2 0 0 7 S N M H I G H L I G H T S L E C T U R E

Molecular Imaging: Thriving AllOver the World

The Highlights Lecture is presented at the closingsession of each SNM Annual Meeting by Henry N. Wagner,Jr., MD. This year’s lecture was presented on June 6 inWashington, DC. To mark the 30th year in which Wagner haspresented this summary of presentations, innovations, andpersonal observations at the meeting, new SNM PresidentAlexander McEwan, MD, and immediate past PresidentMartin Sandler, MD, gave the speaker a plaque and expressedtheir thanks and the thanks of the entire SNM membership.

This year I would like to begin by thanking MartinSandler; Fred Fahey, DSc, chair of the ScientificProgram Committee; Jane Day and all the SNM staff

members responsible for making this meeting a success;and Virginia Pappas for her continued leadership increating what has become a truly exciting organizationand annual meeting. I once heard an attendee remark thatcoming to the SNM meeting is like drinking out of a firehose—an excellent analogy for the challenge of taking inthe enormous amount and variety of material packed intothis meeting.

Each year I pick a theme for this highlights talk, andthis year I’ve chosen the global spotlight on molecularimaging. Molecular imaging is truly thriving all over theworld. This also happens to be the 50th anniversary of theInternational Atomic Energy Agency (IAEA), which wasformed in 1957 in response to a suggestion by PresidentDwight D. Eisenhower in his 1953 ‘‘Atoms for Peace’’speech before the United Nations. The IAEA has playeda major role in the development of nuclear medicinethroughout the world and will continue to do so, increasinglyinteracting with large societies such as the SNM as wellas organizations in developing countries. Since its founding,the IAEA has been responsible for 2,248 fellowships and954 person-years of training in areas related to nuclearmedicine. The agency has sponsored 288 training coursesattended by 4,756 trainees. The IAEA has been in theforefront of global training and, in my judgment, has beenresponsible for the continued spread of nuclear medicinethroughout the world.

Once again the SNM meeting was a truly global event,with 44 countries submitting papers or posters for pre-sentation. Of those presented, half were from outside theUnited States (Table 1). This figure has held steady formany years, fluctuating between 47% and 53%. As I doeach year, I would like to ask those from the United States

to give a round of applause inappreciation for the contributionsfrom those who attend this meetingfrom other countries. We hope youwill all keep coming back.

On April 16, 2007, I had theopportunity to visit Kuwait, where Imet with technologists and physi-cians in nuclear medicine at theMubarak Hospital in Kuwait City.PET systems today are in activeoperation in the Middle East in Egypt, Lebanon, Jordan,Saudi Arabia, and Israel. A PET scanner has been installedin Kuwait and is about to begin operation, and another willsoon be installed in the United Arab Emirates. I wasreminded of the spread of molecular imaging throughoutthe world during my visit to Kuwait, where a $120-millionInstitute for Molecular Imaging in Medicine is being builtat the Kuwait Health Sciences Center. Nuclear medicineresearch and clinical benefits truly constitute a globaleffort, and the papers that I will highlight in this talk clearlyrepresent the extent to which contributions from all over theworld are coming together to advance molecular imagingand therapy.

Much of current nuclear medicine research is of world-wide interest. Koopmans et al. from the University MedicalCentre Groningen (The Netherlands) presented results onthe use of 11C-hydroxytryptophan and 18F-fluoro-L-DOPA(18F-DOPA) PET to image islet cell tumors. Figure 1 showstheir results, indicating that 11C-hydroxytryptophan clearlyshows widespread islet cell tumors in a patient in whom18F-DOPA, CT, and octreotide imaging results were inferior.

Diabetes is a new target for molecular imaging and iscertain to remain a focus of interest for SNM members andtheir colleagues. Today there is a worldwide epidemic of type2 diabetes in both industrialized and developing countries.On my visit to Kuwait, for example, I learned that 5% of thatcountry’s population have type 2 diabetes. High percentagesare also seen in India and China. The challenge to molecularmedicine is to slow or reverse the progressive age-relateddecline in pancreas b-cell function, the major cause of pro-gression of type 2 diabetes. This decline is known to beginlong before it is reflected in blood sugar levels.

Radiolabeled tracers offer one possible way to identifypersons at risk for type 2 diabetes. Kung et al. from the

Henry N. Wagner, Jr.,MD

Newsline 15N

NE

WS

LI

NE

University of Pennsylvania (Philadelphia) reported on 18F-FP-(1)-DYBZ as a PET ligand for measuring b-cell massin the pancreas. This tracer was serendipitously found tobind to the pancreas during studies of vesicular monoaminetransporters in the mouse brain. Avid accumulation wasobserved in the pancreas, as shown with the red arrow inFigure 2. It is because of this and similar excitingapproaches to the study of type 2 diabetes that I have

selected this image as the Small Animal Image of the Yearfor 2007. I’d like to call for a round of applause for theseinvestigators.

Small Animal ImagingFigure 3 shows the growth in the number of small

animal PET and SPECT studies presented at the SNMmeeting over the past 8 years. This provides clear evidencethat we will see a continuing stream of tracers fueling thedevelopment of new techniques. This is a part of the mis-sion of the SNM in translating research from the bench tothe bedside. Animal imaging was the focus of many pre-sentations at this meeting: 126 with PET, 46 with SPECT,17 with PET/CT, 21 with SPECT/CT, and 21 with the in-creasingly important technique of optical imaging. Manu-facturers are supporting this effort with dedicated smallanimal imaging devices. Among those seen at the SNMmeeting were the Bioscan NanoSPECT, Gamma MedicaX-SPECT, GE Healthcare Explore SPECT-CZT, MolecularImaging U-SPECT, Neurophysics MollyQ, and SiemensInveon.

(Continued on page 19N)

TABLE 1Countries Represented in Presentations at SNM 2007

Country*

No. of

presentations

Percentage

of total

United States 776 50.0

Japan 146 9.4Germany 146 9.4

Korea 111 7.1

China 60 3.9

Canada 42 2.7Italy 41 2.6

France 39 2.5

Taiwan 38 2.4

The Netherlands 34 2.2Turkey 31 2.0

United Kingdom 29 1.9

Australia 27 1.7Belgium 27 1.7

Switzerland 25 1.6

India 22 1.4

Austria 19 1.2Spain 16 1.0

Israel 12 0.8

Denmark 11 0.7

Sweden 9 0.6Poland 9 0.6

Brazil 8 0.5

Czech Republic 6 0.4Greece 6 0.4

Iran 5 0.3

Saudi Arabia 5 0.3

Russian Federation 3 0.2Hungary 2 0.1

The Philippines 2 0.1

*Other countries represented in 2007 included Argentina, Chile,

Croatia, Cuba, Ireland, Jordan, Lebanon, New Zealand, Pakistan,

Romania, Slovenia, South Africa, Thailand, and Uruguay.

FIGURE 1. Images acquired in a 54-year-old patient with isletcell tumor: (A) CT; (B) octreotide scintigraphy; (C) 18F-DOPAPET; and (D) 11C-5-hydroxytryptophan PET.

FIGURE 2. Small Animal Image of the Year. Avid traceraccumulation of this promising PET ligand was observed in ratpancreas (arrow), suggesting potential applications in quanti-tative measurement of b-cell mass and research into thepathogenesis and treatment of diabetes (left). In vivo biodis-tribution study results in normal rat (right).

FIGURE 3. Numbers of presentations on small animal imagingresearch with PET and SPECT at the SNM Annual Meeting,2000–2007.

16N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

(Continued from page 16N)

I have drawn examples from a few presentations thatillustrate where and how new tracers are being generated inuniversity pharmacology, biochemistry, and nuclear med-icine departments. Forrer et al. from Erasmus MedicalCollege (Rotterdam, The Netherlands) and the ResearchCentre Julich (Germany) reported on small animal SPECT/CT for in vivo evaluation of radiotracers. Figure 4 showsthe SPECT/CT image of a tumor-bearing rat injected with40 MBq 111In-octreotide. The spatial resolution can be seento be comparable with that of ex vivo autoradiograms(where animals were injected in vivo and killed beforeimaging). Another example came from Gagnon et al. at theUniversity of California, Davis, who presented the resultsof high-throughput screening of molecular imaging agentsusing microPET. The group evaluated 43 peptides radio-labeled with 18F in animals over 11 consecutive days (4 perday). This effort yielded 4 hits; that is, 4 peptides withfavorable pharmokinetics and tumor specificity. With high-throughput microPET imaging, combinatorial laboratoriescan produce millions of compounds on insoluble beads.Peptide sequences are then identified, analyzed using invitro assays, and finally evaluated in animal studies. Thisprocess can greatly accelerate the identification of newtracers and speed their translation to human studies.

van der Have et al. from University Medical Cen-ter Utrecht (The Netherlands) presented work with theU-SPECT II, a versatile sub-half-millimeter–resolution smallanimal SPECT system suitable for gated cardiac perfusionimaging in mice (Fig. 5). When I look at images such as

this I realize how astounded I would have been when I firstentered the field of nuclear medicine to know that 50 yearshence we would be able to do gated cardiac perfusion ina mouse model. I find this quite exciting, and I hope you alldo as well. In Figure 6, acquired with the same apparatus,the right and left ventricles are clearly visible in the rat, andat the bottom is the image obtained in a mouse. On the leftit is apparent that hot sources only .4 mm apart can beresolved. This indicates the great improvements in spatialresolution that we continue to see.

I might note that these spectacular images wereacquired with SPECT. We still hear speculation that PETwill continue to prosper and that SPECT will atrophy.

FIGURE 4. SPECT/CT of tumor-bearing rat injected with 40MBq 111In-octreotide (top). In the bottom image, ex vivo auto-radiography (top row) is compared with in vivo SPECT images(bottom row). The authors concluded that their small animalSPECT/CT device was a highly accurate tool for following phys-iologic processes in the same animal over time with differenttracers.

FIGURE 5. Single frame from dynamic U-SPECT II gatedcardiac perfusion imaging in a mouse. In addition to ultra-high-resolution rat and mouse imaging capabilities, the device hasadvanced data acquisition and detector electronics that cansupport novel research.

FIGURE 6. Rat image (top right) acquired with the U-SPECT IIclearly shows the right and left ventricles. Collimator images(left) and mouse image (bottom right) indicate the increasedspatial resolution possible with this device.

Newsline 19N

NE

WS

LI

NE

However, I look at it as I do the human hands: PET is 1hand and SPECT the other. You can do a lot with 1 handand a lot with the other, but you can do much more withboth. Each of these 2 modalities has unique capabilities.PET has the important advantage of being able to study themost important elements in the body: carbon, oxygen,nitrogen, and hydrogen (with fluoride as an analog). PEThas limitations in spatial resolution. SPECT has unlimitedspatial resolution, and studies can be extended out for days,a capability that is not possible with PET because of theradiation dose. SPECT also facilitates multiple studies withdifferent nuclides in the same animal. New and innovativeapplications will continue to be identified for both PET andSPECT.

Wagenaar et al. et al. from Gamma Medica-Ideas, Inc.(Northridge, CA); the University of California, Irvine; andJohns Hopkins University (Baltimore, MD) presented amultiring small animal cadmium zinc telluride system forsimultaneous SPECT/MR imaging. Figure 7 shows an invivo multiisotope SPECT with CT mouse image acquiredwith an MR-compatible unit. The distinct separations of99mTc for bone, 123I for thyroid, and 201Tl for the heart areclearly seen in the image.

Kadrmas et al. from the University of Utah (Salt LakeCity) presented a study illustrating the feasibility ofmultiple tracer studies with PET—despite the fact that allthe tracers give rise to indistinguishable 511-keV photonpairs. What they showed is that the kinetic behavior of eachtracer obeys certain restraints. When staggered injectionsare used, it is possible to recover from the time–activitycurves signal components related to each tracer. (I note inpassing that another method of performing double-tracer

studies with PET is to use the different half-lives of 11C [20minutes] and 18F [2 hours]).

Judenhofer and a consortium of researchers from theUniversity of Tubingen (Germany); the University ofCalifornia, Davis; Siemens Preclinical Solutions (Knox-ville, TN); and Bruker BioSpin MRI (Ettingen, Germany)presented a fully integrated 7T MR/PET system for animalstudies. This system allows not only PET measurements butalso nuclear MR spectroscopy and functional MR imaging.The system yields high-resolution PET/MR images, as seenin the examples in Figure 8. Figure 9 shows in vivodynamic PET/MR imaging with 11C-methylphenidate over60 minutes in a mouse brain.

Molecular Medicine: AdvancingRadiopharmaceutical Development

A Swedish proverb cautions that ‘‘Naught is had that isnot won.’’ To advance molecular imaging, it takes morethan good science to move research from the lab into rou-tine medical practice. I’ve shown you some examples ofinstruments and screening methods for developing com-pounds with potential clinical use. The question now is howto most effectively, efficiently, and rapidly move these fromthe lab into clinical practice.

All big pharmaceutical companies now have medicalimaging divisions that work to integrate molecular imaging

FIGURE 7. Triisotope SPECT with CT image acquired witha miniaturized, stationary, multiring CZT SPECT system thatprovides 24 unique angular samples and has components thatare functional in MR fields up to 7T. In the image, orange 599mTc targeting of bone; blue 5 123I targeting of thyroid; andgreen 5 201Tl cardiac targeting.

FIGURE 8. Simultaneous PET/MR image acquisition ina mouse: PET (left); MR (middle); and the fused image (right).The fully integrated PET and 7T apparatus also allows nuclearMR spectroscopy and functional MR images.

FIGURE 9. Images from a 60-minute dynamic PET/MR seriesacquired with 11C-methylphenidate in mouse brain. Transverseand coronal views are shown for PET (left); MR (middle); and thefused image (right).

20N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

into drug design and development. Together with advancesin academic medical centers, these programs produce manynew candidates for diagnostic radiopharmaceuticals. Atthe same time, however, their parent companies are moreinterested in the development of blockbuster drugs withbillion dollar sales. Our challenge is to find new ways ofdrawing on research from industry and academia to con-tinue to develop novel tracers, none of which is likely to bea ‘‘blockbuster,’’ and work together to have these approvedand integrated into medical practice.

The nature of the health care system in the UnitedStates (and many other countries) is complex, raising dif-ficult issues involving not only science and technology butalso politics and economics. Major economic gains, forexample, could be realized by decreasing unhelpful surgery,avoiding the administration of unhelpful or even harmfulmedication, practicing diligence in preventing disease aswell as exacerbations or recurrences of disease, and speed-ing the translation of research advances to patient care.

‘‘Personalized medicine’’ describes the health careapproach in which patients are treated as individuals ratherthan basing management decisions solely on statisticalevidence from similar patients. To make personalized med-icine a practical reality, I do not believe that we shouldsearch for the causes of disease; these are, for the most part,impossible to find. What we call ‘‘cause’’ is really only thechoice of a single antecedent over another, almost nevergetting back to a single, verifiable cause. Instead, I believewe can identify ‘‘modifiable molecular manifestations’’(3M) of the patient’s problems and use 1 or more of these3Ms to decide on the course of treatment and follow up onits effectiveness.

At this and past SNM meetings we have seen numerousexamples of ways in which molecular imaging can serve asthe foundation for molecular medicine. ‘‘Molecular med-icine’’ suggests that revolution through which molecularconsiderations will pervade the practice of medicine. Al-though molecular imaging is already here and presentthroughout the world, molecular medicine is only begin-ning. To make molecular medicine truly effective in thecomplex set of variables in health care systems, we willneed to consider the economic advantages and implicationsof moving in this direction. One example is in screeningtechniques. Shibata et al. from the Nishidai Clinic DiagnosticImaging Center and the National Cancer Center Hospital(Tokyo, Japan) presented a study on cancer screening with18F-FDG PET. In initial studies, the authors examined 16,923apparently health individuals, 248 (1.4%) of whom weredetermined to have cancer. Moreover, the group concludedfrom the results of their extensive follow-up studies that peri-odic cancer screening may reduce cancer mortality rates.

Giacomuzzi et al. from the Azienda Ospedaliero-Universitaria (Undine, Italy) presented a study on 18F-FDG PET/CT detection of unexpected primary tumors inpatients with known cancer. The study included 2,344patients, and focal 18F-FDG–avid lesions suggestive of

another new primary malignant tumor were found in 130(5.5%) patients.

Questions arise in connection with these interestingstudies. What is the cost of doing these studies? What is theeconomic consequence of finding unexpected lesions? Ibelieve that if molecular imaging is to advance even morequickly in clinical medicine, we must try first to modifyexisting U.S. Food and Drug Administration (FDA) regu-lations to facilitate approval of new molecular imagingagents for clinical use and also publicize an economicmodel that can justify the high costs of molecular imagingin patient care.

Suleiman and Fejka, from the FDA Office of OncologyDrug Products (Silver Spring, MD), presented an update onRadioactive Drug Research Committees (RDRCs) at theSNM meeting. Under current FDA regulation of RDRCs,research is limited to ‘‘basic science research and is in-tended to obtain basic information regarding the metabo-lism of a radioactively labeled drug or regarding humanphysiology, pathophysiology, or biochemistry, but is notintended for immediate therapeutic, diagnostic, or similarpurposes or to determine the safety and effectiveness of thedrug in humans for such purposes.’’ In 2003 (the last yearfor which I have data), there were 284 research studiesperformed under RDRC regulations, involving 2,797 humansubjects with more than 120 different radioactive tracermolecules—all done in basic research under the presentregulations.

I suggest that these regulations should be modified. Asnoted, under 21 Code of Federal Regulations 361.1, theconditions for RDRC research require that studies must beperformed for basic science research and have no pharma-cologic effect. The RDRC responsibilities include reviewand approval of each research protocol (with InstitutionalReview Board concurrence) and submission of reports tothe FDA. The nuclear medicine community should con-stantly emphasize to the world that diagnostic radiotracerstudies and other tracer studies under development withother modalities do not have pharmacologic effects. Reg-ulation as primarily stable drugs designed and administeredto have a pharmacologic effect is simply not appropriatefor radiopharmaceuticals. Moreover, this approach is theprimary limitation to approval of these agents.

My modest proposal is that clinical studies as well asbasic science research should be permitted under RDRCguidelines; that is, we should be able to have studiesapproved that may help specific patients while the basicstudy is being performed. To facilitate approval of the verylarge numbers of new tracers with potential for clinical use,we must work with the FDA to modify the present RDRCguidelines. Under such an approach, new molecular imag-ing probes would be assessed to determine their value inhelping solve a patient’s or patients’ problem(s). The FDArole would be to create new guidelines for RDRC approvalof novel molecular probes, but the RDRCs themselveswould have a radical increase in their scope of activities.

Newsline 21N

NE

WS

LI

NE

Hospitals would cover the costs of the first clinical researchstudies before approval for payment by insurance. Themedical and economic value of each study would beassessed.

It is interesting and instructive to look at the ways inwhich other countries are managing these challenges, anda strong relationship with the IAEA is helpful in obtainingsuch data. In Japan, for example, local institutional ap-proval regulates the introduction of new radiotracers. Thenational government is involved only in decisions aboutinsurance payment. So when our Japanese colleagues tell usthat 18F-FDG has not yet been approved in their country, itmeans only that the government has not approved it forinsurance reimbursement. Yet for more than a decade, 18F-FDG has been widely used in Japan. Many research grantsor patients themselves paid for the studies before gov-ernment approval. Provincial governments inspect localmanufacturing practices, including environmental safetyinspections. Guidelines for use of the tracers in medicalpractice are provided by the Japanese Society of NuclearMedicine and the Japan Radioisotope Association. Hospi-tals ensure and validate the medical value of tracer pro-cedures in all patient studies. If this seems unusual to a U.S.audience, I should remind you of the analogous status ofsurgery here. Here, hospitals have the same sorts of re-sponsibilities for novel surgical procedures performed inpatients. It is my belief that hospitals can and shouldassume similar responsibility for radioactive tracer studies—a process that would allow performance of studies beforetracers have been officially approved under regulations forreimbursement.

Dr. Maurizio Dondi, head of the Nuclear MedicineSection of the IAEA proposed at this meeting that theIAEA work with national organizations to help developinternational guidelines for approval of clinical molecularimaging tracers, with particular emphasis on ensuring goodmanufacturing practices and environmental radiation safety.Such cooperation is important, but I am also advocating theidea of local production, control, and approval as a way ofrapidly building important research and beneficial clinicalexperience with new radiotracers.

The Economic Value of Our KnowledgeAs I have noted, the economics of health are becoming

increasingly important throughout the world. A relativelynew idea in economics is that knowledge itself can beviewed as a discrete unit of value that must be consideredalong with the classic elements of space, personnel, andmaterials. The number of studies ordered by referring phy-sicians, for example, is an indication of value, where whathas been termed ‘‘creative neglect’’ by referring physicianscan eliminate unhelpful studies.

It is my contention that the value of each patient studycan be measured in terms of hospital days. Under thisassumption, the economic value of molecular imaging canbe expressed as dV/dt 5 fK*, where V 5 value and K 5

knowledge. Correct decision making depends on knowl-edge; knowledge is the basic unit. Increasing our knowl-edge by molecular imaging is expensive, but makingcorrect decisions decreases the overall cost of caring foreach patient. The result may be an overall increase in insti-tutional costs, because increasing productivity also in-creases demand for procedures. This is a key economicpoint with direct relevance to innovation in imaging: whena new and effective technique or tracer appears, demandwill increase, which will increases expenditures but im-prove productivity.

I am proposing the following process to assess the valueof a procedure. First identify a group of patient medicalrecords, along with suitable control patients, and obtainappropriate institutional and regulatory approval for accessto these studies. The first data point is time in days fromdetection of disease to the beginning of treatment (t1).Detection of disease could be defined as the first occurrenceof a symptom or disability, and t1 would be the time fromthat occurrence up to the initiation of treatment. Data pointt2 is the time from the start of treatment to the beginning ofimprovement of the patient; t3 is the length of stay in thehospital; t4 is the time in days required for the patient toreturn to normal activities, and t5 is the time from initiationof treatment until death. It is common to assign economicvalues to patient days, and in this model all times wouldbe assigned such values. In the resulting expression, thenumerator represents the survival of the patient in days,whereas the denominator contains the values that we seekto decrease:

V5fðt5Þ

fðt1Þ1 fðt2Þ1 fðt3Þ1 fðt4Þ:

Translating Research into Clinical ApplicationsThe incidence of many diseases is too low to allow any

single hospital to conduct prospective clinical trials. Con-solidation of data from multiple institutions into a registrycan validate molecular imaging procedures, assess thera-peutic regimens, and help develop knowledge-based prac-tice guidelines. Leung et al., for example, from the LondonHealth Sciences Center, London Regional Cancer Center(Canada), and the Zentralklinik Bad Berka (Germany) re-ported on the design and implementation of the Inter-national Neuroendocrine Therapy (I-NET) Registry, anonline multicenter database of patients with neuroendocrinecancer. Scott et al. from a consortium of health centersreported on the Australian Prospective Multicenter PETData Collection Project, which is similar in intent to theU.S. National Oncologic PET Registry. As a condition forAustralian government funding of PET in 2001, each of8 sites was required to participate in a data collectionprogram. This is 1 example of a program for collecting datathat also facilitates use of procedures and tracers prior tosecuring approval for reimbursement. The data collection

22N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

and reporting process was standardized in a manneranalogous to the RadLex initiative from the RadiologicalSociety of North America. Data collected included patientdemographics, reason for the study, basis for diagnosis(ICD-10-AM), stage of disease (TNM or AJCC), manage-ment plan, and PET effect on decision making. TheAustralian registry’s database included almost 30,000 18F-FDG PET studies at the time of presentation. Indicationsfor scanning included diagnosis (13.7%), staging (34.3%),restaging (49.4%), therapeutic monitoring (2.5%), andother (0.1%). Scott’s group reported specifically on themedical impact of 18F-FDG PET in oncology, epilepsy, andcardiac patients, which they found to be significant. This isanother example where a database could also be used tocalculate the economic effects of such studies.

A number of presentations at the meeting focused onbeneficial applications of tracers that have not yet been ap-proved for clinical use. Fellows et al. from RWTH AachenUniversity (Germany) and the University of Mainz (Germany)reported on the use of 18F-fallypride PET for therapeuticmonitoring in patients with schizophrenia being treated withthe dopamine receptor antagonist ziprasidone. This tracershowed promising results in longitudinal comparison oftreatment results in medicated and unmedicated patients(Fig. 10), but the challenge now is to get 18F-fallypride studiesapproved for clinical use.

Trends and TracersOne of the advantages of having given these highlights

talks for the last 30 years is the ability to look back andidentify trends. If we look at the numbers of oral and posterpresentations on PET and SPECT at the SNM AnnualMeeting over the past 25 years (Fig. 11), we see continuousgrowth. In 2006, PET studies in the United States increasedby 20% from the previous year. The 2006 figures included1.4 million PET, 62 million CT, and 27 million MRimaging studies. The current gap between the numbers forPET and other studies illustrates the tremendous potentialfor expansion of PET applications. It is certainly clear thatPET provides information that is equivalent to or as

significant as CT and MR studies—and we should makesure that the public, political leaders, and industry areaware of the potential of both PET and SPECT. In 2006,240 PET/CT and 50 SPECT or SPECT/CT instrumentswere sold in the United States.

18F remains the dominant tracer in studies presented atthe SNM Annual Meeting (Fig. 12), with the overwhelmingmajority of these being 18F-FDG studies. Thirty-two presen-tations focused on 18F-fluorothymidine (18F-FLT). Otherpositron tracers included in presentations were 11C, 15O,68Ga, 64Cu, 82Rb, 13N, and 124I.

It was not too long ago in our field that we would seepresented and/or published studies with 10 or 5 or evenfewer patients. At this meeting, many presentations in-volved large numbers of patients and long-term follow-upperiods. For example, Baum et al. from the ZentralklinikBad Berka described the results of a 5-year follow-upof 1,150 courses of peptide receptor radionuclide therapyin 360 patients with progressive somatostatin receptor–positive neuroendocrine tumors. At a mean follow-up of24 months, 17% of patients were found to be in partialremission, 76% with stable disease, and 7% with progres-sive disease, with tumor response noted in 88%. Overallclinical benefit was seen in .90% of these patients. Again,

FIGURE 10. 18F-fallypride PET imaging in an untreated patient(left) and a patient treated with ziprasidone (right) suggest thatthis tracer may be useful in therapeutic monitoring of new atyp-ical antipsychotic medications.

FIGURE 11. Presentations on PET and SPECT at the SNMAnnual Meeting, 1983–2007 (left); these presentations as per-centages of total presentations, 1983–2007 (right).

FIGURE 12. Presentations on work with various radiotracersat the SNM Annual Meeting, 1983–2007. 18F remains the pre-dominant radioisotope.

Newsline 23N

NE

WS

LI

NE

I would be interested in learning more about the economicanalysis of the benefits of such a study.

The numbers of papers on oncology presented at theSNM meeting each year continue to grow (Fig. 13).Oncology presentations far outnumber those on neurosci-ence and cardiology (Fig. 14). But if you look at U.S.clinical nuclear medicine practice today, heart studies faroutnumber those for cancer. I predict that we are about tosee a tremendous rise in clinical applications in the neu-rosciences as well. Of the 702 oncology presentations at themeeting, 18F-FDG was by far the dominant tracer (362presentations), followed by radiolabeled FLT (32), hypoxiatracers (22), and radiolabeled choline (13), acetate (11), me-thionine (9), fluoroethyltyrosine (FET) (7), DOPA (7), rituximab(3), fluoro-ß-D-arabinofuranosyl-thymine (FMAU) (3), andfluorodeoxy-1-ß-D-arabinofuranosyl-5-iodouracil (FIAU) (3).

We need another blockbuster tracer. 18F-FDG is won-derful but not perfect, as illustrated in the presentation byMathews et al. from the University of Texas SouthwesternMedical Center (Dallas). They presented diagnostic di-lemmas in 18F-FDG PET/CT, listing a number of mech-anisms that can result in low FDG uptake in tumors,including low metabolic demand for glucose, low expres-sion of Glut 1 transporters, low expression of hexokinases,low availability of FDG for distribution in cells, highexpression of glucose-6-phosphatase, high mucin contentwithin tumors, high content of necrotic material withintumors, and high blood levels of glucose. They identifiednumerous tumor types that may show reduced 18F-FDGuptake, suggesting that our oncology studies would begreatly benefited by the introduction of new tracers.

Can we predict which will be the next Centers forMedicare & Medicaid–approved tracer study? Will it be18F-fluoride, 18F-tyrosine, other 18F-labeled amino acids, ornew single-photon agents? It has been 12 years since thelast radiolabeled tracer was approved for widespread clin-ical use and reimbursement in the United States.

Rutten et al. from CHU of Liege and the University ofLiege (Belgium) reported on an 18F-FET PET/CT study inpatients with meningiomas at the skull base. All meningi-omas were visualized with 18F-FET, with no uptake seen in

normal brain (a difficulty with 18F-FDG). PET’s ability toaccurately delineate the extent of lesions was equal to thatof MR imaging in 54% of patients and greater than that ofMR imaging in 38%.

Antigens constitute a broad area with potential for thedevelopment of new molecular radiotracers. Maresca et al.from Molecular Insight Pharmaceuticals (Cambridge, MA)and the Johns Hopkins Medical Institution reported onmolecular targeting of prostate cancer with small-moleculeinhibitors of prostate-specific membrane antigen (PMSA).Figure 15 shows selective targeting in the positive tumorbut not in the PMSA-negative tumor on the opposite side.Bencherif et al. from the University of Pittsburgh (PA)presented work on the relationship between the histopa-thology estimation of cell proliferation (Ki67 index) and18F-FLT uptake in ovarian tumors. Figure 16 shows a fused18F-FLT PET/CT image in recurrent ovarian cancer. Ki67proliferation indices correlated well with 18F-FLT PET/CTpositivity and negativity.

Think Globally, Produce LocallyFor molecular imaging to move forward steadily in

clinical care, we must be able to produce more clinicallyuseful radiotracers in hospital or regional laboratoriesthrough local production and distribution, including hos-pital laboratories. In 1991, I predicted that in 30 years everymajor hospital would have a cyclotron. I still stick to thatprediction. Tiwari et al. from the Fukui Medical University(Japan) presented an example of work that is possible withan in-house cyclotron. They reported on an automatic 15Olabeling system using hemoglobin vesicles for clinical oxy-gen metabolism studies (Fig. 17). Local production makespossible onsite synthesis of tracers that are not commer-cially available. Oommen et al. from the Christian MedicalCollege (Vellore, India) reported on 14 years of experiencewith in-house preparation of 131I-metaiodobenzylguanidine(131I-MIBG) for scintigraphy. Over the 1993–2006 period

FIGURE 13. Presentations focusing on oncology at the SNMAnnual Meeting, 1983–2007 (left); these presentations as per-centages of total presentations, 1983–2007 (right).

FIGURE 14. Oncology presentations far outnumbered thoseon neuroscience and cardiology at the SNM Annual Meeting.This graph represents only 18F-FDG studies in oncology, theneurosciences, and cardiology.

24N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

of the study, 247 batches were prepared and used in 671clinical studies.

Enclosed systems are being developed to facilitate lo-cal production. Thostenson et al. from the University ofWashington (Seattle) reported on the validation of a dispos-able, closed system for radiolabeling of monoclonal anti-bodies with 131I for high-dose radioimmunotherapy. Theylabeled antibodies with activities up to 1,600 mCi and

validated a closed system with larger volume ($2.5 mL)labeling columns.

Locascio et al. from the Dana–Farber Cancer Institute(Boston, MA) described a program of PET/CT qualityassurance in clinical trials and provided guidance on goodclinical practice. The investigators implemented qualityassurance procedures and a PET/CT database to achieveaccurate and reproducible PET standardized uptake value(SUV) measurements. They reviewed and measured param-eters in 11,000 patients and found that significant errorswould have occurred had parameters not been accuratelymeasured and entered in the database during acquisition.Many papers at the SNM meeting emphasized the impor-tance of quality control and documentation.

Zubal et al. from the Institute for NeurodegenerativeDisorders (New Haven, CT) described an automated programfor analyzing striatal uptake in 123I-b-CIT SPECT imagesin patients with Parkinson’s disease. The graphs in Figure18 showing agreement between automated and manualanalysis of images indicate the utility of such an approach.

18F-fluoride is another useful tracer that is producedlocally. It is not too much to hope that at this time next yearthis tracer will have been approved for use and reim-bursement. Lee et al. from the Yokohama City UniversityHospital (Japan) described the usefulness and limitations of

FIGURE 16. Fused 18F-FLT PET/CT image in recurrent ovar-ian cancer. Ki67 proliferation indices correlated well with 18F-FLTPET/CT positivity and negativity.

FIGURE 17. One group presented an automatic 15O labelingsystem using hemoglobin vesicles for clinical oxygen metabolismstudies. Red blood cell (left); polymerized hemoglobin (right).

FIGURE 18. Results from an automated program for analyzingstriatal uptake in 123I-b-CIT SPECT images in patients withParkinson’s disease. Comparison of 1,388 single scanscomputed by automated and manual analysis (left) andcomparison of 694 repeated scans computed by automatedand manual analysis (right) show excellent agreement betweenthe automated and conventional approaches.

FIGURE 15. Selective molecular targeting of prostate cancerwith small-molecule inhibitors of prostate-specific membraneantigen with in vivo 123I-MIP-1072 SPECT/CT. The techniquemay lead to optimization of treatment strategies for prostatecancer.

Newsline 25N

NE

WS

LI

NE

18F-sodium fluoride for PET bone imaging. Advantagesinclude shorter acquisition times than methylene diphospho-nate (MDP) SPECT tracers, good spatial resolution, moreaccurate evaluation of bone tumors than SPECT, and truesemiquantitative 3D images. Figure 19 is an example ofa whole-body PET image acquired with 18F-sodium fluo-ride and subsequently fused with an MR image. Figure 20provides another example from this group of skeletal im-aging with 18F-sodium fluoride, showing bilateral lesions.

Many types of syntheses can be automated to facilitatelocal production of clinically useful radiotracers. Wang et al.from the National Yang-Ming University Medical School(Taipei, Taiwan) and the Taipei Veterans General Hospitaldescribed a high-yield robotic system for synthesizing5-18F-fluoro-2’-deoxyuridine, which can be used for measur-ing RNA activity. The increasing use of local germanium/gallium generators was also in evidence at the meeting.Kurihara et al. from the University of Texas M.D. AndersonCancer Center (Houston) used their locally generated 68Gato label guanine for PET assessment of cell proliferativeactivity. The arrows in Figure 21 indicate that accumulation

of this 68Ga-EC-guanine tracer is lower in organs and in-flammation than in tumor.

Young Nuclear Medicine ProfessionalsThis year I had the opportunity to attend the meeting of

the SNM Young Professionals Council, which I believe isan outstanding success. It is wonderful to sit in the back ofthe room and see these young nuclear medicine profes-sionals. Werner et al. from Harvard Medical School (Boston,MA) and Eberhard-Karls University (Tubingen, Germany)reported on the effect of respiratory gating on quantitativePET imaging in lung cancer. In this study, 18 patients withand without known cancerous lesions underwent 18F-FDGPET imaging with and without respiratory gating. In theirpresentation, the group showed cine images indicatingclearly that if steps are not taken to correct for motion,quantification errors may result. Figure 22 shows the chartedresults for lesion volumes and SUVs in this study. For thisexcellent work, this study was named ‘‘Best Clinical Paperby a Young Investigator.’’ Respiratory gating enhances thedetermination of tumor gating in lung cancer and increases

FIGURE 19. Whole-body PET image acquired with 18F-sodium fluoride and subsequently fused with an MR image.(A) Original radiographic image; (B) PET image; and (C) T1-weighted MR image (top), T2-weighted MR image (middle), andPET/MR fusion image (bottom).

FIGURE 20. Volume-rendered skeletal imaging with 18F-sodium fluoride and CT. Multiple lesions can be seen in theleft hip.

FIGURE 21. Accumulation of locally generated 68Ga-labeledguanine was lower in organs and inflammation than in tumor inmouse images. This tracer shows promise in assessing cellproliferative activity.

26N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

the measured tracer avidity of the lung lesions. It may soonbecome routine practice to perform gated nuclear studies ofthe lung or in lesions near the diaphragm.

Radionuclide TherapyFigure 23 shows the changes in numbers of radionu-

clide therapy presentations at the SNM meeting since 1997.Imaging techniques for monitoring such therapy were alsoprominent at this meeting. One example was from Chen et al.from the University of California, Los Angeles (UCLA),who reported on metabolic imaging with 18F-FLT PET asa powerful predictor of overall survival in patients withmalignant gliomas treated with bevacizumab and irinote-can. Figure 24 from this study indicates that a dramaticdecrease in 18F-FLT uptake could be seen on PET in only 1week, whereas MR imaging showed little change until 3months after treatment. This is another example where ad-ditional investigation might yield valuable data on eco-nomic benefits. Figure 25 shows median survival times for18F-FLT responders and nonresponders in the study. This is

an example of data on what I previously termed T5, thetime from initiation of therapy until death.

Schiepers et al. from UCLA reported on 18F-FLT ki-netics in brain tumors during treatment. The data displayedin Figure 26 indicate a linear relationship between SUVand influx K in these studies.

Hybrid and Fused ImagingHalf a century ago ‘‘fused imaging’’ was achieved by

superimposing rectilinear scan images on radiographs.Figure 27 is an example of this technique in a patient witha sublingual mass that accumulated iodine. However, the

FIGURE 22. Respiratory gating en-hanced the determination of tumor vol-umes (left) in lung cancer and increasedthe measured tracer activity in lunglesions (right), resulting in greater quanti-tation accuracy in PET imaging.

FIGURE 23. Presentations in radionuclide therapy at the SNMAnnual Meeting, 1997–2007.

FIGURE 24. 18F-FLT PET is a powerful predictor of overallsurvival in patients with malignant gliomas treated withbevacizumab and irinotecan. MR imaging 3 months aftertreatment showed change only at 3 months (left), whereasa dramatic decrease in tracer uptake could be seen on PET aslittle as 1 week after treatment (right).

Newsline 27N

NE

WS

LI

NE

fused imaging indicated that there was no thyroid tissue inthe neck, and surgery was avoided. The principles are thesame today, but the technology continues to advance rap-idly. Table 2 indicates the numbers of papers on varioustypes of fusion imaging at this and the last 6 SNM meet-ings. The 2007 meeting featured 449 presentations onfusion: 318 on PET/CT (hybrid, 301; coregistered, 17), 43on PET/MR imaging (hybrid, 8; coregistered, 35), 80 onSPECT/CT (hybrid, 74; coregistered 6), and 8 on SPECT/MR imaging (hybrid, 3; coregistered, 6).

Instrument development in nuclear medicine also con-tinues at a rapid pace, with new cameras available every

year. Sharir et al. from Procardia (Tel Aviv, Israel), Cedars-Sinai Medical Center (Los Angeles, CA), University Collegeof London Hospital (UK), and Spectrum Dynamics (Haifa,Israel) reported on the D-SPECT system for high-speedmyocardial perfusion imaging. In Figure 28 you can seethe higher resolution results and faster acquisition timesachieved with this system compared with those from adual-detector Anger camera. Slomka et al. from Cedars-Sinai Medical Center and the Cardiovascular Medical Groupof Southern California (Los Angeles) reported on softwarefusion of 64-slice CT angiography (CTA) and myocardialperfusion SPECT. They determined that the fused imagesmodified the definition of the perfused vascular territory in6 of 20 cases. Integrated SPECT/CTA analysis detected 2left anterior descending lesions and 2 right coronary arteriallesions that CTA alone did not. Figure 29 shows the 2diagonal branches, indicating where the blockage was andassociated with the perfusion defect on the left, which wasnot visible by looking at the calcified lesions on CTA. Itwill be interesting to follow the growth and development ofthese very expensive 64-slice CTA instruments in nuclearcardiology.

Molecular Medicine and Disease ProcessesAnother important concept emphasized in presentations

at this meeting is that molecular medicine focuses not onlyon organs but also on disease processes. Involvement of theautonomic nervous system in coronary artery disease is anexample. Fricke et al. from the Institute of Radiology,Nuclear Medicine, and Molecular Imaging and the Heart

FIGURE 25. Median survival times for those defined by early18F-FLT PET as responders (solid line) and nonresponders(dotted line) in the study described in Figure 24. Median survivaltime for nonresponders was 3.4 months and for responderswas 10.8 months.

FIGURE 26. 11F-FLT kinetics in malignant brain tumors showa linear relation between influx rate K (x axis) and SUVs. Theauthors of this study concluded that 11F-FLT uptake appearsappropriate for measuring therapy response.

FIGURE 27. Two views of a 1961 rectilinear scan performedwith 131I and superimposed over a radiograph to reveal a sub-lingual thyroid. ‘‘Fused’’ imaging avoided surgery.

TABLE 2Image Fusion Presentations at Recent SNM Annual

Meetings

Modality 2001 2003 2005 2007

PET/CT 33 96 242 318PET/MR 25 21 28 43

SPECT/CT 8 19 41 80

SPECT/MR 8 11 16 8

28N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

and Diabetes Center North Rhine-Westphalia (Bad Oeynhausen,Germany) looked at myocardial sympathetic innervation indiabetic patients with asymptomatic coronary artery dis-ease. They saw, for example, that preserved sympatheticnerve activity could often be found in regions where theperfusion was decreased (Fig. 30, where the black arrowindicates decreased flow). It is possible that this is aharbinger of fatal ventricular fibrillation. Sasano et al. fromJohns Hopkins University reported on the relationship be-tween sympathetic dysinnervation and invasive electro-physiology in postinfarct ventricular tachycardia in animals.Voltage mapping colocalized the sites of ventricular tachy-cardia in pigs (Fig. 31), with findings that may also supportthe idea that detection of sympathetic nerve abnormalitiesmay be prognostic of ventricular fibrillation. These types ofstudies could conceivably move nuclear medicine intoanother domain, the prevention of disease.

Schindler et al. from UCLA showed the diagnosticvalue of global and longitudinal myocardial flow responsesto sympathetic stress (cold pressor test) to identify abnor-

mal coronary vasomotor activity. The decrease in regional(longitudinal, that is strips of decreased) myocardial bloodflow during a cold pressor test had a higher sensitivity thanchange in global myocardial blood flow in identifying cor-onary vasomotor dysfunction.

Seventeen presentations at the SNM meeting focusedon the atherosclerotic process. Rudd et al. from the MountSinai School of Medicine and Hospital (New York, NY),for example, used 18F-FDG PET to show that atheroscleroticplaque inflammation in 1 arterial territory is highly corre-lated with inflammation elsewhere. In Figure 32, the greenarrows in the upper row indicate calcifications of the as-cending aorta, and the lower image shows uptake in theright carotid artery. Again, atherosclerosis is a generalizeddisease that is not restricted to specific organs and is a veryimportant domain for nuclear medicine research. Hosokawaet al. from Kitano Hospital (Osaka, Japan) and KyotoUniversity and Hospital (Japan) targeted vulnerable plaquesby using an intravascular radiation detector that can bemoved along the vessel looking for evidence of inflam-mation in the plaque. Huyghe et al. from Antwerp Uni-versity (Belgium) introduced a gamma probe device usedin guiding skeletal biopsies. At times the accumulation of99mTc-MDP could not be seen by imaging but could beseen by the probe; that is, if the patient had tenderness ina certain area that showed no lesion on imaging, the probeshowed the lesions (Fig. 33).

The Neurosciences and MoreWill neuroscience be the next big leap forward?

Neuroscience was the focus of 269 papers presented atthe 2007 SNM meeting. Figure 34 indicates the growth innumbers of papers on neurotransmission at this meetingsince 1983. They seem to have plateaued at a high level.Brain dysfunction was also the focus of a growing numberof presentations this year.

Dr. Robert Butler, the respected leader of the Inter-national Longevity Center, now recommends publicly that18F-FDG PET be used in the care of patients with dementia.

FIGURE 29. Results with integrated software fusion of 64-slice CT angiography and SPECT. Stress (left); rest (right).

FIGURE 30. Preserved sympathetic innervation in the stunnedmyocardium in diabetic patients. Arrows indicate severelyreduced flow on stress imaging (left) and preserved 11C-hydroxyephedrine tracer retention (right). The authors con-ducted the study to determine whether coronary flow reservehas a pronounced influence on sympathetic innervation indiabetic patients.

FIGURE 28. Stress and rest images acquired with a dual-detector Anger camera (top) and a new system for high-speedmyocardial perfusion imaging (bottom) in a 66-year-old malewith left circumflex artery ischemia. Stress and rest images withthe Anger camera were acquired over 19 and 11 minutes, re-spectively; these figures were 4 and 2 minutes for the high-speed system.

Newsline 29N

NE

WS

LI

NE

Several papers presented at this meeting addressed suchapplications. Rehani et al. from Kettering Medical Center(OH) and Wright State University (Dayton, OH) reportedon 18F-FDG PET assessment in 96 patients referred forsuspected dementia. PET imaging changed the diagnosis in27% and agreed with previous diagnoses in 73%. The useof PET changed management for 27% of the patients, andanticholinergic drug therapy was initiated in 14 patients onthe basis of imaging results. Again, this is a study that callsfor controls, time criteria, and an analysis of the economicadvantages that accrued to the use of PET in assessingpatients for dementia. Price et al. from the University ofPittsburgh (PA), the University of Michigan (Ann Arbor),and Washington University (St. Louis, MO) evaluated themultiinstitutional reproducibility of 11C-PIB PET studies ofdementia. Figure 35 shows the similarity of results acrossinstitutions in PET imaging of controls and individuals withmild cognitive impairment.

Heiss et al. from the Max Planck Institute for Neuro-logical Research (Cologne, Germany), the University of

Cologne, and the University of Manchester (UK) reportedon the identification of cholinergic brainstem nuclei in PETimages coregistered with MR images. Cholinergic impair-ment was found in patients with Alzheimer’s disease (AD).This impairment was more severe in patients who weresleep deficient. The work promises to shed light on therecognized association between sleep disturbances in ADand cholinergic impairment in brainstem nuclei.

Ellis et al. from Monash University and Austin Health(Victoria, Australia) and Cognitive Drug Research (Goring-on-Thames, UK) used 2-[18F]F-A-85380 PET in a studythat found that galantamine improves cognition but doesnot affect nicotinic receptors in early AD. Galantaminetreatment in early AD significantly enhanced performance

FIGURE 32. Green arrows in 18F-FDG PET images indicatethat atherosclerotic plaque calcifications in the ascending aorta(top) correspond to uptake in the right carotid artery (bottom).

FIGURE 33. Areas of increased focal 99mTc-methylene di-phosphonate uptake on bone scintigraphy (left) often do notcorrelate with any anatomical lesions displayed on conventionalradiography but can be detected and biopsied by the use ofa gamma probe (right).

FIGURE 31. 13N-ammonia PET (left)and 11C-epinephrine PET (second fromleft) perfusion images acquired in a pig at4 and 11 weeks, respectively, after in-duction of myocardial infarction in a studythat showed that regional perfusion/innervation mismatch colocalizes with thesite of ventricular tachycardia inducibility.Cine MR and invasive electroanatomicalmapping (right) were performed within 1week of the later PET imaging.

30N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

on clinical measures of cognitive function and on languageand verbal learning tasks. However, 2-[18F]F-A-85380 PETimaging identified no significant effect of galantamine ona4b2 naChRs. Changes in naChRs alone, then, are notresponsible for cognitive improvements following galant-amine treatment. This is an example of a ‘‘negative’’ resultthat provides important knowledge.

Odano et al. from the Karolinska Institute (Stockholm,Sweden) compared 18F-flumazenil with 11C-flumazenil forcentral benzodiazepine receptor imaging. Although themean values of the binding potentials were almost identical,variance with the 18F-labeled flumazenil was less than that

with 11C. This provides important information for clinicaltrials, indicating that differences in response to therapy aremore significant if the variance is less. Figure 36 shows thatalthough the images are quite similar, 18F would be thesuperior label for monitoring response to therapy.

Lin et al. from the Chang Gung University and Me-morial Hospital (Kwei-Shan, Taiwan) reported on rodentserotonin transport imaging using multipinhole SPECTand correlated their results with those from quantitativeautoradiography. Their studies with dynamic 123I-ADAMSPECT correlated images with time changes.

Huang et al. from the Tri-Service General Hospital andNational Defense Medical Center (Taipei, Taiwan) reportedon 4-18F-ADAM PET imaging of serotonin transporters inhealthy and drug-naıve, depressed individuals. In theUnited States, it is often difficult to recruit patients whohave never received pharmaceutical treatment, but thisstudy was possible in Taiwan. They found that the braindistribution of this tracer in humans appears to be cor-related with the known distribution of serotonin transport-ers. The 18F-ADAM imaging quality is better than that of123I-ADAM. Figure 37 shows decreased distribution vol-ume ratios and binding parameters in depressed subjects.

We have seen many studies over the past few years ondopamine and serotonin transporters, and norepinephrinetransporter studies are now growing in numbers. Takanoet al. from the Karolinska Institute and Eli Lilly and Co.

FIGURE 35. 11C-PIB imaging in healthy individuals (top row)and individuals with mild cognitive impairment (bottom row).Images were acquired at Washington University (left), theUniversity of Pittsburgh (middle), and the University of Michigan(right). Consistent measures of tracer retention and test–retestvariability over the 3 sites are promising for multisite therapystudies in which different scanner systems are used.

FIGURE 36. Comparison of 18F-flumazenil (left) and 11C-flumazenil imaging (right) in central benzodiazepine assess-ment. Summation images (0–93 minutes; top) and interimimages (39–45 minutes, bottom) indicate that 18F is the superiortracer and could provide useful information in clinical studies.

FIGURE 34. Presentations on neurotransmission topics at theSNM Annual Meeting, 1983–2007.

Newsline 31N

NE

WS

LI

NE

(Indianapolis, IN) reported on imaging norepinephrinetransporters with regional pharmacokinetic analysis. Theyshowed activity curves illustrating differences in thalamusand caudate, with accompanying images (Fig. 38).

We now know quite clearly that communication in-volves all cells, not only those in the central nervous sys-tem. At this meeting 110 studies focused on receptors inoncology. Receptor topics included somatostatin (40), in-tegrins (21) bombesin (11), EGFR (8), estrogen (7), folate (4),sigma (3), and others (16).

Macke et al. from the University of Basel (Switzerland),the University of Berne (Switzerland), and the Salk Insti-tute (La Jolla, CA) reported on research designed to answerthe questions: Are radiolabeled somatostatin receptor an-tagonists superior to agonists for in vivo tumor targeting?They found that internalization of the tracer is essential forefficient and persistent targeting of G-protein–coupledreceptors with radioligands. They also found that antago-nists target more binding sites.

Molecular Imaging, Genetics, and PharmacologyMolecular imaging, genetics, and pharmacology are all

key elements in understanding the relationship betweenbrain chemistry and behavior. Dopamine, serotonin, andnorepinephrine are involved in emotions, including aggres-

sive behavior. The emotional ‘‘fight or flight’’ response wasfirst described by Walter Cannon and extended by HansSelye decades ago. At that time, they focused on adrenalepinephrine and steroids in stress reactions. An excellentstudy showing the gene–brain–behavior relationship wasreported on at the meeting by Alia-Klein et al. from theBrookhaven National Laboratory (Upton, NY), King’s Col-lege (London), and the National Institute on Drug Abuse(NIDA; Bethesda, MD). Their studies with 11C-clorgylineshowed an inverse quantitative association between aggres-sion and brain monoamine oxidase A (MAO A) activity; assubjects became more aggressive, MAO A activity de-creased. Significantly, no relationship was found betweengenotype and MAO A activity. These types of studies arebecoming increasingly relevant to psychiatrists and psy-chologists who have long used other types of quantitativetechniques, such as multidimensional personality question-naires. I selected a composite of representative images fromthe Brookhaven study as the Image of the Year (includingthe brain image shown in Figure 39), illustrative of thetypes of studies that may in the future further illuminaterelationships between genotypes and phenotypes. [Editor’snote: the complete Image of the Year was featured in theJuly issue of Newsline (J Nucl Med. 2005;48(7):15N).]

Do such studies as that from Brookhaven suggest thatnorepinephrine and testosterone are the culprits in aggres-sion? Dopamine and serotonin generally are thought to bemore ‘‘calming,’’ whereas norepinephrine is involved in thefight-or-flight response. Lakshmi et al. from the Universityof Pennsylvania and Columbia University (New York, NY)presented work with (R)-N-methy-3-(3-123I-pyridin-2yloxy)-3-phenylpropylamine as an improved SPECT ligand fornorepinephrine transporters.

Addictive smoking is associated with a decrease instriatal dopamine D2 receptor availability. Landvogt et al.from the University of Mainz reported that addictivesmoking is associated with a decrease in dopamine D2receptor availability. This suggests that large amounts ofdopamine are on the receptors to compete with the tracer(which is also the case with other substances of abuse).Figure 40 includes bilateral putamen ratio images compar-ing results first in smokers who are smoking and non-smokers and then comparing smokers after stoppingsmoking with nonsmokers.

Esterlis et al. from the Yale University School of Medi-cine (New Haven, CT) and the VA Connecticut HealthcareSystem used 123I-iomazenil SPECT to look at GABAA ben-zodiazepine receptors in smokers and nonsmokers. Theyfound that these receptors decreased in anxiety disorders,posttraumatic stress disorder, and depression. They con-cluded that smoking may help to reduce anxiety and depres-sive symptoms by stimulating GABA function. Smoking,then, affects those psychological factors that can now beassessed with radiotracer techniques.

Wang et al. from the Brookhaven National Laboratory,NIDA, the National Institute on Alcohol Abuse and Alcoholism

FIGURE 37. CT (top rows of both image sets) and 18F-ADAMPET (bottom rows) in coronal (left), sagittal (middle), and trans-axial (right) views in healthy individuals (top) and depressedindividuals (bottom) suggest that this tracer may be a useful toolin assessing the status of human brain serotonin transporters.

32N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

(Rockville, MD), and St. Luke’s Hospital (New York, NY)used 18F-FDG PET imaging to demonstrate sex differencesin brain metabolic response to cognitive inhibition duringfood stimulation (Fig. 41). When hungry men and womenwere presented with food (that they could see and smell),strikingly greater levels of neuronal activity throughout the

brain were seen in women than in men. In cognitiveinhibition studies (when hungry patients were shown foodbut told not to think about it), men showed slight decreasesin neuronal activity but women showed none at all.

Another example of studies of imaging related tobrain actions was research by Badgaiyan et al. from theMassachusetts General Hospital (Boston, MA) and HarvardMedical School on detection of task-induced extrastriataldopamine secretion using the tracer 18F-fallypride. Theystudied dopamine secretion in association with humanemotional memory processing. They presented neutral

FIGURE 38. Activity curves from initialhuman studies with (S,S)-18F-FMeNER-D2 PET show differentiation of uptake inthalamus and caudate. Preliminary im-aging suggests that this tracer has suit-able characteristics for PET studies ofthe norepinephrine reuptake system.

FIGURE 39. In a part of the image series selected as 2007SNM Image of the Year, 11C-clorgyline PET imaging (an exam-ple of which is shown here) indicated an inverse quantitativeassociation between male aggression and brain monoamineoxidase A activity.

FIGURE 40. 18F-fallypride PET ratio images of the bilateralputamen. Top row: Ratio images compared heavy smokersunder ‘‘consume’’ conditions with nonsmokers. Bottom row:Heavy smokers after overnight withdrawal compared with non-smokers.

Newsline 33N

NE

WS

LI

NE

words, such as ‘‘park,’’ ‘‘pencil,’’ or ‘‘nuclear medicine’’ andemotional words, such as ‘‘fire,’’ ‘‘gun,’’ or ‘‘radiology.’’ Sub-jects indicated their like or dislike for each word. The linesin the middle of the graphs in Figure 42 show the point atwhich the emotional word was spoken to the subject. Im-mediately after hearing ‘‘emotional’’ words, ligand activitydecreased in the amygdala, hippocampus, and prefrontalcortex, interpreted as endogenous dopamine secretion. Timecurves that can be derived for these decreases can beobtained for every voxel in the imaging system.

We must remind ourselves frequently that molecularimaging in patients with thyroid disease began more thanhalf a century ago with the molecule sodium iodide and thattechniques in the detection and treatment of thyroid dis-orders remain important in our field. Eichhorn et al. from

the Johannes Gutenberg University (Mainz, Germany) re-ported on the use of 18F-FDG PET to measure the neuronalcorrelates of overt hypothyroidism. They found that hy-pothyroid patients showed significantly decreased metabo-lism in the bilateral thalamus and posterior insula. After areturn to euthyroidism, FDG accumulation increased in theprimary visual cortex and in the sensory association cortex.

FIGURE 41. 18F-FDG PET imaging in men (top row) and women(bottom row) demonstrated sex differences in brain metabolicresponse to food stimulation (left) and cognitive inhibition (right).In each pair of images, the left represents stimulus imaging andthe right represents baseline for that challenge.

FIGURE 42. 18F-fallypride PET changes (top) in extrastriataldopamine secretion in the amygdala, medial temporal, andventral prefrontal cortex were associated with neural process-ing of neutral and emotion-stimulating words. Time–activitycurves (below) indicate that immediately after hearing emotion-stimulating words, dopamine secretion declined in all 3 areas,most precipitously in the prefrontal cortex.

FIGURE 43. 18F-fluoroethyldiprenorphine PET imaging foundpositive correlations between opioid receptor availability andtest metrics of reward dependence (left) and self-directednessin healthy individuals (right).

FIGURE 44. In vivo PET mapping of the human cerebralcannabinoid-type 1 receptor (here coregistered with 3D MRimages) showed gender-dependent increases in temporal,limbic, and motor areas with age.

34N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

Klega et al., also from the Johannes Gutenberg Uni-versity, reported on central opioid system correlates withpersonality traits in healthy subjects. They found positivecorrelations between opioid receptor availability andreward dependence and between opioid receptor availabil-ity and self-directedness (Fig. 43). Van Laere et al. fromUniversity Hospital (Leuven, Belgium) reported on in vivoPET mapping of the human cerebral cannabinoid-type 1receptor and showed gender-dependent increases withaging. Figure 44 shows increased cannabinoid receptorsin temporal, limbic, and motor areas with age, an increasethat was more pronounced in women. In a separate study,the same group reported that cannabinoid-type 1 receptor

availability is inversely correlated to novelty seeking inhealthy human volunteers (Fig. 45; novelty seekers also showeda higher incidence of substance abuse and eating disorders).These results may be related to known cannabinoid-type 1receptor inhibitory presynaptic modulation.

Data Processing, Display, and AutomationContinuing advances are being made in data processing

and display. Lu et al. from Insightful Corporation (Seattle,WA), the Allen Institute for Brain Science (Seattle), and theUniversity of Washington described FusionViewer, an opensource medical image display, for which additional in-formation is available at http://fusionviewer.sourceforge.net. The display has multiplatform support, uses the InsightToolkit, is customized for PET/CT but is extensible, andwas funded originally by the National Cancer Institute.

Another advance was a Java-based integrated environ-ment for image quantitation presented by Truong et al.from the UCLA School of Medicine. Shiraishi et al. fromthe University of Chicago (IL) presented an automatedmethod for temporal subtraction images (that is, comparing1 image with another in skeletal imaging) to assist radio-logists in assessing interval changes in successive whole-body bone scans. Their clinical trial with this methodbegan in 2006, and Figure 46 shows an example of theresults.

Li et al., also from the University of Chicago, describedan automated computerized scheme for tumor detection inPET images. What was the impetus for the development ofthis system? In some parts of Japan, as noted earlier, 18F-FDG PET is beginning to be used to screen healthy persons.Interpretation of large numbers of studies by subjectivescreen viewing is time consuming. As PET screening isused more, demands for imaging physicians to read andreport on every screening study are likely to becomeoverwhelming. Automated detection schemes are clearly 1

FIGURE 45. In vivo PET mapping of cannabinoid-type 1receptor availability in healthy human volunteers correlated withpsychological test metrics of novelty seeking. Images acquiredin ‘‘low novelty-seeking’’ individuals (top) and ‘‘high novelty-seeking’’ individuals (bottom) suggest a potential role for thecannabinoid receptor system in behavior.

FIGURE 46. An automated method forproducing temporal subtraction witha nonlinear image-warping technique isdesigned to assist radiologists in assess-ing interval changes in successivewhole-body bone scans. Anterior (leftset) and posterior (right set) imagescontain prior (left), later (middle), andautomatically subtracted images (right).

Newsline 35N

NE

WS

LI

NE

answer to this challenge. One example of the ways in whichwe will be challenged to find more efficient ways toevaluate large-scale screening efforts can be found in thepresentation by Nishizawa et al. from the HamamatsuMedical Photonics Foundation (Shizuoka, Japan). Theircancer screening trial to evaluate the efficacy of 18F-FDGPET in healthy volunteers included 1,197 individuals, inwhom 27 cancers were detected (12 by PET and 7 by othermeans; 8 were not detected by screening). We have to ask:what was the cost of all this screening and what was thevalue of detecting cancer in these 12 people? Clearly oneway to make such studies more cost efficient is to generatevalidated detection methods that aid the imaging specialistin rapidly reviewing findings.

New Techniques and InstrumentationNew techniques were numerous at the meeting: 29 pre-

sentations involved corrections for patient motion in fusedimaging, 12 were concerned with intraoperative probes, and10 involved other probes for use beyond the operating room.

Scanners continue to be improved. Townsend et al.from the University of Tennessee (Knoxville, TN) and theUniversity of Surrey (Guildford, UK) reported on a newPET/CT scanner with an extended axial field of view (FOV).The detector volume is increased by 30%, which increasesthe sensitivity by 78%. Increased sensitivity means shorterimaging times, and larger axial FOV means fewer bedpositions. These advances are being made in parallel withreconstruction techniques. Figure 47 shows a high-qualitywhole-body scan acquired in 15 minutes on the BiographTruePoint TrueV unit described by Townsend et al.

Tsui et al. from Johns Hopkins University and PhilipsMedical Systems (Milpitas, CA) reported on improvedmyocardial perfusion SPECT using a rotating multisegmentslant-hole collimator system that does not continuously goaround the subject but instead images in segments (Fig. 48).

We also saw presentations at the meeting on devicesdeveloped for specific organ imaging or specific procedures.One example is that presented by Raylman et al. from WestVirginia University (Morgantown), Thomas Jefferson NationalAccelerator Facility (Newport News, VA), the University ofWashington, and the University of Maryland School ofMedicine (Baltimore, MD), which described initial testingof a positron emission mammography (PEM)/PET breastimaging and biopsy device. PET is used for detection andPEM for verification. The unit consists of 2 sets of rotatingplanar detectors, with 2 · 2 · 15-mm LYSO detector ele-ments and a 15 · 20 FOV. The biopsy system is computercontrolled.

Yamamoto and Hatazawa described a small-FOVLaBr3(Ce) gamma camera for low-energy single-photonimaging. The device uses a pinhole collimator and provideshigh spatial resolution, as seen in the 99mTc-labeled boneimage of a rat (Fig. 49).

Heckathorne et al. from UCLA and IntraMedical Im-aging, LLC (Los Angeles, CA) described a surgical positron

FIGURE 47. High-quality whole-body scans acquired in only15 minutes on a new PET/CT scanner with an extended axialfield of view.

FIGURE 48. Standard myocardial perfusion SPECT withpinhole collimator (left), compared with improved SPECT imagingusing a rotating multisegment slant-hole collimator system (right).

36N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE

imager using novel photodetectors, again an example of animproved special-purpose instrument. Kerrou et al. from theHopital Tenon (Paris, France) and the Laboratoire IMNC–CNRS (Paris) described an intraoperative imaging probefor sentinel lymph node biopsy in breast cancer. The devicehas an FOV of 40 mm and can be used with 99mTc, 111In, or125I tracers, with a spatial resolution of 2 mm. The probecan be used both preoperatively and in the operating room.

Wendler et al. from Technische Universitat Munchen(Germany) reported on validation of navigated b-probeimaging with PET/CT-generated activity surfaces. Thisconstitutes a new approach to radioguided resection for 18F-FDG–positive tumors (Fig. 50). This is representative ofa movement that is seeing more nuclear medicine tech-niques transition into the operating room in image-guidedinterventions.

PET and Patient ManagementAnsquer et al. from University Hospital (Nantes, France),

University Hospital (Nancy, France), and University Hospital(Marseille, France) examined PET/CT as a new tool to

avoid unnecessary surgery of benign adrenal tumors. Theyfound PET imaging to be positive in 65% of lesions thatwere candidates for surgery and negative in 86% of lesionsnot considered candidates for surgery. This, again, is a studythat could provide valuable data for economic analysis.

Rousseau et al. from the Cancer Center (Nantes-SaintHerblain, France) and University Hospital (Nantes) re-ported on the clinical impact of 18F-FDG PET in recurrentovarian carcinoma in patients with elevated serum levels oftumor marker. They found that the use of PET with CTmodified treatment plans in 69% of patients in the study.PET/CT modified treatment plans in 95% of patients inwhom previous CT imaging was negative or equivocal.

18F-FDG PET plays a very important role in identifyingcancer in thyroid nodules. Agrawal et al. from the Radi-ation Medicine Center (BARC; Mumbai, India) reported onthe efficacy of 18F-FDG PET in identifying cancer inpatients in whom fine-needle aspiration biopsy of thesuspected lesion had been indeterminate. Sayed et al. fromSt. Louis University (MO) reported on a study of the addedvalue of PET/CT in patients with cancer of unknownprimary (CUP) origin. 18F-FDG PET detected primarytumors in 21 (56.7%) of the 37 studied patients with CUPand provided detection rates higher than those reported forany other imaging modality. PET/CT should be consideredas standard for imaging in all CUP cases. And, again, I hopethat next year at the SNM meeting we will see a number ofsolid economic analyses based on these excellent studies.

Time is short, and I have a number of other presen-tations I would like to mention briefly. Abouzied et al. fromthe State University of New York at Buffalo and the KingFaisal Specialist and Research Center (Riyadh, Saudi Arabia)reported on the role of 18F-FDG PET/CT in the follow-upof pediatric patients with sarcoma after neoadjuvant che-motherapy. They found that PET was effective in ruling outthe recurrence of local disease, with negative and positivepredictive values of 95% and 67%, respectively.

Piperkova et al. from the St. Vincent’s Medical Centersof New York (NY) reported on the discrepancy between CTand PET in evaluating treatment response and follow-up inpatients with soft tissue sarcomas. In 27 early follow-upstudies at 2–6 months after treatment, discordance betweenPET and CT was observed in 4 studies (15%). In 34 studiesevaluated in long-term follow-up (.6 months), this dis-cordance was seen in 4 studies (12%).

Turcotte et al. from the Centre Hospitalier Universitairede Sherbrooke (Canada) compared PET and CT quantita-tive methods to predict the response of diffuse large B-celllymphomas after a single cycle of RCHOP chemotherapy.They found that the percentage change in tumor SUVmaxcan clearly discriminate responders from nonrespondersafter the first cycle of chemotherapy. CT volumetric mea-sures could not make such distinctions, even after 4 cyclesof chemotherapy. The ability to discriminate responders fromnonresponders opens the door to potential lowered costs oftreatment.

FIGURE 49. Planar image of a 99mTc hot phantom acquiredwith small-FOV LaBr3(Ce) gamma camera low-energy single-photon imaging (top). The device uses a pinhole collimator andprovides high spatial resolution, as seen in the 99mTc-HDMPbone image of a rat (bottom).

Newsline 37N

NE

WS

LI

NE

ConclusionThe range of new techniques, tracers, and devices seen

at the SNM 2007 Annual Meeting makes it imperative thatwe find new approaches to educate the world about the waysin which the knowledge provided by molecular imaging iseconomically as well as clinically valuable. We must showthat molecular imaging can pay for itself by saving morethan it costs to care for specific patients. I stress the individ-ual patient, here, because nuclear medicine will not decreasecosts to the entire system. As the system gets better and morepeople want the health advantages that can be provided, thenthe government must find a way to meet the demand.

We also must decrease the costs of certain studies by usingless expensive systems, such as dedicated imaging devices orprobes that match the spatial resolution of the detection systemto the spatial resolution needs of the procedure. Every patientwith a heart problem does not need CTA.

Finally, we must increase our efforts worldwide toeducate patients, referring doctors, the public, and politicalleaders about the value of molecular imaging.

Henry N. Wagner, Jr., MDThe Johns Hopkins Medical Institutes

Baltimore, Maryland

FIGURE 50. Schematics (left) and sur-face model results (right) of a navigatedb-probe that generates PET/CT activitysurfaces, providing a new approach toradioguided resection of 18F-FDG–positivetumors.

Erratum

In the article ‘‘SNMTS Honors Member Contributions, Achievements’’ (J Nucl. Med. 2007;48[7]:22N), the photocaption incorrectly identified the person accepting the Outstanding Technologist Award. Aileen Carey accepted theaward on Antonella Guardiola’s behalf.

38N THE JOURNAL OF NUCLEAR MEDICINE • Vol. 48 • No. 8 • August 2007

NE

WS

LI

NE