- Email: [email protected]
Cytomegalovirus infection in AIDS. Patterns of disease, response to therapy and trends in survival
Journal of Infection (I99I) z3, I29-I37
C y t o m e g a l o v i r u s i n f e c t i o n in A I D S . P a t t e r n s o f disease, r e s p o n s e to t h e r a p y a n d t r e n d s in s u r v i v a l B. S. Peters,* E. J. Beck, t S. Anderson,* D. Coleman,* R. Coker, jJ J. Main,~ C. Migdal,~ J. R. W. Harris ]] and A. J. Pinching* * Department of Immunology, afAcademic Department of Public Health, Departments of ~ Infectious Diseases and Medicine, ~ Ophthalmology, and []Genito-urinary Medicine, St Mary's Hospital Medical School, Norfolk Place, London W2, U.K. Accepted for publication 25 March I991 Summary Among 347 AIDS patients seen at St Mary's Hospital, London between I983 and I989, cytomegalovirus (CMV) disease was observed in 75 (22 %). Of these, 58 (77 %) had CMV retinitis, 26 (35%) CMV colitis, and I2 (I6%) had CMV infection diagnosed at other sites. Relapse occurred in 7I %. A favourable response to the use of ganciclovir as induction therapy for CMV retinitis was observed in 92 %. Relapse of CMV retinitis occurred in 54 % at a median time of 97 days. Neutropenia was the most frequent and serious side-effect of ganciclovir, 76 % patients having neutrophil counts < I'O X I o g / l and 48% < 0'5 x IO9/1 at some stage of therapy. Thrombocytopenia was also common, and platelet counts of < 5o x IO9/1 occurred in 43 % patients on ganciclovir. The concurrent use of zidovudine made the development of severe neutropenia and thrombocytopenia more likely. Median survival following the diagnosis of CMV disease increased from 5-8 months between I984 and I987, to over I2 months in i988. Patients with CMV colitis had a worse prognosis than patients with CMV retinitis, with median survival of 4"5 and 7 months respectively. In conclusion, CMV is an important opportunist infection in AIDS and both the disease and its treatment cause considerable morbidity. Hence, it is important to develop more effective and less toxic forms of therapy for CMV infection.
Introduction Cytomegalovirus is one of the most i m p o r t a n t viral opportunist infections of A I D S patients. T h e most c o m m o n site of disease is the retina, followed by the large bowel, although other parts of the alimentary tract, the liver and biliary tree, the adrenal glands, the central nervous system and the lungs are among other sites that can be involved. Hence, unlike Pneumocystis carinii, which usually affects one target organ, the lungs, C M V is far more extensive and current treatments for it are only partly effective and m u s t be given intravenously. As the severity of P. carinii p n e u m o n i a (PCP) and its mortality have lessened in response to earlier diagnosis and improved m a n a g e m e n t , C M V has assumed increasing importance as an opportunist infection in A I D S . F o r this reason we decided to study its prevalence in our A I D S population, as well as the consequences of c o n t e m p o r a r y therapy, so that we can identify those aspects of C M V disease which merit f u r t h e r research. oI63-4453/9I/o5o~29 +o9 $03.00/0
© I99I The British Society for the Study of Infection
I30
B. S. P E T E R S E T A L .
Patients and m e t h o d s
T h e hospital records of all patients with A I D S managed at St Mary's Hospital, London between March I983 and December r989, were reviewed. Twelve patients whose case notes were unavailable were excluded from analysis, leaving a population of 347 under study. T h e majority (3Io) of our patients were homosexual men, 23 were bisexual men, and 14 were heterosexual, five of whom were women; seven were intravenous drug users. T h e racial composition was 33o white, Io Afro-Caribbean, four Hispanic and three Asian. T h e following information was recorded for all patients: age, sex, risk factors for H I V infection, date of A I D S diagnosis, index diagnosis and date of death. Patients with CMV retinitis or colitis were identified and the site and date of each episode of CMV disease was recorded. T h e ophthalmoscopic diagnosis of CMV retinitis was based on the findings of pale irregular areas of retinal necrosis with or without accompanying haemorrhages. Retinal examination was performed after mydriasis by both the physician and an ophthalmologist. CMV colitis was diagnosed by the presence of abdominal pain, usually accompanied by peritonism, with or without diarrhoea, and with evidence of CMV inclusion bodies on rectal biopsy. Subsequent episodes of CMV colitis could be diagnosed on clinical criteria alone, providing there was a rapid response to ganciclovir therapy. Diagnosis at other sites was made after histological examination of the appropriate biopsy material. Since I984 we have routinely used ganciclovir as first-line treatment and phosphonoformate as second line, both initially as part of clinical trials. 1' 2 Induction therapy was for 2 weeks, using ganciclovir Io m g / k g / d a y in two divided doses and for phosphoformate 18o m g / k g / d a y in three divided doses. We routinely provided maintenance therapy for all episodes of CMV retinitis, but only for colitis when relapsing. Maintenance therapy consisted of ganciclovir Io m g / k g / d a y Iv three times a week. If CMV infection recurred, the patient was reinduced, but we then used five divided doses weekly for maintenance. Maintenance phosphonoformate therapy was given as a 9o m g / k g / d a y infusion on 5-7 days each week. Ganciclovir was the firstchoice therapy but if there was a failure of initial response or neutrophil count less than o'5o × r&/1 despite stopping zidovudine, phosphonoformate treatment was substituted. Since its licensure in April I987 the majority of patients with A I D S have been prescribed zidovudine in doses of 100o--1200 mg/24 h. Most of those taking zidovudine, 22/28 (79%), were continued on it when ganciclovir was introduced. This contrasts with the practice in the U.S.A. where the two drugs are rarely given together. If the neutrophil count fell below 0-8 × lO9/1 the zidovudine dose was reduced to 5oo rag/24 h and blood counts taken weekly; zidovudine was stopped if the count fell to o'5 x lO9/1 or less. T h e lowest haemoglobin, neutrophil and platelet counts following the use of ganciclovir were recorded. Statistical analysis
Using the log-rank test, survival analyses were performed from the date of diagnosis of A I D S and the date of diagnosis of CMV disease; similarly,
Cytomegalovirus infection in A I D S
I3 1
2t 1.0.
L
/
0'4
0.2
I 0
I
I
I0
I 20
I
50
Months
Fig. I. Two-year survival from diagnosis of A I D S in patients who subsequently developed C M V infection. (D), 84; ( 0 ) , 85"; ([]), 86; (~}), 87; (m), 88.
survival was compared for those patients presenting with C M V colitis and C M V retinitis. Other statistical tests used included the Chi-squared (X2) test and the Fischer exact test, as indicated in the text. Nanostat software was used. Results Incidence and presentation
Cytomegalovirus disease was identified in 75 ( 2 2 ~/o) of our patients. Of these 75 patients 32 (43 %) had C M V retinitis alone, I4 (I9 %) had C M V colitis alone, and 13 (I7~o) had both. Twelve patients (z6 %) had I6 episodes of cytomegalovirus disease diagnosed at other sites, (six oesophageal, one gastric, one oropharyngeal, three pulmonary, four cerebral, one cardiac). T h e average annual n u m b e r of new patients diagnosed with C M V disease between I984 and I989 was I2 per year (range 8-2o). Average age at time of diagnosis was 36 years (range 22--57 ) and no significant changes occurred over time. Four percent (n = I5) of all patients presented with C M V disease as their A I D S defining diagnosis: 2 0 ~/o developed C M V disease within 3 m o n t h s of their A I D S diagnosis, 43 % within I2 m o n t h s and 8o % within 24 months. Overall 7I % (53) patients with C M V disease had recurrent symptomatic episodes: 5 o % of those with colitis had recurrence in the same site and, despite using maintenance therapy in all patients with C M V retinitis since I984, recurrent ocular disease occurred in 55 %. T h e median n u m b e r of days
I32
B. S. PETERS E T A L . 1'2
I'OJ
0.81
> 0"6
#,
0.4 • °
0.2
"
. . . . . . . . . . . . . . . . . . . .
, 0
.,°!
y!
:!___, 20
I0
30
Months
Fig. 2. Two-year survival from diagnosis of CMV infection to death in AIDS patients for various years. (V]), 84; (O), 85; ([]), 86; (~), 87; (m), 88.
to recurrence of C M V retinitis was 97 (range I5-396) compared with 58 days (range 21-216) for colitis. Retinal detachment occurred in I I / 5 8 ( I 9 % ) patients with C M V retinitis. Survival
T h e median survival of patients with C M V disease measured from the date of diagnosis of A I D S has increased from I 4 - I 6 m o n t h s for the I984-I986 cohorts to 24 m o n t h s for the I987 cohort, although this i m p r o v e m e n t did not reach statistical significance, probably because of insufficient n u m b e r s (Fig. I ; log-rank X, 2 = 4"6, P = o'327). Survival from the date of C M V diagnosis has not changed significantly, with a median survival ranging between 5 and I2 m o n t h s for the years I984-I987 (Fig. 2; log-rank X42= I.I, P = 0"89). However, a significant difference in survival was observed in those patients with C M V colitis compared with C M V retinitis; the median survival in the former group was 4"5 months, compared with 7 m o n t h s for those with C M V retinitis (Fig. 3; log-rank X2 = 8"0, P = 0"005). C o m p l i c a t i o n s o f therapy for CMV infection
Bone marrow toxicity was the most frequent complication of ganciclovir therapy (Table I), although nausea developed in three, skin rash in two and fevers in one patient. Severe neutropenia ( < 0"5 x io9/1) occurred in 31/65 (48 %) patients taking ganciclovir, and severe thrombocytopenia ( < 5o x Io9/1)
Cytomegalovirus infection in AIDS
I33
I'0
\
!
0-;
0"2
i !
L
! O.
. . . . . .
.
. . . . n
I
I 6
0
18
12
24
Mont"hs
Fig. 3. Two-year survival from diagnosis of CMV by site in AIDS patients. (--0--), Colitis; ( - - A - - ) , retinitis.
Table I
Haematological complications of ganciclovir therapy
Lowest value* Neutrophils ( x IO9/1) > I'5 I'O-I" 5 O'5--I'O
< o'5 Platelets ( x io9/1) > I5o IOO--I5O 5o-Ioo < 50 Haemoglobin (g/dl) > Io 7-Io < 7 Total
Number (%)
Number on zidovudine (%)
6 (9 %) II (I7%) 17 (26 %) 3I (48 %)
i 3 7 I7
(I7 %) (27%) (41%) (55 %)
9 18 Io 28
(I4%) (28 %) (I5 %) (43 %)
3 (33%) 7 (39 %) 5 (50 %) I3 (46 %)
6 22 37 65
(9 %) (34%) (57 %) (ioo %)
2 8 I8 28
(33 %) (36%) (49 %) (43 %)
* This refers to the lowest recorded value during ganciclovir therapy.
30
134
B. S. P E T E R S
ET AL.
in 28 (43 %); the concurrent use of zidovudine exacerbated these cytopenias (Table I). Frequent haematological monitoring, a reduction in dose or withdrawal of zidovudine and a switch to phosphonoformate when appropriate, meant that serious complications were usually avoided. However, septicaemia occurred in seven patients with neutropenia (six bacterial, one fungal) with one death. Serious haemorrhage occurred in three patients, with the bowel as the main site. All cases were successfully managed with platelet transfusion, stopping zidovudine, and withdrawal of ganciclovir in favour of phosphonoformate therapy. Of 28 patients on ganciclovir and zidovudine, zidovudine had to be stopped in 25 (89 %), with recovery of neutrophil and platelet counts in the majority of cases. Discussion
Cytomegalovirus is the most frequent opportunist infection to involve the eye in A I D S patients in the Western World. T h e prevalence of C M V retinitis, in I7 % of our patients, compares with figures of 6-40 % for series from U.S.A. centres, 3-1°. T h e differences may reflect recruitment biases, especially as several of the series with higher figures refer only to patients attending ophthalmology clinics. Apart from one early study using phosphonoformate, 2 ganciclovir has been our treatment of first choice for C M V disease because its longer serum halflife allows for shorter infusion regimens, and also because several studies demonstrating its efficacy have led to its licensure as first line therapy; hence we are unable to compare its efficacy with phosphonoformate. A favourable clinical response to induction therapy was recorded in 92 % of our patients. Comparable figures from other series vary from 5 7 - I o o %)1-22 This wide variation may reflect the different measures used to assess disease outcome, rather than real differences in therapeutic response. For example, the lowest figure (57 %) belonged to the centre using the widest definition of disease progression, u W i t h o u t maintenance therapy the median onset of clinical relapse has been reported to be between I9 and 47 day s.z~'~ T h e median onset, 97 days, of relapse of C M V retinitis in our patients compares favourably to figures of 54-1o5 days from other centres using similar total weekly dose maintenance therapy. 23'~4 This suggests that five times weekly ganciclovir maintenance, as routinely used by other centres, has no advantage over our three times weekly regimen. Another group has reported similar results using thrice weekly maintenance, with progression to C M V retinitis occurring in 40 % subjects at a mean 123 days. 25 T h e treatment of C M V retinitis has important consequences, both in terms of morbidity and the use of resources. T h e diagnosis of C M V retinitis results in the hospitalisation of the patient for induction therapy and insertion of an indwelling venous catheter (Hickman line or Portacath). Maintenance therapy with ganciclovir is expensive, requires more frequent blood counts, and our patients receive regular monitoring by ophthalmologists. T h e most important toxic effect of ganciclovir is myelotoxicity, particularly neutropenia. Zidovudine also causes bone marrow suppression, 26 and most of
Cytomegalovirus infection in A I D S
r35
our patients had zidovudine withdrawn because of severe neutropenia when used with ganciclovir. One particular consequence of neutropenia in our patients has been an increased rate of infection of indwelling venous catheters, especially with staphylococcal species. T h e median survival of our patients following their diagnosis of CMV for the years I984-I988 ranged between 4 and 9 months, with no significant change over time, and compares with figures from other series of 5"5 and 6 months following a diagnosis of CMV retinitis. 12'18 Although the onset of CMV disease is considered a poor prognostic factor this may partly be because CMV infection tends to occur when CD4 counts drop to very low levels ( < 50 x IO12/1); hence CMV is both a marker and a consequence of advanced immunosuppression. T h e trend towards increased survival of patients with CMV as measured from their diagnosis of A I D S , has been seen amongst our A I D S patients as a whole and is probably related to a reduced mortality from PCP, and the action of zidovudine. 27 In our study the median survival of patients with CMV colitis was less than in those with CMV retinitis. This observation has also been made by others 23 and may be due to greater systemic involvement with CMV in those patients with bowel disease, and the fact that a limited amount of C M V disease in such a critical site as the retina will be evident earlier than in other sites. T h e r e are several areas where further research might improve the management of CMV infection. At present primary prophylaxis is not commonly used. Ganciclovir and phosphonoformate would not be acceptable in this role because of their toxic effects and the need for regular venous access. T h e use of high dose acyclovir, despite its very weak activity against CMV, has reduced the likelihood of this infection developing in patients who have undergone renal transplantationfl 8' 29 An open study looking at the use of high dose oral acyclovir (8oo mg 5 times daily) in A I D S patients has suggested a considerable reduction in the incidence of CMV disease ;30 however, controlled studies are needed to evaluate the role of high dose acyclovir in both the treatment and prophylaxis of CMV infection in AID S. Granulocyte-macrophage colony-stimulating factor ( G M - C S F ) has been shown to improve the neutropenia associated with H I V infection and zidovudine use. TM 32 However, G M - C S F has also been shown to increase H I V replication in vitro. 33 Hence, granulocyte colony-stimulating factor (G-CSF), which does not have this effect, would be the more suitable trial agent to reverse neutropenia associated with ganciclovir therapy. Effective oral therapy for CMV disease would represent a considerable advance, both in terms of patient acceptability and in reducing demands on hospital resources. Although poorly absorbed, one study using oral ganciclovir 6 hourly, has shown that plasma levels sufficient to suppress viral replication were achieved. ~4 T h e role of oral ganciclovir in the management of CMV disease must await further study. Meanwhile, efforts should be made to expand the role of outpatient intravenous therapy; the successful use of outpatient induction therapy by one centre, 35 could be more widely adopted by others. Our study has demonstrated the considerable morbidity associated with
I36
B. S. P E T E R S E T A L .
C M V disease in a p o p u l a t i o n o f A I D S patients. I n o r d e r to r e d u c e this m o r b i d i t y in t h e f u t u r e it is i m p o r t a n t t h a t we give p r i o r i t y to a s t u d y o f t h e n a t u r a l h i s t o r y o f C M V disease a n d to r e s e a r c h into t h e d e v e l o p m e n t o f m o r e effective m a n a g e m e n t r e g i m e n s . References
I. Parkin JM, Thomson M, Pinching AJ. Preliminary results of a therapeutic trial of DHPG for cytomegalovirus infections in AIDS patients. In: Hemmer G, Staquet M, Baert A, Eds. Clinical aspects of AIDS. Oxford: Oxford University Press, x986: I95-2oi. 2. Weber JN, Thom S, Barrison I e t al. Cytomegalovirus colitis and oesophageal ulceration in the context of AIDS; clinical manifestations and preliminary report of treatment with foscarnet (phosphonoformate). Gut I987; z8: 482-487. 3. Holland GN, Pepose JS, Pettit TH et al. Acquired immune deficiency syndrome: ocular manifestations. Ophthalmology I983 ; 9o: 859-873. 4. Rosenberg PR, Uliss AE, Friedland GH et al. Acquired immunodeficiency syndrome: ocular manifestations in ambulatory patients. Ophthalmology I983 ; 9o: 874-878. 5. Schuman JS, Friedman AH. Retinal manifestations of the acquired immune deficiency syndrome (AIDS): cytomegalovirus, Candida albicans, cryptococcus, toxoplasmosis and Pneumocystis carinii. Trans Ophthalmol Soc UK I983; IO3: I77-I9O. 6. Khadem M, Kalish SB, Goldsmith JA et al. Ophthalmologic findings in acquired immune deficiency syndrome (AIDS). Arch Ophthalmol I984; IOZ: 2oi-2o6. 7. Freeman WR, Lerner CW, Mines JA et al. A prospective study of the ophthalmologic findings in the acquired immune deficiency syndrome. Am J OphthalmoI I984; 97: I33-I42. 8. Newsome DA, Green WR, Miller ED et al. Microvascular aspects of acquired immune syndrome retinopathy. Am J Ophthalmol I984; 98: 59o-6oi. 9- Palestine AG, Rodrigues MM, Macher AM et al. Ophthalmic involvement in acquired immunodeficiency syndrome. Ophthalmology I984; 9I: IO92-Io99. Io. Jabs DA, Green WR, Fox R, Polk F, Bartlett JG. Ocular manifestations of acquired immune deficiency syndrome. Ophthalmology I989; 96:Io92-Io99 . xI. Holland GN, Buhles WC, Mastre B, Kaplan HJ. A controlled retrospective study of ganciclovir treatment for cytomegalovirus retinopathy : use of a standardized system for the assessment of disease outcome. Arch Ophthalmol I989; *o7: I759-I766. I2. Jabs DA, Enger C, Bartlett JG. Cytomegalovirus retinitis and acquired immunodeficiency syndrome. Arch Ophthalmol I989; ~o7: 75-8o. I3. Collaborative DHPG Treatment Study Group. Treatment of serious cytomegalovirus infections with 9-(I,3-dihydroxy-2-propoxymethyl)guanine in patients with AIDS and other immunodeficiencies. N Engl J Med I986; 314: 8OI-805. I4. Masur H, Lane HC, Palestine A et al. Effect of 9-( I ,3-dihydroxy-2-propoxymethyl)guanine on serious cytomegalovirus disease in eight immunosuppressed homosexual men. Ann Intern Med I986; IO4: 41-44. I5. Palestine AG, Stevens G, Lane HC et al. Treatment of cytomegalovirus retinitis with dihydroxy propoxymethyl guanine. Am J Ophthalmol I986; IOI : 95-Ioi. I6. Rosecan LR, Stahl-Bayliss CM, Kalman CM, Laskin OL. Antiviral therapy for cytomegalovirus retinitis in AIDS with dihydroxy propoxymethyl guanine. Am J Ophthalmol I986; Iox: 4o5-418. I7. Erice A, Jordan MC, Chace BA, Fletcher C, Chinnock BJ, Balfour HH. Ganciclovir treatment of cytomegalovirus disease in transplant recipients and other immunocompromised hosts. J A M A I987; z57:3082-3087 • i8. Henderly DE, Freeman WR, Causey DM, Rao NA. Cytomegalovirus retinitis and response to therapy with ganciclovir. Ophthalmology I987; 94:425-434 • I9. Holland GN, Sidikaro Y, Kreiger AE et al. Treatment of cytomegalovirus retinopathy with ganciclovir. Ophthalmology 1987; 94: 815-823. 20. Jabs DA, Newman C, de Bustros S, Polk BF. Treatment of cytomegalovirus retinitis with ganciclovir. Ophthalmology I987; 94: 824-83 °. 21. Laskin OL, Stahl-Bayliss CM, Kalman CM, Rosecan LR. Use of ganciclovir to treat serious cytomegalovirus infections in patients with AIDS. J Infect Dis I987 ; I55 : 323-327 •
Cytomegalovirus infection in A I D S
I37
22. Orellana J, Teich SA, Freidman AH, Lerebours F, Winterkorn J, Mildvan D. Combined short- and long-term therapy for the treatment of cytomegalovirus retinitis using ganciclovir (BW B759U). Ophthalmology I987; 94: 831-838. 23. Jacobson MA, O'Donnell JJ, Porteous D, Brodie HR, Feigal D, Mills J. Retinal and gastrointestinal disease due to cytomegalovirus in patients with the acquired immune deficiency syndrome: prevalence, natural history, and response to ganciclovir therapy. Q 57 Med I988; 67: 473-486. 24. Buhles WC, Mastre BJ, Tinker AJ, Strand V, Koretz SH (The Syntex Collaborative Ganciclovir Study Group). Ganciclovir treatment of life or sight threatening cytomegalovirus infection: experience in 314 immunocompromised patients. Rev Infect Dis I988; xo (Suppl 3): $495-$5o6. 25. Jennens I, Stewart K, McLean H et al. Three times a week maintenance ganciclovir to prevent relapse of CMV retinitis in AIDS. 6th International Conference on AIDS, San Francisco, June I99O; (Abstr.) Th.B.433. 26. Richman DD, Fiscl MA, Grieco MH et al. The toxicity of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex. A double-blind, placebocontrolled trial. N Engl57 Med 1987; 317: I92-I97. 27. Peters BS, Beck EJ, Coleman DG et al. Changing disease patterns in people with AIDS in a referral centre in the United Kingdom: the changing face of AID S. Br Med J 1991 ; 302: 203-207 . 28. Meyers JD, Reed EC, Shepp DH et al. Acyclovir for prevention of cytomegalovirus infection and disease after allogenic marrow transplantation. N Engl J Med I988; 3x8: 7o-75. 29. Balfour HH, Beverly AC, Stapleton JT, Simmons RL, Fryd DS. A randomized, placebocontrolled trial of oral acyclovir for the prevention of cytomegalovirus disease in recipients of renal allografts. N Engl J Med I989; 3zo : I38I-I387. 3o. Metroka CE, Josefberg H. Usefulness of high dose acyclovir as prophylaxis for CMV. 6th International Conference on AIDS, San Francisco, June i99o; (Abstr.) 2I I I. 3 i. Mitsuyasu R, Levine J, Miles SA et al. Effects of long-term subcutaneous administration of recombinant granulocyte-macrophage colony-stimulating factor (GM-CSF) in patients with HIV-related leukopenia. Blood I988; 72 (Suppl i): 356a. 32. Levine JD, Allan JD, Tessitore JH, Falcone N, Israel R, Groopman JE. Granulocytemacrophage colony stimulating factor ameliorates the neutropenia induced by azidothymidine in AIDS/ARC patients. Proc A m Soc Clin Oncol I989; 8: I. 33. Koyanagi Y, O'Brien WA, Zhao JQ, Golde DW, Gasson JC, Chen ISY. Cytokines alter production of HIV-I from primary mononuclear phagocytes. Science I988; 24z: I673-I675. 34- Jacobson MA, deMiranda P, Cederberg DM et al. Human pharmacokinetics and tolerance of oral ganciclovir. Antimicrob Agents Chemother I987; 3I: I25I-I254. 35. Miller J, Dolan G, Fowles T, Schreurs E. Feasibility trial for outpatient induction with ganciclovir. 6th International Conference on AIDS, San Francisco, June I99O; (Abstr.) S.D.773.
C y t o m e g a l o v i r u s i n f e c t i o n in A I D S . P a t t e r n s o f disease, r e s p o n s e to t h e r a p y a n d t r e n d s in s u r v i v a l B. S. Peters,* E. J. Beck, t S. Anderson,* D. Coleman,* R. Coker, jJ J. Main,~ C. Migdal,~ J. R. W. Harris ]] and A. J. Pinching* * Department of Immunology, afAcademic Department of Public Health, Departments of ~ Infectious Diseases and Medicine, ~ Ophthalmology, and []Genito-urinary Medicine, St Mary's Hospital Medical School, Norfolk Place, London W2, U.K. Accepted for publication 25 March I991 Summary Among 347 AIDS patients seen at St Mary's Hospital, London between I983 and I989, cytomegalovirus (CMV) disease was observed in 75 (22 %). Of these, 58 (77 %) had CMV retinitis, 26 (35%) CMV colitis, and I2 (I6%) had CMV infection diagnosed at other sites. Relapse occurred in 7I %. A favourable response to the use of ganciclovir as induction therapy for CMV retinitis was observed in 92 %. Relapse of CMV retinitis occurred in 54 % at a median time of 97 days. Neutropenia was the most frequent and serious side-effect of ganciclovir, 76 % patients having neutrophil counts < I'O X I o g / l and 48% < 0'5 x IO9/1 at some stage of therapy. Thrombocytopenia was also common, and platelet counts of < 5o x IO9/1 occurred in 43 % patients on ganciclovir. The concurrent use of zidovudine made the development of severe neutropenia and thrombocytopenia more likely. Median survival following the diagnosis of CMV disease increased from 5-8 months between I984 and I987, to over I2 months in i988. Patients with CMV colitis had a worse prognosis than patients with CMV retinitis, with median survival of 4"5 and 7 months respectively. In conclusion, CMV is an important opportunist infection in AIDS and both the disease and its treatment cause considerable morbidity. Hence, it is important to develop more effective and less toxic forms of therapy for CMV infection.
Introduction Cytomegalovirus is one of the most i m p o r t a n t viral opportunist infections of A I D S patients. T h e most c o m m o n site of disease is the retina, followed by the large bowel, although other parts of the alimentary tract, the liver and biliary tree, the adrenal glands, the central nervous system and the lungs are among other sites that can be involved. Hence, unlike Pneumocystis carinii, which usually affects one target organ, the lungs, C M V is far more extensive and current treatments for it are only partly effective and m u s t be given intravenously. As the severity of P. carinii p n e u m o n i a (PCP) and its mortality have lessened in response to earlier diagnosis and improved m a n a g e m e n t , C M V has assumed increasing importance as an opportunist infection in A I D S . F o r this reason we decided to study its prevalence in our A I D S population, as well as the consequences of c o n t e m p o r a r y therapy, so that we can identify those aspects of C M V disease which merit f u r t h e r research. oI63-4453/9I/o5o~29 +o9 $03.00/0
© I99I The British Society for the Study of Infection
I30
B. S. P E T E R S E T A L .
Patients and m e t h o d s
T h e hospital records of all patients with A I D S managed at St Mary's Hospital, London between March I983 and December r989, were reviewed. Twelve patients whose case notes were unavailable were excluded from analysis, leaving a population of 347 under study. T h e majority (3Io) of our patients were homosexual men, 23 were bisexual men, and 14 were heterosexual, five of whom were women; seven were intravenous drug users. T h e racial composition was 33o white, Io Afro-Caribbean, four Hispanic and three Asian. T h e following information was recorded for all patients: age, sex, risk factors for H I V infection, date of A I D S diagnosis, index diagnosis and date of death. Patients with CMV retinitis or colitis were identified and the site and date of each episode of CMV disease was recorded. T h e ophthalmoscopic diagnosis of CMV retinitis was based on the findings of pale irregular areas of retinal necrosis with or without accompanying haemorrhages. Retinal examination was performed after mydriasis by both the physician and an ophthalmologist. CMV colitis was diagnosed by the presence of abdominal pain, usually accompanied by peritonism, with or without diarrhoea, and with evidence of CMV inclusion bodies on rectal biopsy. Subsequent episodes of CMV colitis could be diagnosed on clinical criteria alone, providing there was a rapid response to ganciclovir therapy. Diagnosis at other sites was made after histological examination of the appropriate biopsy material. Since I984 we have routinely used ganciclovir as first-line treatment and phosphonoformate as second line, both initially as part of clinical trials. 1' 2 Induction therapy was for 2 weeks, using ganciclovir Io m g / k g / d a y in two divided doses and for phosphoformate 18o m g / k g / d a y in three divided doses. We routinely provided maintenance therapy for all episodes of CMV retinitis, but only for colitis when relapsing. Maintenance therapy consisted of ganciclovir Io m g / k g / d a y Iv three times a week. If CMV infection recurred, the patient was reinduced, but we then used five divided doses weekly for maintenance. Maintenance phosphonoformate therapy was given as a 9o m g / k g / d a y infusion on 5-7 days each week. Ganciclovir was the firstchoice therapy but if there was a failure of initial response or neutrophil count less than o'5o × r&/1 despite stopping zidovudine, phosphonoformate treatment was substituted. Since its licensure in April I987 the majority of patients with A I D S have been prescribed zidovudine in doses of 100o--1200 mg/24 h. Most of those taking zidovudine, 22/28 (79%), were continued on it when ganciclovir was introduced. This contrasts with the practice in the U.S.A. where the two drugs are rarely given together. If the neutrophil count fell below 0-8 × lO9/1 the zidovudine dose was reduced to 5oo rag/24 h and blood counts taken weekly; zidovudine was stopped if the count fell to o'5 x lO9/1 or less. T h e lowest haemoglobin, neutrophil and platelet counts following the use of ganciclovir were recorded. Statistical analysis
Using the log-rank test, survival analyses were performed from the date of diagnosis of A I D S and the date of diagnosis of CMV disease; similarly,
Cytomegalovirus infection in A I D S
I3 1
2t 1.0.
L
/
0'4
0.2
I 0
I
I
I0
I 20
I
50
Months
Fig. I. Two-year survival from diagnosis of A I D S in patients who subsequently developed C M V infection. (D), 84; ( 0 ) , 85"; ([]), 86; (~}), 87; (m), 88.
survival was compared for those patients presenting with C M V colitis and C M V retinitis. Other statistical tests used included the Chi-squared (X2) test and the Fischer exact test, as indicated in the text. Nanostat software was used. Results Incidence and presentation
Cytomegalovirus disease was identified in 75 ( 2 2 ~/o) of our patients. Of these 75 patients 32 (43 %) had C M V retinitis alone, I4 (I9 %) had C M V colitis alone, and 13 (I7~o) had both. Twelve patients (z6 %) had I6 episodes of cytomegalovirus disease diagnosed at other sites, (six oesophageal, one gastric, one oropharyngeal, three pulmonary, four cerebral, one cardiac). T h e average annual n u m b e r of new patients diagnosed with C M V disease between I984 and I989 was I2 per year (range 8-2o). Average age at time of diagnosis was 36 years (range 22--57 ) and no significant changes occurred over time. Four percent (n = I5) of all patients presented with C M V disease as their A I D S defining diagnosis: 2 0 ~/o developed C M V disease within 3 m o n t h s of their A I D S diagnosis, 43 % within I2 m o n t h s and 8o % within 24 months. Overall 7I % (53) patients with C M V disease had recurrent symptomatic episodes: 5 o % of those with colitis had recurrence in the same site and, despite using maintenance therapy in all patients with C M V retinitis since I984, recurrent ocular disease occurred in 55 %. T h e median n u m b e r of days
I32
B. S. PETERS E T A L . 1'2
I'OJ
0.81
> 0"6
#,
0.4 • °
0.2
"
. . . . . . . . . . . . . . . . . . . .
, 0
.,°!
y!
:!___, 20
I0
30
Months
Fig. 2. Two-year survival from diagnosis of CMV infection to death in AIDS patients for various years. (V]), 84; (O), 85; ([]), 86; (~), 87; (m), 88.
to recurrence of C M V retinitis was 97 (range I5-396) compared with 58 days (range 21-216) for colitis. Retinal detachment occurred in I I / 5 8 ( I 9 % ) patients with C M V retinitis. Survival
T h e median survival of patients with C M V disease measured from the date of diagnosis of A I D S has increased from I 4 - I 6 m o n t h s for the I984-I986 cohorts to 24 m o n t h s for the I987 cohort, although this i m p r o v e m e n t did not reach statistical significance, probably because of insufficient n u m b e r s (Fig. I ; log-rank X, 2 = 4"6, P = o'327). Survival from the date of C M V diagnosis has not changed significantly, with a median survival ranging between 5 and I2 m o n t h s for the years I984-I987 (Fig. 2; log-rank X42= I.I, P = 0"89). However, a significant difference in survival was observed in those patients with C M V colitis compared with C M V retinitis; the median survival in the former group was 4"5 months, compared with 7 m o n t h s for those with C M V retinitis (Fig. 3; log-rank X2 = 8"0, P = 0"005). C o m p l i c a t i o n s o f therapy for CMV infection
Bone marrow toxicity was the most frequent complication of ganciclovir therapy (Table I), although nausea developed in three, skin rash in two and fevers in one patient. Severe neutropenia ( < 0"5 x io9/1) occurred in 31/65 (48 %) patients taking ganciclovir, and severe thrombocytopenia ( < 5o x Io9/1)
Cytomegalovirus infection in AIDS
I33
I'0
\
!
0-;
0"2
i !
L
! O.
. . . . . .
.
. . . . n
I
I 6
0
18
12
24
Mont"hs
Fig. 3. Two-year survival from diagnosis of CMV by site in AIDS patients. (--0--), Colitis; ( - - A - - ) , retinitis.
Table I
Haematological complications of ganciclovir therapy
Lowest value* Neutrophils ( x IO9/1) > I'5 I'O-I" 5 O'5--I'O
< o'5 Platelets ( x io9/1) > I5o IOO--I5O 5o-Ioo < 50 Haemoglobin (g/dl) > Io 7-Io < 7 Total
Number (%)
Number on zidovudine (%)
6 (9 %) II (I7%) 17 (26 %) 3I (48 %)
i 3 7 I7
(I7 %) (27%) (41%) (55 %)
9 18 Io 28
(I4%) (28 %) (I5 %) (43 %)
3 (33%) 7 (39 %) 5 (50 %) I3 (46 %)
6 22 37 65
(9 %) (34%) (57 %) (ioo %)
2 8 I8 28
(33 %) (36%) (49 %) (43 %)
* This refers to the lowest recorded value during ganciclovir therapy.
30
134
B. S. P E T E R S
ET AL.
in 28 (43 %); the concurrent use of zidovudine exacerbated these cytopenias (Table I). Frequent haematological monitoring, a reduction in dose or withdrawal of zidovudine and a switch to phosphonoformate when appropriate, meant that serious complications were usually avoided. However, septicaemia occurred in seven patients with neutropenia (six bacterial, one fungal) with one death. Serious haemorrhage occurred in three patients, with the bowel as the main site. All cases were successfully managed with platelet transfusion, stopping zidovudine, and withdrawal of ganciclovir in favour of phosphonoformate therapy. Of 28 patients on ganciclovir and zidovudine, zidovudine had to be stopped in 25 (89 %), with recovery of neutrophil and platelet counts in the majority of cases. Discussion
Cytomegalovirus is the most frequent opportunist infection to involve the eye in A I D S patients in the Western World. T h e prevalence of C M V retinitis, in I7 % of our patients, compares with figures of 6-40 % for series from U.S.A. centres, 3-1°. T h e differences may reflect recruitment biases, especially as several of the series with higher figures refer only to patients attending ophthalmology clinics. Apart from one early study using phosphonoformate, 2 ganciclovir has been our treatment of first choice for C M V disease because its longer serum halflife allows for shorter infusion regimens, and also because several studies demonstrating its efficacy have led to its licensure as first line therapy; hence we are unable to compare its efficacy with phosphonoformate. A favourable clinical response to induction therapy was recorded in 92 % of our patients. Comparable figures from other series vary from 5 7 - I o o %)1-22 This wide variation may reflect the different measures used to assess disease outcome, rather than real differences in therapeutic response. For example, the lowest figure (57 %) belonged to the centre using the widest definition of disease progression, u W i t h o u t maintenance therapy the median onset of clinical relapse has been reported to be between I9 and 47 day s.z~'~ T h e median onset, 97 days, of relapse of C M V retinitis in our patients compares favourably to figures of 54-1o5 days from other centres using similar total weekly dose maintenance therapy. 23'~4 This suggests that five times weekly ganciclovir maintenance, as routinely used by other centres, has no advantage over our three times weekly regimen. Another group has reported similar results using thrice weekly maintenance, with progression to C M V retinitis occurring in 40 % subjects at a mean 123 days. 25 T h e treatment of C M V retinitis has important consequences, both in terms of morbidity and the use of resources. T h e diagnosis of C M V retinitis results in the hospitalisation of the patient for induction therapy and insertion of an indwelling venous catheter (Hickman line or Portacath). Maintenance therapy with ganciclovir is expensive, requires more frequent blood counts, and our patients receive regular monitoring by ophthalmologists. T h e most important toxic effect of ganciclovir is myelotoxicity, particularly neutropenia. Zidovudine also causes bone marrow suppression, 26 and most of
Cytomegalovirus infection in A I D S
r35
our patients had zidovudine withdrawn because of severe neutropenia when used with ganciclovir. One particular consequence of neutropenia in our patients has been an increased rate of infection of indwelling venous catheters, especially with staphylococcal species. T h e median survival of our patients following their diagnosis of CMV for the years I984-I988 ranged between 4 and 9 months, with no significant change over time, and compares with figures from other series of 5"5 and 6 months following a diagnosis of CMV retinitis. 12'18 Although the onset of CMV disease is considered a poor prognostic factor this may partly be because CMV infection tends to occur when CD4 counts drop to very low levels ( < 50 x IO12/1); hence CMV is both a marker and a consequence of advanced immunosuppression. T h e trend towards increased survival of patients with CMV as measured from their diagnosis of A I D S , has been seen amongst our A I D S patients as a whole and is probably related to a reduced mortality from PCP, and the action of zidovudine. 27 In our study the median survival of patients with CMV colitis was less than in those with CMV retinitis. This observation has also been made by others 23 and may be due to greater systemic involvement with CMV in those patients with bowel disease, and the fact that a limited amount of C M V disease in such a critical site as the retina will be evident earlier than in other sites. T h e r e are several areas where further research might improve the management of CMV infection. At present primary prophylaxis is not commonly used. Ganciclovir and phosphonoformate would not be acceptable in this role because of their toxic effects and the need for regular venous access. T h e use of high dose acyclovir, despite its very weak activity against CMV, has reduced the likelihood of this infection developing in patients who have undergone renal transplantationfl 8' 29 An open study looking at the use of high dose oral acyclovir (8oo mg 5 times daily) in A I D S patients has suggested a considerable reduction in the incidence of CMV disease ;30 however, controlled studies are needed to evaluate the role of high dose acyclovir in both the treatment and prophylaxis of CMV infection in AID S. Granulocyte-macrophage colony-stimulating factor ( G M - C S F ) has been shown to improve the neutropenia associated with H I V infection and zidovudine use. TM 32 However, G M - C S F has also been shown to increase H I V replication in vitro. 33 Hence, granulocyte colony-stimulating factor (G-CSF), which does not have this effect, would be the more suitable trial agent to reverse neutropenia associated with ganciclovir therapy. Effective oral therapy for CMV disease would represent a considerable advance, both in terms of patient acceptability and in reducing demands on hospital resources. Although poorly absorbed, one study using oral ganciclovir 6 hourly, has shown that plasma levels sufficient to suppress viral replication were achieved. ~4 T h e role of oral ganciclovir in the management of CMV disease must await further study. Meanwhile, efforts should be made to expand the role of outpatient intravenous therapy; the successful use of outpatient induction therapy by one centre, 35 could be more widely adopted by others. Our study has demonstrated the considerable morbidity associated with
I36
B. S. P E T E R S E T A L .
C M V disease in a p o p u l a t i o n o f A I D S patients. I n o r d e r to r e d u c e this m o r b i d i t y in t h e f u t u r e it is i m p o r t a n t t h a t we give p r i o r i t y to a s t u d y o f t h e n a t u r a l h i s t o r y o f C M V disease a n d to r e s e a r c h into t h e d e v e l o p m e n t o f m o r e effective m a n a g e m e n t r e g i m e n s . References
I. Parkin JM, Thomson M, Pinching AJ. Preliminary results of a therapeutic trial of DHPG for cytomegalovirus infections in AIDS patients. In: Hemmer G, Staquet M, Baert A, Eds. Clinical aspects of AIDS. Oxford: Oxford University Press, x986: I95-2oi. 2. Weber JN, Thom S, Barrison I e t al. Cytomegalovirus colitis and oesophageal ulceration in the context of AIDS; clinical manifestations and preliminary report of treatment with foscarnet (phosphonoformate). Gut I987; z8: 482-487. 3. Holland GN, Pepose JS, Pettit TH et al. Acquired immune deficiency syndrome: ocular manifestations. Ophthalmology I983 ; 9o: 859-873. 4. Rosenberg PR, Uliss AE, Friedland GH et al. Acquired immunodeficiency syndrome: ocular manifestations in ambulatory patients. Ophthalmology I983 ; 9o: 874-878. 5. Schuman JS, Friedman AH. Retinal manifestations of the acquired immune deficiency syndrome (AIDS): cytomegalovirus, Candida albicans, cryptococcus, toxoplasmosis and Pneumocystis carinii. Trans Ophthalmol Soc UK I983; IO3: I77-I9O. 6. Khadem M, Kalish SB, Goldsmith JA et al. Ophthalmologic findings in acquired immune deficiency syndrome (AIDS). Arch Ophthalmol I984; IOZ: 2oi-2o6. 7. Freeman WR, Lerner CW, Mines JA et al. A prospective study of the ophthalmologic findings in the acquired immune deficiency syndrome. Am J OphthalmoI I984; 97: I33-I42. 8. Newsome DA, Green WR, Miller ED et al. Microvascular aspects of acquired immune syndrome retinopathy. Am J Ophthalmol I984; 98: 59o-6oi. 9- Palestine AG, Rodrigues MM, Macher AM et al. Ophthalmic involvement in acquired immunodeficiency syndrome. Ophthalmology I984; 9I: IO92-Io99. Io. Jabs DA, Green WR, Fox R, Polk F, Bartlett JG. Ocular manifestations of acquired immune deficiency syndrome. Ophthalmology I989; 96:Io92-Io99 . xI. Holland GN, Buhles WC, Mastre B, Kaplan HJ. A controlled retrospective study of ganciclovir treatment for cytomegalovirus retinopathy : use of a standardized system for the assessment of disease outcome. Arch Ophthalmol I989; *o7: I759-I766. I2. Jabs DA, Enger C, Bartlett JG. Cytomegalovirus retinitis and acquired immunodeficiency syndrome. Arch Ophthalmol I989; ~o7: 75-8o. I3. Collaborative DHPG Treatment Study Group. Treatment of serious cytomegalovirus infections with 9-(I,3-dihydroxy-2-propoxymethyl)guanine in patients with AIDS and other immunodeficiencies. N Engl J Med I986; 314: 8OI-805. I4. Masur H, Lane HC, Palestine A et al. Effect of 9-( I ,3-dihydroxy-2-propoxymethyl)guanine on serious cytomegalovirus disease in eight immunosuppressed homosexual men. Ann Intern Med I986; IO4: 41-44. I5. Palestine AG, Stevens G, Lane HC et al. Treatment of cytomegalovirus retinitis with dihydroxy propoxymethyl guanine. Am J Ophthalmol I986; IOI : 95-Ioi. I6. Rosecan LR, Stahl-Bayliss CM, Kalman CM, Laskin OL. Antiviral therapy for cytomegalovirus retinitis in AIDS with dihydroxy propoxymethyl guanine. Am J Ophthalmol I986; Iox: 4o5-418. I7. Erice A, Jordan MC, Chace BA, Fletcher C, Chinnock BJ, Balfour HH. Ganciclovir treatment of cytomegalovirus disease in transplant recipients and other immunocompromised hosts. J A M A I987; z57:3082-3087 • i8. Henderly DE, Freeman WR, Causey DM, Rao NA. Cytomegalovirus retinitis and response to therapy with ganciclovir. Ophthalmology I987; 94:425-434 • I9. Holland GN, Sidikaro Y, Kreiger AE et al. Treatment of cytomegalovirus retinopathy with ganciclovir. Ophthalmology 1987; 94: 815-823. 20. Jabs DA, Newman C, de Bustros S, Polk BF. Treatment of cytomegalovirus retinitis with ganciclovir. Ophthalmology I987; 94: 824-83 °. 21. Laskin OL, Stahl-Bayliss CM, Kalman CM, Rosecan LR. Use of ganciclovir to treat serious cytomegalovirus infections in patients with AIDS. J Infect Dis I987 ; I55 : 323-327 •
Cytomegalovirus infection in A I D S
I37
22. Orellana J, Teich SA, Freidman AH, Lerebours F, Winterkorn J, Mildvan D. Combined short- and long-term therapy for the treatment of cytomegalovirus retinitis using ganciclovir (BW B759U). Ophthalmology I987; 94: 831-838. 23. Jacobson MA, O'Donnell JJ, Porteous D, Brodie HR, Feigal D, Mills J. Retinal and gastrointestinal disease due to cytomegalovirus in patients with the acquired immune deficiency syndrome: prevalence, natural history, and response to ganciclovir therapy. Q 57 Med I988; 67: 473-486. 24. Buhles WC, Mastre BJ, Tinker AJ, Strand V, Koretz SH (The Syntex Collaborative Ganciclovir Study Group). Ganciclovir treatment of life or sight threatening cytomegalovirus infection: experience in 314 immunocompromised patients. Rev Infect Dis I988; xo (Suppl 3): $495-$5o6. 25. Jennens I, Stewart K, McLean H et al. Three times a week maintenance ganciclovir to prevent relapse of CMV retinitis in AIDS. 6th International Conference on AIDS, San Francisco, June I99O; (Abstr.) Th.B.433. 26. Richman DD, Fiscl MA, Grieco MH et al. The toxicity of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex. A double-blind, placebocontrolled trial. N Engl57 Med 1987; 317: I92-I97. 27. Peters BS, Beck EJ, Coleman DG et al. Changing disease patterns in people with AIDS in a referral centre in the United Kingdom: the changing face of AID S. Br Med J 1991 ; 302: 203-207 . 28. Meyers JD, Reed EC, Shepp DH et al. Acyclovir for prevention of cytomegalovirus infection and disease after allogenic marrow transplantation. N Engl J Med I988; 3x8: 7o-75. 29. Balfour HH, Beverly AC, Stapleton JT, Simmons RL, Fryd DS. A randomized, placebocontrolled trial of oral acyclovir for the prevention of cytomegalovirus disease in recipients of renal allografts. N Engl J Med I989; 3zo : I38I-I387. 3o. Metroka CE, Josefberg H. Usefulness of high dose acyclovir as prophylaxis for CMV. 6th International Conference on AIDS, San Francisco, June i99o; (Abstr.) 2I I I. 3 i. Mitsuyasu R, Levine J, Miles SA et al. Effects of long-term subcutaneous administration of recombinant granulocyte-macrophage colony-stimulating factor (GM-CSF) in patients with HIV-related leukopenia. Blood I988; 72 (Suppl i): 356a. 32. Levine JD, Allan JD, Tessitore JH, Falcone N, Israel R, Groopman JE. Granulocytemacrophage colony stimulating factor ameliorates the neutropenia induced by azidothymidine in AIDS/ARC patients. Proc A m Soc Clin Oncol I989; 8: I. 33. Koyanagi Y, O'Brien WA, Zhao JQ, Golde DW, Gasson JC, Chen ISY. Cytokines alter production of HIV-I from primary mononuclear phagocytes. Science I988; 24z: I673-I675. 34- Jacobson MA, deMiranda P, Cederberg DM et al. Human pharmacokinetics and tolerance of oral ganciclovir. Antimicrob Agents Chemother I987; 3I: I25I-I254. 35. Miller J, Dolan G, Fowles T, Schreurs E. Feasibility trial for outpatient induction with ganciclovir. 6th International Conference on AIDS, San Francisco, June I99O; (Abstr.) S.D.773.