Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Oct 14.
Published in final edited form as: Biol Res Nurs. 2014 Jan 28;16(4):429–437. doi: 10.1177/1099800413519494

Baseline Immune Biomarkers as Predictors of MBSR(BC) Treatment Success in Off-Treatment Breast Cancer Patients

Richard R Reich 1,2, Cecile A Lengacher 2,3, Kevin E Kip 3, Steven C Shivers 4, Michael J Schell 2,5, Melissa M Shelton 3, Raymond H Widen 6, Catherine Newton 5, Michelle K Barta 3, Carly L Paterson 3, Jerrica R Farias 3, Charles E Cox 4,5, Thomas W Klein 5
PMCID: PMC4604564  NIHMSID: NIHMS725863  PMID: 24477514

Abstract

Researchers focused on patient-centered medicine are increasingly trying to identify baseline factors that predict treatment success. Because the quantity and function of lymphocyte subsets change during stress, we hypothesized that these subsets would serve as stress markers and therefore predict which breast cancer patients would benefit most from mindfulness-based stress reduction (MBSR)-facilitated stress relief. The purpose of this study was to assess whether baseline biomarker levels predicted symptom improvement following an MBSR intervention for breast cancer survivors (MBSR[BC]). This randomized controlled trial involved 41 patients assigned to either an MBSR(BC) intervention group or a no-treatment control group. Biomarkers were assessed at baseline, and symptom change was assessed 6 weeks later. Biomarkers included common lymphocyte subsets in the peripheral blood as well as the ability of T cells to become activated and secrete cytokines in response to stimulation with mitogens. Spearman correlations were used to identify univariate relationships between baseline biomarkers and 6-week improvement of symptoms. Next, backward elimination regression models were used to identify the strongest predictors from the univariate analyses. Multiple baseline biomarkers were significantly positively related to 6-week symptom improvement. The regression models identified B-lymphocytes and interferon-γ as the strongest predictors of gastrointestinal improvement (p < .01), +CD4+CD8 as the strongest predictor of cognitive/psychological (CP) improvement (p = .02), and lymphocytes and interleukin (IL)-4 as the strongest predictors of fatigue improvement (p < .01). These results provide preliminary evidence of the potential to use baseline biomarkers as predictors to identify the patients likely to benefit from this intervention.

Keywords: mindfulness-based stress reduction, breast cancer, biomarkers

Following breast cancer treatment, women are at high risk for residual psychological symptoms of stress, anxiety, depression, fear of recurrence, and impaired cognitive functioning as well as physical symptoms of pain, fatigue (FA), and sleep disturbances, all of which can negatively impact quality of life after treatment (Dodd, Cho, Cooper, & Miaskowski, 2010; Von Ah, Habermann, Carpenter, & Schneider, 2013). These symptoms tend to cluster together and may have natural associations, similar shared pathways, and common underlying mechanisms (Dodd et al., 2010). In a study specific to breast cancer patients, Lengacher, Reich et al. (2012) identified three-symptom clusters among breast cancer patients: gastrointestinal (GI), cognitive/psychological (CP), and FA. They also demonstrated that a mindfulness-based stress reduction (MBSR) program (specific to breast cancer; MBSR intervention for breast cancer survivors, MBSR[BC]) was effective in reducing some symptoms. MBSR has been demonstrated to be a robustly efficacious treatment for relief of both psychological and physical symptoms in numerous conditions (Bohlmeijer, Prenger, Taal, & Cuijpers, 2010; Grossman, Niemann, Schmidt, & Walach, 2004; Hofmann, Sawyer, Witt, & Oh, 2010; Ledesma & Kumano, 2009).

As with most psychotherapeutic treatment studies, however, not all MBSR(BC) patients studied by Lengacher et al. (2012) experienced symptom reduction. This variability in results prompted an interest in identifying important patient characteristics that anteceded MBSR(BC) treatment and might help to predict its effects. Discovering baseline markers that identify patients most likely to benefit from treatment has long been a goal of “patient-centered” medicine (Simon, 2008). These identified markers increase the chances that patients will get the treatments that benefit them while decreasing the unnecessary burden of treatments unlikely to work. Typically, in previous patient-centered approaches, investigators have explored biological markers as predictors of medical treatment success (Wistuba, Gelovani, Jacoby, Davis, & Herbst, 2011), while psychosocial factors have been used to predict psychotherapeutic treatment success (Project MATCH Research Group, 1998). In the present study, we took a novel approach by examining baseline immunological biomarkers as predictors of which women who had recently completed breast cancer treatment would most benefit from a psychotherapeutic treatment.

We selected immunological biomarkers because (a) changes in these markers are related to stress (Graham, Christian, & Kiecolt-Glaser, 2006), (b) they have extensive links with symptoms of cancer and cancer treatment (detailed below), and (c) they are easily obtainable. Several biomarkers have been correlated with both physical and psychological symptoms experienced by cancer patients on and off treatment (Gardner, 1999; Kang et al., 2009; van der Most, Currie, Robinson, & Lake, 2006). Changes in immune function and associated bio-marker expression in cancer survivors resulting from psychological distress may directly or indirectly affect survival outcomes (Andersen et al., 2007; Sephton et al., 2009). Additionally, chemotherapy (CT) and radiation therapy (RT) may affect immune recovery and survival outcomes (Kang et al., 2009). RT significantly reduces the levels of white blood cells, lymphocytes, and natural killer (NK) cell activity (Koukourakis, Zabatis, Zacharias, & Koukourakis, 2009; Standish et al., 2008; Yamazaki et al., 2002). Moreover, alterations in immune biomarkers have been shown to persist following treatment (Kang et al., 2009; Mozaffari et al., 2009; Standish et al., 2008). Among breast cancer patients receiving CT and RT, Standish et al. (2008) found decreases in the levels of lymphocytes, NK cell activity, monocyte phagocytic activity, and TNF-α production following RT compared to post-CT; interferon-γ (IFN-γ), however, was unchanged. Among 80 early-stage breast cancer patients who had completed treatment up to 1 year prior to the study, Kang et al. (2009) found that interleukin (IL)-6 and most cluster of differentiation (CD) cell subsets (except CD4+) recovered faster than did NK cell activity, levels of IL-2, IL-4, IFN-γ, and lymphocyte proliferation; CT alone or CT combined with RT significantly delayed IL-2 recovery (T helper (Th)1-type response), while IL-4 (Th2-type response) recovery was significantly delayed in the RT-only group. Studies have also shown that NK cell number and activity and T-cells are reduced following breast cancer treatment (Mozaffari et al., 2007; Solomayer et al., 2003). Mozaffari et al. (2007, 2009) found a lower number of CD4+ cells and lower expression of CD28+ on T cells during extended follow-up for breast cancer patients who received RT and CT; however, IFN-γ, IL-2, and IL-4 were higher in RT and CT patients versus RT-only patients.

The purpose of the present study was to identify baseline immune subsets, as markers of stress, that could be used to indicate patients who would most benefit from MBSR(BC) before the onset of treatment. This study can be considered a “developmental” study, as designated by Simon (2008), because the goal was to identify, in an exploratory manner, candidate bio-markers from a small sample. As such, it is the first step on the path to “validation” studies with specific hypotheses regarding specific biomarkers. Although this study may be considered developmental, it is not completely exploratory, as we chose the biomarkers to measure based on previously demonstrated links to physical and psychological symptoms following cancer treatment (Bower et al., 2009; Wang et al., 2012).

Materials and Methods

Sample and Setting

We recruited 41 female breast cancer survivors, diagnosed with Stage 0, I, II, or III breast cancer, who had undergone surgery (lumpectomy) and received adjuvant RT or RT and CT from the Moffitt Cancer Center. All women had completed treatment 2–12 weeks prior to enrollment. Women were excluded if they (a) had been diagnosed with Stage IV breast cancer, (b) had been treated for a breast cancer recurrence, (c) had undergone a mastectomy, or (d) had severe psychiatric problems. The Institutional Review Board at the University of South Florida in Tampa, Florida, approved the study protocol. Participants received US$100 for participation, US$50 at the beginning and US$50 at the completion of the study.

Study Design and Random Assignment

Using a two-armed randomized controlled design, we randomly assigned participants to either a 6-week MBSR(BC) program or a wait-listed usual care (UC) group. Women in the UC group could elect to receive the MBSR(BC) intervention after 6 weeks (upon study completion). Subjects were stratified by stage of cancer (0, I, II, or III) and type of treatment (RT alone or RT and CT).

Data Collection Procedures

At the orientation session, all participants provided written informed consent, and we collected baseline data on demographics and symptoms. In the MBSR(BC) group, we collected blood for immune analyses at the orientation session prior to the start of the intervention and again at a follow-up session for MBSR(BC) participants within 2 weeks of completion of the 6-week intervention. In the UC group, we collected data and blood samples at baseline during the orientation session and again 6–8 weeks from baseline at a follow-up session for UC participants. We also collected data on symptoms for the MBSR(BC) and UC groups at their respective follow-up sessions.

Intervention and UC

MBSR(BC) is a 6-week program adapted from Kabat-Zinn’s original 8-week MBSR program (Kabat-Zinn, Lipworth, & Burney, 1985) to address concerns of breast cancer survivors by training them to use mindfulness and attention to adapt to emotional/psychological symptoms (anxiety, depression, and fear of recurrence) and physical symptoms such as FA, pain, and sleep disturbance. A psychologist trained in MBSR taught the weekly 2-hr sessions. Participants received a manual that guided the classes. The MBSR(BC) program includes three components: (a) educational materials and exercises related to meditation practices and the mind–body connection, (b) practice time and a CD on which verbal support for four meditative practices was recorded (sitting meditation, walking meditation, body scan, and gentle hatha yoga), and (c) opportunity for group discussion, including time to answer questions related to barriers to formal and informal practice. Participants were asked to spend 15–45 min daily outside of the group sessions on formal and informal practice and to record their practice time in a diary. Formal mindfulness practice includes sitting meditation, walking meditation, body scan, and yoga practice. Informal mindfulness meditation practices, or mindfulness in everyday life, incorporate an awareness of pleasant and unpleasant events and encourage awareness of routine activities.

The UC group had standard posttreatment clinic visits. We specifically asked women in this group not to use or practice meditation, yoga techniques, or MBSR during the study. Following completion of the study, we provided each UC participant with a brief orientation to MBSR(BC) along with a manual and CDs of the complete program and offered them optional scheduled classes.

Measures

Symptoms

We measured symptoms using the M. D. Anderson Symptom Inventory (MDASI; Cleeland et al., 2000). The MDASI measures the severity of symptoms and the degree to which they interfere with daily functioning. The instrument includes 2 items for each symptom, one reflecting the last 24 hours and the other the last 7 days. Respondents are asked to rate the severity of 13 core symptoms from 0 (not present) to 10 (pain as bad as you can imagine). There are also 6 symptom interference items in which respondents rate the level at which symptoms interfere with function. Internal consistency has been shown to range from 0.82 to 0.94 (Cleeland et al., 2000). Use of the MDASI has been supported in studies using patients with multiple cancer diagnoses from multiple populations (Chen & Lin, 2007; Ivanova et al., 2005; Kwon et al., 2006; Okuyama et al., 2003; Tseng, Cleeland, Wang, & Lin, 2008; Wang et al., 2004, 2006). Using the “last 24 hr” items from the MDASI, Lengacher, Reich, et al. (2012) identified three central symptom clusters in breast cancer patients: (a) GI (dry mouth, nausea, vomiting, lack of appetite), (b) CP (trouble remembering, distress, and sadness), and (c) FA (disturbed sleep, and drowsiness). To calculate cluster scores, we categorized items by symptom cluster and summed responses to items included in each cluster.

Peripheral blood lymphocytes

Peripheral blood samples were collected between 9 a.m. and noon into heparinized tubes and processed immediately for lymphocyte analysis. Heparinized blood (0.1 ml) was mixed with 2.0 ml of IOTest 3 lysing solution (Beckman Coulter) to lyse red blood cells, and 1.0 ml of the mixture was counted in a Vi-Cell XR cell viability analyzer (Beckman). The number of viable cells per microliter (μl) was calculated for each sample. The percentage of lymphocyte subpopulations (subsets) in peripheral blood samples was determined by antibody staining and flow cytometry as described previously (Lengacher et al., 2013). Briefly, heparinized blood samples were stained and incubated in two different tubes with antibodies (BD Bioscience) to either CD45, CD4, CD8, and CD3 (Tube 1) or CD45, CD19, CD3, and CD56 (Tube 2). In a third tube, cells were stained with antibodies (BD Bioscience) to the surface antigens CD3 and CD69 (an activation marker) and then processed with Cytofix/Cytoperm (BD Bioscience) for intracellular staining with antibodies to IFN-γ (Th1 T cells) and IL-4 (Th2 T cells).

The stained samples were then analyzed on an FACScalibur flow cytometer (Beckton Dickinson) using MultiSet software, and the percentages of CD3+, CD4+, and CD8+ T cells, CD19+ B cells, and CD56+ NK cells were determined. The absolute number of cells in each subset was determined using these percentages combined with the number of viable white blood cells determined as described previously.

The percentages of Th1 (CD3+CD4+IFN-γ+) and Th2 (CD3+CD4+IL-4+) subsets were also determined on patient blood samples stimulated with two different mitogens and stained for intracellular cytokines as described previously (Lengacher et al., 2013). Briefly, blood samples were incubated for 1 hr with either phorbol myristate acetate (5 ng/ml, Sigma Chemical) and ionomycin (500 ng/ml, Sigma) or with PHA (phytohemagglutinin; 20 μg, Sigma) and then incubated for an additional 4 hr with mitogen and brefeldin A (1 ug/ml, BD Bioscience) to prevent secretion of cytokines from the stimulated cells and thus build up the intracellular levels of IFN-γ and IL-4. Next, the cells were stained with antibodies (BD Bioscience) to the surface antigens CD3 and CD69 (an activation marker) and then processed with Cytofix/Cytoperm (BD Bioscience) for intracellular staining with antibodies to IFN-γ (Th1 T cells) and IL-4 (Th2 T cells). Stained cells were analyzed by flow cytometry and Cellquest software (BD Bioscience) for CD3+CD69+ cells that were also either IFN-γ+ or IL-4+.

Demographics, medical history

We administered standard demographic data and detailed clinical history forms at baseline and recorded any changes to the data at 6 weeks.

Statistical Analyses

Baseline characteristics (demographics and immune biomarker levels) were compared between MBSR(BC) and UC conditions using the χ2 test or Wilcoxon rank sum test (for categorical and continuous variables, respectively). Pre–post symptom differences were analyzed using the Wilcoxon signed ranks test. To calculate symptom change, posttreatment symptom cluster scores were subtracted from pretreatment scores.

The analyses to identify baseline biomarkers that predicted 6-week symptom cluster change were run for both MBSR(BC) and UC patients to determine whether baseline immune biomarkers predicted improvement in general or only MBSR(BC)-specific improvement. Identifying baseline bio-markers that predicted 6-week change in symptom clusters was a two-step process. First, bivariate Spearman correlations were calculated between symptom cluster change scores and each baseline biomarker value. Spearman correlations were chosen because of the tendency for the biomarkers not to be normally distributed (most were highly skewed). Because these initial steps to build prediction models were designed to catch all relevant variables, candidate predictors were identified with the relatively liberal p value < .10 (Hosmer & Lemeshow, 2000). Second, to identify the strongest predictors (those that explained the most independent variance in symptom change scores), a backward elimination regression model was created for each of the symptom clusters. To reduce the aforementioned problem of skew in the biomarkers, each was log and square root transformed. The transformation yielding the lowest skew was chosen for the regression models. All candidate bio-markers were entered in the models and subsequently removed if their p value was greater than .05. Cancer treatment (RT or RT and CT) and time since cancer treatment also were entered into the regression models as covariates because previous studies have demonstrated differential immune biomarker responses by cancer treatment (Kang et al., 2009; Koukourakis et al., 2009; Mozaffari et al., 2007, 2009; Solomayer et al., 2003; Standish et al., 2008; Yamazaki et al., 2002).

Results

Demographic characteristics and immune biomarkers did not differ between UC and MBSR(BC) groups at baseline (Tables 1 and 2). Also, the baseline biomarker levels we obtained in the present study are similar to those other researchers have reported for similar populations. For instance, the number of lymphocytes averaged approximately 1,000/μl of blood (see Table 2), which is within the range of 490–6,000/μl (median of 2,000/μl) as reported by Park and Han (2012). The percentages of CD3+ lymphocytes (~80%), CD16+/CD56+ cells (~10%), and CD19+ cells (~10%) were also in line with other reports (Abud-Mendoza et al., 2012; Carlson, Speca, Faris, & Patel, 2007; Park & Han, 2012). Regarding the ratio of Th1 to Th2 cells, others have reported ratios from 12.5 (Toldi et al., 2011) to 5.1 (Cseh et al., 2012), while we observed ratios ranging from 3.4 to 9.1 in the present study, depending upon the group. So, although all the women in the study received either RT only or RT combined with CT, participants as a whole showed lymphocyte subset values within the normal range upon entering the study, and analysis of the baseline data by cancer treatment type showed no significant differences.

Table 1.

Demographic Characteristics of Female Breast Cancer Survivors at Baseline.

Variable UC (n = 24) MBSR(BC) (n = 17) p value
Age, years (mean, SD) 58.2 (9.5) 58.0 (10.3) .90
% minoritya 29 18 .40
% married 50 77 .08
% employed 54 59 .44
Stage of disease % .25
 0 25 6
 I 50 71
 II 21 12
 III 4 12
% who received chemotherapyb 30 35 .68
Open in a new tab

Note. MBSR(BC) = mindfulness-based stress reduction program for breast cancer survivors; UC = usual care.

a

The category “minority” included White-Hispanic, Black non-Hispanic, Ashkenazi Jew, and Native American.

b

All participants received radiation treatment.

Table 2.

Mean (SD) Levels of Biomarker Lymphocyte Subsets Among Female Breast Cancer Survivors at Baseline.

Biomarker Number of cells/μl of blood
p valuea
UC (n = 24) MBSR(BC) (n = 17)
Lymphocyte subsets
 CD3+ 815 (215.6) 809 (382.4) .57
 CD8+ 247 (110.0) 271 (150.7) .84
 CD4+ 554 (174.8) 527 (265.5) .45
 CD4+CD8+ 11 (6.8) 13 (8.8) .51
 CD16+/CD56+ ratio (NK cells) 121 (74.0) 112 (57.0) .72
 CD19+ (B cells) 89 (37.4) 77 (52.5) .19
Total lymphocytes 1054 (268.2) 1031 (450.8) .44
CD4+/CD8+ ratio 2.7 (1.7) 2.3 (1.5) .40
Mitogen-stimulated subsetsb
 PHA stimulated
  CD3+ 62 (11.4) 60 (14.5) .97
  CD3+CD69+ (activated) 31 (14.1) 29 (17.8) .58
  CD3+IFN-γ+ (Th1) 11 (10.2) 12 (9.2) .44
  CD3+IL-4+ (Th2) 1.3 (1.13) 1.5 (1.42) .85
 PMA/IO stimulated
  CD3+ 70 (11.4) 72 (9.8) .62
  CD3+CD69+ (activated) 51 (24.5) 52 (26.5) .80
  CD3+IFN-γ+ (Th1) 22 (12.6) 21 (9.7) .89
  CD3+IL-4+ (Th2) 6.6 (6.8) 4.9 (4.1) .72
Open in a new tab

Note. SD = standard deviation; CD = cluster of differentiation; IFN = interferon; IL = interleukin; MBSR(BC) = mindfulness-based stress reduction program for breast cancer survivors; NK cells = natural killer cells; PHA = phytohemagglutinin; PMA/IO = phorbol myristate acetate/ionomycin; Th1 and Th2 = T helper 1 and 2 response; UC = usual care.

a

All comparisons made using Wilcoxon rank-sum test.

b

All mitogen-stimulated subsets are percentage of T cells, with the exception of CD3+, which is percentage of lymphocytes.

Symptom cluster scores tended to go down across both UC and MBSR(BC) groups for all three clusters (Table 3). These within-group decreases were significant, however, only in the MBSR(BC) condition for the fatigue cluster (p = .003) and in the GI cluster (p = .035).

Table 3.

Symptom Cluster Scores at Baseline and 6 Weeks.

Symptom cluster (mean score)a UC (n = 24)
MBSR(BC) n = 17
Baseline 6 weeks Baseline 6 weeks
Gastrointestinal 6.0 (8.1) 5.5 (8.1) 4.5 (5.4) 1.9 (3.5)**
Cognitive/psychological 9.1 (8.7) 7.8 (8.5) 7.1 (10.1) 4.3 (5.1)
Fatigue 9.1 (7.1) 7.3 (6.0) 8.8 (6.9) 4.6 (5.3)*
Open in a new tab

Note. MBSR(BC) = mindfulness-based stress reduction program for breast cancer survivors.

a

Symptom-cluster scores were from the M. D. Anderson Symptom Inventory.

*

Within-group difference (p < .05).

**

Within-group difference (p < .01).

Although groups did not differ on mean baseline biomarker levels, we did note substantial within-group variability, as indicated by the standard deviations (Table 2). We assessed the relationship between this variability and 6-week symptom change scores using Spearman correlations. For the UC group, two bio-markers were significant univariate predictors of improvement in the GI symptom cluster (number of CD4+CD8+ T cells; PHA-stimulated CD3+ [% of lymphocytes]), but none were predictors of improvement in the other clusters. For the MBSR(BC) group, five biomarkers were significant univariate predictors of GI symptom cluster improvement, one was a significant predictor of CP symptom cluster improvement, and eight were significant predictors of FA symptom cluster improvement (Table 4). Although the CP cluster had only one significant predictor, number of CD4+CD8+ T cells, the correlation between this predictor and cluster (r = .61) was the strongest when compared to all other predictors.

Table 4.

Correlations (Spearman’s ρ) Between Baseline Biomarkers and Symptom Cluster Improvement by Condition (Usual Care [UC; n = 24] vs. Mindfulness-Based Stressed Reduction for Breast Cancer Program [MBSR(BC); n = 17]) at 6 Weeks of Treatment Among Female Breast Cancer Survivors.

Biomarker Symptom cluster
Gastrointestinal
Cognitive/psychological
Fatigue
UC MBSR(BC) UC MBSR(BC) UC MBSR(BC)
Lymphocyte subsets (# of cells/μl of blood)
 CD3+ −.16 .44* −.31 .31 .10 .58**
 CD8+ .17 .38 .05 .38 −.01 .54**
 CD4+ −.11 .37 −.26 .17 .17 .50**
 CD4+CD8+ .40* .14 −.10 .61*** .07 .44*
 NK cells −.02 .18 −.26 −.20 .00 .21
 B lymphocytes −.21 .44* −.16 −.01 −.01 .34
 Total lymphocytes −.07 .42* −.24 .27 .17 .58**
Mitogen-stimulated subsetsa
 PHA stimulated
  CD3+ −.38* .39 .13 .13 .00 .32
  CD3+CD69+ (activated) −.31 −.25 .00 .04 −.02 −.08
  CD3+IFN-γ+ (Th1) .08 .53** .14 .33 .01 .34
  CD3+IL-4+ (Th2) .18 .43 −.14 .27 .28 .54**
 PMA and IO stimulated
  CD3+ −.22 .43* −.08 .18 .07 .47*
  CD3+CD69+ (activated) .18 −.16 −.03 .26 .23 .08
  CD3+IFN-γ+ (Th1) .12 .37 .04 .18 .03 .40
  CD3+IL-4+ (Th2) −.07 .37 −.02 .35 .22 .48*
Open in a new tab

Note. UC = usual care; MBSR(BC) = mindfulness-based stress reduction program for breast cancer survivors; CD = cluster of differentiation; IFN = interferon; IL = interleukin; NK cells = natural killer cells; PHA = phytohemagglutinin; PMA/IO = phorbol myristate acetate/ionomycin; Th1 & Th2 = T helper 1 or 2 response.

a

All mitogen-stimulated subsets are percentage of T cells, with the exception of CD3+, which is percentage of lymphocytes.

*

p < .10.

**

p < .05.

***

p < .01.

Our initial plan was to introduce all biomarkers that had demonstrated significant univariate prediction capability into the multivariate regression models predicting symptom cluster improvement following MBSR(BC). The fact that several of the biomarkers were highly intercorrelated, however, introduced the potential problem of multicollinearity into our models. Therefore, when two biomarkers were highly correlated with each other (r > .50), we introduced only the strongest predictor into the model. For the GI symptom cluster, two predictors remained statistically significant following the backward elimination procedure: B lymphocytes and PHA-stimulated CD3+, IFN-γ+ [Th1]. The R2 for this two-predictor model was .65, meaning the baseline values of these two biomarkers accounted for 65% of the variance in GI symptom cluster change (Table 4). Two predictors also remained statistically significant for the FA cluster—total lymphocytes and PHA-stimulated CD3+, IL-4+ (Th2)—with an R2 value of .56. For the CP cluster, one predictor remained statistically significant—CD4+ CD8+—with an R2 of .31. Cancer treatment and time since cancer treatment, included as covariates in these regression models, were not significant predictors of symptom cluster improvement in any of the three regression models. As a result, we removed these variables in the backward elimination regression procedure. Final regression models are summarized in Table 5.

Table 5.

Backward Elimination Regression Models Predicting Symptom Cluster Change in Female Breast Cancer Survivors Receiving Mindfulness-Based Stress Reduction for Breast Cancer (MBSR[BC]) Treatment.

Predictorsa p value Modelb R2 Model p value
Gastrointestinal cluster
 B lymphocytesc < .001
 PHA-stimulated CD3+IFN-γ+ .013
  (Th1)c .65 <.001
Cognitive/psychological cluster
 CD4+CD8+ .012
.31 .012
Fatigue cluster
 Total lymphocytesd .008
 PHA-stimulated CD3+IL-4+ .007
  (Th2)d .56 .003
Open in a new tab

Note. CD = cluster of differentiation; IFN = interferon; IL = interleukin; PHA = phytohemagglutinin; Th1 & Th2 = T helper 1 or 2 response.

a

Predictors were retained in model if p < .05.

b

Adjusted R2.

c

Log transformed.

d

Square root transformed.

Discussion

The findings of the present study introduce the possibility of predicting the success of MBSR(BC) treatment using biomarker values obtained at the time of study entry. The central purpose of the study was to identify baseline immune subsets, as markers of stress, that could be used to indicate patients who would most benefit from MBSR(BC) before the onset of treatment. Our study concentrated on lymphocyte subpopulations as biomarkers among two groups of female breast cancer survivors, one of which received UC while the other participated in an MBSR(BC) program. Despite the similarity in baseline measures between the groups, substantial individual variation was apparent in the data. We tested these variations as predictors of MBSR(BC) symptom reduction in a sample where MBSR(BC) treatment had already demonstrated efficacy (Lengacher, Kip, et al., 2012).

Several biomarkers were significant univariate predictors of symptom cluster improvement, such as higher values for total lymphocytes, T cells, B cells, and CD4+CD8+ precursor cells as well as increased potential to respond to a relatively weak stimulus like PHA (Table 3). Of these, CD3+, CD4+CD8+, and total lymphocytes were significant univariate predictors for more than one cluster (e.g., CD4+CD8+ cells were associated with all three-symptom clusters; Table 3). After using backward elimination procedures, we developed multivariate regression models (Table 4) to define the cell subsets that were the strongest independent predictors for improvement in each symptom cluster after MBSR(BC). However, we did not find any significant predictors that were common across all of the symptom clusters.

At this point, it is only possible to speculate as to why different subsets should uniquely predict improvement in a given symptom cluster. Perhaps all the subsets we have observed to have significant associations with symptom cluster improvement are representative of a recovering immune response, from increased total lymphocytes and CD4+CD8+ naive cells to increases in circulating B cells and increases in T cells capable of responding to PHA. Thus, it is possible that the immunity of some cancer patients recovers better than that of others, rendering them better able to benefit from MBSR. This recovering immune response could result in both elevated B lymphocytes in peripheral blood and increases in the ability to produce the Th1 cytokine IFN-γ after stimulation of the patient’s T cells in vitro in the presence of underlying infections against which antibodies are being produced. It makes sense that patients experiencing GI symptoms, especially nausea and vomiting, would have elevations in immunologic cells associated with antibody production. Breast cancer patients who have undergone RT and/or CT often have such symptoms that are related to sensitivity of the cells lining the GI tract to these agents and/ or an underlying immune dysregulation (Andreyev, Davidson, Gillespie, Allum, & Swarbrick, 2012).

The number of baseline CD4+CD8+ T cells was associated with some type of improvement for all three of the defined symptom clusters (Table 3). These double-positive cells with cell-surface markers that are normally mutually exclusive are usually detected only in very small numbers in the peripheral blood (Parel & Chizzolini, 2004). Investigators have observed increases in CD4+CD8+ T cells in patients under a variety of immunologic stresses, such as chronic viral infections (Nascimbeni, Pol, & Saunier, 2011; Nascimbeni, Shin, Chiriboga, Kleiner, & Rehermann, 2004; Weiss et al., 1998) and autoimmune diseases (Parel & Chizzolini, 2004). Although the origin and function of CD4+CD8+ T cells in the peripheral blood remains controversial, there is some evidence that they represent mature antigen-specific effector memory cells that are involved in latent and/or active immune responses (Nascimbeni et al., 2011).

Our results also showed that, along with patients showing elevated levels of total lymphocytes in the peripheral blood, patients whose T cells were able to produce IL-4 (Th2 response) after stimulation in vitro derived the most benefit from MBSR(BC) with respect to fatigue symptoms. Collado-Hidalgo, Bower, Ganz, Cole, and Irwin (2006) demonstrated that decreased activated (CD69+) T lymphocytes in the peripheral blood is highly diagnostic of fatigue in breast cancer survivors. Other studies have demonstrated a bias toward Th2-type responses, particularly IL-4 production after polyclonal stimulation, in patients with chronic fatigue syndrome (Schwarz, Chiang, Muller, & Ackenheil, 2001; Torres-Harding, Sorenson, Jason, Maher, & Fletcher, 2008). Although the results of these studies are consistent with our findings, the mechanistic nature of the relationship between fatigue and Th2-type T cell responses is not clear, though our cumulative findings do suggest that a functional immune response is required.

For women receiving the MBSR(BC) program in the current study, the significant relationships between immunological bio-markers and symptom improvement were exclusively positive. This positive relationship means that the larger the baseline immune biomarker value, the greater the improvement in symptoms. In other words, it appears that symptom improvement was associated with increased immune activity at baseline.

Despite having received RT or RT and CT, all of the women showed lymphocyte subset values within the normal ranges upon entering the study, and analysis of the data by cancer treatment type showed no significant differences. Partially contributing to this normalcy was the fact that two thirds of the participants had completed treatment 1–3 months before the start of the study. A number of reports have examined immune markers as a function of immunosuppressive anticancer therapy (Mills et al., 2008). Most of these studies examined blood cytokine levels as biomarkers, and the constant theme was that cytokines increased during or shortly after treatment (Mills et al., 2008) but returned to pretreatment levels 1–3 months after treatment (Bower et al., 2009; Sepah & Bower, 2009; Wang et al., 2012).

The results of this study should be interpreted with caution due to the small sample size. The parent study included larger numbers, but we had previously shown that time since treatment enhanced the effect of MBSR(BC) on immune biomarker levels (Lengacher, Kip, et al., 2012). Therefore, we limited participation in the present study to patients who were recently off-treatment, greatly curtailing the sample. The large number of significant relationships between the biomarkers and symptom clusters, however, should provide sufficient evidence to move from this exploratory developmental approach to a larger validation approach (Simon, 2008).

If the predictive value of these biomarkers is validated, the cost-effectiveness of MBSR(BC) may be amplified as it could be targeted toward patients with high levels of these biomarkers, particularly within the first few months of recovery. Because MBSR has been successful across many patient populations (Bohlmeijer et al., 2010; Grossman et al., 2004; Hofmann et al., 2010; Ledesma & Kumano, 2009), future research should test the generalizability of these specific biomarker results.

Acknowledgments

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the National Institutes of Health, National Cancer Institute, grant number R21CA109168.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

  1. Abud-Mendoza C, Cuevas-Orta E, Santillan-Guerrero EN, Martinez-Martinez MU, Hernandez-Castro B, Estrada-Capetillo L, Baranda L. Decreased blood levels of B lymphocytes and NK cells in patients with systemic lupus erythematosus (SLE) infected with papillomavirus (HPV) Archives of Dermatological Research. 2012;305:117–123. doi: 10.1007/s00403-012-1258-9. [DOI] [PubMed] [Google Scholar]
  2. Andersen BL, Farrar WB, Golden-Kreutz D, Emery CF, Glaser R, Crespin T, Carson WE. Distress reduction from a psychological intervention contributes to improved health for cancer patients. Brain, Behavior and Immunity. 2007;21:953–961. doi: 10.1016/j.bbi.2007.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Andreyev HJ, Davidson SE, Gillespie C, Allum WH, Swarbrick E. Practice guidance on the management of acute and chronic gastrointestinal problems arising as a result of treatment for cancer. Gut. 2012;61:179–192. doi: 10.1136/gutjnl-2011-300563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bohlmeijer E, Prenger R, Taal E, Cuijpers P. The effects of mindfulness-based stress reduction therapy on mental health of adults with a chronic medical disease: A meta-analysis. Journal of Psychosomatic Research. 2010;68:539–544. doi: 10.1016/j.jpsychores.2009.10.005. [DOI] [PubMed] [Google Scholar]
  5. Bower JE, Ganz PA, Tao ML, Hu W, Belin TR, Sepah S, Aziz N. Inflammatory biomarkers and fatigue during radiation therapy for breast and prostate cancer. Clinical Cancer Research. 2009;15:5534–5540. doi: 10.1158/1078-0432.CCR-08-2584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Carlson LE, Speca M, Faris P, Patel KD. One year pre-post intervention follow-up of psychological, immune, endocrine and blood pressure outcomes of mindfulness-based stress reduction (MBSR) in breast and prostate cancer outpatients. Brain, Behavior, and Immunity. 2007;21:1038–1049. doi: 10.1016/j.bbi.2007.04.002. [DOI] [PubMed] [Google Scholar]
  7. Chen ML, Lin CC. Cancer symptom clusters: A validation study. Journal of Pain Symptom Management. 2007;34:590–599. doi: 10.1016/j.jpainsymman.2007.01.008. [DOI] [PubMed] [Google Scholar]
  8. Cleeland CS, Mendoza TR, Wang XS, Chou C, Harle MT, Morrissey M, Engstrom MC. Assessing symptom distress in cancer patients: The M.D. Anderson Symptom Inventory. Cancer. 2000;89:1634–1646. doi: 10.1002/1097-0142(20001001)89:7<1634::aid-cncr29>3.0.co;2-v. [DOI] [PubMed] [Google Scholar]
  9. Collado-Hidalgo A, Bower JE, Ganz PA, Cole SW, Irwin MR. Inflammatory biomarkers for persistent fatigue in breast cancer survivors. Clinical Cancer Research. 2006;12:2759–2766. doi: 10.1158/1078-0432.CCR-05-2398. [DOI] [PubMed] [Google Scholar]
  10. Cseh A, Farkas KM, Derzbach L, Muller K, Vasarhelyi B, Szalay B, Farkas V. Lymphocyte subsets in pediatric migraine. Neurological Science. 2012;34:1151–1155. doi: 10.1007/s10072-012-1218-3. [DOI] [PubMed] [Google Scholar]
  11. Dodd MJ, Cho MH, Cooper BA, Miaskowski C. The effect of symptom clusters on functional status and quality of life in women with breast cancer. European Journal of Oncology Nursing. 2010;14:101–110. doi: 10.1016/j.ejon.2009.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gardner RV. Long term hematopoietic damage after chemotherapy and cytokine. Frontiers in Bioscience. 1999;4:e47–e57. doi: 10.2741/A479. [DOI] [PubMed] [Google Scholar]
  13. Graham JE, Christian LM, Kiecolt-Glaser JK. Stress, age, and immune function: Toward a lifespan approach. Journal of Behavioral Medicine. 2006;29:389–400. doi: 10.1007/s10865-006-9057-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Grossman P, Niemann L, Schmidt S, Walach H. Mindfulness-based stress reduction and health benefits. A meta-analysis. Journal of Psychosomatic Research. 2004;57:35–43. doi: 10.1016/S0022-3999(03)00573-7. [DOI] [PubMed] [Google Scholar]
  15. Hofmann SG, Sawyer AT, Witt AA, Oh D. The effect of mindfulness-based therapy on anxiety and depression: A meta-analytic review. Journal of Consulting and Clinical Psychology. 2010;78:169–183. doi: 10.1037/a0018555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hosmer DW, Lemeshow S. Applied logistic regression. Hoboken, NJ: John Wiley; 2000. [Google Scholar]
  17. Ivanova MO, Ionova TI, Kalyadina SA, Uspenskaya OS, Kishtovich AV, Guo H, Wang XS. Cancer-related symptom assessment in Russia: Validation and utility of the Russian M. D. Anderson Symptom Inventory. Journal of Pain and Symptom Management. 2005;30:443–453. doi: 10.1016/j.jpainsymman.2005.04.015. [DOI] [PubMed] [Google Scholar]
  18. Kabat-Zinn J, Lipworth L, Burney R. The clinical use of mindfulness meditation for the self-regulation of chronic pain. Journal of Behavioural Medicine. 1985;8:163–190. doi: 10.1007/BF00845519. [DOI] [PubMed] [Google Scholar]
  19. Kang DH, Weaver MT, Park NJ, Smith B, McArdle T, Carpenter J. Significant impairment in immune recovery after cancer treatment. Nursing Research. 2009;58:105–114. doi: 10.1097/NNR.0b013e31818fcecd. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Koukourakis GV, Zabatis H, Zacharias GA, Koukourakis MJ. Post-surgical irradiation causes cellular immune suppression in patients with breast cancer. European Journal of Cancer Care. 2009;18:306–312. doi: 10.1111/j.1365-2354.2008.00995.x. [DOI] [PubMed] [Google Scholar]
  21. Kwon YC, Yun YH, Lee KH, Son KY, Park SM, Chang YJ, Cleeland CS. Symptoms in the lives of terminal cancer patients: Which is the most important? Oncology. 2006;71:69–76. doi: 10.1159/000100450. [DOI] [PubMed] [Google Scholar]
  22. Ledesma D, Kumano H. Mindfulness-based stress reduction and cancer: A meta-analysis. Psycho-Oncology. 2009;18:571–579. doi: 10.1002/pon.1400. [DOI] [PubMed] [Google Scholar]
  23. Lengacher CA, Kip KE, Barta M, Post-White J, Jacobsen PB, Groer M, Shelton MM. A pilot study evaluating the effect of mindfulness-based stress reduction on psychological status, physical status, salivary cortisol, and interleukin-6 among advanced-stage cancer patients and their caregivers. Journal of Holistic Nursing. 2012;30:170–185. doi: 10.1177/0898010111435949. [DOI] [PubMed] [Google Scholar]
  24. Lengacher CA, Kip KE, Post-White J, Fitzgerald S, Newton C, Barta M, Klein TW. Lymphocyte recovery after breast cancer treatment and mindfulness-based stress reduction (MBSR) therapy. Biological Research for Nursing. 2013;15:37–47. doi: 10.1177/1099800411419245. [DOI] [PubMed] [Google Scholar]
  25. Lengacher CA, Reich RR, Post-White J, Moscoso M, Shelton MM, Barta M, Budhrani P. Mindfulness based stress reduction in post-treatment breast cancer patients: An examination of symptoms and symptom clusters. Journal of Behavioural Medicine. 2012;35:86–94. doi: 10.1007/s10865-011-9346-4. [DOI] [PubMed] [Google Scholar]
  26. Mills PJ, Ancoli-Israel S, Parker B, Natarajan L, Hong S, Jain S, von Känel R. Predictors of inflammation in response to anthracycline-based chemotherapy for breast cancer. Brain, Behavior, and Immunity. 2008;22:98–104. doi: 10.1016/j.bbi.2007.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Mozaffari F, Lindemalm C, Choudhury A, Granstam-Bjorneklett H, Helander I, Lekander M, Mellstedt H. NK-cell and T-cell functions in patients with breast cancer: Effects of surgery and adjuvant chemo and radiotherapy. British Journal of Cancer. 2007;97:105–111. doi: 10.1038/sj.bjc.6603840. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Mozaffari F, Lindemalm C, Choudhury A, Granstam-Bjorneklett H, Lekander M, Nilsson B, Mellstedt H. Systemic immune effects of adjuvant chemotherapy with 5-fluorouracil, epirubicin and cyclophosphamide and/or radiotherapy in breast cancer: A longitudinal study. Cancer Immunology, Immunotherapy. 2009;58:111–120. doi: 10.1007/s00262-008-0530-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Nascimbeni M, Pol S, Saunier B. Distinct CD4+ CD8+ double-positive T cells in the blood and liver of patients during chronic hepatitis B and C. PLoS One. 2011;6:e20145. doi: 10.1371/journal.pone.0020145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Nascimbeni M, Shin EC, Chiriboga L, Kleiner DE, Rehermann B. Peripheral CD4(+)CD8(+) T cells are differentiated effector memory cells with antiviral functions. Blood. 2004;104:478–486. doi: 10.1182/blood-2003-12-4395. [DOI] [PubMed] [Google Scholar]
  31. Okuyama T, Wang XS, Akechi T, Mendoza TR, Hosaka T, Cleeland CS, Uchitomi Y. Japanese version of the MD Anderson Symptom Inventory: A validation study. Journal of Pain and Symptom Management. 2003;26:1093–1104. doi: 10.1016/j.jpainsymman.2003.05.003. [DOI] [PubMed] [Google Scholar]
  32. Parel Y, Chizzolini C. CD4+ CD8+ double positive (DP) T cells in health and disease. Autoimmunity Reviews. 2004;3:215–220. doi: 10.1016/j.autrev.2003.09.001. [DOI] [PubMed] [Google Scholar]
  33. Park J, Han K. Single-color multitarget flow cytometry using monoclonal antibodies labeled with different intensities of the same fluorochrome. Annals of Laboratory Medicine. 2012;32:171–176. doi: 10.3343/alm.2012.32.3.171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Project MATCH Research Group. Matching alcoholism treatments to client heterogeneity: Treatment main effects and matching effects on drinking during treatment. Journal of Studies on Alcohol. 1998;5:631–639. doi: 10.15288/jsa.1998.59.631. [DOI] [PubMed] [Google Scholar]
  35. Schwarz MJ, Chiang S, Muller N, Ackenheil M. T-helper-1 and T-helper-2 responses in psychiatric disorders. Brain, Behavior, and Immunity. 2001;15:340–370. doi: 10.1006/brbi.2001.0647. [DOI] [PubMed] [Google Scholar]
  36. Sepah SC, Bower JE. Positive affect and inflammation during radiation treatment for breast and prostate cancer. Brain, Behavior, and Immunity. 2009;23:1068–1072. doi: 10.1016/j.bbi.2009.06.149. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sephton SE, Dhabhar FS, Keuroghlian AS, Giese-Davis J, McEwen BS, Ionan AC, Spiegel D. Depression, cortisol, and suppressed cell-mediated immunity in metastatic breast cancer. Brain, Behavior, and Immunity. 2009;23:1148–1155. doi: 10.1016/j.bbi.2009.07.007. [DOI] [PubMed] [Google Scholar]
  38. Simon R. Development and validation of biomarker classifiers for treatment selection. Journal of Statistical Planning and Inference. 2008;138:308–320. doi: 10.1016/j.jspi.2007.06.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Solomayer EF, Feuerer M, Bai L, Umansky V, Beckhove P, Meyberg GC, Diel IJ. Influence of adjuvant hormone therapy and chemotherapy on the immune system analysed in the bone marrow of patients with breast cancer. Clinical Cancer Research. 2003;9:174–180. [PubMed] [Google Scholar]
  40. Standish LJ, Torkelson C, Hamill FA, Yim D, Hill-Force A, Fitzpatrick A, Martzen MR. Immune defects in breast cancer patients after radiotherapy. Journal of the Society for Integrative Oncology. 2008;6:110–121. [PMC free article] [PubMed] [Google Scholar]
  41. Toldi G, Molvarec A, Stenczer B, Müller V, Eszes N, Bohács A, Tamási L. Peripheral T(h)1/T(h)2/T(h)17/regulatory T-cell balance in asthmatic pregnancy. International Immunology. 2011;23:669–677. doi: 10.1093/intimm/dxr074. [DOI] [PubMed] [Google Scholar]
  42. Torres-Harding S, Sorenson M, Jason LA, Maher K, Fletcher MA. Evidence for T-helper 2 shift and association with illness parameters in chronic fatigue syndrome (CFS) Bulletin of the International Association for CFS/ME. 2008;16:19–33. [PMC free article] [PubMed] [Google Scholar]
  43. Tseng TH, Cleeland CS, Wang XS, Lin CC. Assessing cancer symptoms in adolescents with cancer using the Taiwanese version of the M. D. Anderson Symptom Inventory. Cancer Nursing. 2008;31:E9–E16. doi: 10.1097/01.NCC.0000305728.50098.51. [DOI] [PubMed] [Google Scholar]
  44. van der Most RG, Currie A, Robinson BW, Lake RA. Cranking the immunologic engine with chemotherapy: Using context to drive tumor antigen cross-presentation towards useful antitumor immunity. Cancer Research. 2006;66:601–604. doi: 10.1158/0008-5472.CAN-05-2967. [DOI] [PubMed] [Google Scholar]
  45. Von Ah D, Habermann B, Carpenter JS, Schneider BL. Impact of perceived cognitive impairment in breast cancer survivors. European Journal of Oncology Nursing. 2013;17:236–241. doi: 10.1016/j.ejon.2012.06.002. [DOI] [PubMed] [Google Scholar]
  46. Wang XS, Laudico AV, Guo H, Mendoza TR, Matsuda ML, Yosuico VD, Fragante EP, Cleeland CS. Filipino version of the M. D. Anderson Symptom Inventory: Validation and multisymptom measurement in cancer patients. Journal of Pain and Symptom Management. 2006;31:542–552. doi: 10.1016/j.jpainsymman.2005.11.011. [DOI] [PubMed] [Google Scholar]
  47. Wang XS, Wang Y, Guo H, Mendoza TR, Hao XS, Cleeland CS. Chinese version of the M. D. Anderson Symptom Inventory: Validation and application of symptom measurement in cancer patients. Cancer. 2004;101:1890–1901. doi: 10.1002/cncr.20448. [DOI] [PubMed] [Google Scholar]
  48. Wang XS, Williams LA, Krishnan S, Liao Z, Liu P, Mao L, Cleeland CS. Serum sTNF-R1, IL-6, and the development of fatigue in patients with gastrointestinal cancer undergoing chemoradiation therapy. Brain, Behavior, and Immunity. 2012;26:699–705. doi: 10.1016/j.bbi.2011.12.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Weiss L, Roux A, Garcia S, Demouchy C, Haeffner-Cavaillon N, Kazatchkine MD, Gougeon ML. Persistent expansion, in a human immunodeficiency virus-infected person, of V beta-restricted CD4+CD8+ T lymphocytes that express cytotoxicity-associated molecules and are committed to produce interferon-gamma and tumor necrosis factor-alpha. Journal of Infectious Disease. 1998;178:1158–1162. doi: 10.1086/515674. [DOI] [PubMed] [Google Scholar]
  50. Wistuba II, Gelovani JG, Jacoby JJ, Davis SE, Herbst RS. Methodological and practical challenges for personalized cancer therapies. Nature Reviews Clinical Oncology. 2011;8:135–141. doi: 10.1038/nrclinonc.2011.2. [DOI] [PubMed] [Google Scholar]
  51. Yamazaki H, Yoshioka Y, Inoue T, Tanaka E, Nishikubo M, Sato T, Inoue T. Changes in natural killer cell activity by external radiotherapy and/or brachytherapy. Oncology Reports. 2002;9:359–363. [PubMed] [Google Scholar]

RESOURCES