Alves F, Kalinowski P, Ayton S.
Accelerated Brain Volume Loss Caused by Anti-β-Amyloid Drugs: A Systematic Review and Meta-analysis.
Neurology. 2023 May 16;100(20):e2114-e2124. Epub 2023 Mar 27
PubMed.
The outcomes of the aducanumab and lecanemab Phase 3 trials are small: -0.39 and -0.45 CDR-SB point differences between treatment and placebo, which are rendered as 22 percent and 27 percent slowing of clinical decline relative to placebo. The expected reduction in Aβ plaques with the antibodies is prominently highlighted as a biomarker for clinical course, and serves as the basis for FDA’s accelerated approvals as evidence for the likelihood of clinical benefit. Although MRI brain volumes are also planned outcomes of these trials, they haven’t received as much attention, even though the antibody-placebo differences are nominally statistically significant. No doubt this is because they run counterintuitive to what was expected.
In the context of Phase 3 trials, the relative decreases in MRI brain volumes of 28 percent, and increases in ventricle volumes of 36 percent, are big deals. Relatively speaking, it is an equal and opposite effect to the clinical CDR-SB effect. Yet we seem to either not discuss it or explain the volume changes away as due to the removal of plaques, as an Eisai spokesperson put it to a reporter for the journal Science.
As I mentioned previously in Alzforum (comment on No Easy Answers on Clinical Meaningfulness of Alzheimer’s Treatments, Feb 2023), the Alzheimer’s Association’s hand-picked experts, and the pharma sponsor redefined the very small -0.45 CDR-SB difference at 18 months as being 27 percent better than placebo; said this relative difference will only get bigger and better; and should be expressed as a time savings or gain. Eisai did this with a slope analysis presented at CTAD 2022, where they suggested a 5.3-month delay in clinical deterioration after 18 months, increasing to 7.5 months after 25 months of treatment. This begs the question how they would parse the obviously unwanted and concerning MRI outcomes in the trials. Would they say that the 35 percent relative increase in ventricular volume and 27 percent relative decrease in brain volume with lecanemab projects to inexorably and substantially greater hydrocephalus, atrophy with lecanemab, and an increased clinical progression of about half a year? Is losing about a teaspoon of MRI brain volume (5.2 mL), per the Melbourne group’s analysis, a good thing? A kind of brain diet? We really don’t know, yet we speculate and provide various explanations, as the authors of the Neurology editorial did.
A counter-fact is that the small, -0.45 CDR-SB, clinical difference would have been tremendously compelling if it had been accompanied by MRI brain-volume increases. Without further consideration, we would be celebrating increased MRI volume as neuroprotection. This of course did not happen, and we are left explaining away the unexpected, and unwanted.
The FDA is encouraging and approving treatments based on pure biomarker definitions and outcomes. The Alzheimer’s Association/NIA draft “ATN” research criteria recommends MRI volumes be used for the N, “neurodegeneration,” component (Jack et al., 2016). A finding of “N” moving opposite to A and T might undermine ATN criteria for use in trials or as a surrogate outcome unless there is better understanding.
Unfortunately, the Melbourne group is limited to meta-analyses because the pharmaceutical sponsors will not share their trial data, provide transparent analyses, or allow independent analyses. Indeed, a specific request through the Vivli.org clinical trials data-sharing platform was denied by Biogen. Independent post hoc analyses might discover relationships to the volume changes that make sense, e.g., as Ayton speculated, a relationship with edema. Pharma and academics should share their data.
References:
Jack CR Jr, Bennett DA, Blennow K, Carrillo MC, Feldman HH, Frisoni GB, Hampel H, Jagust WJ, Johnson KA, Knopman DS, Petersen RC, Scheltens P, Sperling RA, Dubois B.
A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers.
Neurology. 2016 Aug 2;87(5):539-47. Epub 2016 Jul 1
PubMed.
This paper is important and deserves study. It calls attention to many previous observations that treatments directed at amyloid often, but not always, are accompanied by brain volume loss. A careful meta-analysis of the data available to the authors is presented. The authors also appear to imply that brain volume loss is accompanied by, or causes, negative consequences to the study participants.
One important limitation is that the authors state that they have not been able to gain access to the data from the aducanumab, lecanemab, and donanemab clinical trials. Of course, these are the most important data, because aducanumab and lecanemab have received accelerated approval by the U.S. FDA, and lecanemab is now being made available to appropriate patients in the Veterans Administration. The Phase 3 trial of donanemab will be completed soon. The companies should report brain volume changes associated with treatment, the relationship to loss of amyloid plaques, ARIA, and cognition.
We do not understand the mechanism of these brain volume changes. I believe that the authors are correct in stating that the volume of amyloid loss would not account for the reductions of brain volume. However, amyloid plaques contain much more than just amyloid. Glial activation, which removes amyloid plaques, may be clearing other materials that might account for a portion of the volume loss. Fluid shifts may also play a role. All of this is speculative.
The authors appear to take the view that brain volume loss in this context is necessarily a bad thing for the patient. It is correct that previous studies have shown a correlation of brain volume (especially hippocampal volume) with neuronal cell counts or neuronal volume. But those observations do not necessarily indicate that the volume loss which occurs in response to anti-amyloid treatment (especially plaque-lowering monoclonal antibodies) is caused by neuronal loss or reductions of neuronal volume. Possibly, changes in volume of astrocytes or glia could account for some of this. Could there be a low level of brain edema in patients with amyloid plaques, and is it possible that there is a reduction of such edema in response to treatment? All of this is pure speculation.
What would be helpful is an analysis that correlates the brain volume changes to cognitive change. This would help answer very important questions: Are reductions in brain volume associated with reductions in cognition? Are the treatments less effective in patients who show brain volume reductions?
In conclusion, the authors have made an important contribution by summarizing the available data and pointing out the need for more study of this issue. Hopefully, the companies performing the clinical trials will provide more data about brain volume changes. All such data should be considered by regulators and should ultimately be available to clinicians and the patients making decisions about the use of these treatments.
This issue of accelerated brain-volume loss after anti-amyloid treatment is a relatively old question that has likely not been given the interest it deserves. The problem with volume/atrophy is that it is a nonspecific marker that might reflect many different underlying processes, including both "negative" (pathological) and "positive" (less inflammation or edema, fewer pathological molecules, etc.), even if greater volume usually correlates with greater cognitive performance.
In 2010, we showed that amyloid-positive asymptomatic elderly had greater brain volume in the anterior temporal cortex and hippocampus compared to amyloid-negative asymptomatic elderly (Chételat et al., 2010). In other words, more amyloid equated to more volume, which is akin to less amyloid equating to less volume found in these anti-amyloid trials.
This amyloid/volume correlation could reflect neuroinflammation due to amyloid, or the presence of amyloid itself, although I agree with the authors that the latter would not be expected to result in such a degree of volume change, but only have a very minor impact. However, we concluded that the volume correlation was rather likely to reflect brain reserve—those with higher temporal/hippocampal volume tolerate amyloid better than those with lower volume, who would exhibit subjective or cognitive decline such as they would be in another clinical group. Indeed, we found a positive link between volume in these regions and episodic memory performance.
Here, we lack the data to further understand the etiology/meaning/underlying mechanisms and consequences of this greater atrophy rate in trials. For sure, this would need further investigation. The question is not simple, since some of those drugs also tend to show better cognitive/clinical outcome, and even lower tau deposition, which are known to be associated with higher brain volume.
Ultimately, we need to know more about i) the factors that influence; ii) the correlates, and iii) the consequences of this anti-amyloid drug-related accelerated atrophy before any conclusion can be made.
References:
Chételat G, Villemagne VL, Pike KE, Baron JC, Bourgeat P, Jones G, Faux NG, Ellis KA, Salvado O, Szoeke C, Martins RN, Ames D, Masters CL, Rowe CC.
Larger temporal volume in elderly with high versus low beta-amyloid deposition.
Brain. 2010 Nov;133(11):3349-58.
PubMed.
We do not understand the biology of this observation completely. The reduction in brain volume is seen with all the monoclonal antibodies and was seen with a BACE inhibitor and with valproate. I would not call it atrophy, since atrophy implies cell loss, and we do not know that to be the case.
We spent a lot of time thinking about this in the verubecestat trial and concluded that the volume loss was regional, nonprogressive, and corresponded to the regions of high amyloid content (Sur et al., 2020). This suggests that amyloid dynamics may be contributing importantly to this brain volume reduction. It is an observation worthy of further attention, but the available data suggest that it is not an indication of brain injury.
References:
Sur C, Kost J, Scott D, Adamczuk K, Fox NC, Cummings JL, Tariot PN, Aisen PS, Vellas B, Voss T, Mahoney E, Mukai Y, Kennedy ME, Lines C, Michelson D, Egan MF.
BACE inhibition causes rapid, regional, and non-progressive volume reduction in Alzheimer's disease brain.
Brain. 2020 Dec 1;143(12):3816-3826.
PubMed.
I think this phenomenon, that anti-Aβ immunotherapies consistently increase the ventricular volumes, is intriguing and should receive more attention. We do not yet have a satisfactory explanation for it. To learn more, it would be great if the companies with access to the individual-level data from anti-Aβ immunotherapies trials would study this phenomenon in greater depth and share their experience with the AD community.
For example, one could study at an individual level whether 1) more efficient removal of Aβ plaques (as measured with amyloid PET) is associated with ventricular enlargement, 2) ARIA frequency and severity are associated with ventricular enlargement, 3) changes in key biomarkers, such plasma p-tau, GFAP, and NfL are associated with ventricular enlargement, and finally, whether 4) change in functional outcomes are associated with ventricular enlargement.
That said, my current hypothesis is that Aβ plaques result in neuroinflammation in AD with consequent mild swelling of the brain, and when the plaques are removed such brain changes are diminished and the brain parenchyma is somewhat reduced. This hypothesis is not proven, but we have recently shown that changes in diffusion MRI of cortical regions are associated with Aβ-plaque pathology and that this association is partly mediated by changes (GFAP) in reactive astrocytes (Spotorno et al., 2022). We hope that companies will analyze this type of diffusion MRI sequences in their clinical trials to unravel whether normalization in cortical diffusion metrics is associated with Aβ-plaque removal and whether this change might mediate the association between Aβ plaque removal and ventricular enlargement.
References:
Spotorno N, Strandberg O, Vis G, Stomrud E, Nilsson M, Hansson O.
Measures of cortical microstructure are linked to amyloid pathology in Alzheimer's disease.
Brain. 2022 Sep 21;
PubMed.
It is good to see consideration of the interesting and unanswered question of why potential Alzheimer’s-disease-modifying therapies have shown increased brain volume reductions relative to placebo. First seen in the AN1792 vaccination trial, and repeatedly since with monoclonal antibodies, this appears the norm rather than the exception for anti-amyloid immunotherapies. But this is not limited to immunotherapies—for example, treatment-related brain volume reduction is also seen with verubecestat (β-secretase inhibition) and with resveratrol (SIRT1 activation).
However, where I think more caution is needed is in equating “excess” volume loss to “excess” cognitive loss—which the evidence just does not support. Nor should volume change be extrapolated to imply “increased rates of atrophy,” or indeed, increased neurodegeneration. We know, for example, that with verubecestat the volume change occurred rapidly (within 13 weeks) and was not progressive. It certainly does not seem justified to take a volume change, translate that into a rate of atrophy and then project it forward to imply (as in figure 6) that progression from MCI to AD would be more rapid.
There are many areas in neurology where we do not equate treatment-related volume change with increased neurodegeneration; for example, when steroids or mannitol are given to reduce brain edema, they also reduce brain volumes. In multiple sclerosis trials, “pseudo-atrophy” did not imply a worse outcome for patients. In individuals undergoing hemodialysis, reversible volume changes of ~3 percent of total brain volume can occur over hours.
The correlations across therapies (at least the immunotherapies) between increased volume loss, amyloid removal, and ARIA are important, but there are lots of cross-confounds. Patient-level data would help disentangle this. We do not understand what causes these treatment-related brain volume changes in AD, and this mandates further exploration. Mechanisms are likely to differ between therapies, but we should be careful in our current interpretation. There is a danger if these issues are taken up by the press in an alarmist way. Patients have enough reasons for anxiety already.
References:
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M.
Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease.
Neurology. 2005 May 10;64(9):1563-72.
PubMed.
Turner RS, Thomas RG, Craft S, van Dyck CH, Mintzer J, Reynolds BA, Brewer JB, Rissman RA, Raman R, Aisen PS, Alzheimer's Disease Cooperative Study.
A randomized, double-blind, placebo-controlled trial of resveratrol for Alzheimer disease.
Neurology. 2015 Oct 20;85(16):1383-91. Epub 2015 Sep 11
PubMed.
Sur C, Kost J, Scott D, Adamczuk K, Fox NC, Cummings JL, Tariot PN, Aisen PS, Vellas B, Voss T, Mahoney E, Mukai Y, Kennedy ME, Lines C, Michelson D, Egan MF.
BACE inhibition causes rapid, regional, and non-progressive volume reduction in Alzheimer's disease brain.
Brain. 2020 Dec 1;143(12):3816-3826.
PubMed.
Walters RJ, Fox NC, Crum WR, Taube D, Thomas DJ.
Haemodialysis and cerebral oedema.
Nephron. 2001 Feb;87(2):143-7.
PubMed.
Zivadinov R, Reder AT, Filippi M, Minagar A, Stüve O, Lassmann H, Racke MK, Dwyer MG, Frohman EM, Khan O.
Mechanisms of action of disease-modifying agents and brain volume changes in multiple sclerosis.
Neurology. 2008 Jul 8;71(2):136-44.
PubMed.
This report provides a sobering reassessment of a previously observed but unexplained consequence of treatment with anti-Alzheimer drugs that target Aβ: many of these drugs are associated with cerebral volume loss.
The clinical correlates of this volume loss are yet to be firmly established, but this report suggests at least one adverse effect in that there is a strong correlation between ventricular volume and the monoclonal antibodies that are associated with amyloid-related imaging abnormalities (ARIA). As treatment of persons with Alzheimer's dementia using anti-amyloid drugs becomes possible in clinical practice, the potential for cerebral volume loss should be considered with ARIA as a potential adverse event.
I’d like to thank the Alzforum readership for their engagement in this research. Our reason for undertaking this work was to raise the profile of this long-reported finding, and hope this level of interest leads to more research. We believe these findings warrant concern, but cannot make firm conclusions about whether this reflects accelerated neurodegeneration or something benign. For several years I have been highlighting this issue as one that requires prioritization as we enter the therapeutic era (Ayton, 2022; Ayton, 2021). There is much data that has been collected that would help us illuminate this finding, but the data hasn’t been explored or released. We are left wondering, and there are speculations in both directions. The available evidence we have already provides some insight.
In the only study that has related drug-induced volumetric changes to clinical outcomes, volume changes caused by verubecestat were associated with worse clinical outcomes (Sur et al., 2020). The only paper that reported an association between drug-induced volume changes and NfL reported that volume changes caused by donanemab were correlated with plasma NfL (Pontecorvo et al., 2022). I regard both of these as preliminary evidence, but the available evidence suggests that volume changes may be detrimental and underscores why we should be conservative. Especially when dealing with the possibility of brain damage, we should be cautious in our interpretation and gather more data before dismissing this finding. Reinterpretation of the volume data requires evidence to support this.
I’m glad to read a consensus that plaque removal cannot account for the volume changes induced by anti-amyloid drugs.
A putative reduction in inflammation caused by anti-amyloid therapies is another potential explanation for the change in volume. This, too, is a speculation that requires evidence. The evidence we currently have argues against this possibility. ARIA is MRI-detectable inflammation, and we show that there is a strong correlation between the frequency of ARIA and the extent of ventricular enlargement. If brain volumes were increasing due to inflammation, we would see the opposite result—ARIA would be associated with increased brain volume. By no means is this the end of the argument, but the only available evidence we currently have regarding inflammation is correlated strongly in the wrong direction. So we should not currently hold this explanation with much certainty. The reverse possibility—that inflammation causes degeneration, is a concept grounded in a large body of literature and is the conservative interpretation that fits the data. Until we have evidence to direct us otherwise, we should be cautious in attributing volume changes to reduced neuroinflammation.
The same can be said for edema due to CSF drainage. It has been put to me that anti-amyloid drugs will cause less water retention in the brain. Yet, they increase lateral ventricular volume.
Why should we care about these results if anti-amyloid drugs improve cognition? The clinical significance of the cognitive findings remains contentious. Brain volumetrics is supportive and objective evidence of disease progression that would have been used to support the clinical result if the finding were in the opposite direction. When the clinical findings are unclear, these secondary outcomes help provide context to the primary result. The volumetric findings do not provide additional support of efficacy.
Additional questions emerge from volumetric data that may guide our questions on efficacy:
Does everyone respond positively to anti-amyloid therapies? It is possible that some (e4?) deteriorate due to volumetric changes? The only study that has investigated volumetric changes when stratified by APOE e4 is bapineuzumab (Novak et al., 2016), and these data show that e4 subjects exhibit greater deterioration. Could this explain why e4 subjects respond less favorably to lecanemab in the exploratory analysis (van Dyck et al., 2023)?
Do drug-induced volume changes impact non-cognitive outcomes? I note that hippocampus is not affected by monoclonal antibodies. We need to look at a range of brain regions. Donanemab has outcomes for 13 brain regions (reported on clinicaltrials.gov: NCT03367403), which stands apart from most studies that report on average two to three. Perhaps clinical symptoms not captured with the instruments used in the clinical trials deteriorate in response to anti-amyloid therapies, and examining volume changes to a range of structures could guide us what to look at.
Do anti-Aβ drugs provide benefit in the long term? It is conceivable that these drugs benefit patients in the short term, but patients regress when the effects of accelerated neurodegeneration take hold. (Note that all drugs except verubecestat showed volume changes that increased over time). The example of tofersen for SOD1 ALS might provide perspective. The reduction in NfL by tofersen was not accompanied by clinical improvement in the blinded phase of the study; rather, clinical benefit was apparently delayed and observed in open-label extension (Miller et al., 2022). If it is argued that slowing neurodegeneration can lead to a delayed clinical benefit in the case of toferson, it is equally plausible that accelerating brain-volume changes would lead to a delayed clinical detriment in the case of anti-amyloid therapies. We need longer-term follow-up to determine this.
This is all speculation. We need more data and analysis.
A final thought—as has been pointed out in the Alzforum article, this finding is not new. We have observed this in scores of trials for more than a decade. Yet we know so little about it, and hardly any exploration into the available data has been performed. It should give us pause that we cannot provide basic explanations into this phenomenon as we now enter an era where anti-amyloid drugs have approval. We should ask, why as a field have we not demanded more questions of these data as they have been published in numerous journals over many years? I hope that now we will see these data more intensively investigated. I sincerely hope that the finding is benign and does not risk patient safety.
References:
Ayton S.
Ventricular enlargement caused by aducanumab.
Nat Rev Neurol. 2022 Jul;18(7):383-384.
PubMed.
Ayton S.
Brain volume loss due to donanemab.
Eur J Neurol. 2021 Sep;28(9):e67-e68. Epub 2021 Jul 16
PubMed.
Sur C, Kost J, Scott D, Adamczuk K, Fox NC, Cummings JL, Tariot PN, Aisen PS, Vellas B, Voss T, Mahoney E, Mukai Y, Kennedy ME, Lines C, Michelson D, Egan MF.
BACE inhibition causes rapid, regional, and non-progressive volume reduction in Alzheimer's disease brain.
Brain. 2020 Dec 1;143(12):3816-3826.
PubMed.
Pontecorvo MJ, Lu M, Burnham SC, Schade AE, Dage JL, Shcherbinin S, Collins EC, Sims JR, Mintun MA.
Association of Donanemab Treatment With Exploratory Plasma Biomarkers in Early Symptomatic Alzheimer Disease: A Secondary Analysis of the TRAILBLAZER-ALZ Randomized Clinical Trial.
JAMA Neurol. 2022 Dec 1;79(12):1250-1259.
PubMed.
Novak G, Fox N, Clegg S, Nielsen C, Einstein S, Lu Y, Tudor IC, Gregg K, Di J, Collins P, Wyman BT, Yuen E, Grundman M, Brashear HR, Liu E.
Changes in Brain Volume with Bapineuzumab in Mild to Moderate Alzheimer's Disease.
J Alzheimers Dis. 2016;49(4):1123-34.
PubMed.
van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, Kanekiyo M, Li D, Reyderman L, Cohen S, Froelich L, Katayama S, Sabbagh M, Vellas B, Watson D, Dhadda S, Irizarry M, Kramer LD, Iwatsubo T.
Lecanemab in Early Alzheimer's Disease.
N Engl J Med. 2023 Jan 5;388(1):9-21. Epub 2022 Nov 29
PubMed.
Miller TM, Cudkowicz ME, Genge A, Shaw PJ, Sobue G, Bucelli RC, Chiò A, Van Damme P, Ludolph AC, Glass JD, Andrews JA, Babu S, Benatar M, McDermott CJ, Cochrane T, Chary S, Chew S, Zhu H, Wu F, Nestorov I, Graham D, Sun P, McNeill M, Fanning L, Ferguson TA, Fradette S, VALOR and OLE Working Group.
Trial of Antisense Oligonucleotide Tofersen for SOD1 ALS.
N Engl J Med. 2022 Sep 22;387(12):1099-1110.
PubMed.
The apparent atrophy related to anti-amyloid treatments has been a matter of discussion since the first active vaccine trial (Fox et al., 2005), and it's crucial to better understand the etiology of such phenomena to guarantee the safety of our patients. While historically atrophy has always been associated to neurodegeneration in the AD field, this may not be the case for other neurological disorders, as previously noted by Dr. Fox. Thus, the reported atrophy in the reviewed studies of this meta-analysis might arise from a beneficial effect of removing amyloid plaques rather than neurodegeneration.
We agree with Alves and collaborators that this atrophy cannot be solely due to the removal of amyloid. Instead, we have previously proposed and validated a model in which early amyloid accumulation leads to neuroinflammation that results in increased thickness. In fact, using in vivo astrocytosis PET, we previously reported a positive association between inflammation and cortical thickening in a cohort of autosomal dominant AD (Vilaplana et al., 2020). Moreover, cortical thickening has been observed in other familial AD cohorts such as DIAN (Montal et al., 2021), or the Colombian cohort (Fox-Fuller et al., 2021) and in sporadic AD (Fortea et al., 2014). Multimodal analyses using DWI data suggest that this increased thickness arises from the increment of cellular bodies, which might be caused by neuronal hypertrophy, morphology expansion in reactive astrocytes and microglia and/or glia recruitment (Montal et al., 2018). Therefore, we do not imply that the increase in thickness is caused by amyloid per se but rather by an amyloid-related inflammation process.
We hypothesize that in symptomatic stages of AD, despite the observed atrophy, this amyloid-related inflammation and consequent relative thickening persists, as previously reported in studies using in-vivo microglia PET (Femminella et al., 2019). As a result, the removal of amyloid and partial resolution of the inflammation will result in brain atrophy, despite not being detrimental.
It is crucial to study this putative effect using clinical trial data and assess if this hypothesis is confirmed at the individual-patient level. We urge the different companies behind these trials to share their data with the community to allow an in-depth comprehension of this apparent detrimental effect in anti-amyloid trials. It is important to continue research on the apparent atrophy related to anti-amyloid treatments to better understand its etiology and to ensure the safety and efficacy of these treatments.
References:
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M.
Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease.
Neurology. 2005 May 10;64(9):1563-72.
PubMed.
Vilaplana E, Rodriguez-Vieitez E, Ferreira D, Montal V, Almkvist O, Wall A, Lleó A, Westman E, Graff C, Fortea J, Nordberg A.
Cortical microstructural correlates of astrocytosis in autosomal-dominant Alzheimer disease.
Neurology. 2020 May 12;94(19):e2026-e2036. Epub 2020 Apr 14
PubMed.
Montal V, Vilaplana E, Pegueroles J, Bejanin A, Alcolea D, Carmona-Iragui M, Clarimón J, Levin J, Cruchaga C, Graff-Radford NR, Noble JM, Lee JH, Allegri R, Karch CM, Laske C, Schofield PR, Salloway S, Ances B, Benzinger T, McDale E, Bateman R, Blesa R, Sánchez-Valle R, Lleó A, Fortea J, Dominantly Inherited Alzheimer Network (DIAN).
Biphasic cortical macro- and microstructural changes in autosomal dominant Alzheimer's disease.
Alzheimers Dement. 2021 Apr;17(4):618-628. Epub 2020 Nov 16
PubMed.
Fox-Fuller JT, Torrico-Teave H, d'Oleire Uquillas F, Chen K, Su Y, Chen Y, Brickhouse M, Sanchez JS, Aguero C, Jacobs HI, Hampton O, Guzmán-Vélez E, Vila-Castelar C, Aguirre-Acevedo DC, Baena A, Artola A, Martinez J, Pluim CF, Alvarez S, Ochoa-Escudero M, Reiman EM, Sperling RA, Lopera F, Johnson KA, Dickerson BC, Quiroz YT.
Cortical thickness across the lifespan in a Colombian cohort with autosomal-dominant Alzheimer's disease: A cross-sectional study.
Alzheimers Dement (Amst). 2021;13(1):e12233. Epub 2021 Sep 14
PubMed.
Fortea J, Vilaplana E, Alcolea D, Carmona-Iragui M, Sánchez-Saudinos MB, Sala I, Antón-Aguirre S, González S, Medrano S, Pegueroles J, Morenas E, Clarimón J, Blesa R, Lleó A, Alzheimer's Disease Neuroimaging Initiative.
Cerebrospinal fluid β-amyloid and phospho-tau biomarker interactions affecting brain structure in preclinical Alzheimer disease.
Ann Neurol. 2014 Aug;76(2):223-30. Epub 2014 Jun 13
PubMed.
Montal V, Vilaplana E, Alcolea D, Pegueroles J, Pasternak O, González-Ortiz S, Clarimón J, Carmona-Iragui M, Illán-Gala I, Morenas-Rodríguez E, Ribosa-Nogué R, Sala I, Sánchez-Saudinós MB, García-Sebastian M, Villanúa J, Izagirre A, Estanga A, Ecay-Torres M, Iriondo A, Clerigue M, Tainta M, Pozueta A, González A, Martínez-Heras E, Llufriu S, Blesa R, Sanchez-Juan P, Martínez-Lage P, Lleó A, Fortea J.
Cortical microstructural changes along the Alzheimer's disease continuum.
Alzheimers Dement. 2018 Mar;14(3):340-351. Epub 2017 Oct 31
PubMed.
Femminella GD, Dani M, Wood M, Fan Z, Calsolaro V, Atkinson R, Edginton T, Hinz R, Brooks DJ, Edison P.
Microglial activation in early Alzheimer trajectory is associated with higher gray matter volume.
Neurology. 2019 Mar 19;92(12):e1331-e1343. Epub 2019 Feb 22
PubMed.
Biological mechanisms in addition to the more-studied proteopathies can hypothetically lead to atrophy, and could be related here. Some clues from literature:
Localized hypoxemia perhaps could be related to the immunotherapy response, e.g., Marchi et al., 2020.
Accumulation of reactive oxygen species in the brain leading to mitochondrial dysfunction and DNA damage, e.g., Manoharan et al., 2016.
Cell stress leading to apoptosis through genetic and epigenetic damage (perhaps from a combination of inflammatory cytokines and/or above factors or others).
Any of the above mechanisms, a combination of them, or other related atrophy causes would prevent cognitive improvement. Neuroenhancement and neuroplasticity, to my knowledge, are possible in elderly but obviously less extensive compared to younger adults who have a greater reserve of neural stem cells. The path forward to determining the underlying biological mechanisms (in this case behind atrophy) likely lies in the disproving of ideas, as Planck, Popper, and others have established in the past.
References:
Marchi NA, Ramponi C, Hirotsu C, Haba-Rubio J, Lutti A, Preisig M, Marques-Vidal P, Vollenweider P, Kherif F, Heinzer R, Draganski B.
Mean Oxygen Saturation during Sleep Is Related to Specific Brain Atrophy Pattern.
Ann Neurol. 2020 Jun;87(6):921-930. Epub 2020 Apr 20
PubMed.
Manoharan S, Guillemin GJ, Abiramasundari RS, Essa MM, Akbar M, Akbar MD.
The Role of Reactive Oxygen Species in the Pathogenesis of Alzheimer's Disease, Parkinson's Disease, and Huntington's Disease: A Mini Review.
Oxid Med Cell Longev. 2016;2016:8590578. Epub 2016 Dec 27
PubMed.
Geloso MC, D'Ambrosi N.
Microglial Pruning: Relevance for Synaptic Dysfunction in Multiple Sclerosis and Related Experimental Models.
Cells. 2021 Mar 20;10(3)
PubMed.
Korbo L.
Glial cell loss in the hippocampus of alcoholics.
Alcohol Clin Exp Res. 1999 Jan;23(1):164-8.
PubMed.
The paper by Alves and colleagues is methodologically sound. It brings a piece of high-level evidence and highlights this poorly investigated but universal finding of high-clearance anti-amyloid immunotherapies: ventricular enlargement +/- gray matter loss. The debate and numerous comments underline the relevance and uncertainty of these findings, and the need for further investigation. I align with my mentor’s (Gaël Chételat) opinion here.
In addition to the arguments already detailed here, I would like to emphasize three elements:
The long-term neuropathological findings from the cited AN1792 trial did not evidence increased synaptic/neuronal loss despite increased ventricular enlargement +/- gray matter loss (Boche et al., 2010; Fox et al., 2005; Holmes et al., 2008; Nicoll et al., 2019). This is indirect, and preliminary, evidence that increased ventricular enlargement +/- gray matter loss in these trials does not reflect accelerated neurodegeneration. Of course, this piece of evidence has several potential biases, but it is reassuring before we get these long-term neuropath data in current trials.
The relationship with plasma NFL in the Phase 2 donanemab trial (Pontecorvo et al., 2022), highlighted by Alves et al. as an argument in favor of neurodegeneration, is very low evidence: amongst the six correlations tested (uncorrected for multiple testing), only one is significant, and Spearman Rho values range between 0.12 and -0.17.
The authors claim that in AD, unless the contrary is proven, ventricular enlargement and gray matter loss reflect neurodegeneration, which is an undisputed assertion. In line with this argument: In AD, ventricular enlargement and gray matter loss correlate very well with cognitive impairment (one example amongst hundreds: Hua et al., 2010). Unfortunately, to date, no data has been disclosed by the sponsors of the high-clearance anti-amyloid immunotherapies trials. Even so, it is highly improbable that the cognitive effect would be positively related to the MRI effect as in the natural course of AD, whereas, during these trials, the means of the two variables go significantly in oppositive directions. Thus, it is very likely that the relationship between ventricular enlargement +/- gray matter loss and cognition in these trials is the opposite as in the natural course of AD. If proven, this would be a strong argument demonstrating that a different biological mechanism underlies ventricular enlargement +/- gray matter loss in high-clearance anti-amyloid immunotherapies trials and in the natural course of AD.
Overall, the current evidence supporting that ventricular enlargement +/- gray matter loss in these trials reflects neurodegeneration is meager. On the other hand, the quality of evidence supporting the opposite is, in my opinion, more convincing. Both remain insufficient. Like Alves et al. and almost all commentators here, I advocate for further transparency and disclosure of available MRI data in the current high-clearance anti-amyloid immunotherapies trials: especially the correlation between cognitive outcomes and MRI metrics during the core phase of the trials and long-term OLE data cognitive outcomes and MRI metrics.
References:
Boche D, Denham N, Holmes C, Nicoll JA.
Neuropathology after active Abeta42 immunotherapy: implications for Alzheimer's disease pathogenesis.
Acta Neuropathol. 2010 Sep;120(3):369-84.
PubMed.
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M.
Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease.
Neurology. 2005 May 10;64(9):1563-72.
PubMed.
Holmes C, Boche D, Wilkinson D, Yadegarfar G, Hopkins V, Bayer A, Jones RW, Bullock R, Love S, Neal JW, Zotova E, Nicoll JA.
Long-term effects of Abeta42 immunisation in Alzheimer's disease: follow-up of a randomised, placebo-controlled phase I trial.
Lancet. 2008 Jul 19;372(9634):216-23.
PubMed.
Hua X, Hibar DP, Lee S, Toga AW, Jack CR, Weiner MW, Thompson PM, .
Sex and age differences in atrophic rates: an ADNI study with n=1368 MRI scans.
Neurobiol Aging. 2010 Aug;31(8):1463-80.
PubMed.
Nicoll JA, Buckland GR, Harrison CH, Page A, Harris S, Love S, Neal JW, Holmes C, Boche D.
Persistent neuropathological effects 14 years following amyloid-β immunization in Alzheimer's disease.
Brain. 2019 Jul 1;142(7):2113-2126.
PubMed.
Pontecorvo MJ, Lu M, Burnham SC, Schade AE, Dage JL, Shcherbinin S, Collins EC, Sims JR, Mintun MA.
Association of Donanemab Treatment With Exploratory Plasma Biomarkers in Early Symptomatic Alzheimer Disease: A Secondary Analysis of the TRAILBLAZER-ALZ Randomized Clinical Trial.
JAMA Neurol. 2022 Dec 1;79(12):1250-1259.
PubMed.
Clinicians have often associated focal or generalized cerebral atrophy seen on imaging studies with loss of brain tissue, or loss of specific cell types, due to injury or degeneration. It is understandable that apparent loss of brain tissue leads to concern, even though cerebral atrophy can be due to reversible causes, such as anorexia nervosa (Heinz et al., 1977), Cushing’s syndrome (Heinz et al., 1977), ACTH/Corticosteroid therapy (Lagenstein et al., 1979; Bentson et al., Apparent cerebral atrophy on computed tomography following steroid use. Presented at the International Symposium on Computer Assisted Tomography in Nontumoral Diseases of the Brain, Spinal Cord, and Eye. National Institutes of Health, Bethesda, Maryland, 11-15 October 1976), alcoholism (Carlen et al., 1978), and multiple sclerosis (De Stafano and Arnold, 2015), and despite the general acceptance that some loss of brain volume is expected due to aging.
It is well-documented that several types of anti-amyloid therapies, from active vaccines to large and small molecules, have also been associated with more loss of brain volume and greater ventricular expansion in treated patients compared to randomized controls. Since the earliest observations of this phenomenon (Fox et al., Apparent cerebral atrophy on computed tomography following steroid use. Presented at the International Symposium on Computer Assisted Tomography in Nontumoral Diseases of the Brain, Spinal Cord, and Eye. National Institutes of Health, Bethesda, Maryland, 11-15 October 1976.), there has been no clear association between clinical status and this treatment-related brain atrophy.
This systematic review and meta-analysis by Alves et al. helps to quantify the extent of this observation in patients who: took various types of anti-amyloid drugs, had at least one biomarker of target engagement, and at least one MRI sufficient to evaluate at least one brain region. It should be noted that their review included only 31 of the 145 studies identified, so additional insights may still be derived from the excluded studies. They attempted to relate atrophy to ARIA at the group level (in studies in which ARIA occurred) using Pearson’s regressions and modeled how much “worse” MCI patients were due to treatment, with the unsupported assumption that rate of atrophy was equal to disease progression.
We have reviewed vMRI data from the recently-reported Graduate studies, and concluded that the observed differences in vMRI measures in gantenerumab-treated patients versus placebo-treated patients, which were similar to those seen with other anti-Aβ monoclonal antibodies (MABs), could be due to a beneficial effect, such as reduced inflammation associated with plaque removal, or changes in CSF dynamics (Barkhoff and Knopman, 2023), and may not be due to accelerated neurodegeneration (data in preparation). Given the clinical benefits seen with some MABs (van Dyck et al., 2023; Mintun et al., 2021), it is difficult to conclude that atrophy seen with those treatments represents diffuse neuronal loss. Further, in our Graduate studies, the reduction of NFL in CSF seen with gantenerumab treatment would argue against accelerated neurodegeneration (Bittner et al., GRADUATE I and II results: Effect of subcutaneous gantenerumab on fluid biomarkers of AD pathology and neurodegeneration. Presented atvADPD, Gothenberg, Sweden, March 30, 2023).
In conclusion, we do not know what causes atrophy in reversible conditions, normal aging, or with various anti-Aβ amyloid treatments. The reasons may be diverse and should not be assumed to be related to brain injury or worsening of the neuropathologic process of AD. In our view, the weight of the evidence to date would argue against hypotheses of detriment.
References:
Heinz ER, Martinez J, Haenggeli A.
Reversibility of cerebral atrophy in anorexia nervosa and Cushing's syndrome.
J Comput Assist Tomogr. 1977 Oct;1(4):415-8.
PubMed.
Lagenstein I, Willig RP, Kühne D.
Reversible cerebral atrophy caused by corticotrophin.
Lancet. 1979 Jun 9;1(8128):1246-7.
PubMed.
Carlen PL, Wortzman G, Holgate RC, Wilkinson DA, Rankin JC.
Reversible cerebral atrophy in recently abstinent chronic alcoholics measured by computed tomography scans.
Science. 1978 Jun 2;200(4345):1076-8.
PubMed.
De Stefano N, Arnold DL.
Towards a better understanding of pseudoatrophy in the brain of multiple sclerosis patients.
Mult Scler. 2015 May;21(6):675-6. Epub 2015 Jan 26
PubMed.
Barkhof F, Knopman DS.
Brain Shrinkage in Anti-β-Amyloid Alzheimer Trials: Neurodegeneration or Pseudoatrophy?.
Neurology. 2023 May 16;100(20):941-942. Epub 2023 Mar 27
PubMed.
van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, Kanekiyo M, Li D, Reyderman L, Cohen S, Froelich L, Katayama S, Sabbagh M, Vellas B, Watson D, Dhadda S, Irizarry M, Kramer LD, Iwatsubo T.
Lecanemab in Early Alzheimer's Disease.
N Engl J Med. 2023 Jan 5;388(1):9-21. Epub 2022 Nov 29
PubMed.
Mintun MA, Lo AC, Duggan Evans C, Wessels AM, Ardayfio PA, Andersen SW, Shcherbinin S, Sparks J, Sims JR, Brys M, Apostolova LG, Salloway SP, Skovronsky DM.
Donanemab in Early Alzheimer's Disease.
N Engl J Med. 2021 May 6;384(18):1691-1704. Epub 2021 Mar 13
PubMed.
With currently available technology, it seems unlikely that the pathophysiology associated with Aβ immunotherapy-related atrophy can be understood without studying human brain tissue. That there has not been systematic postmortem follow-up of patients in recent passive immunotherapy trials represents a missed opportunity. We gathered a postmortem cohort of 22 subjects who had received AN1792, demonstrating removal of Aβ plaques and a relative reduction in both tau and microglial activation. Remarkably, these effects persisted as long as 14 years after active immunization.
Imaging analyses reported that immunotherapies, whether or not associated with amyloid removal and independently of the protocol (active versus passive), have led to either a reduction in whole-brain volume or an increased ventricular volume in treated participants compared to placebo (AN1792 (Fox et al., 2005); Bapineuzumab (Novak et al., 2016), Aducanumab (Budd Haeberlein et al., 2022), Donanemab (Mintun et al., 2021), and now Lecanemab (Swanson et al., 2021; van Dyck et al., 2023). The biological basis for this observation is unknown; however, interestingly, those imaging results are concordant with the analysis describing brain atrophy in our postmortem cohort of immunized AD (iAD) patients compared with non-immunized AD (cAD) patients (Paquet et al., 2015; Serrano-Pozo et al., 2010).
Brain atrophy was defined by increased cerebral cortical neuropil degeneration (cAD 48.6 percent versus iAD 53.3 percent; p=0.013), increased neuronal loss (cAD 73.7 versus iAD 66.8 neurons/field; p=0.036) and inter-neuronal distance (cAD 695 µm vs. iAD 703.6 µm; p=0.006). However, while atrophy in non-immunized AD has been consistently associated with progression of the disease and the severity of pathological changes rather than slowing of symptoms, our neuropathological studies did not observe an association between the number of neurons and the cognitive evolution nor the side effects. Of note, the brain atrophy was not related to markers of neurodegeneration (tangles, Aβ42 load, pPKR), and conversely, we observed improved health of the residual neurons with less neuritic curvature and the presence of fewer pro-apoptotic neurons in the immunized brains Paquet et al., 2015; Serrano-Pozo et al., 2010; Paquet et al., 2017).
Now that we are at the exciting stage of licensing of Aβ immunotherapeutic agents for clinical use, we strongly encourage study of the brains of participants of trials of passive immunotherapy in order to throw light on the important issue of atrophy and whether it is good, bad, or indifferent.
References:
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M.
Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease.
Neurology. 2005 May 10;64(9):1563-72.
PubMed.
Novak G, Fox N, Clegg S, Nielsen C, Einstein S, Lu Y, Tudor IC, Gregg K, Di J, Collins P, Wyman BT, Yuen E, Grundman M, Brashear HR, Liu E.
Changes in Brain Volume with Bapineuzumab in Mild to Moderate Alzheimer's Disease.
J Alzheimers Dis. 2016;49(4):1123-34.
PubMed.
Budd Haeberlein S, Aisen PS, Barkhof F, Chalkias S, Chen T, Cohen S, Dent G, Hansson O, Harrison K, von Hehn C, Iwatsubo T, Mallinckrodt C, Mummery CJ, Muralidharan KK, Nestorov I, Nisenbaum L, Rajagovindan R, Skordos L, Tian Y, van Dyck CH, Vellas B, Wu S, Zhu Y, Sandrock A.
Two Randomized Phase 3 Studies of Aducanumab in Early Alzheimer's Disease.
J Prev Alzheimers Dis. 2022;9(2):197-210.
PubMed.
Mintun MA, Lo AC, Duggan Evans C, Wessels AM, Ardayfio PA, Andersen SW, Shcherbinin S, Sparks J, Sims JR, Brys M, Apostolova LG, Salloway SP, Skovronsky DM.
Donanemab in Early Alzheimer's Disease.
N Engl J Med. 2021 May 6;384(18):1691-1704. Epub 2021 Mar 13
PubMed.
Swanson CJ, Zhang Y, Dhadda S, Wang J, Kaplow J, Lai RY, Lannfelt L, Bradley H, Rabe M, Koyama A, Reyderman L, Berry DA, Berry S, Gordon R, Kramer LD, Cummings JL.
A randomized, double-blind, phase 2b proof-of-concept clinical trial in early Alzheimer's disease with lecanemab, an anti-Aβ protofibril antibody.
Alzheimers Res Ther. 2021 Apr 17;13(1):80.
PubMed.
Correction.
van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, Kanekiyo M, Li D, Reyderman L, Cohen S, Froelich L, Katayama S, Sabbagh M, Vellas B, Watson D, Dhadda S, Irizarry M, Kramer LD, Iwatsubo T.
Lecanemab in Early Alzheimer's Disease.
N Engl J Med. 2023 Jan 5;388(1):9-21. Epub 2022 Nov 29
PubMed.
Paquet C, Amin J, Mouton-Liger F, Nasser M, Love S, Gray F, Pickering RM, Nicoll JA, Holmes C, Hugon J, Boche D.
Effect of active Aβ immunotherapy on neurons in human Alzheimer's disease.
J Pathol. 2015 Apr;235(5):721-30. Epub 2015 Jan 7
PubMed.
Serrano-Pozo A, William CM, Ferrer I, Uro-Coste E, Delisle MB, Maurage CA, Hock C, Nitsch RM, Masliah E, Growdon JH, Frosch MP, Hyman BT.
Beneficial effect of human anti-amyloid-beta active immunization on neurite morphology and tau pathology.
Brain. 2010 May;133(Pt 5):1312-27.
PubMed.
Paquet C, Nicoll JA, Love S, Mouton-Liger F, Holmes C, Hugon J, Boche D.
Downregulated apoptosis and autophagy after anti-Aβ immunotherapy in Alzheimer's disease.
Brain Pathol. 2017 Oct 13;
PubMed.
This study shows that frequently there is not adequate analysis of prior clinical data, an exercise helpful in designing new therapeutic agents and clinical studies. A pertinent case is the AN1792 vaccine for AD, which used the potent pro-inflammatory immune modulator QS-21 despite lack of understanding of its pharmacological properties. The situation was worsened by changes made in Phase 2 without a rigorous scientific analysis. For some time it was assumed that despite AN1792’s serious side effects, it had some beneficial effects but, as reported by von Bernhardi (2010), those receiving only QS-21 as a placebo had an unusually high rate of cognitive decline, 6 to 7 points loss in the MMSE, which exceeds the average 3.5 to 4 points for AD patients. This was likely because QS-21 induces the damaging pro-inflammatory immunity Th17, disrupting the immune homeostasis needed to maintain an anti-inflammatory Th2 immunity and prevent aggravating the disease. Since many AD vaccines used similar immune modulators or adjuvants, one can assume that they elicited the damaging pro-inflammatory immunities Th1 and/or Th17. These effects complicate interpretation of the clinical results.
However, the situation may be different with passive immunotherapy. Aβ insolubilized as plaque is solubilized by IgGs that bind this protein, irrespective of their therapeutic value. By binding to Aβ and forming complexes, the IgGs weaken ionic-hydrophobic interactions that stabilize plaque. Once this occurs, the Aβ-IgG complexes become soluble. Actually, IgGs due to their carbohydrate content are quite soluble in an aqueous milieu. Thus, the degree of plaque solubilization should correlate with the dose of anti-Aβ IgG. Hence, the higher the IgG dose, the less plaque would be left, which may explain the reported brain shrinkage.
This solubilization process is rather like the old prozone effect, where an antigen in the presence of specific antibodies aggregates and precipitates. But upon addition of an excess of those antibodies, the interactions among adjacent antigen molecules bridged by IgGs are disrupted. The process leads to formation of soluble antigen-IgG complexes, and a concomitant loss in the amount of antigen/IgG precipitate. Solubilization of Aβ plaque by antibodies against it was studied by J.Y. Wang et al. (Liu et al., 2015). They called it the “dust-raising effect.” The study concluded that the release from plaque of soluble toxic Aβ oligomers (AβOs), the accepted causative agents of this disease (Cline et al., 2018; Walsh and Selkoe, 2020), increases neurotoxicity. However, those results were reported in 2015, before the availability of antibodies like lecanemab, capable of neutralizing AβOs’ cytotoxicity. Thus, it will be of interest to compare the effects of lecanemab, which has been found to slow Alzheimer’s disease, with brain shrinkage. While lecanemab will solubilize plaque, the released soluble AβOs must be inactivated by this monoclonal antibody, which should minimize the likely neurological damage caused by the solubilized cytotoxic AβOs derived from plaque.
As the authors and many commentators indicate, it will be helpful to compare the results from the clinical trials with the pharmacological effects of the therapeutic agents used. This task will require collaboration of the different companies that ran those studies, to obtain the relevant data. This should not present problems, since many companies discontinued development of the drugs after their clinical studies failed.
References:
von Bernhardi R.
Immunotherapy in Alzheimer's disease: where do we stand? Where should we go?.
J Alzheimers Dis. 2010;19(2):405-21.
PubMed.
Liu YH, Bu XL, Liang CR, Wang YR, Zhang T, Jiao SS, Zeng F, Yao XQ, Zhou HD, Deng J, Wang YJ.
An N-terminal antibody promotes the transformation of amyloid fibrils into oligomers and enhances the neurotoxicity of amyloid-beta: the dust-raising effect.
J Neuroinflammation. 2015 Aug 28;12:153.
PubMed.
Cline EN, Bicca MA, Viola KL, Klein WL.
The Amyloid-β Oligomer Hypothesis: Beginning of the Third Decade.
J Alzheimers Dis. 2018;64(s1):S567-S610.
PubMed.
Walsh DM, Selkoe DJ.
Amyloid β-protein and beyond: the path forward in Alzheimer's disease.
Curr Opin Neurobiol. 2020 Apr;61:116-124.
PubMed.
I am surprised to see no discussion here of T2 weighting. I have no expertise in imaging, but wouldn't the T2/T1 ratio help to determine how much of the immunotherapy-related change is a loss of edema versus macromolecular structure?
Comments
University of Southern California Keck School of Medicine
The outcomes of the aducanumab and lecanemab Phase 3 trials are small: -0.39 and -0.45 CDR-SB point differences between treatment and placebo, which are rendered as 22 percent and 27 percent slowing of clinical decline relative to placebo. The expected reduction in Aβ plaques with the antibodies is prominently highlighted as a biomarker for clinical course, and serves as the basis for FDA’s accelerated approvals as evidence for the likelihood of clinical benefit. Although MRI brain volumes are also planned outcomes of these trials, they haven’t received as much attention, even though the antibody-placebo differences are nominally statistically significant. No doubt this is because they run counterintuitive to what was expected.
In the context of Phase 3 trials, the relative decreases in MRI brain volumes of 28 percent, and increases in ventricle volumes of 36 percent, are big deals. Relatively speaking, it is an equal and opposite effect to the clinical CDR-SB effect. Yet we seem to either not discuss it or explain the volume changes away as due to the removal of plaques, as an Eisai spokesperson put it to a reporter for the journal Science.
As I mentioned previously in Alzforum (comment on No Easy Answers on Clinical Meaningfulness of Alzheimer’s Treatments, Feb 2023), the Alzheimer’s Association’s hand-picked experts, and the pharma sponsor redefined the very small -0.45 CDR-SB difference at 18 months as being 27 percent better than placebo; said this relative difference will only get bigger and better; and should be expressed as a time savings or gain. Eisai did this with a slope analysis presented at CTAD 2022, where they suggested a 5.3-month delay in clinical deterioration after 18 months, increasing to 7.5 months after 25 months of treatment. This begs the question how they would parse the obviously unwanted and concerning MRI outcomes in the trials. Would they say that the 35 percent relative increase in ventricular volume and 27 percent relative decrease in brain volume with lecanemab projects to inexorably and substantially greater hydrocephalus, atrophy with lecanemab, and an increased clinical progression of about half a year? Is losing about a teaspoon of MRI brain volume (5.2 mL), per the Melbourne group’s analysis, a good thing? A kind of brain diet? We really don’t know, yet we speculate and provide various explanations, as the authors of the Neurology editorial did.
A counter-fact is that the small, -0.45 CDR-SB, clinical difference would have been tremendously compelling if it had been accompanied by MRI brain-volume increases. Without further consideration, we would be celebrating increased MRI volume as neuroprotection. This of course did not happen, and we are left explaining away the unexpected, and unwanted.
The FDA is encouraging and approving treatments based on pure biomarker definitions and outcomes. The Alzheimer’s Association/NIA draft “ATN” research criteria recommends MRI volumes be used for the N, “neurodegeneration,” component (Jack et al., 2016). A finding of “N” moving opposite to A and T might undermine ATN criteria for use in trials or as a surrogate outcome unless there is better understanding.
Unfortunately, the Melbourne group is limited to meta-analyses because the pharmaceutical sponsors will not share their trial data, provide transparent analyses, or allow independent analyses. Indeed, a specific request through the Vivli.org clinical trials data-sharing platform was denied by Biogen. Independent post hoc analyses might discover relationships to the volume changes that make sense, e.g., as Ayton speculated, a relationship with edema. Pharma and academics should share their data.
References:
Jack CR Jr, Bennett DA, Blennow K, Carrillo MC, Feldman HH, Frisoni GB, Hampel H, Jagust WJ, Johnson KA, Knopman DS, Petersen RC, Scheltens P, Sperling RA, Dubois B. A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology. 2016 Aug 2;87(5):539-47. Epub 2016 Jul 1 PubMed.
View all comments by Lon S. SchneiderUCSF
This paper is important and deserves study. It calls attention to many previous observations that treatments directed at amyloid often, but not always, are accompanied by brain volume loss. A careful meta-analysis of the data available to the authors is presented. The authors also appear to imply that brain volume loss is accompanied by, or causes, negative consequences to the study participants.
One important limitation is that the authors state that they have not been able to gain access to the data from the aducanumab, lecanemab, and donanemab clinical trials. Of course, these are the most important data, because aducanumab and lecanemab have received accelerated approval by the U.S. FDA, and lecanemab is now being made available to appropriate patients in the Veterans Administration. The Phase 3 trial of donanemab will be completed soon. The companies should report brain volume changes associated with treatment, the relationship to loss of amyloid plaques, ARIA, and cognition.
We do not understand the mechanism of these brain volume changes. I believe that the authors are correct in stating that the volume of amyloid loss would not account for the reductions of brain volume. However, amyloid plaques contain much more than just amyloid. Glial activation, which removes amyloid plaques, may be clearing other materials that might account for a portion of the volume loss. Fluid shifts may also play a role. All of this is speculative.
The authors appear to take the view that brain volume loss in this context is necessarily a bad thing for the patient. It is correct that previous studies have shown a correlation of brain volume (especially hippocampal volume) with neuronal cell counts or neuronal volume. But those observations do not necessarily indicate that the volume loss which occurs in response to anti-amyloid treatment (especially plaque-lowering monoclonal antibodies) is caused by neuronal loss or reductions of neuronal volume. Possibly, changes in volume of astrocytes or glia could account for some of this. Could there be a low level of brain edema in patients with amyloid plaques, and is it possible that there is a reduction of such edema in response to treatment? All of this is pure speculation.
What would be helpful is an analysis that correlates the brain volume changes to cognitive change. This would help answer very important questions: Are reductions in brain volume associated with reductions in cognition? Are the treatments less effective in patients who show brain volume reductions?
In conclusion, the authors have made an important contribution by summarizing the available data and pointing out the need for more study of this issue. Hopefully, the companies performing the clinical trials will provide more data about brain volume changes. All such data should be considered by regulators and should ultimately be available to clinicians and the patients making decisions about the use of these treatments.
View all comments by Michael WeinerCyceron
This issue of accelerated brain-volume loss after anti-amyloid treatment is a relatively old question that has likely not been given the interest it deserves. The problem with volume/atrophy is that it is a nonspecific marker that might reflect many different underlying processes, including both "negative" (pathological) and "positive" (less inflammation or edema, fewer pathological molecules, etc.), even if greater volume usually correlates with greater cognitive performance.
In 2010, we showed that amyloid-positive asymptomatic elderly had greater brain volume in the anterior temporal cortex and hippocampus compared to amyloid-negative asymptomatic elderly (Chételat et al., 2010). In other words, more amyloid equated to more volume, which is akin to less amyloid equating to less volume found in these anti-amyloid trials.
This amyloid/volume correlation could reflect neuroinflammation due to amyloid, or the presence of amyloid itself, although I agree with the authors that the latter would not be expected to result in such a degree of volume change, but only have a very minor impact. However, we concluded that the volume correlation was rather likely to reflect brain reserve—those with higher temporal/hippocampal volume tolerate amyloid better than those with lower volume, who would exhibit subjective or cognitive decline such as they would be in another clinical group. Indeed, we found a positive link between volume in these regions and episodic memory performance.
Here, we lack the data to further understand the etiology/meaning/underlying mechanisms and consequences of this greater atrophy rate in trials. For sure, this would need further investigation. The question is not simple, since some of those drugs also tend to show better cognitive/clinical outcome, and even lower tau deposition, which are known to be associated with higher brain volume.
Ultimately, we need to know more about i) the factors that influence; ii) the correlates, and iii) the consequences of this anti-amyloid drug-related accelerated atrophy before any conclusion can be made.
References:
Chételat G, Villemagne VL, Pike KE, Baron JC, Bourgeat P, Jones G, Faux NG, Ellis KA, Salvado O, Szoeke C, Martins RN, Ames D, Masters CL, Rowe CC. Larger temporal volume in elderly with high versus low beta-amyloid deposition. Brain. 2010 Nov;133(11):3349-58. PubMed.
View all comments by Gael ChetelatUniversity of Nevada Las Vegas
We do not understand the biology of this observation completely. The reduction in brain volume is seen with all the monoclonal antibodies and was seen with a BACE inhibitor and with valproate. I would not call it atrophy, since atrophy implies cell loss, and we do not know that to be the case.
We spent a lot of time thinking about this in the verubecestat trial and concluded that the volume loss was regional, nonprogressive, and corresponded to the regions of high amyloid content (Sur et al., 2020). This suggests that amyloid dynamics may be contributing importantly to this brain volume reduction. It is an observation worthy of further attention, but the available data suggest that it is not an indication of brain injury.
References:
Sur C, Kost J, Scott D, Adamczuk K, Fox NC, Cummings JL, Tariot PN, Aisen PS, Vellas B, Voss T, Mahoney E, Mukai Y, Kennedy ME, Lines C, Michelson D, Egan MF. BACE inhibition causes rapid, regional, and non-progressive volume reduction in Alzheimer's disease brain. Brain. 2020 Dec 1;143(12):3816-3826. PubMed.
View all comments by Jeffrey CummingsLund University
I think this phenomenon, that anti-Aβ immunotherapies consistently increase the ventricular volumes, is intriguing and should receive more attention. We do not yet have a satisfactory explanation for it. To learn more, it would be great if the companies with access to the individual-level data from anti-Aβ immunotherapies trials would study this phenomenon in greater depth and share their experience with the AD community.
For example, one could study at an individual level whether 1) more efficient removal of Aβ plaques (as measured with amyloid PET) is associated with ventricular enlargement, 2) ARIA frequency and severity are associated with ventricular enlargement, 3) changes in key biomarkers, such plasma p-tau, GFAP, and NfL are associated with ventricular enlargement, and finally, whether 4) change in functional outcomes are associated with ventricular enlargement.
That said, my current hypothesis is that Aβ plaques result in neuroinflammation in AD with consequent mild swelling of the brain, and when the plaques are removed such brain changes are diminished and the brain parenchyma is somewhat reduced. This hypothesis is not proven, but we have recently shown that changes in diffusion MRI of cortical regions are associated with Aβ-plaque pathology and that this association is partly mediated by changes (GFAP) in reactive astrocytes (Spotorno et al., 2022). We hope that companies will analyze this type of diffusion MRI sequences in their clinical trials to unravel whether normalization in cortical diffusion metrics is associated with Aβ-plaque removal and whether this change might mediate the association between Aβ plaque removal and ventricular enlargement.
References:
Spotorno N, Strandberg O, Vis G, Stomrud E, Nilsson M, Hansson O. Measures of cortical microstructure are linked to amyloid pathology in Alzheimer's disease. Brain. 2022 Sep 21; PubMed.
View all comments by Oskar HanssonDementia Reserch Center
It is good to see consideration of the interesting and unanswered question of why potential Alzheimer’s-disease-modifying therapies have shown increased brain volume reductions relative to placebo. First seen in the AN1792 vaccination trial, and repeatedly since with monoclonal antibodies, this appears the norm rather than the exception for anti-amyloid immunotherapies. But this is not limited to immunotherapies—for example, treatment-related brain volume reduction is also seen with verubecestat (β-secretase inhibition) and with resveratrol (SIRT1 activation).
However, where I think more caution is needed is in equating “excess” volume loss to “excess” cognitive loss—which the evidence just does not support. Nor should volume change be extrapolated to imply “increased rates of atrophy,” or indeed, increased neurodegeneration. We know, for example, that with verubecestat the volume change occurred rapidly (within 13 weeks) and was not progressive. It certainly does not seem justified to take a volume change, translate that into a rate of atrophy and then project it forward to imply (as in figure 6) that progression from MCI to AD would be more rapid.
There are many areas in neurology where we do not equate treatment-related volume change with increased neurodegeneration; for example, when steroids or mannitol are given to reduce brain edema, they also reduce brain volumes. In multiple sclerosis trials, “pseudo-atrophy” did not imply a worse outcome for patients. In individuals undergoing hemodialysis, reversible volume changes of ~3 percent of total brain volume can occur over hours.
The correlations across therapies (at least the immunotherapies) between increased volume loss, amyloid removal, and ARIA are important, but there are lots of cross-confounds. Patient-level data would help disentangle this. We do not understand what causes these treatment-related brain volume changes in AD, and this mandates further exploration. Mechanisms are likely to differ between therapies, but we should be careful in our current interpretation. There is a danger if these issues are taken up by the press in an alarmist way. Patients have enough reasons for anxiety already.
References:
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M. Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease. Neurology. 2005 May 10;64(9):1563-72. PubMed.
Turner RS, Thomas RG, Craft S, van Dyck CH, Mintzer J, Reynolds BA, Brewer JB, Rissman RA, Raman R, Aisen PS, Alzheimer's Disease Cooperative Study. A randomized, double-blind, placebo-controlled trial of resveratrol for Alzheimer disease. Neurology. 2015 Oct 20;85(16):1383-91. Epub 2015 Sep 11 PubMed.
Sur C, Kost J, Scott D, Adamczuk K, Fox NC, Cummings JL, Tariot PN, Aisen PS, Vellas B, Voss T, Mahoney E, Mukai Y, Kennedy ME, Lines C, Michelson D, Egan MF. BACE inhibition causes rapid, regional, and non-progressive volume reduction in Alzheimer's disease brain. Brain. 2020 Dec 1;143(12):3816-3826. PubMed.
Walters RJ, Fox NC, Crum WR, Taube D, Thomas DJ. Haemodialysis and cerebral oedema. Nephron. 2001 Feb;87(2):143-7. PubMed.
Zivadinov R, Reder AT, Filippi M, Minagar A, Stüve O, Lassmann H, Racke MK, Dwyer MG, Frohman EM, Khan O. Mechanisms of action of disease-modifying agents and brain volume changes in multiple sclerosis. Neurology. 2008 Jul 8;71(2):136-44. PubMed.
View all comments by Nick FoxWashington University School of Medicine
This report provides a sobering reassessment of a previously observed but unexplained consequence of treatment with anti-Alzheimer drugs that target Aβ: many of these drugs are associated with cerebral volume loss.
The clinical correlates of this volume loss are yet to be firmly established, but this report suggests at least one adverse effect in that there is a strong correlation between ventricular volume and the monoclonal antibodies that are associated with amyloid-related imaging abnormalities (ARIA). As treatment of persons with Alzheimer's dementia using anti-amyloid drugs becomes possible in clinical practice, the potential for cerebral volume loss should be considered with ARIA as a potential adverse event.
View all comments by John MorrisFlorey Institute of Neuroscience and Mental Health
I’d like to thank the Alzforum readership for their engagement in this research. Our reason for undertaking this work was to raise the profile of this long-reported finding, and hope this level of interest leads to more research. We believe these findings warrant concern, but cannot make firm conclusions about whether this reflects accelerated neurodegeneration or something benign. For several years I have been highlighting this issue as one that requires prioritization as we enter the therapeutic era (Ayton, 2022; Ayton, 2021). There is much data that has been collected that would help us illuminate this finding, but the data hasn’t been explored or released. We are left wondering, and there are speculations in both directions. The available evidence we have already provides some insight.
In the only study that has related drug-induced volumetric changes to clinical outcomes, volume changes caused by verubecestat were associated with worse clinical outcomes (Sur et al., 2020). The only paper that reported an association between drug-induced volume changes and NfL reported that volume changes caused by donanemab were correlated with plasma NfL (Pontecorvo et al., 2022). I regard both of these as preliminary evidence, but the available evidence suggests that volume changes may be detrimental and underscores why we should be conservative. Especially when dealing with the possibility of brain damage, we should be cautious in our interpretation and gather more data before dismissing this finding. Reinterpretation of the volume data requires evidence to support this.
I’m glad to read a consensus that plaque removal cannot account for the volume changes induced by anti-amyloid drugs.
A putative reduction in inflammation caused by anti-amyloid therapies is another potential explanation for the change in volume. This, too, is a speculation that requires evidence. The evidence we currently have argues against this possibility. ARIA is MRI-detectable inflammation, and we show that there is a strong correlation between the frequency of ARIA and the extent of ventricular enlargement. If brain volumes were increasing due to inflammation, we would see the opposite result—ARIA would be associated with increased brain volume. By no means is this the end of the argument, but the only available evidence we currently have regarding inflammation is correlated strongly in the wrong direction. So we should not currently hold this explanation with much certainty. The reverse possibility—that inflammation causes degeneration, is a concept grounded in a large body of literature and is the conservative interpretation that fits the data. Until we have evidence to direct us otherwise, we should be cautious in attributing volume changes to reduced neuroinflammation.
The same can be said for edema due to CSF drainage. It has been put to me that anti-amyloid drugs will cause less water retention in the brain. Yet, they increase lateral ventricular volume.
Why should we care about these results if anti-amyloid drugs improve cognition? The clinical significance of the cognitive findings remains contentious. Brain volumetrics is supportive and objective evidence of disease progression that would have been used to support the clinical result if the finding were in the opposite direction. When the clinical findings are unclear, these secondary outcomes help provide context to the primary result. The volumetric findings do not provide additional support of efficacy.
Additional questions emerge from volumetric data that may guide our questions on efficacy:
This is all speculation. We need more data and analysis.
A final thought—as has been pointed out in the Alzforum article, this finding is not new. We have observed this in scores of trials for more than a decade. Yet we know so little about it, and hardly any exploration into the available data has been performed. It should give us pause that we cannot provide basic explanations into this phenomenon as we now enter an era where anti-amyloid drugs have approval. We should ask, why as a field have we not demanded more questions of these data as they have been published in numerous journals over many years? I hope that now we will see these data more intensively investigated. I sincerely hope that the finding is benign and does not risk patient safety.
References:
Ayton S. Ventricular enlargement caused by aducanumab. Nat Rev Neurol. 2022 Jul;18(7):383-384. PubMed.
Ayton S. Brain volume loss due to donanemab. Eur J Neurol. 2021 Sep;28(9):e67-e68. Epub 2021 Jul 16 PubMed.
Sur C, Kost J, Scott D, Adamczuk K, Fox NC, Cummings JL, Tariot PN, Aisen PS, Vellas B, Voss T, Mahoney E, Mukai Y, Kennedy ME, Lines C, Michelson D, Egan MF. BACE inhibition causes rapid, regional, and non-progressive volume reduction in Alzheimer's disease brain. Brain. 2020 Dec 1;143(12):3816-3826. PubMed.
Pontecorvo MJ, Lu M, Burnham SC, Schade AE, Dage JL, Shcherbinin S, Collins EC, Sims JR, Mintun MA. Association of Donanemab Treatment With Exploratory Plasma Biomarkers in Early Symptomatic Alzheimer Disease: A Secondary Analysis of the TRAILBLAZER-ALZ Randomized Clinical Trial. JAMA Neurol. 2022 Dec 1;79(12):1250-1259. PubMed.
Novak G, Fox N, Clegg S, Nielsen C, Einstein S, Lu Y, Tudor IC, Gregg K, Di J, Collins P, Wyman BT, Yuen E, Grundman M, Brashear HR, Liu E. Changes in Brain Volume with Bapineuzumab in Mild to Moderate Alzheimer's Disease. J Alzheimers Dis. 2016;49(4):1123-34. PubMed.
van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, Kanekiyo M, Li D, Reyderman L, Cohen S, Froelich L, Katayama S, Sabbagh M, Vellas B, Watson D, Dhadda S, Irizarry M, Kramer LD, Iwatsubo T. Lecanemab in Early Alzheimer's Disease. N Engl J Med. 2023 Jan 5;388(1):9-21. Epub 2022 Nov 29 PubMed.
Miller TM, Cudkowicz ME, Genge A, Shaw PJ, Sobue G, Bucelli RC, Chiò A, Van Damme P, Ludolph AC, Glass JD, Andrews JA, Babu S, Benatar M, McDermott CJ, Cochrane T, Chary S, Chew S, Zhu H, Wu F, Nestorov I, Graham D, Sun P, McNeill M, Fanning L, Ferguson TA, Fradette S, VALOR and OLE Working Group. Trial of Antisense Oligonucleotide Tofersen for SOD1 ALS. N Engl J Med. 2022 Sep 22;387(12):1099-1110. PubMed.
View all comments by Scott AytonBarcelona Supercomputing Center
Hospital of Sant Pau
The apparent atrophy related to anti-amyloid treatments has been a matter of discussion since the first active vaccine trial (Fox et al., 2005), and it's crucial to better understand the etiology of such phenomena to guarantee the safety of our patients. While historically atrophy has always been associated to neurodegeneration in the AD field, this may not be the case for other neurological disorders, as previously noted by Dr. Fox. Thus, the reported atrophy in the reviewed studies of this meta-analysis might arise from a beneficial effect of removing amyloid plaques rather than neurodegeneration.
We agree with Alves and collaborators that this atrophy cannot be solely due to the removal of amyloid. Instead, we have previously proposed and validated a model in which early amyloid accumulation leads to neuroinflammation that results in increased thickness. In fact, using in vivo astrocytosis PET, we previously reported a positive association between inflammation and cortical thickening in a cohort of autosomal dominant AD (Vilaplana et al., 2020). Moreover, cortical thickening has been observed in other familial AD cohorts such as DIAN (Montal et al., 2021), or the Colombian cohort (Fox-Fuller et al., 2021) and in sporadic AD (Fortea et al., 2014). Multimodal analyses using DWI data suggest that this increased thickness arises from the increment of cellular bodies, which might be caused by neuronal hypertrophy, morphology expansion in reactive astrocytes and microglia and/or glia recruitment (Montal et al., 2018). Therefore, we do not imply that the increase in thickness is caused by amyloid per se but rather by an amyloid-related inflammation process.
We hypothesize that in symptomatic stages of AD, despite the observed atrophy, this amyloid-related inflammation and consequent relative thickening persists, as previously reported in studies using in-vivo microglia PET (Femminella et al., 2019). As a result, the removal of amyloid and partial resolution of the inflammation will result in brain atrophy, despite not being detrimental.
It is crucial to study this putative effect using clinical trial data and assess if this hypothesis is confirmed at the individual-patient level. We urge the different companies behind these trials to share their data with the community to allow an in-depth comprehension of this apparent detrimental effect in anti-amyloid trials. It is important to continue research on the apparent atrophy related to anti-amyloid treatments to better understand its etiology and to ensure the safety and efficacy of these treatments.
References:
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M. Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease. Neurology. 2005 May 10;64(9):1563-72. PubMed.
Vilaplana E, Rodriguez-Vieitez E, Ferreira D, Montal V, Almkvist O, Wall A, Lleó A, Westman E, Graff C, Fortea J, Nordberg A. Cortical microstructural correlates of astrocytosis in autosomal-dominant Alzheimer disease. Neurology. 2020 May 12;94(19):e2026-e2036. Epub 2020 Apr 14 PubMed.
Montal V, Vilaplana E, Pegueroles J, Bejanin A, Alcolea D, Carmona-Iragui M, Clarimón J, Levin J, Cruchaga C, Graff-Radford NR, Noble JM, Lee JH, Allegri R, Karch CM, Laske C, Schofield PR, Salloway S, Ances B, Benzinger T, McDale E, Bateman R, Blesa R, Sánchez-Valle R, Lleó A, Fortea J, Dominantly Inherited Alzheimer Network (DIAN). Biphasic cortical macro- and microstructural changes in autosomal dominant Alzheimer's disease. Alzheimers Dement. 2021 Apr;17(4):618-628. Epub 2020 Nov 16 PubMed.
Fox-Fuller JT, Torrico-Teave H, d'Oleire Uquillas F, Chen K, Su Y, Chen Y, Brickhouse M, Sanchez JS, Aguero C, Jacobs HI, Hampton O, Guzmán-Vélez E, Vila-Castelar C, Aguirre-Acevedo DC, Baena A, Artola A, Martinez J, Pluim CF, Alvarez S, Ochoa-Escudero M, Reiman EM, Sperling RA, Lopera F, Johnson KA, Dickerson BC, Quiroz YT. Cortical thickness across the lifespan in a Colombian cohort with autosomal-dominant Alzheimer's disease: A cross-sectional study. Alzheimers Dement (Amst). 2021;13(1):e12233. Epub 2021 Sep 14 PubMed.
Fortea J, Vilaplana E, Alcolea D, Carmona-Iragui M, Sánchez-Saudinos MB, Sala I, Antón-Aguirre S, González S, Medrano S, Pegueroles J, Morenas E, Clarimón J, Blesa R, Lleó A, Alzheimer's Disease Neuroimaging Initiative. Cerebrospinal fluid β-amyloid and phospho-tau biomarker interactions affecting brain structure in preclinical Alzheimer disease. Ann Neurol. 2014 Aug;76(2):223-30. Epub 2014 Jun 13 PubMed.
Montal V, Vilaplana E, Alcolea D, Pegueroles J, Pasternak O, González-Ortiz S, Clarimón J, Carmona-Iragui M, Illán-Gala I, Morenas-Rodríguez E, Ribosa-Nogué R, Sala I, Sánchez-Saudinós MB, García-Sebastian M, Villanúa J, Izagirre A, Estanga A, Ecay-Torres M, Iriondo A, Clerigue M, Tainta M, Pozueta A, González A, Martínez-Heras E, Llufriu S, Blesa R, Sanchez-Juan P, Martínez-Lage P, Lleó A, Fortea J. Cortical microstructural changes along the Alzheimer's disease continuum. Alzheimers Dement. 2018 Mar;14(3):340-351. Epub 2017 Oct 31 PubMed.
Femminella GD, Dani M, Wood M, Fan Z, Calsolaro V, Atkinson R, Edginton T, Hinz R, Brooks DJ, Edison P. Microglial activation in early Alzheimer trajectory is associated with higher gray matter volume. Neurology. 2019 Mar 19;92(12):e1331-e1343. Epub 2019 Feb 22 PubMed.
View all comments by Juan ForteaUniversity of Sydney
Biological mechanisms in addition to the more-studied proteopathies can hypothetically lead to atrophy, and could be related here. Some clues from literature:
Any of the above mechanisms, a combination of them, or other related atrophy causes would prevent cognitive improvement. Neuroenhancement and neuroplasticity, to my knowledge, are possible in elderly but obviously less extensive compared to younger adults who have a greater reserve of neural stem cells. The path forward to determining the underlying biological mechanisms (in this case behind atrophy) likely lies in the disproving of ideas, as Planck, Popper, and others have established in the past.
References:
Marchi NA, Ramponi C, Hirotsu C, Haba-Rubio J, Lutti A, Preisig M, Marques-Vidal P, Vollenweider P, Kherif F, Heinzer R, Draganski B. Mean Oxygen Saturation during Sleep Is Related to Specific Brain Atrophy Pattern. Ann Neurol. 2020 Jun;87(6):921-930. Epub 2020 Apr 20 PubMed.
Manoharan S, Guillemin GJ, Abiramasundari RS, Essa MM, Akbar M, Akbar MD. The Role of Reactive Oxygen Species in the Pathogenesis of Alzheimer's Disease, Parkinson's Disease, and Huntington's Disease: A Mini Review. Oxid Med Cell Longev. 2016;2016:8590578. Epub 2016 Dec 27 PubMed.
Geloso MC, D'Ambrosi N. Microglial Pruning: Relevance for Synaptic Dysfunction in Multiple Sclerosis and Related Experimental Models. Cells. 2021 Mar 20;10(3) PubMed.
Korbo L. Glial cell loss in the hippocampus of alcoholics. Alcohol Clin Exp Res. 1999 Jan;23(1):164-8. PubMed.
View all comments by Phillip JanowiczSorbonne University - APHP - Pitié-Salpêtrière Hospital
The paper by Alves and colleagues is methodologically sound. It brings a piece of high-level evidence and highlights this poorly investigated but universal finding of high-clearance anti-amyloid immunotherapies: ventricular enlargement +/- gray matter loss. The debate and numerous comments underline the relevance and uncertainty of these findings, and the need for further investigation. I align with my mentor’s (Gaël Chételat) opinion here.
In addition to the arguments already detailed here, I would like to emphasize three elements:
Overall, the current evidence supporting that ventricular enlargement +/- gray matter loss in these trials reflects neurodegeneration is meager. On the other hand, the quality of evidence supporting the opposite is, in my opinion, more convincing. Both remain insufficient. Like Alves et al. and almost all commentators here, I advocate for further transparency and disclosure of available MRI data in the current high-clearance anti-amyloid immunotherapies trials: especially the correlation between cognitive outcomes and MRI metrics during the core phase of the trials and long-term OLE data cognitive outcomes and MRI metrics.
References:
Boche D, Denham N, Holmes C, Nicoll JA. Neuropathology after active Abeta42 immunotherapy: implications for Alzheimer's disease pathogenesis. Acta Neuropathol. 2010 Sep;120(3):369-84. PubMed.
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M. Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease. Neurology. 2005 May 10;64(9):1563-72. PubMed.
Holmes C, Boche D, Wilkinson D, Yadegarfar G, Hopkins V, Bayer A, Jones RW, Bullock R, Love S, Neal JW, Zotova E, Nicoll JA. Long-term effects of Abeta42 immunisation in Alzheimer's disease: follow-up of a randomised, placebo-controlled phase I trial. Lancet. 2008 Jul 19;372(9634):216-23. PubMed.
Hua X, Hibar DP, Lee S, Toga AW, Jack CR, Weiner MW, Thompson PM, . Sex and age differences in atrophic rates: an ADNI study with n=1368 MRI scans. Neurobiol Aging. 2010 Aug;31(8):1463-80. PubMed.
Nicoll JA, Buckland GR, Harrison CH, Page A, Harris S, Love S, Neal JW, Holmes C, Boche D. Persistent neuropathological effects 14 years following amyloid-β immunization in Alzheimer's disease. Brain. 2019 Jul 1;142(7):2113-2126. PubMed.
Pontecorvo MJ, Lu M, Burnham SC, Schade AE, Dage JL, Shcherbinin S, Collins EC, Sims JR, Mintun MA. Association of Donanemab Treatment With Exploratory Plasma Biomarkers in Early Symptomatic Alzheimer Disease: A Secondary Analysis of the TRAILBLAZER-ALZ Randomized Clinical Trial. JAMA Neurol. 2022 Dec 1;79(12):1250-1259. PubMed.
View all comments by Nicolas VillainRoche/Genentech
Roche
Genentech
Clinicians have often associated focal or generalized cerebral atrophy seen on imaging studies with loss of brain tissue, or loss of specific cell types, due to injury or degeneration. It is understandable that apparent loss of brain tissue leads to concern, even though cerebral atrophy can be due to reversible causes, such as anorexia nervosa (Heinz et al., 1977), Cushing’s syndrome (Heinz et al., 1977), ACTH/Corticosteroid therapy (Lagenstein et al., 1979; Bentson et al., Apparent cerebral atrophy on computed tomography following steroid use. Presented at the International Symposium on Computer Assisted Tomography in Nontumoral Diseases of the Brain, Spinal Cord, and Eye. National Institutes of Health, Bethesda, Maryland, 11-15 October 1976), alcoholism (Carlen et al., 1978), and multiple sclerosis (De Stafano and Arnold, 2015), and despite the general acceptance that some loss of brain volume is expected due to aging.
It is well-documented that several types of anti-amyloid therapies, from active vaccines to large and small molecules, have also been associated with more loss of brain volume and greater ventricular expansion in treated patients compared to randomized controls. Since the earliest observations of this phenomenon (Fox et al., Apparent cerebral atrophy on computed tomography following steroid use. Presented at the International Symposium on Computer Assisted Tomography in Nontumoral Diseases of the Brain, Spinal Cord, and Eye. National Institutes of Health, Bethesda, Maryland, 11-15 October 1976.), there has been no clear association between clinical status and this treatment-related brain atrophy.
This systematic review and meta-analysis by Alves et al. helps to quantify the extent of this observation in patients who: took various types of anti-amyloid drugs, had at least one biomarker of target engagement, and at least one MRI sufficient to evaluate at least one brain region. It should be noted that their review included only 31 of the 145 studies identified, so additional insights may still be derived from the excluded studies. They attempted to relate atrophy to ARIA at the group level (in studies in which ARIA occurred) using Pearson’s regressions and modeled how much “worse” MCI patients were due to treatment, with the unsupported assumption that rate of atrophy was equal to disease progression.
We have reviewed vMRI data from the recently-reported Graduate studies, and concluded that the observed differences in vMRI measures in gantenerumab-treated patients versus placebo-treated patients, which were similar to those seen with other anti-Aβ monoclonal antibodies (MABs), could be due to a beneficial effect, such as reduced inflammation associated with plaque removal, or changes in CSF dynamics (Barkhoff and Knopman, 2023), and may not be due to accelerated neurodegeneration (data in preparation). Given the clinical benefits seen with some MABs (van Dyck et al., 2023; Mintun et al., 2021), it is difficult to conclude that atrophy seen with those treatments represents diffuse neuronal loss. Further, in our Graduate studies, the reduction of NFL in CSF seen with gantenerumab treatment would argue against accelerated neurodegeneration (Bittner et al., GRADUATE I and II results: Effect of subcutaneous gantenerumab on fluid biomarkers of AD pathology and neurodegeneration. Presented atvADPD, Gothenberg, Sweden, March 30, 2023).
In conclusion, we do not know what causes atrophy in reversible conditions, normal aging, or with various anti-Aβ amyloid treatments. The reasons may be diverse and should not be assumed to be related to brain injury or worsening of the neuropathologic process of AD. In our view, the weight of the evidence to date would argue against hypotheses of detriment.
References:
Heinz ER, Martinez J, Haenggeli A. Reversibility of cerebral atrophy in anorexia nervosa and Cushing's syndrome. J Comput Assist Tomogr. 1977 Oct;1(4):415-8. PubMed.
Lagenstein I, Willig RP, Kühne D. Reversible cerebral atrophy caused by corticotrophin. Lancet. 1979 Jun 9;1(8128):1246-7. PubMed.
Carlen PL, Wortzman G, Holgate RC, Wilkinson DA, Rankin JC. Reversible cerebral atrophy in recently abstinent chronic alcoholics measured by computed tomography scans. Science. 1978 Jun 2;200(4345):1076-8. PubMed.
De Stefano N, Arnold DL. Towards a better understanding of pseudoatrophy in the brain of multiple sclerosis patients. Mult Scler. 2015 May;21(6):675-6. Epub 2015 Jan 26 PubMed.
Barkhof F, Knopman DS. Brain Shrinkage in Anti-β-Amyloid Alzheimer Trials: Neurodegeneration or Pseudoatrophy?. Neurology. 2023 May 16;100(20):941-942. Epub 2023 Mar 27 PubMed.
van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, Kanekiyo M, Li D, Reyderman L, Cohen S, Froelich L, Katayama S, Sabbagh M, Vellas B, Watson D, Dhadda S, Irizarry M, Kramer LD, Iwatsubo T. Lecanemab in Early Alzheimer's Disease. N Engl J Med. 2023 Jan 5;388(1):9-21. Epub 2022 Nov 29 PubMed.
Mintun MA, Lo AC, Duggan Evans C, Wessels AM, Ardayfio PA, Andersen SW, Shcherbinin S, Sparks J, Sims JR, Brys M, Apostolova LG, Salloway SP, Skovronsky DM. Donanemab in Early Alzheimer's Disease. N Engl J Med. 2021 May 6;384(18):1691-1704. Epub 2021 Mar 13 PubMed.
View all comments by Robby WeimerUniversity of Southampton, Faculty of Medicine
Université de Paris Cognitive Neurology Center, GHU APHP Nord Lariboisiere Fernand-Widal Hospital
University of Paris Diderot
University of Southampton
With currently available technology, it seems unlikely that the pathophysiology associated with Aβ immunotherapy-related atrophy can be understood without studying human brain tissue. That there has not been systematic postmortem follow-up of patients in recent passive immunotherapy trials represents a missed opportunity. We gathered a postmortem cohort of 22 subjects who had received AN1792, demonstrating removal of Aβ plaques and a relative reduction in both tau and microglial activation. Remarkably, these effects persisted as long as 14 years after active immunization.
Imaging analyses reported that immunotherapies, whether or not associated with amyloid removal and independently of the protocol (active versus passive), have led to either a reduction in whole-brain volume or an increased ventricular volume in treated participants compared to placebo (AN1792 (Fox et al., 2005); Bapineuzumab (Novak et al., 2016), Aducanumab (Budd Haeberlein et al., 2022), Donanemab (Mintun et al., 2021), and now Lecanemab (Swanson et al., 2021; van Dyck et al., 2023). The biological basis for this observation is unknown; however, interestingly, those imaging results are concordant with the analysis describing brain atrophy in our postmortem cohort of immunized AD (iAD) patients compared with non-immunized AD (cAD) patients (Paquet et al., 2015; Serrano-Pozo et al., 2010).
Brain atrophy was defined by increased cerebral cortical neuropil degeneration (cAD 48.6 percent versus iAD 53.3 percent; p=0.013), increased neuronal loss (cAD 73.7 versus iAD 66.8 neurons/field; p=0.036) and inter-neuronal distance (cAD 695 µm vs. iAD 703.6 µm; p=0.006). However, while atrophy in non-immunized AD has been consistently associated with progression of the disease and the severity of pathological changes rather than slowing of symptoms, our neuropathological studies did not observe an association between the number of neurons and the cognitive evolution nor the side effects. Of note, the brain atrophy was not related to markers of neurodegeneration (tangles, Aβ42 load, pPKR), and conversely, we observed improved health of the residual neurons with less neuritic curvature and the presence of fewer pro-apoptotic neurons in the immunized brains Paquet et al., 2015; Serrano-Pozo et al., 2010; Paquet et al., 2017).
Now that we are at the exciting stage of licensing of Aβ immunotherapeutic agents for clinical use, we strongly encourage study of the brains of participants of trials of passive immunotherapy in order to throw light on the important issue of atrophy and whether it is good, bad, or indifferent.
References:
Fox NC, Black RS, Gilman S, Rossor MN, Griffith SG, Jenkins L, Koller M. Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease. Neurology. 2005 May 10;64(9):1563-72. PubMed.
Novak G, Fox N, Clegg S, Nielsen C, Einstein S, Lu Y, Tudor IC, Gregg K, Di J, Collins P, Wyman BT, Yuen E, Grundman M, Brashear HR, Liu E. Changes in Brain Volume with Bapineuzumab in Mild to Moderate Alzheimer's Disease. J Alzheimers Dis. 2016;49(4):1123-34. PubMed.
Budd Haeberlein S, Aisen PS, Barkhof F, Chalkias S, Chen T, Cohen S, Dent G, Hansson O, Harrison K, von Hehn C, Iwatsubo T, Mallinckrodt C, Mummery CJ, Muralidharan KK, Nestorov I, Nisenbaum L, Rajagovindan R, Skordos L, Tian Y, van Dyck CH, Vellas B, Wu S, Zhu Y, Sandrock A. Two Randomized Phase 3 Studies of Aducanumab in Early Alzheimer's Disease. J Prev Alzheimers Dis. 2022;9(2):197-210. PubMed.
Mintun MA, Lo AC, Duggan Evans C, Wessels AM, Ardayfio PA, Andersen SW, Shcherbinin S, Sparks J, Sims JR, Brys M, Apostolova LG, Salloway SP, Skovronsky DM. Donanemab in Early Alzheimer's Disease. N Engl J Med. 2021 May 6;384(18):1691-1704. Epub 2021 Mar 13 PubMed.
Swanson CJ, Zhang Y, Dhadda S, Wang J, Kaplow J, Lai RY, Lannfelt L, Bradley H, Rabe M, Koyama A, Reyderman L, Berry DA, Berry S, Gordon R, Kramer LD, Cummings JL. A randomized, double-blind, phase 2b proof-of-concept clinical trial in early Alzheimer's disease with lecanemab, an anti-Aβ protofibril antibody. Alzheimers Res Ther. 2021 Apr 17;13(1):80. PubMed. Correction.
van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, Kanekiyo M, Li D, Reyderman L, Cohen S, Froelich L, Katayama S, Sabbagh M, Vellas B, Watson D, Dhadda S, Irizarry M, Kramer LD, Iwatsubo T. Lecanemab in Early Alzheimer's Disease. N Engl J Med. 2023 Jan 5;388(1):9-21. Epub 2022 Nov 29 PubMed.
Paquet C, Amin J, Mouton-Liger F, Nasser M, Love S, Gray F, Pickering RM, Nicoll JA, Holmes C, Hugon J, Boche D. Effect of active Aβ immunotherapy on neurons in human Alzheimer's disease. J Pathol. 2015 Apr;235(5):721-30. Epub 2015 Jan 7 PubMed.
Serrano-Pozo A, William CM, Ferrer I, Uro-Coste E, Delisle MB, Maurage CA, Hock C, Nitsch RM, Masliah E, Growdon JH, Frosch MP, Hyman BT. Beneficial effect of human anti-amyloid-beta active immunization on neurite morphology and tau pathology. Brain. 2010 May;133(Pt 5):1312-27. PubMed.
Paquet C, Nicoll JA, Love S, Mouton-Liger F, Holmes C, Hugon J, Boche D. Downregulated apoptosis and autophagy after anti-Aβ immunotherapy in Alzheimer's disease. Brain Pathol. 2017 Oct 13; PubMed.
View all comments by James NicollThis study shows that frequently there is not adequate analysis of prior clinical data, an exercise helpful in designing new therapeutic agents and clinical studies. A pertinent case is the AN1792 vaccine for AD, which used the potent pro-inflammatory immune modulator QS-21 despite lack of understanding of its pharmacological properties. The situation was worsened by changes made in Phase 2 without a rigorous scientific analysis. For some time it was assumed that despite AN1792’s serious side effects, it had some beneficial effects but, as reported by von Bernhardi (2010), those receiving only QS-21 as a placebo had an unusually high rate of cognitive decline, 6 to 7 points loss in the MMSE, which exceeds the average 3.5 to 4 points for AD patients. This was likely because QS-21 induces the damaging pro-inflammatory immunity Th17, disrupting the immune homeostasis needed to maintain an anti-inflammatory Th2 immunity and prevent aggravating the disease. Since many AD vaccines used similar immune modulators or adjuvants, one can assume that they elicited the damaging pro-inflammatory immunities Th1 and/or Th17. These effects complicate interpretation of the clinical results.
However, the situation may be different with passive immunotherapy. Aβ insolubilized as plaque is solubilized by IgGs that bind this protein, irrespective of their therapeutic value. By binding to Aβ and forming complexes, the IgGs weaken ionic-hydrophobic interactions that stabilize plaque. Once this occurs, the Aβ-IgG complexes become soluble. Actually, IgGs due to their carbohydrate content are quite soluble in an aqueous milieu. Thus, the degree of plaque solubilization should correlate with the dose of anti-Aβ IgG. Hence, the higher the IgG dose, the less plaque would be left, which may explain the reported brain shrinkage.
This solubilization process is rather like the old prozone effect, where an antigen in the presence of specific antibodies aggregates and precipitates. But upon addition of an excess of those antibodies, the interactions among adjacent antigen molecules bridged by IgGs are disrupted. The process leads to formation of soluble antigen-IgG complexes, and a concomitant loss in the amount of antigen/IgG precipitate. Solubilization of Aβ plaque by antibodies against it was studied by J.Y. Wang et al. (Liu et al., 2015). They called it the “dust-raising effect.” The study concluded that the release from plaque of soluble toxic Aβ oligomers (AβOs), the accepted causative agents of this disease (Cline et al., 2018; Walsh and Selkoe, 2020), increases neurotoxicity. However, those results were reported in 2015, before the availability of antibodies like lecanemab, capable of neutralizing AβOs’ cytotoxicity. Thus, it will be of interest to compare the effects of lecanemab, which has been found to slow Alzheimer’s disease, with brain shrinkage. While lecanemab will solubilize plaque, the released soluble AβOs must be inactivated by this monoclonal antibody, which should minimize the likely neurological damage caused by the solubilized cytotoxic AβOs derived from plaque.
As the authors and many commentators indicate, it will be helpful to compare the results from the clinical trials with the pharmacological effects of the therapeutic agents used. This task will require collaboration of the different companies that ran those studies, to obtain the relevant data. This should not present problems, since many companies discontinued development of the drugs after their clinical studies failed.
References:
von Bernhardi R. Immunotherapy in Alzheimer's disease: where do we stand? Where should we go?. J Alzheimers Dis. 2010;19(2):405-21. PubMed.
Liu YH, Bu XL, Liang CR, Wang YR, Zhang T, Jiao SS, Zeng F, Yao XQ, Zhou HD, Deng J, Wang YJ. An N-terminal antibody promotes the transformation of amyloid fibrils into oligomers and enhances the neurotoxicity of amyloid-beta: the dust-raising effect. J Neuroinflammation. 2015 Aug 28;12:153. PubMed.
Cline EN, Bicca MA, Viola KL, Klein WL. The Amyloid-β Oligomer Hypothesis: Beginning of the Third Decade. J Alzheimers Dis. 2018;64(s1):S567-S610. PubMed.
Walsh DM, Selkoe DJ. Amyloid β-protein and beyond: the path forward in Alzheimer's disease. Curr Opin Neurobiol. 2020 Apr;61:116-124. PubMed.
View all comments by Dante MarcianiUniversity of Arkansas for Medical Sciences
I am surprised to see no discussion here of T2 weighting. I have no expertise in imaging, but wouldn't the T2/T1 ratio help to determine how much of the immunotherapy-related change is a loss of edema versus macromolecular structure?
View all comments by Steve BargerMake a Comment
To make a comment you must login or register.