Sexually Transmitted Diseases Including Aids

Mitchell Bernstein, MD
Attending, Division of Colon & Rectal Surgery
St. Luke's/Roosevelt Hospital Center
Assistant Professor of Clinical Surgery
Columbia University College of Physicians & Surgeons
New York, NY

Introduction

Sexually transmitted diseases (STDs) continue to increase worldwide and currently include numerous bacterial, viral, fungal and protozoan agents. More than 50 organisms have been classified as sexually transmitted. In 1995, five of the ten most common diseases reported to the Center for Disease Control and prevention were sexually transmitted. These diseases come with a significant price tag: the estimated cost of treating pelvic inflammatory disease alone in 1990 was 4.2 billion dollars. The United States has among the world's highest rates of STDs with an ever-increasing incidence of STDs affecting the anorectum, making it essential for the practicing colorectal surgeon to have a thorough knowledge of these diseases.

Accurate diagnosis and treatment of STDs of the anorectum is difficult due to the fact that patients often harbor more than one organism. Distinguishing between what is acting as a pathogen and what has merely colonized the anorectum and not causing disease is not always easy. For example, Chlamydia is found is 15% of asymptomatic homosexual men and up to one third of homosexuals have anorectal herpes simplex virus infection. The HIV epidemic has only complicated the diagnosis and treatment of STDs. Of note, it is the population of HIV positive patients with good functional status that are particularly apt to harbor several organisms yet remain asymptomatic.

Below is a table that summarized the symptoms physical findings, diagnostic tests and treatment modalities for these STDs:

Bacterial Infections

Gonorrhea
Gonorrhea is likely the most common STD affecting the anorectum. It is caused by the Gram-negative intracellular diplococcus Neisseria gonorrhoeae. About 3 million people per year are infected with gonorrhea with as many as 55% of homosexual men being infected. At least 50% of male patients and up to 95% of female patients with rectal gonorrhea are asymptomatic, accounting for a large pool of disease carriers in the community. A significant percentage of women diagnosed with gynecologic gonorrhea will also have gonorrhea in the anorectum. Only 4% of women with gonorrhea will have the anorectum as the only site of involvement whereas 40-50% of homosexual men will have the anorectum as the exclusive site of infection.

The incubation period for gonorrhea is 5-7 days but may be as long as 30 days. When symptoms develop they vary in severity. Anorectal inoculation usually produces a distal proctitis with an associated cryptitis. This can manifest as a non-specific pruritus or tenesmus accompanied by a mucoid and sometimes bloody rectal discharge. Left untreated, a more advanced systemic infection can occur resulting in such conditions as endocarditis, pericarditis, and a unilateral migratory gonococcal arthritis of large joints.

Anorectal gonorrhea almost uniformly results in a thick yellow mucopurulent discharge from the anus. This purulent material is often able to be expressed from the anal crypts and can be visualized directly by applying gentle external pressure while viewing the distal anorectum through an anoscope. Sigmoidoscopy often reveals a proctitis extending to no more than about 10 cm from the anal verge. The distal rectum is often congested and edematous with the mucosa being covered by a thick, creamy tenacious pus. Although these findings may be mistaken for an ulcerative or non-specific proctitis, abscesses, fistulae and ulcers are not typical.

Diagnosis is made by culturing the gram-negative organism on a Thayer-Martin medium plate incubated in a carbon dioxide environment. Gram stains of the rectal discharge are unreliable, with a relatively high false negative rate. However, swabbing the pus directly from the anorectum under direct visualization increases the positive yield of a Gram stain to about 80%. When obtaining a specimen for Gram stain or culture from the anorectum the anoscope or proctoscope should not be lubricated with anything other than water as many lubricating agents contain antibacterial compounds. Because of the unreliability of Gram stains a high index of suspicion warrant empiric treatment pending final culture results.

Due to the increasing prevalence of penicillinase-producing Neisseria gonorrhoeae treatment is no longer with an intramuscular injection of 4.8 million units of aqueous procaine penicillin G and 1 gram of oral probenecid. First line treatment is now a single intramuscular injection of 250 mg of ceftriaxone followed by doxycycline, 100 mg orally twice a day for 7 days. Alternative treatments that have been proposed include single intramuscular injections of spectinomycin (2g) or cefoxitin followed by the oral dose of probenecid. Multiple dose regimens of ciprofloxacin, trimethoprim-sulfamethoxazole, tetracycline, ampicillin and ceftriaxone have also been suggested. Even with reliable therapy, recurrence or failure of initial treatment occurs in up to 35% of patients. Thus, a follow-up examination within three months of treatment is necessary. Disappearance of symptoms alone is an unreliable determination of eradication of the organism.

Chlamydia/lymphogranuloma venereum
Chlamydia trachomatis is the most common sexually transmitted bacterial infection worldwide. In the United States, male homosexuals account for the majority of rectal Chlamydial infections. With 15 known serotypes, Chlamydia infection can be asymptomatic or cause proctitis or lymphogranuloma venereum (LGV). In one study 43% of males and 79% of females infected with Chlamydia had no symptoms. Alternatively, Chlamydia proctitis can occur within 10 days of anoreceptive intercourse, with an incubation period ranging from 5 days to 2 weeks. Serotypes D through K are responsible for causing proctitis while serotypes L1-L3 are known to cause LGV.

Symptoms of the non-LGV proctitis include rectal pain, tenesmus, and fever. Examination reveals enlarged matted inguinal lymph nodes and sigmoidoscopy shows an erythematous rectal mucosa without frank ulcerations. In LGV patients also complain of pain, fever and tenesmus but often have a slight mucopurulent discharge and hematochezia. Constipation or loose stools may be associated as well. The inguinal lymph nodes in LGV are often more enlarged as they fuse into a large indurated mass with overlying erythema. In addition, LGV causes perianal ulcerations. Sigmoidoscopy reveals a more severe granular proctitis with mucosal friability and frank ulceration. Left untreated, the disease may progress to fistulae, abscesses and late rectal strictures. In this setting, LGV may be confused for perianal Crohns disease. It is the marked inguinal lymphadenopathy that helps distinguish LGV from Crohns disease.

Diagnosis is difficult as the obligate intracellular nature of Chlamydia make culture unrewarding. When performed, biopsy of the inflamed rectal mucosa should be transported in a sucrose phosphate media on ice for immediate tissue culture. Alternatively, diagnosis may be made with serum antibody titers, however, titer elevation often occurs more than one month after infection.

Treatment of a chlamydial infection is with oral tetracycline or erythromycin, 500 mg four times a day. Treatment of symptomatic strictures should be with at least a three-week course of antibiotics; however, proximal diversion or resection have been reported for treatment failures.

Syphilis
Primary anal syphilis, caused by the spirochete Treponema pallidum, is largely a disease of homosexual men. Colorectal surgeons will typically encounter syphilitic infections as a chancre or proctitis in the primary stage or as condyloma lata in the secondary stage. The organism is introduced into the anorectum during anoreceptive intercourse and causes ulcers within 2-6 weeks. In 10-20% of cases the primary chancre will be hidden within the anal canal. The chancre begins as a maculopapular lesion that soon ulcerates. It may be mistaken for a common fissure. Unlike chancres that appear on the external genitalia, anal chancres are quite painful. Location off the midline, a more external location away from the anal verge or internally above the dentate line help distinguish these lesions from idiopathic anal fissures. In addition, multiple lesions that have irregular borders and appear opposite one another in a mirror image ("kissing") are often seen. Proctitis without associated chancres has been reported.

Left untreated, the chancres regress spontaneously in 3-4 weeks with the secondary stage appearing 2-10 weeks later. A diffuse maculopapular rash classically appears on the palms of the hand and soles of the feet. Alternatively, secondary syphilis can present as a pale brown or pink, flat verrucous lesion: condyloma latum. Many lesions may coalesce and secrete mucus causing a foul odor as well as pruritus.

Diagnosis is by demonstration of the corkscrew-shaped yellow-green spirochetes on darkfield examination. Biopsy of the rectal lesions may show the spirochetes on Warthin-Starry silver stain. Treatment of the early stages of syphilis is with a single dose of 2.4 million units of long-acting benzathine penicillin given intramuscularly. In the later stages or if the patient is HIV+, three consecutive doses given at two-week intervals is recommended. Follow-up testing with VDRL or RPR should be repeated at three-month intervals for one year.

Chancroid (see Table)

Granuloma inguinale (see Table)

Viral Infections

Herpes simplex
Anogenital herpes simplex virus (HSV) infections are on the rise with an estimated ninefold increase between 1966 and 1990. The majority of anorectal herpes infections are caused by HSV-2 with only about 10% being caused by HSV-1. Transmission is through autoinoculation or direct contact with an infected individual. After local inoculation, the virus travels along peripheral nerves to the sensory neuron's nucleus.

A primary infection may cause an initial tingling sensation at the viral entry point with subsequent eruption of the classic vesicles on a mucuos membrane or skin surface. The clinical infection may begin 4-21 days following anoreceptive intercourse. Prodromal symptoms include minor local irritation, burning and paraesthesias in the anorectal area. The irritation escalates to an intense pain with the vesicles being extremely painful to the touch. A painful proctitis, pain radiating down the thigh or to the base of the scrotum and pain with urination are often the result of a sacral radiculitis. Along with proctitis, rectal bleeding and seropurulent discharge are common.

The vesicles are typically red and small and may be clustered or scattered in the perianal skin, anal canal or perineum. Shallow perianal ulcers can coalesce and extend to the sacrococcygeal region in a butterfly pattern. Anoscopy reveals friable mucosa, ulceration and the mucopurulent discharge. These findings will be limited to the distal 10 cm on proctoscopy.

Diagnosis is made on the clinical examination, finding multinucleated giant cells with intranuclear inclusion bodies on Pap smear, a positive Tzank preparation or a positive culture. Viral culture of a vesicle will be positive in 90% of infections. Treatment of an acute infection is with acyclovir. Acyclovir is available in topical, oral and IV forms with IV treatment reserved for severe cases. The oral form is far superior to topical treatment and is given as 200-400 mg five times a day for ten days. Most patients get relief of pain within 2-4 days of beginning oral treatment. Of note, acyclovir-resistant strains of HSV is well documented in the HIV+ subset of patients, however, it has recently been isolated in HIV- patients as well. Foscarnet or vidarabine is the treatment for acyclovir-resistant HSV.

Human papilloma virus
Human papilloma virus (HPV) is one of the most common STDs in the United States with 10-15 million infected people. The resulting condyloma acuminata are the most common STD seen by the colon and rectal surgeon. More than 60 subtypes of HPV have been identified with types 6 and 11 most commonly associated with the benign, exophytic condylomata of the anogenital region. Types 16 and 18 have been associated with more aggressive lesions that can become invasive squamous cell cancers.

The primary mode of transmission of anogenital HPV is sexual intercourse. Inoculation of the anal epithelium allows entry of the HPV into the basal cell layers. As these cells proliferate viral replication occurs in the nucleus. The basal cells migrate toward the surface and infective particles are released in the form of visible warts. Mature infectious particles are found in the surface layers of these lesions.

Condyloma acuminata are easily recognized as epithelialized cauliflower-like projections. They range in size from millimeters to a large Buschke-Lowenstein type lesion. They may be flat, raised, sessile or pedunculated. They may be in clusters or grow to cover the perineum and anal canal in a "carpet-like" fashion. The warts may be asymptomatic are cause pruritus, bleeding, or discharge. Close examination reveals that the warts are confined to the perianal area in just 6% of cases with the majority of patients having internal as well as external warts. Treatment of external warts without concomitant treatment of internal warts leads to treatment failure. Even with appropriate therapy warts are notorious for recurrence.

Treatment options include excisional therapy, destructive therapy and immunotherapy:

Molluscum contagiosum (see Table)

HIV and AIDS
Surgery for anorectal disorders remain among the most common procedures performed in HIV+ individuals. However, unlike in recent years in which AIDS-specific problems dominated the clinical picture, such conditions are becoming increasingly uncommon as HIV+ patients' performance status as a group is improving. In large measure due to highly active antiretroviral therapy (HAART), viral loads are lower and T-cell counts are higher, resulting in fewer anorectal problems that are unique to HIV+ patients. Nonetheless, issues unique to this population arise and must be considered when dealing with the more common anorectal conditions that afflict the population as a whole, such as fissures and condyloma. The anorectal conditions that afflict HIV+ patients generally fall into one of two broad categories. The first is a "routine" condition that occurs in the general population at large and the second is a disorder that is unique to HIV+ individuals. However, even "routine" disorders can present in uncommon ways in HIV+ patients. Moreover, treatment plans must take into consideration the effect of prior repetitive anoreceptive intercourse on sphincter function, diarrhea (infectious or as a side effect of HAART) and prior sphincter cutting surgery.

Anal Condyloma
Anal condyloma is the most common anorectal condition seen in HIV+ patients. The apparent infestation of Human papilloma virus (HPV) in the practicing homosexual population is likely to be a confounding factor in a variety of pathologic conditions in this population, including the increasing cases of anal cancer. As mentioned, subtypes 16 and 18 have been associated with malignant transformation. The E6 and E7 genes of these subtypes have been shown to be incorporated into the host genome and bind to the p53 and Rb protein, respectively. Loss of these tumor-suppressor genes is thought to be a contributing factor in malignant transformation. Moreover, HIV is thought to be a cofactor in E6 and E7 expression.

Clinically apparent warts should be aggressively treated. In our practice we prefer a combination of scissors excision and fulguration with electrocautery following acetic acid staining. In addition, for refractory, recalcitrant warts, a significantly decreased time to and severity of recurrence has been achieved with injection of ten million units of alpha-interferon 2b (Intron(, Schering Corporation, Kenilworth, NJ) into the operative bed following completion of surgery. Interferon injection and fulguration combine the antiproliferative and antiviral properties of interferon with formal cytodestruction. Alternatively, topical Imiquimod(r) is used in conjunction with surgery. Imiquimod(r) induces alpha-interferon and other cytokines. Unfortunately, while these approaches help to maximize eradication of both clinical and subclinical disease, high recurrence rates remain a vexing problem.

More problematic than clinical recurrence, however, is the high incidence of anal intraepithelial neoplasia (AIN) seen in these patients. Initially, many likened AIN to cervical intraepithelial neoplasia (CIN) and like CIN, a precursor to cervical cancer, were concerned about transformation of AIN to squamous cell carcinoma. While two small studies have failed to demonstrate a progression of AIN to squamous cell cancer in these patients, a large longitudinal study that clearly documents the natural history of AIN in HIV+ patients has not been done. What is known is that in the HIV+ patient, HPV infection may occur in with a clinically normal appearing epithelium and that progression to AIN appears to be related to the level of immunosuppression rather than the specific subtype of HPV present.

Such confusion surrounding the issue of AIN has led to controversy about how to treat it. Some advocate ablative surgery. This can lead to debilitating surgery, as the AIN is often extensive. This approach is also technically problematic as the AIN is often within clinically normal appearing tissues achieving "negative margins" in the operating room may be difficult. It is our practice to observe these patients closely at three-month intervals at which time suspicious lesions can be biopsied to check for progression of disease.

Ulcerative Disease
Ulcerative processes of the anorectum remain the most common non-condylomatous anorectal condition leading to surgery in HIV+ patients. Such ulcers run the spectrum from truly "benign" anal fissures to idiopathic AIDS-related ulcerations of the anal canal. Both types of processes can be debilitating in their own right and effective treatment of these ulcers is predicated upon proper identification of the process involved. While "historically" this distinction had not been made and contributed to the poor results initially obtained in treating these conditions, much has been learned in recent years and current literature has helped in clearly defining idiopathic AIDS-related ulcerations of the anal canal as distinct entities.

Similar to the general population, HIV+ patients are susceptible to developing "benign" fissures. Such a fissure is indistinguishable from fissures seen in the seronegative population. The fissure is typically located in the posterior or anterior midline and may be associated with a hypertrophic anal papilla and/or a sentinel tag. Gentle spreading of the buttocks usually allows visualization of the fissure and reveals it to be a shallow ulcer with the circular fibers of the internal anal sphincter forming its base. The benign anal fissure never goes proximal to the dentate line.

Although a benign anal fissure in an HIV+ patient is clinically indistinguishable from a similar fissure in a seronegative individual, the etiology of the fissure is likely to be different. It has long been recognized that hypertonicity of the internal anal sphincter plays a role in benign fissure disease. However, patients with a history of repetitive anoreceptive intercourse are unlikely to have hypertonicity of the internal sphincter as such activity has been shown to decrease anal canal resting pressures. Thus, other etiologies must be considered. Trauma secondary to anoreceptive intercourse, diarrhea secondary to an infectious source or more common today, as a side effect of antiretroviral medication, may predispose HIV+ patients to developing fissures.

Once identified as such, benign anal fissures in HIV+ patients should be treated according to the same treatment algorithm applied to the seronegative population. Treatment begins with a trial of stool softeners, bulking agents, modifications of diet and sitz baths for comfort. In addition, abstinence from anoreceptive intercourse should be encouraged and treatment of any underlying diarrheal state should be instituted. As in the seronegative population, the majority of HIV+ patients will achieve symptomatic relief as well as healing of their fissure with such a regimen.

Should medical management of a benign anal fissure fail to achieve symptomatic improvement, surgery should be considered. However, prior to recommending a lateral internal sphincterotomy, consideration must be given to prior anorectal surgery that may have already compromised the internal anal sphincter, as well as the presence of any concomitant diarrhea. Division of any additional internal sphincter under such circumstances may render a patient incontinent. With these factors in mind, there is now ample evidence to demonstrate that HIV+ patients with clearly identified benign fissures have symptomatic relief and adequate wound healing following sphincterotomy. Even before the development of HAART, low morbidity with respect to both wound healing and incontinence was achievable. Now, as patients' performance status continue to improve and viral loads remain low and T-cell counts high, concerns regarding wound healing will likely become less pressing in HIV+ patients. Thus, sphincterotomy remains a viable option in selected HIV+ patients with benign anal fissures.

True AIDS-related ulcers can be extremely debilitating. Clinical features distinguishing these ulcers from benign anal fissures include proximal location in the anal canal, often proximal to the dentate line, as well as atypical positions in the anal canal; more often than not, they are off the midline. These ulcers are highly erosive in nature resulting in deep, broad-based cavities with over-hanging edges and occult pockets. The erosive nature of these ulcers results in transgression of normal sphincter planes into the deep or superficial post anal space. The base of the ulcer is often the external sphincter or ischiorectal fossa. Invasion into the ischiorectal space may cause life-threatening hemorrhage with erosion into the inferior rectal arteries. Such sphincter destruction often results in a patulous appearing anus. While benign anal fissures occur throughout the spectrum of HIV disease, idiopathic anal ulcers almost universally occur in advanced stages of AIDS, when CD4 counts drop below 200. Thus, what was once a very common condition and source of significant morbidity in HIV+ patients, AIDS ulcers are becoming increasingly uncommon clinical entities due to effective medications resulting in higher CD4 counts for longer periods of time in HIV+ patients.

The most common symptoms caused by these ulcers are pain with defecation and perianal pressure. The pain is in part due to the pain associated with a tear in the anal mucosa. In addition, the erosive nature of these ulcers with mucosal bridging leads to subcutaneous pockets in which stool and pus accumulate. Such collections cause pain and pressure in the anorectum and pelvis that may radiate down the legs. The typical complaint is one of intermittent pain and pressure alternating with spontaneous relief as the pockets decompress themselves.

Potential etiologies of perianal ulcers in HIV+ patients include infectious agents such as herpes simplex virus (HSV), cytomegalovirus (CMV), acid fast bacilli (AFB), HIV, as well as malignant conditions such as squamous cell carcinoma or non-Hodgkin's lymphoma. Despite, a thorough search for a causative factor, a specific etiology cannot be identified in up to 50% of HIV+ patients with anorectal ulcers. It is these ulcers that are deemed "idiopathic" and are the source of such morbidity in AIDS patients.

HSV is the most common virus affecting the anal canal of HIV+ patients, becoming more prevalent in the anal canal as the CD4 count drops. Several authors have implicated HSV as the most common cause of perianal ulceration in HIV+ patients.

While HSV can cause perianal ulceration, it does not cause large, broad-based ulcers. HSV can form vesicles that can coalesce to create large denuded areas of skin. These denuded areas are very shallow superficial ulcers, not deep and erosive. As HSV has been shown to be nearly ubiquitous in AIDS patients, when HSV is isolated from the large, broad-based, deeply erosive ulcers it should be considered more of as a "contaminant" than an etiologic agent of the ulcer.

Similarly, CMV is nearly as ubiquitous in the anal canal of the HIV+ patient as HSV. Although certain authors have attributed the deeply erosive AIDS-related ulcers to CMV infection, others have not been able to implicate CMV at all as a primary cause of ulcerative disease in HIV+ patients. Moreover, the lack of results with therapy aimed at CMV (i.e. failure to heal the ulcers or achieve symptomatic relief), supports the notion that CMV is just another non-invasive inhabitant of the granulation tissue of these ulcers.

Finally, some have suggested the human immunodeficiency virus itself as the causative agent in these AIDS-specific ulcerations of the anal canal. These investigators have identified a higher proportion of HIV particles in the base of these ulcers than in the surrounding tissue.

Following a thorough examination in which an infectious or malignant etiology has been ruled out, management of HIV+ patients with symptomatic idiopathic ulcerations of the anal canal is directed at relieving the debilitating symptoms. Symptoms are most often due to the pocketing of stool and pus in the deep and occult cavities, resulting in pain and pelvic pressure. Surgical management is therefore directed at eliminating these reservoirs. Aggressive debridement of over-hanging edges, unroofing of cavities, removing mucosal bridges and drainage of all collections is key to achieving symptomatic relief. Extrapolating from our early experience with intralesional steroid injections, it is commonplace in the author's practice to routinely inject the base of these ulcers with 80-160 mg Depo-Medrol( (Steris Laboratories, Inc., Phoenix, AZ) following debridement. Patients uniformly report improvement following this approach to the management of these ulcers. If patients develop pain following debridement and examination reveals no evidence of undrained sepsis, additional injections may be given in the office at two-week intervals. It has been our experience that patients rarely need more than two treatments.

A final option in patients with symptomatic idiopathic AIDS ulcers who do not achieve relief with adequate debridement and steroid injection is the use of Thalidomide (available through the FDA, Washington, DC). Thalidomide( acts as a cytokine downregulator. Several patients have not only achieved symptomatic relief with Thalidomide, but have also gone on to heal their ulcers. Whether this is a function of Thalidomide or improved immune status secondary to HAART (or a combination of the two) remains to be seen.

References

1. Modesto VL, Gottesman L. Sexually transmitted disease and anal manifestations of AIDS. Surgical Clinics: Anorectal Surgery 1994; 74: 1433-64.

2. Deheragoda P. Diagnosis of rectal gonorrhea by blind anorectal swabs compared with direct vision via a proctoscope. Br J Vener Dis 1977; 53: 311-313.

3. Schwarca SK, Zenilman JM, Schnell D et al. National surveillance of antimicrobial resistance in Neisseria gonorrhoeae. JAMA 1990; 264: 1413-17.

4. Quinn TC, Gaydos C, Shepard M et al. Epidemiologic and microbiologic correlates of Chlamydia trachomatis infection in sexual partnerships. JAMA 1996; 276: 1737-42.

5. Schacter J. Chlamydial infections. N Engl J Med 1978; 298: 428-35.

6. Hughes E, Cuthbertson AM, Killingback MK (eds). Venereal diseases of the anal canal and rectum. In: Colorectal Surgery. New York: Churchill Livingstone 1983: 203-8.

7. Akdamarr K, Martin RJ, Ichinase H. Syphilitic proctitis. Dig Dis 1977; 22:710-14.

8. Jacobs E. Anal infections caused by herpes simplex virus. Dis Colon Rectum 1976; 19: 151-7.

9. Wexner SD. Sexually transmitted disease of the colon, rectum and anus. Dis Colon Rectum 1990; 33: 521-31.

10 Bernstein MA. Anal fissure and the Human Immunodeficiency Virus. Seminars in Colon & Rectal Surgery 1997; 8: 40-45.

11. Horn RD, Hood AF. Cytomegalovirus is predictably present in perianal ulcers from immunocompromised patients. Arch Derm 1990; 126: 642-644.

12. Gottesman L. Ulcerative disease of the anorectum in AIDS. Int J STD AIDS. 1995; 6: 4-6, (editorial review).

13. Puy-Montbrun T, Denis J, Ganansia R, et al. Anorectal lesions in human immunodeficiency virus infected patients. Int J Colorect Dis 1992; 7: 26-30.

14. Pare A, Gottesman L. Anorectal diseases. Gastroenterol Clin N Am 1997; 26: 367-376.

15. Werness BA. Association of human papillomavirus types 16 and 18 E6 proteins with p53. Science 1990; 248: 76-79.

16. Buonaguro FM, Tornesello ML, Buonaguro L, et al. Role of HIV as cofactor in HPV oncogenesis:In vitro evidence of virus interactions. Anibiot Chemother 1994; 46: 102-109.

17. Surawicz CM, Critchlow CW, Sayer J, et al. High grade anal dysplasia in visually normal mucosa in homosexual men: seven cases. Am J Gastro 1998; 90: 1776-1778.

18. Metcalf AM, Dean T. Risk of dysplasia in anal condyloma. Surgery 1995; 118: 724-726.

19. Critchlow CW, Surawicz CM, Holmes KK, et al. Prospective study of high grade anal squamous intraepithelial neoplasia in a cohort of homosexual men: Influence of HIV infection, immunosuppression and human papillomavirus infection. AIDS 1995; 9:1255-1262.

20. Northfelt DW. Anal neoplasia in persons with HIV infection. AIDS Clin Care 1996; 8: 63-66.