THE AMERICAN JOURNAL Of GASTROENTEROLOOY
`Copyright© 1998 by Am. Coil. of Gastroenterology
`Published by Elsevier Science Inc.
`
`Editorials
`
`Vol. 93, No. 3, 1998
`ISSN 0002-9270/98/$19.00
`PIT S0002-9270(98)00002-l
`
`MENTAL STRESS AND PEPTIC ULCERS: AN
`EARTHSHAKING ASSOCIATION
`
`Stress can be defined as the biological response to an
`adverse physical or mental stimulus that disturbs an _organ(cid:173)
`ism's homeostasis. The term stress (or stressor) ts also
`sometimes used to refer to the stimuli that elicit the biolog(cid:173)
`ical responses (1 ). If the compensating responses are inad(cid:173)
`equate or inappropriate, stress may cause an organism to
`experience distress, disorder, and/or disease.
`Major physical stressors (e.g., bums, trauma, surgery,
`multiorgan failure) clearly can trigger biological responses
`that often result in acute gastric ulcers that bleed ("acute
`stress ulcers") (2). What is less clear is the relationship
`between major mental stressors, biological responses to
`these stressors, and ulcers (3). A cause-and-effect relation(cid:173)
`ship between mental stressors and peptic ulcer disease has
`long been accepted by the public, and individual case stud(cid:173)
`ies support this link (4). However, carefully controlled stud(cid:173)
`ies testing cause-and-effect have been difficult to perform.
`Nevertheless, there is some evidence that mental stress may
`lead to acute ulcer formation.
`For example, during the devastating air raids of London
`in 1940 and 1941, British physicians observed an increase of
`>50% in the incidence of acute perforated gastric ulcer,
`and, to a lesser extent, duodenal ulcer, compared with ulcer
`perforation rates in the same 16 hospitals in the years before
`or after the raids. The authors proposed that anxiety was
`responsible for the increase in ulcer perforation rates (5, 6).
`Weiner et al. (7) performed psychological tests and upper
`GI barium examinations in 120 men (63 serum pepsinogen
`hypersecretors and 57 hyposecretors) before their entering
`basic military training. Hypersecretors tended to be more
`dependent, immature, anxious, and depressed than hypose(cid:173)
`cretors. Four of the hypersecretors had radiologic evidence
`of duodenal ulcer disease at baseline. Upper GI x-rays were
`then repeated 8 to 16 wk into basic training. Five more
`hypersecretors had developed duodenal ulcers.
`More than a decade ago, Walker and I reported that male
`veterans hospitalized for bleeding or perforated duodenal or
`gastric ulcers were no more likely to have had recent stress(cid:173)
`ful events than were patients hospitalized concurrently for
`symptomatic kidney stones or gallstones (8). However, we
`found that ulcer patients had fewer skills and less social
`support to cope with these events than did their counterparts
`with other diseases. Ulcer patients were also more likely
`
`Received Dec. 5, 1997; accepted Dec. 12, 1997.
`
`291
`
`than were patients with stones to have interpreted events
`negatively, and to demonstrate anxiety and/or depression
`(distress). We also found a direct relationship between the
`degree of emotional disturbance and the elevation of serum
`pepsinogen concentrations in ulcer patients (9).
`In the 1980s, the discovery of H. pylori and its close
`relationship with peptic ulcer disease forced us to reconsider
`the role of mental stress in the pathogenesis of peptic ulcers,
`and rightly so. Eradicating H. pylori from the stomach with
`antimicrobial drugs led to very low duodenal ulcer recur(cid:173)
`rence rates regardless of life stresses ( 10), and cure of ulcer
`disease became possible for the first time in history. Al(cid:173)
`though the role of H. pylori gastritis in duodenal and gastric
`ulcer pathogenesis is now firmly established, the basic
`mechanism(s) by which H. pylori leads to these ulcers is not
`certain. H. pylori gastritis is usually acquired in childhood,
`but ulcers in infected individuals usually occur after a la(cid:173)
`tency period of years or even decades, and more often occur
`in the duodenum than in the stomach (11). Furthermore, the
`natural history of untreated or placebo-treated H. pylori(cid:173)
`associated duodenal and gastric ulcem is that they heal
`spontaneously, even though the H. pylori gastritis persists,
`only to recur weeks, months, or even years later. For in(cid:173)
`fected individuals there must be events that trigger these
`episodes of acute ulcer formation that are often accompa(cid:173)
`nied by serious complications such as bleeding or perfora(cid:173)
`tion. Ingestion of NSAIDs is one well established trigger for
`acute ulcer formation (12), but many ulcer episodes are not
`associated with NSAID use. Other triggers, such as mental
`stressors, could contribute to episodic ulcer formation in H.
`pylori-infected individuals. Also, because a considerable
`proportion of individuals with duodenal ulcers ( 10-20%) or
`gastric ulcers (20-40%) is not infected with H. pylori (11),
`some of their ulcers could be caused by mental stress.
`In this issue, Aoyama et at. report on the incidence of
`peptic ulcers after a major earthquake in the Hanshin-Awaji
`region of Japan in January of 1995 (13). This earthquake
`was the most devastating earthquake in Japan in nearly 75
`years, killing more than 5,000 people and leaving more than
`300,000 homeless. In this retrospective study, the investi(cid:173)
`gators recorded the number of patients with gastric or du(cid:173)
`odenal ulcers diagnosed by endoscopy during the 8-wk
`period after earthquake and compared this with the number
`of ulcers diagnosed in the same 61 hospitals during the same
`8-wk period the year before the earthquake. Data were
`analyzed from 23 hospitals near the epic·~nter (approximate(cid:173)
`ly 7 on the Richter scale), from 15 hospitals in the surround-
`
`

`

`292
`
`EDITORIALS
`
`ing area (Richter 5-6), and from 23 hospitals in a relatively
`spared peripheral area (Richter <4). No data were presented
`on ulcer perforations, presumably because ulcer perforations
`were diagnosed by surgeons rather than by endoscopists.
`There are a number of findings from this study that
`suggest that the stress of the earthquake may have affected
`the incidence, clinical presentation, and/or severity of ulcer
`disease. First, although 5,000 fewer endoscopies were per(cid:173)
`formed after the earthquake than in the previous year (33%
`decrease), the total number of ulcers diagnosed was similar
`(465 ulcers in 1994 and 454 in 1995), suggesting that the
`incidence of ulcers increased as a result of the natural
`disaster. Nevertheless, these data do not prove that the
`incidence of ulcers increased, inasmuch as it is possible
`(although not probable) that most individuals with symp(cid:173)
`tomatic ulcers after the earthquake were triaged to receive
`endoscopy.
`The clinical presentation of ulcer disease may have
`changed as well. After the earthquake, there was a shift to a
`higher ratio of gastric ulcers relative to duodenal ulcers,
`especially in the hardest hit regions, where this ratio in(cid:173)
`creased from <2:1 to >3:1. Furthermore, patients diag(cid:173)
`nosed with gastric ulcers after the earthquake were, on
`average, 10 yr older than in the previous year, whereas the
`ages of patients diagnosed with duodenal ulcers did not
`differ from year to year. There was also a trend for gastric
`ulcers diagnosed after the earthquake to more often be the
`initial ulcer episode for a given patient. These data suggest,
`but again do not prove, that the earthquake increased the
`incidence of gastric ulcers, especially in the elderly popu(cid:173)
`lation. Alternatively, the earthquake stress may have shifted
`the location of ulcers from the duodenum to the stomach,
`without necessarily increasing overall peptic ulcer inci(cid:173)
`dence.
`Last, the earthquake appeared to affect the severity of
`ulcer disease. The number of bleeding ulcers diagnosed by
`endoscopy nearly doubled, from 112 before to 222 after the
`earthquake. This increase was accounted for by 97 extra
`bleeding gastric ulcers and 13 extra bleeding duodenal ul(cid:173)
`cers. The increase in bleeding ulcers was not explained by
`increased NSAID use, which was low in both years (4-5%).
`The prevalence of H. pylori infection in ulcer patients was
`also similar before and after the earthquake (80-85%).
`Importantly, the increased incidence of bleeding ulcers was
`most apparent in the hardest hit areas. For example, near the
`epicenter, bleeding gastric and duodenal ulcers increased by
`186% and 138%, respectively, whereas no substantial
`changes were observed in the most remote hospitals. In the
`year before the earthquake, only nine patients with bleeding
`ulcers received blood transfusions (1.1 patients per week),
`whereas 63 ulcer patients received transfusions after the
`earthquake (7.9 patients per week).
`This sentinel study by Aoyama et al. suggests that the
`emotional havoc produced by the Hanshin-Awaji earth(cid:173)
`quake changed the clinical presentation and severity of
`peptic ulcer disease in this region of Japan. The findings
`
`AJG --Vol. 93, No. 3, 1998
`
`complement the data on ulcer perforations during the Lon(cid:173)
`don air raids in World War II, and lend further credence to
`the concept that major mental stresHors increase the inci(cid:173)
`dence of serious peptic ulcer disease. It will be important to
`confirm the findings of these studies prospectively and also
`to evaluate whether this type of stre~.s-associated ulcer can
`be prevented. Physicians and other first aid workers should
`be alert to an increased risk of bleeding gastric (and prob(cid:173)
`ably duodenal) ulcers in the early we;:ks after natural disas(cid:173)
`ters, even in persons who have sustained no physical inju(cid:173)
`ries. Our concept of acute stress ulcers needs to be
`broadened to include those associated with emotional as
`well as physical stressors.
`
`Mark Feldman, M.D.
`Department of Internal Medicine
`Dallas VA Medical Center and The University of
`Texas Southwestern Medical Center
`Dallas, Texas
`
`REFERENCES.
`
`l. Dorland's Medical Dictionary, 27th edition. Philadelphia: WB Saun(cid:173)
`ders, 1988.
`2. Soli AH. Peptic ulcer and its complications. In: Feldman M, Schar(cid:173)
`schmid! BF, Sleisenger MH, eds. Gastrointestinal and liver disease.
`Pathophysiology, diagnosis, management, 6th edition. Philadelphia:
`WB Saunders, 1997:620-78.
`3. Walker PM, Feldman M. Psychosomatic aspects of peptic ulcer dis(cid:173)
`ease: A multifactorial model of stress. Ga;troenterol Inter 1992;5:33-
`47.
`4. Peters MN, Richardson CT. Stressful life events, acid hypersecretion
`and ulcer disease. Gastroenterology 1983:84:114-9.
`5. Stewart DN, de R. Winser DM. Incidence of perforated peptic ulcer.
`Effect of heavy air-raids. Lancet 1942;11:259-61.
`6. Spicer CC, Stewart DN, de R. Winser DM. Perforated peptic ulcer
`during the period of heavy air-raids. Lan<:et 1944;1: 14.
`7. Weiner H, Thaler M, Reiser MF, et al. E1iology of duodenal ulcer. I.
`Relation of specific psychological characteristics to rate of gastric
`secretion (serum pepsinogen). Psychosom Med 1957;19:1-10.
`8. Feldman M, Walker P, Green JL, et al. Life events stress and psycho(cid:173)
`social factors in men with peptic ulcer disease. A multidimensional
`case-controlled study. Gastroenterology 1986;91: 1370-9.
`9. Walker P, Luther J, Samloff IM, eta!. Life event stress and psycho(cid:173)
`social factors in men with peptic ulcer disease. Gastroenterology
`1988;94:323-30.
`10. Helicobacter pylori in peptic ulcer disea!.e. NIH Concensus Confer(cid:173)
`ence. JAMA 1994;272:65-9.
`II. Peterson WL, Graham DY. Helicobacter pylori. In: Feldman M, Schar(cid:173)
`schmid! BF, Sleisenger MH, eds. Gastrointestinal and liver disease.
`Pathophysiology, diagnosis, management, 6th edition. Philadelphia:
`WB Saunders, 1997;604-19.
`12. Cryer B. Nonsteroidal anti-inflammatory drugs and gastrointestinal
`disease. In: Feldman M, Scharschmidt BF, Sleisenger MH, eds.
`Gastrointestinal and liver disease. Pathophysiology, diagnosis,
`management, 6th edition. Philadelphia: WB Saunders, 1997;343-
`57.
`13. Aoyama N, Kinoshita Y, Fujimoto S, et al. Peptic ulcers after the
`Hanshin-Awaji earthquake: Increased in<:idence of bleeding gastric
`ulcers. Am J Gastroenterol 1998;93:311-•5.
`
`Reprint requests and correspondence: Mark Feldman, M.D., VA Med(cid:173)
`ical Center (Ill), 4500 South Lancaster Road, Dallas, TX 75216.
`
`

`

`AJG- March 1998
`
`HOW TO TREAT THE CYTOMEGALOVIRUS
`TROLL
`
`Cytomegalovirus was once called the "troll of transplan(cid:173)
`tation." Trolls are mythical beings that wait in hiding be(cid:173)
`neath a bridge until an opportune time to jump up and attack
`an unsuspecting traveler. Like a troll, cytomegalovirus is
`hiding out in most healthy adults and almost all HIV(cid:173)
`infected patients. The opportune time for jumping out and
`causing clinical disease is when the immune system is
`compromised, either by immunosuppressive drugs (CMV is
`the most important infectious cause of complications and
`death in organ transplant recipients) or in the acquired
`immunodeficiency syndrome (CMV disease occurs in about
`one-third of AIDS patients during the course of their dis(cid:173)
`ease). The attack can be on a variety of organs, but all parts
`of the gastrointestinal tract are at risk. The attack causes
`erosions or ulcerations in the gastrointestinal mucosa and is
`manifested by symptoms referable to the level of the gas(cid:173)
`trointestinal tract involved (i.e., odynophagia or substernal
`pain for an esophageal ulcer; epigastric pain or melena for
`a gastric or duodenal lesion; lower abdominal pain, diarrhea,
`and hematochezia for colonic disease) (1 ). The attack can
`precipitate an abdominal emergency such as bowel perfo(cid:173)
`ration, appendicitis, or cholecystitis. But most often the
`symptoms are slowly progressive and require endoscopic
`diagnosis and medical treatment.
`The controversies about endoscopic diagnosis have been
`reviewed (2). Cultures, antigen assays, and DNA amplifi(cid:173)
`cation techniques performed on blood or urine do not reli(cid:173)
`ably identify patients with active CMV disease in some solid
`organ (3). Multiple biopsies (at least 10) of endoscopic
`lesions should be stained with hematoxylin and eosin and
`carefully examined by experienced pathologists for the
`characteristic cytopathic effect ( 4 ). Immunohistochemical
`staining may be helpful in some cases but is not as important
`as multiple biopsies routinely stained and carefully
`searched. Viral culture has little role. PCR on the biopsy
`specimen (5, 6) shows great potential but is still a research
`tool because of limited availability. There is a dream of
`using PCR-based assays on the blood as a key component in
`the prediction, diagnosis, and treatment of CMV disease, but
`this has yet to be realized (3).
`Cytomegalovirus intestinal disease is treated with either
`of two intravenous drugs: ganciclovir or foscamet. They
`inhibit CMV DNA synthesis by slightly different mecha(cid:173)
`nisms. Published practice guidelines (7) and technical re(cid:173)
`ports (8) indicate that both drugs are equivalent, that induc(cid:173)
`tion at full dose for 3-4 weeks should be used, that the
`choice of drug should be based on toxicity profiles, that
`maintenance should be used after first relapse, that relapse
`can be treated with the same drug, and that further recur(cid:173)
`rence or refractoriness should be treated by switching drugs
`or by using a combination of both drugs. These guidelines
`
`Received Oct. 23, 1997; accepted Oct. 23, 1997.
`
`EDITORIALS
`
`293
`
`are based on published uncontrolled trials that have shown
`that both ganciclovir and foscamet yield clinical improve(cid:173)
`ment in approximately 75% of affected patients. A random(cid:173)
`ized comparison showed similar result1; (9). Relapse rate
`after effective induction therapy is 20-50%. Maintenance
`therapy with once-daily infusions of approximately half the
`induction dose probably delays recurrence, but even with
`maintenance therapy gastrointestinal CMV disease can re(cid:173)
`lapse or progress (9, 10). Because of the high relapse rate,
`some physicians advise maintenance therapy for all patients
`after induction therapy to prevent or delay relapse. The
`disadvantages of placing all patients on maintenance are
`drug toxicity, the need for continuous intravenous access,
`patient inconvenience, and cost. Treating relapses with full
`dose ganciclovir or foscamet is effective in most patients
`( 11 ). The concurrent use of both drugs may lead to response
`in 90% of patients who fail to respond to either or both of
`the drugs when they are used alone (12).
`The study reported by Parente et al. in the current issue of
`the Journal supports several of the practice guidelines on the
`treatment of AIDS-related esophageal CMV disease (13).
`When choosing between the two drugs the main consider(cid:173)
`ation is not efficacy but rather toxicity, convenience, and
`cost. For example, a patient with cytop~nia, or mandatory
`use of other bone marrow toxic drugs, or a recent bone
`marrow transplant may be treated wi,:h foscamet. Con(cid:173)
`versely, a patient with renal insufficiency, mandatory use of
`other nephrotoxic drugs, or a recent renal transplant, will
`likely be treated with ganciclovir. One of the reasons that
`Parente et al. were unable to enroll more patients in their
`study was because almost one-half of the CMV infected
`patients could not ethically be considered for both drugs.
`Side effects occur in one-half of both ganciclovir and fos(cid:173)
`carnet treated patients (9), but switching to the alternative
`drug only is mandated in less than 1% of patients (14) if
`appropriate treatment algorithms are followed (15). Patient
`convenience is another issue. Ganciclovir requires a 30-min
`infusion time and foscamet a 2-h infusion time after a 1-h
`saline load (500-1000 cc of normal saline). Annual costs of
`treatment, including administration and monitoring for tox(cid:173)
`icity, can vary between $50,000 and $100,000 (16).
`The study by Parente et al. shows in the gastrointestinal
`tract what was demonstrated for CMV retinitis quite a few
`years ago: ganciclovir and foscamet have similar efficacy
`and safety (14). Because CMV retinids is 10 times as
`common in AIDS than gastrointestinal CMV disease (17), it
`is not surprising that ophthalmologists and their infectious
`disease colleagues are ahead in other CMV -related areas as
`well, especially in the areas of new dmg testing, mainte(cid:173)
`nance, and treating relapse. Because CMV is such a com(cid:173)
`mon problem for organ transplant services, transplant phy(cid:173)
`sicians have also made important strides in the diagnosis and
`prevention of CMV disease. Gastroenterologists should be
`aware of the following issues on which data and experience
`are being collected.
`Failure of therapy: Approximately 20% of patients re-
`
`

`

`294
`
`EDITORIALS
`
`ported by Parente et al. (as with most other series) did not
`respond to drug treatment. Experience from studies in
`AIDS-related CMV retinitis would suggest that there is no
`single explanation for failure to respond (17). Although
`resistant strains to both foscarnet and ganciclovir have been
`identified ( 18), the vast majority of therapeutic failures are
`not associated with resistant strains (19). Failure of therapy
`must be related to a combination of multiple causes such as
`drug levels (17, 20), compliance, length of therapy, misdi(cid:173)
`agnosis, or, most importantly, progressive immune defi(cid:173)
`ciency. Treatment is likely to be more successful when there
`is a return to more normal immunologic function, as in a
`transplant recipient. Immune therapy, in the form of mono(cid:173)
`clonal anti-CMV antibodies, combined with either foscarnet
`or ganciclovir, are more effective than either drug by itself
`(16).
`Relapsing disease: Parente eta[. don't tell us how they
`managed the patients that responded to therapy. The prob(cid:173)
`lem of relapsing disease in spite of maintenance therapy
`continues to be studied in patients with AIDS-related CMV
`retinitis (17, 21). Three foscarnet/ganciclovir strategies for
`relapsing disease while on maintenance were compared:
`repeat induction with the same drug, switching to the other
`drug, or giving both drugs in combination. The combination
`was superior in halting progression but no different in sur(cid:173)
`vival or sight sparing (22). The combination was also ex(cid:173)
`pensive, time consuming, and toxic.
`New drugs and drug formulations are emerging to deal
`with relapsing disease (17). Although the oral bioavailabil(cid:173)
`ity of ganciclovir is only 6-9%, an oral formulation is
`available, which, at the recommended dose ( 1000 mg 3
`times daily) achieves blood levels close to the median ef(cid:173)
`fective inhibitory dose of most clinical isolates (19). It is
`almost as effective as intravenous ganciclovir in mainte(cid:173)
`nance therapy for CMV retinitis (19, 23), but the frequency
`of neutropenia and catheter related infections is reduced
`(17). Cidofovir, a nucleotide analogue with very potent
`activity against CMV in humans, has been used to prevent
`the progression of CMV retinitis in AIDS patients (24, 25).
`It is equivalent to ganciclovir and foscarnet for this purpose
`(24, 25). Its advantage is that it can be given as a weekly
`intravenous infusion because it has a long intracellular half(cid:173)
`life and prolonged effect against CMV (24). The drug is
`nephrotoxic (17).
`Monitoring treatment: Parente et al. used quantitative
`PCR and quantitative antigen assays (infected cells per
`200,000 polymorphonuclear leukocytes) to monitor the
`treatment. Measurement of the disappearance of leukocyte
`or serum CMV DNA positivity as a measure of therapeutic
`response has been previously studied, but not in association
`with gastrointestinal infection (3). Quantitative PCR has
`shown that standard therapy is associated with a fall in the
`amount of CMV genome in gastrointestinal tissues, but that
`20% of patients don't demonstrate such a fall (26). These
`patients are at higher risk of relapse. PCR predicts relapse of
`GI CMV infection in intestinal transplant recipients (5).
`
`AJG --Vol. 93, No.3, 1998
`
`Prevention: The CMV status of donor and recipient are
`the strongest predictors of survival in liver transplant pa(cid:173)
`tients (27). Therefore, studies have been done to show how
`to prevent disease in this population. Intravenous ganciclo(cid:173)
`vir prevents serious CMV disease m patients after bone
`marrow, cardiac, and liver transplantation (28-30). In AIDS
`patients with low CD4 counts, a year of oral ganciclovir
`reduced the incidence of CMV disease by 50% compared
`with placebo (31). Resistance to ganciclovir was not iden(cid:173)
`tified excessively and toxicity was mild. Treatment required
`taking 12 capsules per day at an annual drug cost of
`$14,000. There are no studies that Hhow that prophylaxis
`yields better outcomes than diagnosis of disease followed by
`treatment. Routine use of oral ganciclovir as prophylaxis has
`not been recommended (32).
`Preemptive treatment: Preemptive therapy is differenti(cid:173)
`ated from nontargeted prophylaxis by selecting a population
`at greatest risk based on some clinical measurement such as
`type of immunosuppression, antibody status, viremia, DNA
`level, etc. (33). This eliminates the need to give prophylactic
`therapy to everyone (saving money, toxicity, and inconve(cid:173)
`nience) while preventing the great cost, morbidity, and
`mortality of treating active CMV di~.ease. Monoclonal an(cid:173)
`tibody-based stains of peripheral blood polymorphonuclear
`leukocytes for CMV antigen or quantitative PCR (like the
`tests used by Parente et al.) may provide a quantitative
`expression of risk of CMV in the transplant setting, and the
`strategy of PCR or antigen-based assay directed preemptive
`therapy is emerging (34-37). In AIDS, there are no vali(cid:173)
`dated tests that accurately predict who will get CMV disease
`(32); however, studies are being done (38-40).
`Highly active antiretroviral therapy: A final but impor(cid:173)
`tant consideration is the role of highly active antiretroviral
`therapy in dealing with AIDS-related CMV infection. The
`impact of this treatment on the frequency and severity of
`other opportunistic infections has been impressive. AIDS
`diarrhea, along with its causative agents, is dramatically
`decreasing in incidence and severity (personal unpublished
`observations). Although such therapy may change many
`things about the presentation, natural history, and response
`to therapy of AIDS-related CMV infe,;tions (41), improving
`the immune status of the patient may well be the most
`important way to drive the CMV troll back under the bridge
`for an extended period of time.
`Richard W. Goodgame, M.D.
`Baylor College of Medicine
`Gastroenterology Section
`Houston, Texas
`
`REFERENCES
`
`I. Goodgame R. Gastrointestinal cytomegalovirus disease. Ann Intern
`Med 1993;199.
`2. Bonacini M, Laine L. Detection of cyton:egalovirus in the gastroin(cid:173)
`testinal tract: Seeking a gold standard. Am J Gastroenterol 1993;88:
`332-3.
`3. Cartwright C. Laboratory diagnosis. In: Masur H, ed. Advances in the
`
`

`

`AJG- March 1998
`
`management of AIDS-related cytomegalovirus retinitis. Ann Intern
`Med 1996;125:126-36.
`4. Wilcox C, Straub R, Schwartz D. Prospective evaluation of biopsy
`number for the diagnosis of viral esophagitis in patients with HIV
`infection and esophageal ulcer. Gastrointestinal Endosc 1996;44:587-
`93.
`5. Kusne S, Manez RF, St. George K, et al. Use of DNA amplification for
`the diagnosis of cytomegalovirus enteritis after intestinal transplanta(cid:173)
`tion. Gastroenterology 1997; 112: 1121-8.
`6. 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;31:2066-9.
`7. Dieterich D, Wilcox C. Diagnosis and treatment of esophageal diseases
`associated with HIV infection. Am J Gastroenterol 1996;91:2265-9.
`8. Wilcox C, Rabeneck L, Friedman S. AGA technical review: Malnu(cid:173)
`trition and cachexia, chronic diarrhea, and hepatobiliary disease in
`patients with human immunodeficiency virus infection. Gastroenter(cid:173)
`ology 1996;111:1724-52.
`9. Blanshard C, Benhamou Y, Dohin E, et al. Treatment of AIDS(cid:173)
`associated gastrointestinal cytomegalovirus infection with foscarnet
`and ganciclovir: A randomized comparison. J Infect Dis 1995;167:
`622-6.
`I 0. Erice A, Chou S, Biron K, et al. Progressive disease due to ganciclovir(cid:173)
`resistant cytomegalovirus in immunocompromised hosts. N Eng!
`J Med 1989;320:289-93.
`II. Dieterich D, Poles M, Dicker M, et al. Foscarnet treatment of cyto(cid:173)
`megalovirus gastrointestinal infections in acquired immunodeficiency
`syndrome patients who have failed ganciclovir induction. Am J Gas(cid:173)
`troenterol 1993;88:542-8.
`12. Dieterich D, Poles M, Lew E, et al. Concurrent use of ganciclovir and
`foscarnet to treat cytomegalovirus infection in AIDS patients. J Infect
`Dis 1993;167:1184-8.
`13. Parente F, Bianchi Porro G, Group TICS. Treatment of cytomegalo(cid:173)
`virus esophagitis in patients with acquired immune deficiency syn(cid:173)
`drome: A randomized controlled study of foscarnet versus ganciclovir.
`Am J Gastroenterol 1998;93:317-22.
`14. Studies of Ocular Complications of AIDS Research Group, AIDS
`Clinical Trials Group. Mortality in patients with the acquired immu(cid:173)
`nodeficiency syndrome treated with either foscarnet or ganciclovir for
`cytomegalovirus retinitis. N Eng! J Med 1992;326:213-20.
`15. Studies of Ocular Complications of AIDS Research Group, AIDS
`Clinical Trials Group. Morbidity and toxic effects associated with
`ganciclovir or foscarnet therapy in a randomized cytomegalovirus
`retinitis trial. Arch Intern Med 1995;155:65-74.
`16. Polls M. Systemic treatment. In: Masur H, ed. Advances in the man(cid:173)
`agement of AIDS-related cytomegalovirus retinitis. Ann Intern Med
`1996;125: 126-36.
`17. Jacobson M. Treatment of cytomegalovirus retinitis in patients with the
`acquired immunodeficiency syndrome. N Eng! J Med 1997;337:105-
`14.
`18. Dunn J, MacCumber M, Forman M, et al. Viral sensitivity testing in
`patients with cytomegalovirus retinitis clinically resistant to foscarnet
`or ganciclovir. Am J Ophthalmol 1995;119:587-96.
`19. Crumpacker C. Ganciclovir. N Eng! J Med 1996;335:721-9.
`20. Piketty C, Bardin C, Gilquin J, et al. Low plasma concentrations
`achieved with conventional schedules of administration of ganciclovir
`in patients with AIDS. J Infect Dis 1996;174:188-90.
`21. Young L. Therapy for cytomegalovirus retinitis: Still no silver lining.
`JAMA 1996;275:149-50.
`22. Studies of Ocular Complications of AIDS Research Group, AIDS
`Clinical Trials Group. Combination foscarnet and ganciclovir therapy
`vs. monotherapy for the treatment of relapsed cytomegalovirus retinitis
`in patients with AIDS: The cytomegalovirus retreatment trial. Arch
`Ophthalmol 1996;114:23-33.
`23. Drew W, Ives D, Lalezari JEA. Oral ganciclovir as maintenance
`treatment for cytomegalovirus retinitis in patients with AIDS. N Eng!
`J Med 1995;333:615-20.
`24. Studies of Ocular Complications of AIDS Research Group, AIDS
`Clinical Trials Group. Parenteral cidofovir for cytomegalovirus reti(cid:173)
`nitis in patients with AIDS: The HPMPC peripheral cytomegalovirus
`retinitis trial. Ann Intern Med 1997:126:264-74.
`25. Lalezari J, Stagg R, Kuppermann B, et al. Intravenous cidofovir for
`
`EDITORIALS
`
`295
`
`peripheral cytomegalovirus retinitis in patients with AIDS. Ann Intern
`Med 1997;126:257-63.
`26. Cotte L, Drouet E, Bailly F, et al. Cytomegalovirus DNA level on
`biopsy specimens during treatment of cytomegalovirus gastrointestinal
`disease. Gastroenterology 1996;111:439-44.
`27. Falagas M, Snydman D, Griffith J, et al. Group at BCfLTCS. Effect of
`cytomegalovirus infection status on first-ye.rr mortality rates among
`orthotopic liver transplant recipients. Ann Intern Med 1997; 126:275-9.
`28. Schmidt G, Horak D, Niland J, et al. A randomized controlled trial of
`prophylactic ganciclovir for cytomegalovirus pulmonary infection in
`recipients of bone marrow transplants. N Eng! J Med 1991;324:1005-
`11.
`29. Merigan T, Renlund D, Keay SEA. A controlled trial of ganciclovir to
`prevent cytomegalovirus disease after heart transplantation. N Eng!
`J Med 1992;326:1182-6.
`30. Winston D, Wirin D, Shaked A, et al. Randomized comparison of
`ganciclovir and high-dose acyclovir for long-term cytomegalovirus
`prophylaxis in liver transplant recipients. Lancet 1995;345:69-74.
`31. Spector S, McKinley G, Lalezari J, et al. Oral ganciclovir for the
`prevention of cytomegalovirus disease in persons with AIDS. N Eng!
`J Med 1995;334:1491-7.
`32. Masur H. Prevention. In: Masur H, ed. Advances in the management
`of AIDS-related cytomegalovirus retinitis. Am Intern Med 1996;125:
`126-36.
`33. Rubin R. Preemptive therapy in immunocompromised hosts. N Eng!
`J Med 1991;324:1057-9.
`34. Imbert-Marcille B, Cantarovich D, Ferre-Aubineau V, et al. Usefulness
`of DNA viral load quantification for cytomegalovirus disease moni(cid:173)
`toring in renal and pancreas/renal transplant recipients. Transplantation
`1997;63: 1476-81.
`35. Brennan D, Garlock K, Lippmann B, et al. Control of cytomegalovi(cid:173)
`rus-associated morbidity in renal transplant patients using intensive
`monitoring and either preemptive or deferred therapy. J Am Soc
`Nephrol 1997;8:118-25.
`36. Abecassis M, Koffron A, Kaplan B, et al. The role of PCR in the
`diagnosis and management of CMV in solid organ recipients: What is
`the predictive value for the development of disease and should PCR be
`used to guide antiviral therapy? Transplantation 1997;63.
`37. Prentice H, Kho P. Clinical strategies for the management of cyto(cid:173)
`megalovirus infection and disease in allogeneic bone marrow trans(cid:173)
`plant. Bone marrow Transplantation 1997;19:135-42.
`38. Dodt K, Jacobsen P, Hofmann B, et al. Development of cytomegalo(cid:173)
`virus (CMV) disease may be predicted in IllY-infected patients by
`CMV polymerase chain reaction and the antig enemia test. AIDS 1997;
`11:F21-8.
`39. Shinkai M, Bozzette S, Powderly W, et al. Utility of urine and leuko(cid:173)
`cyte cultures and plasma DNA polymerase chain reaction for identi(cid:173)
`fication of AIDS patients at risk for developing human cytomegalo(cid:173)
`virus disease. J Infect Dis 1997;175:302-9.
`40. Rasmussen L, Morris S, Zipeto D, et al. Quantit