USING MASS SPECTROMETRY IMAGING TO VISUALIZE DRUG DISTRIBUTION IN PUTATIVE VIRAL RESERVOIRS We are here Angela Kashuba, PharmD Elias Rosen, PhD Eshelman School of Pharmacy School of Medicine University of North Carolina at Chapel Hill David Muddiman, PhD North Carolina State University
Pharmacokinetic-Pharmacodynamic Relationships are Important Design rational drug dose, frequency, and duration
Predicting Drug Efficacy Plasma PK often used as a surrogate for drug exposure at distal locations may not accurately reflect concentrations at multiple tissues/sites of activity Tissue distribution is heterogeneous tissue distribution is nonhomogeneous and tissue specific, with high inter-tissue and intersubject variability target site concentrations may substantially differ from plasma concentrations Tissue drug distribution is nonhomogeneous and tissue specific. Representative PET images of human subjects following the 18F-trovafloxacin are shown. Muller AAC 2004
Quantifying Tissue Concentrations: LC-MS/MS Traditional analytical (LC-MS) based methods of quantification Extraction of drug from tissue homogenate Averaged concentration results in loss of spatial information Isolation of cells from tissues Loss of spatial information Decline in intracellular exposure during processing (except NRTI dp/tp)
Quantifying Tissue Concentrations: Imaging Traditional imaging techniques Quantitative whole body autoradiography (QWBA), positron emission tomography (PET) Require radiolabels ($), doesn t distinguish between parent and metabolite Challenging to administer/evaluate multi-drug therapies New Mass Spectroscopy Imaging (MSI) approaches hold promise. Characterize spatial distribution of compounds in tissues Provide additional information on endogenous compounds and metabolites Solon, DMD 2002 Nature Rev Cancer 2010
Mass Spectroscopy Imaging Approaches MALDI MSI Primarily for large molecules Generally qualitative Requires an organic matrix overlay on tissue for ionization that can cause interference with whole classes of small molecules Performed under a vacuum, leading to dehydration and sample deformation -20 C Cryostat Snap frozen tissue IR Laser Ablation & ES Ionization Orbitrap Collection IR-MALDESI MSI Can detect small molecules Ice used as matrix for all analytes, causing no interference Performed under ambient conditions to retain native water content Translation of Mass Spectra and Abundance into Drug Map Generation of Mass Spectra Kim et al. Sci Rep 2013
Optimizing ARV Quantification: Internal Standard and Calibration Curves Drug response beneath fetal mouse cryosection Normalization of analyte abundance to IS (3TC) FTC no IS * FTC incubated cervical tissue FTC IS normalized ; less variability and greater signal intensity/ reliability Quantification of analyte response using calibration standards MSI Bokhart et al Anal Bioanal Chem. 2015
Quantitative MSI of ARVs in NHP Tissues Objective: investigate MSI utility in describing drug distribution One healthy uninfected rhesus macaque dosed to steady-state with: 30mg/kg tenofovir disoproxyl fumarate (TDF) SQ QD 50mg/kg emtricitabine (FTC) SQ QD 200mg efavirenz (EFV) PO QD 100mg raltegravir (RGV) PO QD 24h plasma PK necropsy/11 tissues isolated flash frozen 10µm cryosections used to compare LC-MS/MS results to IMS IMS discretized into 10-4 mm 3 voxels (resolving 100 µm features) LC-MS/MS Shimadzu HPLC system with AB SCIEX API 5000 Triple Quad MS with turbospray interface; IR-MALDESI source with Thermo Q-Exactive MS CEREBRUM BASAL GANGLIA AXILLARY LYMPH NODES SPLEEN MESENTERIC LYMPH NODES ILEUM COLON ILIAC LYMPH NODE RECTUM INGUINAL LYMPH NODE TESTES Thompson et al. Antimicrob Agents Chemother. 2015
Quantitative MSI of ARVs in NHP Tissues Lower limit of detection LC-MS/MS 20 pg/slice MSI 0.06-2 pg/voxel LC-MS/MS results analyzed by Analyst; MSI results visualized with custom analysis software (MSiReader) Serial sections stained for tissue morphology (H&E) and immunohistochemistry (CD3+ and CD4+) LC-MS/MS: ARVs quantified in all tissue slices MSI: EFV quantified in all tissues TFV and FTC detected but not quantifiable in brain and testes RGV detected but not quantifiable in brain, testes, and 3 of 4 lymph nodes (tissue lipids causing ion suppression?) Thompson et al. Antimicrob Agents Chemother. 2015
ARVs in NHP Tissues: LC-MS/MS Results CEREBRUM BASAL GANGLIA AXILLARY LYMPH NODES SPLEEN MESENTERIC LYMPH NODES ILEUM COLON ILIAC LYMPH NODE RECTUM INGUINAL LYMPH NODE TESTES Concentration Relative to CSF or Plasma 10000 1000 100 10 1 CNS LYMPH NODES GIT Efavirenz Raltegravir Emtricitabine Tenofovir GT 0.1 Cerebrum Basal ganglia Lymph node: Axillary Lymph node: Mesenteric Lymph node: Inguinal Lymph node: Iliac Spleen Ileum Colon Rectum Testes Thompson et al. Antimicrob Agents Chemother. 2015
ARVs in NHP Tissues by MSI: Drug-Specific Distribution Cholesterol MSI H&E CD3 EFV MSI TFV MSI COLO- RECTAL localized to mucosa, lamina propria Intra-tissue concentration gradient 10 fold 3 fold Cholesterol MSI H&E CD3 EFV MSI TFV MSI LYMPH NODE localized to capsule, few follicles medullary sinuses Intra-tissue concentration gradient 3 fold 17 fold Thompson et al. Antimicrob Agents Chemother. 2015 Rosen E, unpublished data
Combining MSI and Other Imaging Modalities: MSI and ISH Registration MSI: Cholesterol Registration of MSI and ISH based on control points ISH: Counterstain MVC vrna Downsampled to MSI resolution Rosen E, Estes J, Unpublished Data
Colocalization of vrna and ARVs in lymph nodes TFV RM 42226 EFV RM 42226 MVC RM 42827 ATZ RM 42827 TFV RM 42971 EFV RM 42971 MVC RM 41380 ATZ RM 41380 Green = drug Red = RNA Rosen E, Estes J, Unpublished Data
Cumulative ARV Coverage in Lymph Nodes TFV Medullary Sinuses PD-1 MVC ATZ Rosen E, Unpublished Data
Predictors of ARV Distribution: MVC & Pgp AXILLARY LYMPH NODES ILEUM RECTUM OVARIES TESTES Rosen E, Thompson C, Unpublished Data
Tissue MSI Summary and Future Directions IR-MALDESI sensitively measures ARVs across tissues Using IS reduces variability Concentrations can be quantified and validated against LC-MS/MS methods Combining MSI spatial distributions of ARVs with other imaging modalities (IHC, ISH) provides: Drug proximity to target cells and virus Insight into mechanisms of drug distribution/localization Future directions Improvements to spatial resolution (10µm) and sensitivity Detection of intracellular NRTI metabolites Semi-automated image registration and processing Identify contributing factors to drug distribution (eg fibrosis, drug transporters)
Additional IR MALDESI Applications: Hair Forensics Porta et al Anal. Chem., 2011
Antiretrovirals in Hair Gandhi et al Clin Infect Dis. 2011 Liu et al PLoS One. 2014
IR MALDESI ARV Analysis in Hair Rosen E et al. Anal Chem 2016
Normalizing for Melanin Reduces Interpatient Variability patient 1 patient 2 patient 3 patient 1 patient 2 patient 3 Rollins et al. 2003 Rosen E et al. Anal Chem 2016 Rosen E et al. Anal Chem 2016
Optimizing IR-MALDESI Response to ARVs in Hair Systematic Analysis of ARV response using Design of Experiments Incubated Strands Dosed Strands ARV response for both Drug Incubated and Dosed hair strands optimized by reducing # of laser shots *TFV requires >20 hair strands for LC-MS/MS analysis Undetectable by IR MALDESI in single strands Fixed-dose combination of TFV+FTC allows FTC detection by IR MALDESI Rosen E, unpublished data
Hair MSI Summary and Future Directions IR-MALDESI MSI can detect the presence of ARVs in hair strands from dosed patients ATZ, DRV, FTC, ABC, RPV, EFV, EVG, DTG, MVC Melanin normalization reduces inter-subject variability Allows for development of drug exposure benchmarks To be evaluated: Influence of hair treatment on ARV response Dose-proportionality and half-life studies for benchmarking adherence patterns HPTN 069 and HIV+ clinic patients Markers of stress (eg cortisol) as related to adherence
MSI Opportunities Potential to perform PK-PD in one tissue sample Provide insight into influences of drug distribution Improve our understanding of species differences in tissue distribution (and efficacy/toxicity) Select evidence-based dosing strategies for preclinical and FTIH studies Provide a noninvasive measure of drug adherence over time
Acknowledgements Kashuba Laboratory Eli Rosen, PhD Heather Prince, PA-C Craig Sykes, MS Nicole White, BS Corbin Thompson, PharmD Amanda Schauer, BS Kimberly Handy, MPH Ashlyn Norris, BS W.M. Keck Fourier Transform Mass Spectrometry Laboratory (NC State) David Muddiman, PhD Jeremy Barry, PhD Guillaume Robichaud, PhD Mark Bokhart, PhD Luciw Laboratory (UC Davis) Paul Luciw, PhD Lourdes Adamson, PhD Frederick National Laboratory For Cancer Research Jake Estes, PhD UNC School of Medicine Collaborators Myron Cohen, MD Elizabeth Geller, MD Nicholas Shaheen, MD Yuri Fedoriw, MD Cindy Gay, MD P30 AI 050410 U01 AI 095031 R01 AI 111891 R01 GM 087964