31 December 2012. Feeling wistful that the 12 days of Christmas will soon be over? Here’s hoping that this year’s 12 trends of Alzheimer’s may have staying power—enough to put the search for therapies on a stronger footing. As your Alzforum editors reviewed the major highlights of the year, we tried to note lessons learned. We close 2012 with a sense of humility for the size of the challenge, but also with hope that the past year’s work on improving drugs and trial designs, and a broadening genetic and research base of this disease, are setting the stage for more success in 2013.
1. Clinical Trials: Drugs and Design
After years of failure, scientists and patients were hoping that clinical trial fortunes would turn on amyloid-β immunotherapies. Alas, whatever positive signs researchers divined in the year’s Phase 3 data were small. In summer, Pfizer and Janssen Alzheimer Immunotherapy conceded that their anti-Aβ antibody bapineuzumab missed both endpoints in people with mild to moderate AD, and halted development of the infused formulation of the drug (see ARF related news story; ARF news story). Bapineuzumab did seem to have a biological effect, reducing phospho-tau in the cerebrospinal fluid (CSF) and preventing further accumulation of Aβ in the brain (see ARF related news story). Some target engagement in the absence of clinical change led many to claim that the therapy was too little too late, supporting the widely held view that treatments aimed at Aβ must begin earlier in the course of disease (see ARF related news story). Many think this particular antibody failed because its side effects prevented it from being dosed high enough.
On the heels of this setback, Eli Lilly announced in August that its Aβ antibody solanezumab also missed primary endpoints in two trials of mild to moderate AD (see ARF related news story). The silver lining here was brighter. Pooled analyses done independently by the Alzheimer's Disease Cooperative Study, as well as by Lilly itself, revealed very small cognitive and functional benefits in people with mild AD, echoing the idea that treatment must begin early (see ARF related news story). This antibody will be tested in a confirmatory Phase 3 trial in mild AD in 2013 and beyond.
Solanezumab did nothing to CSF p-tau, widely seen as a marker of neurodegeneration, raising speculation that it may benefit cognition directly through some action on synapses. Solanezumab and bapineuzumab yielded surprising biomarker signals that led researchers to doubt the theragnostic value of the current biomarker crop, notwithstanding their unquestioned diagnostic and prognostic value (see ARF related news story). They also led to calls on drug sponsors to run larger biomarker substudies as part of their clinical trials, so the field can truly understand how the markers relate to clinical change.
Other therapies left the scene deflated. In Phase 2, Bristol-Myers Squibb’s γ-secretase inhibitor (GSI) avagacestat failed to protect at safe doses, and at higher doses worsened cognition in people with mild to moderate AD (see ARF related news story; ARF news story). Echoing the fate of Lilly's defunct GSI semagacestat (see ARF related news story), BMS’ 2012 results were seen as the death knell for the current suite of γ-secretase inhibitors. Hope for blocking that enzyme, for the time being, has shifted to modulators that allow the secretase to process non-APP substrates. Still, the avagacestat program broke new ground, as it taught researchers how to conduct trials in patients with prodromal AD—in other words, people with MCI-stage cognitive deficits who are ascertained by a biomarker to be on the way to AD dementia.
GSIs are not the only therapies that can make things worse. Showing how a four-month clinical trial of antioxidants came up short, results published in March surprisingly had cognition declining faster in patients who took a cocktail of vitamins E, C, and α-lipoic acid (see ARF related news story). The trial set out to test if antioxidants reduced oxidative stress by looking at biomarkers, something that no prior antioxidant trial had done. The vitamin cocktail failed to budge CSF Aβ or tau, only reducing isoprostane, an oxidation marker, by a fifth. Coenzyme Q, administered separately to a subgroup of patients, moved no biomarker.
The landscape of negative trials features some notable exceptions as well. For example, in a large Phase 2b trial, EnVivo Pharmaceutical’s α7 nicotinic agonist met primary endpoints and most secondary endpoints, as specified prior to the trial. Patients with mild to moderate AD improved over baseline by taking a capsule of EVP-6124 once a day; no Phase 3 trial has been announced yet (see ARF related news story).
All told, the race to new treatments evokes a game of pharmaceutical Chutes and Ladders. Some 2012 forerunners have tumbled, while others further back in the developmental pipeline are now poised to climb into pole position. Solanezumab’s confirmatory trial has sent it down the chute for several more years (see ARF related news story), making room at the top of the ladder for intravenous immunoglobulin being tested in people with mild to moderate AD, and Hoffmann-La Roche's Aβ antibody gantenerumab, which is in Phase 3 for people with prodromal AD. In December, Merck launched a first trial of its BACE inhibitor, MK-8931, as step one of its Phase 3 program for this drug (see ARF Q&A with Johan Luthman; ARF related news story). Other Phase 2 BACE inhibitors are waiting in the wings (see ARF related news story). So are immunotherapies, including some that innovate with Bayesian adaptive trial designs (see ARF CTAD story). Rember®, a dye derivative said to interrupt tau aggregates, will be tested in a Phase 3 trial for AD and frontotemporal dementia (see ARF related news story).
2. Prevention Initiatives
This year saw major advances in the field’s preparation for AD secondary prevention studies—a large, ambitious task. In May, the NIH announced it would give $16 million to the Alzheimer's Prevention Initiative (API) for its $100 million prevention trial of crenezumab, a therapeutic antibody made by Genentech (see ARF related news story. In 2013, the API will start the first-ever clinical trial on outwardly healthy people who are at high genetic risk of Alzheimer’s dementia. It will recruit 300 volunteers, mostly from a large Colombian kindred who harbor a presenilin-1 mutation, but also from families with familial AD mutations in the U.S.
For its part, the Dominantly Inherited Alzheimer Network (DIAN) announced in October that it had chosen gantenerumab, solanezumab, and a BACE inhibitor yet to be determined for its prevention trial (see ARF related news story). The news came shortly after analysis of Phase 3 results suggested solanezumab may have slowed cognitive decline in mild AD (see above). DIAN will recruit 160 people in their thirties and older who carry an FAD mutation.
Leaders of the Anti-Amyloid Treatment in Asymptomatic AD Trial, or A4, expect to announce funding, drug choice, and start date for their trial in early 2013. This trial aims to delay cognitive decline in people who are at risk for AD owing to amyloid biomarker positivity, albeit without a genetic predisposition.
In July, API, DIAN, and A4 announced the formation of an umbrella group called the Collaboration for Alzheimer’s Prevention (CAP). CAP serves as a forum for leaders of the three prevention initiatives to harmonize their respective methods and strategies (see ARF related news story). Cooperation could benefit all three initiatives, as each plans to collect biomarker data, use adaptive design elements, and learn about outcome measures along the way (see ARF related news story). DIAN has expanded its sites beyond the U.S., Australia, and the U.K. to Spain and Germany, and is looking to expand further (see ARF related news story; ARF news story). As longitudinal data on participants continues to grow for DIAN and API (ARF related news story; ARF news story), both initiatives confirm that biomarker changes precede symptoms by up to 30 years (see ARF related news story). Both programs have launched online registries to attract people interested in participating in prevention research (Alzheimer’s Prevention Registry EndAlznow.org; DIAN Expanded Registry).
API and DIAN have inspired similar initiatives in other diseases as well. This year saw the birth of GenFi, a biomarker-cum-trials platform in frontotemporal dementia (see ARF related news story, and the Down's Syndrome Biomarker Initiative was announced just this month (see ARF related news story).
Regulatory agencies on both sides of the Atlantic have recognized the value of attempting treatment in people who are presymptomatic. They have endorsed this endeavor in principle. The rules to be applied to such trials are evolving, not least because monitoring those trials will depend in large part on biomarker outcomes. The goal of presymptomatic treatment studies has sparked intense interest in the interface between AD research and regulatory science (see ARF related news story).
Biomarkers continued to be a major research focus in 2012. Though the problems of standardization and validation of tests still need to be overcome, the Global Consortium for the Standardization of CSF Biomarkers and the Coalition Against Major Diseases are making inroads (see ARF related news story). For example, the matrix effect boggles researchers trying to develop Aβ diagnostics. This is where concentration-dependent binding of Aβ to other proteins in CSF confounds the measurement, such that diluting a sample twofold does not halve the concentration of Aβ, as it is in equilibrium with matrix proteins. Researchers worked on circumventing this and related problems, developing fully automated in-vitro assays and a mass spectrometry approach to create a worldwide reference method (see ARF related news story).
Against that backdrop of applied science, evidence grew strong that biomarker changes in presymptomatic individuals herald impending dementia. In particular, data suggest that CSF Aβ reaches a nadir up to 10 years before symptoms emerge, while CSF tau begins to creep up later as people near disease onset (see ARF related news story). Evaluating the diagnostic value of biomarkers was helped by the DIAN and API initiatives, both of which document biomarker changes long in advance of overt disease (see above). Both the Australian Imaging and Lifestyle (AIBL) Flagship Study of Ageing and the Alzheimer’s Disease Neuroimaging Initiative (ADNI) are generating data to help predict who among cohorts diagnosed with MCI, and even cognitively normal people, will go on to AD dementia, and when.
Amyloid imaging, FDG-PET, and functional and structural MRI continue to be studied intensely. All show changes years prior to symptoms. The trajectory of each marker and how each changes relative to the others are being studied as research groups around the world pressure-test a proposed staging diagram (see ARF Webinar) with data from their longitudinal cohorts; watch for that story to break in 2013. Other markers are less developed but may hold promise. Researchers are calling for measures of synaptic activity; here, electroencephalography may represent the best available (see ARF related news story).
4. Amyloid Imaging
In April, FDA approval of the amyloid imaging ligand Amyvid® for clinical use in cognitively impaired people made a big splash (see ARF related news story). Within months, Lilly and Avid Radiopharmaceuticals, Inc., announced that the ligand would be available at 16 centers within the U.S., and that training for those who plan to interpret scans—a stipulation of the agency’s approval—would begin in June (see ARF related news story). The regulatory nod galvanized scientific efforts to determine how best to use the ligand. The Society of Nuclear Medicine and Molecular Imaging and the Alzheimer's Association created a 14-member Amyloid Imaging Task Force to develop guidelines for how the ligand should be used in clinical practice and in research. The task force has not yet issued guidelines, but appears poised to recommend that, for the time being, scans not be carried out on people who are cognitively normal (see ARF related news story). That reflects what some see as shaky ethical support for providing a diagnostic test for a disease for which there is no cure (see ARF related news story and ARF news story).
The approval also launched a debate over who would pay for the scans, which cost several thousand dollars each. Industry began lobbying the Centers for Medicare and Medicaid Services (CMS) to lift its blanket coverage exclusion on all new PET ligands (see ARF related news story). The CMS is expected to rule on that in July 2013; private insurers often follow its lead in deciding on reimbursement. In the meantime, the Institute for Clinical and Economic Review, which independently assesses the value and clinical benefit of healthcare interventions, concluded in a white paper that more evidence is needed to show that patients benefit from an amyloid scan. Most insurers require improved health outcome data before reimbursing for a procedure (see ARF related news story). In January 2013, the institute will present its finding to a CMS committee.
As stakeholder groups debate the true value of amyloid imaging, data continued to grow that positive amyloid scans do signal cognitive decline further down the road (see ARF related news story). This is true not just for Amyvid (florbetapir), but for several other ligands that are potentially nearing approval, including florbetaben and flutemetamol, all of which seem to accurately label plaques as judged by recent autopsy trials (see ARF related news story). A big science story on PET imaging will break in early 2013 when the first tau ligands generate human data. These include an agent developed by Tohoku University, Japan, and one by Siemens in Culver City, California.
5. Welcome Back, Genetics
After the early 1990s, when scientists found autosomal-dominant AD mutations and the major risk factor ApoE4, genetics fell into a decade-long slump until large genomewide association studies (GWAS) started to turn up reproducible risk variants. The resurrection of genetics continued in 2012, which saw the discovery of a protective, Aβ-lowering mutation in the APP gene (see ARF related news story) and rare but strong risk factors in variants of the TREM2 gene (see ARF related news story). One TREM2 mutation, which substitutes a histidine for an arginine in the protein, confers as much risk as ApoE4. The APP and TREM2 discoveries are fruits of deep sequencing methods that researchers believe will turn up other risk alleles too rare to be caught by GWAS. TREM2, in particular, is an example of how exome sequencing can bridge a gap in human genetics. On one side of the gap are extremely rare, fully penetrant Mendelian mutations, while on the other are common but low odds-ratio variants; both sets have been found. But a large chunk in the middle—the high-risk, somewhat rare variants—remained inaccessible to geneticists. Advances in methodology have changed that. In 2013 geneticists expect to discover more causative variants that, in the years to come, will explain a larger proportion of the population-attributable genetic risk of AD.
This year was a rich one for genetics in other neurodegenerative diseases as well. An expansion in the C9ORF72 gene that causes frontotemporal dementia and/or amyotrophic lateral sclerosis (ALS) dominated that field (see ARF related news story), while a new C9ORF72 mutation unrelated to the expansion complicated interpretation of the gene’s role in disease (see ARF related news story). ALS mutations turned up in the profilin gene (see ARF related news story), while the case for TMEM106 variants predisposing to FTD/ALS grew stronger (see ARF related news story).
In 2012, a curious genetic phenomenon came to the fore. Some people who carry one copy of a given pathogenic variant develop a different, later-onset disease than people who are homozygous for that variant, belying the conventional wisdom that recessive carriage of a disease gene is harmless. For example, TREM2 was reported last October as a risk factor for FTD (see ARF related news story), but homozygous variants in the gene cause Nasu-Hakola disease, a rare form of dementia coupled with bone disease. Homozygous progranulin mutations cause a rare lysosomal storage disease (see Smith et al., 2012), while heterozygous mutations predispose to FTD. Mutations in the gene glucocerebrosidase show the same pattern for Gaucher’s disease and parkinsonism/dementia with Lewy bodies. Are these just the tip of the iceberg?
The genetics successes of 2012 were enabled by the Human Genome Project. Other large genetic initiatives also began to bear fruit. The ENCODE project that catalogs regulatory DNA elements released a flood of data in September (see ARF related news story).This helps genetics interpret GWAS (see ARF related news story), which often turn up variants in non-coding, regulatory elements. Large initiatives to sequence whole genomes of families with AD also received support in 2012 (see ARF related news story).
This year brought several developments on this front. Interest in the role of inflammatory processes in AD had already been growing with the realization that some of the agreed-upon GWAS hits fall into inflammatory and innate immunity pathways, and the TREM2 gene only reinforced this. TREM2 encodes a cell-surface receptor found on immune cells, including macrophages and microglia. Researchers do not know its precise function, but it appears to regulate inflammation and phagocytosis, including of Aβ. Beyond that, scientists found that silencing interleukin (IL) 12 and/or IL-23—cytokines that drive inflammatory glial reactions—prevents amyloid accumulation and improves cognition in mouse models of AD (see ARF related news story and ARF Webinar on neuroinflammation in AD). Ustekinumab, an antibody used to suppress the interleukins, was approved as a treatment for psoriasis, an inflammatory skin condition, meaning it has passed safety hurdles and could be accelerated into a Phase 2 clinical trial for AD. Similarly, putting a damper on the inflammasome, a glial complex that drives cytokine signaling, also protected AD model mice from Aβ pathology and improved their cognition (see ARF related news story).
Despite these advances, in 2012 the field continued to struggle with finding ways to clearly distinguish proinflammatory, potentially detrimental microglia from those that are phagocytic and more likely beneficial to a brain clogged by Aβ. Imaging ligands would help study such cell populations in the human brain, but few are yet at hand. The PET tracer L-deprenyl, which seems to preferentially bind mitochondria in astrocytes, emerged as a contender for an astrocytic imaging ligand, and a deuterated form revealed stronger binding in the brains of people with mild cognitive impairment than in those of controls (see ARF related news story). The work supports the idea that inflammation begins early in the disease process, and may be part of the pathology rather than a reaction to it. PK11195, a reported imaging ligand for microglia, appears to lack the necessary specificity. Scientists are watching newer candidates, which bind to the same microglia protein, a mitochondrial membrane protein dubbed TSPO (see ARF related news story).
The riddle around the king of genetic risk factors—ApoE4—is why scientists still don’t fully know how it causes AD, even 20 years after its discovery. This year saw renewed efforts to solve this. For example, scientists are unsure whether therapeutically increasing or decreasing ApoE would be better for the brain, because the protein can be both beneficial and detrimental, depending on which outcome scientists measure. In February, the approved cancer drug bexarotene made waves when it temporarily cleared amyloid plaques and dramatically improved cognition in mouse models of AD (see ARF related news story). This seems related to a boost in ApoE levels driven by the retinoid X receptor, which binds bexarotene. Yet other research suggests that lowering ApoE levels improves clearance of Aβ (see ARF related news story). How to explain? The lipidation status of ApoE might be key. Highly lipidated forms of the protein may be the most beneficial, helping macrophages and microglia to clear Aβ. Researchers want to test bexarotene in ApoE4-negative patients, though hopes are tempered by the knowledge that, in previous instances where approved drugs worked in AD mouse models, their brain penetrance and pharmacokinetic and -dynamic properties ultimately rendered them unsuitable for chronic administration in people.
Evidence grew that ApoE affects AD pathology in myriad ways. At the molecular level, the E4 isoform promotes oligomerization of Aβ (see ARF related news story), while at a tissue level it appears to disrupt the blood-brain barrier and blood flow to the brain (see ARF related news story), perhaps independently of Aβ. At the organ level, brain imaging analysis suggests that ApoE4 is a detriment in and of itself. ApoE4 carriers who have amyloid plaques have a hypometabolic brain as judged by FDG-PET, but the metabolic loss appears to correlate not so much with Aβ but with ApoE4 (see ARF related news story).
8. Brain Networks and Propagation of Disease
Research solidified the concept that pathology in the brain spreads by templated protein misfolding among interconnected neural networks. Scientists showed for the first time that synthetic forms of Aβ can act as seeds for aggregation in the mouse brain (see ARF related news story). This settled the lingering question of whether previous instances of Aβ seeding could have been due to something other than the peptide carried across in brain extracts.
Scientists reported that α-synuclein spreads in a similar fashion, not only in transgenic mouse models of PD (see ARF related news story), but also in wild-type animals (see ARF related news story). In support of this, two separate groups reported that tau, the principal component of intracellular neurofibrillary tangles, spreads through the brain along connected pathways linking the entorhinal cortex and the hippocampus (see ARF related news story and de Calignon et al., 2012), as had been previously postulated for Aβ.
The concept is holding up at a higher level of investigation, too. Functional imaging data suggested that network communication breaks down as disease progresses (see ARF related news story). Multimodal brain imaging is trending in the same direction (e.g., Förster et al., 2012). Even mathematical models independently created by two groups strongly favor a scenario whereby disease spreads through interconnected networks (see ARF Webinar).
Mechanistically speaking, researchers are puzzled how intracellular proteins such as tau and α-synuclein can spread from cell to cell. Alas, these proteins may not be entirely inside the neuron. Interstitial fluid protein measurement has spotted considerable amounts of tau (Yamada et al., 2011; Kfoury et al., 2012). Scientists from a biotech company used induced pluripotent stem cells derived from people with familial AD mutations to discover a secreted, extracellular form of tau that worms its way into neurons, rendering them hyperexcitable (ARF related news story). Research on this topic bears watching in the coming year.
9. Induced Pluripotent Stem Cell Models
Despite considerable advance fanfare, induced pluripotent stem cells (iPSCs) had a slow start in the field of neurodegeneration. Suddenly in 2012, however, they became a hot area of research. Having now been derived from patients with many kinds of neurodegenerative disease by different groups of scientists, iPSCs are poised to yield new insights on the cellular, even network, level of pathogenesis. On the heels of iPSCs from people with presenilin mutations, in January 2012 researchers reported they had made stem cells using tissue donated from people with APP mutations and from patients with sporadic AD (see ARF related news story). Neurons derived from these cells secrete large amounts of Aβ and produce hyperphosphorylated tau, as do iPSCs from Down’s syndrome patients when differentiated into neurons (see ARF related news story). Going beyond simple cell culture, these neurons fire synchronously, beginning to model networks. While these models are still in their infancy, researchers have begun studies of known drugs and screens for new drugs in them. Large initiatives such as DIAN are making such lines as well.
10. Traumatic Brain Injury and Chronic Traumatic Encephalopathy
In 2012, a consensus coalesced around a definition of chronic traumatic encephalopathy (CTE) as a progressive tauopathy initiated by one or more traumatic injuries to the head. Researchers now believe that CTE stands distinct from stable brain damage seen after traumatic brain injury (TBI), although TBI may precipitate CTE in some people (see ARF conference story and Blennow et al., 2012).
Interest in CTE took off when autopsies of high-profile football players who had committed suicide revealed extensive pathology in the brain. Now, researchers study the natural history and biomarkers associated with the phenomenon in boxing, soccer, and other sports where blows to the head are par for the course (see ARF related news story). Military veterans are also affected. While CTE pathology seems distinct from AD pathology, epidemiological analysis conducted by curators at the AlzRisk database indicated that TBI broadly increases the risk for dementia (see ARF related news story). A report last September suggested that having played in the American National Football League increases fourfold one’s risk of dementia (see ARF related news story). Research on CTE has been far outpaced by media attention, with questions now being asked about head trauma in children who play contact sports (see ARF related news story).
11. Amyloid Precursor Protein and Its Processing
Our yearly news roundup would hardly be complete without the latest on APP and its processing. The amyloid hypothesis may have its detractors, but it got a boost last July when a protective mutation in the APP gene was discovered, firmly tying APP to sporadic as well as familial AD (see above and ARF related news story).
Scientists forged a link between soluble extracellular Aβ oligomers and neuronal hyperactivity. Using two-photon microscopy to peer into the cortices of APP/PS double transgenic mice, researchers found aberrant calcium signaling long in advance of plaque deposition. They reproduced those calcium spikes by bathing the cortex in synthetic Aβ dimers (see ARF related news story). The finding supports the idea that Aβ causes neural dysfunction long before plaques form in the brain.
That vision was buoyed by a report that APP/PS1 transgenic mice cannot compensate when one eye is occluded during development. In wild-type mice, synaptic activity from the uncovered eye typically ramps up and the optic nerve co-opts a bigger swath of the visual cortex. That did not happen in the transgenic mice, indicating very early deficits in synaptic plasticity when there is excess Aβ (see ARF related news story).
The importance of intracellular Aβ has always been controversial, and it was back in the news with new mouse models based solely on production of either Aβ42 or 40 within neurons. The former developed not amyloid plaques but intracellular Aβ aggregates. Their brains atrophied and the animals lost motor control (see ARF related news story). They had substantial neuronal loss, which occurs in AD but has been difficult to recapitulate in mouse models expressing mutant human APP or presenilin. But the field has recognized that APP processing is complex, with a slew of smaller Aβ species ranging in size from 15 to 38 amino acids complicating the picture (see ARF related news story and Nutu et al., 2012).
Aβ was not the only APP fragment in the news. New evidence that BACE cleavage products—soluble APPβ, or β C-terminal fragments (CTFs)—are toxic in a mouse model of familial Danish dementia contrasted with the finding that CTFs might bind a G protein to stimulate neurite growth (see ARF related news story). Along the same vein, β-arrestin 2 turned up as a γ-secretase activator that interacts with G protein-coupled receptors. Knocking down β-arrestin 2 slashed Aβ production, and scientists believe blocking β-arrestin 2 could offer a new therapeutic approach (see ARF related news story). This year saw the debut of appoptosin, an APP-binding protein that promotes programmed cell death. People with AD seem to accumulate the protein in the brain, suggesting it may be pathologically relevant (see ARF related news story).
12. Policy and Funding
This year saw some policy goodwill for Alzheimer’s research, but nonetheless was a dismal year for funding. In early January, the U.S. Department of Health and Human Services (HHS) released a draft of the National Plan to Address Alzheimer's Disease, a stipulation of the National Alzheimer's Project Act. It called on input from researchers and the public (see ARF related news story), prompting a flurry of advocacy (see ARF related news story). In late April, HHS Secretary Kathleen Sebelius provided an update (see ARF related news story), and officially announced the plan on 15 May. The plan set forth five main goals as outlined in earlier drafts (see ARF related news story). The administration invested an additional $130 million in AD research (see ARF related news story), a far cry from the $2 billion that advocates had wanted.
Congressional representatives in February introduced the Spending Reductions Through Innovations in Therapies (SPRINT) Agenda Act. If passed, SPRINT aims to cut healthcare costs over the long term by investing in research up front. It would invest an additional $50 million for research in 2013, requiring twice that in matching private funds (see ARF related news story).
Of course, the Budget Control Act of 2011, endlessly referred to as the "Fiscal Cliff," may yet slash spending across the board, including the NIH budget and research dollars for Alzheimer's. At the time of writing, on 31 December 2012, Congress appeared to be inching toward an eleventh-hour solution of sorts, but there has been no announcement of a deal. That’s a story for 2013.—Tom Fagan and Gabrielle Strobel.