An excisional biopsy should be performed on any pigmented lesion for which no precise diagnosis has been reached, and this is particularly the case when melanoma is suspected. Occasionally, however, an incisional biopsy is performed to rule out malignancy. Although an incisional biopsy might be adequate in other types of skin tumors, melanocytic lesions require consideration of both tissue architecture and cytology for histologic diagnosis. Two of the architectural criteria used are symmetry or asymmetry and the presence or not of diffuse borders. Both these criteria are difficult or impossible to evaluate in partial biopsies. Furthermore, marked cell atypia is often not evenly distributed in the lesion and diagnostic error might result if cytologically pseudobenign zones are biopsied.8

If large melanocytic lesions cannot initially be excised completely, the operator should biopsy from a site that is as representative as possible. The usefulness of dermatoscopy for selecting sites with the greatest Breslow depth (as has also been attempted with confocal microscopy) was recently studied.9 Also under investigation are confocal microscopy criteria that might be used to identify features associated with lentigo maligna when choosing biopsy sites.10,11 We have mentioned that shave biopsy is not the best technique in these cases, although some authors do think that this technique can sometimes be useful.8,12 When deciding between diagnoses of melanoma versus nevus, or sclerosing basal cell carcinoma versus desmoplastic trichoepithelioma (Fig. 6), shave biopsies will be too superficial to allow the dermatopathologist to reach a definitive diagnosis.13

Figure 6.

Desmoplastic trichoepithelioma. The cords of squamous cells that can be seen in the reticular dermis would make it difficult to rule out a sclerosing basal cell epithelioma in an incisional biopsy. Hematoxylin–eosin, original magnification ×20.

(0.47MB).

Finally, to ensure the patient receives the very best medical care possible, a biopsy technique should be chosen not only because it is the simplest one and is associated with the least morbidity but also because it will facilitate a certain diagnosis.

Immunofluorescence is particularly useful for examining biopsy specimens in vesicular and bullous diseases, including bullous pemphigoid, acquired epidermolysis bullosa, gestational pemphigoid or herpes gestationis, cicatricial pemphigoid, dermatitis herpetiformis, and other forms of pemphigoid dermatoses in which blisters develop, such as lupus erythematosus or lichen ruber planus. Direct immunofluorescence (DIF) with antibodies against immunoglobulin (Ig) G, IgA, IgM, C3, and fibrin is required in order to properly characterize lesions in these diseases. DIF, which identifies proteins found at the dermal–epidermal junction, is useful to distinguish erosive lichen planus from lupus erythematosus or mucosal pemphigus, for example, because these lesions may resemble each other clinically or histologically. When Henoch–Schönlein purpura is suspected, evidence of vasculitis should be studied14 (Fig. 7) in biopsy specimens taken within 24hours of a lesion's appearance.15 This approach is also relevant in livedoid vasculopathy, in which a pattern of marked fibrin and C3 deposition in superficial and deep vascular plexuses was recently described.16

Figure 7.

Small vessel leukocytoclastic vasculitis and incipient necrosis of the epidermis. Indirect immunofluorescence confirmed immunoglobulin A deposits, leading to a diagnosis of Henoch–Schönlein vasculitis.

(0.57MB).

Lesions in bullous diseases should also be biopsied early to avoid sampling tissue showing reepithelization. Ideally, healthy skin bordering the lesion (<5mm) should also be sampled.17 When bullous lesions older than 24h are biopsied, the dermatopathologist often observes reparative changes on hematoxylin–eosin staining.18,19 How to proceed when taking samples for DIF is debatable because even though rapid processing of the sample is recommended, some authors have argued that storing the tissue in isotonic saline solution at room temperature for 24 to 48hours eliminates residual fluorescence and improves results.20 When DIF is used to study a bullous lesion, immune deposits can be distorted by tissue separation or degraded owing to the inflammatory response, leading to false negative results.17 As can be guessed, the best DIF results in bullous diseases are obtained with samples from healthy perilesional skin. This is equally true in dermatitis herpetiformis. In lupus erythematosus, on the other hand, the erythematous zone should be biopsied.21 When biopsying epidermolysis bullosa lesions, some authors recommend inducing a fresh blister by rubbing the skin with an eraser just before taking the sample in severe cases or showing parents how to rub the zone to induce erythema in less severe cases.22

It is important to know that a tissue sample for an indirect immunofluorescence (IIF) assay should not be fixed in formol but rather transported wrapped in gauze moistened with physiological saline solution. An alternative is Michel medium, which preserves tissue reactivity for at least 10 days15,23 and possibly for as long as 6 to 8 weeks.22

In some diseases, such as bullous pemphigoid, the area of the body biopsied can also affect the results of DIF: negative results are more likely when the tissue sample comes from the lower limbs.24

Finally, the findings of DIF can be complemented by other diagnostic methods such as IIF, enzyme-linked immunosorbent assay, or western blotting.25Table 1 summarizes measures to take into consideration when preparing biopsy specimens for DIF.

A thorough medical history and physical examination that include an estimate of hair loss, a pull test, and the inspection of follicular openings is necessary to study dermatoses that cause alopecia.26 We will now describe some of the most commonly used techniques.

Signs of inflammation, scaling, erythema, and scarring of the scalp should be looked for during the physical examination. Follicular openings should also be examined.27 Other possible tests to perform are a pull (or pinch) test (in scarring alopecia the hair comes out easily), dermatoscopy, and a trichogram.28

An optical microscope can be used not only to study a biopsy specimen but also to examine the shafts of hairs removed in the pull test; the hairs can be fixed on a slide, hydrated for 10minutes and examined for morphology and growth cycle.

The main indications for taking a scalp biopsy are suspicion of scarring (permanent) alopecia, the presence of tumors on the scalp, inflammatory diseases of uncertain etiology that affect the scalp, and situations in which the growth potential of existing follicles must be determined.

As there are no standard tests for alopecia, each case is considered individually according to clinical signs.29,30

The zone to be biopsied should be shaved and petroleum jelly or tape should be applied to fix the surrounding hairs. The biopsy should be performed in the direction of the axis of the follicle, not perpendicular to the skin, to avoid cutting across follicles. Orentreich punches, which are tools used mainly for hair transplants, are especially appropriate for scalp biopsies as they help the operator take a deeper sample with the same diameter as other punch tools.31 Bleeding is common because the scalp is highly vascularized. For optimal hemostasis, use a local anesthetic with epinephrine. Some consider a 4-mm biopsy to be adequate,30,32 but in our experience samples that small do not always contain enough subcutaneous tissue to allow more than a single follicle to be examined, making it difficult to arrive at a definitive diagnosis. As most dermatopathologists evaluate transverse and longitudinal sections of the sample to be certain of their findings, the operator can take 2 punch biopsies of 3 to 4mm or can section a single punch of 6mm (Fig. 8).

Figure 8.

Punch biopsies from a patient with lichen planopilaris. Hematoxylin–eosin, original magnification ×40. A, Longitudinal section. B, Transverse section.

(0.47MB).

When scalp diseases manifest with pustules, microbiology to rule out bacterial and fungal infections will be useful (Fig. 9).

Figure 9.

This highly damaged biopsy showed folliculitis, an empty central zone (an artifact caused by a clamp), and a sclerosed area at the bottom of the field (a compression artifact). Hematoxylin–eosin, original magnification ×20.

(0.32MB).

If scarring alopecia is suspected, another sample should be taken and sent for DIF.

In scarring dermatoses (mainly lupus erythematosus and lichen planopilaris), take the biopsy from an active margin, selecting a site at the edge of a lesion or in an area with erythema and avoiding areas that show scarring (which should ideally be reflected in no more than a third of the tissue sampled). We feel it is useful to make serial sections and use Alcian blue, periodic acid-Schiff, and elastic fiber stains and even the Ziehl–Neelsen technique in some cases.33

The dermatopathologist should at least examine the following histologic features: hair density and structure (there should be at least 30 terminal hair follicles per 4-mm cross section31); adnexal involvement; epithelial changes; the pattern, grade, and location of fibrosis; changes at the dermal–epidermal junction; and finally, the distribution and predominant cell content of an infiltrate if present (Fig. 10).

Figure 10.

Algorithm for histopathology in cases of alopecia. Adapted from Harries and coworkers27 and Hermes and coworkers.29 PAS refers to periodic acid-Schiff.

(0.44MB).

Biopsy is strongly indicated when nail changes are difficult to diagnose based on clinical examination, radiology, or laboratory methods.

The nail region is unfortunately a portion of the tegumentary system where misdiagnosis is most likely even though nail complaints are the reason for consultation in about 10% of cases.34

Tumors, whether benign or malignant, are the main indication for biopsy, especially in melanoma, which can present as a longitudinal melanonychia at this site. A biopsy is also useful for the diagnosis of inflammatory skin diseases that might be confined to the nail unit (lichen planus, plaque psoriasis), apparent infections or similar conditions (mainly onychomycosis and, very rarely, parasite infestation) in which microbiology has been unable to identify the culprit pathogen; samples can also be of use in the diagnosis and treatment of painful nails.35,36

The nail matrix should be biopsied in a direction parallel to growth. Although the whole nail is often extracted, this procedure is traumatic for both the matrix and the nail bed. Therefore, the decision to take the whole nail should be made carefully37 and sterile conditions, local anesthesia (preferably a local-regional block), and hemostasis measures are required. A digital, periungual, or combined block will be useful, and applying a digital tourniquet can ensure adequate ischemia and facilitate visualization of the surgical field.34 Soaking the finger in soapy water for 10minutes softens the nail, making it easier to remove.37

These biopsies often have artifacts caused by clamps that are similar to those seen on scalp samples. To avoid this problem, the nail can be extracted with special instruments,38 such as delicate Joseph or Gillies hooks. Alternatively, a needle from a syringe can simply be bent and used as a surgical hook.39

Understanding the complex anatomy and physiology of the nail unit is necessary for obtaining optimal results, and trained operators are required for nail matrix biopsies if postoperative nail dystrophy is to be avoided.

Certain changes in nail morphology (hyperkeratosis, dyschromia, onycholysis, pits, clubbing, etc.) can be classified according to the nail zone affected.

The operator should biopsy the nail plate, bed, or matrix (Table 2), as appropriate for the suspected diagnosis. For sampling the nail bed, elevate the nail plate and perform an elliptical biopsy no more than 3mm wide or take 2 punch biopsies. To reveal and biopsy the matrix, first separate a flap of the proximal nail fold; then try to avoid sampling the proximal zone so as to minimize the risk of nail dystrophy.34 The skin surrounding the nail can be sampled like other parts of the integumentary system or by performing a lateral longitudinal biopsy.

It would be helpful to have a register of subungual melanomas that have originated in the nail matrix. Delay in diagnosing melanoma in this location has often been considered to lead to a poor prognosis. Subungual melanomas, which are often found on the great toe, usually present as a melanonychia, but may occasionally take the form of an unpigmented exophytic mass, nail dystrophy, or ulceration. In these situations, a biopsy of the nail matrix will be required for diagnosis and to guide therapy according to tumor thickness.40

It is useful to remember that nail biopsy is the most sensitive diagnostic tool in onychomycosis, as the hyphae will be visible in the sample even in cases where cultures have repeatedly given negative results.41

In 1969 DeCloux and Friederici42 reported seeing similar histopathological changes in skin and nerve tissue biopsies from 6 patients with mucopolysaccharidosis. Even though considerable time has elapsed since then, skin biopsies are seldom performed to study neurological diseases; among the features that can be investigated are the intra- or extracellular deposition of various metabolic byproducts present in these diseases.

Most are metabolic diseases that affect the nervous system, accompanied by biochemical and genetic abnormalities that can be identified during fetal development. The presence of lysosomal deposition, for example, can be detected by electron microscopy.

Skin biopsy can be used to diagnose the neurological diseases discussed in this section (Table 3).

Lafora disease, or lipofuscinosis, is characterized by myoclonic epileptic seizures and is associated with cerebellar syndrome and progressive dementia. Findings include homogeneous perinuclear inclusions called Lafora bodies, consisting of glycogen granules that accumulate especially in the apocrine glands of the skin. A biopsy taken from the axilla, an area rich in apocrine glands, is especially useful.

In the syndrome referred to as CADASIL (from cerebral autosomal dominant arteriopathy with subcortical infarcts, and leukoencephalopathy), electron microscopy reveals granular osmiophilic material around the smooth muscle cells of small arterioles. Such deposits are present in the brain, nerves, and dermis.43

Some peripheral nervous system diseases affect small caliber sensory nerves that are present in the skin, and a biopsy can assist diagnosis.44 In fact, consensus-based guidelines published in January 2010 recommended performing skin biopsy of the distal portion of lower limbs for quantification of nerve fiber density and comparison to age-standardized tables.45 Most publications on this topic consider that small fiber neuropathy can be diagnosed in patients with sensory signs and symptoms when sural nerve conduction findings are normal.

A punch biopsy of 3 to 4mm in diameter and 6 to 8mm deep that includes eccrine glands and hair follicles would be adequate.45 The site to biopsy would be the distal region of the leg, about 10cm above the malleolus or, in some cases, the outside of the upper part of the thigh 20cm below the anterior iliac spine.45 The sample should be frozen immediately and kept at 4°C for 24hours. Later it will be incubated with rabbit polyclonal anti-protein gene product (anti-PGP) 9.5 antibodies in a 2% solution of paraformaldehyde–lysine–periodate45 before cryostat section for study by fluorescence or confocal microscopy, depending on the technique chosen.

Because there is inter-laboratory variability in nerve fiber counts for normal skin, tables standardized for each center must be used and staff must be properly trained to examine the samples and interpret the findings. The technique has not been standardized and only a few specialized laboratories in the world offer it.

Various methods for visualizing and quantifying autonomic nerve fibers have been suggested, and although no preferred method has been established, some authors have reported correlation between neuropathy scales and pilomotor nerve density, even though the density of these nerve fibers does not in turn correlate with that of the small sensory nerve fibers mentioned above.46

The study of peripheral neuropathies has benefited not only from the various methods for visualizing peripheral nerves in the skin but also from techniques for extracting proteins from skin biopsy tissue; in the investigation of the origin of neuropathic pain, this approach facilitates the study of implicated cytokines, which are difficult to visualize by means of immunohistochemistry.47

Dystrophin expression in smooth muscle of the skin, which is to say in the skin's erector muscles, can also be useful in the diagnosis of Duchenne–Becker muscular dystrophy. Immunohistochemistry demonstrates a decrease in or absence of dystrophin expression in these skin muscles even before signs and symptoms of the disease begin to appear.48

Rabies, which is caused by a type virus of the Lyssavirus genus, can also be detected antemortem by DIF complemented by reverse transcriptase polymer chain reaction techniques on skin biopsy material.49 As the disease progresses the virus proliferates in cranial nerve axons and, therefore, in the cutaneous nerves of the face and scalp. As more nerve fibers are present in perifollicular zones, the preference is for a nuchal skin biopsy at least 3mm in diameter.50

The authors of the present article have not yet had direct experience with this type of biopsy, given that the procedures we mention in this section are only handled in a few specialized centers.

Prenatal skin biopsies can be taken under ultrasound guidance during the second trimester and analyzed by immunohistochemistry or electron microscopy.51–53

The inheritable skin diseases that are most often diagnosed prenatally are congenital epidermolysis bullosa, oculocutaneous albinism, and some ichthyoses.

Conventional histopathology is particularly useful when the changes caused by the disease are known but the responsible genetic mutation has not been determined.51 There are risks, however. Between 1% and 3% of fetuses are lost and there may be leakage of amniotic fluid or fetal scarring.51 These methods can also be used to study enzyme defects or keratin gene mutations.52 Chorionic biopsy and amniocentesis, which allow us to study fetal DNA, are also possible.52

When in vitro fertilization is used, diagnostic studies can also be performed before implantation of an embryo.53 DNA analysis of blastomeres from an embryo of at least 6 cells can identify a healthy embryo for transfer to the uterus.

Finally, karyotyping can be performed on keratinocytes isolated from skin biopsy material. This technique is useful in Ito hypomelanosis, linear and whorled hypermelanosis, and McCune–Albright syndrome, in which karyotyping of lymphocytes may not show abnormalities.51,54

Another step forward would be to use a less invasive procedure to diagnose certain diseases in which the fetus produces nucleated erythrocytes that appear in small numbers in the mother's blood.53

The number of fetal skin biopsies required has declined thanks to recent characterizations of the genetic defects that cause a large number of these diseases; amniotic fluid or chorionic villus samples can now be used to obtain fetal DNA for analysis.51

In spite of progress in this field, there are important social, political, and ethical problems to resolve as more robust diagnostic techniques become available. A multidisciplinary team—including a pediatrician, a geneticist, a gynecologist, and a dermatologist—is therefore needed and information must be fully and correctly conveyed to the parents.

As dermatologists we should know that the main indications for prenatal diagnosis are the presence of genetically determined diseases in the family in general or in older siblings. Adequate study requires taking samples from both parents and from family members affected by a disease in order to characterize or rule out de novo mutations, uniparental disomy, or stem cell mosaicism.51