Cytologic 8 Diagnosis of Herpes Simplex Virus Infection Zuher M. Naib. M.D.
From the Department of Pathology, Emory University School of Medicine, Atlanta, Georgia
Alterations in epithelial cells infected with herpes simplex viruses (HSV) have been described repeatedly’-7 since Tzank,B in 1947, advocated the use of cytology for the diagnosis of these infections. In spite of successful use of this technique in the following years by enthusiasm for this diagnostic Blank and his collaborators,gJo method gradually waned because of the emergence of other, more sophisticated, laboratory tests for viral diagnosis (culture, serology, antibody determination, animal inoculation, electron microscopy, etc.).” For the last 20 years, with the use of the Papanicolaou staining method, cytology has proven to be a fairly reliable, rapid, and simple method of diagnosis of HSV infections. The Pap stain gives better nuclear detail of the alcohol-fixed cells than the original Tzank smear, which was air-dried and Giemsa stained. This cytologic diagnosis has been and continues to be especially significant in detecting cervical herpetic infections through the use of Papanicolaou smears for cancer screening of asymptomatic women.12 Scraping Techniques The scraping of the dome of an intact herpetic vesicle will contain only normal squamous cells. These are seldom diagnostic of the infection. Similarly, the aspiration of the fluid from the vesicle will contain only some histiocytes, lymphocytes, and a few epithelial cells, too degenerate for proper morphologic interpretation. The vesicle should be denuded and the margins, rather than the bottom, of the newly created ulcer, should be energetically scraped, preferably with a sharp plastic or metal curette (Fig. 1). In cases of older, encrusted lesions, the crust should be removed by soaking under a dripping wet compress, which should be left over the lesion for at least 30 minutes. After removal of the crust, the margins of the small ulceration can then be scraped. Smearina Techniaues Before scraping the lesion, in order to prevent the cells from air drying, a small ring is drawn on a slide with a wax pencil to create a
FIG. 1. The scraping produces a satisfactory
of the surface of the dome (A) or the aspiration of fluid from the vesicle (B) seldom cytologic specimen. The margins of a fresh ulcer(C) should beenergeticallyscraped.
shallow well. The slide is then covered with 95% ethyl alcohol. The scraped material is spread rapidly and mixed with the alcohol contained in the ring. The alcohol is allowed to evaporate slowly from the surface of the slide, which should be maintained in a horizontal position. After a few minutes, the almost-dry slide is sprayed or dropped gently into a fixative. The presence of the ring prevents loss of the cells, which may occur if they are scattered over too large a surface. It also helps the pathologist in locating the cells. The smears are later stained with Papanicolaou stain. If avery rapid diagnosis is needed, the smear could be covered with methylene blue or any other vital stain, cover-slipped while wet, and immediately examined for the large, multinucleated cells diagnostic of HSV. Cellular Morphology The HSV-infected cells almost always show a combination of viral and degenerative changes. Their morphology may vary slightly, depending on the type and age of the infection and on other factors, such as hormonal influences, the nature of the infected cells, viral strain, etc. The sequential cytologic manifestations that correlate well with those observed in infected tissue culture cells are as follows: 1. The virus seems to have a special predilection for the immature cells (squamous parabasal, columnar reserve, melanin-
containing basal cells of the skin, etc.), although the infection may eventually spread to the mature epithelial cells. The first manifestation of the infection is usually a nonspecific nuclear hypertrophy (from 2 to 4 times, Fig. 2A). The frequent appearance of a thin, perinuclear halo in the processed cells results from the shrinkage of the nucleus during fixation and processing. Similarly, the nucleoli, when present, appear to enlarge for a short period and become granular, probably as a result of the enlargement of their components, the nucleolini or deoxyribonucleoprotein (RNP) bodies. The nucleoli may first appear finely vacuolated, then rapidly scatter into small granules and finally, become invisible. Some of these nucleolar particles and nuclear chromatin granules move to the periphery toward the inner nuclear membrane, making it appear thicker and more prominent. Almost simultaneously, the cytoplasmic ground substance loses its normal thin, transparent appearance. It becomes denser and more opaque and often stains deep purple or blue-green with the Papanicolou stain (Fig. 2B). Some of the cytoplasmic changes are probably related to increased protein and RNA synthesis. Other cytoplasmic changes (vacuolization, ill-defined borders, etc.) are due to early cellular degeneration. With the disappearance of the nucleoli, a progressive increase of the RNP bodies fills the center of the nucleus, giving it a
April-June 1984 Voldme 2 Number 2
of HSV Infection
FIG. 2. Early stage of herpetic infection: A (left). Enlargement of the nucleus and nucleoli (X 1200). B (right). Increased density of the cytoplasm and the beginning of multinucleation (X 1200). C (left). Multinucleation. Note the thick nuclear membrane and the bland appearance of the nuclei, which vary in size and mold against each other (X 1200). D (right). Early inclusion. Note the small particles packing the central nucleus and the coarsely granular inclusions in the other cells (X 1200).
bland look. This also further displaces the nuclear chromatin to the periphery, where it adheres to the inner nuclear membrane, making it appear even thicker. Multinucleation (4-15 nuclei) then occurs in some of the infected cells. The nuclei vary in size and shape and usually mold against one another (Fig. 2C). 5. The centers of these enlarged nuclei often appear to be packed with small, acidophilic particles (l-2/1), giving them a distinct, homogenous, granular, ground-glass texture when observed under low-power magnification. These particles are formed by clusters of masses of RNP, which can only be visualized under oil-immersion
magnification (Fig. 2D). 6. Eventually these fine granules condense and move toward the center of the nuclei to form the typical, single, coarsely granular, acidophilic intranuclear inclusions (type A inclusion) (Fig. 3A). These inclusions are rapidly surrounded by more or less prominent halos. The coarse granularity of these inclusions is distinctive and permits differentiation from other viral infections with type A intranuclear inclusions. Thus, in adenovirus infections the inclusions are amorphous and in cytomegalo viral infections the inclusions often have a mosaic-like texture. 7. The infected cells lose most of their adhe-
FIG. 3. Late stage of herpetic infection: A, 6 (left, right above). Note the single, coarsely granular, acidophilic intranuclear inclusions in the infected cells (X 1200). C, D (left, right below). Early and late stage of cellular degeneration and death of the infected cells. Note the cytoplasmic vacuolization and terminal fuzziness (X1200).
sion to the surrounding noninfected cells and tissue (Fig. 3B). This explains the usual abundance of infected cells in smears, even when the lesion is very small, and their scarcity in tissue sections. 8. The degeneration of the infected cells manifests itself by an increase of cytoplasmic and nuclear vacuolization (Fig. 3C), and irregularity of the shape of the cytoplasm, followed by fuzziness and breakage of the cytoplasmic and nuclear membranes (Fig. 3D). The death and final breakage of some of the infected cells results in an increase in the amount of debris, protein deposits, inflammatory cells, and histiocytes usually found in the background of smears obtained from older lesions.
Differential Diagnosis Contrary to previous reports,‘3 cytology cannot differentiate a primary from a recurrent herpetic infection. The bland nuclei (thought to be an indication of primary infection) and the inclusion-containing nuclei (thought to be an indication of secondary infection) may be found in scrapings from both primary and secondary lesions. The bland nuclei indicate an early stage of HSV infection and the inclusion-containing nuclei a later stage. Similarly, cytology cannot differentiate type 1 HSV infection from type 2. The cellular changes are identical. The multinucleated infected cells in the
April-June 1984 Volume 2 Number 2
of HSV Infection
FIG. 4. A (aboveleft). Scraping of a varicella skin lesion. Note the inclusion-containing small, single, infected cells (X 800). B (above right). Mutlinucleated endocervical syncitial cells. Note the persistence of the nucleolus and the chromatin pattern that excludes HSV infection (X 1200). C (below left). Langhan’s type giant cells with peripheral nuclei and absence of nuclear molding (X 1200). D (below right). Small foci of HSV-infected cells in a tissue section. Note the loose cells with the diagnostic enlarged bland nuclei (X 400).
scrapings of herpes simplex and varicellazoster viral infections are also identical. They cannot be used to differentiate these two lesions; however, the scrapings of varicella type infections are most likely to produce, in addition to the multinucleated cells, a large number of single, small, round cells with scanty cytoplasm and central, single nucleus containing the typical herpesvirus inclusion (Fig. 4A). The HSV-infected multinucleated squamous or columnar cells must also be differentiated from other forms. endocervical cells may show Multinucleated thick nuclear membranes and nuclear molding similar to that observed in herpes infec-
tion (Fig. cytoplasm appearance chromatin
4B). In contradistinction, their remains transparent and lacy in and the nucleoli and granular remain distinct. Multinucleated “repair” cells, associated with the repair or healing reaction of injured tissues, often show a heavy nuclear membrane and dense cytoplasm with pseudopodlike projections similar to those found in the late stage of herpetic infection; however, the nuclei often overlap on each other, in contrast to the molding observed in HSV-infected cells. Their chromatin also remains coarsely granular, nucleoli are often prominent, and type A inclusions are not seen. Foreign
body and granulomatous
Clinics in Dermatology
ban’s” type giant cells are differentiated by the presence of ingested debris in the cytoplasm, the location of the nuclei in a semicircle around the periphery, and the absence of nuclear molding (Fig. 4C).
Specificity and Sensitivitv In smears of good quality, the cytologic diagnosis of herpesvirus infection is usually easy and very specific. The infected cells almost always stand out distinctly because of the density of their cytoplasm, their large size, and the characteristic molding of their nuclei. Their presence definitely indicates a current HSV infection. The specificity of the diagnosis can be further tested if needed with specific fluorescence or peroxidase staining’s of the smears. The clinical diagnosis of HSV infection, especially of older, encrusted lesions, appears to be more difficult. Of 100 consecutive satisfactory scrapings of genital and nongenital lesions strongly suspected of being herpesvirus infection, cytology confirmed the clinical impression in only 42 cases. Candida fungal infection was diagnosed in six, molluscum contagiosum in two, and lymphogranuloma in one. The rest of the smears showed only nonspecific inflammation and repair activities. Viral cultures for HSV were positive in only 3 of such lesions. The sensitivity of cytology depends largely on the technique used to obtain the cells. No amount of skill and experience will enable the cytologist to reach a diagnosis from a poorly obtained or fixed cellular sample. He does a disservice to the patient if he attempts to interpret such a smear. About 20% of the smears received in our laboratory to rule out HSV infection are found to be unsatisfactory either because the cells are too scanty or they were allowed to air dry. After proper instruction, a repeat smear is recommended in such cases. The sensitivity of the cytologic method also depends on the age of the lesion. As reported
in one of our previous studies,‘2 in patients as well as in experimental animals, cytology correctly diagnosed 85% of virologitally proven (positive viral culture) HSV infections when the lesions were less than 4 days old. This percentage decreases rapidly in older lesions, except when the patient is pregnant. The lesions in these cases seem to continue to exfoliate diagnostic, multinucleated cells for a greater length of time. Although the cellular changes in both histologic sections and the smears are identical, the biopsies of suspected lesions were half as diagnostic as the scrapings. In the sections, the characteristic herpetic cells are few and are mainly located at the edge of the ulceration. They are usually loose and unattached to the surrounding tissue (Fig. 4D). Consequently, they are easily lost during the cutting of the paraffin blocks.
Conclusion The higher frequency of reporting genital herpes and the awareness of the general population of the existence of this infection increase the need for a simple and specific diagnostic test that requires a minimal amount of time. Scrapings from suspected lesions, which can be stained quickly with methylene blue or with Papanicolaou stain, seem to fill some of this need. The cellular changes are easy to interpret and are fairly diagnostic. Although viral culture is still the most sensitive means of diagnosing the infection, this procedure usually takes several days. Furthermore, viral culture is not available in many areas because of the need for special laboratory apparatus and personnel. Although simple and primitive by comparison, a positive cytologic finding definitely confirms the existence of a current herpesvirus infection. A negative cytologic finding does not necessarily rule out the disease, and other, more time-consuming, procedures should then be employed.
April-June 1984 Volume 2 Number 2
features of viral 1963:7:295-g.
JH, Bingul 0, Burgoon CB. Cytodiagnosis of inflammatory dermatoses. Pitfalls in evaluation of smears. Arch Dermatol. 1963;87:18-27.
2. Lynch P, Welsh 0, Naylor
B. Cytodiagnosis viral lesions. Cutis. 1970;6:79-83.
3. Naib ZM. Exfoliative
cytology of viral cervicoActa Cytologica. 1966;10:126-9.
ZM. Exfoliative cvtologv in the rapid diagnosis of herpes simplex v&s infection.in: edited by Nahmias A, Dowdle W, Schinazi R, eds. The Human Herpesviruses. New York: Elsevier, 1981:381-6.
ZM, Nahmias AZ, Josey WE, Kramer JH. Genital herpetic infection. Association with cervical dysplasia and carcinoma. Cancer. 1969;23:940-5.
6. Naib ZM, Nahmias
and histopathology infection. Cancer. 7. Stern
Cytoloaic Diagnosis of HSV Infection
AZ, Josey WE. Cytology of cervical herpes simplex 1966;19:1026-32.
E, Longo LD. Identification simplex virus in a case showing
of herpes cytological
Address for correspondence: Zuher Street, S.E.. Atlanta, GA 30303.
immediat en *. Tzanck A. Le cytodiagnostic dermatologie. Bulletin de la Societe Francaise de Dermatologie and Syphiligraphie. 1947; 7168-76. 9. Blank H. Virus disease affecting the skin. A summary of recent advances. Acta Derm Venereol. 1949;29:77-83. ._ IU. Blank H, Burgeon CF, Baldridge GD, McCarthy PL, Urbach F. Cytologic smears in the diagnosis of herpes simplex, herpes zoster, and varicella. JAMA 1951;146:1410-2. antibody 11. Griffin JW. Fluorescent herpes simplex lesions and recurrent Oral Surg 1963;16:945-52.
study of aphthae.
12 Nahmias A, Naib Z, Josey W, Clepper A. Genital herpes simplex infection: virologic and cytologic studies. Obstet Gynecol. 1967; 29:395-400. 13. Ng AB, Reagan JW, Linder E. The cellular manifestations of primary and recurrent herpes genitalis. Acta Cytologica. 1970;14:124-9.