Abstract
Background:
Reflectance confocal microscopy (RCM) allows noninvasive, real-time evaluation of the skin at resolution akin to histopathology (HP), but its application in cutaneous graft-versus-host disease (GvHD) has not been extensively assessed.
Objective:
We describe RCM features of cutaneous GvHD including acute (aGvHD), late acute, chronic (cGvHD; sclerotic and nonsclerotic subtypes), and inactive GvHD and correlate RCM with same-site HP for a subset of patients.
Study Design:
Thirty-two adult and pediatric allogeneic hematopoietic cell transplant (allo-HCT) recipients with cutaneous GvHD received RCM imaging of ≥1 lesions (n=44), 13 of which required skin biopsy. RCM images were de-identified and assessed by 2 RCM experts blinded to clinical and HP findings to reach consensus on features and patterns of inflammatory dermatoses. Major RCM features (present in ≥65% of lesional sites) and patterns were reported. To determine correlation between RCM and HP, detection of cellular features and patterns of inflammatory dermatoses were compared using percent agreement and prevalence adjusted bias adjusted kappa estimates (PABAK).
Results:
Seven early and late aGvHD patients (7 lesions) had irregular honeycombing, spongiosis, dermoepidermal (DEJ) and dermal inflammation, and melanophages; early aGvHD also had hyperkeratosis, dilated vessels, and coarse connective tissue. Both had an interface dermatitis pattern. Eighteen nonsclerotic cGvHD patients (24 lesions) had irregular honeycombing, spongiosis, DEJ and dermal inflammation, dilated vessels, coarse connective tissue and interface and spongiotic dermatitis patterns. Three sclerotic cGvHD patients (7 lesions) had irregular honeycombing, DEJ and dermal inflammation, with an interface dermatitis pattern. Four inactive GvHD patients (6 lesions) showed minimal inflammation. RCM and HP had similar detection for 6 of 13 features and overall patterns important in diagnosis for 2 late aGvHD (2 lesions; 15%) and 10 nonsclerotic cGvHD patients (11 lesions; 85%) requiring skin biopsy.
Conclusion:
RCM can detect features commonly reported in cutaneous GvHD and is comparable to HP. Additional characterization of cutaneous GvHD on RCM may enable future use to diagnose, monitor, or predict disease in real time.
Keywords: GVHD, skin, imaging, Allogeneic HCT
INTRODUCTION
Graft-versus-host disease (GvHD) represents a leading cause of morbidity and mortality after allogeneic hematopoietic cell transplantation (allo-HCT).4–6 GvHD can lead to a high symptomatic burden, decline in functional status and poor quality of life.7,8 Cutaneous GvHD occurs frequently and is the most common presenting sign, highlighting the importance of early dermatologic diagnosis and intervention to mitigate disease progression.9 Acute GvHD (aGvHD) typically presents with a maculopapular rash whereas chronic GvHD (cGvHD) presents with poikiloderma, lichen planus-like, lichen sclerosus-like, morphea-like, and sclerotic features.10 As diverse presentations can mimic other conditions, detection of histopathologic features of aGvHD and cGvHD on skin biopsy is often utilized for diagnostic clarification.11–17
While skin biopsy can confirm cutaneous cGvHD or detect alternative diagnoses, the ability to make a definitive diagnosis or alter management for cutaneous aGvHD remains controversial.18–20 In aGvHD, early skin biopsy following allo-HCT can be nonspecific, show findings typical of other post allo-HCT dermatoses (e.g., drug eruption, viral exanthems, toxic erythema of chemotherapy), and often lack correlation with clinical rash severity.13,15,17,21,22 Moreover, the clinical severity of rash rather than histopathology (HP) often prompts initiation of treatment for suspected GvHD.21 For cGvHD, skin biopsy is encouraged and may be required in the absence of diagnostic findings, but does not provide diagnosis in real time. Skin biopsies may further result in false-negative findings when taken too early in disease onset due to lack of significant cellular change or if focal processes are not sampled. False-positive findings are also possible due to concurrent inflammatory dermatoses such as eczema or drug eruptions.13 Furthermore, the value of serial biopsies during treatment of cGvHD is unknown, and conducting multiple biopsies on various skin sites is impractical.13 Skin biopsies carry risks, such as infection, bleeding, or scarring, and should be reconsidered if the results will not alter clinical management.23,24
Reflectance confocal microscopy (RCM) permits noninvasive, in vivo visualization of the skin at cellular resolution akin to HP. Greyscale images are en face, or in the horizontal plane, and can assess the epidermis to the papillary dermis extending up to 250 μm in depth.25 This modality has proven successful in bedside detection or diagnosis of melanoma and keratinocyte carcinomas.25–28 In addition, RCM can identify characteristic features and patterns of inflammatory dermatoses, including discoid lupus erythematosus, lichen planus, plaque psoriasis, as well as seborrheic, irritant and allergic contact dermatitis.27,29–36 To date, RCM features of 5 aGvHD lesions have been reported and correlate to HP.35,37 We thus investigated the presence of 14 RCM features and major patterns unique to cutaneous aGvHD, late aGvHD, cGvHD (both sclerotic and nonsclerotic subtypes) and inactive GvHD. Agreement between same-site RCM images and HP are reported for a subset of patients.
METHODS
This prospective study took place at a single tertiary academic hospital, Memorial Sloan Kettering Cancer Center, New York, NY and was Institutional Review Board-approved. We accrued 32 adult and pediatric allo-HCT recipients presenting to the dermatology service with a rash from November 2016 to March 2019 who provided written informed consent. Twelve patients (13 lesions) required a skin biopsy for clinical care. Lesional sites for RCM imaging and biopsy were identified by a dermatologist with expertise in cutaneous GvHD. Diagnoses were made based on clinical findings or clinicopathologic correlation when skin biopsies were conducted. The International Bone Marrow Transplant Registry (IBMTR) classification guided the aGvHD grading, except grades A–D were labeled grades 1–4.38 Late aGvHD included those with aGvHD features after day 100 not meeting diagnostic criteria for cGvHD.10 A diagnosis of cGvHD was based on the National Institutes of Health (NIH) Consensus Conference diagnostic criteria and organ-specific severity scoring.10 Presence of maculopapular rash or erythema, lichen-planus like, papulosquamous, ichthyosis, or keratosis pilaris-like, or sclerotic features were assigned a score of 0 (0% BSA), score 1 (1–18% BSA), score 2 (19–50% BSA), or score 3 (> 50% BSA). Additional sclerotic features were assessed to diagnose sclerotic cGvHD.10 These were used to assign a global score of cGvHD (mild, moderate, or severe). Inactive GvHD included those with hypo or hyperpigmented skin changes in the absence of erythema or sclerosis.
Reflectance Confocal Microscopy
Two RCM devices – the VivaScope 1500 and 3000 (Caliber Imaging and Diagnostics, Rochester, NY) – were utilized to image the skin. The wide-probe, arm-mounted VivaScope 1500 (Figure 1) was used to image lesions on flat areas of the body (e.g., trunk and extremities). Following an established clinical imaging protocol, mosaics, or 4 × 4 mm large field-of-view images, were taken at the stratum corneum and spinosum, dermoepidermal junction (DEJ), and papillary dermis (about 200 μm depth). To estimate epidermal thickness, we acquired 5 stacks (500 × 500 μm small field-of-view images) at consecutive depths, 3.0 μm apart, starting from the stratum corneum to the papillary dermis in each quadrant and at the center of the imaged lesion. A minimum of two 30 second videos of dermal blood vessels were captured.
Figure 1. Reflectance Confocal Microscopy.
(a) The VivaScope 1500 device is attached to the skin with adhesive an adhesive window. The laser light source penetrates the skin and is reflected based on refractory indices of cellular structures mainly from melanin. High resolution images are visualized in real time on the attached monitor. (b) RCM takes horizontal plane images at each skin layer from the epidermis to the papillary dermis up to a depth of 250 μm. (c) Greyscale images of each skin layer capture cellular features. The device can take stacks of about 500 × 500 μm images from the epidermis to the papillary dermis. (d) The device can also capture blocks, or mosaics, by collating multiple images into large-scale, 8 × 8 mm images.
The handheld VivaScope 3000, similarly to an ultrasound probe, moves freely over skin, and was used to image lesions on hard-to-reach or curved areas (e.g., face or axilla) or when device attachment was not possible, such as an active pediatric patient. Unlike the VivaScope 1500 that can capture mosaics and stacks, the VivaScope 3000 captures stacks with a larger surface area of 750 × 750 μm and video mosaics. Protocol required 2 stacks in each quadrant and 2 in the center of each lesion from the stratum corneum to papillary dermis at depths 3.0 μm apart. Video mosaics were taken at each skin layer, as well as two 30 second videos of vessels. Patients had same-day RCM imaging for 1 or more lesions using VivaScope 1500 (n = 25; 57%), 3000 (n = 17, 39%) or both (n = 2; 4%), depending on lesion location by an investigator with extensive RCM imaging experience.
Under a retrospective protocol, RCM images were de-identified and assessed by 2 RCM experts blinded to clinical and HP findings to reach consensus on predetermined RCM features of inflammatory dermatoses and incidental features defined in Table 2.27 Established RCM patterns of inflammatory dermatoses helpful in diagnosis (interface, spongiotic, and psoriasiform dermatitis) were assigned (Table 3).27 The smallest punch biopsy used for HP assessment of inflammatory dermatoses is 2 mm, which provides a minimal tissue sample for adequate histopathologic processing. Using this standard, RCM images capturing a surface area of under 2 mm2 were excluded from data analysis.
Table 2.
Cellular Features of GvHD Assessed on Histopathology and Reflectance Confocal Microscopy.
| Cellular Feature | Histopathologic Definition39,48 | Appearance on RCM27,49 | Images |
|---|---|---|---|
| Epidermal Features | |||
| Epidermal acanthosis or atrophy | Thickening (acanthosis) or thinning (atrophy) of the epidermis (mostly stratum spinosum and granulosum) compared to what is expected for that anatomic location. | Increased or decreased thickness of the epidermis assessed by measuring vertical depth of the nucleated cell layers to up to the DEJ on stack analysis. | NA |
| Parakeratosis | Retention of the nuclei in the corneocytes in the stratum corneum (can be focal or diffuse). | Polygonal, hyper-reflective cells in the stratum corneum with possible identification of a small, dark nuclei (green arrow). Can be focal or diffuse. |
|
| Hyperkeratosis | Thickening of the stratum corneum compared to what is expected at that anatomic location. | Bright, hyper-reflective areas of the stratum corneum in the horizontal plane (black arrow), or assessed in the vertical plane on stack analysis as a thickness of > 20 – 40 μm from the stratum corneum to the first nucleated cell layer in the epidermis. |
|
| Irregular honeycombing | NA – no histopathologic correlate. | Variation in thickness, size, and shape of the bright outlines of cells in the stratum spinosum and granulosum create an abnormal pattern. Represents epidermal disarray or atypia. |
|
| Spongiosis (intercellular widening/edema) | Fluid accumulation or edema between cells causing widening of the cell-cell junctions with notable bridges on light microscopy. Can have severe spongiosis with vesicle formation.39 May also note exocytosis or the presence of lymphocytes in the epidermis. | Intercellular widening between keratinocytes usually with infiltrative, single or clusters of bright cells (exocytosis) in the stratum spinosum. Spongiotic vesicle formation is noted as a dark, round space with reflective inflammatory cells in the center. (yellow circle). |
|
| Dendritic Cells | May correlate to either Langerhans cells or melanocytes. Langerhans cells are mildly eosinophilic, irregularly shaped cells in the stratum spinosum while melanocytes are clear cells mostly found at the basal cell layer. | Detected as bright cells with stellate-like extensions (white arrow). Represent either Langerhans cells or melanocytes which cannot be differentiated on RCM. |
|
| Dyskeratotic/necrotic cells | Noted by early keratinization with cytoplasm that is highly eosinophilic and a small, basophilic nuclei. Term use to encompass necrotic, apoptotic, and dyskeratotic (prematurely keratinized) cells. | Greyish to bright, medium-sized cells that appear to be separate from the surrounding epidermis. May be circular or irregularly shaped (red circle). |
|
| Satellitosis | Lymphocytes surrounding necrotic or apoptotic keratinocyte. | Bright, small lymphocytes (blue arrow) surrounding a dyskeratotic/necrotic cell (blue circle). |
|
| Dermoepidermal Junction Features | |||
| Complete or partial DEJ obscuration | Disruption of normal DEJ architecture by inflammatory cell infiltrate (can be focal or diffuse). | Loss of regular DEJ pattern (orange arrow). Appears as a partial or complete blurring of the DEJ (yellow square) from clusters of bright inflammatory cells (green arrow). |
|
| DEJ Inflammation (focal or diffuse) | Lymphocytes at the level of the DEJ obscuring normal architecture. May be vacuolar with clear spaces, or lichenoid with diffuse, band-like lymphocytic infiltration along the DEJ.27 | Includes a diffuse (lichenoid) and abundant infiltrate of bright inflammatory cells at the DEJ (seen in image on right), or focal infiltrate with discrete clusters or single, bright inflammatory cells at the DEJ. |
|
| Perifollicular/adnexal inflammation | Inflammatory cells surrounding hair follicles or adnexal structures. | Bright inflammatory cells (yellow arrow) surrounding adnexal structures (hair follicle marked by yellow asterisk). |
|
| Sub-basalar Vacuolization | Small round, clear spaces lining the DEJ that can range in size.39 | Noted by non-reflective dark, round spaces at the level of the DEJ (red square). |
|
| Dermal Features | |||
| Dilated Vessels | Increased diameter of the microvasculature within the dermis. | Distinct dark tubular structures that may be round or linear (red arrow) found in dermis. Prominent leukocyte trafficking can be visualized on live imaging. |
|
| Dermal Inflammation | Aggregates or solitary inflammatory cells in the dermis. Able to identify lymphocytes and melanophages as well as specific inflammatory cells such as eosinophils, neutrophils, plasma, or mast cells. | Presence of small bright lymphocytes (green arrows) and plump bright melanophages (pink arrow) in the dermis. More specific inflammatory cells are difficult to discern. |
|
| Melanophages | Macrophages with phagocytosed melanin appearing as large, brown cell in the dermis, sometimes dendritic. | Presences of plump, bright, predominantly non-nucleated inflammatory cells in the dermis (blue arrow). |
|
| Coarse Connective Tissue | Widened and bundled collagen fibrils or degenerated elastic fiber/collagen tissue. | Thick, bright fibrillar or bundled structures in the dermis. Diffuse coarse connective tissue noted in the image on the right. |
|
DEJ indicates dermoepidermal junction. RCM image sizes are 1 × 1 mm for DEJ inflammation and 0.75 × 0.75 mm for sub-basalar vacuolization and perifollicular/adnexal inflammation. Remaining images are 0.5 × 0.5 mm.
Table 3.
Major and Minor Criteria for Diagnosis of Inflammatory Dermatoses on Reflectance Confocal Microscopy
| Dermatitis | Major Criteria | Minor Criteria |
|---|---|---|
| Interface Dermatitis |
|
|
| Spongiotic Dermatitis |
|
|
| Psoriasiform Dermatitis |
|
|
DEJ indicates dermoepidermal junction. Adapted from Ardigò et al.27 with permission from Elsevier.
Histopathology
Following RCM imaging, same-day and same-site 3 to 4 mm punch biopsies were taken and submitted in formalin for routine histopathological processing (Figure 2). HP specimens were assessed by a dermatopathologist blinded to clinical and RCM findings for 13 cellular features (Table 5) and assigned patterns. In cases where a feature could not be assessed (e.g., perifollicular/adnexal involvement in a HP sample without adnexal structures present, or collagen on RCM that could not reach the dermis), the sample was excluded from analysis for that specific feature.
Figure 2. Histopathologic Processing.
(a) A surgical punch, or circular blade, is used to penetrate the skin and is twisted downward to extract a 3 to 4 mm punch biopsy sample that includes the epidermis to the subcutaneous tissue. (b, c) Tissue is vertically sectioned into tissue samples that are 5 μm in depth. (d) Thin tissue sections are laid on glass slides for staining with H&E and histopathologic evaluation in the vertical plane.
Table 5.
Comparison of Reflectance Confocal Microscopy and Histopathology Features and Patterns
| Cellular Feature | Percent Agreement, % | Prevalence Adjusted Bias Adjusted Kappa |
|---|---|---|
| Epidermal | ||
| Epidermal acanthosis or atrophy | 61.5 | 0.23 |
| Parakeratosis | 66.7 | 0.33 |
| Hyperkeratosis | 69.2 | 0.39 |
| Spongiosis (intercellular widening/edema) | 76.9 | 0.54* |
| Dendritic cells | 53.9 | 0.08 |
| Dyskeratotic/necrotic cells | 76.9 | 0.54* |
| Dermoepidermal Junction | ||
| Complete/partial DEJ obscuration | 84.6 | 0.69† |
| DEJ inflammation (focal or diffuse) | 76.9 | 0.54* |
| Perifollicular/adnexal inflammation | 75.0 | 0.50* |
| Dermal | ||
| Dilated vessels | 69.2 | 0.38 |
| Dermal inflammation | 53.9 | 0.08 |
| Melanophages | 84.6 | 0.69† |
| Coarse connective tissue | 45.5 | −0.09 |
| Cellular Pattern | ||
| Pattern | 76.9 | 0.54* |
Represents features with moderate agreement of feature detection between RCM and HP
Represents features with substantial agreement of feature detection between RCM and HP
DEJ indicates dermoepidermal junction
Data Analyses
For the analysis, lesional sites from the same individual were considered independent. To determine major RCM features of subtypes of cutaneous GvHD, each imaged lesion was coded as having the absence or presence of 14 RCM features (Table 4). Imaged lesional sites were grouped by diagnoses: aGvHD, late aGvHD, nonsclerotic cGvHD, sclerotic cGvHD, inactive aGvHD and cGvHD (both sclerotic and nonsclerotic types). The prevalence of 14 RCM features was calculated, and those present in ≥65% of lesions within a diagnostic GvHD group were defined as major features for that subtype of cutaneous GvHD. RCM patterns (interface, spongiotic, or psoriatic dermatitis) within diagnostic categories were reported. To determine correlation between RCM and HP feature detection, 13 cellular features of inflammatory dermatoses (Table 5) were identified, and patterns assigned on RCM and HP. RCM and HP findings were compared using percent agreement and prevalence adjusted bias adjusted kappa estimates (PABAK). PABAK values are interpreted like Cohen’s kappa estimates, with ≤ 0 indicating no agreement, 0.01 – 0.20 as none to slight, 0.21 – 0.40 as fair, 0.41 – 0.60 as moderate, 0.61 – 0.80 as substantial, and 0.81 – 1.00 as almost perfect agreement.
Table 4.
Presence of Reflectance Confocal Microscopy Features in Subtypes of Cutaneous GvHD
| RCM Features | Nonsclerotic cGvHD (n=24) | Sclerotic cGvHD (n=7) | Inactive cGvHD (n=3) | Acute GvHD (n=2) | Late Acute GvHD (n=5) | Inactive Acute GvHD (n=3) |
|---|---|---|---|---|---|---|
| Epidermal, n (%) | ||||||
| Epidermal acanthosis or atrophy | 2/23 (9) | 0/7 (0) | 1/3 (33) | 0/2 (0) | 0/5 (0) | 0/3 (0) |
| Parakeratosis | 5/24 (21) | 1/7 (14) | 0/3 (0) | 0/2 (0) | 0/5 (0) | 0/3 (0) |
| Hyperkeratosis | 15/24 (63) | 1/7 (14) | 1/3 (33) | 2/2 (100)* | 3/5 (60) | 2/3 (67)* |
| Irregular honeycombing | 19/24 (79)* | 5/7 (71)* | 0/3 (0) | 2/2 (100)* | 4/5 (80)* | 2/3 (33) |
| Spongiosis (exocytosis or vesicles) | 23/24 (96)* | 3/7 (43) | 2/3 (67)* | 2/2 (100)* | 4/5 (80)* | 2/3 (67)* |
| Dendritic cells | 6/24 (25) | 0/7 (0) | 0/3 (0) | 1/2 (50) | 3/5 (60) | 0/3 (0) |
| Dyskeratotic/necrotic cells | 12/24 (50) | 0/7 (0) | 0/3 (0) | 0/2 (0) | 0/5 (0) | 0/3 (0) |
| Dermoepidermal Junction, n (%) | ||||||
| Complete/partial DEJ obscuration | 9/24 (38) | 1/6 (17) | 1/3 (33) | 0/2 (0) | 2/5 (40) | 2/3 (33) |
| DEJ inflammation (focal or diffuse) | 16/24 (67)* | 6/7 (86)* | 1/3 (33) | 2/2 (100)* | 5/5 (100)* | 0/3 (0) |
| Perifollicular/adnexal inflammation | 9/24 (38) | 0/7 (0) | 0/3 (0) | 1/2 (50) | 1/5 (20) | 0/3 (0) |
| Dermal, n (%) | ||||||
| Dilated vessels | 16/23 (70)* | 3/7 (43) | 1/3 (33) | 2/2 (100)* | 2/5 (40) | 2/3 (67)* |
| Dermal Inflammation | 18/23 (78)* | 5/7 (71)* | 0/3 (0) | 2/2 (100)* | 4/5 (80)* | 2/3 (33) |
| Melanophages | 14/23 (61) | 1/7 (14) | 0/3 (0) | 2/2 (100)* | 4/5 (80)* | 2/3 (33) |
| Coarse connective tissue | 13/19 (68)* | 3/7 (43) | 1/3 (33) | 2/2 (100)* | 2/5 (40) | 2/3 (33) |
Indicates a major RCM feature.
DEJ indicated dermoepidermal junction; GvHD, graft-versus-host disease (cGvHD, chronic)
Imaged sites included for each feature assessment varied and is reflected by the denominator for each feature.
Images were excluded for specific feature assessment when RCM was unable to visualize necessary layers of skin.
For example, when RCM was unable to reach the dermis for collagen assessment, or if the DEJ could not be visualized due to a diffuse inflammatory infiltrate.
Data Sharing Statement
For original data, please contact markovaa@mskcc.org.
RESULTS
Study Population
Table 1 summarizes patients’ characteristics. We evaluated 32 adult (n = 25, 78.1%) and pediatric (n = 7, 21.9%) allo-HCT recipients. Patients were median age 47 years (range 8 months – 73 years) and were 53.1% male and 68.8% white. Diagnoses included acute myeloid leukemia and myelodysplastic syndrome (n = 17, 53.1%), non-Hodgkin lymphoma and chronic lymphocytic leukemia (n = 9, 28.2%), chronic myeloid leukemia (n = 2, 6.2%), and non-malignant hematologic disorders (n = 4, 12.5%). Graft types included peripheral blood (n = 19, 59.4%), bone marrow (n = 8, 25.0%), and cord blood (n = 5, 15.6%). Donors were mostly unrelated (n = 20, 62.5%) and 10/10 HLA-matched (n = 2, 65.6%). Most underwent myeloablative (n = 20, 62.5%) conditioning and GvHD prophylaxis was predominantly calcineurin inhibitor based. Transplants occurred from 2011 to 2018 with 2 patients transplanted in 1990 and 2007 diagnosed with inactive cGvHD.
Table 1.
Patient Demographics and Transplant Characteristics
| Characteristic | Value |
|---|---|
| Age, years | |
| Median (range) | 47 (0.7 – 73) |
| Sex, n (%) | |
| Male | 17 (53.1) |
| Female | 15 (46.9) |
| Race, n (%) | |
| White | 22 (68.8) |
| African American | 6 (18.8) |
| Asian | 2 (6.2) |
| Other | 2 (6.2) |
| Diagnosis, n (%) | |
| AML/MDS | 17 (53.1) |
| NHL/CLL | 9 (28.2) |
| CML | 2 (6.2) |
| Non-malignant hematologic disorders* | 4 (12.5) |
| Graft Type, n (%) | |
| Peripheral Blood | 19 (59.4) |
| Bone Marrow | 8 (25.0) |
| Cord Blood | 5 (15.6) |
| Donor type | |
| Related | 12 (37.5) |
| Unrelated† | 20 (62.5) |
| Donor-recipient HLA match, n (%) | |
| 10/10 | 21 (65.6) |
| 9–8/10 | 5 (15.6) |
| ≤ 7/10 | 6 (18.8) |
| Conditioning Intensity, n (%) | |
| Myeloablative | 20 (62.5) |
| Reduced-intensity | 10 (31.3) |
| Non-myeloablative | 2 (6.2) |
| Year of Transplant, n (%) | |
| 2017 – 2018 | 19 (59.4) |
| 2015 – 2016 | 10 (31.3) |
| < 2012 | 3 (9.3) |
| GvHD Prophylaxis, n (%) | |
| T-cell depletion | 6 (18.8) |
| CNI/MTX +/− sirolimus | 10 (31.3) |
| CNI/MMF | 10 (31.3) |
| PTCy/CNI/MMF | 3 (9.3) |
| CNI/MMF/MTX | 1 (3.1) |
| CNI alone | 1 (3.1) |
| Alemtuzumab | 1 (3.1) |
| Cutaneous GvHD Diagnosis, n (%) | |
| Early aGvHD | 2 (6.2) |
| Late aGvHD | 5 (15.6) |
| Nonsclerotic cGvHD | 18 (56.4) |
| Sclerotic cGvHD | 3 (9.4) |
| Inactive aGvHD | 2 (6.2) |
| Inactive cGvHD | 2 (6.2) |
| Systemic GvHD Therapy at Presentation, n (%) | |
| CNI | 10 (31.3) |
| Corticosteroid | 2 (6.2) |
| Corticosteroid/CNI | 8 (25.1) |
| Corticosteroid/Ibrutinib +/− CNI | 2 (6.2) |
| Corticosteroid/Ruxolitinib | 2 (6.2) |
| Corticosteroid/Imatinib | 1 (3.1) |
| None | 7 (21.9) |
AML indicates acute myeloid leukemia; CNI, calcineurin inhibitor; CLL, chronic lymphocytic leukemia; CML, chronic myeloid leukemia; GvHD, graft-versus-host disease (aGvHD, acute; cGvHD, chronic); HLA, human leukocyte antigen; MMF, mycophenolate mofetil; MTX, methotrexate; MDS, myelodysplastic syndrome; NHL, non-Hodgkin lymphoma; and PTCy post-transplantation cyclophosphamide.
Includes immune deficiencies, aplastic anemia, and Fanconi anemia
Includes cord blood transplants
Cutaneous GvHD diagnoses included early aGvHD (n = 2, 6.2%), late aGvHD (n = 5, 15.6%), nonsclerotic cGvHD (n = 18, 56.4%), sclerotic cGvHD (n = 3, 9.4%), inactive aGvHD (n = 2, 6.2%), and inactive cGvHD (n = 2, 6.2%). Due to history of other organ involvement, 78.1% were on systemic GvHD therapy. Median time of rash presentation post allo-HCT was 40.5 (range 26 – 55) days for early aGvHD, 214 (141 – 329) days for late aGvHD, 417.5 days (72 days – 5.3 years) for nonsclerotic cGvHD, and 2.3 (1.5 – 2.6) years for sclerotic cGvHD. RCM images were taken a median of 152 (152 – 391) days post allo-HCT for inactive aGvHD and 19.7 (11.6 – 27.8) years for inactive cGvHD. Most patients had 1 skin site imaged (n = 23, 71.8%), while 9 (28.2%) had more ≥2. A total of 44 RCM images were collected from the head/neck (n = 4, 9.1%), trunk (n = 16, 36.4%), and upper (n = 17, 38.6%) and lower extremities (n = 7, 15.9%).
Cutaneous Acute GvHD
Two patients with cutaneous early aGvHD had 2 lesional sites imaged on RCM. At assessment, both patients were skin stage 2, grade 2 aGvHD characterized by a maculopapular rash in one patient and erythema and edema in the other. Major RCM features in 2/2 (100%) lesions included hyperkeratosis, irregular honeycombing, spongiosis with exocytosis, DEJ inflammation (1 with focal and 1 with diffuse infiltrate), dermal inflammation, melanophages, dilated vessels, and coarse connective tissue (Figure 3; Table 4). Both lesions (100%) had an interface dermatitis pattern (1 focal and 1 lichenoid pattern) with mild spongiosis.
Figure 3. Reflectance confocal microscopy images of active cutaneous acute GvHD.
(a-d) Acute GvHD: (a) Image at the stratum corneum showing diffuse bright, hyper-reflective hyperkeratosis. (b) Image at the stratum spinosum showing irregular honeycombing. (c) Image at the DEJ showing diffuse DEJ inflammation and the presence of abundant plump, bright melanophages (blue arrow) and loss of DEJ architecture. (d) Image at the dermis showing dilated vessels (red arrow). (e-h) Late acute GvHD: (e) Image at the stratum corneum showing diffuse hyper-reflective hyperkeratosis. (f) Image showing DEJ obscuration from DEJ inflammation (white arrow) (g) Image at the dermis showing dilated vessels (green arrow) and bright dermal inflammatory cells (yellow arrow). (h) Image at the dermis showing coarse connective tissue (orange arrow). All images are 0.5 × 0.5 mm.
Five patients with cutaneous late aGvHD had 5 RCM imaged lesional sites. Three (60.0%) patients had skin stage 3, grade 3 aGvHD, while 2 (40.0%) had skin stage 1, grade 1 aGvHD. All patients presented with an eczematous rash with red, erythematous scaly papules, macules, and/or patches. Major RCM features included 5/5 (100%) with focal DEJ inflammation and 4/5 (80%) with irregular honeycombing, spongiosis with exocytosis (1 had exocytosis and vesicles), dermal inflammation and melanophages (Figure 3; Table 4). One lesion had noted sub-basalar vacuolization. Patterns included 3 (60%) lesions with focal and 1 (20%) with lichenoid interface dermatitis, and 1 (20%) with mixed focal interface and spongiotic dermatitis.
Cutaneous Chronic GvHD
Eighteen patients with cutaneous nonsclerotic cGvHD had 24 imaged lesional sites. NIH global scores included 7 (38.9%) patients with mild, 8 (44.4%) with moderate, and 3 (16.7%) with severe cGvHD, with skin scores of 1 (n = 9, 50.0%), 2 (n = 7, 38.9%), and 3 (n = 2, 11.1%). Twelve (50%) lesions were clinically diagnosed with eczematous cGvHD with red to pink, erythematous scaly patches, plaques, and papules. Ten (42%) had lichen planus-like cGvHD presenting with scattered, scaly pink to violaceous dusky papules and plaques. Two (8%) had poikilodermatous cGvHD presenting as atrophic epidermis with hypo or hyperpigmented brown to pink reticular macules with telangiectasia.
Major RCM features included spongiosis with exocytosis (23/24, 96%; 6/23 with exocytosis and vesicles), irregular honeycombing (19/24, 79%), DEJ inflammation (16/24, 67%; 8/16 with focal, 5/16 with diffuse infiltrate, and 3/16 with both), dermal inflammation (18/23, 78%), dilated vessels (16/23, 70%), and coarse connective tissue (13/19, 68%; Table 4; Figure 4). Patterns included 5 (21%) with focal and 9 (38%) with lichenoid interface dermatitis, 7 (29%) with spongiotic dermatitis, and 3 (12%) with a mixed spongiotic and interface dermatitis pattern. Of the lesions diagnosed as lichen planus-like cGvHD, 7/10 (70%) had a lichenoid interface dermatitis pattern. In lesions diagnosed as eczematous cGvHD, 9/12 (75%) had a spongiotic or mixed spongiotic and interface dermatitis pattern.
Figure 4. Reflectance confocal microscopy images of active cutaneous chronic GvHD.
(a–d) Nonsclerotic cGvHD: (a) Image at the stratum spinosum showing a dyskeratotic bright cell surrounded by dark halo (blue circle). (b) Image at the stratum spinosum/DEJ shows DEJ and perifollicular inflammation (yellow arrow) at a hair follicle (yellow asterisk), dark holes of sub-basalar vacuolization and satellitosis (red circle and inset). (c) Image at the DEJ showing melanophages (blue arrow) and diffuse DEJ inflammation. (d) Image at the DEJ showing diffuse inflammation causing partial DEJ obscuration (orange arrow). (e–h) Sclerotic cGvHD: (e) Image at the stratum spinosum showing irregular honeycombing. (f) Image at the DEJ showing diffuse inflammation causing partial obscuration of the DEJ with distinct melanophages (red arrows). (g) Image of the papillary dermis showing inflammatory cells (green arrows). (h) Imaging in the dermis showing coarse connective tissue. Image (b) is 1mm × 0.75 mm; remaining images are 0.5 × 0.5 mm.
Three patients with sclerotic cGvHD had 7 lesional sites imaged. All 3 had severe cGvHD with skin scores of 3, presenting with morphea-like thick, pink to hyperpigmented, bound-down plaques with skin rippling. Major RCM features included focal DEJ inflammation (6/7, 86%; 1 also with diffuse infiltrate), dermal inflammation (5/7, 71%) and irregular honeycombing (5/7, 71%; Table 4; Figure 4). Patterns included 6 lesions (86%) with a focal (n = 5) or lichenoid (n = 1) interface dermatitis and 1 (14%) nonspecific, mild spongiotic dermatitis.
Inactive Acute and Chronic GvHD
There were 2 patients with inactive aGvHD with 3 imaged lesional sites. Both patients had skin stage 0, and 1 had grade 2 aGvHD based on visceral involvement. Clinically they had normal appearing skin or hyperpigmented patches. Hyperkeratosis, spongiosis with exocytosis, and dilated vessels were found in 2/3 lesions (67%; Table 4). There was minimally detected immunologic activity with only 1/3 (33%) having dermal inflammation or melanophages (Figure 6). All patterns were nonspecific, but 2 lesions (67%) had mild spongiotic dermatitis.
Figure 6. Reflectance confocal microscopy images of inactive cutaneous GvHD.
(a, d) Images at the epidermis showing regular honeycombing. (b, e) Images at the DEJ showing no signs of inflammation and intact DEJ with ringed pattern (e). (c, f) Images showing a bland normal dermis without any inflammatory cells nor melanophages. All images are 0.5 × 0.5 mm.
Two patients had 3 imaged lesional sites with inactive sclerotic (1 site) and nonsclerotic (2 sites) cGvHD. One patient had severe and 1 had moderate cGvHD based on other organ involvement. Disease was quiescent in the skin as both had scores of 0, with hyperpigmented, atrophic patches. The major RCM feature in 2/3 (67%) lesions was spongiosis with exocytosis (Table 4). There was no additional evidence of active inflammation (Figure 6). Patterns included 1 lesion (33%) with focal interface dermatitis, 1 (33%) with nonspecific, mild spongiotic dermatitis and 1 (33%) identified as normal skin.
Reflectance Confocal Microscopy Comparison to Histopathology
Twelve patients had clinically indicated skin biopsies (1 patient had 2 lesional sites biopsied), providing 13 lesional biopsies for comparison to same-day and same-site RCM images. Clinicopathological diagnoses included nonsclerotic GvHD (11 lesions; 85%), and late aGvHD (2 lesions; 15%). Detection of 6 of 13 cellular features showed moderate to substantial agreement between RCM and HP (Table 5). Three of 6 epidermal features were similarly detected on RCM and HP. This included moderate agreement (76.9% agreement, kappa (κ) = 0.54) for detection of spongiosis and dyskeratotic or necrotic cells (Figure 5). Of note, satellitosis was assessed for all samples on HP and was incidentally observed in 3 lesions on RCM with 100% agreement to HP (Figure 5). All DEJ features had moderate or substantial agreement, including DEJ inflammation (76.9%, κ = 0.54), DEJ obscuration (84.6%, κ = 0.69), and perifollicular/adnexal inflammation (75.0%, κ = 0.50). Sub-basalar vacuolization was assessed for all samples on HP and incidentally observed in 5/13 samples on RCM with moderate agreement (80.0%, κ = 0.60; Figure 4). Of the 4 features assessed at the dermis, detection of melanophages had substantial 84.6% agreement (κ = 0.69; Figure 5). Patterns on HP and RCM were in moderate agreement (76.9%, κ = 0.54), with most having an interface dermatitis pattern (HP: 62%; RCM: 69%). Spongiotic dermatitis (HP: 15%; RCM: 23%) and mixed or fibrotic patterns (HP: 23%; RCM: 8%) were also reported.
Figure 5. Comparative features of nonsclerotic GvHD on reflectance confocal microscopy (left) and histopathology (right).
(a) RCM; (b) HP: at the stratum spinosum level shows irregular honeycombing and grey dyskeratotic cells also detected on HP (yellow circles and insets). (c) RCM; (d) HP: at the DEJ shows DEJ and periadnexal inflammation (yellow arrows), dark holes of sub-basalar vacuolization and satellitosis (red circles and inset) confirmed on HP. (e) RCM; (f) HP: at the DEJ/papillary dermis shows diffuse DEJ inflammation appearing as abundant bright cells which corresponds to inflammatory infiltrate of lymphocytes and melanophages on HP. (g) RCM; (h) HP: at the papillary dermis shows abundant, plump bright melanophages which appear as brown cells on HP (blue arrows). RCM image (a) 0.75 × 0.5 mm; (c) 1mm × 0.75 mm; (e) 1.5 mm × 1.25 mm; (g) 0.5 × 0.5 mm. Histopathology is 20x magnification.
DISCUSSION
To date, this is the largest analysis of RCM for assessment of cutaneous GvHD. We determined that RCM was able to identify accepted HP features of cutaneous aGvHD and cGvHD and detect signs of active inflammation.
In the aGvHD and late aGvHD groups, major RCM features of DEJ inflammation, spongiosis with exocytosis, irregular honeycombing, and dermal inflammation aligned with established HP findings of cutaneous aGvHD, and were previously observed on RCM in 5 patients with aGvHD.35,39,40 We additionally detected subacute/chronic inflammatory features including melanophages, hyperkeratosis, and coarse connective tissue. While melanophages and superficial fibrosis are less commonly reported in cutaneous aGvHD, they were previously noted on HP in 13% of aGvHD and late aGvHD lesions.14 Notable dilated blood vessels were also a major feature in the aGvHD group. Dilated blood vessels are commonly detected on RCM in various inflammatory dermatoses, and are a minor criterion for assigning patterns of spongiotic, psoriatic, and interface dermatitis.29 Caruntu and Boda demonstrated that after topical application of capsaicin used to induce cutaneous inflammation, RCM detected significantly larger, dilated capillaries compared to the same anatomic sites in the non-capsaicin controls.41 Therefore, dilated vessels may be indicative of active inflammation. Most aGvHD and late aGvHD lesions had a focal interface dermatitis pattern on RCM, mirroring HP findings of less dense inflammation in aGvHD, compared to a diffuse, lichenoid inflammatory pattern common in nonsclerotic cGvHD.12 In summary, RCM was able to detect commonly reported HP features of aGvHD and signs of active inflammation in both the aGvHD and late aGvHD groups. Further, we identified dilated blood vessels as a novel finding associated with aGvHD.
Major RCM features in cutaneous nonsclerotic cGvHD such as DEJ inflammation, dermal inflammation, and coarse connective tissue reflected accepted HP features of cGvHD.13 Major RCM features of irregular honeycombing, spongiosis with exocytosis, and dilated vessels reflect nonspecific findings common in various inflammatory dermatoses and are associated with disease activity.29 Dyskeratotic cells and perifollicular/adnexal inflammation were more commonly identified in the nonsclerotic cGvHD than the aGvHD group. This may be due to the larger sample size, but is possibly attributed to the 39% of patients who were not on systemic immunosuppressive therapy who may have had more active disease when imaged.13
Patterns detected on RCM highly correlated with final nonsclerotic cGvHD diagnoses. Most lesions with lichen planus-like clinical morphology were found to have a pattern of lichenoid interface dermatitis and those with eczematous cGvHD had patterns of spongiotic dermatitis or a mixed pattern of interface and spongiotic dermatitis on RCM. Both lichenoid interface dermatitis and spongiotic dermatitis support a diagnosis of lichen planus-like or eczematous GvHD in the right clinical context, respectively.39,42,43 This suggests that RCM can detect common features of GvHD that allow for specific pattern recognition of diagnostic significance.
In the sclerotic cGVHD group there was focal DEJ inflammation, dermal inflammation, and irregular honeycomb pattern, with most lesions having an interface dermatitis pattern. These findings are not unusual as remaining inflammatory activity and homogenization and sclerosis of dermal collagen is often reported in sclerotic cGvHD.12,13,44 Coarse connective tissue possibly indicative of dermal sclerosis was reported in under half of the lesions. While connective tissue changes were expected to be found in most lesions, sclerosis and fibrosis typically progresses from the papillary to reticular dermis, subcutis, muscle and fascia in a “top-down” fashion.13 Therefore, connective tissue changes deeper than the superficial reticular dermis reflective of sclerotic cGvHD may not have been visualized on RCM due its depth limitation. While this may limit diagnosis of sclerotic cGvHD on RCM, the ability to identify inflammation may help detect disease activity in clinically diagnosed sclerotic cGvHD to assist with treatment or management decisions.
Inactive aGvHD and cGvHD presented with expected limited DEJ or dermal inflammation and were diagnosed with patterns that were nonspecific and even normal. This distinctively bland appearance on RCM, compared to the marked inflammation present in the active cutaneous GvHD groups, may help determine disease activity. This alludes to the possibility of RCM to act as a cutaneous biomarker with the ability to detect inflammation in real time to guide treatment discissions. For example, in deciding to initiate GvHD treatment, patients with signs of active inflammation may have enhanced response to certain therapeutic agents.45 If confirmed in a larger sample, RCM may hold the potential to monitor disease activity for treatment initiation or cessation of various topical or systemic therapies.
For the subset of nonsclerotic GvHD lesions that underwent biopsy, features detected on RCM were confirmed on HP as measured by moderate or substantial agreement between the two modalities. This included dyskeratotic cells, satellitosis, DEJ inflammation, sub-basalar vacuolization and perifollicular/adnexal inflammation, all which are used to support a diagnosis of cutaneous GvHD on HP.39Additionally, nonsclerotic cGvHD and late aGvHD often have an interface dermatitis pattern, which was the predominant pattern detected on both RCM and HP.39 Our findings parallel literature reporting RCM to be comparable to HP in other inflammatory dermatoses.34 31 32,35,37 Specifically, for 11 patients with cutaneous aGvHD, there was over 88% correlation between RCM and HP for detection of basal vacuolar change, dyskeratotic cells, and dermal inflammation.37 Seven cellular features (epidermal acanthosis or atrophy, hyperkeratosis, parakeratosis, dilated vessels, dendritic cells, dermal inflammation, and coarse connective tissue) had lower agreement between RCM and HP, likely due to inherent difference between the two modalities. Specifically, HP visualizes smaller tissue samples in the vertical plane with H&E staining, while RCM captures much larger area, in the horizontal plane in greyscale. Additionally, HP can detect features from the top layers of the epidermis to the subcutis, while RCM is unable to visualize structures past the deep dermis. Ultimately, these features are nonspecific HP features of GvHD. Dermal inflammation was more readily detected by RCM than HP with a resulting weaker agreement between modalities. This is likely because on RCM, dermal inflammation more easily visualized as distinct, bright cells in large surface area images. Had additional sections of HP been assessed, these features may have had higher agreement. Remaining features with weaker agreement have been detected similarly on RCM and HP in other inflammatory dermatoses, and would have likely shown higher agreement in a larger sample population.31,32,34,35 Overall, the ability of RCM to detect cellular features and patterns of late aGvHD and nonsclerotic cGvHD was comparable to HP. Thus, real-time RCM imaging may be preferred in select cases when biopsy is not feasible such as facial lesions, patients with delayed wound healing, young patients, or when multiple skin sites require biopsies.
Limitations included depth of RCM imaging, hindering observations of cellular changes in deep dermal connective tissue important in sclerotic cGvHD. Furthermore, in cases of epidermal acanthosis, hyperkeratosis, or active inflammation, the thick or crowded epidermis may scatter light and alter resolution of deeper dermal structures on RCM.34 RCM is in greyscale dependent on the refractive index of various tissue structures, mainly from melanin. In this mostly light skin population lacking pigmentation, subtle microstructures may have been missed.25 Additionally, vacuolization and satellitosis were incidentally noted on RCM, and systematic assessment would increase sensitivity to detect these features. Lastly, diagnostic groups had small sample sizes, and the nonsclerotic cGvHD group contained mostly eczematous cGvHD, an uncommon subtype, thus limiting generalizability. Our findings are novel and warrant further investigation in a larger study.
Our study demonstrated that in recipients of allo-HCT, RCM was able to both detect established features and patterns of cutaneous acute and chronic GvHD and was proven comparable to available HP. RCM was further able to detect inflammation in active GvHD lesions. RCM holds the potential to act as a real-time, noninvasive biomarker for the diagnosis and possible prognostication of GvHD and can assist in monitoring disease activity both longitudinally and at various sites when multiple biopsies are not feasible.
HIGHLIGHTS.
Reflectance confocal microscopy (RCM) can noninvasively visualize skin
RCM can detect cellular features of cutaneous graft-versus-host disease (GvHD)
RCM allows identification of signs of inflammation and disease in real time
RCM is comparable to histopathology in detecting features of cutaneous GvHD
Acknowledgements:
From Memorial Sloan Kettering Cancer Center, New York, NY, we thank Allan Halpern, MD (Dermatology Service) and Milind Rajadhyaksha, PhD (Dermatology Service) for their invaluable feedback on this project. They were not compensated. This work was supported in part by the National Cancer Institute of the National Institutes of Health under Award Number P30-CA008748 made to Memorial Sloan-Kettering Cancer Center. This commentary contains a discussion of the investigative use of reflectance confocal microscopy in the clinical assessment of graft-versus-host disease. The contents of this submission are the views of the individual authors and do not reflect the opinions of the author affiliations.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Reingold RE, Monnier J, Ardigò M, et al. Reflectance Confocal Microscopy Correlates with Histopathology in Nonsclerotic GvHD and Non-GvHD Rash in Allogeneic Stem Cell Recipients [abstract 87]. Transplantation & Cellular Therapy Meetings of the American Society for Transplantation and Cellular Therapy and the Center for International Blood & Marrow Transplant Research; 2021 February 11, 2021; Digital Experience. [Google Scholar]
- 2.Reingold RE, Monnier J, Ardigò M, et al. Real-Time Reflectance Confocal Microscopy Features of Cutaneous Graft-Versus-Host Disease [abstract 350]. Transplantation & Cellular Therapy Meetings of the American Society for Transplantation and Cellular Therapy and the Center for International Blood & Marrow Transplant Research; 2021 February 11, 2021; Digital Experience. [Google Scholar]
- 3.Reingold RE, Monnier J, Ardigò M, et al. Real-time Reflectance Confocal Microscopy Features of Cutaneous Chronic Graft-versus-Host Disease [abstract]. 5th International Chronic Graft-versus-Host Disease Symposium Virtual Congress; 2021 April 16, 2021; Zagreb, Croatia [Google Scholar]
- 4.Solh MM, Bashey A, Solomon SR, et al. Long term survival among patients who are disease free at 1-year post allogeneic hematopoietic cell transplantation: a single center analysis of 389 consecutive patients. Bone Marrow Transplant 2018;53:576–83. [DOI] [PubMed] [Google Scholar]
- 5.Wingard JR, Majhail NS, Brazauskas R, et al. Long-term survival and late deaths after allogeneic hematopoietic cell transplantation. J Clin Oncol 2011;29:2230–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Styczyński J, Tridello G, Koster L, et al. Death after hematopoietic stem cell transplantation: changes over calendar year time, infections and associated factors. Bone Marrow Transplantation 2020;55:126–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Lee SJ, Flowers MED. Recognizing and Managing Chronic Graft-Versus-Host Disease. Hematology 2008;2008:134–41. [DOI] [PubMed] [Google Scholar]
- 8.El-Jawahri A, Pidala J, Khera N, et al. Impact of Psychological Distress on Quality of Life, Functional Status, and Survival in Patients with Chronic Graft-versus-Host Disease. Biol Blood Marrow Transplant 2018;24:2285–92. [DOI] [PubMed] [Google Scholar]
- 9.Strong Rodrigues K, Oliveira-Ribeiro C, de Abreu Fiuza Gomes S, Knobler R. Cutaneous Graft-Versus-Host Disease: Diagnosis and Treatment. Am J Clin Dermatol 2018;19:33–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jagasia MH, Greinix HT, Arora M, et al. National Institutes of Health Consensus Development Project on Criteria for Clinical Trials in Chronic Graft-versus-Host Disease: I. The 2014 Diagnosis and Staging Working Group report. Biol Blood Marrow Transplant 2015;21:389–401.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Fischer A, Jakubowski AA, Lacouture ME, et al. Histopathologic Features of Cutaneous Acute Graft-Versus-Host Disease in T-Cell-Depleted Peripheral Blood Stem Cell Transplant Recipients. Am J Dermatopathol 2015;37:523–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hillen U, Häusermann P, Massi D, et al. Consensus on performing skin biopsies, laboratory workup, evaluation of tissue samples and reporting of the results in patients with suspected cutaneous graft-versus-host disease. J Eur Acad Dermatol Venereol 2015;29:948–54. [DOI] [PubMed] [Google Scholar]
- 13.Shulman HM, Cardona DM, Greenson JK, et al. NIH Consensus development project on criteria for clinical trials in chronic graft-versus-host disease: II. The 2014 Pathology Working Group Report. Biol Blood Marrow Transplant 2015;21:589–603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Massi D, Fondi C, Nozzoli C, et al. The impact of histopathologic examination of graft-versus-host disease in the era of reduced-intensity conditioning regimen: a study from the Gruppo Italiano Trapianto di Midollo Osseo. Hum Pathol 2011;42:254–68. [DOI] [PubMed] [Google Scholar]
- 15.Häusermann P, Walter RB, Halter J, et al. Cutaneous graft-versus-host disease: a guide for the dermatologist. Dermatology 2008;216:287–304. [DOI] [PubMed] [Google Scholar]
- 16.Weaver J, Bergfeld WF. Quantitative analysis of eosinophils in acute graft-versus-host disease compared with drug hypersensitivity reactions. Am J Dermatopathol 2010;32:31–4. [DOI] [PubMed] [Google Scholar]
- 17.Kuykendall TD, Smoller BR. Lack of specificity in skin biopsy specimens to assess for acute graft-versus-host disease in initial 3 weeks after bone-marrow transplantation. J Am Acad Dermatol 2003;49:1081–5. [DOI] [PubMed] [Google Scholar]
- 18.Haimes H, Morley KW, Song H, Okhovat JP, Schmidt B, Huang JT. Impact of skin biopsy on the management of acute graft-versus-host disease in a pediatric population. Pediatr Dermatol 2019;36:455–9. [DOI] [PubMed] [Google Scholar]
- 19.Paun O, Phillips T, Fu P, et al. Cutaneous complications in hematopoietic cell transplant recipients: impact of biopsy on patient management. Biol Blood Marrow Transplant 2013;19:1204–9. [DOI] [PubMed] [Google Scholar]
- 20.Kohler S, Hendrickson MR, Chao NJ, Smoller BR. Value of Skin Biopsies in Assessing Prognosis and Progression of Acute Graft-Versus-Host Disease. The American Journal of Surgical Pathology 1997;21:988–96. [DOI] [PubMed] [Google Scholar]
- 21.Zhou Y, Barnett MJ, Rivers JK. Clinical significance of skin biopsies in the diagnosis and management of graft-vs-host disease in early postallogeneic bone marrow transplantation. Arch Dermatol 2000;136:717–21. [DOI] [PubMed] [Google Scholar]
- 22.Firoz BF, Lee SJ, Nghiem P, Qureshi AA. Role of skin biopsy to confirm suspected acute graft-vs-host disease: results of decision analysis. Arch Dermatol 2006;142:175–82. [DOI] [PubMed] [Google Scholar]
- 23.Wahie S, Lawrence CM. Wound Complications Following Diagnostic Skin Biopsies in Dermatology Inpatients. Archives of Dermatology 2007;143:1267–71. [DOI] [PubMed] [Google Scholar]
- 24.Abhishek K, Khunger N. Complications of skin biopsy. J Cutan Aesthet Surg 2015;8:239–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Rajadhyaksha M, Marghoob A, Rossi A, Halpern AC, Nehal KS. Reflectance confocal microscopy of skin in vivo: From bench to bedside. Lasers Surg Med 2017;49:7–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jain M, Pulijal SV, Rajadhyaksha M, Halpern AC, Gonzalez S. Evaluation of Bedside Diagnostic Accuracy, Learning Curve, and Challenges for a Novice Reflectance Confocal Microscopy Reader for Skin Cancer Detection In Vivo. JAMA Dermatology 2018;154:962–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ardigò M, Prow T, Agozzino M, Soyer P, Berardesca E. Reflectance confocal microscopy for inflammatory skin diseases. G Ital Dermatol Venereol 2015;150:565–73. [PubMed] [Google Scholar]
- 28.Guitera P, Menzies SW, Longo C, Cesinaro AM, Scolyer RA, Pellacani G. In vivo confocal microscopy for diagnosis of melanoma and basal cell carcinoma using a two-step method: analysis of 710 consecutive clinically equivocal cases. J Invest Dermatol 2012;132:2386–94. [DOI] [PubMed] [Google Scholar]
- 29.Ardigo M, Longo C, Gonzalez S. Multicentre study on inflammatory skin diseases from The International Confocal Working Group: specific confocal microscopy features and an algorithmic method of diagnosis. Br J Dermatol 2016;175:364–74. [DOI] [PubMed] [Google Scholar]
- 30.Swindells K, Burnett N, Rius-Diaz F, González E, Mihm MC, González S. Reflectance confocal microscopy may differentiate acute allergic and irritant contact dermatitis in vivo. J Am Acad Dermatol 2004;50:220–8. [DOI] [PubMed] [Google Scholar]
- 31.Ardigo M, Cota C, Berardesca E, González S. Concordance between in vivo reflectance confocal microscopy and histology in the evaluation of plaque psoriasis. J Eur Acad Dermatol Venereol 2009;23:660–7. [DOI] [PubMed] [Google Scholar]
- 32.Moscarella E, González S, Agozzino M, et al. Pilot study on reflectance confocal microscopy imaging of lichen planus: a real-time, non-invasive aid for clinical diagnosis. J Eur Acad Dermatol Venereol 2012;26:1258–65. [DOI] [PubMed] [Google Scholar]
- 33.Hicks SP, Swindells KJ, Middelkamp-Hup MA, Sifakis MA, González E, González S. Confocal histopathology of irritant contact dermatitis in vivo and the impact of skin color (black vs white). J Am Acad Dermatol 2003;48:727–34. [DOI] [PubMed] [Google Scholar]
- 34.Ardigò M, Maliszewski I, Cota C, et al. Preliminary evaluation of in vivo reflectance confocal microscopy features of Discoid lupus erythematosus. Br J Dermatol 2007;156:1196–203. [DOI] [PubMed] [Google Scholar]
- 35.Saknite I, Gill M, Alessi-Fox C, et al. Features of cutaneous acute graft-versus-host disease by reflectance confocal microscopy. Br J Dermatol 2019;181:829–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Koller S, Gerger A, Ahlgrimm-Siess V, Weger W, Smolle J, Hofmann-Wellenhof R. In vivo reflectance confocal microscopy of erythematosquamous skin diseases. Exp Dermatol 2009;18:536–40. [DOI] [PubMed] [Google Scholar]
- 37.Saknite I, Gill M, Christi A-F, et al. Key Histopathology Features of Cutaneous Acute Graft-Versus-Host Disease Can be Detected Noninvasively. Blood 2019;134:3278–. [Google Scholar]
- 38.Rowlings PA, Przepiorka D, Klein JP, et al. IBMTR Severity Index for grading acute graft-versus-host disease: retrospective comparison with Glucksberg grade. Br J Haematol 1997;97:855–64. [DOI] [PubMed] [Google Scholar]
- 39.Fung MA. Inflammatory Diseases of the Dermis and Epidermis. In: Busam KJ, ed. Dermatopathology A Volume in the Series: Foundations in Diagnostic Pathology. 2 ed. Philadelphia, PA: Elsevier/Saunders; 2016:11–78. [Google Scholar]
- 40.Elston DM. Interface dermatitis. In: Elston DM, ed. Dermatopathology. 3 ed. Philadelphia, PA: Elsevier; 2019:132–52. [Google Scholar]
- 41.Căruntu C, Boda D. Evaluation through in vivo reflectance confocal microscopy of the cutaneous neurogenic inflammatory reaction induced by capsaicin in human subjects. J Biomed Opt 2012;17:085003. [DOI] [PubMed] [Google Scholar]
- 42.Chan S-L, Wood BA, Mesbah Ardakani N. Eczematous graft-vs-host disease: A report of three cases and review of the literature. Journal of Cutaneous Pathology 2020;47:1085–95. [DOI] [PubMed] [Google Scholar]
- 43.Creamer D, Martyn-Simmons CL, Osborne G, et al. Eczematoid Graft-vs-Host Disease: A Novel Form of Chronic Cutaneous Graft-vs-Host Disease and Its Response to Psoralen–UV-A Therapy. Archives of Dermatology 2007;143:1157–62. [DOI] [PubMed] [Google Scholar]
- 44.Yeung CS, H. M Pathology of Graft vs. Host Disease A Case Based Teaching Guide. Cham, Switzerland: Springer International Publishing; 2019. [Google Scholar]
- 45.Miklos D, Cutler CS, Arora M, et al. Ibrutinib for chronic graft-versus-host disease after failure of prior therapy. Blood 2017;130:2243–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Rajadhyaksha M, González S, Zavislan JM, Anderson RR, Webb RH. In vivo confocal scanning laser microscopy of human skin II: advances in instrumentation and comparison with histology. J Invest Dermatol 1999;113:293–303. [DOI] [PubMed] [Google Scholar]
- 47.Hashemi P, Pulitzer MP, Scope A, Kovalyshyn I, Halpern AC, Marghoob AA. Langerhans cells and melanocytes share similar morphologic features under in vivo reflectance confocal microscopy: a challenge for melanoma diagnosis. J Am Acad Dermatol 2012;66:452–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Ferringer T, Ko CJ. The basics: Diagnostic terms, skin anatomy, and stains. In: Elston DM, ed. Dermatopathology. 3 ed. Philadelphia, PA: Elsevier; 2019:1–35. [Google Scholar]
- 49.Ardigò M, Agozzino M. The Semiology and Patterns of Inflammatory Skin Conditions. In: Hofmann-Wellenhof R, Pellacani G, Malvehy J, Soyer HP, eds. Reflectance Confocal Microscopy for Skin Diseases. Berlin, Heidelberg: Springer Berlin Heidelberg; 2012:349–65. [Google Scholar]