coitus and cancer - Europe PMC

910

COITUS AND CANCER* LAURE AURELIAN, PH.D. Associate Professor, Department of Comparative Medicine, Department of Biochemistry and Biophysics Assistant Professor, Department of Microbiology The Johns Hopkins University Schools of Medicine and Public Health, Baltimore, Md.

T HE study of the etiology of all human cancers is complicated by the nature of the host. Coital, marital, and parity factors have been identified as playing a role in the etiology of squamous cancer of the human uterine cervix, suggesting this neoplasia may be infectious in origin and venereally transmitted. Herpesvirus type 2 (HSV-2), recently identified as biologically and antigenically1 distinct from the virus causing facial lesions (fever blisters), has been associated with cervical cancer on the basis of numerous seroepidemiologic2 and virologic studies.351 The herpesvirus hypothesis, visualizing HSV-2 as the causative factor, merits serious consideration in view of the venereal pattern of transmission of this virus and the observation that its biologic properties are consistent with an oncogenic potential for man. The available data merit discussion from these standpoints, and in the general contexts of the multifactorial nature of neoplasia, as well as of the role of the immune response -in the progression or control of cancer. CERVICAL CANCER: A VENEREALLY TRANSMITTED DISEASE

The analogy of cervical cancer with venereal diseases is not new. In 1842 Rigoni-Stern presented the results of his studies on cloistered nuns and married women and concluded that the susceptibility of the uterus to cervical cancer in the two groups varies according to "the natural exercise of its functions." His findings on the low rate of cervical cancer in nuns have since been confirmed repeatedly7'8 and various other populations at *Presented as part of a Conference on Sexually Transmitted Diseases held by the Committee on Public Health of the New York Academy of Medicine November 5 and 6, 1975. This research was supported by contract Nol-CP-33345 from the Virus Cancer Program, National Cancer Institute, Bethesda, Md.

Bull. N.Y. Acad. Med.

COITUS AND CANCER

COITUS

911

CANCER911

low risk (e.g., Jewish, Moslem) or high risk (e.g., prostitutes) have been identified.9 Mortality from cervical cancer10 and, even more important, its incidence,1l are increased among the married and decreased among those who never married, independent of age. Studies specifically designed to explore coital, marital, and parity factors in cervical cancer by means of personal interviews have been repeatedly reviewed.9'12-14 Briefly, they indicate that the mean age at first coitus is younger and the proportion initiating coitus early is larger among patients with cervical cancer than among controls. The patients tend to have more coital partners than other women and the proportion among them denying coitus is essentially zero, whereas in controls it is small, but real. These data are consistent with the fundamental relation of the disease with coital or venereal factors. Coinciding with marital factors playing a role in cervical cancer are the observations that women who develop the disease are less likely to remain unmarried, tend to marry for the first time at a younger age, and are more likely to have had several marital partners than other women. Extramarital sexual practice by either the patient or her spouse also is a risk factor implicating venereal elements in the etiology of cervical cancer. Significantly, from the standpoint of venereal transmission, age at first coitus appears to be more important than age at first marriage in differentiating between cervical-cancer patients and controls. Associations of cervical cancer with pregnancy appear to be secondary to the marital or coital determinants. That the penis plays a direct role in cervical cancer is not yet definitely established, although penile factors are associated with the disease. Utilizing data from the Central Cancer Registry of Puerto Rico, Martinez noted an unusually high rate of cervical cancer among the wives of 889 men diagnosed as having cancer of the penis.15 A significant correlation between the death rates from cervical cancer and penile cancer has been reported in Japan.16 Possibly the most impressive and direct data on the role of the male consort in cervical carinogenesis recently was presented by Kessler as a preliminary report of a study17 seeking to answer the basic question: "Is the risk of developing cervical cancer increased among the wives of men who at some other time in their lives were married to other women who developed cervical cancer?" To this end, a large cohort of women with epidermoid carcinoma of the cervix during the past 20 years was identified, and the previous and subsequent wives of their husbands were traced. Controls consisted of random samples of wives similar to the

Vol. 52, No. 8, October 1976

912 L. AURELIAN 91..ARLA

case wives in demographic and marital characteristics. The ultimate risk of cervical neoplasia among the two cohorts was ascertained by interview, review of medical records, and cytological screening. Following analysis of 4,178 probands with cervical cancer, Kessler identified a total of 726 husbands and 856 other wives. At the time of the report, 14 other wives with cervical cancer had been detected in the cancer group as compared with a total of 4 cancers among the control wives. Although Kessler emphasises the preliminary nature of his results, these differences between cancer and control groups are highly significant and clearly suggest that certain males may transmit venereally a risk factor for cervical cancer in women. ASSOCIATED VENEREAL DISEASES

Women with cervical cancer are more likely than expected to have venereal diseases, such as syphilis, gonorrhea, and trichomoniasis. During the 1950s at least four epidemiologic studies yielded information on the prevalence of syphilis among cervical-cancer patients.18-21 A positive association between the two diseases was consistently noted, but a more detailed examination of the data compelled one of the investigators "to assume the existence of a particularly potent carcinogenic factor among prostitutes' -a factor which has no relation to syphilis.18 One may only speculate on the extent to which HSV-2 might have constituted the hypothetical carcinogen. Although such associations constitute further evidence that exposure to one or more sexually transmitted factors is an important determinant in cervical cancer, they beg differentiation between a primary association and one resulting from the relation of each to sexual behavior. At least some answers to this question were provided by Royston and myself;22 we studied a single high risk population and reported a similar frequency of syphilis and trichomoniasis in the carcinoma and control groups, as compared to a significantly higher frequency of HSV-2 in the carcinoma group. It now is generally accepted that the association of both trichomoniasis and syphilis with cervical cancer is secondary, resulting from the relation of each to sexual behavior. VENEREAL TRANSMISSION OF HSV-2

In his description of genital herpesvirus, published in 1736, Jean


COITUS AND CANCER CANCER

913

Astruc23 classified the condition as a venereal disease and noted its frequency among practicing homosexuals. Greenough in 1881 noted that: "...the influence of the venereal act as the immediate cause of herpes progenitalis has not been given the importance that it deserves."24 He stressed that in his practice genital herpes was seen only in males. This conclusion soon was challenged by the German physician Unna, who claimed that he was able to identify only 17 males as compared to 423 females treated for this disease. He concluded that women and men are equally susceptible.25 Direct evidence concerning the venereal transmissibility of genital herpetic infections rests on a number of observational studies on the sexual contact of sexual partners. Unfortunately, in many26 the diagnosis was based on clinical observations without supporting laboratory evidence. In others, although laboratory support was provided, few people were studied. Among the latter studies, those of Nahmias and Rawls and their respective colleagues merit special attention. Nahmias et al.27 examined eight female sexual contacts of seven male patients with virologically confirmed HSV-2 penile infections within one week of the initial detection of the herpetic lesions in the men. Seven of the eight female contacts had HSV-2 infections of the vulva or the cervix at that time, and genital virus was isolated from each of the seven women. Rawls et al.28 examined 30 female contacts of 30 males with clinically diagnosed genital herpes infections attending a public venereal-disease clinic. Virus was successfully isolated from the penile lesions of 53% of the male patients and from the genital secretions of 33% of their female contacts. These rates were in excess of the prevalence of HSV-2 infections in the control group of women from the same clinic. One difficulty with these studies rests on the necessity of determining whether the postcoital manifestations of HSV-2 infections are venereal, due to the transmission of the herpesvirus during coitus, or nonvenereal, i.e., due to stress resulting from coitus and triggering the recrudescence of a persisting latent virus.29 The two phenomena could be differentiated by analyzing for the presence of antibody of HSV-2 prior to and after the appearance of the coital herpetic lesions. Lesions appearing postcoitus in an individual without antibody to HSV-2 may be taken as evidence that this individual obtained his infection at the time of coitus as a result of venereal transmission of the virus. Such studies are relatively difficult to conduct because of their longitudinal pattern. Evidence of venereal trans-


914

AUELA 914~L.L. AU-RELIAN

mission of HSV-2 in animal models was presented by Nahmias et al. ;30 they mated uninfected male mice with female mice with laboratory-produced vaginal HSV-2 infections and reported sexual transmission of the genital infection. These studies are not conclusive evidence for the venereal transmission of HSV-2. However, they are not different from most early clinical efforts to demonstrate the venereal transmission of syphilis and gonorrhea. Of particular significance are recent studies that seem to indicate that not all genital herpetic lesions should be attributed to venereally transmitted virus. Thus, the initial findings of a survey of health and disease carried on for the past few years among several orders of nuns and their married siblings (related by birth) revealed that the prevalence rate of neutralizing antibody to HSV-2 is 58.6% among the nuns and 57.4% among their sisters.a There was a complete absence of trichomonads from the cervical smears of the nuns, but not from those of their siblings. The situation may be analogous to syphilis, which generally follows a venereal mode of transmission, but which manifests itself in endemic nonvenereal forms as

well.31 MALE ROLE IN CERVICAL CANCER

That males may serve as a reservoir for HSV-2 is suggested by the data of Centifanto et al.,32 who isolated the virus from the genitourinary specimens of 15% of 190 male patients seen at a urology clinic. None of the subjects had a known history of genital herpes. Successful isolations of HSV-2 also were reported by others33'34 from the urethral discharges of men with nonspecific urethritis who were attending venereal disease clinics. For selected males to serve as asymptomatic carriers of HSV-2 they would probably have to experience chronic infections characterized by continuous virus replication at a slow rate in cells of the genitourinary tract. This type of infection is to be differentiated from latent herpetic infections,' which may be defined as those in which viral replication (and, therefore, viral isolation) does not' occur, except during the brief periods of clinical recrudescence. Because clinically symptomatic herpetic lesions of the penis are extremely painful, venereal transmission of the agent through sexual intercourse by a patient with latent and recurring infection would rarely, if ever, take place. Therefore, it is the asymptomatic males with chronic, low-grade herpetic infections of the genitourinary tract who are the likely carriers and transmittors of the putative carcinogen. Bull. N.Y. Acad. Med.

COITUS AND CANCER

COITUS

AND

915

915

Proceeding with this line of reasoning, one might speculate that the coital partners of women who are destined to develop cervical cancer would characteristically suffer from chronic, low-grade herpetic infections of the genitourinary tract, whereas their consorts would undergo primary infection leading to the latent state, which on biochemical grounds is potentially oncogenic.35 In turn, the male partners themselves would most likely have been infected by females with chronic herpetic infections because the number of women at any given time with active primary infections probably is negligible. Although the major emphasis of this paper is addressed to the role of HSV-2 in cervical carcinogenesis, an alternative theory consistent with the male role in cervical carcinogenesis should be cited. Coppelson and Reed postulated that malignant transformation of the cervix is not likely to occur when the uterus is actively undergoing squamous metaplasia such as in early adolescence and at first pregnancy.36 They suggest that the spermatozoa may be mutagenic in particular circumstances. Singer further concludes that the spermatozoa are the likliest candidates for mutagenesis as they have far more DNA than does the herpesvirus.37

ASSOCIATION OF HSV-2 WITH CERVICAL CANCER The herpesvirus hypothesis was brought into the limelight by the observations that: 1) women with HSV-2 infections have a higher probability of developing cervical anaplasia than women in control groups38 and 2) the prevalence of antibody to HSV-2 is significantly higher in women with cervical cancer than in control groups. These data have been extensively reviewed2'29'39 and are summarized in Table I. Three hypotheses have been proposed to account for the results of the seroepidemiologic studies .40 The preferential hypothesis, based on the observation that herpesviruses grow better in actively replicating cells,41 suggests that infection with the virus follows development of the neoplastic lesion. The roles of the atypia and carcinoma in situ in the pathogenesis of cervical carcinoma still are unsolved; however, studies of incidence and prevalence rates have suggested that all or a large proportion of invasive carcinoma are preceded by carcinoma in situ. 42 More recent data also have shown progression of epithelial changes from atypia and carcinoma in situ to invasive carcinoma.434' Accordingly, the preferential hypothesis predicts that the prevalence of antibody to HSV-2 in patients with precancerous lesions is lower than that in invasive cervical cases. The promiscuity Vol. 52, No. 8, October 1976

L. AURELIAN

916

AURELIAN

TABLE I. PREVALENCE OF ANTIBODY TO HSV-2 IN CANCER AND CONTROL GROUPS

Area

United States Houston Atlanta West Virginia Baltimore Chicago Boston Europe Belgium Hungary Denmark England

Yugoslavia Other New Zealand Colombia Taiwan Uganda Israel

Percentage positive Cancer Control

Assay method*

Predominant race Negro Negro Caucasian Negro Caucasian Negro Caucasian Caucasian

Kinet. neut. Index microneut. Index microneut. PN microneut. Index microneut. MA neut.t Kinet. neut. MA neut.

33 83 17 73 47 85 Mean K value 0.41 0.96 25 40

Caucasian Caucasian Caucasian

Kinet. neut. IF Kinet. neut.

Caucasian Caucasian

Kinet. neut.t Index microneut.

23 35 50 72 27

Caucasian Mestizo Oriental Negro Caucasian

Index Index Index Index Index

72 80 54 83 52 100 30 42

31 37 48 93 38

22 44 25 35 23 67 9 17

Origin of virus strains

American, though not necessarily from the same city.

American

microneut. microneut. microneut. microneut. microneut.

American

*Kinet. neut. = kinetics of neutralization, MA = multiplicity analysis neutralization, PN or Index microneut. = microneutralization analyzed according to the PN' or the index39 procedures, IF indirect immunofluorescence tHSV-l and HSV-2 in one antigenic mass Reproduced by permission from Aurelian, L.: The "viruses of love" and cancer. Am. J. Med. Tech. 40:496-513, 1974. Copyright 1974, the American Society for Medical Technology.

hypothesis envisions the association of HS V-2 with cervical cancer as attributable to the promiscuous nature of the cervical-cancer population, which by definition is more apt to have all venereal diseases.18-21 Accordingly, the incidence of venereal diseases other than HSV-2 should be significantly higher in groups of cancer patients than in control groups. Lastly, the etiology hypothesis interprets the association of HSV-2 with cervical cancer as resulting from the causative role, direct or cocarcinogenic, played by HSV-2. The mean age of populations infected with HSV-2 precedes by six years that for the development of the earliest detectable neoplastic lesion,38 suggesting that if HSV-2 causes cervical cancer, it must act from a persistent state. Thus, the etiology hypothesis predicts that the association of HSV-2 with cervical neoplastic cells reBull. N.Y. Acad. Med.

COITUS AND CANCER

917

917

sembles the association of the virus with cells that harbor it in the interim between recrudescences (when the patient is said to be latently infected). The validity of the preferential hypothesis was tested in studies inquiring into the prevalence of HSV-2 antibody in women with precancerous and cancerous lesions. The antibody-frequency rate was found to be identical in preinvasive and invasive cancer,22'45'60 suggesting that infection with HSV-2 precedes rather than follows neoplastic conversion. Therefore, the data disagree with the preferential hypothesis. Promiscuity is a major variable associated with all venereally transmitted diseases. Accordingly, the observation that in the same population in which HSV-2 antibody is associated with cervical cancer the frequency of two other venereal diseases (trichomoniasis demonstrated cytologically and syphilis demonstrated serologically) is similar in cases and controls22 argues against the promiscuity hypothesis. The etiology or herpesvirus hypothesis predicts that, preceding neoplastic transformation, virus persists in cervical cells in a latent state.3 It also assumes that the virus plays a causative role in the transformation of the cell from normal to neoplastic. On the basis of in vitro studies, it can be concluded that the persistence of virus in infected cells depends on the early arrest of viral multiplication before the virus expresses those functions responsible for the inhibition of macromolecular synthesis in the host.35 Neoplastic transformation of these cells would therefore depend on: 1) maintenance of the persistently infected state and 2) expression of the persistent viral genome resulting in the synthesis of specific viral proteins which are associated directly or indirectly with transformation. By definition, these proteins differ from those associated with the inhibition of host macromolecular synthesis, and their regulation is independent of the regulation of viral proteins associated with the inhibition of host functions. If we proceed with this line of reasoning, selective factors affecting the course of differentiation and evolution of the new (transformed) cell type would constitute sufficient explanation for the formation of a tumor. Such factors could be physiologic (hormonal), environmental, or immunologic in nature. Whether the virus itself plays a role at this level is difficult to assess at present. The predictions resulting from these interpretations are amenable to laboratory investigation. 3'4'5'94653'6071 They include the presence-in human cervical tumor cells of viral nucleic acids, i.e., deoxyriboneucleic acid (DNA, the complete genome or a fragment thereof) and small amounts of Vol. 52, No. 8, October 1976

918 918

L. AURELIAN AURELIAN

messenger riboneucleic acid (RNA), and of one or more coded viral proteins, different from those associated with the inhibition of host functions. The whole spectrum of viral proteins and infectious virus should not be present in the tumor cells. The hypothesis also predicts that the virus can transform cells in vitro and possibly even in a model animal. These predictions do not impose restrictions on the possible molecular mechanism of cellular transformation. It may be argued that, whatever its etiology, transformation is the result of a. perturbation in the cellular genome which is responsible for the loss of regulation and continuous cell replication. Both chromosomal imbalance,54 possibly associated with gene loss, and the modification of the cell surface in turn resulting in unregulated synthesis of DNA,55 have been suggested as possible mechanisms of transformation. Indeed, aneuploidy is characteristic of cervical cancer.43 The most significant aspect of the etiology hypothesis is that in the context of its predictions it easily could fit either of these two possibilities; herpesviruses modify the cell surface56 and, when ultraviolet inactivated, HSV-2 causes chromosomal breakage,56a which is undistinguishable from that associated with carcinoma in situ.43 It is important to remember that fulfillment of all these predictions is not absolute evidence of causation. Final conclusions about the etiology of any human neoplasm will depend on a multidisciplinary approach providing concordant results. Fulfillment of the predictions made by the etiology hypothesis is an essential aspect of this approach. However, this information must be interpreted in conjunction with the data obtained by the epidemiologists, immunologists, and clinicians. THE ETIOLOGY HYPOTHESIS AND ITS PREDICTIONS

The following observations provide evidence in support of the predictions made by the etiology hypothesis. 1) Cellular changes induced by herpesvirus and associated with active viral replication are not detected in neoplastic but rather in normal cervical cells; women with HSV-2 infections have a higher probability of developing cervical anaplasia than control women.38 2) Viral structural antigens detected in immunofluorescence by antisera to HSV-2 virions are not observed in biopsied neoplastic cells.3'46'47 These cells also do not contain virus particles,3,46,47 and cytoplasmic changes that are characteristic of the synthesis of herpesvirus antigens in productively infected cells58 are not observed.46 Biopsied tumor cells bn the surface of Bull. N.Y. Acad. Med.

919

COITUS AND CANCER

919

TABLE II. PERSISTENCE OF THE HSV-2 GENOME IN SQUAMOUS CERVICAL CARCINOMA. BALTIMORE, ATLANTA, AND FOREIGN STUDIES Tumor cells

Baltimore studies3,40'46'47 Exfoliated Biopsied Culture Carcinoma in situ (S332G) Viable Degenerated Invasive cancer (614) Viable Degenerated Invasive cancer (613) Viable Degenerated Atlanta studies6" Viable A Viable B

Foreign studies Tumor cells62 Tumor cells59 Exfoliated 60 Biopsied60

Structural HS V-2 antigens

InJectious HS V-2

Tumorassociated antigens

-

-

ND +

+

+

+

ND

+(surface) TU

-

-

-

+ ND

+(surface)

-

-

TU

-

-

+ ND

+ +(surface)

ND ND

ND ND

ND ND

ND ND ND ND

ND ND ND ND

Virus particles

+

7

+

+

ND

ND

+

+ +

ND ND

+ = positive, - = negative, ND = not done, TU requires viable cells)

=

technically unfeasible (membrane fluorescence

the neoplastic lesions, as well as exfoliated anaplastic cells from most patients, contain viral antigens detectable by antisera to HSV-2 vinons.3 52'60 However, unlike the situation in productively infected cells, the antigens in the tumor cells are localized only in the cytoplasm. Besides confirming the predictions of the etiology hypothesis, these observations suggest that the viral genome persisting in the tumor cells is repressed. Exposure of the tumor cells to secretions of relatively high pH57 induces its expression. 3) The cytoplasmic changes characteristic of the synthesis of herpesvirus antigens58 are observed in the exfoliated cervical tumor cells containing the viral antigens;3'46 however, at least 20 of these cells did not contain complete or incomplete virus particles.46 Herrera et al.59 reported the presence of complete virions in two out of 14 cervical tumor biopsies. Their photographs reveal particles which morphologically were commensurate with herpesviruses, but mostly were unenveloped. These observaVol. 52, No. 8, October 1976

920 920

L. AURELIAN

tions confirm the predictions of the etiology hypothesis. Further, they suggest that the virus-host cell interaction covers a broad spectrum of possibilities. At one end is the persistence of DNA sequences corresponding to a fragment of the viral genome,4 and therefore the synthesis of a restricted number of viral proteins. At the other end is the persistence of the complete viral genome. The latter could be defective or under transcriptional regulation such that it can express little or almost all of its information. 4)Normal differentiated squamous cells do not contain virus antigens,46'47 indicating that the viral genome is present only in the tumor cells. 5) Evidence of virus persistence and expression also was observed when cervical tumor cells were grown in culture. In our laboratory tissue cultures were established from three tumor biopsies. The first culture (S332G) did not reveal the presence of virions or viral antigens when unstressed; however, when exposed to medium or high pH it yielded infectious HSV-2. Virus could not be induced from the other two cultures (614 and 613) and viral antigens and particles could not be detected in these cells. However, HSV-2 specific viral antigens were observed on the surface of 2 to 4% of the cells in these cultures.3 The data suggest that whereas S332G cells contain persisting DNA sequences corresponding to the complete viral genome in a repressed state, 613 and 614 cells contain persisting DNA sequences corresponding to a fragment of the genome or the complete genome in a partially repressed state. The presence in tumor cells of cytoplasmic and surface viral antigens was confirmed.6062 The data are summarized in Table II. 6) Confirming the predictions of the etiology hypothesis, DNA sequences corresponding to 40% of the viral genome were reported in one cervical tumor biopsy,4 sequences corresponding to 12% of the genome in murine cells transformed by HSV-2,62a and small amounts of messenger RNA in a tumor biopsy and in hamster cells transformed by HSV-2 in vitro. 4,63 7) Five tumor-specific, HSV-2 induced antigens have been described.3'49'50'5'470 One such antigen, designated AG-4, was prepared in HSV-2 infected HEp-2 cells, but also was observed in cervical tumor biopsies and in cultures of cervical tumor cells (Table III). AG-4 is not present in biopsies from adenocarcinoma of the cervix and from normal (non-neoplastic) cervices.49 Maximal synthesis of AG-4 occurs in infected


COITUS AND CANCER

921

921

HEp-2 cells early (4 to 8 hours after infection) in the replicative cycle of the virus.67 It is one of the major proteins made under nonproductive conditions such as in the presence of an inhibitor of viral DNA synthesis.51 Thus, AG-4 can be made in the absence of those viral proteins associated with the inhibition of host macromolecular synthesis. Further, AG-4 probably is virus coded, as it appears to be a virion protein 161,000 daltons in molecular weight.51 However, it is not located on the surface of the virion, as it does not block serum-neutralizing potential.67 The tumor specificity of AG-4 was established by inquiring into the presence of AG-4 antibody in patients with cervical cancer and in controls. The results of these studies (Table IV) indicated a significantly higher prevalence of AG-4 antibody (85%) in patients with invasive carcinoma than in matched control groups (12%). Antibody to AG-4 reflect the progression of the neoplastic lesion; 35% of women with cervical atypia, 65% of those with carcinoma in situ, and 85% of those with invasive cancer were AG-4 positive. Antibody to AG-4 was not present in patients with cervical cancer which had been treated successfully prior to the collection of blood and without evidence of recurrent neoplasia, but was present in those women with recurrent cancer following therapy. The data establish AG-4 as a major viral product in nonproductive conditions as well as establishing its tumor specificity. The data also indicate its association with active tumor growth. In this respect it is significant that AG-4 appears to be located on the cell surface.5 1 Four other tumor-specific antigens have been independently described. The nonvirion antigen of Sabin and Tarro64 is made in HSV-2 infected HEp-2 cells late (24 hours) in the replicative cycle of the virus. It was characterized as a nonstructural protein on the basis of adsorbtion experiments, and has been associated with nine human cancers, including cervical. The nonvirion antigen has recently been retracted by Sabin.68 Another antigen (HSV-NVA) was prepared from sonicated membranes of cervical or lip cancer, as well as from human fetal kidney cells infected for 24 hours.65'66'69 The available data do not unequivocally establish HS V- NVA as HS V-2 specific. The results obtained with control antigens prepared from mock-infected cells are not described. Antibody to this antigen is present in 95% of patients with carcinoma of the larynx and 70% of those with squamous cancer of the head and neck including tongue, mouth, palate, tonsils, maxillary sinus, pyriform sinus, and gingiva.69 HS V- NVA is similar or identical to the nonvirion antigen.65


922

922

L. L. AURELIAN

AURELIAN~~~~~~~~~~~~~~~~~~~~~~~~~~~~

CCA was extracted from cervical cancer tissue. It displays a line of identity with HSV-2 in immunodiffusion assays;53 however, it is not clear whether it is virion or nonvirion in nature. Its relation to the development of the neoplasm has not been studied. More recently, another antigen has been associated with cervical cancer.70 This antigen is designated VP 134 and is a polypeptide of 134,000 daltons molecular weight found in HSV-2 infected HEp-2 cells. Like AG-4, VP 134 is made early in the replicative cycle of the virus (three hours) but, unlike AG-4, it does not comigrate in polyacrylamide gels with virion proteins. On the basis of these criteria it has been classified as nonstructural. Sera from patients with carcinoma of the cervix have a higher reactivity to VP 134 than sera from normal women and women with breast cancer. The presence in cervical tumors of viral antigens other than those associated with the inhibition of host functions35 fulfills the predictions of the etiology hypothesis. 8) As predicted by the etiology hypothesis, HSV-2 transforms hamster cells in vitro. Transformation is observed only with inactivated virus, i.e., virus capable of establishing a nonproductive infection. At least one of the transformed clones induces tumors in animals.5 Transformation of mouse cells and tumor induction by these clones also has been reported.92 9) Preliminary studies71 suggest that cebus monkeys may be a valid animal model for HSV-2 oncogenesis. Approximately six months following repeated infection of the cervix with HSV-2 the animals show cervical changes commensurate with atypia. Appropriate controls have not been analyzed yet. CRITICISMS OF EVIDENCE

The studies described above fulfill the predictions of the etiology hypothesis. They indicate that the HSV-2 genome (or fragments thereof) persists in cervical tumor cells and expresses at least part of its genetic information. Some of this information is tumor-specific and apparently is associated with active tumor growth. When inactivated, the virus can transform cells in vitro, and an animal model appears to be established. Nevertheless, the studies are amenable to criticism. The close antigenic relation between herpesvirus type 1 (HSV- 1) and HSV-2 and the presence of HSV- 1 antibody in 80 to 90% of all populations studied29 has given rise to critical speculations on the sensitivity and specificity of the assays used in the seroepidemiologic studies which are summarized in Table I. The


COITUS AND CANCER

923

923

problem is compounded by the observation that HSV-2 strains isolated in foreign countries are antigenically distinct from the strains from the United States which are used in the seroepidemiologic studies;72 it points to another cause for loss in assay specificity. As discussed by Kessler,2 in the absence of absolute criteria it is virtually impossible to determine whether any one assay is truly specific and sensitive. However, the criticisms are valid and they indicate that the assays are not very sensitive nor, at least in some cases, very specific. It is, therefore, extremely significant that, even with such relatively unsophisticated methods, HSV-2 and not HSV-1 is associated with cervical cancer in 11 of the 16 populations studied (Table I). The presence of HSV-2 antigens in cervical tumor cells is confirmed;52,60,62 however, the virus isolation48 and nucleic acid4 studies still are controversial. HSV-2 was isolated only from a culture established from a case of carcinoma in situ; it could not be isolated from two other lines established from cases of invasive carcinoma.3 Similarly, viral DNA sequences were observed in only one biopsy of invasive cancer.4 The results were not reproduced with other biopsy specimens73 or with HSV-2 transformed hamster cells.63 These difficulties have resulted in controversy,73 particularly in view of the reproducible identification in tumor cells of the genome of EBV, a herpesvirus associated with Burkitt lymphoma.74 The possibility cannot be excluded that both the isolation of the virus48 and the identification of the viral DNA4 in cervical tumor cells are artifactual. However, this is unlikely because 1) the tumor isolate differs from the resident virus in our laboratory in a number of biologic properties75 and 2) HSV-2 transformed cells contain viral proteins5 and mRNA63 and sequences corresponding to 12% of the viral genome were found in HSV-2 transformed mouse cells.75a In view of the relative paucity of tumor cells in a biopsy of cervical cancer (estimated at one in 10,000 cells on the average), it can be calculated that even if one complete copy of the viral genome were present per tumor cell, DNA detection would not succeed; the limit of sensitivity of the hybridization technique is 1/10th of a genome. The alternative interpretation is that the interaction of HSV-2 with the host cell is unlike that of EBV and covers a broad spectrum of possibilities: At one end is the persistence of the complete viral genome in every tumor cell and at the other the persistence of fractions of the genome in only some of those cells. The success of Frenkel et al.4 could be attributed to the large (70 gm.) and exdfitic nature of the specimen they studied. Vol. 52, No. 8, October 1976

924

924

L. L. AURELIAN

AURELIAN~~~~~~~~~~~~~~~~~~~~~~~~

TABLE III. PRESENCE OF HERPESVIRUS TYPE 2 (HSV-2) TUMOR-SPECIFIC ANTIGENS IN CERVICAL TUMOR A uthors and name of antigen Hollinshpad et at. 65,69

Cell of antigen production Carcinoma of cervix

Time of antigen production Unknown

Human sample analyzed Sera

HEK and HEB 24 hrs. HSV-NVA cells a. i. HEK and HEB 24 hrs a. i. cells

Aurelian et al.49'50 AG-4

Sabin and Tarro64

Nonvirion antigen

Gall and

Haines53 CCA Anzai et al. 69 VP 134

HSV-2 infected HEp-2 cells

HSV-2 or HSV-I infected guinea pig or HEp-2 cells

4 hrs. a.i.

3 hrs. a.i. in guinea pig and 24 hrs. a.i. in HEp-2 cells

Carcinoma Unknown of cervix vulva, breast, ovary, and colon Normal cervical tissue HSV-2 infected 3 hrs. HEp-2 cells a.i.

Sera

Sera

Tumor extracts

Sera

Group of patients studied

Positive Jor early antigen No. %

Cervical cancer Controls Cervical cancer Cancer of larynx Controls Cervical atypia Controls Carcinoma in situ Controls Invasive cervical cancer Controls Carcinoma of cervix Carcinoma of prostate Carcinoma of bladder Carcinoma of kidney Carcinoma of nasopharynx Carcinoma of breast Carcinoma of stomach Carcinoma of colon Controls Invasive cervical cancer

Carcinoma of cervix and breast Controls

4/4 21/24

100 0 87

36/38

95

2/36

7/20

5 35

0/20 13/20

0 65

1/20

29/34

5 85

4/34 13/13

12 100

0/6

7/8

88

6/8

75

6/8

75

9/10

90

0/5

0

0/5

0/9

0 0

0/57

0

25/25

100

0/25

0

0/25

0

Higher reactivity in 15 cervical cancer

patients than in controls and breastcancer

patients HEK = human embryonic kidney, HEB = human embryonic brain, a.i. = after infection


925 925

COITUS AND CANCER COITUS

TABLE IV. PRESENCE OF ANTIBODY TO AG-4 IN CANCER PATIENTS AND CONTROLS

Gi-oup No. tested

Atypia Matched controls In situ Matched controls Invasive Matched controls Treated invasive cancer Recurrent cancer Cancer at other sitest Recurrent herpes (labial, genital)

20 20 20 20

Studv A Study, B Positiv e Positive Positive Positive J6r AG-4 fho AG-H* No. for AGG 4o4AG0 H* 4c No. No. %k tested No. cl No. /k

35 0 65 5 85 12

0 0 0 0

0 0 0 0

34 34

7 0 13 1 29 4

0 5

0 15

8 NA ND ND 2 NA

26 7

0 6

0 86

0 0

0 0

ND ND

18

0

0

0

0

ND

3 NA

38 NA

0 NA

0 NA

2 NA

100 NA

0 NA

0 NA

20

3y

15

0

0

5

0

0

0

0

NA = not applicable, ND = not done *Complement fixing antibody against control antigen (AG-H) prepared from mock infected cells is assayed simultaneously with AG-4 t Includes lung, breast, pancreas, stomach, vagina, vulva, ovaries, endometrium, and adenocarcinoma of the cervix in Study A, and vulva, bladder, prostate kidney, NPC, oropharynx, tonsils, salivary gland, breast, stomach, colon, and lung in Study B. ypositives are oropharynx, nasopharynx, and colon (one out of two tested).

In view of the mounting evidence associating neoplasia with defective virus,78 it may be argued that the presence of the complete functional viral genome is evidence against the etiology hypothesis and in favor of a carrier-state relation. Considering this interpretation, the successful isolation of HSV-2 from the carcinoma in situ culture can be attributed to the precancerous nature of the lesion and, therefore, its ability to regress.44'77 Regression could be associated with prior viral expression.77 In view of the potential significance of the tumor-associated antigens to the etiology hypothesis, it becomes imperative to establish whether these are virus-coded proteins and whether they are related specifically to the clinical status of the disease. The present data do not fully answer these questions in a satisfactory fashion. Absolute evidence that an antigen is virus coded is the in vitro synthesis of that antigen under the direction of the viral nucleic acid. In the absence of such data it is reasonable to assume that a virion structural protein indeed is virus coded. Among the tumor-associated antigens, AG-4 is the only virion protein." The nonvirion and VP 134 antigens appear to be nonstructural64,70 and the exact nature of the others is unknown. The Vol. 52, No. 8, October 1976

926

9Th

L. L. AURELIAN AURELIAN~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

association of most of these antigens with the clinical status of the patient also is not clear. The data summarized in Table III indicate that a small number of patients was studied in each group. Nevertheless, efforts to associate the antigen with clinical status were made for AG-4 (Tables III, IV). Of particular significance are the longitudinal follow-up studies3'50 which indicate that AG-4 is associated with tumor recurrence. Thus, the data suggest that AG-4 is a virus-coded protein that reflects active tumor growth. According to these criteria, it may be concluded that the virus plays an active role in oncogenesis. However, large-scale longitudinal follow-up studies are necessary before final conclusions can be made. Recent studies confirm the AG-4 results.77a The controversy over the nonvirion antigen remains to be solved. The antigen was retracted by Sabin,68 but not by Tarro.77b Further, Sabin's retraction discredited the HSV-NVA antigen, as it has been shown to be identical to the nonvirion antigen.65 Notter and Docherty recently have confirmed the HSV-NVA antigen,77a but this confirmation will necessitate the critical reconsideration of the nonvirion antigen or of the data indicating its similarity to the HSV-NVA antigen. The transformation studies would be enhanced greatly by evidence that HSV-2 also can transform human cells, and much remains to be done on the animal-model studies now in progress.69

PARADOXES OF THE ETIOLOGY HYPOTHESIS Disturbing in terms of the etiology hypothesis are the following observations and questions: 1) HSV-2 is a ubiquitous infection,29 whereas cancer of the cervix is a rare disease.9 Why do so many women infected with HSV-2 fail to develop neoplasia? 2) Herpesviruses cause productive infections characterized by virus replication and cell death. On the other hand, neoplasia is cell immortality. What is the solution to this paradox? 3) Assuming that HSV-2 causes human neoplasia and in view of the significant nucleic acid homology between HSV- 1 and HSV-2,29 is HS V- 1 potentially oncogenic? THE UBIQUITOus HSV-2 Cancer is a multifactorial disease. If HSV-2 causes cervical cancer, it


COITUS AND CANCER

-~~~~CIU

927

must be only one of a number of etiologic factors. Among the possible associated factors, steroid hormones,76 genetic determinants,51 and type C virus particles79 have received special attention. Comparing cases of cervical cancer and controls in Yugoslavia, Kessler et al.76 noted that fewer than the expected number of cases reported unusually precocious ages of onset of menstruation. Irregularities in menstruation were rarer among the cases observed and fewer than the expected number of cases suffered from unusually short menstrual intervals or unusually long, heavy, or painful flows. The authors suggested the possibility that the endocrinologic status of women at risk of cervical cancer differs from that of control women so as to "normalize the menses, predispose to early sexual activity, reduce coital satisfaction, increase marital inability and encourage multiple marriages, while at the same time creating a physiologic milieu receptive to carcinogenic agents." Munoz et al.80 reported a higher frequency of precancerous and cancerous lesions of the cervix or vagina in mice exposed to HSV-2 and estrogens than in those exposed to HSV-2 alone. The suggestion that genetic factors play a role in oncogenesis is supported by our study of cervical cells infected with UV inactivated HSV-2. Four cultures were established, three from women seen in the infertility clinic at Johns Hopkins Hospital and the fourth from a fetal cervix. The chromosome modality was diploid in all four cultures. However, 48 hours after infection with the inactivated virus, the chromosome modality differed in the various cultures. Normal (2N) metaphases were observed in 85% of the fetal cells and in the cervical cells established from one seronegative woman with a normal Papanicolau smear. Hypotetraploid (4N) metaphases, characteristic of carcinoma in Situ,43 were observed in the cervical cells from two other women who had lesions diagnosed as carcinoma in situ and marked atypia respectively, on the site opposite to that from which the biopsy was taken.36a These studies indicate that the chromosome modality in the HSV-2 infected cervical cells is a function of the particular individual from whom the cells were obtained. Since all cultures were treated identically, the different response to the inactivated virus may be due to differences in genetic makeup. Alternatively, cells from women with neoplastic disease may be made susceptible to chromosomal imbalance independent of their genetic make-up. However, this is rather improbable as all cultures were diploid prior to infection and were kept in culture at least two months before exposure to the virus. The possibility that oncornaviruses are associated factors was suggested Vol. 52, No. 8, October 1976

928

L. AURELIAN

928~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

by the data of Peters et al. indicating they are essential cofactors in Marek's disease, a lymphoma of chickens caused by a herpesvirus. However, the presence of oncornaviruses in human cancers is not yet established. DEFECTIVE HERPESVIRUS PARTICLES

The etiology hypothesis must reconcile the observation that viral infection causes cell death29'35 with its hypothesized ability to cause cancer, i.e., cell immortality. The solution is the association between neoplasia and noninfectious virus. Thus, HSV-2 transforms cells only following UV inactivation,5 and cancer-like chromosomal modalities are caused only by UV inactivated virus; infectious virus pulverizes the chromosomes.8l.56a Further, the mean age for infection with HSV-2 precedes by six years that for the development of the earliest detectable neoplastic lesions,38 suggesting that if HSV-2 causes neoplasia it acts from a latent (noninfectious) state. How does a venereally transmitted agent become noninfectious? One possible interpretation is that the oncogenic virus is a rare defective particle in an infectious virus population. Alternatively, the defective state results from the modification of the infectious virus by the pressures of associated carcinogens. Such pressures would be most likely to operate during latency.35 According to the first interpretation, the woman infected with a strain of HSV-2 which is rich in defective particles will be most likely to develop cervical cancer. Other HSV-2 seropositive women will not develop neoplasia. This interpretation also explains why virus-inducing stress conditions that are operative in latency35 have no effect on the transformed cells. Kessler's data,17 associating cancer transmission to only certain males, supports this interpretation. The nature of the defective state is unknown. Particles that contain viral DNA consisting of tandemly repeated units of different sizes, such as those recently described by Frenkel et al.,82 or particles containing fragments of viral and cellular DNA78'1 would be defective. According to the second interpretation, a latent state is established by the genome of the infectious virus. Modification in the regulation of this genome would result in oncogenesis. Under these conditions a functional viral genome would be present at least in the precancerous lesions. Oncogenesis would be associated with the expression of some viral informa-


COITUS AND CANCER

929

929

tion that is not expressed during latency.3 The isolation of HSV-2 from a culture established from a precancerous lesion but not from cultures established from invasive cancer support this interpretation. HSV-1,

AN

ONCOGENIC VIRUS

HSV-1 has been associated with cancer of the lip by the studies of Wyburn-Mason84 and Kvascnika,85'86 demonstrating the development of carcinoma of the lip at the site of recurrent herpetic lesions, and with laryngeal cancer by those of Hollinshead et al.6566'69 HSV- 1 can transform hamster cells in vitro;5 however, extensive studies on human cancer specimens remain to be done. In view of the significant nucleic acid homology between HSV-1 and HSV-2,29 the possibility that HSV-1 may be potentially oncogenic in humans is not unexpected. This is of particular significance when considering the increased isolation of HSV-1 from the genitalia.87 Are herpesviruses associated with cancers at other sites? Other than the controversial nonvirion antigen data,64'68 HSV-2 has been associated with a vulvar atypia,77 particles of herpesvirus were observed in tumor cells in the prostate and breast,88'89 and HSV-2 was isolated from one culture of prostate tumor cells and shown to transform hamster cells in vitro.90 These associations remain to be reconciled with the noninfectious patterns of these cancers. It should be pointed out that Dmochowski et al.91 did not observe herpesvirus in prostatic tissue from 34 patients with cancer of the prostate. They describe particles resembling type C virus in three of these specimens.

CONCLUSIONS AND PERSPECTIVES

The data we have considered establish a strong link between HSV-2 and cervical cancer. Seroepidemiologic studies as well as virologic and immunologic inquiries into the validity of the predictions made by the etiology hypothesis gave rise to concordant results, thus building a strong case against this virus. However, much remains to be done. The development of technical approaches to overcome the problems raised by the search for viral nucleic acid deserves special effort. Large-scale, statistically significant clinical studies on the association of the various viral antigens with cervical cancer, including longitudinal analysis of the patients starting prior to therapy, are of the utmost significance. These antigens should be purified and their viral nature established. The possi-


930

L. AURELIAN

bility that they may be good immunodiagnostic tools should be considered. The animal studies are promising and might prove particularly important in the eventual consideration of preventive or therapeutic modalities. The associated factors should be identified and their interaction with HSV-2 in pathogenesis should be given special consideration. Humoral and cellular immune responses to specific viral antigens should be studied on a longitudinal basis in patients and controls. Finally, prospective epidemiologic studies2 deserve major attention as possibly the best and most direct approach to the solution of the etiological problem.

1.

2.

3.

4.

5.

6. 7.

8.

9.

REFERENCES ogy and Prevention: Current Concepts, Dowdle, W. R., Nahmias, A. J., HarSchottenfeld, D., editor. Springfield, well, R. W., and Pauls, F. P.: AssociaIll., Thomas, 1975, chap. 1 1. tion of antigenic type of herpesvirus hominis with site of viral recovery. J. 10. Lilienfeld, A. M., Levin, M. L., and Kessler, I. I.: Cancer in the United Immunol. 99:974-80, 1967. States. Cambridge, Mass., Harvard Kessler, I. I.: Perspectives on the epideUniversity Press, 1972. miology of cervical cancer with special references to the herpesvirus hypothesis. 11. Kessler, I. I.: Cervical cancer epidemiology in historical perspective. J. Cancer Res. 34:1091-1110, 1974. Reprod. Med. 12:173-85, 1974. Aurelian, L.: Persistence and expression of the herpes simplex virus type 2 12. Martin, C. E.: Marital and coital factors in cervical cancer. Am. J. Public Health genome in cervical tumor cells. Cancer 57:803-14, 1967. Res. 34:1126-35, 1974. Frenkel, N., Roizman, B., Cassai, E., 13. Rotkin, I. D.: Adolescent coitus and cervical cancer: Associations of related and Nahmias, A. J.: A DNA fragment events with increased risk. Cancer Res. of herpes simplex 2 and its transcripts in 27:603-17, 1967. human cervical cancer tissue. Proc. Nat. 14. Rotkin, 1. D.: Sexual characteristics of a Acad. Sci. 69:3784-89, 1972. cervical cancer population. Am. J. PubRapp, F. and Duff, R.: Transformation lic Health 57:815-29, 1967. of hamster embryo fibroblasts by herpes simplex viruses type 1 and type 2. 15. Martinez, I.: Relationship of squamous cell carcinoma of the cervix uteri to Cancer Res. 33:1524-34, 1973. squamous cell carcinoma of the penis Rigoni-Stern, I.: Fatti statistici relativi among Puerto Rican women married to alle malatie cancerose. G. Servire Prog. men with penile carcinoma. Cancer Pathol. Terap. 2:507-17, 1842. 24:777-80, 1969. rowne, J. E.: Carcinoma of the cervix In nulliparous and celibate women. Am. 16. Kurihara, M. and Asano, K.: The correlation between cancer of uterus and canJ. Obstet. Gvnecol. 68:606-13, 1955. cer of penis. Tohoku J. Exp. Med. Fraumeni, J. F., Lloyd, J. M., Smith, 64:104, 1956. E. M., and Wagoner, J. K.: Cancer mortality among nuns: Role of marital 17. Kessler, I. I.: Human cervical cancer as a venereal disease. Cancer Res. status in etiology of neoplastic disease in 36:783-91, 1976. women. J. Natl. Cancer Inst. 18. Rojel, J.: The interrelation between uter42:455-68, 1969. ine cancer and syphilis: A pathodermoKessler, I. I. and Aurelian, L.: Uterine graphic study. Acta Pathol. Microbiol. Cervix Cancer. In: Cancer Epidemiol-


COITUS AND CANCER

Scand. (Suppl.) 97:13-82, 1953. 19. Levin, M. L., Dress, L. C., and Goldstein, H.: Syphilis and cancer. N.Y. StateJ. Med. 42:1737-45, 1942. 20. Beral, V.: Cancer of the cervix: A sexually transmitted infection? Lancet 1:1037-40, 1974. 21. Meisels, A.: Dysplasia and carcinoma of the uterine cervix. IV. A correlated cytologic and histologic study with special emphasis on. vaginal microbiology. Acta Cytol. 13:224-31, 1969. 22. Royston, I. and Aurelian, L.: The association of genital herpesvirus with cervical atypia and carcinoma in situ. Am. J. Epid. 91:531-38, 1970. 23. Hutfield, D. C.: History of herpes genitalis. Brit. J. Vener. Dis. 42:263-68, 1966. 24. Greenough, F. B.: Herpes progenitalis. Arch. Dermatol. 7:1-29, 1881. 25. Unna, P. G.: On herpes progenitalis, especially in women. J. Cutan. Vener. Dis. /:321-34, 1883. 26. Hruszek, H.: Beitrag sum problem der naturlichen ubertragung des herpes virus von mensch zu mensch. Die venerische ansteckung. Dermatol. Wschr. 36: 1150-52, 1937. 27. Nahmias, A. J., Dowdle, W. R., Naib. Z. M., Josey, W. E., McLone, D., and Domescik, G.: Genital infection with type 2 herpes virus hominis. A commonly occurring venereal disease. Br. J. Vener. Dis. 45:294-98, 1969. 28. Rawls, W. E., Gardner, H. L., Flanders, R. W., Lowry, S. P., Kaufman, R. H., and Melnick, J. L.: Genital herpes in two social groups. Ami. J. Obstet. Gynecol. 110:682-89, 1971. 29. Nahmias, A. J. and Roizman, B.: Infection with herpes simplex virus 1 and 2. N. Engl. J. Med. 289:667-74, 1973. 30. Nahmias, A. J. Naib, Z. M., Highsmith, A. K., and Josey, W. E.: Experimental genital herpes simplex infection in the mouse. Pediatr. Res. 1:209, 1967. 30a. Kessler, I. I. and Aurelian, L. Unpublished. 31. Grin, E. I.: Epidemiology and Control of Endemic Syphilis. Report on a MassTreatment Campaign in Bosnia. Geneva, World Health Organization, 1953, pp. 9-93.


931

32. Centifanto, Y. M., Drylie, D. M., Deardourff, S. L., and Kaufman, H. E.: Herpesvirus type 2 in the male genitourinary tract. Science 178.318-19, 1972. 33. Goodwin, C. S., Gostling, J. V. T., Pead, P. J., and Sims, E. H.: Genital herpes. Br. Med. J. 4:558-59, 1970. 34. Jeansson, S. and Molin, L.: Genital herpes infection and non-specific urethritis. Br. Med. J. 3:247, 1971. 35. Roizman, B.: The Biochemical Features of Herpesvirus-Infected Cells, Particularly as They Relate to Their Potential Oncogenicity-A Review. In: Oncogenesis and Herpesviruses, Biggs, P. M., de The, G. and Payne, L. N., editors. Lyon, Int. Agency Res. Cancer, 1972. 36. Coppleson, M. and Reid, B.: The etiology of squamous carcinoma of the cervix. Obstet. Gvnecol. 32:432-36, 1968. 37. Singer, A.: Genital herpes and cervical cancer. Br. Med. J. 1:458,1971. 38. Naib, M., Nahmias, A. M., Josey, W. E., and Kramer, J. H.: Genital herpetic infection: Association with cervical dysplasia and carcinoma. Cancer 23:940-45, 1969. 39. Rawls, W. E., Adam, E., and Melnick, J. L.: An analysis of seroepidemiological studies of herpesvirus type 2 and carcinoma of the cervix. Cancer Res. 33:1477-82, 1973. 40. Aurelian, L.: Possible role of herpesvirus hominis, type 2 in human cervical cancer. Fed. Proc. 31;1651-59, 1972. 41. Aurelian, L.: Factors affecting the growth of canine herpesvirus in dog kidney cells. Appl. Microbiol. 17: 179-181, 1969. 42. MacGregor, J. E.: Cervical carcinoma. The beginning of the end? Lancet 11:1296-99, 1967. 43. Jones, H. W., Jr., Katayama, K. P., Stafl, A., and Davis, H. J.: Chromosomes of cervical atypia carcinoma in situ and epidemoid carcinoma of the cervix. Obstet. Gynecol. 30:790, 1967. 44. Koss, L. G.: Concept of genesis and development of carcinoma of the cervix. Obstet. Gynecol. Surv. 24:850-60, 1969. 45. Aurelian, L., Royston, I., and Davis, H. J.: Antibody to genital herpesvirus:

932

46.

47.

48.

49.

50.

51.

52.

53.

54. 55.

L. AURELIAN

Association with cervical atypia and carcinoma in situ. J. Nat. Cancer Inst. 45:455-64, 1970. Aurelian, L., Strandberg, J. D., and Davis, H. J.: HSV-2 antigens absent from biopsied cervical tumor cells: A model consistent with latency. Proc. Soc. Exp. Biol. Med. 140:404-08, 1972. Royston, I. and Aurelian, L.: Immunofluorescent detection of herpesvirus antigens in exfoliated cells from human cervical carcinoma. Proc. Nat. Acad. Sci. 67:204-17, 1970. Aurelian, L., Strandberg, J. D., Melendez, L. V., and Johnson, L. A.: Herpesvirus type 2 isolated from cervical tumor cells grown in tissue culture. Science 174:485-88, 1971. Aurelian, L., Davis, H. J., and Julian, C. G.: Herpesvirus type 2 induced tumor specific antigens in cervical carcinoma. Am. J. Epidem. 98:1-9, 1973. Aurelian, L., Schumann, B., and Marcus, R. L.: Antibody in HSV-2 induced tumor specific antigens in sera from patients with cervical carcinoma. Science 181:161-64, 1973. Aurelian, L. and Strand, B.: Herpesvirus type 2 related antigens and their relevance to humoral and cell mediated immunity in patients with cervical cancer. Cancer Res. 36:810-20, 1976 Gall, S. A. and Haines, H.: Cervical cancer antigens (CCA) and their relationship to herpesvirus. V. 5th Ann. Meeting Soc. Gynec. Oncologists. Key Biscayne, Fla., January 1974. Gall, S. A. and Haines, H.: Cervical carcinoma antigens and the relationship to HSV-2. Gynecol. Oncol. 2:451-59, 1974. Rabinowitz, Z. and Sachs, L.: Control of the reversion of properties in transformed cells. Nature 225:136-39, 1970. Burger, M. M.: The significance of surface structure changes for growth control under crowded conditions. In: Grcwth Control in Cell Cultures, Wolstenholme, G. E. W. and Knight, J., editors. Edinburgh, Churchill, Livingstone, 1971, pp. 45-62.

56. Heine, J. W., Spear, P. G., and Roizman, B.: Proteins specified by herpes simplex virus. VI. Viral proteins in the plasma membrane. J. Virol. 9:431-39, 1972. 56a. Katayama, P., Aurelian, L., and Jones, H.: Chromosomal changes in human cervical cells infected with UV-inactivated HSV-2. In preparation. 57. Moghissi, K. S.: Cyclic changes or cervical mucus in normal and progestin treated women. Fertil. Steril. 17:663-75, 1966. 58. Nazerian, K. and Purchase, H. G.: Combined fluorescent antibody and electron microscopy study of Marek's disease virus-infected cell culture. J. Virol. 5:79-90, 1970. 59. Herrera, I., Valenchiano, L., SanchezGarrido, F., and Botella-Llusia, J.: On findings of virus-like structures in uterine cervical carcinoma. Acta Cytol. 18:45-50, 1974. 60. Pacsa, A., Kummerlander, L., Pajtsik, B., and Poli, K.: Occurrence of herpesvirus antibodies and antigens in patients with cervical anaplasia and controls. J. Nat. Cancer Inst. 55:775-81, 1975. 61. Nahmias, A. J., Naib, Z. M., and Josey, W. E.: Herpesvirus Hominis Type 2 Infection Associated with Cervical Cancer and Prenatal Disease. In: Perspectives in Virology. VIII, Pollard, M., editor. New York, Academic Press, 1971, pp. 73-88. 62. Sprecher-Goldberger, S., Thiry, L., Gould, I., Fassin, Y., and Gompel, C.: Increasing antibody titers to herpes simplex virus type 2 during follow up of women with cervical dyplasia. Am. J. Epidem. 97:103-10, 1973. 62a. Boyd, A. L. and Orme, T. W.: Transformation of mouse cells after infection with ultraviolet irradiation-inactivated herpes simplex virus type 2. Int. J. Cancer 16:526-38, 1975. 63. Collard, W., Thornton, H., and Green, M.: Cells transformed by human herpesvirus type 2 transcribe virus specific RNA sequences shared by herpesvirus types 1 and 2. Nature New Biol. 243:264-66, 1973. 64. Sabin, A. B. and Tarro, G.: Herpes


COITUS AND CANCER

simplex and herpes genitalis viruses in etiology of some human cancers. Proc. Nat. A cad. Sci. 70:3225-29, 1973. 65. Hollinshead, A. C. and Tarro, G.: Soluble membrane antigens of lip and cervical carcinoma: Reactivity with antibody to herpes virus nonvirion antigens. Science 179:698-700, 1973. 66. Hollinshead, A. C., Lee, 0. B., McKelway, W., Melnick, J. L., and Rawls, W. E.: Reactivity between herpesvirus type 2 related soluble cervical tumor cell member antigens and matched cancer and control sera. Proc. Soc. Exp. Biol. Med. 141:688-93, 1972 67. Aurelian, L., Strandberg, J. D., and Marcus, R. L.: Neutralization, immunofluorescence and complement fixation tests in identification of antibody to a herpesvirus type 2-induced, tumor-specific antigen in sera from squamous cervical carcinoma. Progr. Exp. Tumor Res. 19:165-81, 1974. 68. Sabin, A. B.: Herpes simplex-genitalis virus nonvirion antigens and their implication in certain human cancers. Proc. Natl. Acad. Sci. 71:3248-52, 1974. 69. Hollinshead, A. C., Lee, O., Chretian, P. B., Tarplay, J. L., Rawls, W. E., and Adam, E.: Antibodies to herpesvirus nonvirion antigens in squamous carcinomas. Science 182:713-15, 1973. 70. Anzai, T., Dreasman, G. R., Courtney, R. J., Adam E., Rawls, W. E., and Benyesh-Melnick, M.: Antibody to herpes simplex virus type 2 induced nonstructural proteins in women with cervical cancer and control groups. J. Natl. Cancer Inst. 54:1051-59, 1975. 71. Palmer, A., London, W. T., Nahmias, A. J., Naib, Z. M., Tunca, Y., Fucillo, D. A., Ellenberg, J. H., and Sever, J. L.: Investigation of the oncogenic potential of herpes simplex virus type II in cebus monkeys: A preliminary report. Cancer Res. 36:807-09, 1976. 72. Aurelian, L., Smerdel, S. S., and Kessler, I. I.: Three Yugoslav herpes simplex viruses: Biologic and antigenic properties and formation of giant cells in vitro by a cervical isolate. J. InJect. Dis. 129:456-69, 1974. 73. Hausen, H. zur, Schulte-Holthausen,


933

H., Wolf, H., Dorries, K., and Egger, H.: Attempts to detect virus-specific DNA sequences in human tumors: II. Nucleic acid hybrizations with complementary RNA of human herpes group viruses. Int. J. Cancer 13:657-64, 1974. 74. Henle, W. and Henle, G.: Evidence for an oncogenic potential of the EpsteinBarr virus. Cancer Res. 33:1419-23, 1973. 75. Aurelian, L., Strandberg, J. D.: Biologic and immunologic comparison of two HSA-2 variants: One an isolate from cervical tumor cells. Arch. Gesamte Virusforsch. 45:27-38, 1974. 75a. Kingsbury, D. and Boyd, A. L.: Unpublished. 76. Kessler, I. I., Kulcar, Zimolo, A., Grgurevic, M., Strnad, M., and Goodwin, E. J.: Cervical cancer in Yugoslavia. II. Epidemiologic factors of possible etiologic significance. J. Natl. Cancer Inst. 53:51-60, 1974. 77. Friedrich, E. G., Jr.: Reversible vulvar atypia. Obstet. Gynec. 39:173-81, 1972. 77a. Notter, M. F. D. and Docherty, J. J.: Comparative diagnostic aspects of herpes simplex tumor-associated antigens. J. Nat. Cancer Inst. In press; Kawana, T., Cornish, J. D., Smith, M. F., and Aurelian, L.: Frequency of antibody to a virus-induced, tumor-associated antigen (AG-4) in Japanese sera from patients with cervical cancer and controls. Cancer Res. 36:1910-14, 1976. 77b. Tarro, G.: Analysis and Description of Procedures Used in the Study of the Relationship of Herpes Simplex Virus "Non-virion" Antigens to Certain Cancers. In: Oncogenesis and Herpesviruses II. De The, G., Epstein, M. A., and zur Hausen, H., editors. Lyon, Int. Agency Res. Cancer, 1975, part 2, pp. 291-97. 78. Tooze, J.: The Molecular Biology of Tumor Viruses. Cold Spring Harbor, N.Y., Cold Spring Harbor Laboratory, 1973. 78a. Docherty, J. J.: Personal communication. 79. Peters, W. P., Kufe, D., Scholm J., Frenkel, J. W., Prickett, C. O., Groupe, V., and Spiegelman, S.: Bio-

934

L. AURELIAN

logical and biochemical evidence for an interaction between Marek' s disease herpesvirus and avian leukosis virus in vivo. Proc. Natl. Acad. Sci.

87.

70:3175-78, 80. Munoz, N.: Effect of herpesvirus type 2 and hormonal imbalance on the uterine cervix of the mouse. Cancer Research 88. 33:1504-08, 1973. 81. Stich, H. F., Hsu, T. C., and Rapp, F.: Viruses and mammalian chromosomes. I. Localization of chromosome aberra- 89. tions after infection with herpes simplex virus. Virology 22:439-45, 1964. 82. Frenkel, N., Jacob, R. J., Honess, R. W., Hayward, G. S., Locker, H., and Roizman, B.: Anatomy of herpes sim- 90. plex virus DNA. III. Characterization of defective DNA molecules and biological properties of virus populations containing them. J. Virol. 16: 153-67, 1975. 83. Aurelian, L. and Wagner, R. R.: Two 91. populations of herpes virus virions which appear to differ in physical properties and DNA composition. Proc. Natl. Acad. Sci. 56:902-09, 1966. 84. Wyburn-Mason, R.: Malignant change following herpes simplex. Br. Med. J. 92. 2:615-16, 1957. 85. Kvasnica, A.: Relationship between herpes simplex and lip carcinoma. III. Neoplasma 10:199-203, 1964. 86. Kvasnicka, A.: Relationship between

herpes simplex and lip carcinoma. IV. Selected cases. Neoplasma 12:61-70, 1965. Smith, 1. W., Peutherer, J. F., and Robertson, D. H. H.: Characterization of genital strains of herpesvirus hominis. Br. J. Vener. Dis. 49:385-90, 1973. Tannenbaum. M. and Lattimer, J. K.: Similar virus-like particles found in cancers of the prostate and breast. J. Virologv 103:471-75, 1970. Centifanto, Y. M., Kaufman, H. E., Zam, Z. S., Drylie, D. M., and Deardourff, S. L.: Herpesvirus particles in prostatic cells. J. Virol. 12:1608-11, 1973. Centifanto, Y. M. and Kaufman, H. E.: Significance of herpesvirus in prostatic carcinoma. Second ntm. Svmp. Oncogenesis and Herpesi-irus. Nuremberg, 1974. Dmochowski, L., Maruyama, K., Ohtuske, Y., Seman, G., Bowen, J. M., Newton, W. A., and Johnson, D. E.: Virologic and immunologic studies of human prostatic. carcinoma. Cancer Clhernother. Rep. 59:17-31, 1975. Boyd, A. L. and Orme, T. W.: Transformation of mouse cells after infection with UV-irradiation inactivated herpes simplex virus type 2. lInt. J. Cancer 16:666-76, 1975.
