On Thursday, June 5, 2003 the National Cancer Institute sponsored a roundtable at the annual meeting of the PsychoNeuroImmunology Research Society (PNIRS) held June 3-7 in Amelia Island, FL. The roundtable served as a forum to discuss scientific issues related to the conduct of psychoneuroimmunology (PNI) in cancer control. Michael Antoni, Ph.D., Professor, Department of Psychology, University of Miami served as the roundtable facilitator. Dr. Antoni opened the forum by asking participants to share their views on the latest psychoneuroimmunology (PNI) paradigms and technological advances relevant to cancer-related processes and outcomes across the following areas:
- Cancer risk: What is the interaction between genetic risk for cancer and stress?
- Cancer initiation/promotion: What are the roles of PNI and psychoneuroendocrinology (PNE) in promoting a precancerous process to a clinically manifest cancer?
- Cancer treatment: Does stress affect ability to recover function from surgery or from chemotherapy? Within cancer treatment, what are the long-term effects of adjuvant therapy 5, 6, 7, or 8 years later on various surveillance functions such as opportunistic pathogens, quality of life, and other biological factors? Similarly, what are the long-delayed effects of chemotherapy as measured by factors such as illness behavior and cytokine regulation?
- Cancer recurrence: What psychological variables might affect metastases and cancer recurrence?
The Executive Summary covers the main points of the roundtable, recapping the participants' comments in the order that they were offered.
Looking at the Big Picture: Effects of PNI Intervention May Be Long-Term
Dr. Antoni began by describing work by his research team in Miami on recovery or reconstitution of immune functioning of cancer patients in the months and years after adjuvant therapy has stopped. The team recruits patients into studies before they start adjuvant therapy, applies randomized designs to study stress management effects during therapy, and finally tracks long-term effects of adjuvant therapy 5, 6, 7, or 8 years later. These post-tests or outcomes reveal the effects of adjuvant therapy on various surveillance functions such as quality of life, illness behavior, and the status of opportunistic pathogens and other biological factors.
Dr. Antoni referred to presentations, given in earlier sessions of this NCI and PNIRS joint meeting, on immune surveillance models of possible cancer recurrence, and tumor biology models–both models studying factors that affect the growth regulation of cancers. He suggested that the same psychological variables being examined in PNI models, where the researcher is "chasing" or surveying the cancer, might be highly relevant in terms of the life of the tumor itself and the cancer-related processes.
Further, Dr. Antoni expressed interest in applying the same concepts of research used by the NCI to his own research areas–how stress is related to a predisposition to cancer; and how PNI can affect the success of cancer treatment, the long-term effects of cancer treatment, and the probability of recurrence. He related that in the past he had become discouraged because of the obstacles to successful research in these areas. With members of the NCI participating at this Roundtable, he felt encouraged to invite comments that might incite interest and create opportunities once again.
The issue of timing came up again when Dr. Antoni further described the work done by his research team. The study randomized a group of women with breast cancer into two groups before chemotherapy or radiation treatment began–either into a 10-week stress management group or a 1-day control condition–then followed up at 3, 6, 9, and 18 months after the treatment had ended. The results revealed that, after the offset of chemotherapy, the women enrolled in the 10-week stress management group recovered into the normal range of natural killer (NK) cell functioning faster than those in the control group. Followup 5-8 years later will determine if the recovery curves of NK cell functioning in the months after chemotherapy and radiation are relevant to later disease progression, relapse, or recurrence.
After reporting that women in the stress management group recovered into normal NK cell functioning faster than those in the control group, Dr. Antoni challenged the Roundtable participants to question the clinical relevance of his findings. He asked if the study results opened a window of opportunity for people who need to recover immune functioning faster by the point in time addressed by the study. Underscoring his questions surrounding the most useful time to apply his discipline to cancer patients, Dr. Antoni asked if perhaps the study of peri-surgery would offer a better window of opportunity to provide more action in terms of protection against metastases. He reminded the participants that discussions surrounding these questions had taken place during the present forum.
In the study group, consisting of 350-400 stage 1-3 patients, tumor characteristics, the stage at which they are treated, and the type of treatment will be important to the findings. Dr. Antoni posed the following hypothetical questions: What would be the most desirable research objective of a grant to study factors beyond clinical manifestation of relapse or recurrence, 5-6 years post-treatment? Would there be disease activity worth studying continuously, inside of the 5 years?
The discussion turned briefly to the health sciences and economics area. One participant mentioned work in a similar intervention with breast cancer patients, which shows that people who went through the stress management group had 25 percent less billing for health care in the following 2 years.
Results of Group Intervention Reach Beyond Stress Management To Generate a Healthy Attitude Called Benefit-Finding
Asked about psychological predictors of recurrence in his studies, Dr. Antoni explained that there are personality questions to measure optimism and a construct called benefit-finding, two positive psychological indicators. Benefit-finding is an attitude that causes people who go through traumas–whether full-blown post-traumatic stress disorder (PTSD) related traumas, or more minor scares–to grow in a sense of meaning, re-prioritize their lives, enrich their sense of relationships with significant others, and establish more spiritual connections.
Questionnaires with benefit-finding scales have revealed that the psychological variables with the biggest effect size in breast cancer groups are in the benefit-finding score. NK cell recovery data are correlated and predicted by bigger increases in benefit-finding; and benefit-finding also has resulted in increased reductions in cortisol levels. These results have held at 1-year followup.
Dr. Antoni stressed that the control group's findings did not match those of the group undergoing the intervention, indicating that the experience of enduring and surviving diagnosis, surgery, and chemotherapy does not spontaneously increase positive feelings in survivors. Clearly, the experimental group that received intervention benefited from the group setting, and from sharing experiences of side effects and ideas for coping strategies. The intervention offered more than stress management–it offered camaraderie.
Dr. Antoni speculated that less energy devoted to vigilance and defense, and similar mobilizations, allowed the experimental group to have different patterns of hormonal and cytokine regulation that control NK cell functioning; or perhaps the outcomes are related to tumor biology. He remarked that Susan Lutgendorf, Ph.D., Associate Professor, Department of Psychology, University of Iowa, has looked at VEGF in terms of angiogenesis and found relationships with such constructs as social support. Dr. Antoni suggested that benefit-finding might open the door for other interesting constructs such as emotional expression and social support.
One participant spoke of markers. He said that psychiatric studies are applying functional brain imaging before and after therapeutic intervention, and comparing the results to pinpoint the anatomical change. In the years after the intervention, the anatomical change becomes a useful marker to follow in tracking survivals, diminished tumors, and recurrences for the researcher looking for a specific functional pathway–to see if that area of activity or change persists in survivors, or is diminished, or does not occur.
Shamgar Ben-Eliyahu, Ph.D., Professor of Functional and Applied Anatomy, Department of Psychology, Tel Aviv University, guided the discussion back to the subject of timeframes. Dr. Ben-Eliyahu expressed his concern about focusing on breast cancer to determine the degree to which the immune system is responsible for preventing recurrence or metastasis, because breast cancer is among the slowest progressing of the cancers. Most other malignancies progress faster. To track the role of events occurring around the time of breast cancer treatment into recurrence or metastasis would take 5-15 years.
Finding Correlations Between Peripheral Functions and Function in Tissue: NK Cell Changes in Breast Cancer May or May Not Be Relevant to Prevention
Another concern Dr. Ben-Eliyahu articulated was that the study of the NK cells in breast cancer may not be relevant to cancer prevention. These blood NK cells are very accessible as they are circulating in the periphery; but they are not the most potent in destroying circulating tumor cells or local micro-metastases in other organs. They come and go quickly from other tissues. Dr. Ben-Eliyahu asserted that, although the circulating NK cells appear active, they are very weak as compared to those which can more easily destroy tumor cells. Thus, the findings of increased NK cell activity in breast cancer may be useless in terms of cancer prevention.
The next participant to speak asked what would be the best way to model the relevance of the NK cells in the case of recurrence? He mentioned that the cancer did not have to be breast cancer, but offered breast cancer as the example because it probably has the longest history of PNI psychosocial funding.
Ronald Glaser, Ph.D., Associate Vice President for Research/Professor Medical Virology, Immunology, and Medical Genetics, The Ohio State University College of Medicine and Public Health, admitted that, whereas NK cell activity may not adequately reflect the situation in the tissues, it was the subject of much completed study for postsurgical treatment. Dr. Glaser pointed out that it was intriguing to see proof of behavioral interventions leading to restoration in the killing activity. On the other hand, the ability to kill K562 or even a breast cancer cell line may not be the most biologically important role for these cells. In a number of studies, it may be worthwhile to consider the release of certain cytokines related to NK cells and observe the effect of psychosocial interventions on the functionality of NK cells. Studying the circulating levels of some of the cytokines could be useful, and such observations could lead to other intermediate endpoints. Dr. Glaser suggested it might be worthwhile to observe serial serum specimens in a cohort of interest, such as that for cytokines or growth factors like inflammatory cytokines. He mentioned angiogenic factors that Dr. Lutgendorf had observed in ovarian cancer studies.
Cytokine Studies Indicate Correlations Between Peripheral and Tumor Site Activity and May Give Credence to Research on Peripheral Measures
Dr. Antoni spoke of changes seen at the cellular level in the cancer patients in his study. His research team looked at stimulated cells, then supernatants, TH1 and TH2 cytokines. During the period of recovery, when NK cell activity levels increased, cells from the people in the experimental group showed an increase in gamma interferon and a decrease in interleukin-4 (IL-4) in the supernatants, when the cells were stimulated with anti-CD3. Although studies of IL-12 had not yet been conducted, IL-2 and IL-10 showed little difference. Dr. Antoni said that the researchers did not look at the circulating serum levels because they often did not see much in them unless the cells were stimulated, either through intracellular staining or assaying the supernatants of the stimulated cells. He asked if any other participants were seeing serum levels of cytokines in cancer-treated patients.
A participant proffered an idea from the field of endocrinology. The participant said she had found that a breast pump could be used to collect breast fluid to assay cytokines to follow up local response, such as in the case of a lumpectomy or another condition that is tissue-specific enough to find regulation on the bilateral side.
The next participant to comment objected altogether to the discussion on NKa, maintaining that, despite the existence of some evidence, the oncology profession as a whole has rejected the study of NKa in cancer prevention. He emphasized the need to take a closer look at the oncology literature in terms of prognostic markers for tumors.
He then defined ANKA (autologous natural killer activity) as "killing one's own primary tumor by one's own NK cells." Although the study of ANKA requires primary cultures, which makes it more difficult to conduct than other studies, recent papers identify ANKA as a reliable prognostic marker with some positive developments–not in cancer situations, but in factors important to tumor growth. The speaker expressed an interest in reviewing cancer models to see if there are any PNI influences, just as VEGF and angiogenic factors are being reviewed.
Dr. Glaser pointed out that comparing NK cell activity to autologous activity is interesting, but logistically difficult, and he revisited the subject of cytokines. His team has recent data on several different circulating cytokines, particularly inflammatory ones that can be measured in the serum or plasma of cancer patients. He expressed his view that the processes used in his research are more biologically relevant to what is happening in vivo than inducing the cells with an artificial stimulant in vitro. For example, using anti-CD3 as a stimulant is not desirable because a blood specimen will be only selectively stimulated, and the NK cells will not be affected.
Dr. Lutgendorf then asked Dr. Glaser what cytokines he had found in the measurable level. Dr. Glaser responded that the main cytokines his team had looked at were IL-6 and TNF interferon, and that VEGF also could be measured in serum. Dr. Antoni mentioned that Dr. Lutgendorf had done work with the ex vivo model, taking the autologous tissue; he asked her if she had tried the NK model. Dr. Lutgendorf replied that, although it was a "difficult process," her team was testing NK activity from lymphocytes derived from tumors against K562 cells. This process is extremely labor intensive.
Dr. Antoni referred to a talk Dr. Lutgendorf had given earlier this spring describing findings of strong correlations between peripheral functions and functions in the tissue. The findings seemed to give credence to research on peripheral measures. Dr. Antoni asked Dr. Lutgendorf to review this subject.
Dr. Lutgendorf stated that she had presented the material on the previous day, but offered the comment that she was surprised to find relationships between psychosocial factors and IL-6 in the ascites fluid. She said that IL-6 in ascites could come from macrophages or from tumor cells. She stated that she found it interesting that their group had also found relationships between IL-6 produced by tumor cells and that found in plasma.
Dr. Antoni asked if there were any ideas about PNI research and cancer.
A participant said that two good studies, one by a Japanese team and another by a European team, show that myocyte response from peripheral blood of a patient against an autologous tumor is probably the best predictor for survival, better even than the stage of the tumor. He suggested that this might be one index to try and assess.
Timing of Therapy Matters: Physiological Stress Can Neutralize PNI Intervention Effects
The participant then spoke about different approaches to and timing of psychological intervention. Any kind of medical treatment, whether chemotherapy or surgery, produces a stress response in the body. If the psychological intervention is conducted simultaneously with the treatment, the offset of the physiological stress response may mask the full benefits of the intervention. Only by simultaneously attempting to block the stress responses from both psychological and physiological origins may benefits occur. Thus, in terms of cancer prevention, the most important time to try and prevent physiological stress responses would be during, or very close to, the time of psychological intervention.
The next speaker suggested that the reason more comments were not forthcoming at this Roundtable is that "people are depressed about ideas." He asked what research avenues even were worth pursuing, and referred to two seminal papers that appeared in the last few years–(1) Pamela Goodwin's paper, which, according to the speaker, largely eliminated the prospects of finding any effective influence of PNI on cancer survival; and (2) Maggie Watson's paper, which similarly cast doubt on the role of psychological factors on survival. The speaker asserted that, although the two researchers produced excellent work, their findings exhausted the approach paradigms of PNI and psychology, in terms of any future, interesting, human PNI cancer studies. He wondered what the next step was for human PNI research in cancer.
A participant responded that her team did an intervention study several years ago with relaxation therapy delivered between the time of diagnosis and surgery, a time period termed "existential crisis" by David Spiegel. She noted that the timing of the intervention indeed might be important.
Type of Therapy Matters: Some Patients Just Want To Be Left Alone
The participant commented that another aspect to consider might be the type of intervention chosen. Contrary to popular assumption, group therapy may not be good for all patients. She suggested tailoring the intervention to the individual. She is considering including full neo-personality measures for each person going into an intervention to ensure that the intervention is the right one.
Dr. Antoni commented further on the issue of matching the personality coping style to the best intervention. His research team is among those that use a large battery of personality tests. He said that his team is under pressure to test statistical interactions after showing randomized effects. If he finds reasonable interaction with the more reticent type of person (group interaction as an intervention is not a good match for this person) versus the very outgoing, social person who feeds off of group interaction, these findings might come through in the data. One challenge he has is in the power of his trials, as he has only 100-150 patients.
If a reasonably large interaction effect is detected, Dr. Antoni continued, the next step would be to try and match interventions to personality types. He identified this matching as a psychotherapy research issue. He asked which of the following two approaches would work better: (1) study individual (one-on-one) therapy versus group therapy, with the same intervention content; or (2) divide the group, and study relaxation therapy in one subgroup versus cognitive behavior therapy in another subgroup.
Dr. Antoni added that, although such studies are difficult to design because of the high number of disparate factors requiring a single standard of measurement, many researchers hope to conduct them in the future.
Antidepressants May Greatly Improve the Cancer Experience, If Not Survivability
The next speaker identified himself as a psychiatrist, and brought up a different type of intervention. He treats a variety of patients, including those with inflammatory bowel disease, heart disease, cancer, and a number of illnesses that clearly activate the immune system. He remarked that cancer also causes some disregulation of the immune system; and with chemotherapy and all the other interventions, cancer causes huge perturbations to the system.
The doctor proposed that what is effective in other illnesses may be helpful in cancer, specifically antidepressants. Double-blind, placebo-controlled trials are already showing antidepressants to be significantly helpful in post-MI patients, as well as people with a variety of inflammatory illnesses. There have been a few modest suggestions that antidepressants may be helpful to cancer patients.
He wished, however, to make the point stronger than a modest suggestion. Cancer is a large area that may respond significantly to peripheral impacts from outside influences, including pharmacological interventions. Antidepressants can have a central impact on the immune system and its interaction with the cancer. He invited the participants to transmit this suggestion to psychiatric colleagues in their departments, whom he believed were already interested in the concept of antidepressants in cancer treatment. As opposed to enormous, complicated, behavioral interventions, antidepressant therapy is something these colleagues can grasp and implement instantaneously; and they are experts at clinical trials. One of the participants remarked that there are people studying this area.
The next speaker stated that his team recently reviewed, in the literature on depression in cancer, a paper produced by a comorbidity symposium sponsored partly by the NIH and partly by private foundations. Although there was a general indication that antidepressant therapy is effective in cancer, very few double-blind studies have examined the efficacy of available treatments. He noted that survivability is a complicated subject, and that any interventions, psychotherapeutic or psychopharmacologic, are difficult to study.
Shifting Paradigms From Cancer Survival to Cancer as a Chronic Illness: PNI Therapy May Be Most Valuable at Endpoints More Proximal Than Survivability
The speaker expressed his failure to understand why improving quality of life in patients is not a sufficient endpoint. The purpose of PNI research in cancer traditionally has been to prove the correlation between mental well being/quality of life and survivability, with the goal of developing a PNI model for cancer prevention. Although the correlation probably does exist, the speaker felt strongly that research focused on more proximal endpoints–mood, fatigue, pain, and other types of behavioral symptoms–could be more valuable, certainly more successful in terms of applicable treatment. The speaker stated that his team had shifted its focus almost entirely to such proximal endpoints, addressing the interim symptoms in a kind of theoretical construct, based in a large part on the study of cytokines and sickness behavior.
He stated the need for these kinds of trials, and expressed the general disappointment in the field over the lack of double-blind, placebo-controlled trials of antidepressants and cancer patients. He endorsed the idea of the pharmacological approach, noting the efficiency of and the comfort level of patients with such therapy. Underscoring his point, he stressed how difficult it was for the average cancer center to implement complex behavioral interventions. It is much easier to instruct a doctor to give the patient a pill based on scores on a scale. Prescribing medication is something that physicians understand, relate to, and can do.
The next participant to comment pointed out that antidepressant therapy does not last forever. Antidepressants could be given in acute trials, with 1-2 year followup. With survivability low and recurrence high, people may enthusiastically cooperate if the therapy is transmitted properly in double-blind, placebo-controlled trials. Patients would probably be enthusiastic, as well as apprehensive, although he said that any apprehension could be dissipated by the reassurance that no increased harm is being done. Then he added that nothing is perfect; he said the issue is complicated, but the treatment is doable.
The next participant spoke about survival and PNI, stating bluntly that defining PNI research in cancer in terms of group intervention's effect on survival outcomes is a big mistake. He asserted that the goal should be a shift in paradigm, converting cancer in research terms to a chronic illness, where PNI could offer tremendous help in a host of quality-of-life issues, such as wound healing, nausea, fatigue, and sickness behavior; and infectious processes that are very relevant to cancer control. The speaker referred to Pam Goodwin and others' studies which have dampened the enthusiasm of researchers for work in PNI and survival, and he pointed out that there are still methodological questions surrounding the research. He expressed his belief that PNI is relevant in many other areas very applicable and valuable to cancer patients.
On this positive note, Dr. Antoni adjourned the meeting, and thanked everyone for his or her attendance and participation.
Note: Writer, Ellen Sanders of the Scientific Consulting Group, Inc, prepared the executive summary.