There is no known brain morphology variant associated with AD. In this study, we examined the prevalence of two EC sulcal patterns from the MRI scans of YNL, ONL, and AD patients. We report that the frequency of a continuous EC rhinal sulcal pattern in the right hemisphere is more prevalent in AD as compared with normal controls. Further, we demonstrate that the EC patterns are unrelated to the ApoE genotype or to dominant handedness, they are reliably determined, and the patterns do not change with progressive atrophy. Our quantitative MRI and postmortem validation studies showed that the EC patterns are not related to EC surface, volume or the number of EC neurons. Overall, these data suggest that a rhinal sulcal pattern in the right hemisphere is a possible risk factor for AD. However, the anatomical substrate and developmental basis of the added risk remains unknown.
Among healthy individuals there is an evidence of normal “right-greater-than-left” medial temporal lobe (MTL) volume asymmetry (Mega et al., 2002
). In AD, the left hemisphere particularly the left MTL and other cortical regions seems to be more vulnerable to atrophy (Rusinek et al., 2004
; Thompson et al., 1998
). It has been shown that among MCI subjects worse cognitive performance is associated with left hippocampal atrophy (Dickerson et al., 2004
). Although our study did not attempt to characterize the laterality of the hippocampal formation atrophy associated with AD, speculatively, our finding may be consistent with the above observations. With the left hippocampal formation presumably affected earlier, the additional burden of a right hemisphere anatomical variation may further contribute to increased risk of expressing cognitive symptoms.
Coronal MRI images are commonly used to measure the EC size in AD (Petrella et al., 2003
) as it is believed to be a site of early NFT pathology and tissue atrophy (Braak and Braak 1991
). Based on prior postmortem validation studies, the MRI defined lateral margin of the EC is defined anteriorly by the rhinal sulcus and posteriorly by the collateral sulcus (Bobinski et al., 1999
; Insausti et al., 1995
). It is well known that there is variability in the patterns of the rhinal and collateral sulci (Goncharova et al., 2001
; Hanke, 1997
; Insausti et al., 1995
; Ono et al., 1990
; Pruessner et al., 2002
; Van Hoesen, 1995
). Specifically, some authors describe distinct rhinal and collateral sulci (Bobinski et al., 1999
; Goncharova et al., 2001
; Xu et al., 2000
) and others define the sulci in the EC region as a double collateral sulcus (Insausti et al., 1995
; Juottonen et al., 1998
). The apparent discrepancy reflects naming of the posterior extent of the lateral boundary of the EC. However, despite different names, our classification relies on the continuity versus discontinuity of two sulci and is unaffected by the naming. Prior work has not examined whether the variability in the continuity of the rhinal and collateral sulci is associated with disease.
In a prior anatomical study, we demonstrated that the rhinal and collateral sulcal boundaries enable sampling of a region that is predominantly EC with some PRC tissue included. However, this rule caused uncertainty because in a pilot study, both the rhinal sulcal pattern and a smaller EC surface area were associated with an increased risk for decline (unpublished observation). Thus, we were unable to conclude that the smaller EC observed in AD patients was due to atrophy, to an anatomical variant, or to an interaction between them.
The data from the present study confirm the hypothesis that the right hemisphere rhinal pattern is over-represented in AD as compared with elderly controls. However, we did not confirm the second hypothesis that EC pattern type is related to EC surface area or volume either as estimated at postmortem or in vivo. The quantitative histology data shows no relationship between EC pattern and either EC neuronal counts or EC size or PRC size. Further, the MRI data show no relationships between EC pattern and either EC surface area or EC volume. As all these validations were conducted in brains from normal subjects, the results suggest that the increased risk for AD conferred by the EC rhinal sulcus pattern is not simply due to a premorbid reduction in tissue reserve. The biological basis for the increased AD risk associated with the rhinal sulcus pattern remains unknown.
Nevertheless, evidence from other disease studies show that distinctive sulcal and gyral patterns are related to cognitive performance (Fornito et al., 2004
) and confer increased life-long risk. Morphological variability in cortical sulcal and gyral patterns has been linked to schizophrenia (Le Provost et al., 2003
; Yucel et al., 2001
). Global measurements (fractal dimensions) of the cortical convolutions have also indicated decreased shape complexity in schizophrenia, obsessive-compulsive (Ha et al., 2005
) and bipolar disorders (Bullmore et al., 1994
). Variability in the intraparietal sulcus was associated with dyscalculia in Turner syndrome (Molko et al., 2003
) and duplicate Heschl's gyri were related to reading disabilities (Leonard et al., 2001
). Twin studies have indicated significant genetic contributions to paralimbic brain structures consistent with the hypothesis that common genetic risk factors could contribute to both the EC sulcal pattern and the risk for AD (Wright et al., 2002
). The genetic and possible intrauterine environmental influences contributing to these morphology findings remain obscure. Moreover, it remains unknown if the EC rhinal sulcal pattern is associated with early brain development or AD related biological changes.
Gyral development starts early in fetal life and is under genetic control (Gilbert et al., 2005
). The time of appearance of the different cortical sulci is orderly (Garel et al., 2001
). Deeper primary sulci emerge first and the collateral sulcus appears at about week 23 of gestation (Chi et al., 1977
) followed by more shallow secondary and tertiary branches (Lohmann et al., 1991
). The timing for the smaller rhinal sulcus is not known but it is potentially later than the collateral. Overall, the later in development a sulcus appears, the stronger are potential environmental causes of variability.
Therefore, there are at least three explanations for our findings. First, there may be common genetic factors which contribute to both the entorhinal cortex rhinal sulcus variant and to late-onset AD. To date, one genetic risk factor for late-onset AD is characterized, the ApoE4 allele, but we did not observe any relationship between this and the EC pattern. However, genetic epidemiological analysis has suggested that there may be at least 4−7 other genes that influence late-onset AD (Hardy, 2006
). Recently, a late onset gene was identified (Rogaeva et al., 2007
) but there is no known relationship to brain morphology. Second, proteomic analysis of gene expression in the EC of MCI patients has shown significant alterations of a number of genes that are involved in cell cycle/proliferation and cytoskeletal maintenance, which could also influence sulcal patterns (Ho et al., 2005
). It remains unknown when in the course of development these proteomic effects are expressed. Furthermore presenilin, a gene associated with early onset AD, is also involved in the regulation of brain development (Fischer et al., 2005
) but there is no information about a possible association between presenilin mutations to any EC pattern types. Third, epigenetic factors can contribute to the variability of gyral patterns (Bartley et al., 1997
; Lohmann et al., 1991
). Cortical folding approximately doubles from birth to early adult life (Armstrong et al., 1995
), thus environment-associated modifications can possibly influence postnatal development.
There were several weaknesses and limitations with this study. The postmortem validation was done on a small sample n = 34 and the majority of data were from the left hemisphere. This limited ascertaining the relationships between pattern and size by hemisphere. Our longitudinal study (limited to 3 years) revealed that the progression of atrophy in AD patients did not alter the classification of the sulcal pattern. However, we did not examine subjects in the transitions between normal and cognitive impairment or between MCI and AD, so that we cannot exclude the possibility that EC pattern changes in the right hemisphere occur early in the atrophic changes of AD. Thus, earlier and longer duration observations are warranted. Finally, there are potential biases related to the normal subjects selected for study. These biases include: (1) old normal subjects who volunteer for memory research may be a product of sampling from two distributions, a proportion of self-referred individuals at greater risk than a randomly sampled cohort as compared with “randomly recruited” normal controls. Potentially this would have a conservative effect when compared with AD, but the actual extent of this bias is unknown; and (2) a young group which includes subjects, perhaps enriched with a family history for AD, that are destined to decline in the future to AD, but have not reached the ages for expression of symptoms. Nevertheless, our observation of an elevated prevalence of the rhinal sulcal patterns in AD and in young normal controls relative to old normal subjects is consistent with a brain characteristic that is determined early in life and which confers a risk late in life. In other words, young controls fully reflect the distributions of EC sulcal patterns. Older controls show reductions in the prevalence of the rhinal sulcal pattern presumably due to a disproportionate number transitioning to AD.
In summary, we found that a right hemisphere rhinal sulcus pattern is more common in AD than in normal controls. The entorhinal cortex sulcus pattern determination is highly reliable, patterns do not change over time and are equivalently determined in vivo with MRI and from postmortem sections. The rhinal sulcus pattern is not related to ApoE e4 genotype (a known genetic risk factor for AD) and in normal subjects we did not find any in vivo MRI or postmortem histological evidence of an association between EC pattern types and EC size or neuronal numbers. It appears that the risk for AD associated with EC pattern type is not associated with decreased neuronal counts or MRI volume reductions. Overall, our results suggest further inquiry into the genetic or early developmental origin of the rhinal sulcus pattern as a potential factor in AD pathogenesis would be worthwhile.