Application of the Diagnostic Evaluation for Alopecia in Traditional Veterinary Species to Laboratory Rhesus Macaques (Macaca mulatta)

Abstract

Alopecia in nonhuman primates in the biomedical research setting is often attributed to compromised psychologic wellbeing. Behavioral causes, mainly hair plucking, have become the unconfirmed and exclusive default diagnosis, and the possibility that alopecia may be secondary to a primary medical or dermatologic disease is often overlooked. Although nonbehavioral causes of alopecia in nonhuman primates are described in the literature, few prospective hypothesis-based studies have investigated medical and behavioral etiologies concurrently. We therefore undertook such a study with the aim of designing a clinical diagnostic guide for approaching cases of nonhuman primate alopecia. Because most cases of alopecia in nonhuman primates in the literature and at our facility are not associated with a definitive diagnosis, the hypothesis we tested was that the well-established diagnostic evaluation for alopecia used for traditional veterinary species is not applicable to nonhuman primates. Discounting differences in histopathology and behavioral assessment, the current study revealed few clinically relevant significant differences between nonhuman primates with and without alopecia. As a result, our hypothesis was confirmed, and we conclude that the standard dermatologic diagnostic plan typically described for alopecia diagnosis in traditional veterinary species and used as the basis for assessment of alopecia in nonhuman primates should be reassessed.

Hair acts as a protective barrier and plays a role in thermoregulation, sensory perception, and sex and species recognition.^12,25 In addition, nonhuman primate infants rely on hair to cling to their mothers.⁶⁹ Because hair performs many essential functions, realistic concerns about alopecia in captive nonhuman primates are justified.

Because its manifestation is readily and consistently identifiable, alopecia in nonhuman primates recently has received an increase in regulatory scrutiny.^8,23 Although numerous causes of alopecia exist, it typically is attributed to compromised psychologic wellbeing in these animals. Behavioral causes, mainly hair plucking, have become the unconfirmed and exclusive default diagnosis.⁶⁶ Hair plucking, which is a poorly understood multifactorial disorder, is defined as the “pulling out of hair from the body using hands or teeth.”^9,47,66,67 However, the possibility that the occurrence of alopecia may be secondary to a primary medical or dermatologic disease is often overlooked.⁵⁵ This oversight is highlighted in a study of cats with a diagnosis of presumptive psychogenic alopecia, an impulse-control disorder comparable to hair plucking in nonhuman primates.⁸³ Reevaluation of the initial diagnosis revealed that a medical condition was the sole cause of alopecia in 76% of the feline cases; 14% of the cats had both a medical and behavior cause; and in only 10% of the cats was a diagnosis of psychogenic alopecia confirmed. If similarly applicable to nonhuman primates, a systematic medical evaluation to rule out medical causes of alopecia is warranted before attributing its occurrence to aberrant behavior.

Published case reports, surveys, and literature reviews attribute alopecia in nonhuman primates to a variety of medical and environmental causes that are congruent with the differential diagnoses for alopecia in traditional veterinary species. Medical causes include nonpathologic processes such as androgenetic alopecia^24,56,81,82 and senescent balding.³⁷ Pathologic causes reported consist of immune-mediated diseases such as alopecia areata^10,26 and atopic dermatitis;^45,48,58 parasitic causes including Demodex⁷⁵ and Sarcoptes spp. mites;^33,54 hormonal abnormalities including hypothyroidism,^46,49,55,57 hyperadrenocorticism,^41,87 and telogen effluvium;³⁷ nutritional causes including vitamin A^50,61,62 and zinc³² toxicity, folate,⁶⁴ zinc,^17,39,79and protein^5,21,89 deficiencies; fungal causes such as dermatophytosis;^6,7,42,59,80 and congenital atrichia with papular lesions.² Environmental causes include seasonal shedding,^25,75,90 heat stress,⁹⁰ and erythema ab igne.⁸⁴

Although many nonbehavioral causes of alopecia in nonhuman primates are described in the literature, few prospective hypothesis-based studies have investigated medical and behavioral etiologies concurrently. We therefore undertook such a study with the aim of designing a clinical diagnostic guide for approaching future cases of alopecia in nonhuman primates. As a starting point, the well-established dermatologic algorithm for evaluation of alopecia in companion animal species, primarily dogs,^51,70 (Figure 1) was used as the basis for the medical evaluation in this study. However, because the majority of nonhuman primate alopecia cases investigated in the literature^36,45,76 and at our facility are not associated with a definitive diagnosis, we tested the hypothesis that this diagnostic algorithm for traditional veterinary species is not applicable in evaluating alopecia in nonhuman primates. We chose rhesus macaques to test the hypothesis because they are one of the most frequently used nonhuman primate species in biomedical research.

Figure 1.

Standard diagnostic plan for evaluation of alopecia in traditional veterinary animal species.

Materials and Methods

Animals.

Subjects included 78 adult rhesus macaques (Macaca mulatta) of Indian and Chinese origin. All animals were reared in large social groups in the outdoor breeding colony at the Tulane National Primate Research Center, which is an AAALAC-accredited facility. Housing and care were provided in accordance with the regulations of the Animal Welfare Act⁴ and recommendations of the Guide for the Care and Use of Laboratory Animals.³⁸ Subjects used for this study were assigned to various infectious disease studies, the majority involving SIV. Animals were housed indoors in stainless steel cages equipped with perches and multiple enrichment devices. Feeding enrichment was provided a minimum of 12 times each month and included fruits, vegetables, and foraging materials. Animals were maintained on a 12:12-h light:dark cycle with ambient temperature of 64 to 72 °F (17.8 to 22.2 °C) and a relative humidity of 30% to 70%. Subjects were fed commercial diet formulated for nonhuman primates (Purina Diet 5037, PMI Feeds, St Louis, MO) twice daily and provided water ad libitum. The study was approved by our facility's IACUC.

Subject selection was restricted to the current singly housed population to avoid the potential confounds of social overgrooming and allo-hair plucking by cagemates. In addition, subjects were chosen from among animals that had been housed indoors for a minimum of 4 mo to ensure acclimation to their housing condition. This period also allowed sufficient time for hair regrowth, ensuring that the hair coat condition did not reflect any transient stress of a move from field-cage breeding groups to indoor-cage housing.^11,25

Degree of alopecia was scored for potential subjects by using the National Primate Research Center Behavioral Management Consortium Alopecia Scoring System.²² To facilitate calculating the percentage of body surface affected by hair loss, the surface area of the body was assessed in 11 approximately equal parts (Figure 2). This scheme is a modification of the Rule of Nines, which was developed to score the area of injury for human burn patients.²² An alopecia score ranging from 0 to 5 was assigned based on the percentage of body surface affected (Figure 3). On the basis of cageside observations of rhesus macaques assigned to research protocols at our facility, 39 animals were assigned to the alopecic group. This group had alopecia affecting at least 25% of the body surface. Another 39 animals with no more than 10% body surface affected by hair loss served as controls, allowing for normal variation in haircoat density. When possible, controls were selected from the same research protocol as alopecic animals. Animals with greater than 10% but less than 25% hair loss were excluded to ensure comparison of 2 clearly defined, discrete populations.

Figure 2.

The National Primate Research Center Behavioral Management Consortium Alopecia Scoring System divides the body into 11 body parts by using a modified Rule of Nines as a guide.²² The tail accounts for 1% of the surface area and is only considered in borderline cases.

Figure 3.

Conversion of the percentage of body surface affected by hair loss into an alopecia score. Figure based on data from reference 22.

Because alopecia can be secondary to systemic disease, animals were eligible regardless of health status. In addition, animals were included regardless of experimental status, including SIV status, to represent the actual population of animals at risk for developing alopecia at biomedical research institutions. At the time of data collection, 16 (41%) animals in each group were infected with one of the following SIV strains: SIVmac239 (4 alopecic; 8 control), SHIV 162 P3 (8 alopecic; 3 control), SIVsm/E660 (4 control), SIVmac251 (4 alopecic), and SIVB670-CL12 (1 control).

The alopecic group consisted of 14 (35.9%) female and 25 male (64.1%) macaques with a mean age of 8.3 ± 3.4 y (3.9 to 16.9 y) and mean weight of 9.9 ± 3.7 kg (4.4 to 21.3 kg). The control group included 12 female (30.8%) and 27 male (69.2%) macaques with a mean age of 7.6 ± 2.4 y (4.8 to 15.8 y) and mean weight of 10.3 ± 3.4 kg (5.8 to 17.3 kg). Data were collected between January and May 2010.

Physical and dermatologic examinations.

Subjects were anesthetized with tiletamine–zolazepam (8 mg/kg IM; Telazol, Fort Dodge Animal Health, Fort Dodge, IA). Physical and dermatologic examinations included body weight, rectal temperature, and body condition scoring. Alopecia degree was calculated by using the National Primate Research Center Behavioral Management Consortium Alopecia Scoring System as previously described.²² The alopecia scoring for each animal was performed by a single person, who showed high interrater agreement with a previously trained person and scoring system developer (linear-weighted κ = 0.83). In addition, the distribution (symmetric or asymmetric) and configuration of the alopecia were noted. Configuration was described as focal (in a single well-defined location), multifocal (in multiple well-defined areas), patchy (indistinct patches with an uneven distribution), diffuse (spread widely in a uniform distribution), or generalized (involving the majority of the body).

Photographs (4 views: left lateral, right lateral, dorsum, and ventrum) were taken of each animal (Lumix FZ28, Panasonic, Secaucus, NJ). In addition, any primary or secondary dermatologic lesions (if present) were photographed.

Hematology.

Samples were collected for CBC and serum biochemistry. In addition, because many of the animals enrolled in this study had been exposed to SIV, peripheral blood CD4⁺ T lymphocyte counts were used to evaluate disease progression. The clinical pathology laboratory at our facility analyzed the CBC (Advia 120, Siemens, Deerfield, IL) and serum biochemistry (Olympus AU400, Beckman Coulter, Brea, CA) panels. The Immunology Core at our facility analyzed the CD4+ T lymphocyte counts by TruCount (BD Biosciences, Palo Alto, CA) technology.

Endocrinology.

Concentrations of basal cortisol and endogenous adrenocorticotrophic hormone (ACTH) were measured to test for hyperadrenocorticism as performed in previous reports.^41,87 Concentrations of total triiodothyronine (T3), free triiodothyronine (FT3), total thyroxine (T4), free thyroxine (FT4), and thyroid stimulating hormone (TSH) were measured to test for hypothyroidism.^46,49,55,57 Because levels of these hormones, especially basal cortisol, can fluctuate throughout the day, samples were taken during the morning (0800 to 1100) in an attempt to address this potential confound. Blood collected for the endocrinology panel was allowed to clot for 20 min at room temperature and then placed in a 4 °C refrigerator until being centrifuged at 2600 × g at 6 °C for 10 min (Allegra 6R, Beckman Coulter), and the serum was stored at −80 °C. All tests were assayed in duplicate by the Endocrine Technology and Support Core at the Oregon National Primate Research Center (Beavertown, OR) by using Immulite 2000 (Siemens Healthcare Diagnostics) technology, a chemiluminescence-based automatic clinical platform that was validated for measurements of nonhuman primate serum. The sensitivity (upper detection) limit for these assays was 5 to 1250 pg/mL for ACTH, 10 to 500 ng/dL for cortisol, 40 to 600 ng/dL for T3, 1 to 40 pg/mL for FT3, 1 to 24 µg/dL for T4, 0.3 to 6 ng/dL for FT4, and 0.004 to 75 µIU/mL for TSH. Intra- and interassay variation with Immulite 2000 is less than 15%. Quality-control samples and validations were performed prior to every batch of hormonal measurements. Results below detectable levels were assigned as one half of the detection limit.

Dermatophyte culture.

For all procedures listed in Figure 1, samples from alopecic animals were obtained from a maximum of 3 separate body locations affected by hair loss. For control subjects, samples were taken from the caudal dorsum, proximal right leg, and distal left arm; these locations were chosen because they are common locations of hair loss. Moreover, the distal arm was particularly important to investigate because of the assumption that hair loss in this location is due to mechanical friction as monkeys extend their arms through the cage feeder hole, an etiology that was suspected in a previous report of alopecia in squirrel monkeys.²⁰

Hair pluck for dermatophyte culture was performed from the periphery of the lesion. To perform this procedure, hair around the lesion was trimmed to 0.25 in. To reduce skin surface contamination, the area was patted with a moist alcohol gauze and then allowed to air dry. The hair was plucked with sterile hemostats, placed in Dermatophyte Test Medium (Shelby Scientific, Macomb, MI), and cultured.

Skin cytology.

Alopecic areas were outlined by using a permanent marker, and any adjacent hair was shaved to avoid contamination. Tape–slide preparations for skin surface cytology were obtained by using Delasco Slides with Pressure Sensitive Adhesive (Council Bluffs, IA). The slides were stained with Diff-Quik stain (Jorgensen Laboratories, Loveland, CO), by using only the stain and counterstain solutions (the fixative step omitted). The slides were examined under a microscope (CX31, Olympus, Center Valley, PA), and the mean number of yeast per 10 separate and random oil-immersion (100×) fields was calculated. Clusters and budding yeast were counted as single items, and fields with hair fibers were omitted from the examination. The results from the 3 sample locations were averaged. Bacterial counts were not performed.

Skin scraping.

Superficial and deep skin scrapings for ectoparasite detection were obtained by using a No.10 stainless steel surgical blade (Becton Dickinson, Franklin Lakes, NJ). For the deep scrape, the skin was squeezed between the sampler's thumb and forefinger and scraped until capillary bleeding was observed. The material obtained on the blade was placed on a microscope slide coated with mineral oil, and a cover slip (Webster Veterinary, Sterling, MA) was applied. Each slide was scanned under a microscope at 40× and 100× magnification to identify whole mites, mite body parts, and mite fecal pellets.

Bacterial culture.

A single culturette for aerobic bacterial culture was used to test all 3 skin sample locations and was inoculated onto a blood agar plate (tryptic soy agar with 5% sheep blood), onto a Macconkey II agar plate, and into thioglycollate broth. The plates were incubated at 37 °C for 48 h or until growth was observed. Cultures with light growth of 3 or more organisms consisting of Staphylococcus, Streptococcus, Corynebacterium, and Neisseria spp. were reported as normal flora.

Histopathology examination.

Skin biopsy samples were collected by using a 6-mm skin biopsy punch (Acuderm, Ft Lauderdale, FL) and aseptically cleaned instruments but without a surgical scrub to avoid disrupting the integrity of the superficial dermal layers. Macaques received analgesics in the form of buprenorphine (0.01 mg/kg IM BID) for 24 to 72 h after the procedure as necessary based on veterinary assessment. Tissues were fixed in buffered zinc formalin fixative (Anatech, Battle Creek, MI), vertically sectioned at 6 µm, stained with hematoxylin and eosin, and viewed under light microscopy (Leica DMLB, Leica Microsystems, Wetzlar, Germany). A board-certified veterinary pathologist who was blinded to alopecia status developed and used a scoring system to evaluate multiple individual characteristics of the epidermis, dermis, and subcutis; values ranged from 0 (normal; Figure 4) to 4 (severe; Figure 5). The number of follicles in anagen, catagen, and telogen were assessed by using a single, random low-power (10×) field. Scores for the individual histopathology characteristics from the 3 biopsy locations were averaged. In addition, individual histopathology characteristic scores were totaled, and a total score ranging from 0 (normal) to 60 (severe) was assigned to each macaque.

Figure 4.

Definitions of characteristics used in the histopathology scoring system.

Figure 5.

Histopathology scoring system for epidermal, dermal, and subcutal characteristics.

Behavioral assessment.

Animals underwent behavioral evaluation to assess the presence or absence of hair plucking. Hair plucking was differentiated from normal grooming by noting whether the animal pulled hair from the body surface with evident force by using exaggerated arm movements; these actions are not associated with normal self-grooming. In addition, hair-plucking animals often were noted to select the particular hairs to be plucked beforehand and generally consumed them afterward. Video sessions involved 6 h of recording (Handycam DCR-SR67, Sony, San Jose, CA) in the home cage over a minimum of 3 time points. The duration of 6 h of videotaping was chosen as a thorough but reasonable amount of effort based on previous experience and with the aim of developing a diagnostic tool that would be practical and feasible for future clinical use. Videorecording was scheduled during times of minimal personnel activity, because previous observations suggested that hair plucking was most likely to occur during these times. Therefore, videorecording was scheduled avoiding daily feedings, routine husbandry, and research procedures. Recording sessions were performed within 30 d before or after medical diagnostics. If performed after medical diagnostics, at least 2 d were allowed for acclimation and to avoid manipulation of biopsy sites, which could resemble hair plucking.

Statistical analysis.

Alopecic and control groups were compared by using Student t tests for numerical comparisons, χ² tests for categorical comparisons, and Fisher exact tests to compare alopecia score with dependent measures (Statistica version 9.0, StatSoft, Tulsa, OK). Results were expressed as mean ± 1 SD, and differences were considered statistically significant at a P value of less than 0.05. In addition, stepwise selection methods were used to build a multiple logistic regression model predicting alopecia from variables showing a univariate association (P < 0.20) with alopecia status (SAS version 9.2, SAS Institute, Cary, NC).

Although published reference ranges for the diagnostic assays performed in this study exist,^16,31,88 many of these used a small sample size or were calculated from clinically ill or stressed animals. Therefore, reference ranges were created, both for the purpose of the study and to use as a clinical tool for future cases of alopecia. Reference ranges traditionally are calculated from a population of healthy subjects by using 2 SD from the mean.¹³ Therefore, we initially calculated these ranges from the group of nonSIV-infected control macaques (n = 23). However, after confirming that there were no significant effects of physical exam findings or infection status on any dependent measure, we used the full set of control animals (n = 39) for calculating the reference ranges, to generate a larger sample size. Because few animals’ values fell beyond this range, it subsequently was reduced to 1 SD.

Results

Animals.

Males were significantly (Fischer exact test; P < 0.005) overrepresented in the alopecic group. Differences in age and weight were nonsignificant (NS) between the alopecic and control groups (age, t = 0.91; NS; weight, t = 0.53; NS). Moreover, no significant differences were seen when the percentage of area affected by hair loss was compared with sex (F = 0.504; NS) or age (F = 1.115; NS) of the animal.

Physical and dermatologic examinations.

The physical and dermatologic examination findings were normal for most animals in both groups (Table 1). The distribution of animals in each alopecia score category is presented in Table 2. No animals at our facility met the criteria of an alopecia score of 5. Photographs of a control macaque with an alopecia score of 0 (Figure 6) and an alopecic animal with an alopecia score of 4 (Figure 7) are provided. In the alopecic group, the extremities (arms, 93.4%; legs, 98.7%) were the most prevalent location for hair loss, followed by the dorsum (53.9%). Alopecia of the distal arm was noted in a minority of the control animals (10.3%) but a majority of the alopecic animals (89.7%). Within the alopecic group, the most common configuration observed was patchy (56.4%), and focal (2.6%) was the least common.

Table 1.

Summary of physical and dermatologic findings in alopecic and control groups (% of all animals in group)

	% Alopecic	% Control
No abnormalities	69.2	76.9
Dental tartar (mild or moderate)	17.9	20.5
Lymphadenopathy	10.3	2.6
Decreased range of joint motion	5.1	0
Vaginal prolapse	2.6	0
Inguinal hernia	0	2.6
Macule	28.2	2.6
Papule	12.8	0
Scale	17.9	0
Crust	15.4	0
Erythema	23.1	0
Erosion	2.6	0
Lichenification	12.8	2.6

Table 2.

Percentage (by group) of animals in each alopecia score category

Alopecia score	% Alopecic	% Control
0	0	53.8
1	0	46.2
2	0	0
3	87.2	0
4	12.8	0
5	0	0

Figure 6.

Photograph of the dorsum of an animal with an alopecia score of 0.

Figure 7.

Photograph of the dorsum of an animal with an alopecia score of 4.

The alopecic group had no significant differences in the location of hair loss between animals that were observed to hair pluck and those that did not (χ² = 3.01; NS). Within the alopecic group, the alopecia configuration differed by less than 10% between the subjects that were observed hair plucking compared with those that were not.

Hematology.

The CBC and CD4⁺ T lymphocyte measures showed no significant differences between the alopecic and control groups (Table 3). In addition, when the analysis included all of the animals enrolled in the study, there was no association between CD4⁺ T lymphocyte counts and SIV infection status (t = 0.19; NS). The only significantly different serum biochemistry value in the alopecic compared with control animals was a decreased creatinine in the alopecic group (t = 2.81; P < 0.05; Table 4). However, means for both groups for creatinine fell within the calculated reference range (that is, mean ± 1 SD of control group).

Table 3.

CBC values (mean ± 1 SD) of alopecic compared with control animals

	Alopecic	Control	P
WBC (×103/µL)	7.02 ± 2.40	7.60 ± 2.36	0.2853
RBC (×106/µL)	5.69 ± 0.45	5.73 ± 0.49	0.6581
Hemoglobin (g/dL)	13.05 ± 1.09	13.00 ± 1.04	0.8324
Hematocrit (%)	40.47 ± 3.35	40.25 ± 2.99	0.7519
Platelets (×103/µL)	346.54 ± 100.67	334.41 ± 89.16	0.5749
Neutrophils (×103/µL)	3.69 ± 1.66	3.83 ± 1.56	0.6853
Lymphocytes (×103/µL)	2.69 ± 1.07	3.07 ± 1.17	0.1452
Monocytes (×103/µL)	0.37 ± 0.16	0.40 ± 0.17	0.5380
Eosinophils (×103/µL)	0.26 ± 0.13	0.28 ± 0.14	0.4024
Basophils (×103/µL)	0.02 ± 0.02	0.03 ± 0.02	0.5657
CD4+ T lymphocytes (cells/µL)	627.03 ± 320.33	722.79 ± 267.03	0.1556

No parameter differed significantly (Student t test, P < 0.05) between groups.

Table 4.

Serum biochemistry values (mean ± 1 SD) of alopecic compared with control animals

	Alopecic	Control	P
Sodium (mEq/L)	146.56 ± 1.80	146.31 ± 1.36	0.4806
Potassium (mEq/L)	3.90 ± 0.28	3.93 ± 0.34	0.6642
Chloride (mEq/L)	109.28 ± 1.97	109.13 ± 1.95	0.7300
Glucose (mg/dL)	65.13 ± 15.53	66.79 ± 12.19	0.5996
Blood urea nitrogen (mg/dL)	17.34 ± 3.62	18.50 ± 2.48	0.1035
Creatinine (mg/dL)	0.79 ± 0.16	0.89 ± 0.14	0.0064
Calcium (mg/dL)	9.78 ± 2.27	9.34 ± 0.39	0.2407
Phosphorus (mg/dL)	4.50 ± 1.24	4.30 ± 1.08	0.4651
Total protein (g/dL)	6.73 ± 0.48	6.80 ± 0.39	0.4708
Albumin (g/dL)	3.94 ± 0.25	3.85 ± 0.39	0.2568
Globulin (g/dL)	2.79 ± 0.38	2.92 ± 0.38	0.1356
Alkaline phosphatase (U/L)	123.23 ± 100.29	131.85 ± 112.88	0.7226
Alanine aminotransferase (U/L)	35.72 ± 27.11	29.41 ± 13.94	0.2002
Asparate aminotransferase (U/L)	31.87 ± 12.68	28.13 ± 5.74	0.0972
Lactate dehydrogenase (U/L)	545.95 ± 258.51	506.13 ± 201.04	0.4450
γ-Glutamyltransferase (U/L)	44.97 ± 13.39	42.87 ± 9.85	0.4319
Total bilirubin (mg/dL)	0.38 ± 1.37	0.18 ± 0.16	0.3597
Cholesterol (mg/dL)	136.74 ± 26.25	138.69 ± 25.93	0.7424
Triglycerides (mg/dL)	61.33 ± 58.73	55.21 ± 28.28	0.5588

Only creatinine values differed significantly (Student t test, P < 0.05) between groups.

Endocrinology.

Alopecic animals had significantly higher cortisol, T3, and FT3 values than did control animals (Student t test; P < 0.05). No significant differences were seen in the other endocrine values (Table 5). Means for cortisol and FT3 fell within the calculated reference ranges for both the alopecic and control groups, whereas the mean for T3 in the alopecic group fell slightly above the upper limit of the reference range. The distribution (symmetric compared with asymmetric) of the alopecia was compared with each of the measured endocrine values; however, there were no significant differences due to distribution in any of these values (Student t test; NS).

Table 5.

Endocrinology panel values (mean ± 1 SD) of alopecic compared with control animals

	Alopecic	Control	P
ACTH (pg/mL)	13.77 ± 13.07	12.03 ± 10.47	0.5190
Cortisol (ng/dL)	245.56 ± 97.70	199.66 ± 66.58	0.0177
FT3 (pg/mL)	6.68 ± 3.27	5.53 ± 1.13	0.0412
T3 (ng/dL)	171.12 ± 66.47	142.08 ± 28.34	0.0142
FT4 (ng/dL)	1.73 ± 2.75	1.45 ± 1.23	0.5575
T4 (µg/dL)	3.17 ± 1.53	3.25 ± 1.60	0.8249
TSH (µIU/mL)	0.56 ± 0.39	0.69 ± 0.67	0.3062

Cortisol, FT3, and T3 values differed significantly (Student t test, P < 0.05) between groups.

Measuring FT4 and TSH to diagnose hypothyroidism, as recommended for canine⁶³ and human⁷⁸ patients, did not result in a diagnosis of this disease in any of the alopecic macaques. In addition, hyperadrenocorticism was not responsible for any cases of alopecia, according to previous criteria used to diagnose this disease in nonhuman primates.⁴¹

Dermatophytes.

The alopecic group had 37 negative dermatophyte cultures, with the medium of one animal growing Aspergillus spp. and another Penicillium spp. The control group had 37 negative dermatophyte cultures, with the medium of one animal growing Cladosporium spp. and another Penicillium spp.

Skin cytology.

The shape and size of yeast seen on the surface cytology results were consistent with Malassezia spp. Compared with the control animals, the alopecic group had significantly higher numbers of yeast per oil-immersion field (alopecic, 0.52 ± 0.42; control, 0.23 ± 0.23; t = 3.79; P < 0.001). The means for both groups fell within the calculated reference range.

Skin scraping.

Skin scrapes for 76 animals were negative for ectoparasites. One alopecic animal was positive for Demodex spp. mites, and an unidentified mite was found on one control animal.

Bacterial culture.

All animals in both groups had bacterial culture results that were considered normal flora and included growth of Staphylococcus, Streptococcus, Corynebacterium, or Neisseria spp. or various combinations thereof.

Histopathologic examination.

Several individual histopathologic characteristics occurred almost exclusively in the alopecic group. These included orthokeratotic hyperkeratosis (Figure 8) and acanthosis (Figure 8) in the epidermis and comedone (Figure 9) and apocrine gland ectasia (not shown) in the dermis and subcutis. Comparison of the histopathologic characteristics in a sufficient number of control animals to permit statistical analysis found that scores were significantly higher in the alopecic group; these characteristics included vessel ectasia (Figure 10), perivascular edema (Figures 8, 10, and 11), and lymphoplasmacytic perivasculitis (Figure 11) in the dermis and subcutis (Student t test; P < 0.05). An etiologic diagnosis could not be made from any of the examined biopsy specimens. The average total histopathology score for the alopecic macaques was significantly higher than that of the control animals (alopecic, 4.81 ± 3.20; control, 0.69 ± 1.35; t = 7.4; P < 0.0001; Figures 8 and 12). In the stepwise logistic regression model, the total histology score was the best predictor of alopecia (χ² = 20.19; P < 0.001). When total histology score was in the model, no other predictors were necessary.

Figure 8.

Histopathology from the caudal dorsum of the alopecic animal with the highest total histopathology score: orthokeratotic hyperkeratosis score, 2; acanthosis score, 4; perivascular edema score, 2; and total histopathology score, 17. Hematoxylin and eosin stain; scale bar, 100 µm.

Figure 9.

Histopathology from the caudal dorsum of an alopecic animal with a comedone score of 2. Hematoxylin and eosin stain; scale bar, 100 µm.

Figure 10.

Histopathology from the caudal dorsum of an alopecic animal with a vessel ectasia score of 2 and perivascular edema score of 2. Hematoxylin and eosin stain; scale bar, 100 µm.

Figure 11.

Histopathology from the caudal dorsum of an alopecic animal with a perivascular edema score of 2 and lymphoplasmacytic perivasculitis score of 2. Hematoxylin and eosin stain; scale bar, 100 µm.

Figure 12.

Histopathology from the caudal dorsum of a control animal with a total histopathology score of 0. Hematoxylin and eosin stain; scale bar, 100 µm.

There were significant differences in vessel ectasia (t = 2.34; P < 0.05) and perivascular edema (t = 2.03; P < 0.05) when sex was used as an independent variable; however, no differences were seen in the 13 remaining histopathologic characteristics. No significant differences in the histopathologic characteristics were noted based on SIV infection status (Student t test; P < 0.05). When we evaluated primary and secondary dermatologic lesions, macaques in the alopecic group with at least one of these lesions had a significantly higher total histopathology score than did alopecic animals without lesions (t = 4.70, P < 0.0001). Finally, when animals in the alopecic group that hair plucked were compared with animals that did not, there were no significant differences in any of histopathologic characteristics (Student t test; NS) or the total histopathology score (t = 0.63; NS).

The number of hair follicles in catagen or telogen was significantly higher in the alopecic group, whereas the number of hair follicles in anagen was lower (Student t test; P < 0.05). However, when these values were converted to percentage of follicles in each phase of the hair cycle, there were no significant differences between the alopecic and control groups (Student t test; NS).

Behavioral assessment.

A significantly higher proportion of macaques were observed to hair pluck in the alopecic group than in the control group (alopecic, 53.9%; control, 12.8%; χ² = 14.77; P < 0.0005). For the logistic regression predicting alopecia, when the total histopathology score was excluded, the presence or absence of hair plucking emerged as the only predictor of alopecia (χ² = 12.90; P < 0.005). The majority of macaques in both groups that were observed to hair pluck also demonstrated trichophagia (alopecic, 90.5%; control, 80.0%), and 54% of the animals in both groups were observed hair plucking within the first 2 h of video footage. The mean time until the observer noted hair plucking was 140.0 ± 108.5 min, with a minimum of 3.17 min and maximum of 345.0 min (n = 26).

Discussion

In the current study, we investigated the cause of alopecia in captive rhesus macaques by using medical and behavioral diagnostics. Medical diagnostics were based on the typical diagnostic plan used for evaluating alopecia in traditional veterinary species. The behavioral assessment was used to determine whether the macaques engaged in hair plucking.

Overall, many of the findings in the current study are congruent with other published reports investigating alopecia in nonhuman primates. A male majority comprised the alopecic group in the current study, which is consistent with the finding that there is a higher percentage of alopecia in males at our facility. However, this finding contrasts to previous reports in which the majority of alopecic animals were female.^45,75,91 This difference could be due to the fact that in the current study, hair loss due to pregnancy was not a variable as it had been in previous reports. Another report of 637 monkeys found no sex-associated differences seen with respect to alopecia.²³ In addition, our lack of finding a significant difference in age between the alopecic and control animals contrasts with previous studies, in which older animals had a greater incidence of alopecia.^{11,23,45,75,91} However, this difference may be attributable to the fact that animals in the current study were younger than in prior studies. Finally, the lack of a significant difference in weight is concordant with a previous study comparing alopecic and normal squirrel monkeys.³⁶

Although macaques were included regardless of health status, the majority of the animals in both groups had normal physical examination results or minor abnormalities that most likely were independent of alopecia. This finding is in agreement with previous studies, in which physical examination of alopecic animals did not reveal noteworthy abnormalities.^45,76 Our dermatologic examination revealed more primary and secondary dermatologic lesions in the alopecic group. These lesions can aid in formulation of a differential diagnosis.^51,70 Macules, the most common lesion in the alopecic animals, can be associated with endocrine dermatoses.¹ However, in the current study, their clinical significance is questionable, because the diagnostic procedures did not confirm this differential diagnosis or any others associated with these lesions. Moreover, this discrepancy may be due in part to an improved ability to identify these dermatologic lesions without hair present. Because they predominantly occurred in alopecic animals with high total histopathology scores, these lesions could be investigated further in future studies to elucidate any significance. Alopecia was found predominately on the arms and legs, which contrasts with reports that noted its prevalence on the dorsum^36,76 and head.⁹¹ Alopecia of the distal arm occurred in a minority of the control macaques but in the majority of alopecic animals. If alopecia in this area was predominately due to mechanical friction, similar proportions of animals from both groups would be expected to have alopecia in this location. Because this was not the case, the attribution of hair loss to mechanical alopecia was not supported. This theory also was discredited in a study of squirrel monkeys that correlated handedness and alopecia location.²⁰

CBC parameters in the current study revealed no significant differences between alopecic and control macaques. These results parallel those of a previous report that found no major CBC abnormalities in either alopecic or normal rhesus macaques;⁴⁵ moreover, the few correlations observed in this report either fell within reference ranges or did not help to elucidate the underlying etiology of the hair loss. However, in contrast with the current and previous⁴³ results, another study⁷⁶ found that animals with hair coat damage were neutrophilic and lymphopenic compared with those without coat damage. This hematologic pattern was assumed to be due to a stress response or physiologic variation.⁷⁵ Alopecic animals in our current study did not show noteworthy indicators of stress, such as hyperglycemia, hypercortisolemia, or a stress leukogram. CD4⁺ T lymphocyte counts were not significantly different in alopecic and control groups in the current study, and the presence of SIV infection did not significantly affect the dependant variables in either group. The practical utility of the significant difference seen in creatinine is unknown. In another report,⁴⁵ a similar result was considered irrelevant, because it did not indicate a cause for the alopecia.

The clinical utility of the differences in cortisol, T3, and FT3 are unknown. Endocrinopathies predominantly manifest as bilaterally symmetric alopecia in companion animal species;⁷⁰ however, in the current study, the distribution of the alopecia did not vary with any of the measured endocrinology values. In addition, the histopathologic results in the current study were nonspecific and did not aid in diagnosing an endocrine disorder. Although hyperadrenocorticism can cause alopecia with follicular atrophy, other common findings of canine and feline Cushing-like syndromes, such as atrophy of the epidermis and follicular infundibular epithelium, thinning of the dermis, and calcinosis cutis, were not present in our macaques. However, we cannot rule out the possibility that skin manifestations of endocrinopathies in rhesus macaques differ from those of traditional veterinary species. In the report cited earlier,⁴⁵ alopecia was not associated with altered cortisol levels or thyroid dysfunction (T4 and FT4). Moreover, the creatinine and endocrinologic results are likely clinically insignificant, because the means largely fell within the calculated reference ranges, other published references ranges, and ranges used by the clinical pathology laboratory at our facility.^16,27,41,57 Therefore, these tests would have minimal utility as screening tools for future cases of alopecia, but they could be investigated in future studies to elucidate any practical significance.

In agreement with another study,⁷⁶ the majority of the dermatophyte cultures for both groups in the current study were negative. Organisms cultured from control and alopecic animals included Penicillium, Cladosporium, and Aspergillus spp., which were the 3 most common fungal isolates in a survey of dermatophyte cultures performed on domestic and laboratory animals.³ The growth of these ubiquitous saprophytic fungi likely derives from the environment, because the hair of animals easily acquires environmental contaminants. In addition, other than alopecia, few dermatologic lesions consistent with dermatophyte infections, including scaling, crusts, and erythema,⁷⁰ were noted in our macaques, supporting the environmental contamination theory.

With respect to the skin cytology results, the number of yeast organisms differed significantly between our alopecic and control macaques. However, the mean for each group fell within the reference range. These organisms were consistent with Malassezia spp., which is considered normal skin flora in many clinically normal animals, including dogs, cats, ferrets, foxes, bears, pigs, horses, birds, and rhinoceroses.^19,40 In addition, other authors did not find an association between fungal isolates and hair coat quality.⁷⁶ Moreover, in dogs, the number of yeast organisms per high-power field varies widely in healthy animals, and a diagnosis of dermatitis is not made until more than 10 yeast are found per one-half square inch area, which corresponds to the area under a 10×-power field.⁶⁰ In the current study, the means for both the alopecic and control groups were less than 1 yeast per 100× field, which supports the conclusion that yeast organisms were not the cause of alopecia.

Ectoparasite detection by superficial and deep skin scraping was unrewarding, except for one alopecic macaque that was positive for a Demodex spp. mite. This animal was an SIV-negative adult rhesus macaque with abnormal physical and clinicopathologic exam findings including hypercortisolemia and axillary and mesenteric lymphadenopathy. The clinical findings in this animal are similar to those found in cases of canine adult-onset demodicosis, which is secondary to an underlying systemic disease that weakens the immune system, such as hypothyroidism, iatrogenic hypercortisolism, hyperadrenocorticism, and neoplasia.⁵¹ Therefore, the Demodex spp. mite-positive animal in our study most likely had an underlying disease that led to secondary mite infestation and possibly contributed to its alopecia. In contrast, a report concerning Demodex spp. mites in rhesus macaques demonstrated that mites were observed regardless of immune status.⁷⁴ In addition, one control animal was positive for a mite that could not be identified but was not consistent with Sarcoptes, Notoedres, or Trixacarus spp. or any of the other genera of mite commonly seen in laboratory animals. Overall, the current study offers minimal support for an association between ectoparasites and alopecia, in agreement with results of a previous study.⁷⁶

The bacterial culture results for all macaques showed growth of Staphylococcus, Streptococcus, Corynebacterium, or Neisseria spp., which is consistent with bacterial flora of normal human and nonhuman primate skin.^12,43,72 Moreover, because we noted minimal primary or secondary dermatologic lesions consistent with pyoderma, we can conclude that these organisms were commensal skin bacteria rather than abnormal overgrowth.⁷⁰

Except for the histopathologic results, the current study is consistent with other reports showing that the use of typical diagnostics to identify the cause of nonhuman primate alopecia has a very low yield.^45,76 The histopathology findings in this study parallel those of another,⁴⁵ which revealed acanthosis, hyperkeratosis, and mild edema in alopecic animals consistent with a chronic hypersensitivity reaction or atopic-like dermatitis. Although still another report found a decrease in epidermal cell layers in animals with alopecia,⁷⁶ our study found acanthosis (epidermal hyperplasia) primarily in this group. In addition, orthokeratotic hyperkeratosis and perivascular dermatitis were seen in a survey of 183 randomly chosen rhesus macaques with no association with hair loss.⁷⁶ Overall, the histologic findings in our study are nonspecific and could therefore be associated with many forms of chronic dermatoses⁷⁶ and may even occur in animals without hair loss. Because hair offers a protective barrier, having exposed, nonhaired skin in an area where hair would normally be found could increase antigenic stimulation, leading to the perivascular inflammation we noted. Consequently, determining whether the histopathologic differences reflect the cause or the result of alopecia is difficult.

The significant differences between our alopecic and control macaques in the numbers of hair follicles in the 3 different hair growth cycles are interesting even though these differences vanish when the raw numbers are converted into percentages of hair follicles found in each of the 3 different phases of the hair cycle. An increase in telogen follicles with a concurrent decrease in anagen follicles in the absence of peribulbar inflammation is suggestive of androgenetic alopecia or telogen effluvium.⁷³ Androgenetic alopecia is an androgen-dependent hair loss from the frontoparietal scalp region.¹⁸ This condition has been documented in a few nonhuman primate species, such as stump-tailed macaques,⁷⁷ but never definitively in rhesus macaques. In addition, because alopecia in the current study affected multiple areas of the body, not just the scalp, androgenetic alopecia is an unlikely diagnosis for the macaques in our study.

Telogen effluvium is another documented cause of alopecia in nonhuman primates.³⁶ A primary or secondary, chronic or acute, disruption of the hair cycle can lead to increased shedding of telogen hairs and, therefore, a diagnosis of telogen effluvium. Secondary telogen effluvium in humans can be due to multiple mechanisms, including pregnancy, aging, hypothyroidism, iron storage depletion, and environmental stress.³⁶ The majority of these causes can be disregarded in the current study, given that there were no pregnant animals, no significant age difference between the 2 groups, no decrease in thyroid levels, and no microcytic hypochromic anemia associated with iron deficiencies. There is minimal evidence to support a theory of stress-related telogen effluvium in the current study. Cortisol levels for both the alopecic and control groups were within normal limits. In addition, in humans, stress is poorly documented as a valid secondary cause of telogen effluvium,³⁵ and stress was considered an unlikely cause of telogen effluvium in a study of alopecic squirrel monkeys.³⁷ Moreover, other mechanisms must be ruled out, given that telogen effluvium can be a multifactorial disease, and the diagnosis of primary telogen effluvium is one of exclusion.^71,86 Therefore, telogen effluvium is a difficult differential diagnosis to confirm in rhesus macaques.⁷⁶ In human medicine, some advocate the review of horizontally sectioned skin biopsies for reliable diagnosis of telogen effluvium to better categorize the hair cycle stage;⁷¹ this review was not attempted in the current study. However, this issue is controversial among dermatopathologists and to circumvent this dilemma, the use of both vertical and horizontal sections might have led to a definitive diagnosis in the current study.^14,28,29

In humans, a diagnosis of trichotillomania, a compulsive disorder associated with hair pulling, can be confirmed or suggested through histopathology results.⁵² Histologic features of trichotillomania include increased catagen and telogen hair follicles, trichomalacia, melanin casts, and lack of significant inflammation. Incomplete, disrupted follicles are diagnostic even though they are not always present.^30,73 In our study, there were no significant differences in histopathologic results between the alopecic macaques that were observed hair plucking and those that were not. This similarity suggests that diagnosing hair plucking in nonhuman primates by using histopathology is difficult. In addition, the pathognomonic histologic features seen in trichotillomania in humans, including trichomalacia and melanin casts, were not present in either the current or a previous⁷⁶ study. However, as with telogen effluvium, the addition of horizontally sectioned hair follicles might provide a more reliable diagnosis.⁵²

Just over half (54%) of the alopecic macaques in the current study were observed to hair pluck. These animals trended to have hair loss from the caudal dorsum as compared with those not observed to hair pluck. This finding contrasts with a report that noted thigh and forearm alopecia in animals with presumed pyschogenic alopecia.³⁴ In our study, aberrant behavior cannot account for all cases of alopecia, as also was noted in previous surveys.^36,76,85 The assumption that a behavioral etiology is responsible for all cases of alopecia would have overestimated the actual number of animals that pulled their hair and led to inappropriate diagnosis in almost half of these animals. Therefore, the true underlying cause of the alopecia would not have been investigated. In addition, hair plucking may not be the primary underlying cause of the hair loss in these animals, because both hair plucking and alopecia are multifactorial diseases.⁶⁷ In addition, physical examination in the current study showed that the anatomic location of hair plucking was not always at or near the site of alopecia. More importantly, 13% of the control animals were observed to hair pluck, suggesting that hair plucking that does not lead to clinical alopecia may be more prevalent than thought. These macaques might ultimately develop visible alopecia as a result of hair plucking; however, this information is not available because these animals were not reevaluated for this study. Nevertheless, the abnormal behavior of hair plucking and the appearance of an animal's hair coat may not be as strongly associated as previously perceived.

The current study has limitations that merit review. Study macaques were housed in various rooms and buildings at our facility, with lighting, temperature, and humidity maintained within appropriate ranges in accordance with the regulations. However, because every animal was not exposed to precisely the same environment, these variables could influence the study results. In addition, subjects included both female and male macaques without accounting for hormonal differences between the sexes. However, given that only 2 of the 15 histopathologic characteristics were significantly different based on sex, the extent to which this difference confounded the histopathology diagnostic results may be minimal. Finally, our study was cross-sectional as compared with longitudinal or prospective, because this design is an accepted method for characterizing a colony and identifying animals that require assessment or intervention for alopecia. Future studies that follow individual animals over time and screen for the need for assessment or intervention according to temporal patterns could be fruitful for further elucidating the pathogenesis of alopecia.

Moreover, although we intended the diagnostic plan to be broadly encompassing, various diagnostic tests that were not performed might have led to a diagnosis. First, the diagnostic yield may have increased by sampling from more than 3 alopecic sites. However, this notion is most likely not applicable to skin biopsy, given that a previous report in humans demonstrated 98% of the cases of alopecia were diagnosed by using 3 biopsies.⁷¹ Second, hyperadrenocorticism might have been assessed more appropriately by performing an ACTH stimulation or dexamethasone suppression test, as is recommended for companion animal species.⁶⁸ However, these tests would have confounded the primary study to which the animals were assigned and would be impractical in many biomedical research facilities. Nutritional causes of alopecia are reported in the literature for human and nonhuman primates^{17,21,32,39,50,53,61,62,63,77} but were not investigated in the current study, because all animals were indoor-housed and received a nutritionally complete commercial monkey chow. An exclusion diet trial to address food hypersensitivity was not performed, and this assessment is part of the diagnostic workup for companion animal species.⁸³ Food allergy dermatitis was unlikely because pruritus typically occurs in companion animals with this disease and was not widely reported in animals on the current study. Finally, although hyperthyroidism is a common differential diagnosis for alopecia in humans, it is not included in the diagnostic work-up for canines, on which we based our plan. In addition, hyperthyroidism is rare in nonhuman primates^15,44 and was therefore not specifically investigated in the current study. However, the extensive endocrinology panel permitted diagnosis of this disease in study subjects. Although the majority of cases of alopecia among our macaques cannot be attributed to hyperthyroidism, one alopecic animal met criteria for a diagnosis of hyperthyroidism with an increased FT4 and a concurrent decrease in TSH; these changes are diagnostic criterion in human medicine.⁶⁵ Overall, these additional diagnostics might have accounted for a minimal number of alopecic cases, but because the associated diseases are uncommon in companion animal species, it is unlikely that they would account for the majority of undiagnosed cases among nonhuman primates.

The total histopathology score was the best predictor of whether an animal had alopecia; therefore, the histopathology results should be investigated further to probe for more specific diagnostic results. This end can be accomplished by performing immunohistochemistry and special staining techniques. In addition to the conventional vertical orientation, tissue could be processed and examined in multiple horizontal sections to increase diagnostic yield.²⁹ Vertical sections demonstrate changes at the dermoepidermal junction, papillary dermis, and subcutaneous tissue and are used to allow for full examination of all levels of hair follicles, although only a few may be identifiable, whereas horizontal orientation allows for visualization of multiple hair follicles in a single section.^29,36 The addition of a horizontal section to the histopathology results will be included in a future study. Calculation of the number of hair follicles in each phase of the hair cycle may not have provided a complete picture of the current condition, given that we evaluated only a single random 10× power field as a screening tool. Typically, several fields are assessed to make a determination of hair growth phase distribution. Again, this finding could be explored further by more extensive evaluation of tissue sections.

In the current study, we used the diagnostic plan for alopecia that was developed for traditional veterinary species as a guide to design a diagnostic algorithm for the condition in nonhuman primates. However, because complete dermatologic evaluation did not yield a definitive diagnosis for the primary cause of alopecia in any of the cases, we were unsuccessful at devising such a tool for nonhuman primate alopecia. The diagnostic guide was helpful in ruling out the most common causes of alopecia in traditional veterinary patients, leading us to conclude that these causes are not congruent with those of nonhuman primates. Therefore, after completion of a minimum initial database, we recommend shifting the focus away from the well-established diagnostic guide for alopecia in companion animals and pursuing unconventional differentials for alopecia in nonhuman primates with the aid of additional advanced diagnostics. Alternatively, a diagnosis can be pursued by treating cases empirically and examining the responses to these treatments. Nevertheless, a proportion of the cases of alopecia will likely not be attributable to either a medical or a behavioral cause and subsequently will be classified as idiopathic. Although our results indicate that many traditional diagnostic methods will have a low diagnostic yield, individual institutions should develop a plan for how to approach cases of alopecia in their colonies. This clinical decision is made by the veterinarian, after balancing the rarity, severity, and treatability of the condition compared with the time, expense, and potential stress to the animals in diagnosing and treating the conditions. However, recently, a trend has emerged in which the decision to pursue diagnostics is based on regulatory concern rather than veterinary expertise.

Alopecia in nonhuman primates has only recently become a pertinent topic to the field of laboratory animal medicine. Therefore, research in this subject lags behind that of other veterinary species. Future studies are needed to elucidate the pathogenesis of alopecia in nonhuman primates so that effective treatment plans can be designed. The results of the current study support the hypothesis that the standard dermatologic diagnostic plan for alopecia in traditional veterinary species has minimal diagnostic value for identifying the causes of alopecia in nonhuman primates.