Lymphocyte Subsets in Defense Against New Pathogens in Patients With Cancer

INTRODUCTION Immunity in cancer patients is modiﬁ ed both by the cancer itself and by oncospeciﬁ c treatments. Whether a patient’s adaptive immunity is impaired depends on their levels of naive lymphocytes and other cell populations. During the COVID-19 pandemic, cancer patients are at greater risk of progressing to severe forms of the disease and have higher mortality rates than individuals without cancer, particularly while they are receiving cancer-speciﬁ c therapies. An individual’s protection against infection, their response to vaccines, and even the tests that determine the humoral immune response to SARS-CoV-2, depend on lymphocyte populations, meriting their study. OBJECTIVE Estimate blood concentrations of lymphocytes involved in the immune response to new pathogens in cancer patients. METHODS We carried out an analytical study of 218 cancer patients; 124 women and 94 men, 26–93 years of age, who were treated at the National Oncology and Radiobiology Institute in Havana, Cuba, March–June, 2020. Patients were divided into ﬁ ve groups: (1) those with controlled disease who were not undergoing cancer-speciﬁ c treatment; (2) those undergoing debulking surgery; (3) patients undergoing chemotherapy; (4) patients undergoing radiation therapy and (5) patients currently battling infection. We evaluated the following peripheral blood lymphocyte subsets via ﬂ ow cytometry: B lymphocytes (total, naive, transitional, memory, plasmablasts and plasma cells); T lymphocytes (total, helper, cytotoxic and their respective naive, activated, central memory and eﬀ ector memory subsets); and total, secretory and cytotoxic natural killer cells and T natural killer cells. We also estimated neutrophil/lymphocyte ratios. Lymphocyte concentrations were associated with controlled disease and standard cancer therapy. For variables that did not fall within a normal distribution, ranges were set by medians and 2.5–97.5 percentiles. The two-tailed Mann–Whitney U test was used to measure the eﬀ ect of sex and to compare lymphocyte populations. We calculated odds ratios to estimate lymphopenia risk. RESULTS All cancer patients had lower values of naive helper and cytotoxic T lymphocyte populations, naive B lymphocytes, and natural killer cells than normal reference medians. Naive helper T cells were the most aﬀ ected subpopulation. Memory B cells, plasmablasts, plasma cells, activated T helper cells, and cytotoxic central memory T cells were increased. Patients undergoing treatment had lower levels of naive lymphocytes than untreated patients, particularly during radiation therapy. The risk of B lymphopenia was higher in patients in treatment. The odds ratio for B lymphopenia was 8.0 in patients who underwent surgery, 12.9 in those undergoing chemotherapy, and 13.9 in patients in radiotherapy. CONCLUSIONS Cancer and conventional cancer therapies signiﬁ cantly aﬀ ect peripheral blood B lymphocyte levels, particularly transitional T helper lymphocytes, reducing the immune system’s ability to trigger primary immune responses against new antigens.


INTRODUCTION
In cancer patients, infections are a risk factor for morbidity and mortality, since these tend to be more severe due to secondary immunodefi ciency that can develop during the course of the disease and its treatment. The adjusted death rate for infections in individuals with cancer may be three times higher than the general population. [1,2] Patients with hematologic malignancies are at increased risk of infection, compared with patients who have solid tumors, especially when undergoing hematopoietic

Original Research
The immune response (IR) to infection is complex and requires a functioning immune system (IS) to achieve eff ective antimicrobial response. Viral infections require special attention, since antiviral treatments are not as eff ective for them as antibiotics are for bacterial infections. [8,9] Viral infection control depends largely on balancing the innate and adaptive IS, which infl uence infection and recovery. [8] The main cells in innate immunity for viral infection control are natural killer (NK) cells, plasmacytoid dendritic cells and neutrophils, which act immediately through extracellular traps. [10,11] Specifi c adaptive immunity requires more time to develop than innate immunity, especially in the case of new pathogens like SARS-CoV-2, requiring establishment of a new primary IR. The mature cells involved in adaptive immunity are: helper T lymphocytes (Th), cytotoxic T lymphocytes (Tc) and B lymphocytes, in their varying stages of diff erentiation (naive, activated, memory, eff ector and terminally diff erentiated cell forms). B lymphocytes diff erentiate into plasma cells that produce antibodies or specifi c immunoglobulins. [12] Naive lymphocytes are mature T or B cells that reside in peripheral lymphatic organs and in circulating blood, which have never encountered their cognate antigen and are therefore charged with recognizing new pathogens. Their concentration in blood can be measured, as they recirculate in their role in immunosurveillance against new antigens.
IS alterations in cancer patients create challenges in diagnosing and treating emerging infections, as has been the case for COVID-19. [13] Quantitative and qualitative alterations have been noted in both innate and adaptive IS cells in these patients, thus considering cancer a secondary immunodefi ciency. Naive T and B lymphocyte populations may be aff ected, which would compromise the primary immune response of Tc lymphocytes and immunoglobulin production in response to new pathogens such as SARS-CoV-2. [14][15][16] Patients with malignant hemopathies treated with stem cell therapy require special attention, due to the time required for IS reconstitution and the quality of IR after stem cell transplantation. [17] Infection as a comorbidity in cancer patients is welldocumented, [2,5,18] but few studies have investigated patient susceptibility to infection during epidemics based on alterations to immunopathogenic mechanisms. Changes to IS cells in cancer patients limit the use of treatment and screening strategies designed for the general population. Two of these limitations are of major concern: 1) serological screening tests for diagnosing infections that, due to their high sensitivity and feasibility of application, can result in false negatives due to the decreased function of B lymphocytes, resulting in fewer circulating antibodies produced in response to infection [15,16] and 2) prophylactic vaccination schedules that are not always as eff ective as in healthy people, due to IS alterations. This could necessitate modifi cations to the number and interval of vaccine doses, as well as deferring their application, in accordance with cancer stage and type. [19,20] Cancer patients are excluded from clinical trials testing new vaccines, but arguments can be made for their inclusion, because of their increased vulnerability to infection. [21] Knowing the eff ect of standard cancer-specifi c treatments on IS cell proportions in cancer patients could help in designing strategies for controlling epidemics such as COVID-19, including vaccination schedules and detection strategies-and help adjust them to the needs of cancer patients. We carried out this study to evaluate the composition of lymphocytes in blood necessary to trigger a primary immune response against new antigens in Cuban cancer patients.

METHODS
Design, participants and sampling We conducted a crosssectional analytical study in 218 cancer patients with various tumor stages and locations, from March through June 2020. We included 124 women and 94 men 26-93 years of age who were treated in the National Oncology and Radiobiology Institute (INOR) in Havana, Cuba. Patients with malignant hemopathies were not included.

Patients were divided into fi ve groups:
• Patients with controlled cancer without oncospecifi c treatment (n = 39): followup patients with no evidence of active disease were considered 'controlled' after at least 12 weeks since the end of their primary/adjuvant treatment. Breast cancer patients who fulfi lled this criteria and were on hormonal adjuvant therapy for 5 years or prophylactic treatment with zoledronic acid were also included. • Patients who had not started chemotherapy or radiotherapy, and underwent cytoreductive surgery as a therapeutic standard, 1-7 days post-surgery (n = 54). • Patients undergoing chemotherapy, regardless of the cycle and type of chemotherapy used, who did not undergo surgery or radiotherapy (n = 67). • Patients undergoing radiotherapy who had not received prior or concomitant chemotherapy or surgery (n = 44). • Patients diagnosed with acute infections confi rmed via testing (related or unrelated to oncospecifi c therapies) (n = 14).
Peripheral blood samples were obtained by antecubital venipuncture, 4 mL of which were deposited in Vacutainer tubes (Becton Dickinson, USA) with ethylene aminotetraacetic acid added as an anticoagulant. Samples were processed within the fi rst six hours after extraction.
Samples were prepared according to manufacturer specifi cations for cell surface immunophenotyping, with an unwashed reed blood cell (RBC) lysis protocol (VeraLyse; Beckman-Coulter RBC

Original Research
Lysis Buff er, France). We used a 10-color cytometer (Beckman-Coulter, France). 100 μL of blood was dispensed for staining with fl uorochrome-conjugated monoclonal antibodies ( We performed daily quality controls of the Flow-Check fl uorosphere cytometer, aligning the lasers and checking the water system. Fluorescence intensity was monitored with Flow-Set fl uorospheres from the same company. Data was processed with Kaluza Analysis V1.5a software (Beckman-Coulter, France), with a minimum of 50,000 events acquired. We used a manual and logical-sequential window, and the guidelines recommended by the Human Immunology Project's immunophenotype standardization. [22] We used published reference values. [23][24][25] Statistical analysis To defi ne value ranges, we evaluated normal distribution of variables via the Shapiro-Wilk test. Most variables did not follow a Gaussian distribution. Ranges were set through medians and 2.5-97.5 percentiles. We analyzed the eff ect of age on lymphocyte populations with a simple linear regression model, and applied the two-tailed Mann-Whitney U test to evaluate the eff ect of sex as well as the comparison between lymphocyte subpopulations. To evaluate association between variables, we calculated odds ratios. All tests were performed with an associated signifi cance level of p <0.05.

Ethics
The study was approved by the INOR ethics committee. Participants provided written informed consent and we followed the principles for human subject research established by the Declaration of Helsinki.
[26] Identifying information was kept confi dential. Diagnostic means were selected under the guiding principle of maximum benefi cence, the ethical norm of 'do no harm', and material accessibility.

RESULTS
Age and sex did not signifi cantly infl uence patient lymphocyte ratios.
Most naive T and B lymphocyte and NK cell populations were signifi cantly lower in cancer patients than the normal reference median. However, memory lymphocytes and activated Tc were elevated. The neutrophil-to-lymphocyte ratio (NLR) was within normal range (Table 1).
Signifi cant diff erences were observed in the estimated medians for all lymphocyte populations' relative concentrations in the fi ve groups ( Table 2). We found a wide range inside the 2.5-97.5 percentiles for all lymphocyte subpopulations in the diff erent groups, most notably in patients with untreated controlled disease and B lymphocyte subpopulations.
Patients with controlled disease had low total values of T lymphocytes and eff ector memory T lymphocytes (Th and Tc), but their naive T and B cells were not aff ected. Cytotoxic-type NK cells were lower than reference values.
Activated T lymphocytes, central memory, eff ector memory (Th and Tc) and transitional B lymphocytes were lower in patients who underwent cytoreductive surgeries. Cytotoxic-type NK cells were signifi cantly lower. Neutrophil values were higher in patients with infection, with a mean of 10,025 cells/μL, increasing the NLR. However, neutrophil values were normal in other groups and the NLR did not rise ( Table 2).
Patients undergoing chemotherapy had lower levels of naive Th and Tc cells, total and transitional B lymphocytes, and cytotoxic NK cells.
Patients with infections had lower levels of total T lymphocytesat the expense of naive Tc-of total and naive B lymphocytes, and of cytotoxic-type NK cells. The NLR was higher, with increased neutrophils.
In the standard treatment and infection groups, proportions of patients with low total B lymphocyte values were higher than those in the untreated group (Table 3). Only the radiotherapy group showed a signifi cant increase in patients with low levels of naive B lymphocytes compared to the untreated group. The proportion of patients with low naive Th and cytotoxic NK values was signifi cantly higher in the treated and infection groups. In the treated groups, the proportion of patients with low levels of total T lymphocytes was lower compared to untreated patients (Table 3).   In all treatment groups, transitional B lymphocytes were low, but were high in patients with infections and in untreated patients (variability was high in the latter group). In all groups, naive B lymphocyte medians were normal but signifi cantly lower in patients who had not undergone treatment and in patients with infections compared to treated patients (p = 0.001, Mann-Whitney U) ( Figure 1A).

Peer Reviewed
Patients with infections had the lowest total lymphocyte medians and heterogenous values with bimodal distributions around the first and third quartiles. More than half of chemotherapy patients had decreased naive Th lymphocytes, and those who had undergone surgery had a high dispersion of percentage values for these lymphocytes with a bimodal distribution similar to that described for total T lymphocytes. Naive Th lymphocytes were higher than established reference values in the untreated group. Patients in chemotherapy had low Tc lymphocyte values.
Cancer patients undergoing surgery had an eight-fold higher risk of low transitional B lymphocytes than those who did not receive oncospecifi c treatments. Cancer patients who underwent surgery had a lower risk for low concentrations of memory B lymphocytes, Th and Tc, with risk reductions of 7.8, 5.5 and 8.3 times, respectively (Table 4).
Chemotherapy patients had a 12.9-fold greater risk than untreated patients for total B-cell lymphopenia and were 32.5 times more likely to have decreased naive Th lymphocyte levels. Risk of transitional B-cell lymphopenia was 2.5 times higher in chemotherapy patients. However, chemotherapy was associated with protection against total and memory T lymphocyte depletion. The risk of lymphopenia was reduced 3.8 times for Th, 9.1 times for Tc, and 33.3 and 11.1 times for central memory Tc and eff ector memory Th lymphocytes, respectively.
Radiation therapy was associated with a 13.9-fold increased risk of total B-cell lymphopenia. Additionally, it was associated with 10.1-and 7.5-times higher risks of transitional B-cell lymphopenias and naive Th lymphopenias respectively (Table 4), but was a protective factor against central memory and eff ector memory Th and Tc lymphocyte lymphopenias.

DISCUSSION
Cancer patients suff er from IS dysfunction due to a failure in immune surveillance of malignant tumors.
As the disease progresses, IS defi ciency worsens, which explains immune tolerance of increasing tumor burdens. For this reason, cancer is considered a cause of secondary immunodefi ciency. [27] When standard therapies are applied-chemotherapy, [28,29] radiotherapy [30] and cytoreductive surgery [31,32]either alone or in combination, they lead to major

Original Research
Peer Reviewed immunodefi ciency. [29] Chemo-and radiotherapy, which mainly aff ect proliferating cells, can destroy IS cells, especially those of developing leucocyte populations. [27,33] The selective cytotoxic eff ect on these cells could infl uence the fact that no signifi cant decreases were found between men and women, or were associated with age.
In the case of lymphocytes, naive subpopulations require cycles of proliferation and diff erentiation to give rise to eff ector and memory cells, and are thus more aff ected than memory cells by chemotherapy and radiation. [3] Memory cells are mostly quiescent and are more abundant, as they have already undergone clonal expansion. [27,34] In our study, the decrease in naive lymphocytes was related to the eff ects of cancer and cancer-related therapies. These cells were aff ected by a double depletion mechanism, that of their precursors during maturation and that of naive clones when they are activated by cognate antigen recognition, as they undergo proliferation cycles in both conditions. This is why cancer patients have compromised defenses against infectious agents that are coming into contact with the body for the fi rst time. [6,29,34] The increase in memory B and T lymphocytes could be relative since their naive counterparts decreased in percentage. Memory Tc cells may increase, mainly in cancers that involve regional lymph nodes depending on disease stage, such as in breast cancer. [35,36] In these cases, tumor antigens arriving from IRs are systematically introduced to secondary lymphoid structures that generate memory clones that circulate in blood.
There is evidence that increases in memory B lymphocytes and Th lymphocytes are interrelated in some cancers, especially when there is a high density of B cells in tertiary lymphoid structures developed in the microenvironment. Increase of these populations is due mainly to clonal expansion, stemming from a great diversity of specifi cities, particularly in patients younger than 68 years. [37] Montfort concludes that the increase in memory B lymphocytes and antibody-producing plasma cells is due to antigenic diversity generated by the tumor, favored by chemotherapy. [38] The risk cancer patients will suff er IR alterations is increased when they undergo oncospecifi c therapies, manifested by a decrease in leukocytes and an increased risk of infection. Our fi nding of a normal NLR median for all cancer patients was related to a decrease in neutrophils, which could be due to the infl uence of cytotoxic therapies, as they have a rapid turnover in circulation, with continuous replacement of new cells produced in bone marrow. [39] However, patients with infections had neutrophilia with high indices, contributing to the wide range we observed-up to 11.4.
NLR is an important biomarker for prognosis of these patients, since its increase is associated with poor outcomes. [40] This correlates with the literature, since patients undergoing treatment with chemo-and radiotherapy generally have lymphopenia and a low NLR, sometimes requiring treatment with granulocyte colony stimulating factor. In patients with infections, this index increased, which suggests poorer prognoses. In patients that recover, cell restoration kinetics are diff erent for leukocytes and lymphocytes, as lymphocytes require more time to recover. Adaptive immunity takes anywhere from six months to a year to restore itself following chemotherapy. This delay can negatively impact patient prognosis after completion of cytotoxic therapies, as neutrophils normalize much faster than lymphocytes and thus NLR rises at the expense of persistent lymphopenia. [30,41] Other research evaluating the impact of cancer and its treatments on IS cell ratios is based on lineage analysis, but few studies examine distribution of naive or memory populations. Our study shows that the cell population most aff ected by cancer treatment is B lymphocytes, which helps explain why the humoral response is also aff ected. This coincides with publications by other authors; [31,[42][43][44] however, the main lymphocyte subpopulations involved in primary IR were also decreased in some study groups.
Recent research indicates that in chemotherapy patients, not only are naive cells lost, but memory cells (both B and T) are also diminished, manifested by a decrease of antibody titers against previously administered vaccines. This has been the case in antigen-dependent responses. [41,43] However, protection is preserved against latent viruses like cytomegalovirus, suggesting memory cytotoxic T cell preservation. This could be due to the fact that memory cells are mostly quiescent, and this type of therapy acts especially well on proliferating cells. [33,44] We observed the opposite in the defense against new infections produced by viruses like Zika, SARS-CoV-2 or West Nile, in which the naive cell repertoire is compromised and they are the ones generating the primary immune response. [30,43] Our results coincide with these fi ndings, as there was a signifi cant decrease in naive populations, which resulted in a compromised primary response. In the case of total T lymphocytes, no risk of

Original Research
lymphopenia was found in the chemotherapy group, possibly due to the relative estimation of their subpopulations, and the decrease in naive cells did not cause a percentage decrease in T cells. These results are especially important during the COVID-19 pandemic, when cancer patients are more susceptible to severe forms of the disease and death. [2,5] The tissues most aff ected by radiation are bone marrow and intestinal mucosa. Both are essential in the IS, since the fi rst produces naive lymphocytes and the second is one of the largest secondary lymphoid organs. Radiation's cytotoxic eff ects increase with dose and treatment time. The main mechanisms of radiationinduced death in lymphocyte populations are primary necrosis for T lymphocytes, secondary necrosis for B lymphocytes, and apoptosis for NK cells. NK lymphocytes are the most radiosensitive immune cells, followed by B lymphocytes and lastly, T cells. [38] Within T cell and B cell subpopulations, naive cells are the most radiosensitive, because they proliferate when activated. Although radiation therapy is localized, its eff ects on lymphocytes are systemic. [44] Dovšak showed radiotherapy's immunosuppressive eff ects in oral cancer patients and its negative impact on naive cells, even when irradiation was local. NK cells were also aff ected. This depression can persist longer than a year following treatment cessation. [31] Transitional B lymphocytes come from bone marrow, are precursors of naive cells, and complete their maturation in other organs. Cells with the CD38 hi subphenotype are known to be tolerant to circulating autoantigens, like tumor antigens, and the mechanism behind this tolerance is anergy, which often ends in apoptosis. [45] This could explain the depletion found in this lymphocyte subphenotype. In all groups studied, lymphopenia has multiple repercussions for patients ( Figure 1A). On one hand, the repertoire of naive cells derived from transitional cells is reduced and the generation of humoral responses to new antigens is compromised-as is the case for new pathogens like SARS-CoV-2-but on the other hand this could lead to better cancer prognoses, as the cells that produce interleukin-10 are reduced at a systemic level, decreasing antitumor responses.
Cancer patients who underwent surgery had less risk of lower B lymphocyte, Th and total Tc lymphocyte values, which suggests tumor removal modifi es inter-department lymphocyte distribution, although it does not necessarily favor production of naive lymphocytes. Cytoreductive cancer surgery improves patient immune status as it eliminates the tumor microenvironment that produces both local and systemic immunosuppressive eff ects, but it does so by reducing suppressive lymphocytes. [46,47] However, these surgeries are a stressor and they produce a decrease in certain lymphocyte populations, although this decrease is temporary. [47] The decrease in transitional B lymphocytes after debulking surgeries could be related to infl ammatory response and surgical healing processes. After the operation, bone marrow increases production of red blood cells, neutrophils and T lymphocytes related to homeostasis. B lymphocytes are perhaps less necessary to this process and transitional cells belong to the fi nal stage of this lymphocytic lineage's maturation. Naive Th cells had widely dispersed percentage values and exhibited a bimodal distribution similar to that described for total T lymphocytes. This could be due to heterogenicity in the tumor microenvironment's cellular composition, which has a systemic impact and depends on tumor type and disease stage. [48] The prognosis for an unfavorable NLR evolution is greater when dependent on a decrease in lymphocytes. NLR increase denotes dysfunctional and sometimes suppressive infl ammation, which translates clinically into reduced patient survival and poor tumor cell response to therapies inhibiting PD1 receptors. [49] It is also associated with poor prognoses in infections with uncontrolled infl ammatory reactions like those seen in COVID-19, [50] so this indicator should be measured in cancer patients, especially during oncospecifi c therapies. Absence of a high NLR in all study participants is because although patients with infections had neutrophilia and increased NLR, patients who received chemoand radiotherapy had neutropenia and decreased NLR.
Antibody levels against pretreatment infections, including COVID-19, are lowered in cancer patients treated with chemotherapy. [15,16] In hematologic malignancies like leukemia, the antibody response to vaccines is also aff ected before starting chemotherapy, so the defi cit caused by the disease is added to that caused by the therapy, implying that patients should be reimmunized three to six months after fi nishing treatment. [51][52][53] Antibody production is aff ected in cancer patients, so it would be advisable during new epidemics to carry out diagnostic tests that detect the causal agent-rather than estimating antibody presence in response to the causal agent-as this could be decreased or absent and result in false negatives, even if the patient has the infection.
Vaccination strategies in cancer patients should be based on sound scientifi c evidence, to not deprive them of vaccines or subject them to unnecessary risks. Indication for patient vaccination depends on vaccine design, cancer type, the state of the patient's immune system, and treatment timing and type. In our experience with cancer patients, protective eff ects of immunization are obtained one month after immunosuppressive treatment cessation. However, in the case of malignant blood diseases like leukemia, vaccination is not always recommended, as treatments are usually prolonged. Malignant B cell neoplasms such as lymphoma, chronic lymphocyte leukemia and multiple myeloma will likely have diff erent antibody responses to other cancer types in which the cells responsible for producing antibodies in response to infection are not aff ected by treatment. This is important when deciding whether to administer vaccines, where protection is related to antibody production, so indication for vaccination depends on vaccine design and expected response.
In addressing pandemics like COVID-19, vaccination strategies for cancer patients must be considered since high percentages of the population require vaccination and the decision whether to vaccinate patients or their cohabiting relatives should be prioritized. This decision must take into account cancer type, stage, the type of oncospecifi c treatment, the possible response of IS cells to these conditions and other preexisting comorbidities, as well as the type of vaccine to be administered.
Considering our results, it is recommended that vaccinesparticularly those with attenuated live agents-not be administered during convalescence from major surgeries, in the fi rst cycles of chemotherapy, or during full-dose total body irradiation. The

Original Research
proposal would be to vaccinate with a fi rst dose two weeks before treatment or two weeks after treatment, at minimum, and fi nish the immunization schedule before starting another treatment cycle. [54,55] Although this study was aimed at obtaining an overview of cancer as a group of diseases and establishing the eff ects of modifi able factors like therapeutic standards and infections, a stratifi ed study examining age, sex, cancer type, disease stage and particularities within treatment groups is needed. Non-inclusion of these variables and stratifi cation criteria constitute the main limitations of the present study.

CONCLUSIONS
Cancer and its therapeutic standards signifi cantly aff ect levels of NK cells and B lymphocytes in peripheral blood, particularly transitional B cells, and reduce percentages of naive T helper lymphocytes. The radiotherapy group was the most aff ected. These alterations reduce the IS's ability to trigger eff ective immune responses to new antigenic challenges, including when a patient fi rst encounters a new virus or a new vaccine. Due to IS eff ects in these patients, management of new infections and epidemics must diff er from those established for the general population, including vaccination strategies and diagnostic methods that rely on detecting antibodies against specifi c pathogens.