Relation of Cytohistopathology of Genital ... - Cancer Research [PDF]

[CANCER RESEARCH 33, 1452-1463, June 1973]

Relation of Cytohistopathology Cervical Anaplasia1

of Genital Herpesvirus Infection to

Zuher M. Naib,2 AndréJ. Nahmias, William E. Josey, and Saleh A. Zaki Department of Pathologv, Cvlologv Division [Z. M. N., S. A. Z.], Departments of Pediatrics and Preventive Medicine [A. J. N.\, and Department of Gynecology and Obstetrics [W. E. J.], Woodruff Medical Center, Emory University School of Medicine, Atlanta, Georgia 30322

Since 1959, we havenotedcellularchanges suggestive of herpesvirus infection during the screening of routine cervicovaginal smears submitted for cancer detection. Such morphological changes in Giemsa-stained smears from scrapings of herpetic skin vesicles were described by Lipschutz in 1921 (11) and, later, by various dermatologists (3). Nevertheless, early cytopathologists misdiagnosed the cha racteristic inclusion-bearing giant cells in cervical smears as other types of cells, such as trophoblastic elements (10). It was not until 1963 that the herpetic etiology of these cellular changes was definitely demonstrated by Stern and Longo (30), who isolated herpes simplex virus from the genitals of a woman showing the cytological features of herpetic infection in a routine cervicovaginal smear. About the same time, we began our virologica!, clinical and epidemiological studies on genital herpes. One of our first objectives was to validate the specificity of the morpho logical changes by means of virological studies of women with such cytological findings in their smears. Parallel studies were also performed in mice genitally inoculated with HSV3 (18). We found that women with cytological cervicovaginal changes consistent with herpetic infection had positive isolation for HSV, when virological cultures were obtained within a few days of the cytological interpre tation. Table 1 summarizes the older mouse studies and updates the human studies, comparing the results of cyto logical and virological examinations performed at the same time. It was found that the cytological technique was almost completely specific for HSV. On the other hand, the sensitivity of the cytological technique, as compared with the virological method for detecting genital HSV infection, is of the order of 65 to 75%. In view of the above findings, we felt more confident of the association we found between cytologically detected genital herpetic infection and cervical anaplasia ("anaplasia" is used herein to denote cervical dysplasia, in situ carcinoma, or invasive carcinoma) (23). To the best of our knowledge, that report was the first to propose that genital HSV was possibly a cervical carcinogen, although other 1Presented at the American Cancer Society Conference on Herpesvirus and Cervical Cancer, December 8 to 10, 1972, Key Biscayne, Fla. This research was supported by grants from the American Cancer Society and the National Cancer Institute, USPHS. 2Presented by. 3The abbreviations used are: HSV, herpes simplex virus; HSV-2, herpes simplex virus type 2; RNP. ribonucleoprotein; HSV-1, herpes simplex virus type 1.

1452

workers had suspected a nonspecified, venereally transmit ted agent on the basis of epidemiological studies of cervical cancer (12, 28, 29). At the time we first observed the association of genital herpes with cervical anaplasia in 1964, it had not yet been demonstrated (a) that genital herpes is most commonly caused by HSV-2, a different virus from that associated with most nongenital herpetic infections (15); (b) that the cervix is the principal site of infection in women (9); (c) that HSV-2 is most often spread venereally (16); or (d) that so many other viruses in the herpes group are associated with cancers in humans and animals (4). Table 2 presents our earlier observations (23) on the association of genital herpes and cervical anaplasia in 62 women (16.1%), our later studies (24) on 245 women (23.7%), and our most recent findings in a study of 673 women (15.6%). This rate of cervical anaplasia is several times higher than that found in our general hospital population (24) or that noted in most populations studied by others, the rate being approached only in women attending venereal disease clinics or those in prison (6, 14). Table 3 presents the prevalence of cytologically detected herpetic infection, as well as the rates of cervical anaplasia diagnosed cytologically or histologically, as noted in pub lished reports. Our comparative studies of HSV detection by cytological and virological methods (Table 1) suggest that the prevalence rates of cytologically detected genital herpes noted in this report probably represent only twothirds of the actual cases of genital HSV infection. This summary table emphasizes the differences in prevalence rates found in various populations; e.g., the low rate (0.02%) in private hospital or clinic patients, compared to the higher rate (0.26 to 0.31%) found in indigent patients, and the highest rate (3.3%) noted in women attending venereal disease clinics. Another important variable is the age of the population screened. Our median age, for instance, is 22 years. It is also worth mentioning that, in 2 hospitals, the rate found in pregnant women was about 3 times greater than that in nonpregnant women (17, 25). Table 4 reviews 105 cases of histologically diagnosed cervical anaplasia in relation to the time at which the genital herpes was cytologically detected. Note that 23 cases of cervical anaplasia were found prior to the first herpes de tection, 47 within the same year, and 35 after the first year of herpes detection. We now present in greater detail the cytological struc tural changes associated with genital herpetic infection, as well as the pitfalls in cytological diagnosis. We also describe CANCER

RESEARCH

Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

VOL. 33

Cervical Anaplasia Table I

Comparison of virologica! and cytological methods to identify genital HSV infection in cervicovaginal specimens obtained from humans and mice Cellular changes in mice"Negative1043Total3949 findingsPositiveNegativePositive593CWomenNegative31»84Total9087Positive296Female

Total

62

115

177

35

53

88

" Experimentally infected genitally with HSV (18). *Twenty-one had prior positive cytological findings. cTwo had prior positive virological findings. Table 2 Association of cytologically delected genital herpes and histológica/!}-confirmed cervical anaplasia No. of women with Cervical anaplasia situcarcinoma41225Invasivecarcinoma0411Total1058105%"16.123.715.6 Period1963-1965"1963 1968r1963 1972Genitalherpes62245673Dysplasia64269In " With genital herpes and cervical anaplasia. 6Naib et al. (23). c Naib et al. (24).

the histological changes associated with cervical herpes, reported in less detail by ourselves earlier (23) and by others (26). Cellular Structural Changes of Genital Herpetic Infection

Different stages of the characteristic cellular changes of herpetic infection usually will be seen in any single smear. This is due to the fact that the exfoliation of the infected cells can occur at different stages and because the response of various cells to the infection varies according to their type, age, and other factors, such as hormonal influences. The sequential cytological manifestations of cervical herpesvirus infection, which correlate well with those ob served in tissue culture cells of cervical epithelium (32) or other human tissues (31), can be summarized as follows. 1. The virus appears to have a predilection for the more immature or altered cells. The normal squamous basal or parabasal cells and the young cuboidal reserve cells of the endocervical glandular epithelium appear to be the main targets of the virus. 2. The first visible light-microscopic change is the hyper trophy of the cytoplasm and nucleus of infected cells. This increase in cell size, from 2 to 15 times that of normal cells, probably represents fluid displacement resulting in nuclear and cytoplasmic edema. This is indicated by the frequent appearance of the perinuclear halo, as a consequence of the shrinkage of the enlarged nucleus and cytoplasm due to the dehydration of the cell during fixation and staining (Fig. \A). 3. The nucleoli also enlarge early for a short period and

become granular, probably as a result of the enlargement of their components, the nucleolini or RNP bodies. With the expulsion of these RNP bodies into the nucleoplasm, the nucleoli become vacuolated, then scatter into debris and are no longer visible. 4. Simultaneously, marked disturbance of the cytoplas mic ground substance occurs. From its normal thin, trans parent, delicate appearance, the cytoplasmic ground sub stance becomes dense and opaque and stains deeply purple or blue-green with the Papanicolaou stain (Fig. 2/1). These changes are probably related to the increased protein and RNA synthesis occurring in infected cells. 5. With the disappearance of the nucleoli, there is a progressive increase of RNP bodies, also sometimes called B-bodies, to differentiate them from A-inclusion bodies which develop later (Fig. 2A). These B-bodies occupy the center of the nucleus and displace the nuclear chromatin to the periphery, where it becomes adherent to the inner nuclear membrane. This results in the central nuclear clearing and thickening of the nuclear membrane. 6. Multinucleation (4 to 15 nuclei) is either a conse quence of amitotic division of the infected cells that have lost their capacity for normal mitotic division or is due to cell fusion resulting from alterations in the composition of the cytoplasmic membrane. This is indicated by the marked molding and variation in size and shape of the nuclei. (Fig. 3/1). Early degenerative changes that consist of minute cytoplasmic vacuolization and fuzziness of the cytoplasmic and nuclear membranes can also be demonstrated at this stage. 7. The enlarged pale nuclei appear to be packed with small acidophilic particles (1 to 2 ¿/m)so that a distinct, 1453

JUNE 1973 Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

Naib, Nahmias, Josey, and Zaki homogeneous, granular ground-glass texture can be ob served under low-magnification examination. These parti cles are probably formed by clusters of masses of deoxyribonucleoprotein and can be readily visualized under oil-immersion high-power magnification (x 1200) (23). 8. The granules appear to condense and move toward the center of the nuclei to form the typical, single, coarsely granular, acidophilic intranuclear inclusions (A-inclusion), soon to be surrounded by a more-or-less prominent halo (Fig. 4/1). The granularity of these inclusions is distinctive and permits differentation from other viral infections with intranuclear inclusions. Thus, in the case of adenovirus infection, the inclusions are amorphous (22) and, in the case of cytomegaloviral infection, the inclusions often have a mosaic-like texture (20). 9. The infected cells separate from the surrounding normal cells and exfoliate rapidly. This may help to explain the usual abundance of infected cells in a cervical smear, even when the gross lesions are very small. 10. Degenerative changes then occuj in the form of vacuolization of the cytoplasm and nuclei (ballooning changes), bizarre shapes, various formations (long tails or Prevalence rales of cytologically

streamers), and fuzziness and breakage of the cytoplasmic and nuclear membranes. The death and rupture of the infected cells and the resulting epithelial ulcérationare associated with the increased amount of debris, protein deposit, inflammatory cells, and histiocytes that are usually found in the background in smears obtained from older herpetic cervical lesions. Cytological Differential Diagnosis

The HSV-infected squamous or columnar cervical cells need to be differentiated from (a) multinucleated endocervical cells, (b) repair cells, (c) trophoblastic cells, (d) anaplastic cells, and (e) foreign-body giant cells, which may be found in routine cervical smears, as follows, (a) The normally multinucleated endocervical cell may show nuclear molding similar to that observed in an early herpetic infection but, in contradistinction, its cytoplasm remains transparent and lacy in appearance, and the nucleoli and granular chromatin remain distinct (Fig. 5A). (b) The squamous or endocervical "repair" cells, associated

Table 3 delected genital herpelic infection and associated

cervical anaplasia

withcervical

withgenitalherpes718372342220167673%°3.30.5-1.50.030.090.310.020.09 anaplasiaand genitalherpesNot AuthorsBeilby

screenedLondon,

(2)Nowakovsky et al.

England; venereal dis clinicChicago; ease statedHalifax, not Canada: routine screen clinicsNew ing York City; planned parent clinicsCleveland; hood patients'Private indigent

(26)And)Wolinska et al. (33)Ng

and Melamed

(25)Jordan et al.

statedNot

stated54,00043,33175,184119,85322,21421,810643279,587No." stated2"2»IS6'"6"4"105e%II5.422.56

patientsNew ofvarious Mexico; indigent patients groups:IndianCaucasianNegroAtlanta0; ethnic

(7)Naib el al.

et al. (this series)Population

indigent

patients'No.screened209Not

"On the basis of a sensitivity of Cytological detection of genital herpes of around two-thirds (Table I), these numbers could be multiplied by 1.3. * By Cytological examination only. ' Rate in pregnant women was 3 times higher than that in nonpregnant women. "Cervical anaplasia was detected before genital herpes in 18 of 80 women (22.5%); in 4 of 80 women (5%), the anaplasia did not regress (those 4 were detected by histological examination). ' By histological examination. Table 4 Time relation between Cytological detection of HS y and histological diagnosis of cervical anaplasia anaplasiaType of cervical anaplasiaDysplasia In carcinomaInvasive situ cancerTotalLess

1454

No. of cases of cervical than 1 yr 1 yr of prior to HSV detectiong7823WithinHSV detection36 8347More

of cervical ana plasia in 673 women with genital herpes10.2

than 1 yr after HSV detection25 10035Total692511105%

3.31.515.6

CANCER RESEARCH

Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

VOL. 33

Cervical Anaplasia

with the repair or healing reaction of injured tissues, may show the heavy nuclear membrane and the dense cytoplasm with pseudopod projections similar to those found in the late stage of herpetic infection (21). However, the nuclei often overlap each other, in contrast to the molding observed in HSV-infected cells. The chromatin of the repair cells remains coarsely granular, and their nucleoli are often retained and prominent, (c) The giant multinucleated trophoblastic cells (Fig. 5B), which were confused for a long time with herpesvirus-infected cells (10), have a similar dense cytoplasm and occasional nuclear molding. However, their cytoplasmic borders are smooth, distinct, heavy and sharp, instead of appearing fuzzy, and their nuclei retain coarse chromatin granularity with occasional prominent nucleoli, (d) Anaplastic cervical or endometrial cells may resemble herpetic cellular changes by their nuclear molding, thickened nuclear membrane, and irregular cytoplasm. However, the chromatin remains coarsely irregular with occasional prominent abnormal nucleoli (Fig. 5C). (e) Foreign-body giant cells can be differentiated from HSVinfected cells by the presence of ingested debris in their cytoplasm, the location of the nuclei in a peripheral semicircle, and the absence of nuclear molding (Fig. 5D).

predominant cells are lymphocytes and plasma cells. Al though sometimes there is a tendency to perivascular lymphocytic orientation, this is not a constant feature (Fig. 2ß).Relatively few neutrophiles are associated with the erosion. The characteristic herpetic cells are present mainly at the edge of the erosion but can also be found as unattached, loose cells in the midst of inflammatory and cellular debris (Fig. 4B). In some cases, multiple small herpetic lesions in nonulcerative areas can be noted, apart from the main ulcerative lesion. In the presence of squa mous anaplasia, the foci of herpetic infection are located at the surface of the atypical epithelium; these foci are surrounded by noninfected anaplastic cells (Fig. 6). In contrast, viral infection of the normal squamous epithelium involves the basal immature layer only. Although in a review of the histological slides, the location of the herpetic lesions was sometimes difficult to ascertain definitively because of the size and orientation of the blocked tissue, a significant preponderance of lesions was found to be localized at or near the squamocolumnar junction. Thus, in 31 of the 45 specimens, the main location of the herpetic involvement was at the squamocolumnar junction. In about one-half of the specimens, there were scattered foci of herpetic lesions, each made up of a few infected cells located in the endocervical and squamous Histológica! Changes of Cervical Herpetic Infection epithelium. In 8 specimens, the lesions were found in the endocervical Of the 673 patients in this study who had cytological epithelium only and, in 3 other specimens, the lesions were evidence of genital herpes, 126 had cervical punch biopsies, noted only in the basal layers of the squamous epithelium of cone biopsies, or hysterectomy specimens that were ob the external os. In 2 cases, the characteristic herpetic cells tained simultaneously or within a period of 5 weeks after or were present in minute fragments of tissue separate from the before the cytological specimen was taken. The histológica! main biopsy specimen, making it difficult to determine the specimens were routinely fixed in formalin, and slides were exact location of the herpetic lesion. stained with hematoxylin and eosin. Whereas the diagnosis Of the 81 histological specimens in which no herpetic of herpetic infection was reported originally in only 9 cases, lesions could be demonstrated, the majority showed chronic upon reexamination, we found that a total of 45 cases had nonspecific cervicitis, 21 showed evidence of granulation characteristic herpetic lesions. In 4 biopsies, the characteris tissue, and 11 demonstrated a nonspecific ulcération.The tic herpetic lesion was revealed only after deeper cuts were study of multiple deeper sections of the paraffin tissue made into the paraffin block and were not available to the blocks will be done and will be the subject of a later report. original examiner. Table 5 summarizes the results of histological studies that In almost all of the tissues showing herpetic changes, the revealed herpetic infection and/or cervical anaplasia in 126 lesions were ulcerating, and "intact" vesicles could not be specimens obtained within 5 weeks before or after (or at the demonstrated. Cellular criteria for the histological diagnosis same time as) the cytological diagnosis of genital herpetic of cervical herpetic infection are similar to those described infection. Of the 45 biopsy specimens showing herpetic previously for diagnosis in cytological smears. The early changes histologically, 43 were obtained within 7 days of the changes are characterized by a basal cell hyperplasia, with 5 time a cytological smear was taken that was positive for to 10 layers of reactive basal squamous or reserve cells and herpes. In one case, characteristic herpetic cellular changes prominent nucleoli (Fig. IB). In most cases, it is difficult to were noted within an area of invasive squamous cell detect the characteristic herpetic cells by means of low- carcinoma, apparently in malignant squamous cells (Fig. 6). power microscopic examination. In the presence of ulcéra It was interesting to find single cells scattered in the tion, the herpetic lesions are often detected by examination endocervical columnar epithelium, either in the endocervical of the edges of the ulcerated areas. The lesions thus appear lining or in the endocervical glands, which showed\unusual to be formed essentially by superficial erosion. The base of changes. The cytoplasm appeared more dense, and the the erosion is made up of fibrinous proteinaceous material nuclei (often binucleated) were reactive, enlarged, and containing scattered necrotic inflammatory cells (Fig. 4B). slightly hyperchromatic (Fig. 7). No inclusions were noted The depth of the necrosis varies beneath the erosion. An in these cells. Of the 81 cases, 36 showed these single, active proliferated endothelial cell lining is found in the altered, endocervical cells; they were not found in cervical underlying vascular areas (Fig. 3ß).A variable but often biopsies of noninfected cases. heavy chronic inflammatory cellular reaction can be noted One case (not included in Table 5) was detected in which around the erosion and deeper into the necrotic base. The the cytological examination did not reveal herpetic changes, JUNE 1973 Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

1455

Naib, Nahmias, Josey, and Zaki —¿

>Sì"

o

—¿ o

whereas a biopsy taken at the sar.ie time showed a herpetic lesion that was nonulcerating. The herpetic involvement in this specimen was superficial in location and was noted in anaplastic squamous epithelium at the squamocolumnar junction.

—¿

> E » 5

Discussion

È2 ë.

-j I

o

—¿ o

—¿

—¿ —¿ rg

a E •¿~ | "••• ^"

D C^

O

--

eu

x

E-

00

u

—¿

—¿

X

o g. Z

e-

8 U O. ,

E 'ob

1456

E o

Papanicolaou smears offer a rapid method for diagnosing cervical herpes simplex virus infection. Although infection with varicella-zoster virus produces similar changes, isola tion of this virus from the cervix has not been reported. Cytomegalovirus has been recovered from the cervix, particularly in the latter part of pregnancy, but the cytological changes can be readily differentiated from those pro duced by HSV (21). The cytological technique appears to be two-thirds as sensitive as virological isolation methods (Table 1). Never theless, because of the rapidity of cytological tests (within 30 min) and the fact that Papanicolaou smears are being widely obtained for cancer-screening purposes, they are particu larly useful in detecting cervical herpetic infection, which is asymptomatic in most cases (8, 9). The rate of detection in part reflects both the awareness of cytotechnologists and cytopathologists of the various cellular morphological changes induced by HSV infection and their ability to differentiate them from multinucleated endocervical, repair, trophoblastic, anaplastic, and foreign-body giant cells. In large part, however, as demonstrated in Table 3, the detection rates are greatly influenced by the type of population being screened, a reflection in turn of the venereal mode of transmission of most genital herpetic infections. The exact duration of the cytologically detectable infec tion in the genital tract is hard to determine for several reasons. The time of onset of the infection before its cytological detection is usually unknown, and there is also the possibility of negative intervals between 2 periods of positive findings, in the case of recurrent infections. Except in immunologically compromised women, the infection is most often self-limiting. The herpetic lesion usually heals within 2 or 3 weeks, leaving no residual evidence. However, during pregnancy, genital herpetic infection tends to persist for longer periods, particularly if the infection is a primary one ( 17). This may explain in part why cytological detection of HSV is more frequent in pregnant than in nonpregnant women (17, 25). However, it is also likely that the pregnant state itself induces more frequent recurrences. In our experience, genital herpes is more often detected cytologi cally during the postovulatory period. Unlike other workers (21), we have not been able to distinguish primary from recurrent infections by cytological examination. Neither has it been possible for us to distin guish cytological differences between HSV-2 infections and HSV-1 infections. The latter type causes approximately 5% of cervical infections. The histological diagnosis of cervical HSV infection appears to be more difficult than the cytological detection of the infection. As evidenced by our findings, unless special CANCER RESEARCH

Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

VOL. 33

Cervical A naplasia emphasis is placed on the examination, the diagnosis can be overlooked. This is especially true if only low-power screening is used. The few herpetic cells, often unattached in the midst of inflammatory cells and debris, are mainly found at the edges of cervical erosion. Examination of these areas by high-power microscopy permits the detection of characteristic cells. When a superficial erosion is visualized (especially one localized at the squamocolumnar junction), and when neutrophile reaction is mild or absent, deeper cuts into the block will probably reveal more lesions which would otherwise have been missed in initial sections. The chance of diagnosing herpetic lesions on histological examination is much less if the specimen is obtained one or more weeks after onset of the infection. The best cor relation between cytological and histological observations occurs when both types of specimens are obtained at the same time (Table 5). Our finding that the most common location of cervical herpetic lesions is the squamocolumnar junction, an area where most cervical cancers originate, would favor the genital herpes-cervical anaplasia hypothesis. However, the review of other series (Table 3) does not indicate the same high rate of association between the infection and cervical neoplasia. This may in part be explained by the small numbers of reported infections in most other series. Our high rate of cervical neoplasia in women with genital herpes compares more with that noted in women who are in prison (14) or who attend venereal disease clinics (6). These data, therefore, are compatible with either of 2 hypotheses: (a) that genital herpes plays a causal role in cervical carcinogenesis or (b) that both entities most often occur in similar populations of promiscuous women. Another observation (Table 4) was that herpetic infection was detected cytologically around the same time, or one or more years after, cervical anaplasia was diagnosed. In addition, although herpetic changes could not be observed in cytological smears in cervical anaplasia cells, when the 2 types of cells were present in the same specimen (23), viral changes could be demonstrated in histological sections containing anaplastic cells (Table 5). This difference may be attributed to the difficulty in recognizing cytologically the anaplastic changes in herpes-infected cells. The presence of foci of cells showing herpetic changes in the midst of solid tumor strongly suggests that the altered neoplastic cells were infected by the virus. These observations might be interpreted as reflecting the more ready susceptibility of cervical anaplastic cells to herpetic infection. Several argu ments can be presented against this interpretation. First (on epidemiological grounds) is the fact that the peak age of genital herpes is at least 5 years earlier than the peak age for cervical dysplasia, 10 to 15 years earlier than that for in situ carcinoma, and 20 to 40 years earlier than that for invasive cancer (24). Second is the finding that no single case of serologically proven primary HSV-2 infection has been found so far, in our experience, in women with cervical anaplasia. On the other hand, we have almost always found prior serological evidence of HSV-2 antibodies in women with current cervical anaplasia. These observations suggest an alternative hypothesis, based on the known recurrent properties of HSV-1 or HSV-2, to explain the concomitant

detection of genital herpes and cervical anaplasia or the detection of the viral infection after the cervical anaplasia was diagnosed. We have more fully discussed this interpre tation in another report (19) in this Symposium. Cytological changes associated with virologically con firmed HSV infection have been noted in the urinary bladder (13), and histological changes compatible with herpetic infection have been observed in the endosalpinx (27). Our finding of scattered, single altered cells in the endocervix, suggestive of early herpetic infection, has also been noted in the endometrium by us and by other workers (5). Summary and Conclusion

The cytological changes associated with herpes simplex virus infection have been detected in 673 of 279,589 cervicovaginal smears (0.26%) in an indigent hospital popu lation, and has varied from 0.02% to 3.3% in other workers' studies. The prevalence rate depends on age, race, and socioeconomic characteristics of the populations studied (7) and probably reflects the venereal pattern of transmission of genital herpes. These rates probably represent two-thirds of the actual rates, since we have found that the sensitivity of cytological techniques that detect herpetic infection is of that order. The diagnostic cytological features and their pitfalls are described in detail. Histopatholgical features of herpetic infection in 45 biopsies and cone specimens are best observed with highpower magnification of the margin of ulcerated cervical areas, when multiple sections of the tissue block are examined. Most of the lesions were located at the squamo columnar junction, and diagnostic cells were found at the base of the normal squamous epithelium or at the surface of anaplastic epithelium. Most herpetic lesions were multifocal. Peculiar, scattered, single unattached cells with en larged, hyperchromatic nuclei were noted in the endocervical glands of sections from patients with herpetic infections but were not noted in uninfected sections. The inflammatory reaction was composed mainly of lymphocytes and plasma cells. Although herpetic changes could not be identified within anaplastic cells by cytological examination, they could be observed in histological sections of cervical anaplasia com bined with an HSV infection. The presence of foci of cells showing herpetic changes in the midst of solid tumor suggests that the anaplastic cells could be infected by the herpesvirus. The original association of genital herpes with cervical anaplasia, which showed that 16% of 62 cases of cytologi cally detected genital herpes had histologically diagnosed cervical anaplasia, was corroborated when 673 cases of genital herpes were studied. The presence of the anaplasia in about two-thirds of the cases (either before or at the same time as the detection of herpetic infection), in addition to the histopathological findings noted above, suggests that anaplastic tissue is more susceptible to HSV infection. How ever, the ability of HSV infection to recur, the serological findings in women with cervical anaplasia and herpes, and

JUNE 1973 Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

1457

Naib, Nahmias, Josey, and Zaki the epidemiological information demonstrating that the age of acquisition of genital herpes is 5 to 30 years earlier than that of diagnosis of cervical dysplasia, in situ carci noma, and invasive carcinoma, would still make the causalrelation hypothesis tenable.

17.

18.

References 1. An, S. H. Herpes Simplex Virus Infection Detected on Routine Gynecologic Cell Specimens. Acta Cytol., 13: 354 358. 1969. 2. Beilby. J. O. W., Cameron, C. H., Catterall. R. D., and Davidson, D. Herpesvirus hominis Infection of the Cervix Associated with Gonor rhea. Lancet, /: 1065-1066, 1968. 3. Blank. H. Virus Diseases Affecting the Skin. Acta Dermato-Venereo!. 29: 77 107, 1949. 4. de Thé,G.. Biggs. P. M., and Payne, L. N. (eds.) Oncogenesis and Herpesvirus. Lyon, France: International Agency for Research on Cancer, WHO, 1972. 5. Goldman, R. L. Herpetic Inclusions in the Endometrium. Obstet. Gynecol., 36: 603 605, 1970. 6. Greene, H. J., Oppenheim, A., and Olswang, A. Results of Routine Papanicolaou Smears in Two Venereal Disease Clinics in New York City, Acta Cytol., 9: 319-321, 1965. 7. Jordan, S. W., Evangel. E., and Smith, N. L. Ethnic Distribution of Cytologically Diagnosed Herpes Simplex Genital Infection in a Cervical Cancer Screening Program. Acta Cytol., 16: 363 365. 1972. 8. Josey, W. E., Nahmias, A. J., and Naib, Z. M. Genital Infection with Type 2 Herpesvirus Hominis: Present Knowledge and Possible Relation to Cervical Cancer. Am. J. Obstet. Gynecol., 101: 718 729, 1968. 9. Josey, W. E., Nahmias, A. J., Naib. Z. M.. Utley, P. M., McKenzie, W. J.. and Coleman, M. T. Genital Herpes Simplex Infection in the Female. Am. J. Obstet. Gynecol., 96: 493-501, 1966. 10. Koss. L. G. Diagnostic Cytology and Its Histopathologic Bases. Ed. 1. Philadelphia: J. B. Lippincott Co., 1961. 11. Lipschutz. B. Untersuchungen über die Aetiologie der Krankheiten der Herpesgruppe. Arch. Dermatol. Syphilol., 136: 428-482, 1921. 12. Martin, C. E. Marital and Coital Factors in Cervical Cancer. Am. J. Public Health, 57: 803 814, 1967. 13. Masukawa, T.. Garancis, J. C.. Rytel. M. W., and Mattingly. R. F. Herpes Genitalis Virus Isolation from Human Bladder Urine. Acta Cytol., 16: 416-428, 1972. 14. Moghissi, K. S., and Mack, H. C. Epidemiology of Cervical Cancer. Study of a Prison Population. Am. J. Obstet. Gynecol., 100: 607-614, 1968. 15. Nahmias, A. J.. and Dowdle, W. R. Antigenic and Biologic Differ ences in Herpesvirus Hominis. Progr. Med. Viro!., IO: 110 159, 1968. 16. Nahmias, A. J., Dowdle, W. R., Naib, Z. M., Josey, W. E., Melone, D., and Domescik, G. Genital Infection with Type 2 Herpesvirus

1458

19.

20. 21. 22. 23.

24.

25.

26. 27.

28. 29. 30.

31.

32.

33.

Hominis: A Commonly Occurring Venereal Disease. Brit. J. Venereal Diseases, 45: 294-298, 1969. Nahmias, A. J., Josey, W. E., Naib, Z. M., Freeman, M. G., Fernandez, R. J., and Wheeler, J. H. Perinatal Risk Associated with Maternal Genital Herpes Simplex Virus Infection. Am. J. Obstet. Gynecol., 101: 825 834, 1971. Nahmias, A. J.. Naib, Z. M.. Josey, W. E., and Clepper, A. C. Genital Herpes Simplex Infection. Virologie and Cytologie Studies. Obstet. Gynecol., 29: 395 400, 1967. Nahmias, A. J., Naib, Z. N., Josey, W. E., Franklin, E. Prospective Studies of the Association of Genital Herpes Simplex Infection and Cervical Anaplasia. Cancer Res., 33: 1491-1497, 1973. Naib, Z. M. Cytologie Diagnosis of Cytomegalic Inclusion-Body Disease. Am. J. Diseases Children, 105: 153 159, 1963. Naib, Z. M. Exfoliative Cytopathology. Boston: Little, Brown & Co., 1970. Naib, Z. M. Exfoliative Cytology of Viral Cervicovaginitis. Acta Cytol., 10: 126-129, 1966. Naib, Z. M., Nahmias, A. J., and Josey, W. E. Cytology and Histopathology of Cervical Herpes Simplex Infection. Cancer, 19: 1026 1031, 1966. Naib, Z. M., Nahmias, A. J., Josey, W. E., and Kramer, J. H. Genital Herpetic Infection. Association with Cervical Dysplasia and Car cinoma. Cancer, 23: 940-945, 1969. Ng, A. B. P., Reagan, J. W., and Yen, S. S. C. Herpes Genitalis: Clinical and Cytopathologic Experience with 256 Patients. Obstet. Gynecol., 56:645 651, 1970. Nowakovsky, S., McGrew, E. A., Medak, H., Burlakow, P. and Nanos, S. Manifestations of Viral Infections in Exfoliated Cells. Acta Cytol., 12: 227-236, 1968. O'Neill, R. T., and Czernobilsky. B. Morphologic Changes in Endosalpinx Suggestive of Viral Etiology. Report of Two Cases. J. Reproductive Med., 3: 77 80, 1969. Ravish, A. The Role of Viral Infections in Genital Cancers. Cancer Cytol., S: 28-35, 1967. Rotkin, I. D. Adolescent Coitis and Cervical Cancer: Association of Related Events with Increased Risk. Cancer Res., 27: 603-617, 1967. Stern, E.. Longo, L. D. Identification of Herpessimplex Virus in a Case Showing Cytological Features of Viral Vaginitis. Acta Cytol., 7: 295-299, 1963. Teplitz, R. L., Valco, Z., and Rundall, T. Comparative Sequential Cytologie Changes following in Vitro Infection with Herpesvirus Type I and 2. Acta Cytol., 15: 455-459, 1971. Wilbanks, G. D., Campbell, J. A., and Kaufmann, L. A. Cellular Changes of Normal Human Cervical Epithelium Infected in Vitro with Herpesvirus Hominis Type 2 (Herpes Simplex). Acta Cytol., 14: 538-543, 1970. Wolinska, W. H., and Melamed, M. R. Herpes Genitalis in Women Attending Planned Parenthood of New York City. Acta Cytol.. 14: 239 242, 1970.

CANCER RESEARCH

Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

VOL. 33

Cervical A naplasia

Fig. I. Reactive basal cell hyperplasia in early cervical herpetic infection. A, exfoliated basal cell with fragmented nucleoli, clearing of the nuclei, binucleation. and beginning of a perinuclear halo. Papanicolaou stain, x 1200. B, corresponding biopsy section with basal cell hyperplasia and monocytic cell infiltrate. H & E, x 350. Fig. 2. Later stage of the herpetic infection. A, hyalinization of the cytoplasm and blandness of the nuclei. No inclusions are yet seen, but B-bodies can be noted next to the nuclear membranes. Papanicolaou stain, x 850. B, corresponding section. Note the occasional multinucleation and abundance of cytoplasm in the infected cells at the center. H & E, x 350. Fig. 3. Later stage of the herpetic infection. A, note the multinucleation. fine granularity, blandness, and molding of the nuclei and the beginning of intranuclear inclusion formation. Papanicolaou stain, x 850. B, corresponding section showing the same cells starting to detach from the noninfected cells. H & E, x 350. Fig. 4. Latest stage of the herpetic infection. A, giant multinucleated cells, with molded nuclei and prominent multinuclear inclusions. Papanicolaou stain, x 850. B, corresponding section of the cervical infection. Note the few scattered, loose, multinucleated cells in the middle of cellular debris and the mixed inflammation. These cells are located only at the margin of the ulcérationand are easy to miss. H & E. x 350. Fig. 5. Pitfalls in cytological diagnosis: A. degenerated endocervical cells; note the nuclear molding but retention of the chromatin. x 850. B, trophoblastic cells; note the thick cytoplasm, overlapping multiple nuclei, and prominent nucleoli, x 850. C, anaplastic adenocarcinoma; noie the nuclear molding, retention of the abnormal chromatin and fragmentation of the nucleoli x 850. D. foreign body giant cells; note the vesicular cytoplasm and retention of the chromatin granularity x 850. Fig. 6. Foci of herpetic infection in the surface of a well-differentiated invasive squamous cell carcinoma of the cervix. H & E, x 350. Fig. 7. Scattered single endocervical cells, often binucleated, with dark nuclei and dense cytoplasm in the endocervical gland, as seen in the cervical biopsies of proven genital herpetic infection. H & E, x 350.

JUNE 1973

1459

Downloaded from cancerres.aacrjournals.org on January 29, 2019. © 1973 American Association for Cancer Research.

Naib, Nahmias, Josev, and Zaki

1A

<

IB

l,r -

t

'; ^

1

t" .:. Mf« •¿