Abstract
Memory disorders in children and adolescents with brain tumors are rare, but devastating to social, academic, and vocational development. Many pediatric patients with intracranial germ cell tumors (GCTs) complain of memory difficulties. This study investigated memory across a series of GCT patients. A total of 33 GCT patients were retrospectively examined for diagnosis, imaging results, intelligence quotient, treatment variables, evidence of increased intracranial pressure at diagnosis, and memory. The incidence of amnesia in GCT patients was 55%. Memory disturbance could not be predicted by intelligence quotient, treatment, location of lesion, or hydrocephalus at diagnosis. The high incidence of memory deficits in GCT patients suggests a risk to memory in patients with GCT. Formal memory assessment should be considered in all patients with central nervous system GCTs. Specific counseling and planning to assist in adjustment and to ensure safety should be considered standard care for those with memory deficits. The nurse should be instrumental in facilitating understanding of this specific injury in the brain tumor population.
Isolated, acquired organic memory deficits, or amnesia, have been investigated and described, stimulated by the seminal work of Scoville and Milner in 1957. Their studies of adult patients posthippocampectomy for treatment of intractable epilepsy encouraged analysis of and research on the neuroanatomy and phenomenology of profound memory deficits in adults. Isolated memory deficits in children are more poorly characterized. The incidence, etiology, neuroanatomy, natural history, and the impact of acquired amnesia on daily functioning, specifically on school progress and emancipation, in children and adolescents are unclear. This study examines memory functioning in a group of latency-aged children and adolescents who have an acquired memory disorder in the context of a specific central nervous system (CNS) disease—germ cell tumors (GCTs).
It is important to better understand these patients. Their memory disorder is associated with changes to brain areas not typically considered to be involved in memory, and, because it is unanticipated, it may not be recognized. In addition, this circumscribed deficit has a pervasive and profound impact on daily functioning disrupting school performance, social interactions, sense of self, and psychological equilibrium. Though we are unable to cure the memory disorder, understanding the various etiologies, presentations, and ramifications of profound memory deficits for individuals and their families will be useful in assisting in their adjustment and helping families to make appropriate adaptations. The incidence and presentation of profound memory disorders in children must be considered when caring for children and young adults with CNS GCTs. Appropriate nursing care can help the families make accommodations to protect physical safety and aid in emotional adjustment.
Memory Disorders
Individuals with memory difficulties have shared characteristics (Milner, Squire, & Kandel, 1998). They have a disturbance in anterograde memory; they are unable to acquire new information, either verbal or nonverbal in nature. They can have entirely normal intelligence and are able to maintain information in working, or short-term memory (eg, as a group, they do not have difficulty repeating sentences or series of numbers when they are tested). However, they do not encode information for retrieval after a delay or interference, or recognize as familiar stimuli to which they have previously been exposed. They may be able to repeat 7 items from a list of 15 items with practice, but this is typically either the beginning or the end of a list. After a brief period during which they do not practice, they are less able to recall the items they have learned than most others their age. In addition, those with memory deficits cannot reliably discriminate the items they have repeatedly heard from distracters. Typically, there is a component of a retrograde memory deficit: an inability to remember information and events that occurred in the past, though this is more variable and graded. Amnesiacs with relatively intact intellectual function may confabulate, producing erroneous responses when unable to recall information that is requested (Butters & Miliotis, 1985).
Individuals with amnesia have a disturbance of declarative memory. They experience failure in the explicit retrieval of information when effort and conscious awareness are required. Declarative memory is of 2 types: episodic or semantic memory (Squire, Knowlton, & Musen, 1993). Episodic memory refers to autobiographical knowledge that is acquired about events that maintain a specific context, such as knowing what gift you purchased for your father last week or where you purchased it. Semantic memory refers to general knowledge about the world, such as facts, or data. On psychological testing, patients with amnesia have difficulty acquiring and recalling lists and learning and maintaining paired associates (Butters & Miliotis, 1985; Ignelzi & Squire, 1976). Even severely amnesic patients can acquire limited semantic information with a great deal of repetition (Squire et al., 1993).
Much of what we know about memory disorders comes from the study of individuals who have had surgical resections of the medial temporal areas for control of medically intractable epilepsy. The classic description of amnesia is based on observations of the patient, HM, reported by Scoville and Milner (1957). HM was a 29-year-old male, with chronic seizures, who had bilateral medial temporal-lobe resections for treatment of intractable epilepsy. He had significant deficits postsurgically. He could no longer learn or retrieve sets of paired words or remember what he had just drawn or copied, though his intelligence was normal (full-scale intelligence quotient [FSIQ] = 112). On a daily basis, this was seen as inability to experience a sense of familiarity about someone to whom he had been introduced once his focus of attention was altered. A profound memory disorder persisted throughout his lifetime.
HM’ s memory disorder was related to bilateral resection of the hippocampus (Scoville & Milner, 1957). Other brain areas have been implicated in producing amnesia or are required in concert to produce amnesia. The anatomic basis of the memory system, explored in animal models, and expressed in humans who have sustained specific damage, includes the medial temporal lobes, the adjacent related cortex, and areas within the diencephalon, especially the medial thalamus (Butters & Miliotis, 1985). The medial temporal cortex and the medial thalamus both project to the frontal lobe. Though frontal damage does not produce specific amnesia, it mediates the manner in which damage to the diencephalon or hippocampus is expressed (Squire et al., 1993). Differences in the presentation of hippocampal amnesia versus diencephalic amnesia are reported, though not consistently (McKee & Squire, 1992).
The neuroanatomy of memory has been described in adults with discrete acquired memory disorders, but isolated disorders of memory in children are rarer, and the critical structures involved are less clear. Changes in memory are reported in children with acquired injuries, for example, post traumatic-brain-injury (Lajiness-O’Neill, Erdodi, & Bigler, 2010), postradiation (Armstrong et al., 2010), in association with language disorders (Baird, Dworzynski, Slonims, & Simonoff, 2010), and after premature birth (Baron, Erickson, Ahronovich, Baker, & Litman, 2011). These disorders are often associated with multiple possible sources for the memory disorder, for example, hypoxia, developmental/genetic abnormalities, or more diffuse injuries. There is some specificity reported in the head injury literature, with the temporal cortex more vulnerable in acceleration–deceleration injuries, but typically, the memory deficit is a specific component of more widespread damage. There is evidence that children who receive CNS radiation to the temporal lobes for treatment of brain tumors are at greater risk for memory deficits than those receiving radiation to other areas of the cerebral cortex (Armstrong et al., 2010), but again, there are multiple deficits identified in this set of patients.
Study Background
We have observed an isolated memory disorder in a subset of children and adolescents diagnosed with CNS GCTs. GCTs are characteristically located in the midline; in the pineal or suprasellar regions. Presenting problems are related to location of disease. Pineal region tumors often compress the aqueduct of Sylvius, leading to symptoms associated with hydrocephalus. Pineal tumors may present with Parinaud syndrome secondary to compression of the pretectal area (Cho et al., 1998). These children demonstrate paralysis of upward gaze, and loss of pupillary light reflexes, accompanied by retraction–convergence nystagmus. The role of the normal pineal gland is obscure. Cyclic secretion of melatonin appears to be the primary role of the pineal. In humans, the pineal receives input from photic stimuli and influences the circadian cycle. Pineal tumors do not secrete melatonin (Ropper, 2005). Tumors in the suprasellar region typically present with neuroendocrine deficits, visual disturbances, or signs of increased intracranial pressure (Suri, Narang, Sharma, & Mahapatra, 2008).
Classification of CNS GCTs is based on histology. GCTs of the CNS are divided into germinomas (GERs) and nongerminomatous germ cell tumors (NGGCTs), the latter includes sembryonal carcinoma, yolk sac tumors, choriocarcinoma, teratomas, and mixed GCTs (Abeloff, Armitage, Niederhuber, Kastan, & McKenna, 2008). Pure germinomas are highly sensitive to radiation therapy and have a favorable prognosis with a 10-year survival rate of approximately 85% (Ogawa et al., 2004; Sawamura, Ikeda, Shirato, Tada, & Abe, 1998). Nongerminomatous germ cell tumors are resistant to radiation therapy and exhibit poorer outcomes, with a 10-year survival rate of 34% (Ogawa et al., 2004). Memory deficits are not commonly reported in patients with GCTs; there have been isolated case reports of patients with memory deficits in association with germinomas. Sato et al. (2003) describe a 19-year-old who presented with insidiously developing amnesia and a right hemiparesis resulting from a germinoma in the septum pellucidim. The lesion involved the fornices, a recognized component of the memory circuit. This young man showed resolution of symptoms subsequent to treatment, which the authors attributed to recovery from mass effect. Arita et al. (1995) reported a case of a 32-year-old who presented with severe amnesia. The crura of the fornix were affected. Svoboda, Richards, Polsinelli, and Guger (2010) report their treatment of an 18-year-old woman who developed amnesia at 13 years of age in the context of a suprasellar germinoma.
We report a series of patients with GER, and NGGCTs, who presented at diagnosis with complaints of memory difficulties or who developed a pattern of performance consistent with amnesia early in their course while maintaining an adequate or unchanged level of intellectual functioning. These individuals or their parents spontaneously described that the child/young adult was routinely unable to recall events with which he or she had been involved or conversations in which he or she had engaged. The memory disorder was widely disruptive of social, psychological, educational, and vocational functioning.
Methods
Patient Query
The records of patients treated at Children’s Hospital Colorado between 1994 and early 2011 were retrospectively reviewed for diagnosis, age at diagnosis, gender, treatment variables, the presence of hydrocephalus at diagnosis, magnetic resonance imaging (MRI) results, intelligence quotient (IQ), and memory functioning. All studies were conducted in compliance with local and federal human research protection guidelines and institutional review board regulations. Tumors were grouped according to the World Health Organization histological tumor classification (Louis et al., 2007).
Memory Data
Patient FSIQ was measured using the Wechsler Adult Intelligence Scale III or IV, the Wechsler Intelligence Scale for Children III or IV, or the Wechsler Abbreviated Scale of Intelligence (Pearson, San Antonio, TX). Memory was evaluated with the California Verbal Learning Test (CVLT, Pearson, San Antonio, TX). The CVLT requires the child or young adult to learn a list of items that is longer than typically can be recalled with a single presentation. The items come from 3 or 4 categories (eg, fruits, items of clothing), though they are not segregated during presentation. The list is presented in its entirety 5 times with the request that the patients tell the examiner as many items as they can recall, in any order, after each presentation. After a brief delay (of about 3 minutes during which the patient is presented with another list of the same length, including some items from the categories previously presented), the patient is asked to recall the initial list. Twenty minutes later, without being warned that this would be requested, the patient is again asked to recall the repeatedly practiced list. Immediately after this the patient is presented with a list of items from the original list, the interference list, and a list of semantically and phonemically related and unrelated items and asked to identify which items were on the list that they had heard 5 times. Scores are age corrected.
Memory disorders were identified using either a discrepancy model, or a loss of information model. Amnesia was identified using a discrepancy model if there was a difference of > 2.0 standard deviations between concurrent measured IQ and memory performance, with memory the weaker skill. Memory performance was based on the Total T (eg, how much information was acquired over 5 trials), long delay free recall (LDFR, recall after 20 minutes) or recognition discriminability (RD, how much information was retained, as demonstrated by the ability to discriminate previously presented stimuli from distracters, after a 20-minute delay with heterogeneous interference). Amnesia was identified based on loss of previously acquired information if there was a decrement of >1 standard deviation between Total T and either LDFR or RD, independent of overall cognitive functioning. The decision to use these definitions of amnesia is consistent with other studies indicating that a discrepancy of >2 standard deviations between a measure of memory and intellectual ability is a rare occurrence (Binder, Iverson, & Brooks, 2009) or that a loss of >1 standard deviation is significant.
Statistical Analysis
Assessment scores were standardized according to population norms and converted to z scores to facilitate comparisons. Statistical calculations were analyzed with Microsoft Excel (Microsoft, Seattle, WA) or with Graph Pad Quick Calcs (Graph Pad Prism, La Jolla, CA). For all tests, a level of P < .05 was considered statistically significant. All correlations were calculated using the Pearson correlation coefficient.
Results
A total of 73 patients were diagnosed and/or treated for GCT over a period of 17 years. Of these, 33 had sufficient data and were included in the study. The average age of the GCT study cohort at diagnosis was 13.5 years with a male to female ratio of 8:1. A total of 23 patients were diagnosed with GER, 9 were diagnosed with NGGCT, and 1 patient had a pineal teratoma and was included in the NGGCT group for analysis. The average age at diagnosis of patients with GER was 14 years of age (range = 2-26 years at diagnosis; male to female ratio of 10:1), and the average age of patients with NGGCT was 12 years of age (range = 7-16 years at diagnosis; male to female ratio of 4:1). Patients were tested on referral, and not all patients were evaluated at the same point in their course, with consequent variability in the intervals between diagnosis, treatment, and assessment. The average interval between diagnosis and testing was 15 months, with a modal interval of 9 months. Table 1 includes pertinent group characteristics.
Group Characteristics
NOTE: GER = germinoma; NGGCT = nongerminomatous germ cell tumor; IQ = intelligence quotient.
No significant differences.
No significant differences were found between patients diagnosed with GER or NGGCT in age at diagnosis, intellectual ability, or memory function as defined by Total T score. The mean FSIQ of the total GCT cohort was 94 with a standard deviation of 16 (population mean = 100; population standard deviation = 15). The mean composite memory score (Total T) of the GCT cohort expressed as a z score was −1.4, with a standard deviation of 1.3 (population mean = 0; population standard deviation = 1). The modal Total T z score was −3; median was −1.7.
The incidence of amnesia in the GCT cohort was 36% (35% of GER patients, n = 23, and 33% of NGGCT patients, n = 10) when defined as a difference of 2 standard deviations between measured intellectual performance and Total T (learning), LDFR or RD (eg, based on discrepancy from intellectual ability). If amnesia was defined as an unusually rapid loss of previously acquired information (eg, a decrement of >1 standard deviation between learning and either LDFR or RD), a greater level of memory difficulty was observed, with 39% of the GCT cohort demonstrating a deficit. In all, 18 of the 33 patients (55%) had memory difficulties when the requirement for this was meeting just one of the definitions of a memory disorder. If a stricter definition was used, requiring both poor learning relative to IQ and loss relative to learning as the definition of a memory disorder, 7/33 or 21% of the population of GCT patients with CNS lesions demonstrated a specific memory impairment. There was a significant relationship between specific diagnostic category and likelihood of having a profound memory disorder with only GER patients simultaneously meeting both criteria.
There was a significant relationship between IQ and Total T, learning over 5 trials (P < .04). There was no predictable relationship between IQ and RD or LDFR. There was no relationship between the likelihood of having a memory disorder, as defined by either the discrepancy or loss model, and simply having cognitive deficits, as defined as an IQ < 85. There was similarly no predictable relationship between symptoms of confusion, and/or a specific difficulty attending and processing information quickly, as measured by a discrepancy between the perceptual speed quotient, and any other IQ index of 15 points or greater and the presence of a specific memory disorder in the subgroup of patients for whom these data were available (n = 19).
Germ cell tumor patients who developed amnesia were significantly older at diagnosis than those who did not develop amnesia, when defined by meeting the standard of just one of the definitions (P < .04). Patients who met one definition of amnesia had an average age of onset of 15 years of age compared with patients without a memory disorder, who had an average age of onset of 11 years.
Overall, 25 of the 33 subjects, had evidence of increased intracranial pressure at presentation. There was no significant relationship between evidence of increased intracranial pressure at presentation and amnesia on exam. Of the 33 patients for whom we had pertinent data, 12 had not had radiation prior to evaluation. There was no relationship between having had, or not having had, radiation prior to assessment and the presence or absence of a specific memory disorder across the group as a whole. Of the 18 patients with a significant memory disorder, 5 had lesions involving structures in the area of the third ventricle, an area that has been associated with amnesia (Denby et al., 2009; Iizuka, Suzuki, & Mori, 2007). Of the 17 patients with isolated pineal involvement on imaging, 11 (65%) met one of the definitions of amnesia.
Tables 2 and 3 provide diagnosis, age at diagnosis, interval between diagnosis and evaluation, tumor location, treatment variables (chemotherapy and radiation), FSIQ, presence or absence of hydrocephalus at diagnosis, and follow-up status if known for the patients included in the GCT cohort who experienced amnesia. How their memory disorder was manifested is noted. Table 4 presents the relative frequency and type of memory disorder observed in GER and NGGCT patients. Patients with GER were the only patients to simultaneously demonstrate both abnormal learning or retention and loss of data. The history and presentation of one patient who developed a memory disorder in the context of a NGGCT is described below.
Patient Characteristics of Those With GER and Memory Disorders a
NOTE: GER = germinoma; FSIQ = full-scale intelligence quotient; LDFR = long delay free recall; RD = recognition discriminability; CVLT = California Verbal Learning Test; CSI = craniospinal irradiation; SSI = Supplemental Security Income.
The population mean for FSIQ is 100 with a standard deviation of 15.
Abnormal memory based on loss after a 20-minute delay.
Abnormal memory based on discrepancy between memory and IQ.
Patient Characteristics of Those With NGGCT and Memory Disorders a
NOTE: NGGCT = nongerminomatous germ cell tumor; FSIQ = full-scale intelligence quotient; CSI = craniospinal irradiation; LDFR = long delay free recall; RD = recognition discriminability.
The population mean for FSIQ is 100 with a standard deviation of 15.
Abnormal memory based on discrepancy between memory and IQ.
Abnormal memory based on loss after a 20-minute delay.
Patients With Memory Disorders Across Groups; Relative Risk
NOTE: GER = germinoma; NGGCT = nongerminomatous germ cell tumor.
No significant differences.
Case Report
Amnesia is disruptive of daily functioning. The impact is not robustly described by the measured psychometric deficit. Patients with amnesia lack ongoing recollection of their interactions, have an imperfect sense of continuity, cannot keep track of what they have read, and so on. Amnesic disorders fracture episodic experience. The presentation of a young man with amnesia is reported to underscore the pervasively disruptive nature of profound memory defects.
Case 1
AM was a 16-year-old male with a history of dyslexia who failed to come home the Christmas eve prior to diagnosis, having forgotten he was expected at a family holiday dinner. This lapse precipitated medical investigation. He had demonstrated memory problems for the 3 to 4 months prior to diagnosis. He had been unable to remember his school schedule and was getting to classes on time only by referring to a written schedule. He had been unable to remember his destination when driving. He had been surprised to find dogs at a close relative’s home, though they’d been present for several months and he had seen them numerous times. These memory difficulties had been attributed to depression. His initial MRI (Figure 1A) revealed a well-circumscribed mass in the pineal region, with evidence of leptomeningeal enhancement in the lateral and third ventricles, masses in the anterior third ventricle and left foramen of Monro, and abnormal enhancement of the optic chiasm. Physical exam showed no papilledema, gaze palsy, or other neurologic abnormality. Serum and cerebrospinal fluid were positive for the tumor marker β-human chorionic gonadotropin but negative for α-fetoprotein. The patient was diagnosed with NGGCT. Endocrine work-up was normal. Sodium level was normal.
Sagittal T1 magnetic resonance images with gadolinium: (A) at diagnosis and (B) 3 years after diagnosis
AM was treated with chemotherapy (6 alternating courses of carboplatin/etoposide and ifosfamide/etoposide) and radiation (23.4 Gy to the craniospinal axis with a boost of 55.72 Gy to the pineal). Endocrine follow-up revealed peripheral hypothyroidism, hypogonadotropic hypogonadism, and growth hormone deficiency, and these were appropriately treated.
AM was initially evaluated prior to any treatment. He demonstrated normal intellectual ability (verbal IQ = 78, performance IQ = 94, FSIQ = 85). His pattern of performance was consistent with his previously identified language-based learning disorder with adequate math but weak reading and spelling. Repetition of sentences and digits was intact. On memory assessment, he acquired new information poorly and lost it quickly. He had an unusual number of intrusion errors, identifying as familiar stimuli that were not on any list, i.e., confabulating. During testing, AM was frustrated when asked to learn a story and given 5 opportunities to learn that story. He poignantly noted that he could keep in mind part of the story if he focused on just that detail, but when he switched his attention to another detail, the first disappeared. He tearfully described feeling that he had “Alzheimer’s” disease.
AM’s MRI scan 3 years after diagnosis demonstrated mild ventricular enlargement and prominent sulci with no evidence of recurrent disease (Figure 1B). He had no evidence of disease on laboratory testing. Despite disease regression, memory impairment remained. Intellectual functioning was unchanged (FSIQ = 86). On the CVLT, he recalled primarily from the end of the list, reporting what he could maintain in working memory. He showed no increased recall over multiple trials. He lost 50% of the information he had acquired after a 20-minute interval. Recognition memory was abnormal. His academic skills remained unchanged.
AM reported a sense of chronic sadness. He complained that he could not judge how long ago an event occurred and was uncertain of what had transpired in the interval. He was dependent, unable to recall plans, or predict what was going to occur as he did not know what had happened in the past. Social relationships were limited by his inability to recall past interactions. He became confused and he did not remember earlier conversations or altercations. He drew the same picture repeatedly, presenting it as a gift over and over again. He attempted to complete high school. His school provided accommodations, such as open-book tests on material that needed to be explained not memorized, and limited his school program to classes where his preserved academic skills were emphasized. He became disorganized, anxious, and depressed nevertheless. His anxiety precipitated a number of hospitalizations. He wanted to learn to repair large machinery, but his anxiety in situations with which he was not familiar prior to the onset of his disease made vocational education difficult for him to tolerate. He has recently made progress in a rehabilitation program that is teaching him specific skills for recording daily information and addressing the anxiety he experiences when he feels confused and unable to recall what has occurred. Though now an adult, he cannot live independently, as he is unable to recall discussions, whether he has eaten a meal, where he is driving, or what has been planned for the future. He cannot evaluate the passage of time, and he is unclear as to what has happened to him over any interval longer than a few minutes. AM has remained unable to acquire new information for more than 5 years postdiagnosis. He has required intensive interaction with both the nurse and the nurse practitioner on the neuro-oncology team to help him and his mother navigate the psychiatric and medical system for a brain tumor survivor with severe memory impairment.
Discussion
Understanding the etiology, factors that affect outcome, and the nature of memory difficulties is essential for the appropriate assessment and guidance of children and adolescents with brain tumors. Nurses play an important role in education and counseling for this population. We found memory deficits to be evident on specific testing in many children with GCT, even those with normal intellectual functioning. The disorder was often present at diagnosis though frequently misidentified as representative of a psychiatric disorder or evidence of “adolescent behavior.” These GCT patients would not all have been anticipated to experience memory difficulties given the location of their tumors. Memory difficulties are disruptive of daily independent, social, vocational, and school functioning. The prognosis for many patients with GCT is good (Ogawa et al., 2004), and it is important to identify amnestic disorders early to provide counseling and intervention.
To investigate the prevalence and nature of amnesia in GCT, pediatric patients were retrospectively examined for memory impairment. Overall, 38% to 55% of GCT patients were found to have a memory disorder, depending on the definition of amnesia used. The presence of amnesia was substantially unrelated to intellectual ability. GCT patients who developed amnesia were significantly older at diagnosis than those who did not develop amnesia. Amnesia was unrelated to hydrocephalus at diagnosis or to radiation preceding assessment. The identified memory deficits were not reflective of more diffuse disorganization, confusion, or slowed speed of processing. Many GCT patients with amnesia did not have disease associated with classic memory structures. Those with GERs had more significant memory deficits.
The diverse treatment regimens used, our inability to test all patients, and the inconsistency in the interval from diagnosis to assessment makes it difficult to address the issue of treatment effects. Many of the patients, however, complained of memory difficulties, or were assessed early in their course and prior to radiation or other treatment and had psychometric evidence of a memory disorder. The early emergence of memory deficits suggests that this is not entirely related to chemotherapy or radiation variables.
Not all patients have been assessed longitudinally. Those that have been assessed over time often continue to demonstrate a stable memory disorder that limits career and social options. We completed repeat, formal assessment on 6 of the 18 patients who demonstrated memory difficulties at their initial exam. In this group, there was no measurable memory change in 3 of the 6 patients, and 1 of the 3 who changed had increased difficulties. Informal and descriptive follow-up suggest that those with profound memory disorders rarely fully recover, and they are unable to care for themselves or others, work at their prior level of competence, or succeed academically.
Adolescents and young adults (AYA) were more likely to develop amnesia than younger children. Research with a variety of populations has suggested that early-acquired lesions are more devastating, and diffuse deficits, not specific disorders, are more likely in patients who sustain early CNS injury across etiologies (Gordon, 2009). With age, neuropsychological functioning becomes more differentiated, plasticity more limited, and the consequences more circumscribed. Perhaps it is only in AYA that a discrete amnesia can occur. Specific evaluation of memory may be more important in the AYA population, a growing area for nursing awareness and research.
Eighteen patients without lesions involving classic memory structures such as the hippocampus, amygdala, thalamus, or fornix had measured memory deficits. Reports by Ignelzi and Squire (1976) and Carmel (1985) suggest that amnesia associated with third ventricular lesions may be due to destruction of the fornices. This may explain the deficits identified in GTC patients with suprasellar lesions. Memory deficits associated with pineal lesions are not described in the literature. Animal studies specifically exclude a role for the pineal in the acquisition of spatial memory evaluated via a Morris water maze performance. A study by Lecourtier, Saboureau, Kelly, Pévet, and Kelly (2005) suggests that damage to the habenular nuclei, adjacent to the pineal, can cause memory impairment in rats. It is possible that concurrent damage to the habenular nuclei that is not apparent by standard imaging is implicated in the development of amnesia in our patients. Further, investigation is needed to help us better understand the neuroanatomoy of memory difficulties in younger individuals with brain tumors, especially those with GERs.
Discussions with families, schools, and caregivers suggest that the impact of memory deficits is poorly appreciated. Amnesic individuals carry on adequate conversations, repeat what they are told, complete puzzles, make sandwiches or cappuccinos, read, and do math without difficulty. They come into the clinic for their visits and are alert and responsive. They express themselves normally. They often confabulate and give reasonable but inaccurate answers to questions about what has been happening in their lives; the uninitiated will not notice anything amiss. There is no relationship between memory functioning and intellectual ability. Yet these individuals cannot keep track of the disagreements that they have had with family members or friends and are dumbfounded when they receive an icy response when they call or visit later in the day. Caregivers and teachers have difficulty understanding why a child can demonstrate comprehension of information without remembering facts. These patients cannot remember to turn off the stove and are unsafe cooking when alone. They become anxious in new situations, and for them, most environments that were not familiar prior to disease onset are new. Without continuous memory the environment lacks context, and patients are unable to predict what might occur. Memory-disordered patients have difficulty creating or maintaining long-term relationships (including marriage) because they have not encoded and cannot recall what has occurred during prior interactions.
Identification of isolated memory disorders is essential for long-term and educational planning. Norm-based assessment of the acquisition, retention over time, and recognition of learned data is not part of typical school exams (completed for Individualized Education Plans); this range of memory functions is difficult to effectively evaluate in a clinical setting. Identification of memory disorders via specific assessment can be helpful in understanding and containing the social impact of failing to maintain a continuous sense of what is occurring.
Identification of changes in memory is essential due to the devastating academic, social, psychological, and vocational consequences of memory disorders. The bedside nurse is in an ideal position to monitor new brain tumor patients for memory loss. Nurses should be aware of the suprasellar and pineal locations of brain tumors as potential areas inducing memory changes. Referral for neuropsychological assessment should be considered in these patients. As we assess the quality of life of pediatric brain tumor survivors, the ability to make new memories is an important area of function. AYA survivors of intracranial GCTs may have good survival, but with severe memory impairment, they may never live independently. This study suggests that careful assessment of memory should be considered even in the absence of CNS disease that impinges on the classical memory areas.
Footnotes
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The author(s) received no financial support for the research, authorship, and/or publication of this article.