Do ACoA Patients Demonstrate Evidence of Retrograde Amnesia?

The current findings confirm that patients who sustain rupture and who undergo surgical repair of anterior communicating artery aneurysms are at risk post-operatively for retrograde memory problems. These results are consistent with other studies that have demonstrated significant retrograde amnesia in ACoA patients. In comparison to nonamnesic control participants, ACoA patients demonstrated deficient recall of photographs of famous people and popular news events. However, the magnitude of the remote memory problems of ACoA patients was less than that seen in association with temporal amnesia.

Performance on two tests of memory for information in the public domain was examined. These tests differed with respect to frequency of exposure of their respective events and the nature of the retrieval cues. All groups, including nonamnesic control participants, were more adept in recalling FFT than TNET items. These findings indicate that recurrent exposure improves retention, a pattern noted in previous studies of long-term memory (e.g., Bahrick, 1984; Conway et al., 1992). The FFT and TNET both focus on recall of factual information and, to that extent, both tests involve semantic memory. However, because TNET items are from a circumscribed time period they also may involve episodic memory. It is likely that successful retrieval of TNET items depends on access to contextual cues. For instance, recall of the TNET item Fawn Hall (who was in the news frequently, but for a limited time period, in the late 1980's) would evoke specific contextual associations (i.e., the Contragate affair, Oliver North, Senate hearings, etc.) whereas recall of the FFT item Princess Diana would be contextually independent.

ACoA patients performed worse than NC but better than TA participants on FFT recall. With cuing, ACoA patients performed similarly to NC participants suggesting that their deficient memory was due to a retrieval problem rather than a knowledge deficit. Cuing was very beneficial for TA patients but, even with cuing, they demonstrated impoverished memory for FFT items in comparison to ACoA and NC participants. Consequently, their deficient knowledge of FFT items could not be simply seen as a consequence of deficient retrieval.

The second test of retrograde amnesia, the TNET, focused on knowledge of events that were in the news for a time-limited period. Both amnesic groups demonstrated significant impairments on tasks of free recall, with ACoA performing significantly better than TA patients. On the recognition test ACoA patients performed as well as NC participants but TA patients performed significantly worse than both groups. The failure of TA patients to respond to TNET recognition probes is further confirmation that their remote memory problems are qualitatively different from those seen in the ACoA group.

Does Cuing Improve Performance on Tasks of Retrograde Amnesia?

As in previous studies (e.g., Gade & Mortensen, 1990; Kopelman et al., 1999), cuing and recognition probes significantly enhanced the performance of memory impaired patients in our study. ACoA patients did not differ significantly from NC participants when cues were given but, even with cuing and recognition probes, the TA group performed significantly worse than ACoA and NC participants. This pattern of findings suggests that remote events were represented in memory for ACoA patients whose failure to recall items was due to retrieval difficulties. The finding that TA patients were unable to remember items, even when provided with cues and recognition probes, suggests that their problem was not solely due to a disruption of retrieval. It may well be that the TA patients could not access remote items due to the degradation of this information secondary to lesions in medial and lateral temporal cortices.

Do Executive Deficits Contribute to Retrograde Amnesia?

The contribution of executive dysfunction to various components of retrograde amnesia was examined in amnesic patients. Contrary to our prediction that executive deficits would underlie retrograde amnesia, performance on Trails–B and two indices of the WCST did not correlate with any measure of retrograde amnesia for either group. However, FAS performance correlated significantly with free recall of FFT and TNET items in the ACoA group. TA patients did not show significant correlations between FAS and recall of information from remote memory.

Our failure to detect a significant association between many aspects of executive abilities and retrograde memory may be due to the particular tasks used to assess executive dysfunction. The FAS, the Trail-Making Test, and the WCST are thought to be mediated by dorsolateral prefrontal cortices whereas the damage following rupture and surgical repair of ACoA aneurysms typically involves orbitofrontal brain regions. Had we chosen tests more sensitive to orbitofrontal damage, a different pattern of findings might have emerged. At the same time, it should be noted that other investigators demonstrated evidence of executive impairment in the ACoA patient population using a test battery similar to ours (D'Esposito et al., 1996).

The pattern of findings noted above can also be understood by examining the neuropsychological processes common to the selected tasks of executive functions and retrograde amnesia. The Trail-Making and WCST tests focus on components of cognition (e.g., mental flexibility, set maintenance, and problem solving) that may not be critical in terms of recall or recognition of remote events. In contrast, the processing demands of the FAS task and the tests of retrograde amnesia used in this study are quite similar. All three tests rely on the lexical retrieval of information previously stored in memory. It is, therefore, not surprising to note significant associations between performance on these tasks.

Subsequent analyses provided further support for the idea that the retrograde amnesia of the ACoA patient group was largely due to impaired lexical retrieval. Even though patient groups were matched according to naming on the BNT, performance on this test correlated only with retrograde memory in ACoA patients. TA patients did not show significant correlations between BNT scores and recall on the FFT (r = .28; p > .05) or TNET (r = .20; p > .05). In contrast, there was evidence of a significant correlation between the BNT performance of ACoA patients and recall of items from the FFT (r = .53; p < .05) and TNET (r = .78; p < .001).

Do Neuroanatomical Factors Explain Differences in Profiles of Retrograde Amnesia?

Previous studies have shown that lesions in frontal regions, the basal forebrain, and the striatum contribute to the density of anterograde amnesia in the ACoA patient group (Damasio et al., 1985; Irle et al., 1992). Moreover, severity of amnesia has been associated with a specific pattern of lesion: combined lesions in basal forebrain and striatal regions have resulted in severe memory loss whereas isolated damage to either of these brain regions resulted in milder memory deficits. This finding was partially confirmed in the current study where 3 patients with damage in these brain regions demonstrated severe anterograde amnesia. However, there were exceptions to this pattern: at least 1 patient with damage in these brain regions did not have significant anterograde amnesia and 2 others without damage in those regions did have dense anterograde amnesia. Based on Irle's findings, we also speculated that combined lesions in the basal forebrain and striatum would predict density of retrograde amnesia. However, our findings were mixed with regard to this hypothesis. Several patients with basal forebrain and striatal damage had severe retrograde amnesia but others demonstrated dense memory loss in the absence of these lesions.

There are several differences between Irle's investigation and our study that are worth noting. Irle et al. (1992) examined a total of 57 ACoA patients, 30 with well-defined lesions and 27 who did not have evidence of lesions on imaging studies. Irle's conclusions regarding the critical substrates of anterograde amnesia were based on averaged data from patients grouped according to lesion site. We had access to imaging data for only 10 ACoA patients, several with no imaging evidence of lesions. Our small sample was further fractionated by heterogeneity of lesion site. The small sizes of our respective groups precluded statistical analysis of the relationship between damage to focal brain regions and retrograde amnesia. Close scrutiny of individual ACoA patients' performance in both studies revealed that basal forebrain/striatal lesions are not always seen in conjunction with anterograde or retrograde amnesia: as in our study, several of Irle's ACoA patients had severe memory difficulties even though there was no imaging evidence of brain lesions.

Previous Studies of Retrograde Amnesia With ACoA Patients

There have only been two similar group studies that have focused on the profile of retrograde amnesia in ACoA patients. Our findings are consistent with those of Gade and colleagues who showed that ACoA patients, assessed 2 years post onset, had retrograde amnesia. Our findings are also consistent with those of D'Esposito and colleagues who examined the performance of ACoA patients on the FFT at two consecutive points in time. Initial testing (1–2 months post onset) revealed evidence of impaired FFT recall for those patients with executive dysfunction. Repeat testing (2–4 months post onset) revealed improved FFT recall in conjunction with generally improved executive abilities. These findings supported the hypothesis that retrograde amnesia may result from a combination of memory impairments and executive deficits. Our patients also demonstrated impaired FFT recall in the presence of executive dysfunction. In our study, FAS was the only executive test that correlated with remote recall. This suggests that an inability to access stored information is the common underpinning in the remote memory problems and FAS deficits in ACoA patients.

Limitations

Several limitations of this study may have affected the outcome. The first limitation pertains to selection biases. All of the patients in this study were referred to the Memory Disorders Research Center on the basis of memory difficulties in everyday life. Consequently, the present findings do not necessarily generalize to the larger population of individuals who have suffered rupture and surgical repair of an anterior communicating artery aneurysm. A second limitation concerns the fact that only a small number of patients participated in this investigation; it may well be that findings would differ in a larger cohort. This issue may have been particularly problematic with regard to the neuroanatomical analyses because data from only 10 patients were available. A third limitation concerned a possible ceiling effect regarding the performance of NC participants' on tasks of cued recall and recognition. A ceiling effect may have obscured a possible difference between ACoA and NC participants on these measures. Future studies should use more challenging tasks. Finally, it should be acknowledged that reliability and validity analyses of the FFT and TNET have not, to date, been undertaken. This shortcoming is particularly problematical in light of previous work indicating that even dense amnesics may demonstrate inconsistent recall on such tasks (Cermak & O'Connor, 1983).

Summary

Findings from the present study are consistent with previous studies indicating that ACoA patients have remote memory problems. In the present investigation ACoA patients performed worse than nonamnesic control participants on two tasks of retrograde amnesia. In addition, ACoA patients performed better than TA participants. Findings suggested that the poor performance of ACoA patients on tasks of remote memory was associated with impaired lexical retrieval, a problem that also affected performance on tasks of verbal fluency. Regardless of etiology of amnesia, all patients benefited from cuing and recognition probes. However, even with cuing, TA patients did not perform at the same level as ACoA or NC participants suggesting that their remote memory deficits could not be wholly attributed to a retrieval deficit.