Neuro-radiology

Neuroradiology (1992) 35:75-78

9 Springer-Verlag 1992

CT and MRI: prognostic tools in patients with AIDS and neurological deficits A . Mundinger 1, T. A d a m 2, D. Ott 2, E. Dinkel 1, A. B e c k 1, H. H. Peter 3, B. Volk 4, and M. Schumacher 2

Departments of 1Diagnostic Radiology, 2Neuroradiology, 3Rheumatology and Clinical Immunology, 4Neuropathology, University of Freiburg, Freiburg, Federal Republic of Germany Received: 28 September 1991

Summary. To determine the prognostic value of CT and M R I in A I D S we studied the survival of patients with neurological involvement, in relation to the initial imaging results. Twenty-six initial CT and 15 M R I examinations of 41 patients were reviewed for the presence of cerebral atrophy and/or focal lesions. The m e a n survival time of patients with initially normal imaging was longer (700 _+89 days) than that of patients with isolated cerebral atrophy (326 + 65) or isolated focal lesions (202 _+97). The shortest survival (78 + 44 days) was found in patients with both cerebral atrophy and focal lesions. The risk of death in patients with atrophy alone was 3.6 times higher, that in patients with focal lesions alone 6.4 times higher, and in patients with both changes 19.3 times higher than in patients with initially normal imaging. Cerebral imaging with CT and/or M R I thus allows identification of A I D S related cerebral changes and m a y contribute to assessment of prognosis.

K e y words: A I D S - Encephalopathy - C o m p u t e d tomogr a p h y - Magnetic resonance imaging

The central nervous system is frequently involved in h u m a n immunodeficiency virus ( H I V ) infection [1, 2]. In patients with A I D S neurological symptoms and signs can develop due to opportunistic infection or malignancy as well as to the neuropathic effects of H I V [3-6]. Patients with A I D S who display any neurological symptoms or signs are reported to have a reduced probability of survival [7]. The purpose of our study was to determine whether CT and/or M R I can be of m o r e prognostic value than the clinical status in patients with A I D S and neurological symptoms or signs.

Materials and m e t h o d s

Survival times were related to the initial CT and/or MRI findings in 41 patients with proven HIV infection fulfilling the CDC criteria for AIDS and neurological symptoms or signs. Twenty-six patients were

first examined by CT, 15 by MRI. CT studies were performed after intravenous injection of contrast media. MR images were obtained on a 0.23T resistive magnet or a 2T superconductive magnet system using a 256 x 256 pixel matrix. Axial Tl-weighted (500/40) and proton density- and T2-weighted (3000/30 . . . . 120) spin-echo images using a Carr-Purcell-Meiboom-Gill multislice-multiecho sequence were routinely acquired. Additional coronal or sagittal images and gradient echo or RARE (rapid acquisition with relaxation enhancement) images were obtained according to the pathological findings. Three experienced neuroradiologists operating without clinical data except for sex and age defined the presence or absence of cerebral atrophy and diffuse or focal lesions. Considerable loss of grey and white matter, and widening of cerebrospinal fluid spaces, as compared to age-related normals, were the criteria used to define cerebral atrophy. Dehydration or obstructive hydrocephalus as possible alternative causes of widened cerebral ventricles were excluded in our study before the diagnosis of cerebral atrophy was accepted. Kaplan-Meier plots were used to illustrate the survival times of different groups of patients depending on the results of the initial imaging examination: normal, atrophy only, focal lesion(s) only, both atrophy and focal lesion(s). A stepwise Cox regression model was used to determine the relative hazard associated with these imaging results. As usual in survival analysis, 2 patients lost during follow-up were included with the date of the last observation.

Results

The survival times of patients with A I D S and neurological symptoms or signs were significantly different depending on the initial CT and/or M R I findings. The m e a n survival time of patients with initially normal findings was 700.5 _+89.2 days, patients with isolated cerebral atrophy 326.5 _+64.9 days, and patients with isolated focal lesions 202.5 + 97.4 days. The shortest survival time (77.6 _+ 42.7 days) was in patients with both cerebral atrophy and focal lesions (Fig. 1). The most important factor for a p o o r prognosis was any pathological finding on CT or MRI. C o m p a r e d to patients with normal imaging results, patients with atrophy alone had a relative hazard of x 3.6, those with only focal lesions x 6.4, and those with both atrophy and focal lesions x 19.3.

76 proportion of survival

of their H I V infection, compared with a mean of 0.58 episodes per survivor. Patients with atrophy alone and atrophy combined with focal lesions at initial presentation showed the highest frequency of severe pulmonary complications, indicating a more advanced stage of immunodeficiency (Table 1). The most frequent cerebral findings at autopsy were toxoplasmosis and H I V encephalopathy (Table 2). In 7 of 18 patients submitted to necropsy, cerebral disease was the direct cause of death. The last imaging study before death was abnormalin all but 2 of the 18: atrophy and focal lesions were seen in 10, focal lesions in 4, and atrophy in only 2.

i

O,S

i

0,6

[ I rlor rnal

I

0,4

t..... 0,2

I .... 7

1

1

0 0

a_. . . . .

;

atrophy

focal

*focapl

=

=

400

600

200

]

I

atrophy i

1000

800

1200

days of survival

Fig.1. The survival times of patients with AIDS and neurological symptoms or signs were significantly different depending on the initial findings on CT and/or MRI, as illustrated by Kaplan-Meier plots: normal - atrophy- focal - atrophy + focal Table 1. Frequency of severe pulmonary complications in patients

with AIDS within the observation time related to the initial neuroradiological findings

Total number of patients Living Deceased Absolute frequency of severe pulmonary complications Relative frequency of severe pulmonary complications

Initial imaging results Normal Atrophy Focal lesion 17 12 4 10 2 0 7 10 4

Atrophy + focal lesion 6 0 6

13

5

6

1.25

1.67

0.76

23

1.92

Table 2. Cerebral lesions at autopsy in 18 patients related to last

antemortem imaging results Autopsy Diagnosisa Toxoplasmosis HIV encephalopathy PDL PML CMV encephalitis Bacterial abscess Malignant lymphoma

Last imaging results before autopsy Normal Atrophy Focal Atrophy + lesion focal lesion 2 5 2 2 1b 2 2 3 1 1 2

Several diagnoses per patient possible uAdditional toxoplasmosis HIV, Human immunodeficiency virus; PDL, progressive diffuse leucoencephalopathy; PML, progressive multifocal leucoencephalopathy; CMV, cytomegalovirus Due to the limited number of patients the relative risk of additional covariables, such as the meningeal changes seen in 3 patients, or of infectious and neoplastic complications could not be introduced into the statistical model. However, the 27 patients who died had a mean of 1.59 severe episodes of pulmonary complications in the course

Discussion

Prognostic predictions are useful in the care of fatally ill patients. In full-blown AIDS the impact on prognosis of demographic variables, immunological dysfunction, physiological deficits due to H I V infection, or associated infection and malignancy has been emphasized [7-11]. Neurological deficits alone have been demonstrated to be a predictor of poor outcome [12]. In our study we concentrated on the impact of the findings on modern cross-sectional cerebral imaging on the prognosis of patients with AIDS and diffuse and/or focal neurological symptoms or signs. We found statistically significant different survival times depending on the results of CT and/or MRI: the single strongest adverse effect on survival was any pathological finding on CT or MRI. The risk of death during the follow-up period was 19.3 times as high in patients with a combined pathology (atrophy and focal lesions) as in patients with normal initial imaging. In addition, the frequency of severe pulmonary complications was highest in patients showing both atrophy and focal lesions and lowest in patients with normal imaging, suggesting more pronounced immunological dysfunction in the former. We postulate that the natural course of AIDS leads to a stepwise progression of cerebral changes from normal to either atrophy or focal changes and finally to a combination in both. The survival times in our series are within the ranges reported in other studies [7, 13, 14]. Justice et al. reported that the median survival of 31 patients with any neurological manifestation (2.1 months) within a cohort of 117 patients with AIDS was worse than that of the cohort as a whole (5 months) [7]. A median time of 4 months from onset of neurological symptoms or signs to death was compared with a median time of 9 months from diagnosis of AIDS to death in 122 patients in central London. In that study, however, AIDS dementia, CNS opportunistic infections and other neurological deficits were associated with only minor changes in the survival time [14]. Cerebral atrophy, alone or in combination with focal lesions, was our most frequent finding, as in other studies [1, 2, 15-18]. Patients with cerebral atrophy appear to be at higher risk to develop focal lesions than patients with initially normal imaging. Levy et al. [1] observed new focal lesions in 16 % of 75 patients who initially showed atrophy and in 5 % of 81 patients whose initial CT was normal.

77

Fig.2. Proton density-weighted (3000/30) axial MRI demonstrating a hyperintense lesion centered on the left thalamus due to toxoplasmosis. A second lesion is seen in the medial left frontal lobe Fig.3. Gadolinium-DTPA-enhanced axial Tl-weighted (400/30) MRI image of a left hemisphere toxoplasmosis lesion showing central and peripheral hyperintensity, with surrounding oedema and mass effect Although the pathophysiology is not completely understood, cerebral atrophy is probably due to HIV encephalitis and to associated neurotoxic factors; however, other pathogens including cytomegalovirus (CMV) and toxoplasma, are frequently implicated [4, 19]. Microglial nodules with multinucleated giant,cells are reported as the hallmark of HIV encephalitis, alone or in combination with CMV [16, 20-22]. CT or MRI are insensitive in the early stages of H I V encephalitis due to the microscopic level of pathological changes [16, 23]. However, Post et al. [23] found secondary atrophy in 20 of 21 autopsies of patients with pathologically proven HIV encephalitis; cortical atrophy was a useful indicator of cortical involvement by the H I V infection. HIV-induced loci of demyelination are found in a minority of patients with pathologically proven H I V encephalitis [16, 20]. Both focal and diffuse, large, confluent white matter lesions are part of the spectrum of HIV-induced encephalitis, the latter defining the picture of progressive diffuse leucoencephalopathy (Fig.4) [24]. These conditions have to be distinguished from progressive multifocal leucoencephalopathy (PML), caused by the Jakob Creuzfeldt papovavirus [25, 26]. PML in AIDS leads to a more focal pattern of demyelination of the white matter and, less often, to lesions of the grey matter [24, 25]. Focal cerebral lesions in unselected autopsy series of patients with AIDS are usually due to other opportunistic infections and malignancies [17, 18, 27]. Many can be detected ante-mortem by modern imaging modalities [1, 3, 6, 15, 24, 25]. Unenhanced MRI is more sensitive to focal cerebral changes than contrast-enhanced CT (Fig.2). Intravenous gadolinium-DTPA aids characterization of these lesions and improves the specificity of MRI, so that CT may not be required [23] (Fig.3). One limitation of our study is that both CT and MRI were used at the initial presentation of the neurologically

Fig.4a, b. T2-weighted (3000/120) axial MRI showing diffuse hyperintensity of the frontal and temporal white matter, more extensive on the right, involvingsubcortical and cortical layers, due to progressive diffuse leucoencephalopathy

affected patients: several patients who on CT showed only atrophy did have focal lesions on MRI. However, comparison of initial and follow-up examinations of 19 patients examined by CT and MRI at the same time showed that only 1 patient with normal findings and I patient with isolated atrophy on CT had focal lesions in MRI requiring reclassification. Five patients with focal lesions on CT had more lesions on MRI, but stayed within the same category. To assess cerebral atrophy by comparison with agerelated standard images is less sensitive but more specific than quantification of atrophy based on linear, planimetric or volumetric measures and in accordance to common practice in neuroradiology. Quantitative measurements, however, have the advantage of detecting minor changes in the cerebrospinal fluid spaces of a given patient more accurately and may be preferred in longitudinal studies [28-31].

Acknowle@rnent. We are indebted to Professor Schulte-M/Snting, Department of Medical Biometry and Statistics,University of Freiburg.

References

1. Levy RM, Rosenbloom S, Perret LV (1986) Neuroradiologic findingsin AIDS: a review of 200 cases. AJR 147:977-983 2. Stavrou D, Mehraein R Mellert Wet al (1989) Evaluation of intracerebral lesions in patients with acquired immunodeficiency syndrome. Neuropathological findings and experimental data. Neuropathol Appl Neurobio115: 207-222 3. Cirillo SF, Rosenblum ML (1990) Use of CT and MR imaging to distinguish intracranial lesions and to define the need for biopsy in AIDS patients. J Neurosurg 73:720-724 4. Federle IMP(1988) A radiologist looks at AIDS: imaging evaluation based on symptom complexes. Radiology 166:553-562 5. Ho DD, Rota TR, Schooley RT, Kaplan JC, Allan JD, Groopman JE, Resnick L, Felsenstein D, Andrews CA, Hirsch MS

78 (1985) Isolation of HTLV-III from cerebrospinal fluid and neural tissues of patients with neurologic syndromes related to the acquired immunodeficiency syndrome. N Engl J Med 313: 14931497 6. Jarvik JG, Hesselink JR, Kennedy C, Teschke R, Wiley C, Spectot S, Richman D, MeCutchan A (1988) Acquired immunodeficiency syndrome. Magnetic resonance patterns of brain involvement with pathologic correlation. Arch Neuro145: 731-736 7. Justice AC, Feinstein AR, Wells CK (1989) A new prognostic staging system for the acquired immunodeficiency syndrome. N Engl J Med 320:1388-1393 8. Red field RR, Wright DC, Tramont EC (1986) The Walter Reed staging classification for HTLV-III/LAV infection. N Engl J Med 314:131-132 9. Fahey JL, Taylor JM, Detels R, Hofman B, Melmed R, Nishanian R Giorgi JV (1990) The prognostic value of cellular and serologic markers in infection with human immunodeficiency virus type 1. N Engl J Med 322:166-172 10. Rothenberg R, Woelfel M, Stoneburner R, Milberg J, Parker R, Trumen B (1987) Survival with the acquired immune deficiency syndrome: experience with 5833 cases in New York City. N Engl J Med 317:1297-1302 11. Marasca J, McEvoy M (1986) Lenght of survival of patients with acquired immune deficiency syndrome in the United Kingdom. BMJ 292:1717-1729 12. Jakobsen J, Gyldensted C, Bruhn R Helweg-Larsen S, ArlienSborg P (1989) Cerebral ventricular enlargement relates to neuropsychological measures in unselected AIDS patients. Acta Neurol Scand 79:59-62 13. Enzensberger W (1989) Neuro-manifestations of AIDS. Followup examinations in 215 patients with HIV infection. Schwehr, Stuttgart, pp 55-60 14. Guiloff RJ, Fuller GN, Roberts A, Hargreaves M, Gazzard B, Scaravilli E Harcourt-Webster JN (1988) Nature, incidence and prognosis of neurological involvement in the acquired immunodeficiency syndrome in central London. Postgrad Med J 64: 919925 15. Tosch U, Iglesias-Rozas JR, RufB, Witt H (1990) Histopathologisch verifizierte cerebrale CT-Befunde bei AIDS. Fortschr R6ntgenstr 152:196-199 16. Chrysikopoulos HS, Press GA, Grafe MR, Hesselink JR, Clayton AW (1990) Encephalitis caused by human immunodeficiency virus: CT and MR imaging manifestations with clinical and pathologic correlation. Radiology 175:185-191 17. Lang W, Miclossy J, D eruaz JR Pizzolato GP, Probst A, Schaffner T, Gessaga E, Kleihues P (1989) Neuropathology of the acquired immune deficiency syndrome (AIDS): a report of 135 consecutive autopsy cases from Switzerland. Acta Neuropatho177: 379390

18. Nielson SL, Petito CK, Urmacher CD, Posner JB (1984) Subacute encephalitis in acquired immune deficiency syndrome: a postmortem study. Am J Clin Patho182:678-682 19. Flowers CH, Mafee MF, Crowell R, Raofi B, Arnold B, Dobben G, Wycliffe N (1990) Encephalopathy in AIDS patients: evaluation with MR imaging. AJNR 11:1235-1245 20. Lantos PL, Scholtz CL, Mc Laughlin JE, Berry CL, Tighe JR (1989) Neuropathology of the brain in HIV infection. Lancet I: 309-311 21. Petito CK, Cho ES, Lemann W, Navia BA, Price RW (1986) Neuropathology of acquired immunodeficiency syndrome (AIDS): an autopsy review. J Neuropathol Exp Neuro145: 635646 22. Navia BA, Cho ES, Petito CK, Price RW (1986) The AIDS dementia complex II. Neuropathology 19:525-535 23. Post MJD, Tate LG, Quencer RM, Hensley GT, Berger JR, Sheremata WA, Maul G (1988) CT, MR, and pathology in HIV encephalitis and meningitis. AJNR 151:373-380 24. Olsen WL, Longo FM, Mills CM, Norman D (1988) White matter disease in AIDS: findings at MR imaging. Radiology 169: 445-448 25. Padgett BL, Walker DL, Zu Rhein GM, Eckroade RJ (1971) Cultivation of papova-like virus from human brain with progressive multifocal leukoencephalopathy. Lancet I: 1257-1260 26. Mark AS, Atlas SW (1989) Progressive multifocal leukoencephalopathy in patients with AIDS: appearance on MR images. Radiology 73:51%520 27. Center of disease control (CDC) (1987) Leads from the MMWR, supplement IS, revision of CDC-surveillance case definition for acquired immunodeficiency syndrome. JAMA 258:1143-1154 28. Schroth G, Remmes U, Schupmann A (1985) Computertomographische Verlaufsuntersuchung yon Hirnvolumenschwankungen vor und nach Alkoholentzug. Fortschr R/3ntgenstr 142: 361369 29. Pascual-Leone A, Dhuna A, Anderson DC (1991) Cerebral atrophy in habitual cocaine abusers: a planimetric CT study. Neurology 41:34-38 30. Malko LA, Hofman JC, Green RC (1991) MR measurement of intracranial CSF volume in 41 elderly normal volunteers. AJNR 12:371-374 31. Kitani M, Kobayashi S, Yamaguchi S (1990) Computerized tomography with longitudinal follow-up of brain atrophy in patients with Parkinson's disease. Gerontology 36:361-368 Dr. A. Mundinger Abteilung R6ntgendiagnostik der Radiologischen Universit~tsklinik Freiburg Hugstetterstrasse 55 W-7800 Freiburg, Federal Republic of Germany