A substantial number of cortical biopsies studied here, obtained from 95 subjects with iNPH, exhibited ADNC, with Aβ in 63 and HPτ in 61%. The frequency of subjects with Aβ is in line with previous reports whereas the percent of subjects with HPτ is significantly higher [5, 13, 23, 25, 31, 35]. One explanation to these discrepant results is the age of the cohort investigated. We chose to include subjects within the age range of 75–79 years, compared to other studies that included subjects within an age range from 28 up to 87 years [23]. We excluded the oldest to be confident that Age Related Tau AstroGliopathy (ARTAG) pathology was not present [21].
Another factor that might have influenced the finding of a high SAF/HPτ in our cohort is the methodology. Visualization of HPτ applying IHC and automatic platform is currently relatively secure, whereas the assessment of IR might vary. Here we used morphometric method and assessed all pathology disregarding the size of the alteration, i.e., also tiny granules were included. The SAF/HPτ in our study varied from 0.02 (only few grains) to a SAF value of 19%. A low value of SAF (0.05) is pictured in Fig. 1e-f, visualizing both the sparse extent and focal distribution of IR that might be disregarded without the use of imaging technique. To obtain a result in line with previous reports (10% of subjects with HPτ pathology), all subjects with sparse HPτ/SAF (0.02–1.5%) should have been excluded in our cohort [23]. In summary, three aspects that have influenced the outcome are the age of the cohort, method used to visualize the lesion and assessment strategy implemented.
Our observation of a relatively high prevalence of Aβ- and HPτ- pathology (63 and 61%) is in line with the biology of ADNC, known for increases with chronological age [9, 14]. Noteworthy, even in a relatively narrow age span of 5 years, from 75 to 79 years, the extent of both HPτ and Aβ correlated with the age at a borderline level.
In line with what has been reported in PM studies, a significant correlation was observed between the extent of hallmark lesions of AD, Aβ, and HPτ. The two protein alterations of ADNC progress, following distinctive neuroanatomical regions; Aβ, starting in the cerebral cortex, progresses through the limbic regions of the brain toward the cerebellum, whereas HPτ is seen initially in the subcortical areas, following in the limbic structures and finally in the neocortex [3, 6, 7, 40]. Thus, when both these alterations, i.e., ADNC, are observed in the biopsy obtained from the cerebral cortex, the finding suggests a stage of ADNC pathology corresponding to at least low/intermediate level, following the current staging strategy [17, 27]. These observations of a substantial number (~ 60%) of iNPH displaying low/intermediate level of ADNC is in line with a previous report indicating that 50% of subjects with iNPH develop dementia over time, presumably of the AD type [20].
Surprisingly, when assessing the neuronal marker NeuN, we noted an increase in the SAF with age, i.e., we observed a higher number of neurons within a defined area, i.e., more neurons in relation to the matrix/neuropil. This finding suggests that in our cohort of iNPH subjects, a significant alteration not previously reported is depletion of the matrix/neuropil. One explanation for this outcome could be consolidation of neuropil due to an eventual degenerative process, including loss of neuronal- and glial processes or depletion of extracellular matrix proteins. Both a loss of neuronal processes and a depletion of matrix proteins have been previously reported to be observed in subjects with AD but not in subjects with iNPH [36,37,38]. Thus, it is not clear whether this observation, i.e., depletion of the matrix/neuropil is due to the ADNC alterations, seen in a high number of our cohort, or whether it is an alteration related to iNPH. Interestingly the NeuN correlated with Aβ at a borderline level but only in females, and this finding might be due to a more severe depletion of neuropil, i.e., neuronal processes and matrix proteins in females parallel with the increase of Aβ (Fig. 2).
Synaptic density, a measure of the depletion of the neuropil, has been studied for many decades and is reported to occur in normal aging and is accelerated in different neurodegenerative diseases, particularly in AD [26, 37, 38]. Synaptic damage is seen early in the disease process; furthermore, synapses are supposed to be involved in the propagation of altered proteins, and the loss of synapses is suggested as being causative regarding CI [18, 33, 37, 38]. The changes in the synapses are suggested as being induced by ADNC, but the exact process is yet to be discovered [18, 33, 37, 38].
Here, we selected only a few samples to assess the SAF/SYP. The selection was based on neuronal counts, two with the highest and two with the lowest SAF/NeuN. These four cases were selected, disregarding the ADNC pathology or age. We noted a slight decrease in the SAF/SYP in patients with a higher neuronal density. This finding suggests that a depletion of the matrix might be an early change, preceding a neuronal loss, i.e., the processes are altered prior to the cell loss. This observation is certainly questionable due to the low number of subjects assessed, but it emphasizes that studies on matrix versus neuronal somata in degenerative diseases are certainly warranted.
Based on our and previous results, non-invasive-, in vivo-, techniques to assess ADNC in iNPH is certainly of interest. Currently, there is the Positron Emission Tomography (PET), technique applying specific tracers to assess the deposits of Aβ and HPτ in the brain [19, 29]. The most widely used is Carbon-11 labelled thioflavin-T derivate, Pittsburgh compound B [11C] PIB, to visualize Aβ, considered as being a reliable marker of Aβ pathology even in biopsied patients with iNPH [19, 34, 39]. In contrast, a HPτ selective tracer [18F]THK-5117 did not reflect the HPτ pathology in the cortical biopsies from iNPH subjects [24, 29]. The latter observation is due to the extremely low load of HPτ pathology as observed here.
The low SAF/HPτ and the observation of high neuronal density in a setting of iNPH suggest that neuroprotective-, antiapoptotic-, and HPτ targeting therapies may benefit these patients by delaying or slowing down the neurodegenerative process that eventually will lead to AD, as previously reported [2, 12, 20, 25, 41, 43]. Furthermore, there is an urgent need to learn more about the constitutions of the matrix to eventually identify new targets for therapy.
Two-thirds of all patients affected by AD are females; thus, the female gender, next to age and apolipoprotein E status, has been reported as being one of the risk factors of AD [4]. In line with the above, in our cohort of subjects with iNPH, a higher number of females displayed ADNC when compared with men. This was noted particularly for HPτ (both incidence and extent) but also at a borderline level for Aβ (both incidence and extent). Thus, our observations detected in a minimal cortical biopsy are in line with previously published data from several PM studies, where the whole brain has been investigated [15, 30].
Most studies assessing ADNC and neuronal population are performed on PM brain tissue, animal models, or cell cultures. We have unique material to study ADNC and neuronal cell counts in surgical brain biopsies from patients with iNPH. Due to the relatively large number of cases that have been sampled during 10 years, we were able to select a relatively homogenous cohort regarding the age. The absolute benefit while assessing surgical samples in contrast to PM tissue is that alterations to be looked for in the tissue are not affected by factors caused by agonal state or PM events. Furthermore, as previously reported, prospective analysis of tissue alteration, in relation to the clinical progression can be carried out [22, 25].
ADNC pathology is associated with cognitive decline and neuronal damage as well as neuronal death, which are observed in AD [10, 16, 28, 38]. These alterations are not expected to be observed in subjects with iNPH. The observation of ADNC in a cortex biopsy is however in line with the observation that a substantial number of subjects with iNPH develop AD over time [20]. Twenty years of experience while assessing tiny brain biopsies obtained during VPS operation of patients with iNPH and clinical prospective studies of these patients have increased our understanding of both iNPH and as well of AD related neurodegeneration. Based on the above, iNPH seems to be an intriguing and reliable model of AD. The histopathological assessment of the brain tissue of iNPH subjects was carried out to assign a patho-anatomical diagnosis, i.e., with or without ADNC. It is thus surprising that this assessment is seldom performed in this setting.
In this study, we assessed the extent of protein expression in the tissue, applying a computerized morphometrical analysis. We chose to measure the SAF for all the proteins, independent of the compartment of expression to facilitate comparison. For the final SAF, pixels of adequate staining intensity were included. Thus, we feel that the methodology for morphometrical analysis is reliable.
The assessment of a surgical biopsy is certainly informative. However, a tiny biopsy represents only a fraction of a brain and does not represent all the neuroanatomical regions with various vulnerabilities displaying different pathologies. Thus, one needs to acknowledge the influence of sampling deficit on all the above.
In conclusion, here we assessed ADNC and neuronal markers in surgical brain biopsies from patients with iNPH. We identified remarkable neuronal preservation but also a substantial depletion of matrix/neuropil within our cohort. These findings are intriguing as they suggest that a loss of matrix/neuropil might be a hallmark lesion in iNPH. This is in line with the cardinal feature of iNPH, enlarged ventricles. Further, the loss of matrix parallel with preservation of neurons, might be an early sign of neurodegeneration. A substantial number of our patients displayed either Aβ or HPτ or both pathologies, even if the extent of HPτ, was low. These observations are indicative of a progressive neurodegenerative process, result in line with the finding that a substantial number of iNPH patients develop dementia. Additionally, females displayed more advanced ADNC and especially HPτ density in their biopsies, in line with the interpretation that the female gender is a risk factor for ADNC and AD. Surprisingly, we noted a positive correlation between neuronal count and Aβ, but only in females. These findings indicate that even subjects with Aβ in cortex might benefit of neuroprotective therapy. Two decades of experience while assessing brain biopsies from subjects with iNPH, certainly suggest that studying this patient category is a reliable model of AD. Our findings further suggest that a histological assessment of a brain biopsy obtained during a VPS procedure should always be carried out as a diagnostic procedure as it reveals information that cannot be obtained otherwise; hence, the obtained information can lead to a more tailored treatment strategy.