SEARCH

SEARCH BY CITATION

Keywords:

  • morbillivirus;
  • paramyxovirus;
  • measles;
  • disease;
  • infection

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

Morbilliviruses are a group of viruses that belong to the family Paramyxoviridae. The most instantly recognizable member is measles virus (MV) and individuals acutely infected with the virus exhibit a wide range of clinical symptoms ranging from a characteristic mild self-limiting infection to death. Canine distemper virus (CDV) and rinderpest virus (RPV) cause a similar but distinctive pathology in dogs and cattle, respectively, and these, alongside experimental MV infection of primates, have been useful models for MV pathogenesis. Traditionally, viruses were identified because a distinctive disease was observed in man or animals; an infectious agent was subsequently isolated, cultured, and this could be used to recapitulate the disease in an experimentally infected host. Thus, satisfying Koch's postulates has been the norm. More recently, particularly due to the advent of exceedingly sensitive molecular biological assays, many researchers have looked for infectious agents in disease conditions for which a viral aetiology has not been previously established. For these cases, the modified Koch's postulates of Bradford Hill have been developed as criteria to link a virus to a specific disease. Only in a few cases have these conditions been fulfilled. Therefore, many viruses have over the years been definitely and tentatively linked to human diseases and in this respect the morbilliviruses are no different. In this review, human diseases associated with morbillivirus infection have been grouped into three broad categories: (1) those which are definitely caused by the infection; (2) those which may be exacerbated or facilitated by an infection; and (3) those which currently have limited, weak, unsubstantiated or no credible scientific evidence to support any link to a morbillivirus. Thus, an attempt has been made to clarify the published data and separate human diseases actually linked to morbilliviruses from those that are merely anecdotally associated. Copyright © 2006 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

At first sight, the title of this review is surprising as morbilliviruses, with the exception of measles virus (MV), are not known for their propensity to cause disease in humans. Nevertheless, the current controversies about the involvement of MV in various human diseases are best reviewed in the wider context of the pathology of this group of viruses in their natural hosts and the suggested involvement of morbilliviruses in human disease.

Morbilliviruses obtained their collective name from the diminutive form of morbus, meaning plague, and historically, the term was particularly useful in differentiating measles from smallpox and scarlet fever infections 1. As one of the six genera in the Paramyxoviridae family, morbilliviruses are responsible for many severe human and animal diseases. Thus, members infect man: measles virus (MV); cattle: rinderpest virus (RPV); sheep and goats: peste des petits ruminants virus (PPRV); a wide range of carnivores: canine distemper virus (CDV); seals: phocine distemper virus (PDV); and porpoises and dolphins: cetacean morbillivirus (CeMV) 2. As this review focuses on the involvement of these viruses in human disease, we can be very brief about RPV, PPRV, CeMV, and PDV, since in spite of their clear evolutionary relationship with MV and CDV and their potential to use the same cell surface receptors 3, 4, they have not been implicated in any human disease. Domesticated animals infected with RPV or PPRV could potentially transmit very large doses of infectious virus to those who care for them. However, this exposure has not led to any reported or recognizable zoonosis. Possible explanations for the lack of zoonotic infections will be reviewed in a later section in which we review the suggested, but unproven, involvement of CDV in human disease. Thus, we will primarily focus on the basic clinical aspects, molecular properties, pathogenesis, and involvement of MV in human disease.

Measles, the disease and the virus

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

The earliest descriptions of MV are found in Arab writings of Al Rhazes in the tenth century AD 5. Hippocrates did not mention the disease amongst his otherwise careful and accurate descriptions of childhood diseases. This has been interpreted as evidence that the disease was not present in early human populations, although we cannot be sure about this as measles was often confused with other diseases, such as smallpox and scarlet fever. At the time of Hippocrates, human populations were probably too small and too isolated to support endemic MV, which requires interacting populations of 200–300 000 people to supply susceptible children at a sufficient rate to sustain infection 6, 7. Measles is one of the most contagious diseases known. The basic reproduction number (R0), which indicates the number of secondary cases that can be generated from an index case in a susceptible population, is estimated to be more than 15 8 and could be much higher. Hence, the supply of susceptible children is rapidly exhausted in small populations.

The infection is usually not life-threatening and apart from rare but severe sequelae, there are usually no lasting after-effects in individuals who live in the developed world. Here, the mortality rate is approximately 1 : 1000 cases. Complications that require hospitalization such as bronchiolitis and pneumonitis are frequent, occurring in up to 10% of cases. The situation is very different in developing countries. There, measles can have very high mortality rates (5–10%) and severe morbidity is associated with the infection 9, 10. The causes of these differences in clinical outcome between the developed and the developing world are not well understood. Several hypotheses have been put forward such as the dose rate of infection 11, immunological challenge by other infections, vitamin status, and strain differences 12. However, at present, none of these offers a satisfactory explanation for the striking differences in the clinical symptoms manifested in infected children.

Measles is a very serious disease worldwide, with a mortality of still nearly 800 000 children per annum 13. This is primarily associated with malnutrition and secondary bacterial and parasitic infections. Perhaps as many as 7–8 million childhood deaths occurred annually before the introduction of the current live attenuated vaccines and global eradication campaign co-ordinated by the WHO 14. Worldwide vaccination has reduced the number of cases by about 90%. The virus can infect many primate species, including the great apes. However, these are dead-end hosts and the virus can only be maintained in human populations. It does not appear to infect other animals by natural routes.

Molecular biology of the virus

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

In order to understand the pathology of MV and its involvement in human disease, it is first necessary to describe the molecular biology of MV. Apart from a small number of exceptions, the life-cycles of the other morbilliviruses are identical.

MV has a lipid membrane and as such, the virions range in size from 100 to 300 nm (Figure 1A). Six structural proteins are present in the virions. Two glycoproteins span the membrane and form oligomeric spikes which are visible by electron microscopy. The haemagglutinin (H) protein binds cellular receptors (see later) and the fusion (F) protein mediates entry into permissive cells by fusion of the virion and plasma membranes. The F protein is formed as an inactive precursor (F0), which is biologically activated by a proteolytic cleavage involving furin-like proteases 15. In spite of the fact that both proteins form the spike, as illustrated in the schematic representation of the virion (Figure 1), the majority of neutralizing antibodies recognize epitopes on H 16, 17. Following receptor recognition, structural changes in the F glycoprotein initiate virus entry, which occurs in a pH-independent manner. However, even though the structure of the post-fusion ectodomain of a paramyxovirus fusion protein has been recently resolved, the exact molecular details of the rearrangement remain unclear 18. A third hydrophobic viral protein coats the inside membrane of the virion and is thus termed the matrix (M) protein. It is a multifunctional protein having a role in virus budding 19 and transcription regulation 20. Three remaining structural proteins are associated with the genomic RNA and form a helical ribonucleoprotein (RNP) complex (Figure 1B). This is the basic unit of infectivity and is 1 µm in length and 18–21 nm in diameter 21. The nucleocapsid (N) protein is the major component of the RNP, whereas the phospho- (P) and large (L) proteins are present in lower amounts, as illustrated in the schematic representation of the RNP (Figure 1). Functioning as both a transcriptase and a replicase, the L protein is an RNA-dependent RNA polymerase (RdRp) which must be incorporated into all progeny virions.

thumbnail image

Figure 1. Schematic representation of a morbillivirus virion, ribonucleoprotein (RNP) complex, genome organization, and the six mRNAs produced by the RNA-dependent RNA polymerase (RdRp) during transcription. (A) Electron micrograph of a nascent measles virus virion budding from the plasma membrane of an infected cell. Virus spikes composed of the haemagglutinin (H) and fusion (F) glycoproteins are visible on the outside of the virion and a dense area beneath the membrane is where the matrix (M) protein interacts with the cytoplasmic tails of F and H. (B) Electron micrograph of the measles virus RNP showing the characteristic herring-bone configuration. This is also represented diagrammatically in yellow. The RNP is composed primarily of the nucleocapsid (N) protein and it associates with the RdRp, which is made up of the large (L) and phospho- (P) proteins. (C) Diagrammatic representation of the nucleoprotein-encapsidated measles virus genome showing the gene order of the N, PV/C, M, F, H, and L structural proteins and the V and C non-structural proteins which are translated from the PV/C gene. Messenger RNAs are indicated beneath the genome. Their abundance represents the gradient of transcripts that arises due to the start–stop mechanism of transcription from the single promoter in the 3′ end of the genome

Download figure to PowerPoint

Transcription from the non-segmented 15 894 nucleotide-long, negative sensed RNA genome commences once the RNP is liberated into the cytoplasm. The genes encoding the structural proteins are present in six transcription units (TUs) which are separated by non-transcribed intergenic (Ig) sequences. A single promoter in the 3′ leader is recognized by the RdRp and transcription proceeds sequentially by a start–stop process producing polyadenylated mRNAs. As the polymerase can detach from the RNP after transcribing any of the genes and can only re-access it at the promoter, a gradient of transcripts is generated and relative mRNA amounts reflect the viral gene order, ie N > P > M > F > H > L (Figure 1C). Two additional non-structural proteins, V and C, are encoded within the P gene and sometimes this gene is referred to as PV/C. These are generated from a downstream overlapping reading frame (C) and from an ‘edited’ mRNA containing an additional non-templated nucleotide inserted by RdRp slippage (V) 22. Reverse genetics approaches have shown that neither is required for replication in vitro23, 24, although perturbations in the replication of both viruses are observed in primary cells and animal models 25–27. Inhibition of the type I interferon response by C has been reported 28.

Replication involves the RdRp reading across the Ig sequences to produce a full-length positive sensed copy of the genome which is concomitantly encapsidated with N protein. The resulting (+)RNP contains a strong promoter in the 3′ end which directs the production of many more genomic (−) RNPs. Virus assembly occurs in lipid rafts at the plasma membrane to where the F and H glycoproteins move from the Golgi apparatus. The M protein associates with the cytoplasmic tails of the glycoproteins and the (−) RNP. Virions are released by the cell by budding (Figure 1A), although the stages of this process are not particularly well understood.

Many genomic sequences have been obtained for both wild-type and vaccine strains of MV. The virus is monotypic, existing as a single serotype, and infection with one strain appears to provide life-long protection from the disease. With the development of RT/PCR and DNA sequencing techniques, it has become clear that virus isolates vary in their nucleotide sequences, especially in those encoding the last 150 amino acids of the N protein and the entire H protein 29–32. At least 22 genotypes exist and these have been grouped into eight clades. All vaccine viruses are genotype A. Although these viruses may once have been geographically restricted in their distribution, the increase in travel by children, who are the prime reservoir for virus replication, has now allowed worldwide distribution of most of the genotypes.

Pathogenesis of MV infection

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

Many aspects of MV infection are well understood and viral pathogenesis is reviewed comprehensively by Diane Griffin 33. MV spreads from person to person within aerosol droplets. The virus is extremely infectious (see above) and a single tissue-culture infectious dose can cause disease in monkey models 34. After inhalation, the virus replicates in the upper respiratory tract, in the epithelia of the trachea and the lung, although it is not entirely clear in which cell types initial replication occurs. The older literature describes that virus replicates in epithelial cells that line the lungs and the upper respiratory tract. However, there are also cells of the immune system in these tissues and it is possible that these cells are infected first. In this context, it is important to consider what is currently known about MV receptors, as receptor distribution is an important determinant of pathogenesis for many viruses. CD46, also known as membrane cofactor protein, which plays a role in protecting cells from complement lysis, is a cellular receptor for laboratory-adapted and vaccine strains of MV 35, 36. In cells infected with these strains, CD46 is down-regulated from the cell surface and this renders the cells more susceptible to complement lysis 37, but this is not the case in cells infected with wild-type MV. In general, wild-type viruses use CD150 (SLAM) as an entry receptor instead of CD46 3, 38, 39. There is still some controversy about the usage of different receptors by different MV strains. If wild-type viruses exclusively use CD150 38, 40, which is not expressed on epithelial cells, this means that they cannot be infected by the inhaled virus. In that case, initial replication of the virus can only take place in tissue resident lymphocytes, macrophages or dendritic cells that express SLAM. If the wild-type viruses are also able to use CD46, then a wider variety of cells could be infected 41. Recently, it has been shown that haematopoietic stem cells express CD150 and these SLAM-positive cells have been detected in sinusoidal endothelium in spleen and bone marrow 42. This opens up the possibility that MV may be able to replicate in hitherto unrecognized cells and tissues, and, in the future, it will be important to examine what effect infections of such stem cells might have on viral pathogenesis.

The second phase of the infection takes place in the immune system as the virus infects macrophages, lymphocytes, and possibly dendritic cells. Very few free virions can be isolated directly from blood, which is consistent with the highly cell-associated nature of the virus. Most of the infectivity appears to be associated with a small subset of macrophages and T and B lymphocytes, and the number of infected lymphocytes is very small (3–5%). The growth of the virus in cells of the immune system leads to an amplification of the number of virions. This causes MV to spread to tissues either directly in the blood or via the transfer of cell-associated virus particles from cell to cell during the intimate contacts that are formed between cells of the immune system during their normal function. The mechanism by which cell-associated virus infects other cells is not entirely clear. It is possible that loosely attached virions on the outside of infected cells can fuse with uninfected target cells. It is also possible that this leads to fusion of the cell to which the particle was attached and the newly infected cell. This would lead to the formation of multi-nucleated cells or ‘syncytia’.

Lymphoid organs such as the spleen, the lymph nodes, the appendix, and the tonsils are very important sites for primary virus replication 43. From here, the virus can spread to a large number of organs including the skin, the conjunctiva, the kidney, the lung, the gastrointestinal tract, respiratory mucosa, genital mucosa, and the liver. Infection appears to initiate in these tissues either by infiltration of infected macrophages and lymphocytes or by infection of epithelial or endothelial surfaces of these tissues that then allows propagation of the virus and penetration to the inside of the tissues. These processes take place during the asymptomatic phase. By 10 days post-infection, the patient becomes symptomatic, developing the rash that is characteristic of the infection.

Symptoms and transmission

MV infections are characterized by a maculopapular rash, dry cough, coryza, fever, conjunctivitis, and photophobia. A characteristic feature is the appearance of Koplik spots, which are similar to those in the skin but appear in the mucosal surfaces of the mouth 1–2 days earlier. Delayed-type hypersensitivity responses to pre-existing antigens, such as tuberculin, transiently disappear during acute measles infection and convalescence. The description of this phenomenon by von Pirquet in 1906 made MV the first recognized immunosuppressive agent. The immunosuppression does not affect the development of an immune response to the virus itself or to co-administered antigens.

Much of the knowledge that has been obtained in the area of pathogenesis comes from experimental infection in non-human primates (see below) or ex vivo studies, and precise details of the pathogenesis in humans are scant. Spread to the mucosal surfaces of the lung probably allows transmission to a new host. The virus is also detectable in the urine of the patient, probably as cell-associated virus, for approximately 1 week after the rash appears. This does not contribute to person-to-person transmission of the virus. The main route of transmission probably involves virus shedding from immune system cells in the lung of infected children and the aerosolization of these during coughing.

Central nervous system (CNS) complications of measles

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

Children with a well-functioning immune system overcome the acute infection successfully and few long-term sequelae are known. The three most serious and severe complications occur in the CNS. These are acute demyelinating encephalomyelitis (ADEM), measles inclusion body encephalitis (MIBE), and subacute sclerosing panencephalitis (SSPE); the last two are invariably fatal.

Acute demyelinating encephalomyelitis (ADEM)

ADEM occurs about 5–6 days after the appearance of the rash, when 1 : 1000 infected children start to suffer from acute encephalitis 44. It is three orders of magnitude less common in vaccinees. Infection of the CNS has been suggested to be a normal occurrence, as wide-spread electroencephalogram (EEG) changes are seen in almost half of patients. However, cytokine effects from the infection may be an alternative explanation for these. No virus has ever been demonstrated in the brains of children who died from the acute encephalitis 45. ADEM is associated with widespread perivascular demyelination of the myelin sheath of neurones near small blood vessels in the brain, although what and how this occurs is not clear. One suggested mechanism is that it is an auto-immune reaction. However, at present, there is no consensus opinion about the precise mechanism(s) involved. Due to its rarity, it has been very hard to study ADEM.

Measles inclusion body encephalitis (MIBE)

The second type of CNS infection, MIBE, occurs in immunocompromised patients between 2 and 6 months after acute infection 46. This can follow wild-type virus infection as well as vaccination. Dysgammaglobulinaemia or a pre-existing undiagnosed immune abnormality has been found to play a role in the two cases that were associated with vaccine strains 47, 48. This immune diathesis provided the children with only limited ability to restrict virus infection and may have allowed the virus to persist and reach their brain. The mechanism of virus spread and persistence in the brain in MIBE patients are not well understood.

Unlike SSPE (see later), MIBE is not associated with a hyper-immune antibody response to measles proteins and oligoclonal bands are not detected in CNS. Recently, MIBE caused the death of a 13-year-old boy who had been treated for chronic granulomatous disease by stem cell therapy. Neither the patient nor the stem cell donor had apparent recent measles exposure or vaccination, and neither had visited a region of the world where MV was endemic 49. The simplest explanation may be persistence of the virus in either the donor or the recipient. Nevertheless, repeat exposure cannot be ruled out in this case as the virus is highly infectious.

Subacute sclerosing panencephalitis (SSPE)

Much rarer still is SSPE, which affects one in 10 000–25 000 children after acute measles 50. SSPE predominates in males at a 2.5 : 1 ratio 51. This disease is caused by a persistent MV infection 52. The factors that turn a normal acute infection into one that will many years later manifest itself as SSPE are unknown. Acquiring the acute infection under the age of 2 is a risk factor 53. A potential role for maternal antibody has been postulated in the process due to the fact that antibodies have been considered to be able to generate persistent infection and modulate acute infection in vitro in cell cultures 54, 55. However, no direct evidence for this suggestion is available. Early epidemiological studies also showed a rural prevalence for SSPE, but that has not been confirmed in later studies and may simply reflect changes in the patterns of infection 56. All the reliable studies on SSPE have been carried out in countries in which measles vaccination rates have been high. One reason for the rural prevalence has been suggested to be the involvement of co-infection with an animal virus, especially CDV. However, there is no evidence to support this notion. The persistent brain infection leads to a hyper-immune antibody response and this is a pathognomic feature of the disease. There are extremely high titres of neutralizing antibody in both the serum and the cerebrospinal fluid (CSF) of patients. CNS resident B cells give rise to MV-specific oligoclonal bands and it has been demonstrated that the antibodies present in these have undergone affinity maturation 57. Antibodies to structural proteins such as N, F, and H are present in amounts that exceed by ten-fold the normal range observed in the acute infection 58. Antibodies to the M and P proteins are much less prevalent and the lack of an immune response to M in particular has been taken as an indicator that there are mutations in this gene. There is no evidence for abnormalities in cell-mediated immune responses to MV in SSPE patients.

On average, it takes approximately 8 years after an acute infection for the first symptoms of SSPE to appear 59. In the early stages, children lose attention span and show severe neurological symptoms such as ataxia. Then, in all cases, the disease progresses to a situation where the child slides into a vegetative state and dies from the infection. The disease manifests itself in severe demyelination and profound infection of neurones. MV-specific inclusions are present both in the cytoplasm and in the nuclei of infected neurones 60. In the latter stages of SSPE, small numbers of oligodendrocytes, astrocytes, and endothelial cells have been shown to be infected 61. It is assumed that the intra-cytoplasmic inclusions represent the sites of transcription and replication. However, the nature of the nuclear inclusions has not been confirmed. These could be viral RNPs but are more likely to be complexes of viral N protein, which has been shown to be able to translocate to the nucleus 62. One of the longest recorded patients carried the persistent infection for over three decades before the onset of symptoms. In terms of the duration of the symptomatic period, there is a report of a 52-year-old who died 4 years after diagnosis 63. The incidence of SSPE has been significantly reduced by measles vaccination programmes.

The defective nature of virus in SSPE and MIBE brains

Much is known about the molecular characteristics of SSPE viruses (reviewed by Rima and Duprex 64). In general, no infectious virus can be recovered from the CNS tissue at autopsy or biopsy. Mutations in the virus genomes have been documented by sequencing RT/PCR products amplified from autopsy brain material of the patients. In essence, the main mutations observed in SSPE-derived MV render the M, F, and the H proteins non-functional (see below), whereas the N, P, and L proteins are relatively unaffected. This is not surprising as the latter proteins are essential for replication and hence absolutely necessary to maintain the persistent state 65, 66. Mutations that reduce the ability of the virus to fuse with cells or decrease the budding from cells are typical and it appears that such mutations are selected in cases of SSPE. These mutations may allow the virus to grow in the special environment of the CNS in which neurons are in close contact and where there may be particular mechanisms to transfer molecules from the cytoplasm of one cell to another as part of the normal function. Indeed, the virus may exploit these mechanisms to ‘piggy back’ from cell to cell.

A number of mechanisms have been identified which lead to a decrease in the amount of M protein expressed. Generation of bicistronic mRNAs from the P and M genes during transcription by polymerases which ignore the Ig sequences leads to a significant decrease in the number of monocistronic transcripts which can be utilized to produce the M protein 67, 68. As the second open reading frame (ORF) in the bicistronic mRNA is inaccessible to the ribosomes, this should further decrease M protein concentrations. In addition, there is a general perturbation in the relative levels of mRNAs during MV transcription in cells of the CNS. This affects the steepness of the gradient of gene expression, leading to significantly lower levels of F, H, and L proteins. This could allow the virus to evade detection by the immune system and limit the release of virions 67, 69. Sequencing studies have identified many nucleotide changes in the M gene such as alterations to the initiation codon, hypermutation events, and insertion of premature stop codons 65, 68, 70–72. Biochemically, the resulting proteins are more susceptible to proteolytic degradation and have decreased stability. Recombinant viruses have been generated in which the M gene has been deleted. The viruses are highly cell-associated, are more neuro-invasive, but less virulent in a transgenic mouse model 73. This leads to a decrease in the mortality of the animals.

Mutations are also common in the gene that encodes the F glycoprotein. These tend to cluster in the cytoplasmic tail, which is highly conserved in all morbilliviruses. Sequencing studies have identified premature stop codons and nucleotide deletions that cause frame-shifts 74. These truncations and extensions of the cytoplasmic tail are predicated to interfere with an, as yet, unidentified function which may be connected to virus budding or lipid raft targeting. For example, it is known that a specific tyrosine is important as a basolateral sorting signal 75. In a recombinant virus in which the MV F and H glycoproteins have been replaced by the vesicular stomatitis virus G glycoprotein 76 the M protein is absent from the virions. Although the virus grew to low titres this suggests that the cytoplasmic tails of F and H interact with M and that the protein is not essential for budding. Other recombinant viruses which lack only the cytoplasmic tails also showed lower levels of virulence in vivo73, 77.

Recognition of MV genotypes allows sequences obtained from SSPE material to be compared with wild-type and vaccine viruses. To date, no sequences from SSPE cases have been recognized as vaccine-like (ie genotype A). Furthermore, the sequences are typically related not to the currently circulating wild-type viruses, but to those in circulation when the individual developed the acute infection as an infant. These observations confirm that SSPE represents the prime example for the long-term persistence of an RNA virus in the human and that it is not caused by MV re-infection 78, 79 (P Rota, personal communication). This is also consistent with the fact that an acute MV infection confers life-long immunity to the individual. Interestingly, although it is clear that SSPE is caused by MV, what has not been established is where the virus resides and replicates in an individual during the intervening 8–12 years. Thus, the commonly held belief that the virus is present in the CNS has not been formally proven. What is currently known about the viral receptors argues that the site of persistence may not be in the CNS, as the major cell types in which the virus replicates (neurones and oligodendrocytes) do not express SLAM at detectable levels 80.

Other known complications of measles

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

One of the main complications of acute measles is interstitial pneumonitis and mucosal inflammation in the respiratory tract 81. Giant cell pneumonia is much more clinically significant and is primarily a feature in immunocompromised people. Large ‘giant’ cells are caused by fusion of infected cells with neighbouring cells. The resulting destruction of the epithelium leads to bronchial pneumonia and this is usually associated with damage to the lungs and hospitalization of the children. Normally, this will resolve and no long-term damage ensues.

MV induces a clinically significant immunosuppression 82–84. The mechanism by which immunosuppression occurs is not clear but both long-lived cytokine imbalances and direct effects on the proliferation of lymphocytes have been demonstrated. It is also possible that haematopoietic stem cells are deleted in the infected individual as these express CD150 42. The latter two mechanisms may explain the marked lymphopenia associated with the disease.

Immunosuppression has been suggested as the probable cause of many of the complications of measles. For example, it is not clear whether the occurrence of diarrhoea is caused by virus-induced damage to the epithelia in the gastrointestinal tract or whether it results from the virus-induced immunosuppression and subsequent secondary infections. Diarrhoea is most common in the developing world, where inter-current bacterial and parasitic infections are frequent, and is much less common in the developed world. Otitis media and laryngo-tracheo-bronchitis are possibly other manifestations of secondary infection associated with the virus-induced immunosuppression.

Blindness and MV infection

Although the mortality associated with MV infections is often cited, it is much less commonly reported that the virus is the leading cause of blindness in the developing world 85. Conjunctivitis is common in nearly all affected individuals, irrespective of their socio-economic status. Patients demonstrate swelling around the eyes and are photophobic. Viral RNA can be detected in tear secretions by RT/PCR 86. Corneal inflammation (keratitis) affects a significant proportion of individuals. If the patient has a balanced diet and lives in the developed world, long-term ocular defects are rare 87. However, oftentimes this is not the case in developing countries, where malnourishment and particularly vitamin A deficiency are associated with severe keratitis. This is due to the fact that the integrity of the eye is compromised, as, for example, mucins are present in lower levels in such individuals. Secondary infections, both bacterial and viral, are known to exacerbate MV-induced corneal complications. The spectrum of clinical manifestations associated with MV infection is similar to that linked to vitamin A deficiency; thus, high-dose capsules (200 000 IU) are provided twice a year and food fortification programmes have been promoted 88, 89. Furthermore, vitamin A is prescribed to children hospitalized due to a MV infection on admission, re-administered the following day, and again in the following month. Most vaccination programmes in the developing world also include co-administration of vitamin A. This simple treatment is highly efficacious in facilitating the resolution of corneal ulceration. The precise mechanism of vitamin A action in MV-infected individuals is not clear but several clinical trials have confirmed its efficacy in reducing mortality.

Infection of the cornea highlights one of the key conundrums of MV pathogenesis, namely that cells which lack the currently identified receptors (specifically SLAM) can be infected by wild-type viruses. It is clear that wild-type viruses utilize SLAM and this receptor is not detected on cells of the corneal epithelium. To the best of our knowledge, no study has presented unequivocal double-stained immunohistochemical data demonstrating direct infection of corneal epithelial cells. This paradox is mirrored in the CNS, where again neurones, which also are not known to express SLAM, can be infected in SSPE (see above). MV-induced pathology has also been observed in the retina and optic nerve, and some SSPE patients initially present with a loss of vision 90, 91. In one case, an initial diagnosis of an acute multifocal placoid pigment epitheliopathy-like lesion was made, as a white infiltration of the macular area and papillary oedema were observed 92. However, subsequently, progressive neurological defects were observed and the diagnosis was revised to SSPE. Multifocal sub-retinal lesions have also been reported in patients previously diagnosed with SSPE and histopathological analysis has identified RNP-like structures in retinal samples 93.

Hearing loss and measles infection

Loss of hearing is commonly associated with MV infections 94. A key complication is otitis media, which occurs in up to 10% of individuals who have been hospitalized due to the acute infection 95–97. The mechanism by which this occurs is unknown and again it could possibly be due to the virus-mediated immunosuppression, which allows opportunistic secondary infections in the middle ear by decreasing levels of IgG and IgM within this compartment. However, whether this occurs directly or indirectly is unclear 98. Pathological changes have been reported in four individuals with severe necrotizing otitis media who died due to secondary bronchopneumonia, indicating that the virus may directly infect the middle ear 99. Moreover, multi-nucleated cells were detected in one of these cases, although neither MV antigen nor RNA was detected.

The middle ear fluid (MEF) is kept sterile by immune mechanisms involving B lymphocytes and antibody. The B lymphocytes that produce the immunoglobulins may originate from the mucosal-associated lymphoid tissue or the cells may enter the middle ear mucosa from the vasculature. If antibody secreted into the MEF is derived from serum antibody, then it is unlikely that localized measles infection will be able to suppress this significantly. However, Sloyer et al100 suggested that the mucosa lining the middle ear cavity is an active epithelium that locally produces antibodies and the skewed ratio of MEF to serum IgA supports this. Direct MV infection in the antibody-producing cells might compromise the sterility of the MEF. Evidence for local production of IgA against MV in 16 out of 41 cases of otitis media was reported 100. However, in an equal number of cases, the IgA detected was against poliovirus, and in four and five cases, against mumps and rubella virus, respectively. This suggests that the reaction is probably not MV-specific.

Damage to the cochlear sensorineural elements or to the cochlear nerve is known to result in permanent deafness. Historically, sensorineural loss has been linked to ADEM 101 and pathological alterations have been reported to occur in the inner ear 94, 102–104. However, to this day, the precise role that MV plays in irreversible deafness is uncertain, as neither viral antigen nor RNA has been detected in samples from the inner ear. In addition, it is not clear if the pathological changes are due to the virus directly infecting particular cells or are caused indirectly due to the associated inflammatory response and bystander effects or to virus-induced autoimmunity 105.

Thrombocytopenia

Thrombocytopenia and seizures occasionally occur after measles infection. Both resolve quickly and long-term consequences are rare. The live attenuated vaccine induces thrombocytopenia in less than 1 per 20 000 vaccine doses 106, 107. These complications are much more common after infection with wild-type MV strains and thus the vaccine has a protective effect against this complication. An auto-immune mechanism has been proposed to explain this phenomenon but the evidence for this remains scarce 108, 109. It has long been recognized that MV infection and other viruses are associated with the generation of auto-antibodies to a number of antigens including smooth muscle antigens and nuclear antigens. In general, this appears to be a non-specific transient effect that disappears 3–4 weeks post-infection.

Diseases that have been speculatively associated with MV infection

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

Due to its ability to establish persistent infection and the paradigm of SSPE, MV has been suggested to have an association with a large number of diseases (Table 1). Thus, often single and unconfirmed reports have suggested an association with achalasia, insulin-dependent diabetes, thyroiditis, and Kawasaki's disease. These suggestions have not been investigated extensively. Other suggested sequelae of MV infection which have been subject to more study are multiple sclerosis, epilepsy, systemic lupus erythematosis, chronically active auto-immune hepatitis, Paget's disease, otosclerosis, autism, inflammatory bowel diseases such as Crohn's disease, and ulcerative colitis 110. However, no confirmed evidence has been presented to substantiate these associations, let alone prove a causal relationship between these diseases and MV infection. These unsubstantiated associations have had a significant effect on the uptake of measles vaccines and hence the studies that suggest these links are reviewed here in detail.

Table 1. Human diseases in which morbilliviruses have been suggested to play a role
Link establishedLink possibly due to virus-associated immunosuppressionLinked on the basis of largely unconfirmed or flawed evidenceLinked on the basis of single unconfirmed reportsAnecdotally linked or no evidence using modern diagnostic techniques
Acute demyelinating encephalomyelitisOtitis mediaMultiple sclerosis (MV and CDV)EpilepsyAchalasia
DiarrhoeaPaget's disease of bone (MV and CDV)AutismThyroiditis
Measles inclusion body encephalitisOtosclerosisSystemic lupus erythematosisJuvenile-onset diabetes
Subacute sclerosing panencephalitisAuto-immune chronically active hepatitisKawasaki's disease
ThrombocytopeniaCrohn's disease
Ulcerative colitis

Systemic lupus erythematosis

Measles has been suggested to play a role in systemic lupus erythematosis (SLE). The only evidence to support this claim is the detection of MV nucleic acids in tissue samples, and no other virological data, serological or epidemiological evidence exist to support the link. The disease is characterized by the presence of anti-nuclear antibodies and antibodies to the La protein 111. An early study demonstrated the presence of MV-specific RNA in peripheral blood mononuclear cells (PBMCs) 112. However, this has not been confirmed 113 and it is most likely that the probes used in the early studies were contaminated with cellular sequences, as they were made by direct labelling or reverse transcription of RNA from ‘purified’ virus or from purified MV-specific mRNA. One study using a cloned probe 114 indicated by dot blot the presence of MV RNA in 28 out of 34 SLE patients and not in 29 controls. A study using the more sensitive S1 nuclease protection assay was not able to confirm the presence of MV-specific RNA in the PBMCs of SLE patients 115. However, to the best of our knowledge, RT/PCR has not yet been applied. The absence of other strands of evidence and lack of confirmation of the earlier data have probably diminished enthusiasm for carrying out such studies.

Epilepsy

A single case report from the Kawashima group 116 presents data obtained from a patient who presented with epilepsy 9 years before samples were taken (PBMCs and CSF). The patient had been vaccinated at 12 months with MV. However, the virus detected was a wild-type that had circulated at the time of first presentation. All controls from patients with multiple sclerosis (one), measles encephalitis (3 years after first presentation; one), epilepsy after measles infection (one), HIV encephalopathy (one), and Lennox Gastaut syndrome (two) were negative. These samples were obtained from individuals who had been vaccinated against or naturally infected with MV within the previous year. Control PBMCs and CSF from an SSPE patient were positive. This study has not been confirmed, nor have there been follow-up studies from the same group.

Auto-immune chronically active hepatitis

MV was first implicated in this condition when dot-blot hybridization was used to detect genome in lymphocytes using a 50-nucleotide-long oligonucleotide probe complementary to a region of the nucleocapsid gene 117. This study has several technical shortcomings and it is generally accepted that dot-blot hybridization is not sensitive, is susceptible to artefacts, and is not a robust method for demonstrating what the authors claim to have shown—specifically that 12/18 auto-immune chronically active hepatitis patients were MV RNA-positive as well as 1/3 SLE, 2/4 cryptic cirrhosis, and 0/37 controls.

Another study examined samples from two adult and four paediatric cases 118. A sequence corresponding to a wild-type strain only circulating in the early 1990s was identified from a 53-year-old individual. As this person would almost certainly have been exposed to measles in Japan, finding such a virus is very surprising. The other sequence obtained from a 60-year-old could not be typed at the time. However, further analysis shows that this sequence is most similar to clade A, to which only vaccine viruses belong. This person would not been vaccinated, therefore they could only have been infected at an early age by an unidentified wild-type strain from clade A. Limited sequence information was provided from the samples obtained from the children. All of these sequences corresponded to vaccine viruses.

Multiple sclerosis

Multiple sclerosis (MS) is the most common neurological disease of young to middle-aged adults. Its prevalence can be has high as 1 in 200 women 119 but wide geographical differences occur with an increasing prevalence the further one is from the equator. There are genetic factors involved in the disease and it is especially prevalent among Caucasians 120. These have not yet been characterized, although specific HLA types are present at a higher than expected frequency in patients. Epidemiological evidence based primarily on migration studies has been consistent in indicating that an environmental factor encountered before the age of 15 is involved in triggering the disease itself or generating a predisposition for it to be acquired at a later age 121.

A large variety of DNA viruses, especially human herpes virus 6, and RNA viruses from many different families, for example, human T-cell leukaemia virus 1 and human endogenous retroviruses, have been studied for their association with MS. None has received more attention than MV 122, 123. Increased levels of antibodies to a number of viruses including measles have been observed in serum and CSF samples 124, 125. However, these appear to be non-specific reactions, as antibodies to a number of other viruses are also increased. MV-specific T-cell responses in MS are less clear as both increases 126 and decreases 127 have been reported. The patients appear to have a constant Th1-biased cell-mediated response 128 and evidence that administration of interferon gamma exacerbates the disease has been presented 129. Direct evidence for the presence of viral nucleic acids is scant and where they have been found, they were almost always also present in control brains.

Haase et al showed that samples from 1/4 MS patients were positive for MV sequences using in situ hybridization (ISH) 130. RNAse pretreatment was shown to abrogate the signal. However, the probe was generated from radioactively labelled cDNA made from viral genomic RNA. This methodology is unacceptably non-specific, as ribosomal and cellular RNAs cannot be reliably removed. Our group has demonstrated positive ISH in four out of 14 MS cases and one case of cytomegalovirus infection but not in 56 neurological and non-neurological controls 131. Positive signals were only detected in a small number of foci in the MS brains and attempts to confirm these data with other techniques such as RT/PCR have been unsuccessful 132, 133. This may be because viral RNAs could be diluted extensively by cellular RNA. However, the very small number of foci of infection detected by ISH calls into question the aetiological and pathogenic role of MV. At a recent meeting 134, it was still considered likely that an infection triggers MS in genetically predisposed individuals. However, there is a consensus that persistent infection by MV as in SSPE or MIBE is unlikely to be involved because it has been difficult to confirm isolated reports of MV-positive reactions in ISH or immunocytochemistry (ICC). Whether the virus can act alone or in concert with other agents, or as part of a series of infections in triggering auto-immune reactions to myelin components remains to be investigated further, although the involvement of any agent in such a ‘hit and run’ scenario is experimentally very difficult to establish.

Otosclerosis and Paget's disease of bone

Otosclerosis (OS) is a disease that leads to partial or severe deafness due to the irregular growth of bones, primarily the stapes, in the middle ear. Bones grow abnormally due to a defect in metabolism that affects the fine balance between bone deposition and resorption. A similar imbalance gives rise to Paget's disease (PD), which can affect many other bones in the body. Both diseases have a significant genetic component and chromosomal changes have been identified in each 135, 136. However, the genes involved differ between OS and PD. Some HLA haplotypes are associated with OS, as would befit a viral disease. Pathological assessment of tissues from OS patients suggest that a long-lived inflammatory process may cause spongeolytic bone 137, 138. Interestingly, MV has been proposed to play a part in both diseases and virus persistence has been suggested to cause this inflammation 139, 140. However, studies which examined MV antibody levels have been at variance 141–143. A number of studies have identified paracrystalline arrays in osteoblast cells isolated from OS patients 144. These have been suggested to be paramyxovirus-like RNP structures (Figure 1B), although such macromolecular assemblies have been identified in familial expansile osteolysis, a disease with a clear genetic cause 145. Such structures have also been identified in osteoclasts in tissues obtained from patients with PD 146. When samples from both diseases were examined by immunohistochemistry, conflicting data were obtained which suggested that MV, mumps virus or respiratory syncytial viruses were present 144, 147 and these observations were not confirmed in other studies 148. Molecular approaches have been used to detect viral RNA in PD and OS samples 149, 150. However, when the amplicons were sequenced, this showed in all cases that the viruses belonged to genotype A. No wild-type sequences were obtained, which argues against vaccination being responsible for a delay in OS presentation 151. In PD, some nucleotides differed from the published vaccine strain, although when five of these were examined more closely 152, they were due to errors in the sequence used for comparison. Others have not been able to identify MV RNA in samples from patients with PD 152–154. The situation in OS is equally unresolved, as some have provided evidence that MV RNA is present in bone samples 140, 143, 155 and others have not been able to confirm these data 156. Therefore, at present, the involvement of MV in these diseases is at best unsubstantiated.

Inflammatory bowel disease

The link between MV and inflammatory bowel disease (IBD) has mainly been propagated by Andy Wakefield and his colleagues 157, 158. Although they themselves have published negative findings in relation to this link using RT/PCR 159, they considered that the ultimate sensitivity and positive proof were given in a paper by the Kawashima group 160. A recent collaborative study indicated that the sensitivity of the methods employed in the Kawashima laboratory was low, leading to the suggestion that the previous results were either false positives or due to cross-contamination 161. This is supported by the fact that no evidence of MV RNA was found in the gut samples of IBD cases 162 and Wakefield and colleagues consigned them to the control group in the latter paper.

Autism

Wakefield et al also published an inference in The Lancet in 1998 about a putative link between the measles, mumps and rubella (MMR) vaccine and autism 163. This inference has been withdrawn by the majority of the original authors 164, but Wakefield and colleagues continued attempts to find evidence for the presence of measles in the gut of autistic children. They proposed that the presence of the virus would cause a leaky gut epithelium which would allow opioid peptides to enter the bloodstream. These peptides, after crossing the blood–brain barrier, were suggested to interfere with the normal function and/or development of the CNS. No supporting or confirmed evidence for this chain of events has been provided. A single paper directly relating to this 162 has been extensively commented upon on the journal Molecular Pathology's website. From a virological viewpoint, the data are unconvincing as the authors have used research tests that were not validated. This applies especially to the in situ RT/PCR assays, which lack specificity and are notoriously difficult to perform, as evidenced in the paper by the very poor signal in the SSPE controls. Furthermore, only the F gene was detected, which is surprising as it is in a promoter distal position and therefore its mRNA levels would be much lower than the preceding three genes (Figure 1C). As such, there is no evidence to suggest that MV vaccination plays any part in the development of autism.

The role of canine distemper virus (CDV) in human disease

CDV has been postulated to play a role in MS primarily on the basis of epidemiological evidence that has, as yet, not been confirmed. Thus, it is hypothesized that the geographical distribution in incidence is due to the stability of CDV, which would decrease towards the equator 165. An MS link was also promulgated as a result of the so-called MS ‘outbreak’ that followed the invasion of the Faroe Islands by British troops during the Second World War. It was hypothesized that infected dogs brought CDV to the Islands and that this was responsible for the ensuing MS ‘outbreak’. However, evidence for the existence an ‘outbreak’, and for the diagnosis of canine distemper, has been questioned. The evidence for the ability of CDV to cause systemic infection in human beings is also very weak. Morbilliviruses are readily cross-neutralized and this has been exploited in the past in using measles vaccine to protect cattle from rinderpest when no appropriate vaccine was available. Hence the fact that most of the human population would have been exposed to MV at an early age and probably would have carried a maternal antibody against MV that could also neutralize CDV would hinder or prohibit infection by CDV. Cook et al166 suggested a link to CDV but this has not been confirmed. Their studies were confounded by the cross-reactivity of CDV with MV-specific antibodies and the general non-specifically elevated levels of viral antibodies found in MS patients. Recently, brain samples from patients with MS were examined by immunocytochemistry (ICC) using MV and CDV monoclonal antibodies 167. All sections were negative except for one antibody, which bound to 8/9 MS plaques and 2/5 herpes simplex virus encephalitis brain samples, but not six controls or four patients with ischaemic stroke. However, no evidence for the presence of MV in MS plaques was obtained by RT/PCR, calling into question the specificity of the antibody. The authors suggested that the monoclonal antibody may recognize an epitope on an as yet unrecognized morbillivirus present in the human CNS that might be implicated in MS pathogenesis or represent a protein that is up-regulated during inflammation. The latter is most likely the correct interpretation, as Sheshberadaran and Norrby identified a cross-reacting epitope on a 79 kD stress protein 168.

CDV has also been proposed to play a role in PD based on epidemiological evidence of higher than expected dog ownership in the North of England amongst patients, compared with age- and sex-matched controls 169. Other studies were not able to confirm these observations 170. Nevertheless, it led to a number of studies looking for CDV in Pagetic bone tissues. RT/PCR on RNA samples extracted from decalcified bone indicated that CDV was present in 8 of 13 samples but MV was not detected. When the amplicons were sequenced, they were identical to the Onderstepoort strain and not to wild-type strains of CDV. The significance of these results is not clear and contamination cannot be ruled out. Several other studies have not been able to confirm the presence of CDV-specific RNA in nucleic acids extracted from bone, but of course in science it is formally impossible to prove the absence of anything 152, 154. ISH studies using CDV-specific probes were carried out 171. The data appear consistent but the specificity was not high and only 41% of the PD samples were positive with probes for the N and F mRNAs of CDV, even though probes for all genes and anti- and genomic RNA were used. The latter may be explained by the much higher copy numbers of mRNA than genomic RNA present in infected cells. Alternatively, it may be a technical defect in the study, as the authors were similarly unable to detect genomic RNA in dogs with metaphyseal osteopathy, whereas mRNA probes showed positive results. Signals were observed not only in osteoclasts, but also in osteocytes and osteoblasts, whereas the paracrystalline arrays (see above) were observed only in osteoclasts. Recently, Hoyland et al compared in situ RT/PCR, ISH, and RT/PCR for CDV detection in Paget's disease 172. Only in situ RT/PCR was positive in all ten bone samples and not in controls. As commented on above, it is notoriously difficult to obtain specific responses using this technique and hence it must be concluded that the link between CDV and PD is far from established.

The value of animal model systems

One of the major drawbacks in research on MV is that the available animal models only mimic part of the symptoms and pathogenesis of the disease in the natural host. Mouse models have only shown viral replication after intracerebral infection and mostly only using rodent neuro-adapted strains. Transgenic mouse models in which an MV receptor is expressed show little pathology or peripheral infection. When the expression is combined with knock-out of the type I interferon receptor, limited pathology occurs 173. Mice expressing SLAM under control of the lck proximal promoter, which directs gene expression exclusively in T cells, also do not show a great deal of pathology 174. These results have been explained by the observation that mouse cells appear to lack specific factors required from MV replication 175. The cotton rat model mimics aspects of immunosuppression of MV in the human 176, 177. The most realistic model that mimics MV infection in humans is that of infection in macaques 34, 178–184, but the cost of these animals is high and small numbers are available. Rash is shown as a reliable indicator of the value of a model. More studies in macaques are needed of the infection and pathogenesis of wild-type MV viruses that have been passaged in such a way as not to lose virulence. So far, the groups have concentrated on immunological parameters or pathogenesis but more holistic studies that integrate pathogenesis, virological, and immunological aspects are needed.

The pathogenesis of CDV has been studied in dogs and more recently in ferrets, as the latter provides a more tractable animal model. Moreover, CDV has been suggested to be a model for MS 185. RPV infection has been analysed in cattle to a limited extent as this is a rather expensive model 186. Furthermore, studies can be hampered by the lack of immunological reagents.

The development of systems during the last decade which permit the generation of recombinant MVs, ‘virus rescue’ 187, has opened the door to mechanistic studies, not only of MV persistence and its role in neurological sequelae of MV infection and the role of specific genes in this process, but also dissection of the roles of individual genes in isolation or in combination with others, in attenuation, virulence immunosuppression, and interaction with the innate immune system.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References

Morbilliviruses have been associated with an eclectic assortment of human diseases. In this article, we have aimed to review the unequivocal, circumstantial, anecdotal, and often erroneous evidence which links morbilliviruses to particular diseases beyond an uncomplicated acute infection, a hallmark rash, and distinctive immunosuppression. Thus, it is clear that a direct link has been unequivocally established between MV and ADEM, MIBE, SSPE, thrombocytopenia, blindness, and hearing loss. It is also apparent that continuing to study rare conditions such as SSPE may shed valuable light on long-term persistent viral infections. This is all the more important as we further develop and utilize stem cell therapeutic approaches, as is indicated by the individual who developed MIBE after such an intervention. Indirect links to MV via the characteristic immunosuppression caused by the virus can be made to otitis media and diarrhoea. Once again, this demonstrates the utility of continuing to address the underlying molecular mechanisms of the profound immunosuppression induced my MV. We are not convinced from the published data that it is valid at this point in time to link any morbillivirus with multiple sclerosis or Paget's disease. Likewise, the evidence associating the MV with otosclerosis, auto-immune chronically active hepatitis, Crohn's disease or ulcerative colitis remains unconfirmed. Single unsubstantiated reports have associated MV with epilepsy, autism, and systematic lupus erythematosis. No other groups have been able to replicate the work and as such, these reports can be given little credence. Finally, the anecdotal or historical association of MV infection to juvenile-onset diabetes and Kawasaki's disease needs to be either compressively examined using current molecular biological techniques or consigned to the increasing set of reports that have circumstantially linked all kinds of viruses to all kinds of human disease.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Measles, the disease and the virus
  5. Molecular biology of the virus
  6. Pathogenesis of MV infection
  7. Central nervous system (CNS) complications of measles
  8. Other known complications of measles
  9. Diseases that have been speculatively associated with MV infection
  10. Conclusion
  11. References