Plasmacytoid dendritic cells diminution in peripheral blood is prevalent in drug reaction with eosinophilia and systemic symptoms and may precede human herpesvirus 6 reactivation

Shao-Hsuan Hsu; Che-Wen Yang; Yi-Chun Hsieh; Kai-Lung Chen; Yung-Tsu Cho; Jau-Yu Liau; Chia-Yu Chu

doi:10.4103/ds.ds_37_21

ORIGINAL ARTICLE

Year : 2021 | Volume : 39 | Issue : 4 | Page : 175-181

Plasmacytoid dendritic cells diminution in peripheral blood is prevalent in drug reaction with eosinophilia and systemic symptoms and may precede human herpesvirus 6 reactivation

Shao-Hsuan Hsu¹, Che-Wen Yang², Yi-Chun Hsieh³, Kai-Lung Chen⁴, Yung-Tsu Cho³, Jau-Yu Liau⁵, Chia-Yu Chu³
¹ Department of Dermatology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei; Department of Dermatology, National Taiwan University Hospital Yun-Lin Branch, Yunlin, Taiwan
² Department of Dermatology, National Taiwan University Hospital and National Taiwan University College of Medicine; Department of Dermatology, Cathay General Hospital, Taipei, Taiwan
³ Department of Dermatology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan
⁴ Department of Dermatology, National Taiwan University Cancer Center, Taipei, Taiwan
⁵ Department of Pathology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan

Date of Submission	27-Jun-2021
Date of Decision	28-Aug-2021
Date of Acceptance	07-Sep-2021
Date of Web Publication	16-Nov-2021

Correspondence Address:
Dr. Chia-Yu Chu
Department of Dermatology, National Taiwan University Hospital and National Taiwan University College of Medicine, No. 7, Chung-Shan South Road, Taipei 10002
Taiwan

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/ds.ds_37_21

Abstract

Background: Plasmacytoid dendritic cells (pDCs) are implicated in the reactivation of human herpesvirus 6 (HHV-6) in drug reaction with eosinophilia and systemic symptoms (DRESS). Objectives: We aimed at identifying the alterations of circulating pDCs, basophils, and myeloid dendritic cells (mDCs) during the active stage of drug eruptions and the relationship between pDC alteration and HHV-6 reactivation in DRESS patients. Methods: Nine DRESS patients and twelve patients of other drug eruptions (ODE, including five cases of maculopapular eruptions (MPEs), five cases of Stevens–Johnson syndrome, and two cases of fixed drug eruption) were prospectively recruited. Blood samples were collected weekly for 2 consecutive weeks for flow cytometry of pDC, basophil, mDC, and detection of HHV-6 reactivation by serum anti-HHV-6 immunoglobulin G, or whole blood HHV-6 viral DNA by quantitative polymerase chain reaction. Results: The HHV-6 reactivation was exclusively seen in DRESS, and the pDC levels decreased in DRESS compared with ODE (P = 0.002) and (MPE, P = 0.021), with the timing of such decrease preceding HHV-6 reactivation. Conclusion: The decrease of pDC levels is significantly more common in DRESS patients and occurs before HHV-6 reactivation. The chronological sequence implies that the decrease of pDC may be crucial in the pathogenesis of viral reactivation.

Keywords: Basophil, drug reaction with eosinophilia and systemic symptoms, human herpesvirus 6, myeloid dendritic cell, plasmacytoid dendritic cell

How to cite this article:
Hsu SH, Yang CW, Hsieh YC, Chen KL, Cho YT, Liau JY, Chu CY. Plasmacytoid dendritic cells diminution in peripheral blood is prevalent in drug reaction with eosinophilia and systemic symptoms and may precede human herpesvirus 6 reactivation. Dermatol Sin 2021;39:175-81

How to cite this URL:
Hsu SH, Yang CW, Hsieh YC, Chen KL, Cho YT, Liau JY, Chu CY. Plasmacytoid dendritic cells diminution in peripheral blood is prevalent in drug reaction with eosinophilia and systemic symptoms and may precede human herpesvirus 6 reactivation. Dermatol Sin [serial online] 2021 [cited 2023 Mar 27];39:175-81. Available from: https://www.dermsinica.org/text.asp?2021/39/4/175/330496

Introduction

Drug reaction with eosinophilia and systemic symptoms (DRESS) is one of the drug-induced severe cutaneous adverse reactions (SCARs) characterized by maculopapular eruption (MPE), fever, lymphadenopathies, peripheral eosinophilia, atypical lymphocytosis, and internal organ involvements in which the DNA of human herpesvirus 6 (HHV-6) has been detected in skin and blood specimens.^{[1],[2],[3],[4]}

Plasmacytoid dendritic cells (pDCs) participate in a variety of immune activities,^[5] and their instantaneous production of type I interferons following the activation of toll-like receptors 7 and 9 implies their role in anti-viral activity. pDCs are also reported to be involved in the pathogenesis of some of the autoimmune diseases, such as systemic lupus erythematosus and Sjögren disease, due to their interferon-α-producing capability, which may be a possible explanation for the autoimmune diseases observed among the long-term sequelae of DRESS.^[6] In a previous study regarding DRESS, we found significantly reduced levels of several crucial pro-inflammatory cytokines, namely, interleukin (IL)-1β, IL-2, IL-6, interferon-γ, and tumor necrosis factor-α, within 10 days after the onset of skin rash, a phenomenon possibly related to the decreased amount of circulating pDCs. This timing happens to be before or during the reactivation of HHV-6.^[7]

An almost identical clinical condition termed drug-induced hypersensitivity syndrome (DIHS) would also present with prolonged, extensive maculopapular rash, lymphadenopathy, fever, atypical lymphocytosis, eosinophilia, and hepatitis.^[8] However, DIHS is proposed to be an immune reaction of virus-specific and nonspecific T cells to herpesvirus reactivation, most notably HHV-6, hence its incorporation into the diagnostic criteria of DIHS,^[9] while the diagnosis of DRESS does not require viral reactivation.^[10] Decreased peripheral blood pDCs (lineage⁻ CD123⁺) and increased skin pDCs in DIHS have previously been documented in the literature, specifically around the time window of the reactivation of HHV-6,^[11],[12] and all of the cases reported in one study of DIHS had HHV-6 reactivation.^[11] Therefore, the question of whether the decrease in circulating pDCs is a prevalent phenomenon in DRESS or only occurs in conjunction with HHV-6 reactivation, as in DIHS, is an intriguing one.

In addition, the presence of peripheral eosinophilia in DRESS implies the importance of Th2-type immune responses in the pathogenesis of DRESS. The significant elevation of thymus and activation-regulated chemokine (TARC/CCL17), one of the ligands for the chemokine receptor 4 (CCR4) on type 2 helper T cells, found in previous studies further supports this hypothesis.^[13],[14] The source of TARC was inconclusive, but immunohistochemical staining of lesional skin samples in DIHS demonstrated TARC expression in CD11c⁺ dendritic cells (myeloid dendritic cells [mDCs]) in the dermis.^[13] Basophils have also been reported to participate in Th2-type immune responses based on their production of thymic stromal lymphopoietin, IL-4,^[15] and TARC in a mouse model.^[16] However, the alterations in circulating basophils and mDCs in DRESS have not been addressed in previous studies.

This study aimed to verify whether decreased levels of blood pDCs are more prevalent in DRESS than in other types of drug eruptions and to examine the relationships between HHV-6 reactivation and the alterations of blood pDCs, basophils, and mDCs. The absolute quantity and ratio of pDCs in the lesional skin of DRESS and other phenotypes of drug eruption were also demonstrated.

Materials and Methods

Study population

Patients hospitalized or treated in the dermatologic outpatient clinic at National Taiwan University Hospital from July 2014 to October 2017 who received a diagnosis of any type of drug eruptions (DRESS, Stevens–Johnson syndrome [SJS], fixed drug eruption [FDE], or MPE) were recruited, although patients who presented in the late disease phase (i.e., more than 4 weeks after the onset of the drug eruption) were excluded according to the previously reported timing of pDC level alteration and HHV-6 reactivation.^{[7],[11],[12]} This study has been approved by the Research Ethics Committee of National Taiwan University Hospital (201008030R) and conformed to Declaration of Helsinki. All participants gave their informed consent prior to their inclusion in the study.

Clinical evaluation

Clinical profiles of the patients, including age, gender, culprit medication, the index date for the drug eruption (the date of the onset of skin rash for DRESS, FDE, and MPE, or the date of the first blister or erosion for SJS), and the date of initiation and daily dosages of systemic corticosteroid until the date of discontinuation, were collected. The scoring by the RegiSCAR diagnostic criteria for DRESS was calculated by the presence of fever ≥38.5°C, lymphadenopathies, peripheral eosinophilia, atypical lymphocytosis, suggestive skin rash for DRESS, internal organ involvement, prolonged disease course ≥15 days, and lack of other potential causes for the skin rash, as previously described [Supplementary Table 1].^[10]

Routine blood tests were conducted on the day of diagnosis and weekly thereafter, including tests for hemogram and leukocyte differential counts; hepatobiliary enzyme levels and renal function; mycoplasma, herpes simplex virus-1/2, Epstein–Barr virus, cytomegalovirus, hepatitis B virus, and hepatitis C virus serology, as well as autoimmune screening with anti-nuclear antibody, and ferritin levels.

Detection of human herpesvirus-6 reactivation

Whole blood was sampled on the day of diagnosis and weekly thereafter for 1–2 weeks according to the duration of active disease. Peripheral blood mononuclear cells (PBMCs) were purified, and the serum was separated immediately after blood sampling. DNA was extracted from the whole blood samples using a QIAamp DNA Blood mini-kit (Qiagen Inc., Valencia, CA, USA). The extracted samples were stored at-80°C and then subjected to quantitative polymerase chain reaction (qPCR) for the detection of HHV-6 DNA using the LightCycler® 480 System (Roche Applied Science, Indianapolis, IN, USA). The serum samples from the patients were stored at-80°C until use. Levels of anti-HHV-6 immunoglobulin G (IgG) were measured by enzyme-linked immunosorbent assays (ELISA) (Advanced Biotechnologies, Columbia, MD, USA) as described before.^[7],[17] HHV-6 reactivation was defined as either the detection of viral DNA in whole blood samples with >20 copies per microgram of DNA or a more than four-fold elevation in the serial anti-HHV-6 IgG serum levels compared with the first ELISA result.

Flow cytometry

The PBMC samples were stained and prepared with a 4-color assay for the differentiation of basophils (lin1⁻, HLA-DR⁻, CD123⁺), pDCs (lin1⁻, HLA-DR⁺, CD123⁺), and mDCs (lin1⁻, HLA-DR⁺, CD11c⁺) as previously described (BD Biosciences, San Jose, CA, USA).^[18] Flow cytometry was performed using a BD FACSVerse, and the results were analyzed with the BD FACSuite software. The proportions of pDCs, basophils, and mDCs to PBMCs were determined using the gating protocol shown in [Figure 1]a. When the proportion of the cell population of interest increased in consecutive samplings, the alteration of the cell population was designated as “increase,” whereas a diminishing proportion was labeled as “decrease.” If the proportion was 0% throughout the sampling period, the cell population was also categorized as “decrease.”

Figure 1: Flow cytometric analysis of pDCs, mDCs, basophils and the relationship of decreased pDCs with elevated HHV-6 IgG levels. (a) Demonstration of gating strategy for calculating the frequencies of pDCs (green rectangle), mDCs (brown rectangle) and basophil (blue rectangle). (b) The alterations of pDC ratio in different disease groups. Significant decreasing trend of pDC is shown in DRESS versus ODE or MPE [Table 1 and Supplementary Table 3]. (c) The pDC frequency (closed items) and HHV-6 IgG ELISA (open items) or DNA positivity (labeled as asterisks) against disease course in cases 13, 15, and 20. A decrease in pDC frequencies and correlated escalation of HHV-6 IgG level are shown. APC: Labeling CD11c, FITC: Labeling T- and B-cell linages (Linage Cocktail-1 including CD3, CD14, CD16, CD19, CD20, and CD56), FSC: Forward scatter, lin 1 dim, Lineage Cocktail 1 diminished, PE: Labeling CD123, PED: Prednisolone-equivalent dose; PerCP: Labeling HLA-DR, SSC: Side scatter, pDCs: Plasmacytoid dendritic cells, mDCs: myeloid dendritic cells, HHV-6: Human herpesvirus 6, IgG: Immunoglobulin G, DRESS: Drug reaction with eosinophilia and systemic symptoms, ODE: Other drug eruptions, MPEs: Maculopapular eruptions.

Click here to view

Histology assessment

Histological specimens were stained with conventional hematoxylin and eosin stains. Immunohistochemical staining of the specimens from a selection of patients whose histological samples were still available (DRESS [n = 5], MPE [n = 2], SJS [n = 5]) with CD123 (BD Biosciences, Franklin Lakes, NJ, USA; clone 7G3; dilution 1:600) was performed to highlight the pDCs as previously described.^[19] The number of CD123⁺ cells were labeled and calculated with the digital images from optical microscopy at four representative high-power fields (HPF) of each specimen through ImageJ software (National Institutes of Health, United States).

Statistical analysis

Statistical analyses were performed using software (IBM SPSS Statistics Subscription, IBM Corp, Armonk, NY, USA). Fisher's exact test was used for dichotomous categorical data analysis, while the two-sample Student's t-test was used for continuous variables.

Results

Clinical characteristics of the patients with different drug eruptions

Twenty-one patients were included in this study, including nine with DRESS and twelve with other drug eruptions (ODEs), which consisted of five cases of MPE, five cases of SJS, and two cases of FDE [Table 1] and [Supplementary Table 1]. The day of initiation and the total dosage of systemic corticosteroids showed no difference between the groups, while a longer duration of treatment (39.6 ± 35.0 days vs. 11.4 ± 9.0 days, P = 0.046) was shown in the DRESS patients compared with the MPE patients.

Table 1: Demographic data, human herpesvirus 6 reactivation, and alterations of plasmacytoid dendritic cells, basophils and myeloid dendritic cells in relation to different drug eruption phenotypes

Click here to view

Although not statistically significant among different types of drug eruptions, HHV-6 reactivation was exclusively seen in the DRESS patients [Table 1]. Three out of nine (33%) DRESS patients in this study showed viral reactivation [Supplementary Table 2], a rate which was compatible with reports in the previous literature.^[3]

The ratio of pDC decrease was significantly higher during follow-up in the DRESS patients compared with the ODE patients (78% vs. 9%, P = 0.002) and with the MPE patients (0%, P = 0.021 [Table 1], [Supplementary Table 3] and [Figure 1]b]). In contrast, the basophil and mDC levels did not differ between the different types of drug eruptions. Although the increases were not statistically significant, all five of the SJS patients demonstrated increased mDCs.

Clinical characteristics of the patients with increased or decreased pDCs

The corticosteroid dosage during the 1^st week and the accumulated dosage on the timing of pDC alterations did not show significant differences between the pDC increase and decrease groups [Table 2], and a subpopulation analysis of the pDC increase and decrease groups among the DRESS patients still showed no significant differences. Further analysis showed significant correlations between the decrease of pDCs and the presence of atypical lymphocytes (P = 0.019) and HHV-6 reactivation (P = 0.042), while the quantity of eosinophils was not associated with the decreased pDC levels.

Table 2: Correlation of the alteration of plasmacytoid dendritic cell frequencies with dosage of systemic corticosteroid, the presence of atypical lymphocytes, eosinophilia and human herpesvirus 6 reactivation

Click here to view

Temporal relationship of plasmacytoid dendritic cell decrease and human herpesvirus-6 reactivation

When examining the timing of the decrease in pDCs versus the HHV-6 reactivation, we found a consistent tendency for the decrease in pDCs to occur concurrently with or even before the HHV-6 reactivation (decrease in pDCs vs. HHV-6 reactivation: Day 10 vs. day 76 [Case 13], and day 5 vs. day 21 [Case 15] by serology; day 16 vs. day 16 [Case 20] by qPCR, [Figure 1c]). The determined time window for HHV-6 reactivation thus mostly concurred with previous reports.^{[7],[11],[12]}

The number and ratio of plasmacytoid dendritic cells in the skin

We calculated the mean of the absolute cell count or the ratio of CD123⁺ cells to inflammatory mononuclear cells per HPF under optical microscopy [Supplementary Figure 1]. No significant correlation of the number or ratio of the tissue CD123⁺ cells to the phenotypes of drug eruption, HHV-6 reactivation, accumulated dosage of systemic corticosteroids, or circulating pDC levels [Supplementary Table 4].

Discussion

Although currently considered to be the same presentation as one of the drug-induced SCARs, the diagnostic criteria for DIHS and DRESS have nuanced differences,^[8],[10] most notably is the emphasis on HHV-6 reactivation in the diagnosis of DIHS.^[1],[2],[3] Sugita et al. hypothesized that the paucity of circulating pDCs might lead to decreased antiviral responses and, consequently, to HHV-6 reactivation.^[11] However, whether the reactivation of HHV-6 occurs before or after the reduction in the quantity of circulating pDCs was still disputable due to the difficulty in determining the exact timing of the viral reactivation. Meanwhile, the relationship between DRESS and HHV-6 reactivation during the disease course had been demonstrated in the literature. We thus aimed to verify that the decrease of pDCs is also common in DRESS and attempted to establish the temporal relationship between pDC alteration and HHV-6 reactivation.

In this study, the decrease in pDCs was significantly more prevalent in the DRESS patients, even though only one-third of these patients had HHV-6 reactivation. Reversely, while all the patients with HHV-6 reactivation showed decreased pDC levels, decreased pDCs also occurred in patients without viral reactivation. These results implied that decreased pDC level is an important necessary condition instead of a sufficient condition for HHV-6 reactivation. This implication was further supported by the fact that the decreases of pDCs in our patients occurred before or concurrently with HHV-6 reactivation.

The decrease of pDCs significantly correlated to the presence of atypical lymphocytosis during the disease course. The presence of atypical lymphocytosis was shown to be mostly early in the disease course and to precede the decrease of pDCs. This finding concurred with a previous report, which showed that the atypical lymphocytosis was induced by expansion of HHV-6-specific CD8⁺ T cells and nonantigen-specific CD4⁺ and CD8⁺ T cells.^[20]

Considering that pDCs were labeled as lineage⁻ CD123⁺ in previous report by Sugita et al., which might also incorporate basophils (HLA-DR⁻, CD123⁺), we applied an additional staining of HLA-DR to differentiate these two cell populations. The alteration of circulating basophils did not show significant differences among the different drug eruption groups, suggesting that the lineage⁻ CD123⁺ cells might still mostly consist of pDCs. Thus, despite their ability to produce TARC and the fact that they may, therefore, be an important participant in Th2-type immune responses, basophils may not play a pivotal role in the pathogenesis of DRESS.

The CD11c⁺ mDCs have been implicated in the processes of many other inflammatory dermatoses, such as psoriasis and atopic dermatitis,^[21] and in modulating regulatory T cells. Despite the lack of an obvious tendency of alteration in mDCs among DRESS patients, the increases in mDCs were more prevalent in our SJS patients. The influx of IL-10-producing CD4⁺ Foxp3⁺ regulatory T cells to the lesions of FDE had been proposed to be a possible mechanism for the containment of local inflammation and keratinocyte apoptosis.^[22],[23] In one transgenic mouse model of toxic epidermal necrolysis (TEN), Azukizawa et al. demonstrated that the presence of CD11c⁺ dendritic cells was required in collaboration with CD4⁺ CD25⁺ regulatory T cells to mitigate the disease phenotype.^[24] The prevalence of mDCs in our SJS patients might reflect their relatively short disease duration and milder phenotype compared with TEN patients. However, a comparison with TEN patients in terms of the amount of mDCs would be required to support this observation.

The accumulated dosage of systemic corticosteroids was not higher in the group of patients with decreased circulating pDCs, which implied that the immunosuppressive effect from corticosteroids might not be the primary factor leading to the alteration of pDCs. Although both are common features in DRESS, we did not find a significant correlation between decreased pDCs and peripheral eosinophilia, suggesting that these may not be directly interactive events in the disease.

Unlike the previous report by Sugita et al. showing an inverse correlation of the ratio and number of pDC in the blood versus those in the skin, our data among the DRESS patients did not reveal similar results. The time of skin biopsy along the disease course and the accumulated systemic corticosteroid dosage could possibly be some of the influencing factors. Our analysis of a relatively small population somehow failed to demonstrate the correlation. However, low ratio of pDC in the skin specimen was shown to correlate with HHV-6 reactivation.

There were several limitations in this study. First, our cohort included a relatively small number of patients, which might have diminished the power of the statistical analysis or have led to certain biases. More patients with each of the drug eruption phenotypes will be required for better interpretation. Second, the blood examinations were arranged weekly due to ethical concerns and clinical practicality, which might have rendered the detection of alterations in pDCs/basophils/mDCs and viral reactivation less sensitive and instantaneous. Finally, further qualitative and quantitative analyses of the labeled pDCs were beyond the scope of this study.

Conclusion

This study demonstrated that a decrease in pDCs is more common in DRESS than in other types of drug eruptions and that such decrease in pDCs is significantly related to, and may possibly precede, the reactivation of HHV-6. Meanwhile, the alterations in basophils and mDCs were not significantly related to any of the different types of drug eruptions. Further research is mandatory to elaborate the pathophysiological mechanism of decreased pDC levels and their influence on HHV-6 reactivation.

Financial support and sponsorship

This study was supported by the National Taiwan University Hospital (NTUH103-S2366, 106-S3535, 107-S3845) and National Taiwan University Hospital, Hsin-Chu Branch (HCH104-050).

Conflicts of interest

Prof. Chia-Yu Chu, an editorial board member at Dermatologica Sinica, had no role in the peer review process of or decision to publish this article. The other authors declared no conflicts of interest in writing this paper.

Supplementary Material

References

1.	Shiohara T, Ushigome Y, Kano Y, Takahashi R. Crucial role of viral reactivation in the development of severe drug eruptions: A comprehensive review. Clin Rev Allergy Immunol 2015;49:192-202.
2.	Chen YC, Cho YT, Chang CY, Chu CY. Drug reaction with eosinophilia and systemic symptoms: A drug-induced hypersensitivity syndrome with variable clinical features. Dermatologica Sinica 2013;31:196-204.
3.	Yang CW, Cho YT, Chu CY. The role of viral reactivation in drug reaction with eosinophilia and systemic symptoms and other cutaneous adverse drug reactions (CADRS). Curr Dermatol Rep 2016;5:5-11.
4.	Hsu SH, Yang CW, Hsieh YC, Cho YT, Liau JY, Liang CW, et al. Fever, eosinophilia, and abnormal liver function are early signs suggestive of DRESS: A comparative study between DRESS and MPE. Dermatologica Sinica 2018;36:25-9.
5.	Karrich JJ, Jachimowski LC, Uittenbogaart CH, Blom B. The plasmacytoid dendritic cell as the Swiss army knife of the immune system: Molecular regulation of its multifaceted functions. J Immunol 2014;193:5772-8.
6.	Chen YC, Chang CY, Cho YT, Chiu HC, Chu CY. Long-term sequelae of drug reaction with eosinophilia and systemic symptoms: A retrospective cohort study from Taiwan. J Am Acad Dermatol 2013;68:459-65.
7.	Chen YC, Chiang HH, Cho YT, Chang CY, Chen KL, Yang CW, et al. Human herpes virus reactivations and dynamic cytokine profiles in patients with cutaneous adverse drug reactions – A prospective comparative study. Allergy 2015;70:568-75.
8.	Shiohara T, Iijima M, Ikezawa Z, Hashimoto K. The diagnosis of a DRESS syndrome has been sufficiently established on the basis of typical clinical features and viral reactivations. Br J Dermatol 2007;156:1083-4.
9.	Shiohara T, Inaoka M, Kano Y. Drug-induced hypersensitivity syndrome (DIHS): A reaction induced by a complex interplay among herpesviruses and antiviral and antidrug immune responses. Allergol Int 2006;55:1-8.
10.	Kardaun SH, Sidoroff A, Valeyrie-Allanore L, Halevy S, Davidovici BB, Mockenhaupt M, et al. Variability in the clinical pattern of cutaneous side-effects of drugs with systemic symptoms: Does a DRESS syndrome really exist? Br J Dermatol 2007;156:609-11.
11.	Sugita K, Tohyama M, Watanabe H, Otsuka A, Nakajima S, Iijima M, et al. Fluctuation of blood and skin plasmacytoid dendritic cells in drug-induced hypersensitivity syndrome. J Allergy Clin Immunol 2010;126:408-10.
12.	Ishida T, Kano Y, Mizukawa Y, Shiohara T. The dynamics of herpesvirus reactivations during and after severe drug eruptions: Their relation to the clinical phenotype and therapeutic outcome. Allergy 2014;69:798-805.
13.	Ogawa K, Morito H, Hasegawa A, Daikoku N, Miyagawa F, Okazaki A, et al. Identification of thymus and activation-regulated chemokine (TARC/CCL17) as a potential marker for early indication of disease and prediction of disease activity in drug-induced hypersensitivity syndrome (DIHS)/drug rash with eosinophilia and systemic symptoms (DRESS). J Dermatol Sci 2013;69:38-43.
14.	Komatsu-Fujii T, Kaneko S, Chinuki Y, Suyama Y, Ohta M, Niihara H, et al. Serum TARC levels are strongly correlated with blood eosinophil count in patients with drug eruptions. Allergol Int 2017;66:116-22.
15.	Chirumbolo S. State-of-the-art review about basophil research in immunology and allergy: Is the time right to treat these cells with the respect they deserve? Blood Transfus 2012;10:148-64.
16.	Watanabe M, Satoh T, Yamamoto Y, Kanai Y, Karasuyama H, Yokozeki H. Overproduction of IgE induces macrophage-derived chemokine (CCL22) secretion from basophils. J Immunol 2008;181:5653-9.
17.	Yang CW, Cho YT, Hsieh YC, Hsu SH, Chen KL, Chu CY. The interferon-γ-induced protein 10/CXCR3 axis is associated with human herpesvirus-6 reactivation and the development of sequelae in drug reaction with eosinophilia and systemic symptoms. Br J Dermatol 2020;183:909-19.
18.	Willmann K, Dunne JF. A flow cytometric immune function assay for human peripheral blood dendritic cells. J Leukoc Biol 2000;67:536-44.
19.	Liau JY, Chuang SS, Chu CY, Ku WH, Tsai JH, Shih TF. The presence of clusters of plasmacytoid dendritic cells is a helpful feature for differentiating lupus panniculitis from subcutaneous panniculitis-like T-cell lymphoma. Histopathology 2013;62:1057-66.
20.	Suzuki Y, Inagi R, Aono T, Yamanishi K, Shiohara T. Human herpesvirus 6 infection as a risk factor for the development of severe drug-induced hypersensitivity syndrome. Arch Dermatol 1998;134:1108-12.
21.	Zaba LC, Krueger JG, Lowes MA. Resident and “inflammatory” dendritic cells in human skin. J Invest Dermatol 2009;129:302-8.
22.	Teraki Y, Shiohara T. IFN-gamma-producing effector CD8⁺T cells and IL-10-producing regulatory CD4+ T cells in fixed drug eruption. J Allergy Clin Immunol 2003;112:609-15.
23.	Mizukawa Y, Yamazaki Y, Shiohara T. In vivo dynamics of intraepidermal CD8+ T cells and CD4+ T cells during the evolution of fixed drug eruption. Br J Dermatol 2008;158:1230-8.
24.	Azukizawa H, Sano S, Kosaka H, Sumikawa Y, Itami S. Prevention of toxic epidermal necrolysis by regulatory T cells. Eur J Immunol 2005;35:1722-30.