Investigating Associations Between Psychological Distress and Changes in Peripheral Blood Monocytes in Patients With CLL/SLL Managed With Active Surveillance Authors: Carolyn Y. Fang, Alexander W. MacFarlane, Henry C. H. Fung, Nadia Khan, Stefan K. Barta, Jill S. Hasler, Hatcher J. Ballard, Daniel J. Landsburg, Sunita D. Nasta, Stephen J. Schuster, Rashmi Khanal, Richard I. Fisher, Adam D. Cohen, Jakub Svoboda, Kerry S. Campbell Published: 2026 ABSTRACT ABSTRACTAsymptomatic patients with a new diagnosis of chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) are commonly managed by active surveillance as per International Workshop on Chronic Lymphocytic Leukemia (iwCLL) criteria; however, many patients report elevated distress during surveillance. Psychological distress may have biologic implications given that natural killer [NK] cells, T cells, and monocytes are not only sensitive to distress, but also critical in controlling disease progression. Therefore, the objective of this study was to explore associations of distress with immune biomarkers in patients with CLL/SLL being managed by active surveillance. We enrolled 76 participants and assessed perceived stress, anxiety, and depressive symptoms at study entry and every 6 months for up to 2 years during active surveillance. Blood samples were collected at the same time points to assess biomarkers on T cell, NK cell, and monocyte subsets. In multivariable regression analyses, higher levels of perceived stress and anxiety were significantly associated with lower percentages of intermediate monocytes (IM); anxiety was also significantly associated with higher percentages of classical monocytes (CM). In longitudinal analyses, higher stress and anxiety were significantly associated with higher percentages of CM at 12 months. This is one of the first studies to observe that psychological distress is associated with differential patterns of monocyte distribution in patients with CLL/SLL managed with active surveillance. Introduction In 2025, an estimated 80,350 cases of non‐Hodgkin lymphoma (NHL) were diagnosed in the United States, with approximately 19,390 deaths attributed to this disease [1]. NHLs are a heterogeneous group of lymphoid malignancies that can be clinically grouped into slow‐growing, indolent lymphomas (including chronic lymphocytic leukemia/small lymphocytic lymphoma [CLL/SLL]) or aggressive forms (such as diffuse large B‐cell lymphoma) [2]. Approximately 35% of patients with NHL will present with indolent disease, which is often asymptomatic at diagnosis and can progress very slowly, but is considered incurable [3]. A number of studies have established that initiating therapy for patients with asymptomatic CLL/SLL does not confer a survival benefit [4, 5, 6, 7, 8]. As a result, asymptomatic patients are commonly managed by active surveillance as per the guidelines established by the International Workshop on Chronic Lymphocytic Leukemia (iwCLL) [9, 10]. This approach involves monitoring patients at regular intervals with blood tests, physical examinations, and imaging. The benefits of active surveillance include reduced healthcare costs and avoidance of treatment‐related side effects; however, the psychological consequences of surveillance and repeated monitoring over time (e.g., high anxiety, distress) should not be underestimated. Patients often report elevated levels of anxiety during active surveillance [11, 12] and sometimes refer to this phase as “watch and worry.” [13, 14] Among 32 patients with follicular lymphoma, more than 25% reported clinically elevated levels of psychological distress, and 60% reported moderate to severe anxiety [15]. In a sample of 51 patients, greater than 50% experienced borderline or excessive levels of anxiety, depressive symptoms, or both [12]. Patients with CLL/SLL have reported being fearful about the course of their illness and what the future might bring [16, 17]. The psychological consequences of active surveillance may have biologic correlates. Prior studies provide evidence that psychological distress is associated with immunologic changes in populations with cancer [18, 19]. For example, among patients with ovarian cancer, higher levels of distress were associated with lower natural killer cell cytotoxicity (NKCC) [20]. Similarly, in a cross‐sectional study of patients with relapsed or refractory CLL/SLL, self‐reported stress was correlated with greater levels of inflammatory cytokines (TNF‐α, IL‐6) and chemokine markers that are generally associated with CLL/SLL disease progression [21]. These findings are of interest in light of studies suggesting that immune factors play a role in cancer surveillance and anti‐tumor responses [22, 23, 24]. Immune deficiencies have been identified as one of the strongest known risk factors for NHL [25], with some studies focusing on the role of NK cells (CD56+CD3−) in controlling NHL [26]. Numerous studies in both mice and humans indicate that B‐cell lymphomas, including CLL/SLL, can be controlled by NK cells [27, 28, 29, 30, 31, 32, 33, 34, 35]. Of note, we previously showed that NK cells become dysfunctional in CLL/SLL patients, which was associated with progressive loss of viability and frequency of the mature NK cell pool in peripheral blood, presumably due to activation‐induced cell death as they attack the tumor cells [29]. Studies have also demonstrated that mice lacking perforin, which is produced mainly by NK cells and CD8+ T cells, spontaneously develop B‐cell lymphomas [26, 36]. Other immune cells in the blood, such as monocytes, can play important roles in controlling NHL. Monocytes are circulating myeloid cells that are attracted into tissues at sites of inflammation. Upon tissue entry, monocytes can differentiate into dendritic cells (DCs) or macrophages that mediate antigen presentation to T cells and production of pro‐inflammatory cytokines [37]. Upon entry into the tumor microenvironment (TME), however, monocytes differentiate into tumor‐associated macrophages (TAMs) and myeloid‐derived suppressor cells (MDSCs) that can support the survival and growth of tumors and induce immunosuppression [38]. CLL promotes the development of a macrophage population, which has been called nurse‐like cells, that associate with the leukemic cells and support their growth and survival through the production of a variety of factors [39]. Together, these studies suggest that monocytes, NK cells, and T cell subsets are relevant in disease processes. Thus, the goal of this prospective study was to explore whether psychological distress (as characterized by anxiety, depressive symptoms, and perceived stress) is associated with immune subsets over time in adult patients with asymptomatic and untreated indolent CLL/SLL who were being managed with active surveillance. To address this goal, we assessed psychological distress in a sample of patients with indolent CLL/SLL and correlated these measurements with phenotypes of lymphocytes and monocytes in peripheral blood at regular intervals over time. Methods Participants and Procedures Participants were enrolled between April 2016 and February 2020. Participants were untreated adult patients (age ≥ 21 years) diagnosed with CLL/SLL using standard clinical criteria based on the 2008 International Workshop on Chronic Lymphocytic Leukemia (iwCLL) for patients recruited before 2018 and the 2018 iwCLL for patients recruited from 2018 to 2020 [9, 10]. Patients were eligible if they had asymptomatic disease, were untreated, and had low tumor burden (i.e., no bulky disease). These patients were deemed to be appropriate candidates for active surveillance consistent with established guidelines for the clinical management of such patients [9, 40]. Participants were excluded if they: (1) had a history of previous cancer except basal cell carcinoma; (2) self‐reported a known active HIV, hepatitis B, or hepatitis C infection or autoimmune disorder; (3) were currently pregnant; or (4) could not provide informed consent. The study was approved by the Institutional Review Boards at Fox Chase Cancer Center (IRB# 11–886) on September 28, 2015, and subsequently at the University of Pennsylvania (UPCC# 01416) in accordance with federal regulations and consistent with the principles outlined in the Declaration of Helsinki. Informed consent was obtained from each participant prior to study participation. Participants for this analysis were drawn from a larger study of patients diagnosed with indolent B‐cell‐derived neoplasms (including follicular lymphoma, CLL/SLL, marginal zone lymphoma, smoldering myeloma, and B‐cell lymphomas) managed with active surveillance. A total of 218 patients were enrolled in the study, of which 105 were patients with untreated asymptomatic CLL/SLL (see Figure S1). Following informed consent, participants completed the study survey; we were able to obtain viable blood samples from 76 patients at baseline (study entry). Subsequently, participants were re‐contacted at six‐month intervals for follow‐up assessments during which they completed a brief survey and provided a blood sample. Participants completed up to a maximum of five study visits (i.e., study entry, 6‐month, 12‐month, 18‐month, and 24‐month follow‐up) unless they were withdrawn early from the study for disease progression or other exclusion criterion (e.g., subsequently developed another cancer) or no longer wished to participate. Study Measures Demographic and Clinical Characteristics Demographic data, including participant age, sex, race and ethnicity, education level, and marital status, were obtained via self‐report at study entry. Participants were categorized as engaging in current smoking behavior if they reported smoking ‘every day’ or ‘some days’ (vs. no smoking) and were classified as having heavy alcohol consumption if they reported drinking five or more drinks (for males) or four or more drinks (for females) on the same occasion in the past 30 days as defined by the US Centers for Disease Control and Prevention [41]. We obtained absolute lymphocyte count (ALC) at study entry and date of diagnosis via medical chart review conducted by study staff. Perceived Stress Perceived stress was measured using the 10‐item Perceived Stress Scale (PSS) [42, 43], which captures the degree to which an individual perceives events in their life to be stressful or difficult to manage. Response options range from ‘0 = Never’ to ‘4 = Very Often.’ Items were summed to create a total score (possible range = 0 to 40), whereby higher scores reflect greater perceived stress. The PSS demonstrated high reliability in the present study (Cronbach's α = 0.87). Anxiety and Depressive Symptoms The 14‐item Hospital Anxiety and Depression Scale (HADS) was used to measure anxiety and depressive symptoms [44]. Each item is scored on a four‐point scale ranging from 0 (not present) to 3 (considerable). Seven of the 14 items are summed to compute an anxiety score, and 7 items are summed to create a score of depressive symptoms, with each score ranging from zero to 21. Higher scores represent higher anxiety/depressive symptoms; a cut‐off score of 8 or greater has been shown to reliably identify clinically elevated levels of anxiety or depression [45]. This well‐validated scale has been widely used in studies of cancer patients [45] including patients with indolent NHL [46]. Cronbach's α was high in the present sample (α = 0.84). Immunologic Markers Peripheral blood (20 mL) was collected in heparin Vacutainer tubes (BD) and transported at room temperature to the laboratory for same‐day processing and analysis, as previously described [29]. Peripheral blood mononuclear cells (PBMC) were isolated by centrifugation on Lymphoprep (Axis‐shield POC AS, Oslo, Norway), and immune cell status was freshly assessed by staining aliquots of 1 million PBMCs on ice for 20 min with a panel of commercially obtained fluorophore‐conjugated monoclonal antibodies (Table S1). Samples were washed twice with ice‐cold buffer and 100 ng/mL propidium iodide in the second rinse, and cells were analyzed on a BD FACSAria II flow cytometer, which contained four diode lasers with excitation wavelengths at 633 nm, 488 nm, 405 nm, and 355 nm. Between 100,000 and 400,000 events were acquired from each sample at 500–2,500 events/s using 70 psi pressure. The cytometers were housed in the CLIA (The Clinical Laboratory Improvement Amendments of 1988)‐certified and CAP (College of American Pathologists)‐certified FCCC Clinical Pathology Flow Cytometry Laboratory, where they were calibrated daily with fluorescent BD CompBeads for uniform clinical diagnostic use. Compensation and PMT voltage settings were consistent for the acquisition of all samples, as optimized at the beginning of the project based upon analysis of unstained, single‐stained, and multi‐stained PBMC samples. Flow cytometry‐based immune phenotyping analysis of fresh peripheral blood lymphocyte and monocyte subpopulations was performed by manual gating using FlowJo software (v10.7.1, BD) as shown in Figure S2 and Figure S3 to assess 27 distinct immune parameters listed in Table S2. Expression levels of individual biomarkers were quantified as geometric mean fluorescence intensity (GMFI). Monocyte subsets were defined based upon expression of the surface markers CD14 and CD16 as classical (CM; CD14++CD16−), intermediate (IM; CD14++CD16+), and non‐classical (NCM; CD14+CD16+) monocytes [47]. The flow cytometry panel also assessed expression of activating receptors, including NKp30, DNAM‐1, and NKp80, which can stimulate responses by NK and T cells toward tumor cells [48, 49]. Finally, we assessed the inhibitory checkpoint receptor, PD‐1, and its ligand, PD‐L1. PD‐1 is upregulated on activated T cells, and PD‐L1 is normally upregulated by interferon on antigen‐presenting cells, such as monocytes. However, PD‐L1 can also be expressed on many tumor cells, which prevents attack by PD‐1‐expressing T cells and can contribute to rendering the T cells exhausted [50]. Expression of PD‐L1 has also been reported on T and NK cells by several groups, including upregulation on activated NK cells [51, 52, 53]. Analytic Plan Preliminary analyses were performed to characterize the sample in terms of demographic and psychosocial variables. Normality of the variables was assessed through Shapiro–Wilk normality tests and visual inspection of histogram plots, and model diagnostics were assessed through inspection of residual plots and normal q‐q plots. Spearman's correlation and T‐tests/ANOVA were used to examine whether demographic or clinical characteristics were associated with psychosocial measures at baseline. All p‐values were two‐sided. To refine and narrow down the number of immune markers to be included in the analysis, the markers were first categorized as NK cell, T cell, and monocyte markers based on cell type. Next, relationships between immune markers were assessed using Spearman's correlation. Hierarchical agglomerative clustering with complete linkage was used to identify groups of highly correlated immune markers; a single marker from each cluster (defined by Spearman's correlation > 0.8) was then selected for inclusion in the subsequent analyses. Multivariable linear regression models were used to assess the association between the selected immune markers and psychosocial measures. Analyses controlled for covariates including age, time since diagnosis, sex, and absolute lymphocyte count (ALC). Multivariable regression analyses were performed separately for each marker, and model diagnostics were used to assess fit. Where lack‐of‐fit was found, in sensitivity analyses, we confirmed findings using appropriate transformations (e.g., log, logit). Statistical significance was defined as two‐sided p‐values < 0.05. Separately for each outcome, Benjamini‐Hochberg (BH) corrections for multiple testing were performed in secondary analyses (27 total markers per psychosocial outcome). Finally, longitudinal analyses were conducted to examine overall trends in immune markers and psychosocial measures over time using polynomial (lowess) splines. N = 62 patients had multiple samples over time and were included in the longitudinal models. We assessed associations between immune markers and psychosocial measures over time using linear mixed effects models with random intercepts for patients and fixed effects of age, sex, time from diagnosis to baseline, ALC, and study time treated as a discrete variable. Models also included an interaction between psychosocial measures and time point to assess whether the relationship between psychosocial measures and immune markers differed over time. Separate models were fit for each psychosocial and immune measure pair. Estimated marginal trends were used to evaluate the relationship between each psychosocial and immune marker pair at each time point. Models were assessed for fit, and as above, transformations were used in sensitivity analyses where warranted. The linear mixed effects models assumed a missing at random (MAR) data mechanism. A sensitivity analysis including only the patients who remained on active surveillance for the duration of the study period was also performed. Statistical significance was defined as a two‐sided p‐value < 0.05. All statistical analyses were performed using R v4.5.1. Results Participant Characteristics The study sample included 76 patients with CLL/SLL with corresponding blood samples. Overall, participants were predominantly male (60.5%) with a mean age of 63.7 years (SD = 9.1 years) (Table 1). Approximately 5% self‐reported their race as Black or African American, and 4% were of Hispanic ethnicity. At enrollment, the median time since diagnosis was 10.6 months. At study baseline, mean levels of anxiety and depressive symptoms were within the normal range (scores of 0 to 7), although 25% of the participants reported elevated anxiety (scores of 8 or greater). TABLE 1Characteristics of the study Sample (n = 76).VariableMean (SD) or n (%)Age in years (SD)63.7 (9.1)SexMale46 (60.5%)Female30 (39.5%)RaceWhite69 (90.8%)Black or African American4 (5.3%)More than one race3 (3.9%)Hispanic ethnicity3 (3.9%)Marital status* Married/living with partner57 (75.0%)Divorced/widowed/single15 (19.7%)Education* High school or less17 (22.4%)Some college/vocational school15 (19.7%)College degree19 (25.0%)Postgraduate degree19 (25.0%)Median Time since diagnosis (months)10.6ECOG Performance Status0–159 (77.6%)NA/unknown17 (22.4%)Current smoking (yes)6 (7.9%)Heavy alcohol consumption6 (7.9%)Median ALC (cells/μL)1,032Median WBC (cells/μL)11,100Median hemoglobin (g/dL)14Rai StageStage 0–II46 (60.5%)Stage III–IV9 (11.8%)NA/unknown21 (27.6%)Genetic abnormalitiesFISH negative18 (23.7%)Del (13q)29 (38.2%)Del (11q)4 (5.3%)Del (17p)1 (1.3%)Trisomy 128 (10.5%)Anxiety5.3 (3.6)% with elevated scores (scores ≥ 8)25.0%Depressive symptoms2.9 (2.3)% with elevated scores (scores ≥ 8)2.6%*Due to missing data, some percent values do not add up to 100%. Characteristics of the study Sample (n = 76). Due to missing data, some percent values do not add up to 100%. We examined if demographic factors were associated with anxiety, depressive symptoms, or perceived stress using Spearman's correlation for continuous variables (age, time since diagnosis) and T‐tests or ANOVA for categorical variables (sex, marital status, education level). None of the demographic factors were significantly associated with anxiety, depressive symptoms, or perceived stress. Cross‐Sectional Associations Between Psychological Factors and Immune Markers at Study Entry Perceived Stress Multivariable regression analyses revealed that higher levels of perceived stress were associated with lower percentages of IM (β = −0.79, p < 0.001; see Table 2 and Figure 1). Perceived stress was also associated with higher percentages of CM (β = 0.82, p = 0.02) and lower expression of both the activating receptor NKp80 on CD8+ T cells (β = −8.04, p = 0.02) and DNAM‐1 on CD4+ T cells (β = −33.41, p = 0.02), but these associations were attenuated after correction for multiple testing. No associations were observed between perceived stress and expression of checkpoint receptors on NK and T cells. TABLE 2Baseline associations Between Perceived Stress, Anxiety, and Depressive Symptoms with Immune Parameters in Patients with CLL/SLL.Immune parameterPerceived stressAnxietyDepressive symptomsβ p β p β p Monocyte subsets % NCM0.050.87−0.880.15−0.310.72% IM−0.79< 0.001* −1.36< 0.01* −0.810.22% CM0.820.022.36< 0.001* 1.330.17 Innate Immune Receptors NKp30 expression on CD56dim NK cells−3.890.23−9.570.12−20.440.02NKp30 expression on CD4+ T cells−2.720.06−3.620.19−9.820.01NKp30 expression on CD8+ T cells−2.470.11−3.110.28−10.040.01NKp80 expression on CD56dim NK cells−5.400.56−16.040.35−37.920.11NKp80 expression on CD4+ T cells−2.940.13−4.960.18−1.890.72NKp80 expression on CD8+ T cells−8.040.02−14.370.03−5.560.56DNAM‐1 expression on CD56dim NK cells50.650.2357.570.4775.890.50DNAM‐1 expression on CD4+ T cells−33.410.02−40.430.16−73.440.07DNAM‐1 expression on CD8+ T cells−21.110.22−7.110.83−38.310.40 Checkpoint Receptors PD‐L1 expression on CD56dim NK cells−2.320.27−1.480.71−13.970.01PD‐L1 expression on CD4+ T cells−3.900.11−3.140.50−16.120.01PD‐L1 expression on CD8+ T cells−3.730.16−2.160.67−17.320.01 Note: The models are adjusted for patient age, time since diagnosis, sex, and ALC.Abbreviations: ALC, absolute lymphocyte count; CM, classical monocytes, IM, intermediate monocytes; NCM, non‐classical monocytes.*Associations remained significant after BH correction for multiple testing. Baseline associations Between Perceived Stress, Anxiety, and Depressive Symptoms with Immune Parameters in Patients with CLL/SLL. Note: The models are adjusted for patient age, time since diagnosis, sex, and ALC. Abbreviations: ALC, absolute lymphocyte count; CM, classical monocytes, IM, intermediate monocytes; NCM, non‐classical monocytes. Associations remained significant after BH correction for multiple testing. Associations of perceived stress with percentages of total macrophages that are either classical (CM) (p = 0.02) or intermediate (IM) (p < 0.001*) subtype in individual patients. Multivariable linear regression analysis was used to determine indicated p‐values, and p‐values with an asterisk passed Benjamini‐Hochberg (BH) corrections for multiple testing. Lines represent the least squares fit to the data. Anxiety Similar to the patterns observed with perceived stress, higher levels of anxiety were associated with lower percentages of IM (β = −1.36, p < 0.01), but higher percentages of CM (β = 2.36, p < 0.001; Table 2 and Figure 2). Higher anxiety was also associated with lower expression of the activating receptor NKp80 on CD8+ T cells (β = −14.37, p = 0.03), but this association was no longer statistically significant after correction for multiple testing. No associations emerged between anxiety and expression of checkpoint receptors on NK and T cells. Associations of anxiety level with percentages of total macrophages that are either CM (p < 0.001*) or IM (p < 0.01*) subtype in individual patients. Statistical analyses are described in the legend of Figure 1, and p‐values with an asterisk passed Benjamini‐Hochberg (BH) corrections for multiple testing. Depressive Symptoms Depressive symptoms were not significantly associated with any monocyte subsets. Higher levels of depressive symptoms showed preliminary associations with lower expression of NKp30 on CD56dim NK cells (the predominant, mature, and cytolytic subset; β = −20.44, p = 0.02) and CD4+ (β = −9.82, p = 0.01) and CD8+ T cells (β = −10.04, p = 0.01; see Table 2) and were negatively associated with expression of PD‐L1 on these NK and T cell populations (p‐values < 0.05). However, these associations did not reach statistical significance after correction for multiple testing. In addition, we ran the models including two key prognostic variables, Rai stage and presence of del13q. Results were similar between the two sets of models, and the findings did not change substantially (results not shown), but there was a further reduction in sample size. Longitudinal Associations The median length of follow‐up was 22.3 months. Levels of stress, anxiety, and depressive symptoms remained relatively stable over time (Table 3). During the course of the study, 23 participants were withdrawn for the following reasons: 6 (7.9%) were withdrawn due to non‐response (unable to contact), 9 (11.8%) had disease progression and/or initiated treatment, 6 (7.9%) moved out of the area or transferred their care to another healthcare facility, and 2 (2.6%) were diagnosed with another cancer (kidney, prostate) during the study. Thus, 53 participants (69.7%) completed the final assessment at the 24‐month follow‐up time point. Analyses were performed comparing baseline characteristics of study completers (n = 53) with non‐completers (n = 23). Study completers did not differ from non‐completers with regard to age, sex, race or ethnicity, or any of the psychosocial factors (stress, anxiety, and depressive symptoms). In addition, we conducted a sensitivity analysis excluding the 9 patients who had disease progression from the model. The results were similar and did not change. TABLE 3Perceived stress, Anxiety, and Depressive Symptoms Over Time in Patients with CLL/SLL.VariableFollow‐up time pointT2 (6 m)T3 (12 m)T4 (18 m)T5 (24 m) p Perceived stress12.3 (7.1)14.2 (8.4)16.0 (8.3)13.7 (7.8)0.19Anxiety4.7 (3.4)4.5 (3.9)4.6 (3.6)4.9 (3.9)0.79Depressive symptoms3.0 (2.8)2.9 (3.4)3.1 (3.3)3.4 (3.3)0.95 Perceived stress, Anxiety, and Depressive Symptoms Over Time in Patients with CLL/SLL. Examination of psychological distress with immune markers over time revealed several stable associations at follow‐up. At 12‐months post‐baseline, higher levels of stress remained significantly associated with lower percentages of IM (β = −0.51, p‐value < 0.02), consistent with the pattern observed at study entry. But the positive association of stress with higher percentages of CM was attenuated at 12‐months (β = 0.51, p = 0.06). Anxiety was also significantly associated with higher percentages of CM at 12‐month post‐baseline (β = 2.00, p‐value < 0.01). However, associations of anxiety with IM that were observed at baseline were attenuated to non‐significance. Similarly, associations of depressive symptoms with NKp30 and PD‐L1 expression on NK and T cell populations remained negative but were not statistically significant at follow‐up time points. Discussion The present study is one of the first to report associations of psychological distress with differential monocyte distribution in a cohort of patients with CLL/SLL followed for up to 2 years. Notably, we observed that stress was associated with a lower percentage of IM, and higher anxiety was significantly associated with both a lower percentage of IM and a greater percentage of CM. Importantly, several of these associations continued to be observed over time. Specifically, patients with higher levels of stress still had a lower percentage of IM 12 months later. Similarly, patients with higher levels of anxiety had a higher percentage of CM at 12‐month follow‐up, although other associations were attenuated over time. These results are consistent with the pattern of findings observed in a prior cross‐sectional study of stress in patients with relapsed/refractory CLL [21], while extending prior work by capturing longitudinal connections over time. The current findings suggest that higher levels of anxiety and perceived stress were associated with a significant shift from a lower percentage of intermediate to a higher percentage of classical monocytes. Whereas intermediate monocytes are responsible for the proliferation and stimulation of T cells, classical monocytes play a primary role in supporting inflammation [54, 55]. Hence, the observed shift to a higher percentage of classical monocytes could reflect a more pro‐inflammatory environment, which may have negative implications for health and is consistent with prior studies of stress and inflammation [56, 57]. Indeed, the ratios and functions of different monocyte subsets in peripheral blood are known to be altered by a variety of physiologic conditions, including infections, tissue injury, and inflammation [37]. These changes in proportions of peripheral monocytes can be induced by short‐term production of cytokines (such as those induced by psychological stress [58]) or longer‐term epigenetic changes that alter monocyte production or induce migration into tissue [37]. While we can only speculate on the mechanism, our results indicate that increased anxiety and perceived stress in CLL/SLL patients is increasing the pool of the early stage, short‐lived “sensing” CM in the blood, which could result from either greater monopoiesis in bone marrow, enhanced release from peripheral stores into blood, reduced migration into tissues, or decreased rate of differentiation into antigen‐presenting IM and vascular patrolling NCM subpopulations [47]. Regardless, these findings suggest that anxiety and stress in the context of indolent CLL/SLL are associated with physiological alterations in peripheral monocyte biology, which may have consequences for overall immune fitness and anti‐tumor responsiveness. Over time, patients with high anxiety continued to have higher percentages of CM, which may be detrimental for disease progression. A recent study reported that the percentages of CM were higher in CLL patients with advanced disease [59]. In light of those findings, given that higher anxiety is associated with skewing toward an increased percentage of CM, the shift away from IM and NCM may contribute to reduced anti‐tumor immunity and thereby promote disease progression. It is interesting to note that anxiety and perceived stress showed similar patterns of associations across markers, which were separate from depressive symptoms. Although anxiety and depression are closely associated and often co‐occur, some symptoms are distinct, and these states can be experienced differently by individuals [60]. Whereas anxiety is often characterized by nervousness or apprehension with physiologic symptoms of increased heart rate and blood pressure, depressive features are more likely to include a lack of energy, feeling hopeless, or moving or talking more slowly than usual. Such differences could manifest in distinct immune profiles, as demonstrated in a study of individuals with depression, anxiety disorders, or both [61]. In that study, individuals with depression had a more pro‐inflammatory profile, as well as higher levels of triglycerides, unlike those with anxiety, who did not show evidence of greater inflammation or metabolic alterations [61], suggesting that the physiologic pathways likely differ. In the present study, the associations observed with depressive symptoms were attenuated following correction for multiple testing. This could be due not only to the modest sample size but also to the relatively low levels of depressive symptoms reported by study participants, which likely limited the ability to discern any associations if such associations do exist. Thus, these novel findings, which should be viewed as hypothesis‐generating, are presented primarily to help inform and guide future studies with larger patient cohorts. Overall, the findings from the present study are consistent with a biobehavioral framework of cancer stress proposed by Andersen and colleagues [18] and others [19, 62, 63] and add to our understanding of immunologic correlates of psychological distress in an untreated patient population. Indeed, our finding offers a potential biologic mechanism to explain in part the observation from large‐scale studies that stress and depression can be associated with disease recurrence [64, 65, 66] and worse survival in patients with cancer [67, 68]. The present findings are also of interest given that immune deficiencies have been identified as one of the strongest known risk factors for lymphoma [25]. Studies have demonstrated that NK cells and T cells play important roles in the immune surveillance of indolent NHL [69]. Among 166 newly diagnosed, untreated CLL patients, greater host immunity (as assessed by NK cells) was associated with a more indolent disease course compared to patients with lower NK‐cell ratios [70], and it has been proposed that this was due to superior immune regulation of their disease. Although the clinical implications of the observed shift in monocyte distribution are unknown, the findings from this longitudinal study offer an interesting examination of how psychological distress may be associated with blood‐based immune parameters cross‐sectionally and over time in untreated patients. Due to the exploratory nature of the current findings, these associations should be validated in larger studies. Interestingly, participants in the present study were relatively younger than the broader population of patients with CLL/SLL (national statistics report the median age at diagnosis to be around 70 years) [71], and the younger age could have influenced study findings. Although analyses adjusted for age, it is well‐established that age is linked to changes in various components of innate immunity [72, 73]. Several studies have also suggested that psychological stress may amplify immunosenescence, thus contributing to greater vulnerability among older individuals to age‐related conditions [74]. If older individuals have lower immune capacity, this could compound the detrimental effects of stress on physical health and well‐being. Thus, potential age‐related effects and how they might influence associations between psychological factors and immune markers warrant further investigation. This study had several strengths, including the prospective longitudinal study design and repeated assessments over time, but we also acknowledge several limitations of the study, including the relatively brief length of follow‐up (i.e., up to 2 years). However, the focus on a well‐defined patient population, combined with serial sampling of comprehensive immunologic measures over time, offers an especially unique and informative context within which to assess the contribution of psychosocial factors on novel immunologic pathways in a sample of untreated patients with cancer. In addition, in an effort to minimize participant burden, study assessments were focused primarily on selected psychological constructs of anxiety, depressive symptoms, and perceived stress, given that the bulk of evidence on psycho‐immunologic pathways is centered on these factors [19, 66]. The selection of these measures also enables potential comparisons with previously published studies that included the same assessments [15, 46]. Due to the limited sample size and some missingness of clinical variables, however, the models did not incorporate key prognostic variables (e.g., IGHV status, TP53 mutations) that may influence biologic mechanisms and disease progression. As a result, the observed associations should be interpreted with caution. Finally, the study was not designed to adjust for multiple testing; as such, we were underpowered to detect associations after correction for multiple testing. Thus, as expected, in secondary analyses using Benjamini‐Hochberg correction, many associations were no longer statistically significant. Despite these limitations, the present study is among the first to explore novel biobehavioral pathways in an untreated patient population that often experiences high distress. In conclusion, our findings suggest that psychological distress is associated with shifts in monocyte distribution in patients with untreated CLL/SLL. Future studies are needed to investigate whether these immunologic changes are associated with subsequent clinical outcomes such as disease progression or the occurrence of second primary tumors, given the relevance of immune surveillance in cancer. Author Contributions Carolyn Y. Fang: conceptualization (lead), data curation (supporting), funding acquisition (lead), investigation (equal), methodology (equal), project administration (equal), resources (equal), supervision (equal), writing – original draft (lead). Alexander W. MacFarlane IV: data curation (equal), formal analysis (supporting), investigation (supporting), methodology (supporting), visualization (equal), writing – review and editing (equal). Henry C. H. Fung: investigation (equal), project administration (supporting), writing – review and editing (supporting). Nadia Khan: investigation (equal), writing – review and editing (supporting). Stefan K. Barta: investigation (equal), writing – review and editing (supporting). Jill S. Hasler: data curation (equal), formal analysis (equal), writing – original draft (equal), writing – review and editing (equal). Hatcher J. Ballard: data curation (supporting), investigation (supporting), writing – review and editing (supporting). Daniel J. Landsburg: investigation (equal), writing – review and editing (supporting). Sunita D. Nasta: investigation (equal), writing – review and editing (equal). Stephen J. Schuster: investigation (equal), writing – review and editing (supporting). Rashmi Khanal: investigation (equal), writing – review and editing (supporting). Richard I. Fisher: investigation (equal), writing – review and editing (supporting). Adam D. Cohen: conceptualization (supporting), data curation (equal), funding acquisition (supporting), investigation (equal), methodology (equal), project administration (equal), resources (equal), supervision (equal), writing – review and editing (equal). Jakub Svoboda: conceptualization (supporting), data curation (equal), funding acquisition (equal), investigation (equal), methodology (equal), project administration (equal), resources (equal), supervision (equal), writing – review and editing (equal). Kerry S. Campbell: conceptualization (lead), data curation (equal), formal analysis (equal), funding acquisition (lead), investigation (equal), methodology (equal), project administration (equal), resources (equal), supervision (equal), visualization (equal), writing – original draft (equal). Funding This work was supported by the National Cancer Institute, P30 CA06927, R01 CA194263, and U54 CA221705. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Antibody Staining Panel. Each column lists the antibodies added to a distinct staining tube of 1 million PBMCs, with target antigens listed and conjugated fluorophore in the first column. Commercial suppliers, catalog numbers, and monoclonal antibody clones are listed. Abbreviations: BL = BioLegend; BD = BD Biosciences; eB = eBiosciences; SB = Southern Biotechnology; M = Miltenyi; Nivo = nivolumab (anti‐PD‐1); hu = human; Ms. = mouse; PI = propidium iodide. Table S2: Immune parameters quantified in different immune subsets were defined as percentage or absolute number of cells exhibiting the listed characteristic or expression level of a biomarker on the indicated cell subset as geometric mean fluorescence intensity (GMFI). Abbreviations: Non‐Classical Monocytes (NCM), Intermediate Monocytes (IM), Classical Monocytes (CM). Figure S1: CONSORT Flow Diagram. Figure note: *Eligibility criteria for the parent study included patients newly diagnosed with indolent B‐cell‐derived neoplasms (including follicular lymphoma, CLL/SLL, marginal zone lymphoma, smoldering myeloma, and B‐cell lymphomas) managed with active surveillance. §Participants who missed an assessment could still complete subsequent follow‐up assessments. Figure S2: Lymphocyte Gating Strategy. A representative sample of the strategy used to gate lymphocyte populations from staining tubes T/NK #1 and T/NK #2 (Table S1) is shown. Prior to these gates, single cells were selected on a 2‐dimensional plot of forward scatter area × forward scatter height, and viable cells were gated as propidium iodide‐negative. (A–F) Labels on X‐ and Y‐axes indicate the cell parameters (SSC = side scatter) or cell surface biomarker expression in each 2‐dimensional dotplot or histogram. Individual boxed populations of cells were sequentially selected at each step and subgated to isolate the indicated lymphocyte populations. These populations were then used to examine expression levels of additional markers described in the manuscript. Figure S3: Monocyte Gating Strategy. A representative sample of the strategy used to gate monocyte populations from the myeloid staining tube (Table S1) is shown. Prior to these gates, single cells were selected on a 2‐dimensional plot of forward scatter area × forward scatter height, and viable cells were gated as propidium iodide‐negative. (A–E) Labels on X‐ and Y‐axes indicate the cell parameters (SSC = side scatter) or cell surface biomarker expression in each 2‐dimensional dotplot or histogram. (C) CD66b‐negative cell gating was used to remove neutrophils. (D) CD11c+ cell gating was used to further define monocytes. Individual boxed populations of cells were sequentially selected at each step and subgated to isolate the indicated myeloid and (E) monocyte populations (CM = classical monocytes, IM = intermediate monocytes, NCM = non‐classical monocytes).