers potential to optimize treatment through individualized, real-time dosimetry based on hemodynamic responses.

38. PDT at Penn Laser Specialist/Manager Carmen Rodriguez Biostatistics Rosie Mick Mary Putt Radiation Oncology Eli Glatstein Stephen Hahn Robert Lustig James Metz Harry Quon Neha Vapiwala Keith Cengel Veterinary Medicine Lilly Duda Jolaine Wilson Surgery Douglas Fraker Joseph Friedberg Scott Cowan Bert O’Malley S. Bruce Malkowicz Ara Chalian Nursing Coordinators Debbie Smith Susan Prendergast Melissa Culligan Medicine Dan Sterman Colin Gilespie Andrew Haas Gregory Ginsberg Physicists Timothy Zhu Jarod Finlay Andreea DiMofte Pre-clinical Researchers Theresa Busch Sydney Evans Cameron Koch Stephen Tuttle Keith Cengel Arjun Yodh Xioaman Xing Dermatology Steve Fakharzadeh

39. Acknowledgements Radiation Oncology Steve Hahn Eli Glatstein Keith Cengel Cameron Koch Sydney Evans Statistics/Image Analysis E. Paul Wileyto Mary Putt Kevin Jenkins Physics and Astronomy Arjun Yodh Xiaoman Xing Guoqiang Yu Hsing-Wen Wang Medical Physics Timothy Zhu Jarod Finlay Ken Wang Carmen Rodriguez Andreea Dimofte Busch lab Elizabeth Rickter Shirron Carter Min Yuan Amanda Maas Grant Support (NIH) R01 CA 85831 P01 CA 87971