NEUROLOGICAL REVIEW Imaging Technology for Neurodegenerative Diseases Progress Toward Detection of Specific Pathologies

Chester A. Mathis, PhD; William E. Klunk, MD, PhD; Julie C. Price, PhD; Steven T. DeKosky, MD

dvances in neuroimaging over the past 2 decades are products of breakthroughs in imaging technology, developments of more powerful computers and image- processing software, and expanding knowledge in basic and clinical neuroscience. In addition to the insights into normal brain structure and function that such methods Aprovide and the information that can be gained from disease-related changes in structure and func- tion, the promise of achieving diagnostic specificity through neuroimaging lies with the potential identification of pathognomonic . Recent advances in imaging ␤- plaques, one of the hallmarks of Alzheimer disease, offer such a technological breakthrough and the possibility for more efficient assessment of antiamyloid interventions as well as specific noninvasive diagnos- tic capabilities. Arch Neurol. 2005;62:196-200

In the last several decades of the 20th cen- and other pathological disturbances. Tech- tury and the first several years of the 21st, niques for the rapid assessment of blood research and the clinical application of flow advanced functional imaging and the neuroimaging have proceeded at a dizzy- localization of brain activity or function. ing and accelerating rate. Structural im- Magnetic resonance spectroscopy contin- aging was initially limited to skull and ues to develop as another method for the spine radiographs and subsequently to in- noninvasive assessment of brain metabo- ference about what was happening in the lism in development, aging, and disease. brain by analyzing the patterns of arte- The development of positron emis- rial, capillary, and venous blood flow on sion tomography (PET) has followed a cerebral angiography. The advent of the similar rapid evolution since implemen- computed tomography scan in the 1970s tation in the 1970s. These advances in- allowed visualization of brain paren- clude improved detector systems, in- chyma for the first time, and advances in creases in computer processing power and hardware, computer power, and imaging programs to use such power, and ad- programs led to improved resolution. Mag- vances in radiochemistry. These ad- netic resonance imaging added to the vances in turn led to breakthroughs in our bounty of new structural imaging tech- understanding of cerebral blood flow, en- niques in the 1980s and also progressed ergy metabolism, and neurotransmitter through increased imaging resolution and function in humans both with and with- new sequences that led to better visual- out neurological disease. Patterns of ab- ization of white matter, water diffusion, normalities in blood flow or glucose me- tabolism, although nonspecific, have Author Affiliations: Departments of Neurology, Psychiatry, and Radiology, provided significant diagnostic aid in Alz- University of Pittsburgh and Western Psychiatric Institute and Clinic, Pittsburgh, Pa. heimer disease, dementia with Lewy bod- Financial Disclosure: GE Healthcare (formerly Amersham Health) (Chalfont 1 St. Giles, United Kingdom) provided research grant support to the University of ies, and frontotemporal dementia. The Pittsburgh and entered into a license agreement with the University of Pittsburgh ability to devise imaging ligands that bind based on the technology described in this article. Drs Klunk and Mathis are to central nervous system enzymes, re- coinventors of and, as such, have a financial interest in ceptors, transporters, or other compo- this license agreement. nents of neurotransmitter systems has been

(REPRINTED) ARCH NEUROL / VOL 62, FEB 2005 WWW.ARCHNEUROL.COM 196

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021 the special province of PET. Alterations in these PET H3C HO markers over the course of time has aided diagnosis and S CH3 S H added information about the progression of disease. N N + N CH N Recently, PET research has begun to focus on the abil- 3 CH3 CH ity to image the abnormal proteins associated with spe- 3 cific neurodegenerative diseases. The ability to image the Thioflavin T PIB and neurofibrillary tangles of Alzhei- mer disease (AD), the Lewy bodies of Parkinson disease Figure 1. Derivation of Pittsburgh Compound B (PIB) from the fluorescent or dementia with Lewy bodies, or other pathognomonic histochemical dye thioflavin T. proteins would enable unprecedented accuracy of diag- nosis; it would also allow staging of disease and a basis ing affinity of the derivatives for A␤, deletion of the charge for the efficient evaluation of therapeutic interventions from thioflavin T dramatically increased the binding af- to halt the accumulation of, decrease the levels of, or re- finity of the derivatives for A␤. Further derivatization of move abnormal proteins. these neutral analogues of thioflavin T (termed benzo- For amyloid, the potential benefits of a pathology- thiazole anilines) indicated that substitution of the driven imaging approach are the following: 6-methyl group with a 6-hydroxyl group provided a po- tent A␤ ligand with excellent in vivo pharmacokinetic v Improvements in early diagnosis and the possibil- imaging properties.3 This 6-hydroxyl–substituted ben- ity of a presymptomatic diagnostic biomarker, depend- zothiazole aniline is known as Pittsburgh Compound B ing on the time course of amyloid deposition (PIB) (Figure 1). When radiolabeled with carbon 11, v Improved understanding of the natural history of PIB provides a radioligand for PET imaging studies of A␤ amyloid deposition and insights into the pathophysiol- in the brains of living human subjects.4 Pittsburgh Com- ogy of AD, along with a potential tool to further evalu- pound B retains the highly fluorescent properties of its ate the amyloid cascade hypothesis parent, thioflavin T, and has been used to image A␤ in v The capability to directly measure the effects of newly the brains of living transgenic mice using multiphoton developed antiamyloid therapies (eg, secretase inhibi- microscopy techniques.5 tors, immunotherapy, and “plaque breakers”) Critical to the demonstration of the binding targets The ability to directly measure a pathological change of were binding assays that used homog- believed to be intimately related to the pathophysiology enates of postmortem brain because these assays could of the disease might have advantages over less specific be performed using the low nanomolar concentrations structural or metabolic measures in all of these areas, but of amyloid probe that are attainable during in vivo PET the advantage is clearest in the ability to directly mea- studies.6 In brain areas known to contain high concen- sure the efficacy of antiamyloid therapies. Toward this trations of A␤ deposits (eg, the frontal and temporal cor- end, our group set out to develop PET tracers specifi- tex), tissue homogenate studies showed greater than 10- cally for the ␤-amyloid peptide (A␤) in plaques and cere- fold more binding of compounds to AD brovascular amyloid in AD and related conditions. Our brain compared with age-matched control brain. In con- goal was to apply PET technology to detect and quantify trast, no increase in binding was seen in areas of AD brain a specific pathological change in the brain. without fibrillar amyloid deposits (eg, cerebellum). Al- though neurofibrillary tangles could be stained with fluo- A␤-SPECIFIC RADIOTRACERS rescent benzothiazoles applied at a high concentration (1 µM), a 1-nM concentration of benzothiazole com- The design and biological evaluation of agents for PET pounds resulted in no increase in binding over back- and single-photon emission computed tomography to im- ground in brain tissue homogenates that contained large age A␤ plaques in the living brain have been reviewed numbers of tangles but no plaques (ie, Braak II control recently.2 The most successful radiotracers to date for these brain transentorhinal cortex). This suggested that un- purposes have proven to be relatively small molecules der in vivo PET conditions (at nM tracer levels), benzo- (Ͻ600 d), which display better in vivo pharmacokinetic thiazoles cannot detect tangle deposits and that the sig- imaging properties than large proteins such as A␤ pep- nal observed is primarily due to A␤ deposits. Finally, tides and monoclonal antibodies. Well-known A␤- multiphoton microscopy studies of PIB binding in trans- specific histological dyes, such as Congo red and thio- genic mouse models of amyloid deposition clearly dem- flavin T, have provided a starting point for the onstrated specific retention of PIB in plaque and cere- development of these agents, from which many series of brovascular amyloid deposits, confirming both brain entry analogues have been synthesized and evaluated as A␤ im- and plaque specificity in a living animal model of AD.5 aging agents for PET and single-photon emission com- puted tomography. However, the ionic charges on Congo IN VIVO PHARMACOKINETIC MODELING red and thioflavin T prevented good penetration of the blood-brain barrier by the radiotracer analogues. Neu- A primary objective of PIB PET modeling studies has been tral, lipophilic derivatives were necessary to achieve suf- to define a simple and valid method for the in vivo mea- ficient quantities of the radiotracer in the brain for sub- surement of amyloid deposition across the AD spec- sequent PET or single-photon emission computed trum, one that can be readily applied in the clinical set- tomography imaging studies. Although removal of the ting. Because postmortem binding assays have shown PIB ionic charge from Congo red analogues lowered the bind- binding characteristics to be consistent with classical li-

(REPRINTED) ARCH NEUROL / VOL 62, FEB 2005 WWW.ARCHNEUROL.COM 197

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021 gand-binding interactions,3 our analytic approaches were concentrations of the amyloid probe typically used in those well-established for the quantification of PET im- tracer PET studies. aging studies of ligand-receptor binding. Data were ac- The third attempt to study human in vivo amyloid im- quired in AD and control subjects in a fully quantitative aging used PIB and was presented in preliminary form manner that allowed flexibility for subsequent analysis by Engler et al10 followed by the full report in 2004.4 This by multiple approaches that vary in complexity, accu- first human PIB study included 16 patients with mild AD racy, and reliability. These approaches yield PIB reten- (Mini-Mental State Examination scores, 18-28) and 9 tion measures that range from fully quantitative (distri- healthy control subjects. Three of the 9 controls were bution volume) to semiquantitative (standardized uptake young (21 years old) to guarantee the absence of plaques values) and include the following: compartmental mod- in these subjects. There was a robust difference in PIB eling (fully dynamic PET, arterial input: distribution vol- retention between the AD patients and the healthy con- ume), graphical methods (simplified dynamic, with and trols. Compared with controls, AD patients showed without arterial input: distribution volume), and late- marked retention of PIB in areas of the brain association scan PET uptake measures (standardized uptake val- cortex known to contain large amounts of amyloid de- ues). Good agreement has been found across the ana- posits in AD. In AD, PIB retention increased most promi- lytic methods with respect to the regional pattern of PIB nently in the frontal cortex (Ͼ2-fold; PϽ.002). Equally retention in AD, distinct differences in PIB retention be- important, retention was equivalent in AD patients and tween AD and control subjects in areas known to con- controls in areas known to be relatively unaffected by amy- tain amyloid in AD, and similar levels of nonspecific PIB loid deposition (such as the subcortical white matter, pons, retention in AD and control subjects.7 The latter simi- and cerebellum; PϾ.2). Studies in young and old healthy larity in reference PIB retention supports our investiga- controls showed very low PIB retention in cortical areas, tion of image-based analyses that eliminate the neces- independent of subject age. Pittsburgh Compound B re- sity for arterial blood sampling. In fact, good agreement tention correlated inversely with cerebral glucose me- has been found between the fully quantitative arterial- tabolism in the parietal cortex, determined with 2-[F- based methods and simplified reference-based analyses 18]fluoro-2-deoxy-D-glucose (FDG) across all subjects. that allow rapid generation of image maps of PIB reten- The group difference in PIB retention was of greater mag- tion. The next step is to apply the regional results to vali- nitude than FDG detected hypometabolism in all brain date volumetric statistical assessments of parametric im- areas (Figure 2). ages maps of PIB retention that are based on the As we have previously suggested,4 it is important to anatomical standardization of each individual’s mag- avoid the circular reasoning inherent in the association netic resonance and PIB PET data to a common refer- of amyloid deposition with both the diagnosis and the ence. Preliminary parametric statistical assessments have etiology of AD. Therefore, at the outset, it might be best provided results that agree with the regional evalua- not to equate amyloid deposition with clinical diagnosis tions. These stepwise analyses of the PIB PET method (independent of the subject’s clinical status). Instead, it should lead to the identification of a simple and valid might be best to first think of PIB retention more fun- method that will be feasible for routine use and provide damentally as a method to detect and quantify brain “␤- a pragmatic compromise between methodological accu- ,” a phrase first used in reference to AD by racy and precision. Glenner.11 Several fundamental and unbiased questions can then be asked regarding the following: HUMAN AMYLOID IMAGING STUDIES v The cross-sectional variation in ␤-amyloidosis across the spectrum of dementia severity in AD Friedland et al8 reported the first attempt to image amy- v The natural history of ␤-amyloidosis, its onset rela- loid using a radiolabeled anti-A␤ antibody Fab frag- tive to clinical symptoms of dementia, and its progres- ment. Insufficient brain entry of the large macromol- sion in individual patients across the course of the de- ecule made this approach impractical. The second in vivo mentia human amyloid imaging study used the tracer 2-(1-{6- v The correlation of ␤-amyloidosis with clinical di- [(2-[F-18]fluoroethyl)(methyl)amino]-2-naphthyl} ethy- agnosis lidene) malononitrile (FDDNP) to quantify amyloid in v The usefulness of ␤-amyloidosis as a surrogate 9 AD patients and 7 controls.9 The time-activity data from marker of efficacy for antiamyloid therapeutics an AD patient included in that study indicated that, at late time points (90-120 minutes), the absolute reten- Perhaps the most important short-term role of amy- tion of the tracer FDDNP in neocortical areas exceeded loid imaging will be in the development of several new that in the reference region, the pons, by 10% to 15%. classes of antiamyloid therapies, which recently have en- The area of highest retention at late time points was the tered or soon will enter human trials (eg, passive immu- hippocampus/amygdala/entorhinal cortex region, an area nization, ␥-secretase and ␤-secretase inhibitors, and that exceeded the reference region by approximately 30%. ␤-sheet breakers). All of these therapies focus on de- The tracer FDDNP was reported to image both plaques creasing the amyloid load in brain. Without a surrogate and neurofibrillary tangles; however, validation for the biomarker to assess the efficacy of these therapeutic agents latter claim is based on tissue-staining studies. Rel- on their intended central nervous system amyloid tar- evance of such staining studies for activity in vivo in hu- get, one cannot properly interpret the outcome of a thera- mans must be interpreted with caution because the his- peutic trial. For example, a negative clinical outcome in tological procedures employ greater than 1000-fold higher a therapeutic trial of passive immunization (expected to

(REPRINTED) ARCH NEUROL / VOL 62, FEB 2005 WWW.ARCHNEUROL.COM 198

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021 v Optimization of the pharmacokinetic modeling Control AD analysis of the in vivo data and simplification of the im- 1.7 aging protocol v Cross-sectional studies of amyloid deposition across all stages of disease severity in AD and mild cognitive im-

PIB pairment v Studies assessing amyloid deposition in normal ag- ing and relating the asymptomatic presence of amyloid 0 to the likelihood of developing AD at a later time Logan BP v Longitudinal studies of amyloid load progression in AD and mild cognitive impairment v Assessing the contribution of amyloid deposition to dementia in Parkinson disease and dementia with Lewy bodies Studies also are being initiated in special populations MRI such as nondemented Down syndrome subjects and asymptomatic members of families with autosomal domi- nant AD gene mutations. In addition, preliminary stud- ies are being initiated in early-phase therapeutic trials of putative antiamyloid drugs. We also are actively work- ing to develop a fluorine-18–labeled derivative of PIB. This 25 derivative will allow for wider distribution and applica- tion of this amyloid imaging technology. In summary, the ability of PET to identify and quan- ␤ FDG titate -amyloid in vivo in humans has significant im- plications for presymptomatic detection, differential di- agnosis, and therapeutic interventions in AD. Certainty 0 of diagnosis is not only important clinically; it also in- MRDglc creases the certainty of diagnosis when exploring other behavioral or biological studies in the disease. Visual- Figure 2. Sample image maps depicting the binding potential of Pittsburgh ization of an abnormal in a neurodegenerative Compound B (PIB) (top row) and glucose metabolism (bottom row) measured in a 70-year-old healthy control subject (Mini-Mental State disease in vivo provides both proof of principle and pow- Examination score, 29) and a 75-year-old subject with Alzheimer disease erful motivation to develop neuroimaging agents for other (AD) (Mini-Mental State Examination score, 26). Also shown are the neurodegenerative diseases. corresponding anatomical magnetic resonance images (MRIs) (middle row). Greater retention of PIB is evident in frontal, temporoparietal, and posterior cingulate cortices, and less PIB retention is evident in the sensory-motor strip of the AD subject, relative to the control. The 2-[F-18]fluoro-2-deoxy- D-glucose (FDG) images depict the cerebral metabolic rate of deoxyglucose Accepted for Publication: October 20, 2004. (MRDglc), (lumped constant=1), an index of glucose metabolism. Logan BP Correspondence: Steven T. DeKosky, MD, Department indicates Logan binding potential. of Neurology, University of Pittsburgh, 3471 Fifth Ave, Suite 811, Pittsburgh, PA 15213 (DeKoskyST@upmc .edu). remove ␤-amyloid from brain) would be interpreted very Author Contributions: Study concept and design: Ma- differently if no effect on amyloid load was observed at a this, Klunk, and DeKosky. Acquisition of data: Mathis and given dose than if a widespread reduction in amyloid load DeKosky. Analysis and interpretation of data: Mathis, was coupled with a negative clinical outcome. Likewise, Klunk, Price, and DeKosky. Drafting of the manuscript: a positive clinical outcome could not be accepted as proof Mathis, Klunk, Price, and DeKosky. Critical revision of of the “amyloid cascade hypothesis” until a significant the manuscript for important intellectual content: Mathis, effect on amyloid deposition could be coupled with the Klunk, Price, and DeKosky. Obtained funding: Mathis, clinical effect. Klunk, and DeKosky. Administrative, technical, and ma- Amyloid imaging technology holds promise as a di- terial support: Mathis, Klunk, Price, and DeKosky. Study agnostic technique as well. The added value of diagnos- supervision: Mathis and DeKosky. tic potential will not be fully realized until an effective Funding/Support: This work was supported in part by treatment exists for AD. At that time, the ability to iden- grants IIRG-95-076 and TLL-01-3381 from the Alzhei- tify very early (and perhaps presymptomatic disease) will mer’s Association, Chicago, Ill (Dr Klunk); AG05133 from become essential for early and optimal initiation of the National Institute on Aging, Bethesda, Md (Dr De- therapy. Additionally, assessment of amyloid load changes Kosky), AG01039 and AG20226 from the National In- over time will likely be a necessary surrogate marker of stitute on Aging (Dr Klunk); AG18402 from the Na- treatment efficacy. tional Institute on Aging (Dr Mathis); MH070729 from To prepare for these upcoming applications of amy- the National Institute of Mental Health, Bethesda, Md (Dr loid imaging, we are initiating fundamental studies, in- Price); and DE-FD02-03ER63590 from the Department cluding the following: of Energy, Washington, DC (Dr Mathis).

(REPRINTED) ARCH NEUROL / VOL 62, FEB 2005 WWW.ARCHNEUROL.COM 199

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021 6. Klunk WE, Wang Y, Huang GF, et al. The binding of 2-(4Ј-methylaminophenyl) REFERENCES benzothiazole to postmortem brain homogenates is dominated by the amyloid component. J Neurosci. 2003;23:2086-2092. 7. Price JC, Klunk WE, Lopresti BJ, et al. Quantitative PIB PET imaging studies of 1. Herholz K, Heiss WD. Positron emission tomography in clinical neurology. Mol amyloid deposition in subjects with Alzheimer’s disease, mild cognitive impair- Imaging Biol. 2004;6:239-269. ment, and healthy controls. Neuroimage. 2004;22(suppl 2):T49. 2. Mathis CA, Wang Y, Klunk WE. Imaging beta-amyloid plaques and neurofibril- 8. Friedland RP, Kalaria R, Berridge M, et al. Neuroimaging of vessel amyloid in lary tangles in the aging human brain. Curr Pharm Des. 2004;10:1469-1492. Alzheimer’s disease. Ann N Y Acad Sci. 1997;826:242-247. 3. Mathis CA, Wang Y, Holt DP, Huang GF, Debnath ML, Klunk WE. Synthesis and 9. Shoghi-Jadid K, Small GW, Agdeppa ED, et al. Localization of neurofibrillary tangles evaluation of 11C-labeled 6-substituted 2-arylbenzothiazoles as amyloid imag- and beta-amyloid plaques in the brains of living patients with Alzheimer disease. ing agents. J Med Chem. 2003;46:2740-2754. Am J Geriatr Psychiatry. 2002;10:24-35. 4. Klunk WE, Engler H, Nordberg A, et al. Imaging brain amyloid in Alzheimer’s dis- 10. Engler H, Nordberg A, Blomqvist G, et al. First human study with a benzothiazole ease with Pittsburgh Compound-B. Ann Neurol. 2004;55:306-319. amyloid-imaging agent in Alzheimer’s disease and control subjects. Neurobiol 5. Bacskai B, Hickey GA, Skoch J, et al. Four-dimensional multiphoton imaging of Aging. 2002;23(suppl 1):S429. brain entry, amyloid binding, and clearance of an amyloid-beta ligand in trans- 11. Glenner GG. Alzheimer’s disease: the commonest form of amyloidosis. Arch Pathol genic mice. Proc Natl Acad Sci U S A. 2003;100:12462-12467. Lab Med. 1983;107:281-282.

Call for Papers

ARCHIVES Express

The ARCHIVES launched a new ARCHIVES Express section in the September 2000 issue. This section will enable the editors to publish highly selected papers within approxi- mately 2 months of acceptance. We will consider only the most significant research, the top 1% of accepted pa- pers, on new important insights into the pathogenesis of disease, brain function, and therapy. We encourage authors to send their most exceptional clinical or basic research, designating in the cover letter a request for ex- pedited ARCHIVES Express review. We look foward to pub- lishing your important new research in this accelerated manner. Roger N. Rosenberg, MD

(REPRINTED) ARCH NEUROL / VOL 62, FEB 2005 WWW.ARCHNEUROL.COM 200

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021