Christina A. Meyers, Ph.D.
Professor
Neuro-Oncology
The University of Texas MD Anderson Cancer Center
Hi, I am Christina Meyers. I'm a neuropsychologist at the University of Texas MD Anderson Cancer Center.
Cancer patients suffer from a number of adverse symptoms including cognitive impairment, which will be the focus of this presentation; sleep disturbance, pain, fatigue, mood disturbance, sexual dysfunction, and others.
Of course, symptom assessment is mandated as a part of routine patient care. Pain is the fifth vital sign across the country and MD Anderson rolled out distress as the sixth vital sign fairly recently. And cancer treatment is only truly successful if these symptoms are managed. But successful management is hampered by the lack of knowledge of the pathophysiology of these symptoms and improved knowledge may lead to targeted therapy.
Now cognitive dysfunction occurs in the majority of cancer patients on active therapy and it is often the first symptom of cancer pre-diagnosis. Often times, patients will tell me that they've suffered from fatigue or distractibility, which they attributed to stress or overwork, and it turned out to herald the cancer diagnosis. These symptoms persist in a sub --- substantial number of patients after treatment is discontinued and is often referred to by patients as chemobrain or, if you're at the UK or Canada, chemofog And it is a manifestation of central nerv --- nervous system toxicity.
Now what are the components of cognitive dysfunction across cancer pat--- across cancers and across treatments? One is restriction of working memory capacity. By that I mean the amount of information the person is able to handle is reduced. So if I give the person 12 words to learn, they will only learn eight. They will remember all eight later, so it is not rapid forgetting of new information, but just how much the person can handle at once. Also, people suffer from inefficient memory retrieval. By that I mean they block on words in conversation or they'll forget the name temporarily of somebody they know very well, but it always comes back to them at some point. Also, there's a variability of focus attention. A person will be working on something and it's like the lights go out and then when the lights come back on in their brain, they're strug --- struggling to get back on track. They also suffer from an impairment of divided attention and by that I mean multitasking. So people tell me that they no longer can multitask and could really only do one thing at a time. However, they tend to have in --- intact reasoning and problem-solving, so people will have difficulty doing their normal routine even though their intellectual functions are not compromised.
So what is the effect on daily life? If it's hard to multitask, people become overwhelmed when more than one thing is happening. So people will tell me they can't go to parties anymore because they're too many conversations. Or they can't have all the grandchildren over at once because it's just too much. Also, people get easily distracted; they can't concentrate for long periods of time. So they'll do a task and get distracted and do another one and get distracted. And at the end of the day they've had five things started and none of them are completed. People are much slower to do things, so they miss points in conversations or they can't complete tasks in the time that they normally could. And probably the worst thing for this is increased mental effort. It takes more effort to do everything and that just exacerbates all other symptoms.
So we have the chicken and egg problem. If somebody is suffering from cognitive dysfunction and it takes more effort to do everything, they'll get tired at the end of the day. Or if the person has cancer-related fatigue, it's hard to think straight when you're exhausted. And, of course, if you have pain or sleep disturbance this all makes everything worse.
So, what can cause cognitive dysfunction in cancer patients? The cancer itself can, the cancer treatment, certainly every kind of treatment can do it, medical complications such as anemia or infection. Sometimes the person has something else going on unrelated to their cancer. It's not --- it's not infrequently we will see a person for memory problems, an elderly person, who ends up having Alzheimer's disease. Or people can have preexisting psychiatric or other medical conditions. Adjuvant medications can also cause these symptoms. People could be on steroids or antiemetic medication that will do it. Or they can have reactive mood and adjustment disorders. So if somebody is distressed and upset by the situation they're in and are preoccupied, they may not pay attention to what people are telling them. And it may look like a memory problem when it really isn't.
So, we assess cognitive function in cancer patients so we can learn what the cognitive problems are before their treatment ever is started. We want to know what different cancer therapies do to cognitive function and most importantly how we can help our patients who suffer from this.
So, the impact of cognitive dysfunction, we kind of --- we look at the World Health Organization rubric for this. Impairment is what is happening in the brain. And, as a neuropsychologist, I give tests of brain function to determine what's going on. So for instance they may have inefficient memory retrieval. The disability is the impact of that impairment on daily life. So if you have inefficient memory retrieval you may have word-finding pauses in conversation. But the handicap really depends on the person, where that person is in their life, what the demands are on them. So, for instance, a person who refreshes hard drives and may not talk a lot, this may not be such a problem, but if the person is a lawyer in a courtroom setting he may now be disabled from his normal work.
So, I think of the causes of cognitive dysfunction as the seed, which is the cancer; the soil, the person that the cancer is in; and the pesticides that we offer for therapy.
So, in terms of brain tumors, the neurocognitive effects, of course, will be where in the brain the tumor is, it'll disrupt different neural networks and cause different symptoms, how big the tumor is, more importantly how fast it's growing, lesion momentum, and of course treatment is directed at the brain.
So this is a post-op scan of a young woman of 25 who underwent an extensive surgery in her left hemisphere and it removed most of the areas that we consider important for learning and memory and speech. And when I tested her after this massive surgery she was perfectly normal, in fact, she was better than normal, because a very bright young lady. And when I probed a little bit further, her mother who is a teacher told me that she had switched handedness from right to left when she was five years old, so she probably had this tumor growing very, very slowly her whole life and her brain re-organized.
And this is a post-op scan of a person with a much smaller surgical resection in the left hemisphere, who was in fact a courtroom attorney who had spontaneous garbling of his speech and within two weeks was globally aphasic. He couldn't speak, he couldn't understand, he could not read and he couldn't write, a global aphasia in two weeks from a rapidly growing glioblastoma.
But non-brain cancers can cause cognitive deficits before treatment is initiated.
For instance, in our studies of breast cancer patients, we find that more than a third have impaired cognitive function prior to treatment; 26% had mild-to-moderate affective distress, but the cognitive dysfunction was not rala --- related to this distress. And we had trends toward, in terms of explaining this, menopausal status, hormone replacement treatment and the extent of their breast cancer surgery.
And this shows you in just what areas the women were impaired. And it tended to be in learning and memory and attention and speed of processing.
We've also found in people with small cell lung cancer, prior to any treatment, that there's a high prevalence of cognitive impairments. 70% of the people we tested had memory problems, 38% had frontal lobe executive dysfunction, and a third had fine motor coordination problems. But again reasoning and simple attention was not affected.
And in leukemia, we find the same --- same pattern. Pretreatment in people with acute --- acute myelogenous leukemia or myelodysplastic syndrome, again more than 40% have memory difficulties, and about a third have fine motor dexterity problems and executive dysfunction. Also, in this population, fatigue is highly-highly prevalent.
So, now the cancer may have caused some problems and now we got --- offer you treatment for your cancer.
So radiation therapy to non-brain sites such as breast cancer or prostate cancer causes ment --- brain-mediated fatigue. There can be physical fatigue certainly, but the brain madi --- brain-mediated fatigue is very prevalent. And people are vulnerable to distraction and impaired complex information processing. And again problems with frontal lobe executive function, which is organization and multitasking. Fortunately, this tends to be self-limited. And, once radiation therapy is completed, people tend to get better.
But not so with chemotherapy. For breast cancer, standard dose chemotherapy for breast cancer, as I mentioned, about a third of women have cognitive impairment before treatment started. But 65% of women declined during treatment and this is in the setting of an expected practice effect. What I mean by that is, when you have been given this battery of tests a second time, you should do somewhat better because you know what to expect. So, these women are declining when they should have gotten somewhat better because of exposure to the tests. This is after six courses of standard FAC chemotherapy. And many of them had difficulty maintaining their usual work and routine.
So, as I said, about a third of women had problems prior to treatment, but 65% declined, and in mostly learning and memory and executive function.
And it really does have an impact on real life. The women who had no cognitive decline, 86% of them were able to maintain their usual work and routine. And, if they did have cognitive decline, about 84% were either somewhat or unable to maintain their normal work.
And this cognitive impairment persists. When we tested these women one treatment after --- one year after treatment was completed, we found 45% had gotten better, but 45% had not, and 10% had a mixed pattern of --- of recovery, and their baseline level of performance, distress and demographic variables, were not related to the on-treatment decline.
this goes along with other retrospective trials from other institutions showing that persistent cognitive dysfunction can be seen even 10 years posttreatment.
Now we know chemotherapy effects, well, chemobrain has been described for a long time clinically and it wasn't necessarily accepted by oncologists as being a real effect of the chemo because it shouldn't cross the blood-brain barrier. But now we're seeing in animal models a better understanding of what these drugs are doing. So this is a lot of standard chemotherapy agents that are used for a variety cancer. And it turns out that, in animal models, it causes cognitive impairment. And it causes cell death in neurogenic regions, that is regions that are important for neurogenesis, which is important for memory. And I think the thing that is most surprising is that these agents are more toxic to nondividing oligodendrocytes, brain cells, than they are to cancer cell lines. So these agents are more toxic to brain cells than they are to cancer cells.
And in this groundbreaking paper looking at a single dose of ---, of a therapeutic dose of 5-FU in rodents, this caused acute CNS inflammation and vascular damage, which was expected, but the difference here was that they also found delayed myelin damage. So extensive myelin pathology occurred after the acute effects of 5-FU on the brain were seen. And this goes along with a study that we had recently done in breast cancer patients when we were following women after treatment for a period of time and noted persistence of cognitive impairments as I mentioned in about 45%, but then we were starting to see some women start to decline in cognitive function. And we actually did not publish this paper initially because we couldn't explain it and we felt that maybe it would be dismissed as an aberration or something. But, after this came out showing that there is a path --- pathological basis for it, we did publish this study in women with breast cancer.
Hormone therapy can also do it and, consistent with all of our other studies, we tend to see three groups of people pre and follow-up. One group gets better over time, which is the expected practice effect. One group has moderate neurotoxicity, which is very handleable, but they are declining over time. And there is a small subgroup of women who have severe neurotoxicity. And I can tell you every woman in that group was removed from tamoxifen and placed on an alternative drug.
So across cancer types and across treatments, we tend to see the same symptoms: mot --- memory, motor and executive dysfunction, which, as a neuropsychologist, implies a frontal subcortical dysfunction in the brain. And so we're now using functional magnetic resonance imaging to help determine what is actually happening in the brain. And we can also then use it to check intervention efficacy.
So, this is a an example, -- this is a paradigm developed by our colleagues at Emory University looking at -- the paradigm is that the brain loves novelty. So, if you have something predictable happening, nothing much happens; and if you have something unpredictable happening your brain responds pretty robustly. So in this case, the people --- these people have syringes in their mouth, one which shoots water in their mouth and one which shoots fruit juice in their mouth. And if it's predictable nothing happens, but if it is unpredictable which you're going to get, the frontal subcortical cortex lights up like a Christmas tree. And this is your brain on interferon which is used for hepatitis and also for some cancers, and actually nothing is happening. And if you look at the subtraction image you can see that areas of the brain that should be activated are under-activated. And areas of the brain that should not be activated are, which may help explain the symptoms that we are seeing in our patients.
So, why does this happen? What are the mechanisms? We don't know for sure, but there are a couple of lines of inquiry that we are following in our research.
One is hormonal. As I mentioned, in breast cancer patients, there was some suggestion that menopausal status might have been related. So if you are postmenopausal you were twice as likely to have cognitive impairment than if you were not menopausal.
Also, we are looking --- and this is probably the largest area of symptom research, is in the inflammatory response, the induction of cytokines. And we know that in sickness behavior, such as when you have the flu, you have fatigue, you have fuzzy thinking, loss of appetite, lack of motivation, and that's very similar to the symptoms that cancer patients experience. And sickness behavior is due to the fact that cytokines are induced to fight the infection, but they also go to the brain to induce this behavior, which helps you rest up so your body can heal.
And we know, for instance, this is in patients with AML or MDS, these are the level of circulating cytokines in their blood before treatment. 0 would be normal in a healthy normal person. And, as you can see, Interleukin-8 is 22 standard deviations on average elevated in their serum. Interleukin-6 is about 13 standard deviations and so on. So, there's very, very high levels of circulating cytokines in these patients.
And these are highly related to both cognitive problems and symptoms. So you can see here fatigue is highly related to interleukin-6 levels, executive dysfunction is also. Interference with daily life is highly related and so on. And so Interleukin-6, Interleukin-8, Tumor Necrosis Factor, and Interleukin-1 are all involved.
Also, some cancers cause an autoimmune response similar to type 1 diabetes when a child may have some kind of illness and the antibodies they mount for that illness also attacks the islets cells in their pancreas and they develop diabetes. We know that paraneoplastic syndromes occur especially in small cell lung cancer. The person's immune response to the cancer causes an antibody that also attacks the brain and, in small cell lung cancer, it's called the anti-Hu antibody. Now this can cause a very florid syndrome of paraneoplastic limbic encephalitis, where the person has psychiatric symptoms and amnesia, but low level titers --- low titer levels have been found in many cancer patients and, in fact what we found is that in 56% of the patients we looked at the serum, we found antineuronal antibodies. There's also other antineuronal antibodies associated with other cancers including anti-Yo in ovarian cancer, breast cancer, lymphoma and now testicular cancer also.
So, you have the cancer and you have the treatment, but now we have the person. And, as I mentioned earlier, some people have no symptoms whatsoever. Most patients have moderate symptoms that are distressing and some have extremely severe symptoms that really compromise their treatment.
And so we're looking at genetic polymorphisms in the patient to determine if they have any vulnerability to develop these symptoms. And interestingly we're looking at many of the genetic polymorphisms that are looked at in tumors to de --- help predict tumor response. So, we're looking at neurocognitive genes like ApoE4, if people who carry an ApoE4 allele are likely --- are more likely to develop Alzheimer's disease later in life. So we're looking at genes that may be related to other neurologic conditions. We're looking at inflammatory pathway genes. Now, if you have an allele of an inflammatory gene that causes a robost --- robust inflammation in your tumor, that might be good. But if you have that same allele in your normal brain that might not be so good. Also, the metabolizing genes, if you have the allele in your tumor that doesn't allow toxins to metabolize that might be good for your tumor. But again if you have that allele in your normal brain that might not be so good. We're also looking at DNA repair genes. If you have the allele of a DNA repair gene in your tumor that doesn't allow DNA to repair, that might be good for tumor response. But again if it's in your normal tissue it might be a bad deal. So, in the age of personalized medicine what I envision in - in the very near future is that, if you have cancer, your tumor will be genetically analyzed to determine what types of treatments might be best. But you will also be analyzed yourself to determine your specific vulnerability to develop symptoms so that we can be much more proactive about that.
And here, just to give you an example, these are brain tumor patients post-radiation, and if you have at-risk allele of this DNA repair gene, you're much more likely to have a decline over time in your memory function than if you don't have that gene.
And this is a metabolizing gene and again if you have the at-risk allele, you're much likely to decline over time than if you don't --- don't have that allele.
So, the predictors of cognitive impairment are different for different cancers, treatments, and people. So you have the person and their own genetic makeup, their own immune robustness and their nutrition and other things. You have the disease, which may cause an inflammatory response, how fast it's growing, if it causes hormonal problems. You have the treatment, which also can induce an inflammatory response. It can just poison you actually, or there may be a direct effect on the brain with a specific mechanism. And of course, the bottom-line is it's an interaction between the seed, the soil, and pesticides.
But now that we have this, what do we do about it? Well one is to identify and correct underlying medical conditions. We're often so focused on the cancer treatment that we forget a regular general internal medical evaluation. And our patients may be suffering from endocrine problems, thyroid problems, borderline anemia, diabetes, sleep disturbance; all of those are treatable and can improve symptoms.
We also can --- will use medications; methylphenidate or Ritalin® can be extremely helpful for people who have fatigue and difficulties with concentration and so on. And we're also looking at other stimulants like modafinil and so on. Cytokine antagonists to the degree that it's an inflammatory response are being looked at. In fact, for instance, IL-6 antagonist is in the early clinical trials as an antineoplastic agent and it would be very, very helpful if these trials also included cognitive and other symptom endpoints. Anti-inflammatory agents might also be helpful. And agents used to treat other neurologic diseases even though the --- the cognitive impairment associated with cancer is nothing like Alzheimer's disease, for instance, we're just trying everything we can, so we now have studies looking at memantine in brain met patients and donepezil in brain tumor patients after radiation.
We'd love to have neuroprotective strategies. That is, if we could identify people who are vulnerable, is there someway we can protect them from developing symptoms, such as using anti-inflammatory agents or erythropoietin agents? Erythropoietin is very interesting because, in addition to it - it's a treatment for anemia, there are a lot of Epo receptors in the brain and it is thought to be a neuroprotective agent and there have been some studies in stroke that show that it might be helpful. Unfortunately, it has been used in nonanemic cancer patients to improve fatigue and quality of life and it turned out survival might be shorter. So we have to be very careful that anything we do proactively for neuroprotection does not feed the tumor or undermine treatment.
However, there's animal models that are being done now that may very well help guide us in terms of neuroprotective strategies. This is a paradigm where rats are given either saline or they're given Adriamycin® and cyclophosphamide, fairly standard agents, or they're given the chemotherapy with an antioxidant. And they use a passive avoidance test. So what they do is: they put the rats --- there is a box here that's brightly lit or dark. They put the animal in the brightly lit part and it immediately runs into the dark, that's their nature. But then they get shocked, really good, and then the next day they are put in the light part again. If the rat had saline, they --- they went right in the box the first time, but they hung around a long time, the second time because they remembered it wasn't good in there. If the animals got the chemotherapy the next day they ran into the dark part of the box as fast as they did earlier. They had forgotten. And if they had gotten the chemotherapy with the antioxidant, they acted like they had saline. That is, they remembered that it wasn't good in there and they stayed out. Now the interes --- this was done in non-tumor bearing rodents and so we're hopefully going to take this to the next step. And we have collaborations with other institutions who have good animal models to do these kinds of studies in tumor-bearing animals, so we can also see the effect of these agents on the cancer and try to develop neuroprotective agents that work, but do not compromise treatment.
We do a lot of neuroprotective strategies for brain tumor surgery. Right now, we do preoperate --- preoperative functional imaging and dif --- diffusion tensor imaging to determine how the fiber tracts look and what parts of the brain are doing what, so the surgeon can see where the brain is working, vis-a-vis where the tumor is. We do a lot of intraoperative monitoring, that is, we test people awake during their surgery for a p --period of time. We're hoping maybe we can do something before surgery to reduce postsurgical morbidity. And we also provide for some patients prehabilitation instead of rehabilitation, that is, giving people strategies to cope before they start having problems, that is, teaching them compensatory strategies.
So this is an example of a functional imaging paradigm that we use here. This is a working memory paradigm, where the person is given a number of pictures to look at before they go into the scanner. And then, while they're getting their brain scanned, they're shown the pictures and they press a button when they recognize one. And so this activates visual cortex because the patient is looking at something, but it also activates an area in the right frontal lobe, which is responsible for working memory. So the surgeon can see where this is activated and where a tumor might be, so they can maximize their surgery and minimize the possible harm to the patient.
We also do a lot of intervention strategies that are behavioral. We do relaxation training for people who get overwhelmed easily, so they can kind of get a grip and calm down. Exercise is very important, not going to the gym exercise, I'm talking about minor exercise. There's a lot of good evidence out there that it is very helpful. For instance, one study showed that for people who are in germ-free isolation for bone marrow transplant, that 5 minutes on a stationary bike every day reduced length of hospital stay and medical complications. We do cognitive rehabilitation and compensatory strategy train -- training. And we also do alterations of the work environment, so, for instance, a courtroom lawyer who has word-finding problems and can't really function in the courtroom anymore, that person still has a lot of skills and talents that might be able to be used in a different venue. Also, we get accommodations in the workplace and in school because many of our patients are covered by the Americans with Disabilities Act. We also do vocational retraining and, of course, we're working with the caregiver milieu, employers and schools all the time.
What doesn't work is repetitive mental exercises or drills. Now this is very caché now, If you do Sudoku puzzles or video games, it should stave off dementia or whatever, and of course being mentally stimulated is a good thing. You can always be cognitively not at your optimum if you aren't stimulating yourself, but it doesn't undo chemobrain symptoms. For instance, I had a gentle - a young man who, when I was speaking at a bone marrow transplant survivor's conference, managed to get through Law School, but had failed his Bar Exam twice and he was quite nervous about that. And his wife was having him do Sudoku puzzles to try to help make his brain better, which he didn't even like doing. It was like doing Sudoku puzzles will not help you pass the Bar Exam. You really have to be much more organized about it. So doing video games or crosswords, what happens is you get better at that thing, but it doesn't stop you from blocking on a word in conversation. So if you enjoy doing it, do it, but it doesn't generalize to other aspects of your life.
So, in conclusion, a longitudinal design, that is testing people over time is essential, so we can attribute a cause of --- of a problem and determine a change within a specific individual. Cognitive outcome is particularly relevant in diseases where survival is not tremendously useful or we don't have good treatment, such as glioblastoma or pancreatic cancers. In fact, enhancing quality of life may be the main impact we have in certain populations, but the assessment tools need to be sensitive. You have to be able to catch it. And, of course, therapeutic interventions can improve function and quality of life.
And this is a really interesting area of research right now. When I started at MD Anderson over 25 years ago, I was kind of all by myself doing memory testing on patients and I --- it didn't go very far, but now it's a really interesting multidisciplinary area, where we're working with animal models and drug discovery, molecular epidemiology, imaging, and so on. It's an exciting area.
But I believe that optimizing the quality of life of cancer patients is essential and should be as proactive and aggressive as their anti-cancer therapy. Thank you very much for your attention and please let us know if this was helpful to you.
Cognitive Function of Cancer Survivors video
©2011 The University of Texas MD Anderson Cancer Center
1515 Holcombe Blvd, Houston, TX 77030
1-800-392-1611 (USA) / 1-713-792-6161