Peripheral T cell profiling reveals downregulated exhaustion marker and increased diversity in lymphedema post-lymphatic venous anastomosis

Summary Lymphedema is a progressive condition accompanying cellulitis and angiosarcoma, suggesting its association with immune dysfunction. Lymphatic venous anastomosis (LVA) can provide relief from cellulitis and angiosarcoma. However, the immune status of peripheral T cells during lymphedema and post-LVA remains poorly understood. Using peripheral blood T cells from lymphedema, post-LVA, and healthy controls (HCs), we compared the profile of T cell subsets and T cell receptor (TCR) diversity. PD-1+ Tim-3 + expression was downregulated in post-LVA compared with lymphedema. IFN-γ levels in CD4+PD-1+ T cells and IL-17A levels in CD4+ T cells were downregulated in post-LVA compared with lymphedema. TCR diversity was decreased in lymphedema compared with HCs; such TCR skewing was drastically improved in post-LVA. T cells in lymphedema were associated with exhaustion, inflammation, and diminished diversity, which were relieved post-LVA. The results provide insights into the peripheral T cell population in lymphedema and highlight the immune modulatory importance of LVA.


The influence of lymphatic venous anastomosis on peripheral T cells was elucidated T cells after lymphatic venous anastomosis demonstrated upregulation of diversity
Our results highlight the importance of LVA apart from its edema-reducing effect

INTRODUCTION
Lymphedema results in the swelling of limbs because of lymph retention following resection, radiotherapy, and lymph node dissection as part of cancer therapy. 1 It is estimated that 20%-40% of patients who undergo treatment for solid malignancies, such as breast cancer, melanoma, gynecological or urological tumors, or sarcomas, develop lymphedema. 2 Patients with lymphedema develop progressive fibroadipose deposition in the affected limb and are at an increased risk of developing cellulitis and angiosarcoma, suggesting the coexistence of immune dysfunction. 3 The incidence of cellulitis in patients with lymphedema is 10%-40%, 4,5 and approximately 20% of these patients experience recurrence of cellulitis that occasionally leads to sepsis and mortality. 6,7 Angiosarcoma is a rare but fatal tumor that develops 10-15 years after primary cancer therapy, including surgery and radiotherapy. 8 Even with the treatment available for angiosarcoma, the prognosis remains poor, with the median duration to mortality reported to be as short as 10.5 months. 8 Lymphatic venous anastomosis (LVA) is a surgical treatment that improves lymphatic drainage by anastomosing the lymphatic vessels to a cutaneous vein under surgical microscopy. 9 LVA can reduce the circumference of the lymphedematous extremity with low invasiveness, and cellulitis infections are significantly reduced after treatment with LVA. 10,11 Notably, Koshima et al. reported the effectiveness of LVA that can treat lymphedema-related angiosarcoma, including lung metastatic lesion. 12 Increased risk of cellulitis and angiosarcoma has been considered a manifestation of immune dysfunction related to lymphedema, and LVA can influence the immunity of lymphedema. 13 The analysis of lymphedematous tissue revealed that T cells, particularly Th2 cells, contribute to the development of lymphedema. 14 Recently, a clinical trial reported on IL-4/IL-13-neutralizing antibodies targeting Th2 cells 15 ; treatment with IL-4/IL-13-neutralizing antibodies improved quality of life measurements, skin stiffness, and histological changes in the lymphedematous extremity; however, it appeared to be less effective for the reduction of edema. Hence, lymphedema is intimately associated with the formation of a particular immunological environment, especially that of T cells; however, the immune profiling of peripheral T cell populations in lymphedema has not been investigated comprehensively. To enhance our understanding of the immune status in patients with lymphedema and those who have undergone post-LVA, a more detailed investigation of peripheral T cells is warranted. The principal aim of this study was to elucidate the characteristics of peripheral T cell subpopulations and T cell receptor (TCR) repertoire in patients with lymphedema and investigate the alterations in T cell profiles after treatment with LVA.

Clinical efficacy
A median number of 6 (5.5-7) LVAs were performed per patient. The upper extremity lymphedema index in post-LVA decreased compared with that in lymphedema, although it was not significant (126. 7 Figure 1C). (C) Comparison between lymphedema and post-(LVA) in terms of severity of the EL index. *p < 0.05, **p < 0.01. Nonparametric, independent, and paired continuous variables were compared using Mann-Whitney U test and Wilcoxon's rank-sum test, respectively. Data are represented as median +/À interquartile range. iScience Article Leukocyte count, lymphocyte count, and CD4/CD8 ratio in CD3 + cells in peripheral blood were not significantly different among lymphedema, post-LVA, and healthy controls (HCs) (Table S2).
PD-1, Tim-3, Lag-3, and PD-1 + Tim-3 + expression on peripheral CD4 + and CD8 + T cells in patients with lymphedema, post-LVA, and HCs To understand the role of the exhaustion marker on CD4 + T cells in lymphedema, we examined the expression patterns of programmed death-1 (PD-1), T cell immunoglobulin and mucin domain-containing-3 (Tim-3), lymphocyte activation gene-3 (Lag-3), and PD-1 + Tim-3 + , which is known as the exhaustion marker, on peripheral CD4 + T cells in lymphedema, post-LVA, and HCs. PD-1, Tim-3, and PD-1 + Tim-3 + expression on CD4 + T cells showed significant downregulation in post-LVA compared to lymphedema; 30.  -2D and Table S3). Additionally, PD-1, Lag-3, and PD-1 + Tim-3 + expression on CD4 + T cells showed a significant upregulation in patients with lymphedema compared to HCs. Despite the downregulation of PD-1 expression on CD4 + T cells in post-LVA compared with lymphedema, PD-1 expression post-LVA remained significantly higher than the expression in HCs. Next, we investigated the correlation between the expression of the exhaustion marker and the clinical severity of lymphedema as well as the correlation between the change in expression of the exhaustion marker and the reduction in edema by LVA. We could not identify any correlation between the expression of the exhaustion marker and the severity of EL index. Furthermore, we could not identify any correlation between the rate of change in the expression of the exhaustion marker and the rate of improvement in the EL index by LVA ( Figure 2E). We next examined the expression patterns of PD-1, Tim-3, Lag-3, and PD1 + Tim-3 + on peripheral CD8 + T cells in lymphedema, post-LVA, and HCs. PD-1, Tim-3, and PD-1 + Tim-3 + expression on CD8 + T cells showed significant downregulation in post-LVA compared to lymphedema; 17.1 (11.7-25)% vs. 15 -3D and Table S3), respectively. PD-1 + Tim-3 + expression on CD8 + T cells was positively correlated with the severity of the EL index ( Figure 3E). The rate of improvement Treg population in patients with lymphedema, post-LVA, and HCs We compared Tregs and three distinct subpopulations of Tregs (Treg I, Treg II, and Treg III) between patients with lymphedema, post-LVA, and HCs to understand the relationship between Treg and lymphedema. The frequency of Tregs and the three Treg subpopulations had not changed significantly between lymphedema and post-LVA (Table S4). The total proportion of Tregs (including the proportion of Treg I, Treg II, and Treg III) in CD4 + T cells was significantly higher in lymphedema compared to HCs ( Figure 4A). Although the proportion of Treg I in CD4 + T cells was similar between lymphedema and HCs ( Figure 4B), it was notable that the proportion of Treg II and Treg III in CD4 + T cells was significantly higher in lymphedema compared to HCs ( Figures 4C and 4D). We could not identify any correlation between the Treg populations and the EL severity index ( Figure 4E).

Comparison of naive and memory CD4 + T cells between patients with lymphedema, post-LVA, and HCs
To understand the features of the activation state in peripheral CD4 + T cells in lymphedema, we compared the proportions of naive and memory CD4 + T cells among the total CD4 + T cells (Table S5). The number of CCR7 + CD4 + T cells was lower in lymphedema compared with that in HCs, and that the number of CCR7 À CD4 + T cells was higher in lymphedema compared with that in HCs; however, significant changes were not observed between lymphedema and post-LVA ( Figure 5A). The proportion of naive, stem celllike memory T cells (Tscm), memory T cells with naive phenotype (Tmnp), central memory (CM), and effector memory (EM) cells among CD4 + T cells was not changed between lymphedema and post-LVA but was accompanied by an increasing proportion of terminal effector (TE) cells in post-LVA ( Figures 5B-5H). The proportion of naive T cells was decreased in patients with lymphedema compared with HCs, and the proportion of EM cells was increased in patients with lymphedema compared with HCs. We could not identify any correlation between the populations of naive and memory phenotypes among CD4 + iScience Article T cells and severity of the EL index ( Figure S1). We also compared the naive and memory subpopulations among CD8 + T cells (Table S5). In contrast to CCR7 expression in CD4 + T cells, the number of CCR7 + CD8 + cells was significantly decreased, whereas that of CCR7 À CD8 + cells was significantly increased in post-LVA compared with that in lymphedema ( Figure 6A). The population of naive, Tscm, and CM cells was decreased, whereas that of TE cells was increased in post-LVA compared with that in lymphedema ( Figures 6B-6H). The population of naive and Tmnp cells was decreased, whereas that of TE cells was increased in patients with lymphedema compared with that in HCs. Similar to that in CD4 + T cells, we could not identify any correlation between the population of naive and memory phenotypes in CD8 + T cell and severity of the EL index ( Figure S2).

Comparison of cytokine production in CD4 + and CD8 + T cells among patients with lymphedema, post-LVA, and HCs
Inflammatory cytokines are strongly involved with the progression of lymphedema. 20 To understand the correlation between cytokine production by T cells and lymphedema, we compared the expression of IFN-g, IL-4, and IL-17A on CD4 + and CD8 + T cells in lymphedema, post-LVA, and HCs (Table S6)  To understand the diversity of peripheral T cells in lymphedema, we analyzed the TCR b repertoire using peripheral blood mononuclear cells in patients with lymphedema, post-LVA, and HCs. No significant difference was noted in sequence reads among patients with lymphedema, post-LVA, and HCs, except that unique reads were significantly lower in patients with lymphedema and those with post-LVA compared to HCs (  -9D). Furthermore, these indices indicated a significantly lower diversity of T cells in patients with lymphedema compared with that in HCs. The diminished variety of TRBV and J combination in each patient with lymphedema compared with post-LVA and HCs is shown in Figure S3.
Expression profile of the TCR b clones in patients with lymphedema, post-LVA, and HCs To examine the features of the TCR b clonotypes that may characterize the immune environment of lymphedema, we compared the frequency of TRBV and TRBJ in patients with lymphedema, post-LVA, and HCs ( Figures 10A and 10B). The frequency of TRBV3-1 was reduced, whereas that of TRBV6-1, 6-5, and 18, and TRBJ2-6 was increased in patients with post-LVA compared with that in lymphedema. With (H) Total demographic of naive and memory phenotype frequencies in CD4 + T cells. *p < 0.05, **p < 0.01, ***p < 0.001. Non-parametric, independent, and paired continuous variables were compared using Mann-Whitney U test and Wilcoxon's rank-sum test, respectively. Data are represented as median +/À interquartile range. iScience Article regard to combinations of TRBV and TRBJ, three combinations (TRBV5-1/J1-1, TRBV5-1/J2-7, and TRBV29-1/J1-2) reduced TRB in lymphedema compared with post-LVA. In addition, the three combinations reduced TRB clonotypes in patients with lymphedema compared with that in HCs. The TRBV5-1/ J1-1 and TRBV5-1/J2-7 combinations were reduced in patients with lymphedema compared with that in HCs and increased in post-LVA compared with that in lymphedema. We could not identify specific increased clones in lymphedema. The increased and reduced TCR clones between lymphedema and post-LVA are noted in Tables S8 and S9.

DISCUSSION
In the present study, we analyzed fresh peripheral blood samples and characterized the phenotypes and functional profiles of T cells with comprehensive analysis of the TCR repertoire in patients with lymphedema through post-LVA as a consequence of surgical cancer treatment. In the past, several studies have indicated that local T cells deposited in the tissue are related to the progression of lymphedema. [21][22][23][24] In particular, Th2-biased inflammatory responses in lymphedema tissue promote tissue fibrosis via increased collagen deposition and aggravate inflammation. 14 In contrast, increased Treg deposition in lymphedema tissue ameliorates lymphedema development and inflammation. 25,26 In a human analysis of the influence of LVA on lymphedema, LVA reduced the proportion of CD4 + T cells in lymphedematous tissue, which may be directed back into circulation in the peripheral blood. 27 Thus, the association between tissue-deposited local T cells and lymphedema inflammation and progression is strongly suggested. However, to the best The PD-1 regulatory pathway plays indispensable roles in downregulating the immune response and in promoting tolerance to self-antigens by suppressing T cell activation through B7-CD28 co-stimulatory molecules that deliver critical inhibitory signals. 28 Tim-3, a member of the T cell Ig and mucin domain-containing molecule superfamily, is a key regulatory molecule for the Th1 response. 29 PD1 + Tim-3 + co-expressing T cells exhibit reduced proliferation and impaired cytokine production, which is called terminal exhaustion. 30 Lag-3 exhibits high affinity to major histocompatibility complex class II and regulates the proliferation, activation, and function of T cells. 31 In the present study, the expression of PD-1, Tim-3, and PD-1 + Tim-3 + on CD4 + and CD8 + T cells in lymphedema was significantly downregulated in post-LVA, which still showed upregulated expression, compared with that in HCs. Two theories can be considered to explain Treg cells compete for the T cell growth factor IL-2 via the expression of high-affinity IL-2 receptor complexes and exert direct suppressive activity by secreting immunosuppressive cytokines such as TGF-b iScience Article and IL-10. 34 Treg I cells proliferate themselves upon T cell receptor stimulation and convert to Treg II cells. 35 The Treg II subset is functionally important with regard to its potent suppressive function, which is related to its high expression of CTLA-4 and CD25 and its higher sensitivity to IL-2 than that of other Treg subpopulations. Treg III secretes a high amount of effector cytokines (IL-2, IL-17, and IFN-g) without suppressive activity. Treg III cells may be a heterogeneous subset between Treg cells and effector T cells. 35 The proportion of peripheral Treg cells is increased in patients with cancer even after tumor resection compared with HCs. 36 In the present study, Treg, in particular Treg II, was upregulated in patients with lymphedema; however, the upregulation was not significantly changed post-LVA. In this study, all patients with lymphedema had undergone cancer treatment. The upregulation of Treg in patients with lymphedema may be influenced by the history of cancer treatment.
Recent evidence suggests that by guiding cells to and within lymphoid organs, CCR7 essentially contributes to both immunity and tolerance. 37 Furthermore, CCR7 + expression is downregulated with aging, chronic inflammation, and experience of cancer treatment. [38][39][40] In the present study, CCR7 + expression in CD8 + T cells was downregulated in post-LVA compared with that in lymphedema. In contrast, CCR7 + expression was not significantly changed in CD4 + T cells. The downregulation of CCD7 + expression in CD8 + T cells may be influenced by aging. In the present study, the number of naive T cells was significantly lower in patients with lymphedema compared to age-matched HCs. However, the depletion of naive T cells was not correlated with lymphedema severity. Hence, the depletion of naive T cells in patients with lymphedema may be influenced by the fact that patients have undergone cancer treatment. Tmnp in CD8 + T cells, which express increased levels of CD49d and CXCR3, produce cytokines but maintain a naive phenotype. 41 The frequency of Tmnp increases with age and after severe acute infection; however, the frequency of Tmnp in this study was reduced in patients with lymphedema and post-LVA compared with HCs. 41 Further sequencing of this population is warranted to detect the cause of this reduction in the frequency of Tmnp in patients with lymphedema. The inflammatory cytokines IFN-g, IL-4, and IL-17A have been reported to negatively control the formation of lymphatic vessels involved in the progression of lymphedema. 20,42,43 Furthermore, Th1, Th2, and Th17 cells are reported to be correlated with the progression of fibrosis and fibrotic diseases. 44 Th1 cells are considered to play both pro-fibrotic and anti-fibrotic roles, while Th2 and Th17 cells are considered to play dominantly pro-fibrotic roles. 44 Elevated Th1 cytokine expression inhibits Th2-dominated immune response; however, Th1 cell infiltration leads to cardiac fibrosis. 44 In patients with lymphedema, Th2 cells in lymphedematous skin are considered to play an important role in the progression of lymphedema. 14 The analysis of plasma without stimulation could not reveal a significant difference in IFN-g, IL-4, and IL-17A, whereas some other cytokines (such as MIP-1b and IL-6) were found to be predictive markers of lymphedema when comparing breast cancer-related lymphedema and breast cancer without lymphedema. 45,46 We examined the inflammatory cytokine production in CD4 + T cells between lymphedema and post-LVA and found that IFN-g in CD4 + PD-1 + T cells and IL-17A in CD4 + T cells were downregulated iScience Article post-LVA compared with that in lymphedema, while IL-4 in CD4 + T cells was not significantly different. It is plausible that IFN-g has both pro-fibrotic roles for progression of lymphedema and anti-fibrotic roles for counter balancing the Th2-dominant milieu in lymphedematous skin. Hence, IFN-g is the most affected cytokine by lymphedema. Furthermore, inflammatory cytokine production from CD4 + T cells indicated that LVA can ameliorate chronic inflammation in lymphedema.
T cell diversity is required for protective immune responses. 47 Elderly patients with common variable immunodeficiency or autoimmune diseases show decreased T cell diversity in peripheral blood. [47][48][49] Diminished T cell diversity is thought to result in impaired antipathogen responses. 50 In the present study, compared with HCs, patients with lymphedema were associated with decreased TCR repertoire diversity, and this TCR skewing was drastically improved post-LVA. These results suggest that LVA can restore the TCR repertoire diversity in lymphedema and thus promote antipathogen responses. This phenomenon may also explain the mechanisms via which LVA reduces the frequency of cellulitis and provides relief from angiosarcoma in lymphedema. [10][11][12] Among the TCR b chains in the repertoire, we observed less frequent usage of TRBV5-1/J1-1 and TRBV5-1/J2-7 in patients with lymphedema, and its restoration to a pattern similar to that in HCs after LVA. These findings of diminished TCR repertoire diversity and decreased specific TCR chains might enhance our understanding of the immunodeficiency in patients with lymphedema.
To the best of our knowledge, this is the first study to demonstrate the peripheral T cell condition in lymphedema and to elucidate the influence of LVA on peripheral T cells. The upregulation of the exhaustion marker, IFN-g, and IL-17A and downregulation of the TCR repertoire diversity in patients with lymphedema compared with post-LVA and HCs can be associated with immune dysfunction and progression of lymphedema. T cells in lymphedema tend to lead to exhaustion and inflammation with diminishing diversity, which may be associated with immune dysfunction in lymphedema. Notably, LVA can release the exhaustion and inflammation status with upregulation of the diversity of T cells; therefore, this study highlights an advantage of LVA apart from its edema-reducing effect.

Limitations of the study
This study has some limitations. First, the relatively small number of patients consisting entirely of females is a cause of bias in this study. Second, 16 patients (76.2%) in this study had undergone chemotherapy. Waidhauser et al. reported that chemotherapy did not alter the population of T cells, while Krantz et al. reported that neoadjuvant chemotherapy reduced the expression of CD8 + exhaustion marker and the frequency of Treg. 51,52 The influence of chemotherapy on circulating T cells is not well known; hence, a future lymphedema study completed the background of chemotherapy is warranted. Third, contracting COVID-19 or receiving a vaccination for COVID-19 has a possible influence on TCR diversity for the timing of this study. We excluded participants with a history of COVID-19 from this study; however, we could not eliminate the possible effect of anti-COVID immune responses during the pandemic. 53 Fourth, we did not analyze similar T cell profiling data of age-matched patients with cancer but without lymphedema. However, we investigated T cell profiling of patients with lymphedema and after treatment of lymphedema; hence, it is plausible that the comparison of patients with lymphedema and post-LVA indicates the T cell profiling influenced by lymphedema. Yang et al. reported that lymphedema leads to dysregulated gene expression in circulating monocytes, which was restored post-LVA. 54 Lymphedema is not only a localized disease but also accompanied by systemic disease and LVA can restore healthy condition. A large-scale comprehensive analysis of systemic and local effects of lymphedema with regard to immunology is warranted for understanding the pathophysiology of lymphedema and developing new treatments.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
We thank Nanae Nakaju, Sachiko Fukumoto, Sajeda Chowdhury, Emi Nakai, Yoko Matsuzaki, and Masako Ninomiya for their excellent technical and secretarial assistance. We thank Gaku Aoki and Tomoyuki Akita for their excellent statistical assistance. This work was supported by grants from JSPS KAKENHI (Number JP21K09768 to H.I.), Takeda Science Foundation, and the Program of the network-type Joint Usage/ Research Center for Radiation Disaster Medical Science of Hiroshima University, Nagasaki University, and Fukushima Medical University (to T.I.). T.I. has received research funding from Repertoire Genesis Inc. The study sponsor had no role in the data collection, article writing, and submission process.

AUTHOR CONTRIBUTIONS
H.I. and T.K. contributed to study design, executed the experiments, and wrote the manuscript. T.I. and I.K. reviewed the study and proofread the manuscript. All other authors contributed to this study by collecting patient data.

DECLARATION OF INTERESTS
T.I. has received research funding from Repertoire Genesis Inc. No other authors have any association or financial involvement with any organization or commercial entity having a financial interest in or financial conflict with the subject matter or research presented in the manuscript.

Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Hirofumi Imai (imai_h61@yahoo.co.jp).

Materials availability
This study did not generate new unique reagents.

Data and code availability
The raw data of TCR b sequences has been deposited ad Zenodo (Zenodo: https://doi.org/10.5281/ zenodo.7854980), and is publicly available as of the date of publication. This paper does not report original iScience Article code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Patients and healthy controls
This prospective study included patients with a diagnosis of lymphedema and age-and sex-matched HCs without a history of cancer. All participants were consisting entirely of females and the median age of the enrolled patients was 54 (45-59.8) years. Written informed consent was obtained from the participants. This study was approved by the institutional review board of Hiroshima University (number: E-2019-9241) and conforms with the Declaration of Helsinki. The patient cohort was enrolled between August 2019 and September 2021. Lymphedema was diagnosed by histological examination and indocyanine green (ICG) lymphography ( Figures S4A and S4B). 57 The recruitment criteria were as follows: (a) more than 2 years passed since cancer treatment; (b) no active infection; (c) no tumor recurrence or metastasis; (d) no history of comorbid medical disorders (heart failure, renal failure, hepatic failure, endocrine abnormality, immunological disease); (e) unilateral lymphedema; and (f) no history of having undergone LVA. We excluded patients and HCs with a history of COVID-19 from this study based on the questionnaire due to the possible influence on immune function. Data on the type of cancer that caused lymphedema, duration of edema, the radiation therapy used for cancer, frequency of cellulitis, the Campisi clinical staging of lymphedema, 16

LVA operative technique
A small amount (0.25 mg/0.1 mL) of ICG (Diagnogreen Injection, Daiichi Pharmaceutical, Tokyo, Japan) was injected subcutaneously into the first web space in upper extremity lymphedema, lateral malleolus, and the lateral side of the superior edge of the knee in lower extremity lymphedema. Furthermore, 12-18 h after the injection, we observed the ICG lymphography results using a near-infrared imaging device (Photodynamic Eye; Hamamatsu Photonics, Hamamatsu, Japan) and classified them into types I to V, as reported previously. 17 LVA was performed under local anesthesia in all cases along a linear pattern or along the ulnar side of the upper extremity in upper extremity lymphedema, and greater saphenous vein course in lower extremity lymphedema in the area of the dermal backflow pattern. 58 The LVA procedures were performed in an end-to-end manner using 11-0 or 12-0 nylon micro sutures under a surgical microscope ( Figures S4C  and S4D). 9

Preparation of peripheral blood mononuclear cells (PBMCs)
A 15-mL aliquot of fresh peripheral blood was collected from each patient and HCs. PBMCs were isolated with Lymphoprep gradient (Axis-Shiel PoC AS, Oslo, Norway) and used for each experiment after confirming viability >95%, as determined by trypan blue staining.

Flow cytometry analysis
Cell surface marker staining of freshly isolated PBMCs was performed using appropriate combinations of fluorescein-conjugated anti-human antibodies. Briefly, cell suspensions (1 3 10 6 to 2 3 10 6 ) were incubated with a cocktail of the antibodies in the dark for 30 min at 4 C. Intracellular staining of Foxp3 was performed with the FOXP3 Fix/Perm Buffer Set (Biolegend) according to the manufacturer's instructions. After surface labeling, PBMCs were permeabilized in 1 mL of fixation/permeabilization buffer at 20 C for 45 min in the dark. The samples were then stained with the anti-human FoxP3 monoclonal antibody and incubated in the dark for 30 min at 20 C. For analysis of cytokine production, PBMCs were stimulated with PMA (50 ng/mL) and ionomycin (750 ng/mL) in the presence of brefeldin A (10 mg/mL) (BioLegend). Cells were surface-stained, fixed, permeabilized, and stained with anti-cytokine antibodies. Isotype-matched ll OPEN ACCESS iScience Article of CDR3 with the other sequence reads. The copy number of identical USR were automatically counted using RG software in each sample and then ranked in order of the copy number. Percentage occurrence frequencies of sequence reads with V and J genes in total sequence reads were calculated.

QUANTIFICATION AND STATISTICAL ANALYSES
Data are shown as median (IQR). JMP statistical software (SAS Institute Inc., Cary, NC, USA) was used for all statistical analyses. Non-parametric, independent, and paired continuous variables were compared using Mann-Whitney U test and Wilcoxon's rank-sum test, respectively. The Chi-square test was used for categorical variables. Statistical analyses between lymphedema and post-LVA were primarily conducted, and subsidiary analyses were conducted between lymphedema and HCs, and post-LVA and HCs. Spearman's rank correlation was used to evaluate the association between the quantitative indicators. Two-sided p-values <0.05 were considered statistically significant.