Depressive Pseudodementia with Reversible AD-like Brain Hypometabolism: A Case Report and a Review of the Literature

Round 1

Reviewer 1 Report

P 2: one should discourage the use of “organic origin” to differentiate depression linked cognitive deficits  as opposed to neurodegenerative-linked CD: what else than organic can depression be? Of course, it plays out at the molecular level, but that’s as organic as you’ll ever find

P3 and subsequent: an MMSE of 30 after an initial measurement of 16 is quite surprising, and raises doubts as for a diagnosis of AD. The authors mention “His functional independence was mildly compromised for the most complex daily living activities”, which is somewhat vague but suggests minimal cognitive dysfunction at most (and the type of cognitive dysfunction, with an MMSE of 30, according to table 1, seems to be limited to the Stroop test and Recall of ROCF), another remarkable reversal as compared to the patient’s initial presentation of Major CI. In that light, the authors nevertheless conclude that “Therefore, the final neuropsychological diagnosis was initial memory impairment of likely neurodegenerative origin, in a patient with recent major depressive disorder and anxiety”. One cannot but wonder why the “likely neurodegenerative origin” is still … likely. And in fact, it was later replaced by “The conclusion was “minimal cognitive impairment in a patient with a history of major depressive episode”, with the only clinical improvement being in the Stroop test. This conclusion could have been made sooner.

P6: “FDG-PET abnormalities seemed to be stronger driver of cognitive decline,…”: driver should definitely be replaced by predictor, as of course FDG anomalies do not “drive” pathology, but simply reflect one of its components.

P7: The discussion on ECT and its potential application in AD is highly speculative, and one element that could be added as a caveat, especially when citing ref. 24, is that so far, and despite interesting developments, neurogenesis is not established in human adults (for a good review on this: Front Neurosci. 2022 Aug 5;16:933947. doi: 10.3389/fnins.2022.933947). Given that ECT is a very long shot for treating AD, and that it is not central to the paper, I think that this paragraph should suppressed.

P 8: “The authors hypothesized that the improved cerebral metabolism could be the mechanism by which ECT improved neuropsychological symptoms, causing sustained functional neural changes[18].”They are mostly re-interpreting others authors but still need to be rigorous in their interpretation. A quite frequent misunderstanding is exemplified here: the sentence suggests that cerebral metabolism drives neuronal function. While that could happen under extreme conditions such as life-threatening hypoglycemia, under essentially every other condition, it’s the other way around: neuronal activity drives cerebral glucose metabolism.

Conclusion: the authors conclusion is absolutely right. However, it could have taken another angle. They should emphasize that the AD FDG pattern found in typical AD (but not in vAD) in large parts reflects dysfunction in the default mode network of the brain. The centrality of this network in brain function relates amongst others to its pattern of reciprocal innervations with other networks. That it would show anomalies not linked to its own neurodegeneration in a number of conditions, but associated with dysfunction in other networks, or even to a non-degenerative lesion in a part of itself (for instance, in subjects who show anomalies in the posterior cingulate cortex/precuneus post hippocampectomy for epilepsy), hardly comes as a surprise. Therefore, Bayesian analysis of the potential usefulness of FDG PET for the diagnosis of AD in an individual case (well exemplified in section 3.4 on amyloid imaging) should consider the presence of conditions other than AD. An important conclusion of their study is that when confronted with a clinically complex case (presence of extensive vascular damage, trauma, prior surgery, hydrocephalus, age below 65, …), it should be made clear to the requesting physician that a positive for AD study carries a lower specificity than in more usual presentations of AD.

Author Response

Thank you for reviewing our manuscript and proving useful considerations. We hope to address them all satisfactorily.

P 2: one should discourage the use of “organic origin” to differentiate depression linked cognitive deficits  as opposed to neurodegenerative-linked CD: what else than organic can depression be? Of course, it plays out at the molecular level, but that’s as organic as you’ll ever find

Thank you for your observation. We corrected the term “organic” with the more accurate “neurodegenerative”.

P3 and subsequent: an MMSE of 30 after an initial measurement of 16 is quite surprising, and raises doubts as for a diagnosis of AD. The authors mention “His functional independence was mildly compromised for the most complex daily living activities”, which is somewhat vague but suggests minimal cognitive dysfunction at most (and the type of cognitive dysfunction, with an MMSE of 30, according to table 1, seems to be limited to the Stroop test and Recall of ROCF), another remarkable reversal as compared to the patient’s initial presentation of Major CI. In that light, the authors nevertheless conclude that “Therefore, the final neuropsychological diagnosis was initial memory impairment of likely neurodegenerative origin, in a patient with recent major depressive disorder and anxiety”. One cannot but wonder why the “likely neurodegenerative origin” is still … likely. And in fact, it was later replaced by “The conclusion was “minimal cognitive impairment in a patient with a history of major depressive episode”, with the only clinical improvement being in the Stroop test. This conclusion could have been made sooner.

We agree with you that this conclusion could have been reached sooner. However, the neuropsychologist that made the first diagnosis was highly impressed by the FDG-PET results (he was not blinded to the scan results), to the point that they may have influenced the diagnosis she attributed to the case. When we discussed the case before the second FDG-PET scan and neuropsychological evaluation, some of our colleagues argued that concomitant depression could have temporarily worsened the presentation of an otherwise mild cognitive impairment due to AD (as they expected the FDG-PET to confirm the first pattern). In order to acknowledge this potential confounder, we added a clarification in the text, which you may find in from line 143: “Possibly influenced by the knowledge of the FDG-PET results, the final neuropsychological diagnosis was initial memory impairment of likely neurodegenerative origin, in a patient with recent major depressive disorder and anxiety. “

P6: “FDG-PET abnormalities seemed to be stronger driver of cognitive decline,…”: driver should definitely be replaced by predictor, as of course FDG anomalies do not “drive” pathology, but simply reflect one of its components.

We changed the wording according to your useful observation.

P7: The discussion on ECT and its potential application in AD is highly speculative, and one element that could be added as a caveat, especially when citing ref. 24, is that so far, and despite interesting developments, neurogenesis is not established in human adults (for a good review on this: Front Neurosci. 2022 Aug 5;16:933947. doi: 10.3389/fnins.2022.933947). Given that ECT is a very long shot for treating AD, and that it is not central to the paper, I think that this paragraph should suppressed.

Thank you for this useful observation. We would like to maintain this brief paragraph, as we believe it could stimulate further research and it has some preclinical data available, but we acknowledged that it is highly speculative in the text, adding at lines 247-249: “If this is true, our case would suggest that restoring synaptic function would benefit brain metabolism and cognition altogether, with interesting potential consequences, which are at the time highly speculative.”

We also added the reference you kindly provided as a caveat for reference 24. The new section at lines 255-259 is now: “Interestingly, ECT induced temporary hippocampal neurogenesis and reduced aberrant exploratory behavior[24], consistent with studies on behavioral and psychological symptoms of dementia, for which ECT seems to be at least temporarily useful[25] (however, it must be noticed that neurogenesis has not been confirmed yet in the human brain in vivo[26]).”

P 8: “The authors hypothesized that the improved cerebral metabolism could be the mechanism by which ECT improved neuropsychological symptoms, causing sustained functional neural changes[18].”They are mostly re-interpreting others authors but still need to be rigorous in their interpretation. A quite frequent misunderstanding is exemplified here: the sentence suggests that cerebral metabolism drives neuronal function. While that could happen under extreme conditions such as life-threatening hypoglycemia, under essentially every other condition, it’s the other way around: neuronal activity drives cerebral glucose metabolism.

Thank you for your observation. We re-elaborated the sentence as follows, in order to clarify the relation between neural function and cerebral metabolism: “The authors hypothesized that the improved cerebral metabolism could represent the effect of ECT on neuropsychological symptoms, through sustained functional neural changes[18].”

Conclusion: the authors conclusion is absolutely right. However, it could have taken another angle. They should emphasize that the AD FDG pattern found in typical AD (but not in vAD) in large parts reflects dysfunction in the default mode network of the brain. The centrality of this network in brain function relates amongst others to its pattern of reciprocal innervations with other networks. That it would show anomalies not linked to its own neurodegeneration in a number of conditions, but associated with dysfunction in other networks, or even to a non-degenerative lesion in a part of itself (for instance, in subjects who show anomalies in the posterior cingulate cortex/precuneus post hippocampectomy for epilepsy), hardly comes as a surprise. Therefore, Bayesian analysis of the potential usefulness of FDG PET for the diagnosis of AD in an individual case (well exemplified in section 3.4 on amyloid imaging) should consider the presence of conditions other than AD. An important conclusion of their study is that when confronted with a clinically complex case (presence of extensive vascular damage, trauma, prior surgery, hydrocephalus, age below 65, …), it should be made clear to the requesting physician that a positive for AD study carries a lower specificity than in more usual presentations of AD.

Thank you for this interesting observation. We added part of your considerations in section 3.5 (Learning points and future directions), as we thought they would fit better in the consequentiality of the first paragraph and allow to modify the conclusion below. You may find the modified paragraph as such (from line 385): "What we could learn from this case is that baseline FDG-PET (and probably, even amyloid PET) might not reliably differentiate between DP and AD. The pattern of hypometabolism found in typical AD largely reflects a dysfunction in the default mode network of the brain, which might be also affected by other neurodegenerative pathologies[41] or even by non-neurodegenerative conditions, as our case seem to show.”

Accordingly, we added the reference by Drzezga, which is a recent review on the network degeneration hypothesis, since this was not introduced in other sections of the manuscript.

We also modified the last part of the conclusion, adding (lines 414-418) “Our findings suggest caution when using FDG-PET or even pathology-specific biomarkers to differentiate between DP and AD in elderly patients with cognitive impairment and concomitant depression, as these AD biomarkers might show reduced specificity in such atypical cases, compared to the more typical amnesic presentation.”

Reviewer 2 Report

The case report titled " Depressive Pseudodementia with Reversible AD-Like Brain Hypometabolism: A Case Report and a Review of the Literature" is a very interesting article. The authors discussed this case extensively with strong literature support. The article is overall well written but could be improved by avoiding some very long sentences. 

Although the most likely cause of global hypometabolism is from depression as the authors indicated, the possibility of pharmacological effect from the delorazepam cannot be excluded. I suggest the authors to search literature regarding the potential hypometabolism from benzodiazepine and add a brief discussion.

Author Response

The case report titled " Depressive Pseudodementia with Reversible AD-Like Brain Hypometabolism: A Case Report and a Review of the Literature" is a very interesting article. The authors discussed this case extensively with strong literature support. The article is overall well written but could be improved by avoiding some very long sentences. 

Although the most likely cause of global hypometabolism is from depression as the authors indicated, the possibility of pharmacological effect from the delorazepam cannot be excluded. I suggest the authors to search literature regarding the potential hypometabolism from benzodiazepine and add a brief discussion.

Thank you for your observations.

We tried to improve the article as you suggested, by reducing some of the long sentences you mentioned.

Regarding your second excellent suggestion, we performed a literature search about the potential effect of benzodiazepines on brain metabolism. The literature is quite scarce, but it clearly shows that benzodiazepines could cause a global reduction of cortical FDG-PET uptake values, due to a widespread depression of neuronal activity (which however in a study is reported to be more pronounced in frontal areas - which would further confirm that delorazepam was not the cause, since frontal areas were somehow spared in the first scan). However, both the visual assessment and the semiquantitative analysis use regional differences to evaluate patterns of hypometabolism, and these are not affected by the global uptake reduction caused by benzodiazepines. In order not to make the section too much longer, we summed up all of this in a few sentences, using only one significant reference to sustain this discussion. You may find the relevant paragraph between lines 276-282: “It is known that benzodiazepines may cause at most a cortical global reduction in brain metabolism due to a widespread depression of neuronal activity, but both visual assessment and SPM analysis rely on evaluation of regional differences. Although the effect of chronic use of delorazepam on brain metabolism in humans has not yet been studied, a recent work in rats confirmed that chronic administration of diazepam, another benzodiazepine, does not alter the patterns of FDG-PET uptake[29]. Therefore, it is unlikely that treatment could have influenced the scan results in our case.”

Reviewer 3 Report

The authors present a case of severe depression with  hypometabolism in AD pattern which was  reversible on second line therapy.

The fact that amyloid deposition was present actually questions the relevance of this case and findings. The authors have not mentioned whether precuneus hypometabolism was present. Definitely the presence of AD pattern of hypometabolism in patients with severe depression should be interpreted carefully. 

Author Response

The authors present a case of severe depression with  hypometabolism in AD pattern which was  reversible on second line therapy.

The fact that amyloid deposition was present actually questions the relevance of this case and findings. The authors have not mentioned whether precuneus hypometabolism was present. Definitely the presence of AD pattern of hypometabolism in patients with severe depression should be interpreted carefully. 

Thank you for reviewing our manuscript.

You may find the mention that bilateral precuneus hypometabolism was present at line 131: “In the same month he underwent brain MRI, revealing only mild diffuse cortical atrophy and age-related gliosis, and brain FDG-PET, showing severe hypometabolism in bilateral precuneus, temporal and parietal lobes[…]”

Regarding the amyloid deposition, we wrote extensively in section 3.4 why we believe that this could not have driven the metabolic changes observed in the patients first FDG-PET, and why the probability of AD would have been low (but not null) according to the whole clinical picture and age of the subject. Together with the consistent literature on reversible FDG-PET hypometabolisms in different examples of cognitive impairment, we believe that this case could be relevant and inspire new research on depressive pseudodementia.