| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
|
INTRODUCTION |
Malarial retinopathy is a recently described set of retinal signs, which can be accompanied by papilledema. Some of the retinal signs are unique to malaria and their detection is a candidate diagnostic test for severe malaria.
The retina is embryologically part of the central nervous system with an analogous cellular structure and blood-tissue barrier. Therefore, examination of the retina provides a unique opportunity to study an infected microvasculature and its effects on neurologic tissue in vivo. In this report, we review the features of malarial retinopathy, how it can be best identified, and its use in the diagnosis of cerebral malaria. We also summarize research on the pathophysiology of retinal signs and the insights gained into the pathogenesis of cerebral malaria.
|
METHODS |
|
FEATURES OF MALARIAL RETINOPATHY |
Malarial retinopathy consists of four main components: retinal whitening, vessel changes, retinal hemorrhages, and papilledema (Table 1
). The first two of these abnormalities are specific to malaria, and are not seen in other ocular or systemic conditions.
|
and 2).11 It frequently extends temporally between the vascular arcades (Figure 2).
|
|
), or it can be in a more mosaic pattern in the peripheral fundus. The appearance is distinct from cotton wool spots, which also occur, but less frequently. Patches of peripheral whitening associated with severe malaria are less well demarcated, less brightly white, and more widely distributed than cotton wool spots. Retinal whitening is similar in appearance to patchy ischemic retinal whitening, an uncommon finding in central retinal vein occlusion,14 but has a different retinal distribution.
Vessel changes.
Vessel changes are manifest as discoloration of retinal vessels to orange or white, mainly in the peripheral fundus (Figure 3). Either discrete sections of vessels, or peripheral trees, can be involved. White or orange tramlining within larger vessels can occur (continuous or interrupted), delineating an apparently narrowed blood column (Figure 4). Capillary whitening refers to whitening of retinal capillaries and post-capillary venules such that they become prominent against the choroidal background. When viewed with a high magnification indirect lens (x 5.5), capillary whitening appears to be more widely distributed in the fundus than has been previously recognized.
|
|
). In severe cases, these can be extremely numerous (>120 in each eye) and overlapping (Figure 5). Flame and large blot hemorrhages also occur frequently. Occasionally, hemorrhages can extend into the pre-retinal space.
|
Prognostic significance and outcome.
A large prospective study of Malawian children with cerebral malaria has found that the severity of retinal signs, including the number of retinal hemorrhages, is related to fatal outcome and length of coma in survivors.12 The relative risk of death conferred by the presence of each fundus sign is shown in Table 2. This concurred with earlier studies that found that ocular fundus signs were associated with an increased risk of death.8,9 In a smaller number of Kenyan children with cerebral malaria, retinal hemorrhages were associated with deep coma and severe anemia.10 In children in Mali with malaria, retinopathy was related to severity, but this study is difficult to compare with other studies in African children because it does not state the method of ophthalmoscopy used and uses a more inclusive definition of cerebral malaria.15
|
|
DETECTION OF MALARIAL RETINOPATHY |
). Pupil dilation can be achieved with tropicamide 0.5% or 1% eye drops, with the addition of 2.5% phenylephrine if required. The wider field of view afforded by the indirect ophthalmoscope allows visualization of the peripheral retina where the unique vascular changes are mostly situated, beyond the view of the direct ophthalmoscope. This perhaps explains how they had evaded attention.
|
Pharmacologic dilatation of the pupils of patients with acute neurologic illness may be a concern to physicians who are then temporarily unable to monitor pupil size and reaction. The pupil reactions should be carefully examined and recorded prior to instilling dilating drops. Clinicians should be made aware that drops have been given, and the time of administration recorded. The effect of tropicamide wears off after approximately two hours. Since the severity of malarial retinopathy is usually similar in both eyes, most information can be gained by dilating one pupil.
Evidence for the utility of examining pupillary responses comes mostly from head injury and intracranial hemorrhage.22–26 In contrast, cerebral malaria is a diffuse cerebral disease in which the role of tentorial herniation is as yet unreported. Classic studies suggest that pupil responses are less helpful in diffuse parenchymal lesions (which may cause central transtentorial herniation) than in lateralized, rapidly expanding lesions such as hematomas, which lead to uncal herniation.27 It should also be noted that pupil dilation in coma may be due to mechanisms other than compression of the oculomotor nerve by herniation, such as brain stem ischemia.22,28,29 In medical comas, pupil dilation is a late event preceded or accompanied by other signs of brainstem dysfunction, such as deepening of coma, abnormal posturing, abnormal oculocephalic (Doll’s eye) responses, and disordered respiratory patterns. These signs should alert the clinical team to a serious deterioration in a patient’s condition.
Studies using undilated funduscopy are likely to miss many abnormalities and all those peripheral to the posterior pole.19,30 In a parasitemic, comatose child, the additional diagnostic information obtained from dilated funduscopy is arguably of greater utility than maintaining constant pupil monitoring.7 There is no additional intervention that is likely to be of benefit in a patient diagnosed with cerebral malaria should they develop pupil abnormalities. Therefore, we believe the diagnostic benefit gained by dilated funduscopy outweighs the disadvantage of temporarily interfering with pupil signs pharmacologically after the initial pupil examination. More than 1,000 children with cerebral malaria in Malawi have had pupils dilated for retinal examination, without identifiable detriment to clinical decision-making.13
|
DIAGNOSTIC VALUE OF RETINOPATHY |
A large study in 10 Tanzanian district hospitals showed that 54% of people treated for severe malaria had no parasitemia.2 In this study, most children (55%) less than five years of age in moderate and low transmission areas thought to have severe malaria were misdiagnosed because blood films were inaccurate or ignored. The rate of malaria diagnosis without parasitemia was lower in high transmission areas (31%), probably because a higher proportion of children who came to the hospital were parasitemic.
Limitations of current diagnostic methods. Since severe malaria without peripheral parasitemia is rare, a correctly examined blood film provides an important contribution to diagnosis. Given the required microscope, reagents, and trained technicians, the limitation of correctly identifying a parasitemia is that it may be incidental to another cause of the severe illness.
Rapid diagnostic tests based on the detection of Plasmodium-specific proteins are simple to use, but less simple to interpret. These tests rely on detecting different parasite proteins, and vary in sensitivity between malaria species. They do not give any information on the density of infection, and have lower sensitivities at low parasite densities.31 Fundamentally, they cannot distinguish a parasitemia causing a malarial illness from an incidental parasitemia.
In a malaria-endemic area, any child with signs of severe disease may have a parasitemia that is not the cause of the critical illness. Recent autopsy data suggests that this situation may be more common than is generally recognized.7 The parasitemia obviously requires treatment, but a child will not recover unless the main pathology is first suspected, and then addressed. Most of the clinical features of severe malaria in children, notably coma, convulsions, abnormal posturing, respiratory distress, anemia, and jaundice, are not specific. There is considerable overlap between the clinical signs of severe malaria and pneumonia for instance, and studies in east African children have concluded that they cannot be reliably distinguished on clinical grounds.5,6
Most children who die of malaria in Africa have cerebral malaria manifest as profound coma.32 Even in tertiary African hospitals facilities are often not available to reliably exclude other causes of coma in a parasitemic child. In a prospective clinicopathologic study of Malawian children diagnosed with cerebral malaria, 23% of the 31 examined postmortem had an alternative cause of death, with no histopathologic findings typical of cerebral malaria.7 Since all these children satisfied the clinical case definition of cerebral malaria and would have been classed as such had an autopsy not been performed, this number of other diagnoses was unexpected. Of all the clinical and laboratory features documented, only malarial retinopathy distinguished malarial from non-malarial coma.
Malarial retinopathy. In this Malawi autopsy study, 10 different ophthalmologists demonstrated a collective sensitivity of 95% and specificity of 90% when using the detection of retinopathy as a diagnostic test for cerebral malaria in subsequently fatal cases of coma.7 This compares with a specificity of 61% for the diagnosis of cerebral malaria using all other available clinical and laboratory data applied to WHO diagnostic criteria for strictly defined cerebral malaria. The reference standard used in this study was sequestered, parasitized erythrocytes in cerebral blood vessels at post-mortem examination. Patients with parasitemia, but no malarial retinopathy, were found at post mortem to have evidence of alternative causes of death and no significant cerebral sequestration. One patient with a non-malarial coma was deemed to have a trace of retinal whitening and no other retinal abnormalities. If one considers only parasitemic patients, detection of any malarial retinopathy improved the specificity of the cerebral malaria diagnosis from 61% to 100%. Overall, when using malarial retinopathy to diagnose cerebral malaria in any fatal coma, these data produce a positive predictive value (PPV) of 95% and a negative predictive value of 90%, compared with a PPV of 77% for parasitemia with exclusion of other causes.
The combination of retinal abnormalities associated with severe malaria is unique. To our knowledge, retinal whitening and vessel changes are not seen in any other ocular or systemic condition. None of these changes have been seen in Malian febrile controls,15 critically anemic Malawian children without malaria (hematocrit = 7–8%), or Malawian children with meningitis.12 The definite presence of malarial retinopathy of any severity confirms a diagnosis of severe malaria, and is particularly useful in a parasitemic child in a coma (Figure 7).
|
Malarial retinopathy is absent in approximately one-third of children diagnosed with cerebral malaria. From autopsy data of Taylor and others, in which other causes of death were established,7 it can be deduced that the cause of coma in some of these patients with normal fundi may not be malaria. Alternatively, they may have a prolonged postictal state after a febrile convulsion caused by otherwise uncomplicated malaria. Without a reliable reference standard applicable during life, assessing the sensitivity of malarial retinopathy in non-fatal cases is difficult. However, a normal fundus in children meeting the clinical definition of cerebral malaria is associated with a shorter duration of coma, and a lower risk of death.12 The absence of malarial retinopathy should alert the clinician to the possibility of other causes of coma, particularly if the coma is prolonged.
Attempting to use malarial retinopathy to predict children who might deteriorate is difficult because fundus examination in alert children is more demanding, and only a small minority will develop cerebral complications on treatment. In a study in Gambia, retinopathy on admission was associated with later convulsions, but this association was confounded by pre-admission convulsions.20 We have found that a large increase in the number of retinal hemorrhages can be related to clinical deterioration and even death.12
|
RETINAL CHANGES IN ADULTS |
Cerebral malaria in adults is common in southeast Asia and India, where studies of ocular findings have been undertaken. Retinal hemorrhages have been found to be less frequent than in children, and to be associated with severe disease, but not death, in Thai and Indian adults.17,35,36 Retinal edema and exudates have been described as infrequent findings, but these terms presume etiologies that are unproven, and it is not clear whether they refer to retinal whitening seen in children. Macular retinal whitening has been observed in two Malawian adults with cerebral malaria in a prospective study of patients admitted with fever.37 However, malarial retinopathy, as described in African children, does not appear to be a prominent feature of severe malaria in adults. In particular, the unique vessel changes have not been documented. Systematic studies are required to establish the prevalence and significance of malarial retinopathy in adults.
|
PATHOGENESIS OF RETINOPATHY |
The number of retinal hemorrhages seen on funduscopic examination correlates with the number of cerebral hemorrhages in fatal cerebral malaria.39 In common with cerebral hemorrhages, fibrin thrombi are seen in the small vessel at the center of hemorrhages and hemorrhaged red blood cells rarely contain parasites.
Functional studies have been conducted to investigate the pathophysiology of retinal signs.10,16,40 These suggest that macular whitening is caused by oncotic swelling of second-order neurones in the inner retina due to metabolic or hypoxic stress. A study using fundus fluorescein angiography suggests that metabolic steal by intravascular parasites may play a role, rather than capillary obstruction, which was not seen in association with macular whitening.10 Hero and others performed fluorescein angiography on 12 Kenyan children with cerebral malaria and also found that fluorescein did not breach the blood retina barrier or accumulate as extracellular edema.10 The typical distribution of macular whitening, in sites of high metabolic demand and in vascular watershed zones, supports causation by metabolic steal or hypoxia over toxic malarial products, cytokines, or nitric oxide.
Further histopathologic studies of malaria retinopathy are underway or planned, including the electron microscopic appearances of retinal whitening, and detailed comparison of retinal photographs to flat-mounted, whole retinal specimens. These are in conjunction with cerebral histologic examination. However, an important benefit of studying retinopathy in severe malaria is that it can be done in vivo. Vessels with sequestration can be studied in children at any time point, repeatedly, in those who survive as well as those who die. With this in mind, a study using digital fluorescein angiography is underway to investigate retinal perfusion and integrity of the blood retina barrier in Malawian children with cerebral malaria. In contrast to Hero and others, this study benefits from the advantages of digital over film-based technology, and the improved resolution of a modern fundus camera.
|
FUTURE DEVELOPMENTS |
We hope that funduscopy will become more widely used to improve the diagnosis of severe malaria in Africa. An inexpensive, lens-free direct ophthalmoscope has recently been developed and launched in the United Kingdom.41 This may enable increased use of direct ophthalmoscopy in resource-poor settings. It remains to be established how effective direct ophthalmoscopy alone can be in detecting malarial retinopathy when pupils are consistently dilated, and when observers undergo specific training. Detection of malarial retinopathy is potentially a low-cost, bedside, diagnostic method, which could improve the precision of the clinical diagnosis of severe malaria.
|
CONCLUSIONS |
Studies on malarial retinopathy have been conducted mainly in Malawi, Kenya, and Gambia, all with involvement of ophthalmologists. There are not enough ophthalmologists in Africa to perform these examinations. Further work is required to assess whether fatal coma with parasitemia has a similar rate of alternative diagnoses at other settings and transmission intensities, and whether detection of malarial retinopathy can improve diagnosis when fundus examination is performed by non-ophthalmologists.
To a clinician detecting malarial retinopathy, its presence confirms a diagnosis of severe malaria, and in cerebral malaria the likelihood of death is strongly related to its severity. If retinopathy is absent in a child thought to have cerebral malaria, the clinician should bear in mind the possibility of other causes of coma, particularly if the coma is prolonged. This knowledge could improve care of critically ill patients with parasitemia in disease-endemic areas, and also improve the power of studies to assess interventions aimed at reducing the persistently high mortality of severe malaria.
Received February 17, 2006. Accepted for publication May 7, 2006.
Financial support: Nicholas A. V. Beare is supported by a Wellcome Trust project grant 074125/Z/04/Z. Malcolm E. Molyneux is supported by the Wellcome Trust grant 074124. Terrie E. Taylor is supported by the National Institutes of Health grant no. RO1 AI34969. The American Committee on Clinical Tropical Medicine and Travellers’ Health (ACCTMTH) assisted with publication expenses.
Disclosure: The Wellcome Trust and the National Institutes of Health NIH, while providing financial support, played no direct part in the preparation, review, or approval of this article. There are no competing financial interests.
* Address correspondence to Nicholas A. V. Beare, Malawi-Liverpool-Wellcome Trust Clinical Research Programme, College of Medicine, PO Box 30096, Blantyre, Malawi. E-mail: nbeare{at}btinternet.com 
Authors’ addresses: Nicholas A. V. Beare and Malcolm E. Molyneux, Malawi-Liverpool-Wellcome Trust Clinical Research Programme, College of Medicine, PO Box 30096, Blantyre, Malawi, E-mail: nbeare{at}btinternet.com. Terrie E. Taylor, Department of Internal Medicine, Michigan State University, East Lancing, MI 48824. Simon P. Harding, St. Paul’s Eye Unit, Liverpool University Hospital, Prescot Street, Liverpool L7 8XP, United Kingdom. Susan Lewallen, Kilimanjaro Centre for Community Ophthalmology, Kilimanjaro Christian Medical Centre Hospital/Tumaini University, PO Box 2254, Moshi, Tanzania.
Reprint requests: Nicholas A. V. Beare, Malawi-Liverpool-Wellcome Clinical Research Programme, PO Box 30096, Blantyre 3, Malawi.
|
REFERENCES |
This article has been cited by other articles:
|
|
 |
|
  T. W. Yeo, D. A. Lampah, R. Gitawati, E. Tjitra, E. Kenangalem, K. Piera, R. N. Price, S. B. Duffull, D. S. Celermajer, and N. M. Anstey Angiopoietin-2 is associated with decreased endothelial nitric oxide and poor clinical outcome in severe falciparum malaria PNAS, November 4, 2008; 105(44): 17097 - 17102. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gwer, C. R.J.C. Newton, and J. A. Berkley Over-Diagnosis and Co-Morbidity of Severe Malaria in African Children: A Guide for Clinicians Am J Trop Med Hyg, December 1, 2007; 77(6_Suppl): 6 - 13. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Eisenhut ROLE OF FOLATE DEFICIENCY IN THE PATHOGENESIS OF RETINAL AND CEREBRAL HEMORRHAGES IN CEREBRAL MALARIA Am J Trop Med Hyg, May 1, 2007; 76(5): 793 - 793. [Full Text] [PDF]
|
|
|
|
|
|
N. A. V. Beare, T. E. Taylor, S. P. Harding, S. Lewallen, and M. E. Molyneux ROLE OF FOLATE DEFICIENCY IN THE PATHOGENESIS OF RETINAL AND CEREBRAL HEMORRHAGES IN CEREBRAL MALARIA Am J Trop Med Hyg, May 1, 2007; 76(5): 793 - 794. [Full Text] [PDF]
|
|
TOP
ABSTRACT
INTRODUCTION
