BOOK EXCERPT
Esophageal Infections and Other Human Immunodeficiency Virus-Associated Esophageal Disorders
C. Mel Wilcox, MD

CHAPTER 14

As the indications for organ Transplantation expand, the number of chemotherapeutic regimens increases, and our population ages, esophageal infections will remain important. However, since 1996, the incidence of esophageal infections in patients with the acquired immunodeficiency syndrome AIDS) has fallen dramatically,¹ and the prognosis of these patients has improved² due to the widespread use of highly active antiretroviral therapy (HAART). Nevertheless, these human immunodeficiency virus (HIV)-infected patients will continue to present with esophageal infection as the index of manifestation of HIV infection or will fail HAART, leaving them at risk for these esophageal disorders. The number of patients with esophageal infections complicating organ Transplantation remains relatively low, primarily due to the targeted use of antifungal and antiviral prophylaxis in high-risk patients as well as improvements in immunosuppressive regimens. ^3-7 The ability to effectively treat essentially all of these esophageal disorders underscores the importance of an accurate diagnosis.^8,9 Because esophageal infections are best diagnosed endoscopically, the focus here is on clinical features relevant to the endoscopist, the endoscopic manifestations of these diseases, appropriate diagnostic methods, and the recognition and treatment of complications from these disorders. The endoscopic and histopathologic features of a specific disorder may vary based on the underlying cause of immune suppression. Thus, where possible, contrasts and comparisons will be drawn.

Approach to the Patient

A variety of factors impact the approach to the endoscopic examination. The differential diagnosis should be based on the cause and severity of immunodeficiency, character of esophageal complaints, and findings on physical examination, particularly of the oropharynx. For HIV-infected patients, the absolute value of the CD4 lymphocyte count stratifies the risk for an opportunistic disorder.¹⁰ Odynophagia is the most common presenting symptom of esophageal infection. Dysphagia is less frequent but may be reported, particularly with Candida esophagitis. Bleeding associated with esophageal complaints may be the initial manifestation and suggests esophageal ulceration. Importantly, disorders causing ulceration almost uniformly cause odynophagia. In contrast to esophageal candidiasis where oropharyngeal candidiasis (thrush) is common, oropharyngeal ulceration is rarely observed in patients with esophageal ulcers.^11-17 Nevertheless, the presence of oropharyngeal candidiasis does not prove Candida is the only cause, nor does the absence of oropharyngeal candidiasis exclude Candida esophagitis.^12,17 In patients with AIDS, multiple coexisting esophageal disorders are frequent, further complicating management.^12,13,18

Depending on the clinical setting, a complete blood count and prothrombin and partial thromboplastin time should be obtained prior to the procedure. This is important in several regards. First, in the neutropenic patient with esophageal symptoms, prophylactic antibiotics may be important to prevent bacteremia.¹⁹ Second, thrombocytopenia should be recognized prior to endoscopy, because this may limit the ability to take mucosal biopsies. Likewise, an underlying coagulopathy may prohibit biopsies. Pretreatment with platelet transfusions or fresh frozen plasma may therefore be required before proceeding with endoscopy.

All diagnostic equipment that may be required should be readily available, including a cytology brush, formalin, and media for electron microscopy, especially in patients with AIDS. Culture media for virus as well as sterile saline and containers for fungal and mycobacterial cultures may be required. Universal precautions, including gloves, gowns, and protective eyewear, should be used with all patients. Cleaning of the endoscope after these cases should follow standard guidelines. Gluteraldehyde and other solutions used in scope washers are sufficient to destroy bacteria and mycobacteria; HIV is easily killed by these routine cleaning methods.²⁰ For any identified abnormalities, a differential diagnosis should be constructed based on the appearance of the lesion(s). This differential diagnosis can be formulated based on a pattern approach to the lesions (Table 14-1). The differential diagnosis will dictate how the lesion should be biopsied and the number of samples that should be obtained, as well as the requirement for other diagnostic testing (eg, culture) on the biopsy specimens. The location, size, and appearance of all endoscopic lesions should be well documented, which will be useful for comparison on follow-up endoscopic examinations.

Fungal Esophagitis

Candida Esophagitis

EPIDEMIOLOGY

Candida esophagitis is by far the most common esophageal infection. Although a well-known complication of immunosuppression, under selected conditions, Candida may cause disease in the immunocompetent host. Nevertheless, the recognition of any esophageal infection in a presumed immunocompetent patient should alert the physician to the possibility of an underlying immunodeficiency. Conditions predisposing to Candida esophagitis in the normal host include antibiotic use,²¹ inhaled or ingested corticosteroids, ^22,23 antisecretory therapy or hypochlorhydric states,^24-26 and diabetes mellitus.²⁷ Alcoholism, malnutrition, and advanced age have also been linked to Candidal infection.²⁸ Candida esophagitis may also result from radiation therapy to the neck and chest and following photodynamic therapy.²⁹ Esophageal motility disturbances (achalasia, scleroderma) and structural abnormalities (benign and malignant strictures, diverticula) may be complicated by Candida esophagitis due to stasis of esophageal contents.^30-32 In this setting, Candida generally represents an incidental finding. Rarely, no specific underlying disease can be found.³³

Diagnosis

ENDOSCOPIC FEATURES

The endoscopic appearance of Candida esophagitis is well recognized. Kodsi et al³⁴ reported the endoscopic findings in 27 immunosuppressed nonHIV-infected patients with Candida esophagitis. The diagnosis was established primarily by brush cytology, with very few patients undergoing endoscopic mucosal biopsy for histopathologic assessment, and thus viral esophagitis may not have been adequately excluded. Nevertheless, despite these shortcomings, this study forms the basis for most of the grading scales currently used to assess severity of Candida esophagitis. More recently, a grading scale has been proposed based on the severity of plaque formation.³⁵ Grade I was said to be present when rare plaques were seen throughout the esophagus to Grade IV, which was said to be present with severe plaque formation resulting in luminal narrowing. In this endoscopic-pathologic study, esophageal ulceration was not seen in these AIDS patients unless coexisting diseases were present.

Candida esophagitis is characterized endoscopically by the presence of multiple isolated or confluent plaques (Figure 14- 1). Histopathologically, this plaque material is composed primarily of desquamated epithelial cells mixed with fungi, bacteria, and inflammatory cells (Figure 14-2).^35,36 In the early stages of infection, plaques have minimal elevation that may mimic an erosion, and the disease may be focal.³⁷ With disease progression, these plaques coalesce, resulting in circumferential mucosal involvement.³⁸ Barium esophagrams in such a patient may be described as having a “shaggy” appearance to the esophageal mucosa due to barium intercalating with this plaque material.³⁹ Given the severity of immunodeficiency in patients with AIDS, esophagitis may be striking, with the plaque material resulting in luminal narrowing (Figure 14-3A) or even an obstructing lesion on barium studies⁴⁰ as well as endoscopy.⁴¹ Any portion of the esophagus may be involved. In most patients, there appears to be a predilection for the proximal esophagus with distal involvement as the disease progresses. However, isolated involvement of the mid or distal esophagus may occur. Candida esophagitis frequently coexists with other esophageal disorders.^11,42,43 Prospective studies have documented multiple etiologies in up to 50% of HIV-infected patients.^13,15 Viruses were the most frequent copathogens. Candida rarely causes endoscopic or histopathologic ulceration in patients with AIDS;¹⁸ thus, the presence of well-circumscribed ulceration(s) associated with Candida esophagitis of any severity should be a clue to the presence of coexisting viral disease or some other process (Figure 14-4). Histopathologic ulceration has been associated with esophageal candidiasis in other immunocompromised hosts,⁴⁴ although the prevalence with which this occurs cannot be accurately determined. Based on these observations, in any patient with Candida esophagitis and coexistent ulcer, a thorough histopathologic examination of ulcer tissue is mandatory.

DIAGNOSTIC METHODS

Although Candida esophagitis has a characteristic endoscopic appearance, in most situations, objective documentation of infection is prudent. Brush cytology of plaque material has a high sensitivity, approaching 100%.^45,46 For the patient with mild disease (scattered plaques with normalappearing intervening mucosa) in whom histopathologic examination of underlying squamous tissue may not be necessary, brush cytology represents the most cost-effective diagnostic method. With more severe disease, biopsy has a high sensitivity but plays a more important role in excluding coexisting processes. Characteristic histopathologic features of Candida species, including pseudohyphae and true hyphae, can be appreciated by standard hematoxylin and eosin (H & E) staining.

In patients with AIDS and severe (Grades III and IV) disease, endoscopic evaluation of the underlying mucosa is very important as coexisting disorders have been identified in 25% of such patients.⁴⁷ When the plaque material obscures visualization of the underlying mucosa (see Figure 14-3A), we remove this material as thoroughly as possible to detect any underlying mucosal ulceration and perform biopsies of these lesions. To remove this plaque material, we perform the technique of tipping the endoscope into the esophageal wall, applying suction, and advancing the endoscope the desired length. The endoscope is then withdrawn with suction to the original position, and the esophageal wall is visually examined. ⁴⁷ Using this technique, two to three passes usually remove enough debris to adequately expose the underlying mucosa. This technique may result in mucosal friability but will not cause mucosal defects (ulcers) (Figure 14-3B).⁴⁷ Multiple mucosal biopsies should be obtained of any endoscopic abnormalities (erosions, ulcer) to exclude coexistent viral disease. Fungal culture of plaque material is neither widely available nor necessary to establish the diagnosis.

Histoplasmosis

Esophageal involvement caused by histoplasmosis has been reported to occur from mediastinal lymphadenopathy in immunocompetent hosts as well as in patients with underlying neoplasms.^48,49 A single case of histoplasmosis involving the esophagus has been reported in a patient with AIDS.⁵⁰ In this patient, who presented with upper gastrointestinal bleeding, several volcano-appearing lesions were observed in the distal esophagus. Mucosal biopsy with fungal staining was diagnostic. As with most infectious diseases of the esophagus, adequate sampling by mucosal biopsy with appropriate histopathologic staining yields the diagnosis based on the morphology of the infecting pathogen.

Miscellaneous Fungi

Other nonCandidal fungi have been reported to involve the oropharynx and esophagus, primarily in patients with AIDS.^51-53 These include the mold Penicillium chyrsogenum⁵³ and Exophiala jeanslemei.⁵¹ These pathogens were identified by fungal cultures. In general, the identification of a nonCandidal species is of no clinical significance, given the similar response to standard antifungal therapy. Mucormycosis esophagitis in AIDS has also been described;⁵⁴ the endoscopic and histopathologic features in this case were similar to Candida esophagitis. The diagnosis was established by fungal staining of a biopsy specimen, in which large nonseptate hyphae typical for mucormycosis were observed, and by culture of biopsy specimens. Cryptococcal esophagitis has been reported in three patients at autopsy.⁵⁵.

Complications

Candida esophagitis coexisting with viral esophagitis has been reported to cause tracheoesophageal fistula.⁵⁶ This probably represented fungal colonization of preexisting ulcer rather than a primary etiology of fistula. Upper gastrointestinal bleeding has also been attributed to severe Candidal esophagitis.⁵⁷ In some of these patients, an underlying coagulopathy probably initiated and/or potentiated the bleeding episode. Because of the superficial nature of this infection, bleeding is more likely to be diffuse. However, despite the usual precaution of mucosal biopsy in the setting of acute bleeding, brushings and mucosal biopsy may be required to exclude coexisting disorders, as a definitive diagnosis in this setting has important treatment implications. Transmural necrosis with perforation requiring surgery was reported as a complication of Candida esophagitis in two patients with neutropenia.⁵⁸

Viral Esophagitis

Epidemiology

Viral pathogens represent the next most important group of infectious disorders causing esophagitis, with the herpes family of viruses being the most prevalent. In patients without underlying immunosuppression, herpes simplex virus (HSV) type 1 esophagitis is the most common viral cause of esophageal disease. In the absence of immunodeficiency, cytomegalovirus (CMV) esophagitis is rare.^59,60 Although Candida esophagitis is more frequent, HSV and CMV esophagitis may complicate corticosteroid therapy, especially when used chronically.⁶¹ The use of corticosteroids in the immunosuppressed patient may also precipitate viral esophagitis.⁶² Varicella zoster esophagitis has been described in nonHIV-infected transplant patients.⁴⁴

Distinguishing primary viral infection from reactivation in the “normal” host can be accomplished by serologic testing. However, serologic testing plays no important role in immunosuppressed patients, because disease usually results from reactivation. With acute infection, IgM antibodies should be present in high titer while IgG antibodies represent prior infection or reactivation. In most immunosuppressed patients, seropositivity to both HSV type 1 and CMV is common, representing prior infection. In HIV-infected patients, particularly homosexual men, seropositivity to these viruses is almost uniform.⁶³ Following Transplantation, viral esophagitis usually occurs from 1 to 4 months postoperatively. HSV and CMV esophagitis occur with equal frequency, including bone marrow and liver Transplantation recipients.⁴⁴ The incidence of viral esophagitis may be higher in transplant recipients (especially following bone marrow Transplantation) than in other immunocompromised nonHIV-infected hosts because of the greater level of immunosuppression. No association has been found between corticosteroid use or blood levels of cyclosporine and the presence of esophagitis in patients undergoing liver or renal Transplantation.^64,65 As previously mentioned, the routine use of antifungal and, in some cases, antiviral prophylaxis for the transplant patient in the perioperative period has significantly reduced the occurrence of these opportunistic esophageal infections. Viral esophagitis is less common than Candida esophagitis in immunocompromised nonTransplantation patients, such as in those with an underlying neoplasm.⁶⁶

Viral esophagitis is an important cause of morbidity in HIV disease. In these patients, both HSV and CMV disease typically occur in the latter stages of immunodeficiency (absolute CD4 lymphocyte count <100/µL).^2,67 Although some studies of symptomatic patients undergoing endoscopy prior to any therapy have documented an equivalent prevalence of CMV and HSV esophagitis,^12,13 a prospective study of 100 HIV-infected patients with well-defined esophageal ulcer showed CMV esophagitis alone to be significantly more common than HSV esophagitis alone, occurring in 45% and 5% of patients, respectively.¹⁸ This low prevalence of HSV esophagitis was not related to the use of acyclovir therapy. This study also demonstrated HSV and CMV co-infection in four patients. Other reported viral causes of esophageal disease in patients with AIDS include Epstein-Barr virus⁶⁸ and human papilloma virus.⁶⁹ HIV has also been proposed as a cause of esophageal disease and is discussed separately.

Diagnosis

ENDOSCOPIC FEATURES

In contrast to Candida, viral esophagitis characteristically produces mucosal ulceration. The location, size, and appearance of these ulcerative lesions varies considerably based on the cause of immunodeficiency, etiology of ulcer, and severity of disease. These differences likely relate to the underlying pathogenesis of disease. HSV infects squamous epithelial cells^36,70; thus, superficial mucosal injury is common. The earliest lesion of HSV esophagitis is the vesicle. This can be seen in the oropharynx in some patients, although vesicular lesions in the esophagus will infrequently be encountered. With disease progression, HSV typically results in multiple well-circumscribed superficial lesions that may coalesce^71-75 (Figure 14-5). HSV ulceration may be scattered throughout the esophagus or, when severe, may result in a diffuse erosive esophagitis. Exudate overlying small shallow ulcers may mimic Candida esophagitis (Figure 14-6). Large, deep, wellcircumscribed ulcers are almost never observed with HSV, regardless of the cause of immunodeficiency.

In contrast to HSV esophagitis, CMV appears to cause mucosal disease through both infection of endothelial cells (vasculitis)⁷⁶ and cytokine mechanisms. CMV esophagitis is characterized by large, well-circumscribed ulcerations with sparing of the intervening mucosa^77,78 (Figure 14-7). In the largest study to date characterizing 141 CMV ulcers in 33 patients with AIDS,⁷⁸ disease was more common in the mid (57%) and distal (32%) esophagus. Ulcers were multiple in 58%, with three patients having more than 10 ulcers. Ulcers varied in size from less than 1 cm in 43% to greater than 3 cm in 18%. The depth was also variable, with shallow lesions seen in 46% and deep lesions in 8%. A diffuse superficial pattern of ulceration resembling HSV was observed in 6%. Esophagoesophageal fistulas have been noted (Figure 14-8). These fistulas have rarely been found in other settings. Ulcerated mass lesions presumably from an exuberant inflammatory response resembling a neoplasm have been described in the esophagus⁷⁹ and even colon⁸⁰ in AIDS patients. In nonHIV-infected immunocompromised hosts, ulcerations are usually smaller in size (Figure 14-9), and the large deep ulcers typical for AIDS have not been observed.⁴⁴

Differential Diagnosis

In the immunosuppressed nonHIV-infected patient with esophageal ulceration, distinguishing CMV from HSV esophagitis based on endoscopic appearance alone may be difficult. In general, multiple small ulcers are most likely due to HSV whereas CMV results in larger ulcers or a diffuse esophagitis. In HIV-infected patients, large (>2 cm) well-circumscribed ulcer(s) are almost always caused by either CMV or are an idiopathic esophageal ulcer (IEU) (discussion follows). A diffuse superficial ulceration may be caused by either HSV, CMV (Figure 14-10), or a combination of both. Segmental mucosal injury or large well-circumscribed ulcers in the midesophagus may suggest a pill-induced esophagitis. Distal ulcerative disease caused by gastroesophageal reflux disease (GERD) is usually suggested by the clinical history and is easily distinguished endoscopically. However, a severe diffuse erosive pattern in the distal esophagus in the immunosuppressed host may be the result of GERD or viral disease. Occasionally, however, a well-circumscribed ulcer or multiple ulcerations in the distal esophagus may be caused by reflux disease; the presence of a hiatal hernia will help suggest GERD. In this setting, mucosal biopsy to exclude viral disease and clinical follow-up after aggressive antisecretory therapy will be required to establish a definitive diagnosis.

Diagnostic Methods

ESOPHAGEAL BRUSHING

Brushings of esophageal lesions for cytologic analysis play a useful diagnostic role when HSV esophagitis is suspected, although the true sensitivity is unknown. HSV cytopathic effect is identified most reliably in squamous epithelial cells⁷⁰ (Figure 14-11). Thus, brushings of the ulcer edge, the site of active viral replication, rather than the ulcer base will have the highest yield. In the HIV-infected patient with large, well-circumscribed ulcerations with normal-appearing intervening mucosa, esophageal brushings are rarely helpful because HSV is unlikely to be etiologic. Candidal infection is often found by cytologic analysis of esophageal ulcerations, particularly in HIV-infected patients; in this setting, however, Candida represents superficial colonization of ulcer tissue. Brushings of ulcerative esophageal lesions with culture of the brush tip has no advantage over culture of biopsy material alone. Cytologic identification of CMV is poor because the infected cells are located within the granulation tissue and may be few in number.⁸¹

ENDOSCOPIC MUCOSAL BIOPSY

Mucosal biopsy of ulcerative lesions is the most reliable way to diagnose viral esophagitis. In patients with AIDS, mucosal biopsy is also necessary to distinguish IEU from viral disease. There has been much discussion regarding the optimal location to biopsy an ulcer. As discussed previously, in the setting of a diffuse superficial esophagitis or multiple well-circumscribed shallow ulcers, biopsy of the ulcer edge is important to exclude HSV esophagitis. With large, well-circumscribed ulcerations, multiple biopsies of the ulcer base are essential to diagnose CMV esophagitis. CMV cytopathic effect is present in endothelial cells, ganglion cells, and other mesenchymal cells in the granulation tissue (Figure 14-12). CMV does not infect squamous epithelial cells;^82,83 thus, biopsy of the ulcer edge is usually nondiagnostic, unless a significant portion of granulation tissue accompanies the biopsy specimens. When performing mucosal biopsies of the esophagus, we rely on an adaptation of a technique previously reported to increase biopsy size in the stomach.⁸⁴ Briefly, this technique involves passing the biopsy forceps into the esophageal lumen, opening the forceps and withdrawing them to the endoscope, tipping the endoscope with the open forceps into the ulcerative lesion, applying suction to collapse the lesion into the open forceps, advancing the forceps slightly until resistance is met, closing the forceps, directing the endoscope back to the midline, and then withdrawing the forceps, avulsing the tissue. We believe this technique increases biopsy size of esophageal lesions, which may be important to increase diagnostic yield, especially for CMV disease. The appropriate number of biopsy samples required to diagnose viral esophagitis is unknown. In addition, no prospective studies have compared the yield of esophageal brushings to biopsy for HSV esophagitis. In a study by Wilcox et al,⁸⁵ it was shown that in HIV-infected patients with esophageal ulcer, up to 10 biopsies were required to diagnose viral esophagitis, particularly CMV disease. In that study,⁸⁵ the sensitivity of three biopsies for the diagnosis of CMV esophagitis was 80%, 90% for six biopsies, and 99% for 10 biopsies. Ten biopsies were also required to detect HSV/CMV co-infections. Based on this study, we obtain 10 biopsies primarily from the ulcer base. Biopsy of the ulcer edge is performed when HSV esophagitis is suspected (diffuse erosive esophagitis; multiple small, shallow ulcers). Clinical follow-up after therapy provides further confirmatory evidence of etiology. Advancements in diagnostic histopathologic techniques have aided significantly in the diagnosis of viral disease. The use of in situ hybridization and immunohistochemistry have made the identification of viral cytopathic effect much easier. ^70,82,86 This has been especially useful when the inclusions are atypical in appearance or few in number.^86,87 These techniques often highlight more infected cells than are detectable by routine staining methods. Polymerase chain reaction (PCR) has also been used as a diagnostic aid.^88,89 Although highly specific, this technique may be too sensitive for routine use.⁸⁹ The role of viral culture of mucosal biopsy specimens is controversial. At some centers, viral culture of biopsy specimens is routinely performed, particularly for Transplantation patients. Culture positivity of HSV from mucosal biopsies in immunocompromised patients represents true disease. In contrast, the possibility of blood-borne contamination with CMV, especially in patients with AIDS, may complicate the predictive value of a positive viral culture.⁸³ In HIV-infected patients with esophageal ulceration, we rely exclusively on histopathologic examination of mucosal biopsies to diagnose viral esophagitis. In addition, at least for CMV disease, the sensitivity of viral culture appears to be less than histopathologic examination of multiple biopsies^70,89 and requires several days to return positive. At most institutions, routine histopathologic processing of mucosal biopsies can be completed within 24 hours. Given the similarities in endoscopic appearance with CMV and prevalence and therapy of HIV-associated IEU, a definitive diagnosis at the time of initial endoscopy is especially important.

Complications

HSV esophagitis is a well-recognized cause of gastrointestinal hemorrhage; bleeding is usually selflimited, because primary infection spontaneously remits in immunocompetent patients.⁹⁰ In patients with AIDS, large esophageal ulcerations caused by CMV may occasionally bleed.⁹¹ As with peptic ulcer bleeding, unless an active bleeding point is identified in the ulcer bed, endoscopic hemostatic therapy is often ineffective. If bleeding is the presenting manifestation of ulcer, mucosal biopsy is critical to make a definitive diagnosis (CMV vs HSV vs IEU), so that specific therapy can be initiated. If a bleeding point in the ulcer can be identified, multiple biopsies from another portion of the lesion or from a nonbleeding ulcer should be performed. If no specific bleeding point or active oozing is seen, ulcer biopsy is probably safe. Depending on the endoscopic appearance, brush cytology may be useful to exclude HSV when a coagulopathy is present. A tracheoesophageal fistula complicating HSV esophagitis has been reported.⁹² Spontaneous esophageal perforation has also been described.⁹³ Food impaction at the site of extensive ulceration has been observed as well.⁹⁴ Despite the large size of these CMV ulcerations in patients with AIDS, perforation is rare. Fistula to the bronchus was reported in a patient with AIDS.⁹⁵ Esophageal strictures may result from viral esophagitis.⁹⁶ With persistent ulceration and stricture, aggressive medical therapy for the underlying cause of ulcer is required. As with GERD, until ulcerative disease is healed, strictures will persist. Dilation may be necessary to relieve dysphagia, thereby enhancing nutritional intake. There is no evidence that standard bougienage of these strictures is more likely to be complicated by perforation or bleeding.

Mycobacterial Esophagitis

Mycobacterium tuberculosis

Despite the worldwide epidemic of tuberculosis associated with HIV infection, esophageal involvement remains rare as an extrapulmonary manifestation.^97-99 Even prior to the AIDS epidemic, esophageal tuberculosis was uncommon and was usually suggested by radiographic studies or documented at autopsy.¹⁰⁰ Except for HIV-infected patients, almost all other reported patients had spontaneous disease apparently unassociated with immunodeficiency. Corticosteroids have been felt to precipitate disease in some patients.¹⁰¹ A strong clinical suspicion of esophageal tuberculosis should be raised in the patient with known pulmonary disease who develops esophageal symptoms. Almost uniformly, esophageal disease results from contiguous spread from involved mediastinal lymph nodes rather than from direct mucosal infection. This belief is based on the following: pulmonary findings generally antedate esophageal symptoms¹⁰²; chest radiography usually reveals pneumonia and/or mediastinal adenopathy at the time of presentation;¹⁰³ and prominent mediastinal adenopathy can be found by computed tomography (CT) scanning.¹⁰⁴ In these patients, routine chest radiography may reveal old apical disease, pleural effusions, mediastinal adenopathy, and active pulmonary disease, or it can be normal. Even in those patients with a normal chest radiograph, sputum staining and culture may be diagnostic. Primary esophageal infection without apparent adenopathy or fistula has, however, been described.¹⁰⁵ Esophageal disease has not been associated with miliary tuberculosis. The endoscopic appearance of esophageal tuberculosis is variable. An ulcerated mass lesion resembling a malignancy has been most frequently described.^106,107 Other endoscopic manifestations include ulceration alone, displacement of the esophagus from mediastinal adenopathy, and stricture.^106,107 Sinus tracts are also common and are best defined by barium esophagography. Mucosal biopsy of identified lesions demonstrates ulcer tissue; necrotizing granulomata may be found. Mycobacterial staining of biopsy specimens and culture is usually diagnostic. Perforation and fistula have been reported as complications.^108-110

Mycobacterium avium Complex

Despite the prevalence and systemic nature of Mycobacterium avium complex (MAC) disease in AIDS,^111,112 the esophagus remains a rare site of infection. In the reported cases, the disease was endoscopically manifested by an esophagoesophageal fistula¹¹³ or ulcerations.^114-116 In some cases, it is possible that MAC may not have been the primary pathogen, given the absence of mycobacteria on histopathologic examination, with identification only by positive cultures of biopsy specimens.¹¹³ It is also possible that esophageal disease occurred secondarily through mediastinal lymph node involvement. As with M. tuberculosis, histopathologic findings include ulceration, although granulomas are absent. Mycobacterial staining of biopsy specimens was diagnostic in some of the cases

Bacterial Esophagitis

Bacterial infection of the esophagus is rare. The largest series to date reported the clinical and pathologic features of 23 patients in whom 20 were studied at postmortem.¹¹⁷ In these patients with hematologic malignancies, neutropenia (white blood cell count of 100/1 L) was uniform. The causative bacteria identified histopathologically included gram-positive cocci in 14 patients, gram-negative bacilli in three, and mixed infections or gram-positive bacilli in the remaining patients. Associated bacteremia was documented in four patients. Endoscopic examination in the three index cases revealed distal linear shallow esophageal ulcers with associated plaques suggestive of Candida esophagitis. Histopathologic examination of biopsy specimens demonstrated this plaque material to consist of necrotic squamous epithelium with bacteria invading normal-appearing squamous tissue. Bacterial esophagitis has been found at autopsy in immunosuppressed patients with cancer¹¹⁸ and following bone marrow Transplantation.⁴⁴ Bacterial esophagitis has also been described in one HIV-infected patient with concomitant diabetic ketoacidosis.¹¹⁹Other unusual causes of bacterial esophagitis reported in HIV-infected patients include actinomycosis¹²⁰ and nocardia. ¹²¹ In the one reported patient with esophageal actinomycosis, ¹²⁰ barium esophagography revealed marked ulceration, fistulas, and diverticula involving the distal esophagus. Endoscopically, multiple white plaques were observed diffusely throughout the esophagus without frank ulceration. Histologic examination revealed sulfur granules and filamentous gram-positive bacteria consistent with Actinomyces. Actinomyces has also been reported in combination with CMV in a patient with a solitary large esophageal ulcer.¹²² In the reported patient with nocardial esophagitis,¹²¹ endoscopy revealed a single large elliptical ulceration. Mucosal biopsy was diagnostic, demonstrating gram-positive branching filamentous organisms typical of nocardia.

Protozoal Esophagitis

Protozoal infection of the esophagus in immunocompromised, nonHIV-infected patients has not been described. To date, esophageal disease has been reported in HIV-infected patients from cryptosporidia,¹²³ Pneumocystis carinii,¹²⁴ and Leishmania donovani.^125-128 One patient with AIDS and a large solitary ulcer was found to have both CMV and Leishmania.¹²⁵ The patient with P. carinii also had HSV and CMV viral cytopathic effect and yeast forms consistent with Candida in the biopsy specimens.¹²⁴ The endoscopic appearances in these cases consisted of a diffuse exudative esophagitis with superficial ulceration (Pneumocystis) and an appearance compatible with Candida esophagitis plus diffuse linear ulceration, both shallow and deep (Leishmania). In these cases, the etiologic pathogen was identified by appropriate staining of mucosal biopsies. Trichomonas has been reported as a cause as well in a patient with AIDS.¹²⁹

Specific HIV-Related Esophageal Disorders

Disorders Associated with Seroconversion

Primary HIV infection is largely asymptomatic. In some patients, however, a mononucleosis-like illness associated with a maculopapular rash occurs at the time of seroconversion.¹³⁰ Oropharyngeal and esophageal ulceration has been observed during this seroconversion illness. ¹³⁰ These esophageal ulcerations appeared endoscopically as multiple, small, and shallow. ¹³¹ Mucosal biopsy demonstrated ulcer tissue without any etiologic agents. In some of these patients, electron microscopic examination of biopsy specimens revealed enveloped virus-like particles with morphologic features compatible with retroviruses.¹³¹ These lesions resolve spontaneously. This seroconversion syndrome should be suspected in any patient at risk for HIV infection who develops a flu-like illness associated with a rash and esophageal symptoms.¹³² The diagnosis can be established by testing for HIV RNA. Serologic testing usually becomes positive within 3 to 18 months after the illness.¹³⁰ Candida esophagitis has also been described as a presenting manifestation of the seroconversion syndrome.¹³³ As with the esophageal ulcers seen in this syndrome, the esophagitis usually resolves spontaneously. Candida esophagitis is perhaps the result of a transient but significant immunodeficiency that occurs at the time of seroconversion.¹³⁴

Idiopathic Esophageal Ulcer

Early on in the AIDS epidemic, large esophageal ulcerations were recognized where no specific etiology could be determined despite extensive histopathologic examination of ulcer tissue.^43,135 These were termed idiopathic or apthous. Initially, these ulcers were assumed to be of viral etiology, such as from CMV, that were “missed.” We now recognize these as a specific entity. IEU is seen in the later stages of immunodeficiency when the CD4 lymphocyte count is less than 100/µL.¹⁸ It has been suggested that HIV may be the direct cause of these lesions. However, using a variety of histopathologic techniques, HIV has been observed in ulcer tissue from a variety of esophageal lesions and not specifically with IEU136-138 and the infected cells (inflammatory cells and not squamous cells). Taken together, these studies suggest that HIV does not cause IEU, at least based on a direct cytopathic mechanism.

ENDOSCOPIC FEATURES

Early on, isolated giant ulcers were commonly reported,¹³⁹ but it is now appreciated that these ulcers are variable in size and appearance. In a prospective study¹⁴⁰ characterizing 68 IEU in 23 patients, 37% of the ulcers were described as less than 1 cm in greatest dimension, with 34% greater than 2 cm (Figures 14-13 and 14-14). Multiple ulcers were common, seen in 57%. This report also documented the variable depth from shallow lesions (31%) to giant deep ulcerations (7%). These ulcers were described as having a heaped-up appearance in 40% (see Figure 14-14). Uniformly, these lesions appeared as well-circumscribed ulcerations with normalappearing intervening mucosa. The diffuse superficial ulceration typical for HSV or other viral diseases has not been reported to result from IEU. In some patients, these lesions may cause esophagoesophageal fistulas, resulting in a mucosal bridge (see Figure 14-8).

DIFFERENTIAL DIAGNOSIS

The most important lesion to differentiate from IEU is CMV esophagitis. An isolated ulceration in the proximal esophagus may infrequently be a drug-induced esophagitis, because these lesions are typically shallow and multiple. The presence of a large ulcer in the distal esophagus may suggest GERD; in general, a distinction can be drawn based on history (heartburn, regurgitation), associated endoscopic findings (erythema, linear erosions arising from the gastroesophageal junction), and/or the presence of a hiatal hernia. The histopathologic features alone cannot distinguish IEU from GERD.

DIAGNOSTIC METHODS

IEU is a diagnosis of exclusion. Because the endoscopic appearance of IEU is not pathognomonic, mucosal biopsies are necessary to exclude other etiologies. A recent study suggests that at least 10 biopsies may be necessary to reliably distinguish viral esophagitis from IEU.¹³⁸ Biopsies should be performed in the fashion described previously to sample the ulcer base. The presence of a superficial Candidal infection overlying a large well-circumscribed lesion with histopathologic findings of granulation tissue without viral cytopathic effect should still lead to the diagnosis of IEU. Histopathologic processing of mucosal biopsies should be similar to that for the patient with suspected CMV esophagitis. Immunohistochemical staining to confirm viral cytopathic effect may be necessary, especially when viral cytopathic effect is atypical in appearance or few in number.^36,86 Additional stains for fungi and mycobacteria may be necessary, depending on the clinical setting and histopathologic findings.

COMPLICATIONS

As with any ulcerative esophageal disorder, bleeding may occur. This may be seen in patients with a coexistent coagulopathy or after nonsteroidal antiinflammatory drug use. IEUs have been reported to fistulize to the stomach.^141,142 As described for CMV, esophageal fistulas and esophagoesophageal fistulas with a mucosal bridge have been observed, although rarely causing a clinical complication. Fistula formation to the lungs has not been described. Despite the size of these ulcers, esophageal strictures are uncommon.⁹⁶

Neoplasms

Although the incidence of nonHodgkin's lymphoma and Kaposi's sarcoma in HIV-infected patients is high, involvement of the esophagus remains uncommon.¹⁴³ As with disease in other organ systems, these neoplasms involve the esophagus during the latter stages of immunodeficiency. Cutaneous Kaposi's sarcoma may be observed with a CD4 lymphocyte count of greater than 200 per cubic millimeter, whereas nonHodgkin's lymphoma usually occurs with more advanced immunodeficiency.

Kaposi's Sarcoma

Gastrointestinal Kaposi's sarcoma is common in those with cutaneous disease, and esophageal lesions are generally an incidental finding at the time of endoscopy performed for other reasons.^144,145 However, isolated gastrointestinal disease without cutaneous disease has been noted.¹⁴⁶ Gastric or duodenal involvement appears to be more common than esophageal disease.

ENDOSCOPIC FEATURES

The endoscopic features of esophageal Kaposi’s sarcoma are similar to their cutaneous appearance as violatious macular or plaque-like lesions. They have not been reported as circumferential mass lesions like squamous cell tumors or adenocarcinoma. With tumor progression, the lesions became nodular, or even a solitary mass may be found. Ulceration may occur when the lesions become large.

MUCOSAL BIOPSY

These tumors typically involve the submucosa; therefore, mucosal biopsy must sample deeper tissue. To increase diagnostic yield, biopsies may be obtained with “jumbo” biopsy forceps, or techniques such as bite-on-bite or tunnel biopsies may be used. Despite the composition of the tumor by multiple vascular channels, there is no evidence that mucosal biopsy precipitates bleeding. The presence of cutaneous disease in combination with the typical endoscopic appearance may obviate the need for mucosal biopsy. However, when the lesion appears atypical or when cutaneous disease is absent, mucosal biopsy to confirm the diagnosis may have important treatment and prognostic implications. CMV viral cytopathic effect may occasionally be identified in the tumor and probably warrants no specific treatment unless CMV disease is found in other sites (retina, colon). Bleeding may be observed when the tumor becomes large and ulcerated.^91,147 In this setting, temporizing measures employing injection therapy may halt bleeding. Angiographic embolization of esophageal lesions is difficult.

NonHodgkin's Lymphoma

NonHodgkin's lymphoma is a well-recognized complication of HIV-associated immunodeficiency, occurring primarily in homosexual men.¹⁴⁸ Gastrointestinal involvement is particularly common, although esophageal disease remains rare.^149,150 In the largest endoscopic series reported to date of four patients with primary esophageal nonHodgkin's lymphoma, ¹⁴⁹ the endoscopic appearance of these lesions was described as an ulcerated polypoid mass, often with a central ulceration. The size of these lesions varied from 1 cm to extensive disease, resulting in luminal narrowing similar to adenocarcinoma. To definitively establish the diagnosis, multiple biopsies are required. Histopathologic confirmation of lymphoma may be difficult, and, as with Kaposi's sarcoma, CMV viral cytopathic effect may be incidentally identified in the tumor. Delivering fresh tissue on a saline-dampened cloth for frozen section is important not only in distinguishing the cytologic features of the tumor, but also to perform flow cytometry. These studies assist in determining the clonality and specific subtype of the tumor, which may influence the choice of chemotherapeutic regimen.

Miscellaneous Neoplasms

Adenocarcinoma of the esophagus associated with Barrett's epithelium has been reported in HIV-infected patients where the tumor appeared as a fleshy ulcerated mass lesion at the gastroesophageal junction.¹⁵¹ Typical Barrett's epithelium was identified and confirmed at the time of resection. Squamous cell esophageal carcinoma is uncommon in HIV-infected patients, probably due to their relatively young age. To date, squamous cell esophageal carcinoma has been reported in two AIDS patients.¹⁵² Both patients were men, older than 50 years of age, and with established risk factors for this tumor, including alcohol and tobacco use. The tumor appeared as a small plaque-like lesion in one patient. In the other patient, diffuse subtle nodularity was seen radiographically, although endoscopy only revealed erythema throughout the esophagus. The approach to biopsy of an esophageal mass in this setting should be similar to that for any patient with a suspected esophageal neoplasm. Multiple biopsies should be performed and appropriate pathologic support utilized, particularly to identify lymphoma.

Miscellaneous

Pill-Induced Esophagitis

Given the plethora of medications that HIV-infected patients receive, it is surprising that pill-induced esophagitis is not more common. Specific for HIV-infected patients, pillinduced esophagitis caused by zidovudine (AZT) and zalcitabine (ddC) has been noted.^153,154 In these patients, a thorough history was suggestive of the diagnosis. Pill-induced esophageal ulceration typically causes multiple, shallow ulcers in the proximal esophagus near the aortic arch. However, large ulcers have been reported with ddC.¹⁵⁴ Resolution of ulcerations occurs with discontinuation of the offending medication alone. When encountering such a patient, multiple biopsies with appropriate histopathologic evaluation to exclude other potential diseases such as HSV or CMV should be performed. Strictures could also suggest a pill-induced etiology.¹⁵⁵ As HIV-infected patients are living longer, consideration for other causes of drug-induced esophageal ulceration, such as alendronate,¹⁵⁶ must be kept in mind. History, endoscopic findings, and follow-up after drug discontinuation may be required to distinguish IEU from a pill-induced esophagitis.

Gastroesophageal Reflux Disease

Hypochlorhydria has been variably observed in patients with AIDS^157,158 and was speculated to be one reason for the relatively low prevalence of GERD. In a series of 100 HIVinfected patients with esophageal ulcer, reflux disease was the etiology in four.¹⁸ More recently, however, with the use of HAART, GERD is becoming more common.¹⁷ GERD in HIV-infected patients presents similarly both clinically and endoscopically to other patients. Distinguishing gastroesophageal reflux from other HIVassociated diseases may occasionally be difficult, especially in the setting of an isolated ulcer at the gastroesophageal junction or severe distal erosive disease. The absence of typical reflux symptoms and a clinical and endoscopic response to PPIs may assist in distinguishing an HIV-associated esophageal disease from reflux disease if biopsy samples are inconclusive. In some patients, ambulatory 24-hour esophageal pH monitoring may be required to establish a definitive diagnosis.

T R I C K S  O F  T H E  T R A D E

• In the setting of severe Candida esophagitis where the mucosal surface is obscured, removal of the Candidal plaque is important to expose any underlying well-circumscribed ulcerations. In this setting, biopsies should be obtained of the ulcer to exclude viral disease.
• Multiple circumscribed small ulcerations are more likely due to herpes, whereas large well-circumscribed ulcers with normal-appearing intervening mucosa suggest CMV esophagitis.
• Biopsy of the ulcer edge is important to exclude herpes viral esophagitis. CMV viral cytopathic effect resides in the granulation tissue in the ulcer base.

REFERENCES

1. 1. Monkemuller KE, Call SA, Lazenby AJ, Wilcox CM. Declining prevalence of opportunistic gastrointestinal disease in the era of combination antiretroviral therapy. Am J Gastroenterol. 2000;95:457-468.
2. Bini EJ, Micale PL, Weinshel EH. Natural history of HIV-associated esophageal disease in the era of protease inhibitor therapy. Dig Dis Sci. 2000;45:1301-1307.
3. Goodrich JM, Bowden RA, Fisher L, et al. Ganciclovir prophylaxis to prevent cytomegalovirus disease after allogeneic marrow transplant. Ann Intern Med. 1993;118:173-178.
4. Grossi P, Farina C, Fiocchi R, Dalla Gasperina D. Prevalence and outcome of invasive fungal infections in 1,963 thoracic organ transplant recipients: a multicenter retrospective study. Italian Study Group of Fungal Infections in Thoracic Organ Transplant Recipients. Transplantation. 2000;70:112-116.
5. Singh N. Antifungal prophylaxis for solid organ transplant recipients: seeking clarity amidst controversy. Clin Infect Dis. 2000;31:545-553.
6. Slavin MA, Osborne B, Adams R, et al. Efficacy and safety of fluconazole prophylaxis for fungal infections after marrow Transplantation—a prospective, randomized, double blind trial. J Infect Dis. 1995;171:1545- 1552.
7. Wade JJ, Rolando N, Hayllar K, et al. Bacterial and fungal infections after liver Transplantation: an analysis of 284 patients. Hepatology. 1995;21:1328-1336.
8. Laine L, Dretler RH, Conteas CN, et al. Fluconazole compared with ketoconazole for the treatment of Candida esophagitis in AIDS: a randomized trial. Ann Intern Med. 1992;117:655-660.
9. Smith D, Midgley J, Gazzard B. A randomised, double-blind study of itraconazole versus placebo in the treatment and prevention of oral or esophageal candidiasis in patients with HIV infection. Int J Clin Pract. 1999;53:349-352.
10. Mocroft M, Youle M, Phillips AN, Halai R, Easterbrook P, Johnson MA, Gazzard B. The incidence of AIDS-defining illnesses in 4883 patients with human immunodeficiency virus infection. Arch Intern Med. 1998;158:491-497.
11. Baehr PH, McDonald GB. Esophageal infections: risk factors, presentation, diagnosis, and treatment. Gastroenterology. 1994;106:509-532.
12. Bonacini M, Young T, Laine L. The causes of esophageal symptoms in human immunodeficiency virus infection: a prospective study of 110 patients. Arch Intern Med. 1991;151:1567-1572.
13. Connolly GM, Hawkins D, Harcourt Webster JN, et al. Oesophageal symptoms, their causes, treatment, and prognosis in patients with the acquired immunodeficiency syndrome. Gut. 1989;30:1033-1039.
14. Lopez Dupla M, Sanz PM, Garcia VP, et al. Clinical, endoscopic, immunologic, and therapeutic aspects of oropharyngeal and esophageal candidiasis in HIV-infected patients: a survey of 114 cases. Am J Gastroenterol. 1992;87:1771-1776.
15. Porro GB, Parente F, Cernuschi M. The diagnosis of esophageal candidiasis in patients with acquired immune deficiency syndrome: is endoscopy always necessary? Am J Gastroenterol. 1989;84:143-146.
16. Rolston KVI, Rodriguez S. Upper gastrointestinal disease in human immunodeficiency virus-infected individuals [Letter]. Arch Intern Med. 1992;152:881-882.
17. Wilcox CM, Straub RF, Clark WS. Prospective evaluation of oropharyngeal findings in human immunodeficiency virus-infected patients with esophageal ulceration. Am J Gastroenterol. 1995;90:1938-1941.
18. Wilcox CM, Schwartz DA, Clark WS. Esophageal ulceration in human immunodeficiency virus infection: etiology, response to therapy and long term outcome. Ann Intern Med. 1995;123:143-149.
19. Bianco JA, Pepe MS, Higano C, et al. Prevalence of clinically relevant bacteremia after upper gastrointestinal endoscopy in bone marrow transplant recipients. Am J Med. 1990;89;134-137.
20. Hanson PJV, Dor G, Jeffries DJ, et al. Elimination of high titre HIV from fiberoptic endoscopes. Gut. 1990;31:657-659.
21. Hachiya KA, Kobayashi RH, Antonson DL. Candida esophagitis following antibiotic usage. Pediatr Infect Dis. 1982;1:168-170.
22. Chocarro MA, Galindo TF, Ruiz-Irastorza G, Gonzalez LA, Alvarez NF, Ochoa SC, Martin AMI. Risk factors for esophageal candidiasis. Eur J Clin Micro Infect Dis. 2000;19:96-100.
23. Simon MR, Houser WL, Smith KA, Long PM. Esophageal candidiasis as a complication of inhaled corticosteroids. Ann Allergy Asth Immun. 1997;79:333-338.
24. Brooks JR, Smith HF, Pease FB. Bacteriology of the stomach immediately following vagotomy: the growth of Candida albicans. Ann Surg. 1974;6:859-862.
25. Hendel L, Svejgaard E, Walsoe I, et al. Esophageal candidosis in progressive systemic sclerosis: occurrence, significance, and treatment with fluconazole. Scand J Gastroenterol. 1988;23:1182-1186.
26. Larner AJ, Lendrum R. Oesophageal candidiasis after omeprazole therapy. Gut. 1992;33:860-862.
27. Parkman HP, Schwartz SS. Esophagitis and gastroduodenal disorders associated with diabetic gastroparesis. Arch Intern Med. 1987;147:1477-1480.
28. Trier JS, Bjorkman DJ. Esophageal, gastric, and intestinal candidiasis. Am J Med. 1984;77:39-43.
29. Luketich JD, Nguyen NT, Weigel TL, Keenan RJ, Ferson PF, Belani CP. Photodynamic therapy for treatment of malignant dysphagia. Surg Lapar Endosc Percutan Tech. 1999;9:171-175.
30. Das K, Kochhar R, Goenka MK, et al. Obstruction, not cancer, is responsible for esophageal Candidal overgrowth. J Clin Gastroenterol. 1995;20:330-331.
31. Poland JM. The role of Candida infections as an adverse effect upon head and neck cancer patients undergoing therapeutic radiation and the effect of antimycotic treatment. Mycoses. 1989;32(Suppl 2):39-41.
32. Zamost BJ, Hirschberg J, Ippoliti AF, et al. Esophagitis in scleroderma— prevalence and risk factors. Gastroenterology. 1987;92:421-428.
33. Ortuno Cortes JA, Tovar MA, Ruiz Riquelme J, Garcia GA. Esophageal candidiasis in HIV-negative patients. Rev Espan Enferm Digest. 1997;89:503-510.
34. Kodsi BE, Wickremesinghe PC, Kozi NNPJ, et al. Candida esophagitis: a prospective study of 27 cases. Gastroenterology. 1976;71:715-719.
35. Wilcox CM, Schwartz DA. Endoscopic-pathologic correlates of Candida esophagitis in AIDS. Dig Dis Sci. 1996;41:1337-1345.
36. Monkemuller KE, Bussian AH, Lazenby AJ, Wilcox CM. Special histologic stains are rarely beneficial for the evaluation of HIV-related gastrointestinal infections. Am J Clin Pathol. 2000;114:387-394.
37. Farman J, Tavitian A, Rosenthal LE, et al. Focal esophageal candidiasis in acquired immunodeficiency syndrome (AIDS). Gastrointest Radiol. 1986;11:213-217.
38. Levine MS, Macones AJ Jr, Laufer I. Candida esophagitis: accuracy of radiographic diagnosis. Radiology. 1985;154:581-587.
39. Roberts L Jr, Gibbons R, Gibbons G, et al. Adult esophageal candidiasis: a radiographic spectrum. Radiographics. 1987;7:289-307.
40. Bier SJ, Keller RJ, Krivisky BA, et al. Esophageal moniliasis: a new radiographic presentation. Am J Gastroenterol. 1985;80:734-737.
41. Bhalodia MV, Vega KJ, DaCosta J, Trotman BW. Esophageal candidoma in a patient with acquired immunodeficiency syndrome. J Assoc Acad Minor Phys. 1998;9:69-71.
42. Lai YP, Wu MS, Chen MY, Chuang CY, Shun CT, Lin JT. Timing and necessity of endoscopy in AIDS patients with dysphagia or odynophagia. Hepatogastroenterol. 1998;45:2186-2189.
43. Martinez EJ, Nord HJ, Cooper BG. Significance of solitary and multiple esophageal ulcers in patients with AIDS. South Med J. 1995;88:626-629.
44. McDonald GB, Sharma P, Hackman RC, et al. Esophageal infections in immunosuppressed patients after marrow Transplantation. Gastroenterology. 1985;88:1111-1117.
45. Bonacini M, Laine L, Gal AA, et al. Prospective evaluation of blind brushings of the esophagus for Candida esophagitis in patients with human immunodeficiency virus infection. Am J Gastroenterol. 1990;85:385-389.
46. Brandt LJ, Coman E, Schwartz E, et al. Use of a new cytology balloon for diagnosis of symptomatic esophageal disease in acquired immunodeficiency syndrome. Gastrointest Endosc. 1993;39:559-560.
47. Wilcox CM. Technique to examine the underlying mucosa in patients with AIDS and severe Candida esophagitis. Gastrointest Endosc. 1995;42:360-363.
48. Fifer WR, Woellner RC, Gordon SS. Mediastinal histoplasmosis— report of three cases with dysphagia as the presenting complaint. Dis Chest. 1965;47:518-526.
49. Goodman MI, Maher E. Four uncommon infections in Hodgkin's disease [Letter]. JAMA. 1966;198:203.
50. Forsmark CE, Wilcox CM, Darragh TM, et al. Disseminated histoplasmosis in AIDS: an unusual case of esophageal involvement and gastrointestinal bleeding. Gastrointest Endosc. 1990;36:604-605.
51. Cappell MS, Armenian BP. Esophagitis from Candida or exophiala? [Letter]. Ann Intern Med. 1991;115:69.
52. Coleman DC, Bennett DE, Sullivan DJ, et al. Oral Candida in HIV infection and AIDS: new perspective/new approaches. Crit Rev Microbiol. 1993;19:61-82.
53. Hoffman M, Bash E, Berger SA, et al. Fatal necrotizing esophagitis due to Penicillium chrysogenum in a patient with acquired immunodeficiency syndrome. Eur J Clin Microbiol Infect Dis. 1992;11:1158-1160.
54. Margolis PS, Epstein A. Mucormycosis esophagitis in a patient with the acquired immunodeficiency syndrome. Am J Gastroenterol. 1994;89:1900-1902.
55. Washington K, Gottfried MR, Wilson ML. Gastrointestinal cryptococcosis. Mod Pathol. 1991;4:707-711.
56. Rusconi S, Meroni L, Galli M. Tracheosophageal fistula in an HIV1 positive man due to dual infection of Candida albicans and cytomegalovirus. Chest. 1994;106:284-285.
57. Parente F, Cernuschi M, Valsecchi L, et al. Acute upper gastrointestinal bleeding in patients with AIDS: a relatively uncommon condition associated with reduced survival. Gut. 1991;32:987-990.
58. Gaissert HA, Breuer CK, Weissburg A, Mermel L. Surgical management of necrotizing Candida esophagitis. Ann Thor Surg. 1999;67:231-233.
59. Altman C, Bedossa P, Dussaix E, Buffet C. Cytomegalovirus infection of esophagus in immunocompetent adults. Dig Dis Sci. 1995;40:606- 608.
60. Venkataramani A, Schlueter AJ, Spech TJ, Greenberg E. Cytomegalovirus esophagitis in an immunocompetent host. Gastrointest Endosc. 1994;40:392-393.
61. Wiest PM, Flanigan T, Salata RA, et al. Serious infectious complications of corticosteroid therapy for COPD. Chest. 1989;95:180-184.
62. Buss DH, Scharyj M. Herpes virus infection of the esophagus and other visceral organs in adults: incidence and clinical significance. Am J Med. 1979;66:457-462.
63. Levy E, Margalith M, Sarov B, et al. Cytomegalovirus IgG and IgA serum antibodies in a study of HIV infection and HIV related diseases in homosexual men. J Med Virol. 1991;35:174-179.
64. Alexander JA, Cuellar RE, Fadden RJ, et al. Cytomegalovirus infection of the upper gastrointestinal tract before and after liver Transplantation. Transplantation. 1988;46:378-382.
65. Brouillette DE, Alexander J, Yoo YK, et al. T cell populations in liver and renal transplant recipients with infectious esophagitis. Dig Dis Sci. 1989;34:92-96.
66. Wheeler RR, Peacock JE, Cruz JM, et al. Esophagitis in the immunocompromised host: role of esophagoscopy in diagnosis. Rev Infect Dis. 1987;9:88-96.
67. Wilcox CM, Diehl DL, Cello JP, et al. Cytomegalovirus esophagitis in patients with AIDS. A clinical, endoscopic and pathologic correlation. Ann Intern Med. 1990;113:589-593.
68. Kitchen VS, Helbert M, Francis ND, et al. Epstein-Barr virus associated oesophageal ulcers in AIDS. Gut. 1990;31:1223-1225.
69. Schechter M, Pannain VLN, Oliveira AV. Papovavirus associated esophageal ulceration in a patient with AIDS [Letter]. AIDS. 1991;5:238.
70. Rotterdam H, Tsang P. Gastrointestinal disease in the immunocompromised patient. Mod Pathol. 1994;25:1123-1140.
71. Byard RW, Champion MC, Orizaga M. Variability in the clinical presentation and endoscopic findings of herpetic esophagitis. Endoscopy. 1987;19:153-155.
72. Levine MS. Radiology of esophagitis: a pattern approach. Radiology. 1991;179:1-7.
73. Levine MS, Loevner LA, Saul SH, et al. Herpes esophagitis: sensitivity of double contrast esophagography. AJR. 1988;151:57-62.
74. McBane RD, Gross JB. Herpes esophagitis: clinical syndrome, endoscopic appearance, and diagnosis in 23 patients. Gastrointest Endosc. 1991;37:600-603.
75. Ramanathan J, Rammouni M, Varan J Jr, Khatib R. Herpes simplex virus esophagitis in the immunocompetent host: an overview. Am J Gastroenterol. 2000;95:2171-2176.
76. Golden MP, Hammer SM, Wanke CA, Albrecht MA. Cytomegalovirus vasculitis: case reports and review of the literature. Medicine. 1994;73:246-255.
77. Balthazar EJ, Megibow AJ, Hulnick D, et al. Cytomegalovirus esophagitis in AIDS: radiographic features in 16 patients. AJR. 1987;149:919-923.
78. Wilcox CM, Straub RF, Schwartz DA. Prospective endoscopic characterization of cytomegalovirus esophageal ulceration in patients with AIDS. Gastrointest Endosc. 1994;40:481-484.
79. Laguna F, GarciaSamaniego J, Alonso MJ, et al. Pseudotumoral appearance of cytomegalovirus esophagitis and gastritis in AIDS patients. Am J Gastroenterol. 1993;88:1108-1111.
80. Wisser J, Zingman B, Wasik M, et al. Cytomegalovirus pseudotumor presenting as bowel obstruction in a patient with acquired immunodeficiency syndrome. Am J Gastroenterol. 1992;87:771-775.
81. Teot LA, Ducatman BS, Geisinger KR. Cytologic diagnosis of cytomegaloviral esophagitis: a report of three acquired immunodeficiency syndrome-related cases. Int Acad Cytol. 1993;37:93-96.
82. Culpepper Morgan JA, Kotler DP, Scholes JV, Tierney AR. Evaluation of diagnostic criteria for mucosal cytomegalic inclusion disease in the acquired immunodeficiency syndrome. Am J Gastroenterol. 1987;82:1264-1270.
83. Goodgame RW. Gastrointestinal cytomegalovirus disease. Ann Intern Med. 1993;119:924-935.
84. Levine DS, Reid BJ. Endoscopic biopsy technique for acquiring larger mucosal samples. Gastrointest Endosc. 1991;37:332-337.
85. Wilcox CM, Straub RF, Schwartz DA. Prospective evaluation of biopsy number for the diagnosis of viral esophagitis in HIV infection. Gastrointest Endosc. 1996;44;587-593.
86. Schwartz DA, Wilcox CM. Atypical cytomegalovirus inclusions in gastrointestinal biopsy specimens from patients with the acquired immunodeficiency syndrome: diagnostic role of in situ nucleic acid hybridization. Hum Pathol. 1992;23:1019-1026.
87. Beaugerie L, Cywiner-Golenzer C, Monfort L, Girard PM, Carbonnel F, Ngo Y, Cosnes J, Rozenbaum W, Nicolas JC, Chatelet FP, Gendre JP. Definition and diagnosis of cytomegalovirus colitis in patients infected by human immunodeficiency virus. J Acquir Imm Defic Syndr Hum Retrovir. 1997;14:423-429.
88. Cotte L, Drouet E, Bissuel F, et al. Diagnostic value of amplification of human cytomegalovirus DNA from gastrointestinal biopsies from human immunodeficiency virus-infected patients. J Clin Microbiol. 1993;31:2066-2069.
89. Goodgame RW, Genta RM, Estrada R, et al. Frequency of positive tests for cytomegalovirus in AIDS patients: endoscopic lesions compared with normal mucosa. Am J Gastroenterol. 1993;88:338-343.
90. Rattner HM, Cooper DJ, Zaman MB. Severe bleeding from herpes esophagitis. Am J Gastroenterol. 1985;80:523-525.
91. Bini EJ, Micale PL, Weinshel EH. Risk factors for rebleeding and mortality from acute upper gastrointestinal hemorrhage in human immunodeficiency virus infection. Am J Gastroenterol. 1999;94:358-363.
92. Cirillo NW, Lyon DT, Schuller AM. Tracheoesophageal fistula complicating herpes esophagitis in AIDS. Am J Gastroenterol. 1993;88:587.
93. Cronstedt JL, Bouchama A, Hainau B, et al. Spontaneous esophageal perforation in herpes simplex esophagitis. Am J Gastroenterol. 1992;87:124-127.
94. Marshall JB, Smart JR, Elmer C, et al. Herpes esophagitis causing an unsuspected esophageal food bolus impaction in an institutionalized patient. J Clin Gastroenterol. 1992;15:179-180.
95. Chalasani N, Parker KM, Wilcox CM. Bronchoesophageal fistula as a complication of cytomegalovirus esophagitis in AIDS. Endoscopy. 1997;29:S28.
96. Wilcox CM. Esophageal strictures complicating ulcerative esophagitis in patients with AIDS. Am J Gastroenterol. 1999;94:339-343.
97. Hill AR, Somasundaram P, Brustein S, et al. Disseminated tuberculosis in the acquired immunodeficiency syndrome era. Am Rev Respir Dis. 1991;144:1164-1170.
98. Hsiao CH, Huang SH, Huang SF, Song CL, Su IJ, Chuang CY, Yao YT, Lin CT, Hsu HC. Autopsy findings on patients with AIDS in Taiwan. J Microbiol Immun Infect. 1997;30:145-159.
99. Shafer RW, Dong DS, Weiss JP, Quale JM. Extrapulmonary tuberculosis in patients with human immunodeficiency virus infection. Medicine. 1991;70:384-397.
100. Lockard LB. Esophageal tuberculosis: a critical review. Laryngoscope. 1913;23:561-584.
101. Conjalka MS, Usselman J, Hassidim K, Freedman S. Successful medical treatment of a tuberculosis BE fistula. Mt Sinai Med (NY). 1980;47:283-284.
102. Porter JC, Friedland JS, Freedman AR. Tuberculosis bronchoesophageal fistulae in patients infected with the human immunodeficiency virus: three case reports and review. Clin Infect Dis. 1994;19:954-957.
103. McNamara M, Williams CE, Brown TS, Gopichandran TD. Tuberculosis affecting the esophagus. Clin Radiol. 1987;38:419-422.
104. Allen CM, Craze J, Grundy A. Tuberculosis bronchoesophageal fistula in the acquired immunodeficiency syndrome. Clin Radiol. 1991;43:60- 62.
105. Seivewright N, Freehally J, Wicks ACB. Primary tuberculosis of the esophagus. Am J Gastroenterol. 1984;79:847-853.
106. Damtew B, Frengley D, Wolinsky E, et al. Esophageal tuberculosis: mimicry of gastrointestinal malignancy. Rev Infect Dis. 1987;9:140-146.
107. Rosario MT, Raso CL, Comer GM. Esophageal tuberculosis. Dig Dis Sci. 1989;34:1281-1284.
108. Adkins MS, Raccuia JS, Acinapura AJ. Esophageal perforation in a patient with acquired immunodeficiency syndrome. Ann Thorac Surg. 1990;50:299-300.
109. Goodman P, Pinero SSM, Rance RM, et al. Mycobacterial esophagitis in AIDS. Gastrointest Radiol. 1989;14:103-105.
110. Ravera M. Tuberculous bronchoesophageal fistula in a patient infected with the HIV virus. Endoscopy. 1997;29:146.
111. Nightingale SD, Byrd LT, Southern PM, et al. Incidence of Mycobacterium avium intracellulare complex bacteremia in human immunodeficiency virus-positive patients. J Infect Dis. 1992;165:1082- 1085.
112. Torriani FJ, McCutchan JA, Bozzette SA, et al. Autopsy findings in AIDS patients with Mycobacterium avium complex bacteremia. J Infect Dis. 1994;170:1601-1605.
113. Stoopack PM, de Silva R, Raufman JP. Inflammatory double barreled esophagus in two patients with AIDS. Gastrointest Endosc. 1990;36:394-399.
114. Calore EE, Cavaliere JM, Perez NM, Campos Sales PS, Warnke KO. Esophageal ulcers in AIDS. Pathologica. 1997;89:155-158.
115. de Silva R, Stoopack PM, Raufman JP. Esophageal fistulas associated with mycobacterial infection in patients at risk for AIDS. Radiology. 1990;175:449-453.
116. Gray JR, Rabeneck L. Atypical mycobacterial infection of the gastrointestinal tract in AIDS patients. Am J Gastroenterol. 1989;84:1521.
117. Walsh TJ, Belitsos NJ, Hamilton SR. Bacterial esophagitis in immunocompromised patients. Arch Intern Med. 1986;146:1345-1348.
118. Isaac DW, Parham DM, Patrick CC. The role of esophagoscopy in diagnosis and management of esophagitis in children with cancer. Med Ped Oncol. 1997;28:299-303.
119. Ezzell JH, Bremer J, Adamec TA. Bacterial esophagitis: an often forgotten cause of odynophagia. Am J Gastroenterol. 1990;85:296-298.
120. Spencer GM, Roach D, Skucas J. Actinomycosis of the esophagus in a patient with AIDS: findings on barium esophagograms. AJR. 1993;161:795-796.
121. Kim J, Minamoto GY, Grieco MH. Nocardial infection as a complication of AIDS: report of six cases and review. Rev Infect Dis. 1991;13:624-629.
122. Poles MA, McMeeking AA, Scholes JV, Dieterich DT. Actinomyces infection of a cytomegalovirus esophageal ulcer in two patients with acquired immunodeficiency syndrome. Am J Gastroenterol. 1994;89:1569-1572.
123. Kazlow PG, Shah K, Benkov KJ, et al. Esophageal cryptosporidiosis in a child with acquired immune deficiency syndrome. Gastroenterology. 1986;91:1301-1303.
124. Grimes MM, LaPook JD, Bar MH, Wasserman HS. Disseminated Pneumocystis carinii infection in a patient with acquired immunodeficiency syndrome. Hum Pathol. 1987;18:307-308.
125. Gutierrez Macias A, Alonso Alonso JJ, Aguirre Errasti C. Esophageal Leishmaniasis in a patient with the human immunodeficiency virus [letter]. Clin Infect Dis. 1995;21:229-230.
126. Rosenthal E, Marty P, del Giudice P, Pradier C, Ceppi C, Gastaut JA, Le Fichoux Y, Cassuto JP. HIV and Leishmania coinfection: a review of 91 cases with focus on atypical locations of Leishmania. Clin Infect Dis. 2000;31:1093-1095.
127. Sollima S, Pizzuto M, Bonetto S, Ravasio L, Tosoni A, Vago L, Corbellino M, Antinori S. A case of esophageal leishmaniasis indicating visceral leishmaniasis in a patient with AIDS. Eur J Clin Microbiol Infect Dis. 1999;18:752-755.
128. Villanueva JL, Torre Cisneros J, Jurado R, et al. Leishmania esophagitis in an AIDS patient: an unusual form of visceral Leishmaniasis. Am J Gastroenterol. 1994;89:273-275.
129. Borczuk AC, Hagan R, Chipty F, Brandt LJ. Cytologic detection of Trichomonas esophagitis in a patient with acquired immunodeficiency syndrome. Diag Cytopathol. 1998;19:313-316.
130. Fusade T, Liony C, Joly P, et al. Ulcerative esophagitis during primary HIV infection. Am J Gastroenterol. 1992;87:1523.
131. Rabeneck L, Popovic M, Gartner S, et al. Acute HIV infection presenting with painful swallowing and esophageal ulcers. JAMA. 1990;263:2318-2322.
132. Schacker T, Collier AC, Hughes J, Shea T, Corey L. Clinical and epidemiologic features of primary HIV infection. Ann Intern Med. 1996;125:257-261.
133. Pena M, Martinez Lopez MA, Arnalich F, et al. Esophageal candidiasis associated with acute infection due to human immunodeficiency virus: case report and review. Rev Infect Dis. 1991;13:872-875.
134. Fisher J. Acute immunosuppression with HIV seroconversion [letter]. N Engl J Med. 1993;328:288-289.
135. Strohlein S, Posner G, Nicholas A, et al. Giant esophageal ulcers in patients with AIDS-related complex. Dysphagia. 1986;1:84-87.
136. Chawla PK, Ramani K, Chawla K, et al. Giant esophageal ulcers of AIDS: ultrastructural study. Am J Gastroenterol. 1994;89:411-415.
137. Smith PD, Eisner MS, Manischewitz JF, et al. Esophageal disease in AIDS is associated with pathologic processes rather than mucosal human immunodeficiency virus type 1. J Infect Dis. 1993;167:547-552.
138. Wilcox CM, Zaki SR, Coffield LM, et al. Evaluation of idiopathic esophageal ulcer for human immunodeficiency virus. Mod Pathol. 1995;8:568-572.
139. Levine MS, Loercher G, Katzka DA, et al. Giant human immunodeficiency virus-related ulcers in the esophagus. Radiology. 1991;180:323- 326.
140. Wilcox CM, Schwartz DA. Endoscopic characterization of idiopathic esophageal ulceration associated with human immunodeficiency virus infection. J Clin Gastroenterol. 1993;16:251-256.
141. Kimmel ME, Boylan JJ. Fistulous degeneration of a giant esophageal ulcer in a patient with acquired immunodeficiency syndrome. Am J Gastroenterol. 1991;86:898-900.
142. Kotler DP, Reka S, Orenstein JM, et al. Chronic idiopathic esophageal ulceration in the acquired immunodeficiency syndrome: characterization and treatment with steroids. J Clin Gastroenterol. 1992;15:284-290.
143. Cappell MS, Botros N. Predominantly gastrointestinal symptoms and signs in 11 consecutive AIDS patients with gastrointestinal lymphoma: a multicenter, multiyear study including 763 HIV seropositive patients. Am J Gastroenterol. 1994;89:545-549.
144. Friedman SL, Wright TL, Altman DF. Gastrointestinal Kaposi's sarcoma in patients with acquired immunodeficiency syndrome. Gastroenterology. 1985;89:102-108.
145. Rose HS, Balthazar EJ, Megibow AJ, et al. Alimentary tract involvement in Kaposi sarcoma: radiographic and endoscopic findings in 25 homosexual men. Am J Radiol. 1982;139:661-666.
146. Barrison IG, Foster S, Harris JW, et al. Upper gastrointestinal Kaposi's sarcoma in patients positive for HIV antibody without cutaneous disease. Br Med J. 1988;296:92-93.
147. Kadakia SC, Kadakia AS, Westphal KW. Gastrointestinal Kaposi's sarcoma as the first manifestation of acquired immunodeficiency syndrome. South Med J. 1992;85:37-39.
148. Reynolds P, Saunders LD, Layefsky ME, Lemp GF. The spectrum of acquired immunodeficiency syndrome (AIDS)-associated malignancies in San Francisco, 1980-1987. Am J Epidemiol. 1993;137:19-30.
149. Bernal A, del Junco GW. Endoscopic and pathologic features of esophageal lymphoma: a report of four cases in patients with acquired immune deficiency syndrome. Gastrointest Endosc. 1986;32:96-99.
150. Radin DR. Primary esophageal lymphoma in AIDS. Abdom Imaging. 1993;18:223-224.
151. Chalasani N, Parker K, Wilcox CM. Barrett's adenocarcinoma in a patient with acquired immunodeficiency syndrome. J Clin Gastroenterol. 1997;24:184-186.
152. Frager DH, Wolf EL, Competiello LS, et al. Squamous cell carcinoma of the esophagus in patients with the acquired immunodeficiency syndrome. Gastrointest Radiol. 1988;13:358-360.
153. Edwards P, Turner J, Gold J, et al. Esophageal ulceration induced by zidovudine. Ann Intern Med. 1990;112:65-66.
154. Indorf AS, Pegram PS. Esophageal ulceration related to zalcitabine (ddC). Ann Intern Med. 1992;117:133-134.
155. Shapiro BD, Ehrenpreis ED, Tomaka FL, Bonner GF, Secrest KM, Cheney LM. Idiopathic mid-esophageal stricture: a new cause of dysphagia in a patient with AIDS. South Med J. 1997;90:80-82.
156. Colina RE, Smith M, Kikendall JW, Wong RK. A new probable increasing cause of esophageal ulceration: alendronate. Am J Gastroenterol. 1997;92:704-706.
157. Belitsos PC, Greenson JK, Yardley JH, et al. Association of gastric hypoacidity with opportunistic enteric infections in patients with AIDS. J Infect Dis. 1992;166:277-284.
158. Wilcox CM, Waites KB, Smith PD. No relationship between gastric pH, small bowel bacterial colonisation, and diarrhea in HIV-1 infected patients. Gut. 1999;44:101-105.