Ehrlichia and Anaplasma are bacteria related to Rickettsiae. They are obligate intracellular bacteria that grow within membrane-bound vacuoles in human and animal leukocytes. The obligate intracellular bacteria proliferate in white blood cells (monocytes: E. chaffeensis) or granulocytes [Ehrlichia sp. related to E. equi (horses) and E. phagocytophila (cattle)].
Name
The generic name refers to Paul Erhlich (1854-1915), the famous German bacteriologist (Nobel Chemistry Prize 1908), the discoverer of salvarsan, an arsenical preparation active against syphilis. In 1954 the first human ehrlichiosis was described in Japan, caused by E. sennetsu. Since this initial report, several tick-borne infections have now been recognized. Human monocytic ehrlichiosis (HME) was first described in 1986 and is caused by Ehrlichia chaffeensis. The name refers to the American army base Fort Chaffee in Arkansas.
Human granulocytotropic anaplasmosis (HGA) was described in 1993 and is caused by Anaplasma phagocytophilia. Ehrlichia ewingii was described in 1999 as an agent of human ehrlichiosis. E. ewingii provokes “human granulocytic ehrlichiosis “.
Transmission and symptoms
The organisms are transmitted by ixodid ticks. Amblyomma americanum (Lone star tick) is the main vector for E. chaffeensis. In the USA white-tailed deer and coyotes form the most important reservoir. It has been shown that ticks on migrating birds can be infected with Erhlichia sp. and can thus be transported over long distances. Anaplasma phagocytophila in the broad sense is found in rodents such as the dusky-footed wood rats and mice. The reservoir of E. ewingii is still unknown. Transmission occurs predominantly by the bite of infected ticks, but mother-to-child transmission and transmission by blood transfusion or slaughtering of infected animals is reported.
Common symptoms include fever with or without chills, headache, myalgia, arthralgia, weakness, nausea, leukopenia and thrombocytopenia. Rash is uncommon. Liver test abnormalities can be found in about 50% of cases. In rare cases human monocytic ehrlichiosis can be associated with neurological lesions or meningitis. Post-infection asthenia can continue for months. In HIV patients infection can be overwhelming.
Diagnosis
White blood cell and platelet abnormalities are almost always present, so normal values virtually rule out this infection. Anemia is commonly present, so pancytopenia can be suggestive of anaplasmosis or ehrlichiosis.
Probably many infections are missed since laboratory testing is not widely available. The diagnosis of human granulocytic anaplasmosis is made by microscopic examination of a peripheral blood smear or serologic testing. A 4-fold rise in antibody titer between the acute an convalescent phases of infection confirms the diagnosis. Microscopy is labor intensive and the sensitivity of microscopy ranges from 20 to 80% depending on the degree of expertise: bacteria are observed in the cytoplasm of leukocytes as 0.5 to 1.5 µm large inclusions which are combined in groups (morulae). Today, PCR has gained diagnostic importance in high resource settings. Culturing of this intracellular bacteria is complex and it’s the most accurate method, and it is reserved for research purposes.
Human granlulocytic anaplasmosis: Wright’s staining of a peripheral blood smear. Neutrophils show nonspecific toxic granulation and some have intracytoplasmic inclusions (arrow). Source: N Engl J Med 2020;382:1258-66. DOI: 10.1056/NEJMcpc1916
Differential diagnosis:
The differential diagnosis includes rickettsiosis, typhoid fever and several arboviral infections, such as dengue.
The diagnosis of HGA can be overlooked if there is simultaneous infection with B. burgdorferi. In such cases, the typical rash of early Lyme disease (erythema migrans) may mislead the clinician into ignoring possible coinfection with ehrlichia or anaplasma. Findings that may suggest coinfection include leukopenia, thrombocytopenia, and high fever (all relatively uncommon in Lyme disease) and abnormal liver enzyme tests accompanying the erythema migrans.
Treatment
Treatment is based on administration of tetracyclines, e.g. doxycycline 100 mg twice daily for 7 days.
South American-bartonellosis, (Carrión’s disease, Oroya fever, verruga peruana, verruga peruviana) results from infection with the bacterium Bartonella bacilliformis and is transmitted by sandflies. The infection manifests itself in two very different clinical forms with the causal connection being recognised by the young Peruvian doctor Daniel Alcides Carrión.
Pre-Columbian mummies with histologically confirmed verruga lesions have been discovered in Peru and bartonellosis occurred in Francisco Pizarro’s army (1471-1541). During the Inca era, the disease was called “Sirki,” which means “warts in blood.” In Peru between 1869 and 1873 more than 7000 workers building the Lima-La Oroya railway died from the disease at Cocachacra, 65 kilometers from Lima, 1600 meters above sea level. The name “Oroya fever” refers to this, although in the mining town of La Oroya (altitude 3800 m), strangely enough there was no transmission of Oroya fever. In 1936 a large epidemic was seen in the Guaitara valley on the border between Colombia and Ecuador. An epidemic occurred in 1980 in Ecuador and another in 1987 in Peru with a death rate of 88% in the untreated patients. Now and then there have been isolated cases or small outbreaks. In 1997 there was an outbreak in the area of Cuzco, Peru. In an outbreak in Zumba, Ecuador (1995-96), large numbers of dead rodents were found around the places where the cases had occurred. This finding led to the hypothesis that bartonellosis could have an animal reservoir.
Daniel Alcides Carrión
Daniel Alcides Carrión (1858-1885) was a medical student in Lima, Peru. He was required to prepare an original thesis and choose to study the epidemiology and clinical manifestations of verruga peruviana. His home was in Cerro de Pasco, a mining town high in the Andes where he had seen many cases. These left a deep impression on the young man. He told a classmate that he hoped “to make an important contribution to aching humanity”. He became concerned with the difficulty in diagnosing verruga peruviana before the typical eruption started. The appearance of the skin lesions was preceded by fever and anaemia, but there was a lot of confusion between the prodromal phase of verruga and other febrile disorders such as malaria. Carrión wanted to determine the incubation period and early symptoms of verruga, so he decided to inoculate himself with some fluid from a chronic skin lesion of a verruga patient. Many friends and professors tried to dissuade him. On the morning of August 27, 1885, Carrión was in the Nuestra Senora de las Mercedes ward of the Dos de Mayo Hospital in Lima. A 14-year-old boy named Carmen Paredes was admitted with verruga on his right eyebrow. Assisted by Dr Chavez, a young ward physician, Carrión used a lancet to inoculate his own arm with blood taken from that verruga. He kept a diary afterwards. The first symptoms started after 21 days, with discomfort and pain in his left ankle. Two days later he developed fever, chills, abdominal pain and generalised pain in bones and joints. He had anorexia and noted severe thirst. His urine became dark red and scanty. He developed jaundice. A week later, he became too ill to continue his diary. His classmates took over this task and were surprised at how quickly anaemia developed. A systolic heart murmur developed and grew in intensity. A few days later, muscle fasciculations appeared in his arm muscles. He said to his friends: “Up to today, I thought I was only in the invasive stage of the verruga as a consequence of my inoculation, that is, in the period of anaemia that precedes the eruption. But now I am deeply convinced that I am suffering from the fever that killed our friend, Orihuela. Therefore, this is the evident proof that Oroya fever and the verruga have the same origin, as Dr Alarco once said.” This insight was the essence of Carrión’s experiment. He had not set out to prove the single cause of verruga peruviana and Oroya fever. He only intended to study the incubation period and prodrome of verruga. When a completely different disease developed, he was lucid enough to understand the full meaning of his experiment. On October 3, he became delirious and two days later he fell into a coma and died at midday. He became a hero of Peruvian medicine and is remembered to this day. The day of his death, October 5, is celebrated yearly as the “Dia de la Medicina Peruana”. The Peruvian National University in Cerro de Pasco carries his name.
Ponte verrugas in the Andes, a railway bridge on the trail Lima – La Oroya (Peru). The name refers to a bartonellosis epidemic in 1869-1873. Copyright ITMPicture of Dr Daniel Alcides Carrion, on the road Lima – La Oroya. In this area of the Western Andes in Peru, there was a bartonellosis epidemic in 1869-1873. The disease is also known as Oroya fever or Carrion’s disease.
Photo Dr Van den Enden. Copyright ITM
Aetiology
Barton described the pathogen in 1909, but he thought that it was a protozoon. The Japanese bacteriologist Hideyo Noguchi demonstrated the bacterial nature of the pathogen. Bartonella bacilliformis is a small Gram-negative coccobacillus (0.6-1 µm), which takes Giemsa and Warthin-Starry stain. The pathogen has one or more polar flagella. It replicates within the vascular endothelium and erythrocytes. The bacterium is related to rickettsiae. The bacillus grows quickly (extracellularly) on non-living culture media with blood or on chicken embryos at 25-28°C. Numerous related organisms are animal pathogens.
Distribution
The disease caused by Bartonella bacilliformisonly occurs in certain narrow high valleys of the western-most slopes of the Andes at altitudes between 500 and 3200 meters in Peru, Ecuador and Colombia, between 2° N and 13° S. Whether endemic cases occurred in Chili, Bolivia, Guatemala and Honduras is very doubtful. Sporadic cases of so-called “bartonelloses” have been reported in Africa, (Niger, Sudan), in Asia (Pakistan) and in the USA, but it is still not clear whether there is a connection with Carrión’s disease. Our knowledge about Bartonella and related bacteria has largely increased in recent years but is still very incomplete.
Transmission
A sandfly, Lutzomyia verrucarum, and perhaps a few related species, is responsible for transmission. Transmission only occurs at night and is seasonal, particularly during the rains. It was formerly assumed that the reservoir was purely human, but this was recently cast into doubt (there may be a rodent reservoir). In some of the inhabitants in the endemic valleys bacteria can be found in the blood, but these carriers are usually without any symptoms. These latent infections which are likely to have been contracted in childhood probably give stable immunity. It is only if non-immune populations enter the endemic area that epidemics occur, sometimes on a large scale, such as in wars or when large public works are being carried out. Tourists may be at risk for the disease.
Clinical aspects
The clinical range is wide, going from asymptomatic infections via serious febrile forms with acute haemolytic anaemia, to the angiomatous skin lesions which can be present from the onset or can be preceded by the febrile stage. The mortality of untreated cases varies between epidemics and ranges from 10-40% after 2-3 weeks. The disease is less severe in children and the mortality is far lower. If the course of the disease is favourable, the fever can last for 3 to 4 months. In 40-50% of cases of Oroya fever, concurrent salmonellosis (generally Salmonella typhimurium) complicates the illness and makes the prognosis less favourable. The superinfection causes fever with gastrointestinal symptoms and a deterioration of the patient’s general condition.
Acute stage or Oroya fever
1. Incubation takes approximately 3 to 8 weeks (range 10-210 days). It begins insidiously with:
2. Irregular intermittent febrile attacks with shivering
3. Rapidly worsening anaemia with tachycardia, pallor and (sub)icterus
4. Severe headache with bone and joint pain. This may persist after the fever has ended
5. Enlargement of the liver and spleen, slightly painful on palpation
6. Generalised painful swollen lymph nodes
7. Myocarditis, pulmonary oedema and anasarca (generalised oedema)
8. Haemorrhagic diathesis as a result of the endothelial lesions: petechiae and tendency to thrombosis. Necrotic foci are found in the liver, spleen and bone marrow.
9. Neutrophilia
10. Spontaneous abortion, foetal death or transplacental transmission can occur.
11. Neurobartonellosis due to involvement of the CNS takes the form of meningo-encephalitis with or without convulsions and with high mortality. Myelitis also occurs with spastic or flaccid paraplegia with sequelae which can be permanent. There is pleiocytosis of the CSF. More focal and transient lesions of the spinal cord or of the cranial nerves are seen at the verruga stage.
Cholera is an acute infectious disease, characterised by profuse watery diarrhoea. It is caused by a Gram-negative bacterium: Vibrio cholerae O1 (the characters O1 indicate the serogroup). It is a very small, motile, curved bacterium (vibrio is the Greek word for comma). Various subtypes exist, with classification according to biological and biochemical behaviour (biotypes) and serological characteristics (serotypes). Until 1992 it was thought that bacteria causing cholera must belong to V. cholerae, serogroup O1 and that they must be toxicogenic (must possess and express the genes for toxins). It was known that non-O1 Vibrio cholerae could sometimes cause mild gastro-enteritis or even bloodstream infection in immune depressed patients, but not cholera. In October 1992 in Madras (India), a mutated pathogenic bacterium (a new serogroup) was discovered, which also causes cholera. The isolate was given the name Vibrio cholerae O139 (nicknamed Bengalen). After a short bloom, the traditional strains (O1) have become more common again. A few years later, V. cholerae O139 Calcutta was identified.
Antibiotic resistance genes in V. cholerae are often positioned on plasmids and can be transmitted to vibrios from non-pathogenic intestinal flora.
V. cholerae can only cause disease if there are pili [Lat.: “hairs”] present. Pili are shorter and thinner than flagella. The pili adhere to the intestinal mucosa.
Sometimes other Vibrio species are responsible for diarrhoea, e.g. Vibrio cholerae non-O1, V. parahaemolyticus, V. hollisae, V. minicus and V. fluvialis. Our knowledge of these latter bacteria is clearly insufficient. Vibrio vulnificus is an aggressive species present in seawater and filter-feeding organisms such as oysters. This bacterium may cause bloodstream infection and wound infections, certainly in patients with liver cirrhosis.
Biotypes
There are 2 biotypes: classic Vibrio cholerae and V. cholerae biotype El Tor. Biotype El Tor agglutinates chicken erythrocytes and causes lysis of sheep erythrocytes, unlike the classic biotype. The name El Tor originates from the Egyptian town and quarantine camp El Tor in the Southern Sinai dessert, where the bacterium was isolated for the first time in 1905 (during the 6th pandemic) from an asymptomatic Hajj pilgrim from Mecca. The importance of this germ was long disputed (until 1961). At present El Tor has replaced the classic variant in most places, except in the Ganges and Brahmaputra delta. El Tor may also survive longer in the environment, is less dependent on transmission via water and produces more asymptomatic infections (symptomatic/asymptomatic infections = 7/100).
Serotypes
V. cholerae O1 of both biotypes can be subdivided into serotypes according to the structure of the O antigen. If only O antigen A and C are present, the bacterium is known as serotype Inaba. If only A and B are present, the bacterium is known as serotype Ogawa. If A, B and C are present, the name Hikojima is given. Serotype shift seldom occurs (from Ogawa to Inaba and vice versa).
The difference between these serotypes is only of importance for epidemiological studies. For example: in 1991 all cases of cholera in South America were caused by toxin-producing Vibrio cholerae, serogroup O1, biotype El Tor, serotype Inaba. The cholera epidemic in the Rwandan refugees in DRC (July and August 1994) was caused by El Tor, serotype Ogawa. These bacteria were resistant to tetracyclines, cotrimoxazole, chloramphenicol and ampicillin. The outbreak in Haiti in October 2010, 9 months after an earthquake was due to V. cholerae O1, Biotype El Tor.
Flagella, pili and fimbriae
Most motile bacteria move about with structures called flagella (spirochaetes move with the help of axial filaments). Do not confuse active bacterial movement with random Brownian movement. Do not confuse a bacterial flagellum with the flagellum of a eukaryote such as Giardia (cf. also the remark concerning cilia in Balantidium coli). The flagella are too thin (0.2 µm) to be observed with a standard light microscope. The bacterial flagellum carries out a rotating movement. Some bacteria have several flagella. When the flagella rotate anti-clockwise, they form a coherent bundle, so that the bacterium moves in a straight line. On the other hand when the flagella turn clockwise, there is no longer any co-ordination and the bacteria move randomly. By timing the duration of clockwise and anti-clockwise spinning, this mechanism can be used in chemotaxis. The motor is in the membrane and the immediate driving power is not ATP, but a proton gradient. The bacterial flagella must not be confused with fimbriae, thread-like appendages which have no function in movement, but play a part in adherence to cells or tissues (important for virulence). Flagella rotate, fimbriae do not. Pili (singular pilum) are important in conjugation, the bacterial equivalent of sex. These hollow rigid tubes permit DNA transfer between bacteria. F-pili [Fertility] are important in the spread of resistance to antibiotics. Pili may also act as receptors for bacteriophages.
Epidemiology
Epidemics, Pandemics
Cholera has always been endemic in India and Bangladesh, in the huge delta formed by the confluence of the Ganges, Brahmaputra, Jamuna and Meghna rivers. Probably there was no cholera in Europe or America before the 19th century.
Between 1817 and 1923 there were various great pandemics, probably caused by the classic V. cholerae (there is no certainty as to the exact strain). The first pandemic which started in 1817 did not reach Western Europe. In 1829 the bacterium was introduced into the countries around the Persian Gulf via a British army unit stationed in India. From Iran the infection spread to Iraq, Syria, Georgia and Astrakhan (north of the Black Sea). It then travelled towards Odessa, Moscow, Vienna, Warsaw and Hamburg reaching England via the port of Sunderland. The first cases in London were seen in February 1832. The third pandemic merged with the second and was amplified by the miserable conditions during the Crimean war.
When each pandemic began and ended is rather unclear. There was cholera in Belgium in 1832, 1848, 1854, 1859, 1866 and 1892. In 1866, 1 in 100 Belgians died of cholera.
The pathogen was discovered in 1884 by Robert Koch during the fifth pandemic (first work in 1883 in Alexandria, Egypt, confirmation followed by research in India in 1884, with isolation of the bacterium in culture). In fact the bacterium had already been described in 1849 by Pouchet and in 1854 by Filippo Pacini, an Italian physician. However the latter’s work on this was not known outside Italy. The germ theory and in particular the work of Koch were attacked by Pettenkofer. Pettenkofer was a proponent of the “ground water theory” believing that the fermentation of organic matter in the subsoil (“miasma”) released cholera into the air (no transmission from person-to-person) which then infected the most susceptible e.g. those with poor diet, constitution, etc. Both Pettenkorfer and his loyal student Emmerich drank a vial filled with cholera bacteria as proof against Kochs type transmission of V. cholerae. Amazingl, Pettenkofer did not then get cholera, but Emmerich suffered severe diarrhoea for 48 hours.
After the sixth pandemic there was a strange silence for about 40 years, for which no good explanation exists. The seventh pandemic was caused by El Tor. It started in 1961 in Celebes (Sulawesi), Indonesia, reached India in 1964 and Africa in 1970. In 2 years the infection passed through 29 African countries. In 1973 it arrived in the Gulf of Mexico. Early in 1991 the infection spread rapidly in Peru. In 3 weeks there were 30,000 cases. The bacterium then spread further into South America, causing 360,000 cases within the year. In the summer of 1992 a second, less severe outbreak occurred. Nevertheless by August 1992 “only” 5,000 deaths had been reported (from an estimated total of 600,000 cases), thanks to the wide-spread use of rehydration therapies. The case-fatality ratio varied depending on the region. After 1993 the disease assumed an endemic character in several countries, sometimes with local outbreaks. At the end of 1993 the cumulative total amounted to 900,000 cases in three years (1991-1993), with a cumulative mortality of 8,000. According to one hypothesis cholera bacteria infected the marine plankton off the Peruvian coast via the ballast water from a Chinese freighter. The possible role of changes in the nutrient-rich von Humboldt current is still unclear.
Spread of the seventh cholera pandemic
About 80% of the cholera in 1997 occurred in Africa, chiefly in the horn of Africa (118,000 cases were reported officially). The increase in cholera in this region followed heavy rains and flooding (possibly associated with the El Niño weather phenomenon).
Since 1992 V. cholerae O139 is recognised as a cause of a disease which is clinically identical to classic cholera, but which also occurs frequently in adults. Classic cholera in India, on the other hand, is common in children. There is no cross immunity with V. cholerae O1. Bacteria of the O139 serogroup have a polysaccharide capsule (unlike V. cholerae O1), which may explain the increased risk of bloodstream infection.
Cholera O139
After 1992 this new serogroup spread across Bangladesh, India, Pakistan and Southeast Asia. By the end of March 1993 more than 100,000 cases had been reported in Bangladesh. Further spread continued, but somehow diminished again, as the classic form and El Tor took over, reducing the incidence of the new serogroup. The reason is unknown. Therefore it is difficult to make then new Bengalen serogroup responsable for an 8th pandemic. It was observed in India that, after the first spread of V. cholerae O139, new variants (clones) of V. cholerae O1 El Tor once more gained the upper hand.
Cholera also surfaces regularly in Madagascar. From the beginning of December 1999 until the end of February 2000 more than 12,400 cases were reported. The disease can thus certainly not be regarded as an entity which only existed “in the past”.
Recent Epidemics
At the end of 2008 a large cholera outbreak appeared in Zimbabwe. By February 2009, this led to more than 60,000 cases with a mortality of more than 5%, reflecting the general degradation of the nation’s basic infrastructure and the crumbling Zimbabwean health care system. By mid-April 2009 the official count was 96,591 cases with 4,201 deaths.
In 2010, more than 38,000 cases of cholera were identified in Nigeria.
In January 2010, a devastating earthquake hit Haiti, with its epicentre 25 km from Port-au-Prince, the capital. A couple of weeks before Nepalese United Nations peacekeepers arrived in Haiti, a cholera outbreak occurred in Kathmandu, the Nepalese capital. The forces were stationed in Mirebalais, 60 km north-east of Port-au-Prince. Late October 2010, patients with cholera were recognized in some Haitian rural areas. In less than 6 weeks, more than 10,000 cholera cases were identified. The disease quickly spread to the capital, where many people were still living in temporary shelters and tents, without access to safe drinking water or proper sanitary facilities. By January 1, 2011, the Ministry reported 171,304, with a cumulative mortality of 3651. The hospital case fatality rate was too high, and a target of hospital CFR of < 1% should be achievable. A possible epidemiological connection with the Nepalese forces was suspected and created tension between the local population and the UN forces. The current Haitian strain of cholera was identified as a virulent hybrid of the El Tor O1 biotype and the classic type, serotype Ogawa.
Transmission
Cholera is spread by the faecal-oral route, via contaminated water and food. The infectious dose of bacteria required to cause clinical disease varies according to the mode of transmission and varies according to bacterial strain, with hyper-infective strains occurring immediately after gut passage. In people with normal gastric function and if ingested with water, the infectious dose is one thousand to one million vibrios. When ingested with food, it is lower about one hundred to ten thousand vibrios. The low pH of stomach acid kills most vibrios. When a person uses antacids, proton pump inhibitors or ranitidine, a lower infectious dose is required to trigger infection. The same applies to chronic atrophic gastritis and status post-gastrectomy. Asymptomatic infections are common, especially in case of El Tor. People excrete bacteria for about 10 days. This is sufficient time to ensure continued contamination of the environment. Chronic carriers are very rare, but occur, sometimes with vibrios lodging in in the biliary tract.
In third world countries many people have no chlorinated, filtered, treated, pure drinking water. The lack of good toilets and sewers leads to contamination of the surface or ground water. Too often untreated sewage water is still poured into surface water. Sometimes sewage pipes and drinking water pipes are laid in the same trench, which may result in contamination if there are leaks or greatly varying water pressures in the pipes. If drinking water is contaminated in this way, bacteriological checks of the drinking water when it leaves the pumping station will not show anything amiss. In houses, drinking water containers with a wide openings often become contaminated, because people are inclined to scoop up water in their (dirty) hands. Containers with a small spout, from which water must be poured are safer.
There is also direct transmission from person to person, but it is rare. The number of bacteria on dirty hands is usually lower than the minimum infectious dose necessary for direct transmission. Health workers who respect basic hygiene are at extremely low risk. Filter feeders such as mussels or oysters (especially in estuaria) concentrate the bacteria in their bodies. When the organisms adhere to food particles (e.g. the chitin of crustaceans) and in the case of hypochlorhydria, lack of gastric acid due to gastric surgery, antacids, anti-ulcer drugs or atrophic gastritis, the number of organisms needed to trigger infection is much smaller. Food may be infected by dirty hands during or after preparation. The bacteria can survive and reproduce in food such as cereals, rice or lentils and crustaceans. This intermediate replication step is very important. If someone dies of cholera and a meal is made for the mourners at the funeral by the persons who have washed the corpse, the risk of further transmission is very real. The bacteria are very sensitive to drying out, sunlight and acid. Meals which contain acid e.g. tomatoes and/or lemon, are much less dangerous than neutral or alkaline meals. Vegetables and fruit on the market are often sprayed with water to make them appear fresher and more attractive. If this is done with contaminated water, transmission may occur.
Historical note: John Snow and contaminated water
In the first half of the nineteenth century a cholera epidemic occurred in London. In 1848, a cholera outbreak started which would kill more than 14,000 people in London. Another outbreak in 1853 killed more than 10,000 people. The physician John Snow, already well known in 1853 as anaesthetist to Queen Victoria. Dr Snow had also a special interest in cholera. In 1854 he examined the various families presenting cases and calculated that the mortality in the houses that were supplied with water by the Southwark and Vauxhall Water Company was 31/1000 houses. This was 8.5 times higher than in houses supplied by the rival Lambeth Company. Although neither of the two companies offered purified water, the first company took its water from the Thames near London Bridge, downstream from the city sewage outlets while the second company pumped its water upstream from the city at Thames Ditton. In 1849 there had been no difference in mortality between the families that received water from Lambeth or Southwark. Before 1851 the Lambeth Company drew its water from a highly contaminated stretch of the Thames near Hungerford Market. It was this spectacular change (1849 compared to 1854) which made Snow conclude that contaminated water had a causal connection with cholera. Although both companies delivered water in the same streets, the water used in any particular house could be identified by its salt content (London Bridge is closer to the sea and its water is saltier than that at Thames Ditton). Adding silver nitrate leads to precipitation of silver chloride, proportionate to the amount of salt in the water. This was the basis of a simple test that could be carried out in every house.
Similar findings were made in Hamburg in 1890. The incidence of cholera was 34/1000 in Hamburg, where the drinking water was drawn from the river Elbe, and 3.9/1000 for the surrounding areas where other sources were used. In Altona, to the west and downstream from Hamburg, contaminated water was also taken from the Elbe, but there was less cholera. How could this be explained? If anything, more cholera would be expected in Altona. The difference was that in Altona the water was first filtered slowly through sand before being supplied for consumption. These observations led to attempts to provide cities with clean drinking water and to construct adequate sewers. Cholera was the first disease for which surveillance was set up and because of this the disease still has code number 001 in the international classification list of diseases.
Reservoir
Humans are the only vertebrate hosts. Vibrio cholerae can survive long-term and probably permanently in brackish water, especially if there is a neutral or slightly alkaline pH and the water contains minerals and organic material. The bacteria are concentrated in phytoplankton (certain algae) and zooplankton which live in this water. Among the latter, copepods, a group of crustaceans, are important.
Cholera is clearly seasonal. A chronic aquatic reservoir is likely and this might be independent of continuous human faecal pollution. V. cholerae excreted by humans can be cultured in the laboratory. These bacteria however may assume a living form which cannot be cultured in vitro and which do not multiply in the environment However, those bacteria are not dead as they are known to multiply when instilled in a rabbit’s ileum. They are called ‘viable but not culturable’. It may revert to a replicating form in its natural environment when there are favourable environmental factors and this has important epidemiological implications. The living, but non-reproducing form of V. cholerae can probably cause disease. Traditional culture methods for tracing V. cholerae in water miss these “dormant” bacteria. Tests based on fluorescent antibodies may offer a practical solution, as they stain both dormant and active bacteria.
Physiopathology
The bacterium multiplies in the small intestine, where it adheres to the mucosal brush border. The bacterium is not invasive, in other words it does not penetrate the intestinal wall or pass into the blood. It excretes a very powerful toxin which causes active fluid secretion towards the lumen. This fluid is isotonic, ion-rich and protein free. There are no intestinal ulcerations and the faeces do not contain blood. There is little if any fever. There is no tenesmus. The faeces contain significant amounts of sodium, potassium and bicarbonate. Because of this the intestinal content is slightly alkaline (V. cholerae thrives best in a slightly alkaline environment and the bacteria are therefore producing the conditions which are optimal for their own survival). The loss of large amounts of alkaline faeces results in metabolic acidosis. People with blood group O have an equal risk of infection but are at a significantly higher risk of clinically severe cholera if they become infected. The reason is unknown.
Hypervirulent and hyperinfective strains play an important role in epidemics. Passage of Vibrio cholerae through the gastrointestinal tract results in a short-lived, hyperinfectious state of the organism that decays in a matter of hours into a state of lower infectiousness. Such strains have a much lower ID50 (the number of micro-organisms that will disease 50% of a population in normal conditions = measurement for virulence) than strains occurring in natural water reservoirs. The classic strain is associated with more severe illness. Faecal excretion of V. cholerae for up to two weeks has been documented and occasional asymptomatic carriers occur. Asymptomatic patients typically shed bacteria in their stools at about 1000 V. cholerae bacteria per gram of stool, which is a low level of shedding, compared with the 100 million bacteria per gram in case of ricewater stools.
Toxins
Vibrio cholerae produces several toxins: cholera toxin (Ctx), the zona occludens toxin (Zot) and the accessory cholera enterotoxin (Ace). The role of the two latter toxins is not entirely clear. The Ctx enterotoxin of V. cholerae consists of 2 parts: A and B, where A stands for active and B for binding. They stimulate adenylate cyclase. Adenylate cyclase increases intracellular cyclic-AMP, which inhibits salt absorption by the microvilli and promotes active chloride excretion by the crypt cells. Water and potassium bicarbonate passively follow the chloride. In the end there is an overall water loss to the intestinal lumen. Fluid loss originates in the duodenum and upper jejunum, the ileum is less affected. The colon is insensitive to the toxin and cannot absorb the large amount of fluid quickly enough. Catastrophic diarrhoea follows.
Cholera toxin
Part A is a monomer, while part B consist of 5 identical subunits (a pentamer). The polypeptides of part B bind to a receptor (Gm1 ganglioside, a glycolipid) on the epithelium of the small intestine, after which part A can penetrate the cell. Part A binds covalently to an intracellular protein (Gs-protein; s for stimulatory) which irreversibly activates it, leading to the persistent stimulation of another intracellular enzyme, adenylate cyclase.
The toxic A-subunit also has other effects such as disturbing the expression of some genes, increasing inflammatory cytokines and inhibiting antigen presentation by macrophages. On the other hand, the B-subunits of cholera toxin have anti-inflammatory properties. These are under intense study at present for possible therapeutic use in immune abnormalities. While cholera toxin adheres to the intestinal villus cells and disables the cellular saltwater pumps, the Zot toxin loosens the junctions that binds intestinal epithelial cells together. This contributes to the loss of water to the intestinal lumen.
The in-vivo detailed mechanism is probably more complicated. Cholera toxin also stimulates the nervous system in the intestinal wall, the myenteric plexus. This results in the release of 5-hydroxytryptamine (serotonin) from the enterochromaffin cells, leading to in additional fluid loss to the lumen. Granisetron, a 5-HT3 receptor blocker, partially reverses this effect. More research is needed to determine the role of this mechanism in the physiopathology.
Clinical aspects
The incubation period is brief: sometimes only hours, more commonly 1 to 5 days (average 2 days). It is one of the few infectious diseases where -in case of a very severe infection- you can be well in the morning and dead by sunset the same day. Asymptomatic infections are common (about 93%), but chronic carriers are very rare. Sometimes there is an initial transient fever (more seen in children). Massive watery diarrhoea starts suddenly. The faeces very rapidly look like water in which rice has been boiled: watery with flakes of mucus. The faeces have a fish-like smell. The volume of faeces may rise to 500 ml per hour. Vomiting is common, but abdominal cramps are unusual. The onset of thirst, oliguria or anuria and weakness is rapid. In a short time the patient develops severe dehydration and can die within 24 hours. In other cases the diarrhoea is less severe, especially with infections with El Tor. As the patient’s condition deteriorates, hoarseness of the voice and temporary deafness are often observed. Children with severe cholera may present with drowsiness or coma.
‘Cholera feces which look like ”rice water”. Copyright Alexander von Humboldt Institute, Peru’
The signs of dehydration are thirst, dry mouth and lips (if the patient has not vomited recently), hollow eyes and sunken fontanel in children. The skin turgor diminishes. The skin becomes wrinkled (washerwoman’s hands). Often, the voice becomes weak and hoarse, the pulse quickens and is difficult to feel. The radial pulse might be impossible to detect. Blood pressure falls. There is little or no urine production (prerenal failure). Respiration becomes faster due to metabolic acidosis secondary to loss of bicarbonate in the faeces (bicarbonate is alkaline). This acidosis causes vomiting and muscle cramps. There is also significant potassium loss in the faeces. If rehydration is carried out using fluid without potassium, severe hypokalaemia may result. Nevertheless, quite often normokalaemia is found, together with an increased anion gap. The increase in anions (= negative ions) is multifactorial due to the hyperproteinaemia (hemoconcentration), hyperphosphataemia (internal shifts and renal failure) and lactate acidosis (shock). Ketones play little if any role. An elevated hematocrit (hemoconcentration) can be found in in nonanemic patients, as can neutrophil leukocytosis in severe cases.
The mortality from classic cholera may reach 50 %, but can be brought down to < 1 % with correct therapy. Mortality is chiefly due to dehydration with kidney failure, hypokalaemia, hypoglycaemia and aspiration pneumonia during vomiting.
Diagnosis
Cholera should be suspected in acute massive rice-water diarrhoea, certainly if there have been several cases in a short time (epidemic). The clinical picture of severe cholera is so spectacular that differential diagnosis does not present many difficulties. Milder cholera may be similar to other forms of gastro-enteritis (but not to dysentery). A child above the age of five years who develops acute dehydration, or dies as the result of acute diarrhoea, is always suggestive for cholera.
The vibrios are very small and can best be seen in a fresh faecal specimen with the help of dark field microscopy. There is characteristic motility (“star shooting”) which stops immediately after adding anti-O1 antiserum. This does not give any information on possible toxin production.
Confirmation is best made via a bacteriological culture.
Culture of Vibrio cholerae
Culturing should preferably be on a special medium in a bacteriology lab, e.g. TCBS-agar [=Thiosulphate-Citrate-Bile salts-Sucrose], polymyxin mannose tellurite agar (PMT) or another selective medium. TCBS agar is green before inoculation; sucrose-fermenting organisms such as V. cholerae turn it yellow. TCBS agar is important for rapid isolation and identification, but V. cholerae also grows on routine agar media. For routine media, large numbers of bacteria per gram of stool should be present to allow detection. Patients or carriers with low burden of bacteria will be missed with routine culture media. Overgrowth by normal faecal flora limits recovery of colonies. In order to identify the serogroup and the serotype one subsequently finds out to which antibodies (antiserum) the colonies obtained exhibit an agglutination reaction. It is also possible to find out whether the vibrios are toxicogenic (produce toxin), e.g. by a PCR variant called a loop-mediated isothermal amplification (LAMP) assay. Definitive identification is made in a reference laboratory.
Specimens may be transported in a transport medium, e.g. Cary-Blair. This is a kind of mild alkaline buffered gelatine in seawater with low redox potential in which the bacteria will survive for 4 weeks. If it is not available, a filter paper can be soaked with faeces and transported in an airtight bag to a well-equipped laboratory. A sample treated in this way remains usable for 1 week, but the recommendation is “the faster the analysis, the more reliable”. Blotting paper, soaked with liquid faeces and if possible placed in a 1% saline solution, can be kept for several weeks at 37°C (not in the freezer). This is useful if there are initial transport problems. Nevertheless it is better to have a fresh faecal specimen. For specimens from the environment or from food, in which the number of bacteria is much lower than in faeces, enrichment is necessary. The specimen can be incubated for 8 hours in alkaline peptone water, after which a TCBS agar is used.
About 10 days after infection with V. cholerae O1 the patient produces vibrocidal antibodies. They start diminishing after only one month and disappear within the year. Antibodies against cholera toxin are produced more slowly and remain for years. However these cross-react with enterotoxin produced by ETEC bacteria [enterotoxic Escherichia coli]. The immune response to V. cholerae O139 is not well understood. The detection of antibodies is not important for the urgent care of the individual patient but does permit retrospective diagnosis.
Other Vibrios
Sometimes other Vibrio species are responsible for diarrhea, e.g. Vibrio cholerae non-O1, V. parahaemolyticus, V. hollisae, V. minicus and V. fluvialis. Our knowledge of these latter bacteria is clearly insufficient. Vibrio vulnificus is an aggressive species present in seawater and filter-feeding organisms such as oysters. This bacterium may cause bloodstream infection and wound infections, certainly in patients with liver cirrhosis or otherwise immune depressed.
Treatment
Rehydration is essential and must be instituted as soon as possible. Two phases are distinguished. First it is important to replenish what has been lost in the previous hours or days. Then one must compensate the persistent fluid loss (e.g. the amount of fluid that is lost every hour). In mild cholera without vomiting oral rehydration may suffice. In severe forms IV fluids should be administered.
There are several possible compositions of rehydration fluids. Solutions containing salt, sugar, potassium and bicarbonate are recommended. Lactate is also good because it is converted in the body to bicarbonate. In cholera it is preferable to use Ringer’s lactate (= Hartmann’s solution). Normal physiological saline is second choice because it does not correct the acidosis nor does it contain potassium. Severe hypokalaemia may occur, with cardiac arrhythmias, kidney damage, paralytic ileus and significant muscle weakness with reduced or absent tendon reflexes. Dextrose (= glucose) 5 % without electrolytes is not advised as a rehydration fluid. A reminder: 1 gr KCl = 13 mEq KCl. So: Hartmann = Ringer’s lactate >Ringer> physiological saline>>> not glucose infusion if there is an alternative.
In severe cholera (fluid loss > 10 % of weight) the missing fluids should be administered quickly, e.g. 6 litres over 4 hours for a patient weighing 60 kg. The first 3 litres may each be administered in 10 minute boluses (total therefore 30 minutes). After administration of the lost volume, losses are compensated with further IV and/or PO fluids (faeces volume + urine volume + 500 ml). Vomiting may make oral administration of fluids difficult. Generally a total of 6 to 10 litres per patient is necessary. When patients start to drink and stop vomiting, it is advised to leave IV lines in place for a while until you are sure rehydration will not pose any more problems.
Cholera epidemic in Congo. Refugees live in very poor conditions. Photo courtesy Els De TemmermanCholera bed used by AIDS patients with chronic diarrhoea. With special thanks to Prof. Dr R. Colebunders. Copyright ITM
Special cholera beds are useful: they have a central opening to allow the liquid faeces to pass through, and they can be collected in a bucket. This makes it possible to quickly determine the amount of fluid loss. During an epidemic people who can still hold themselves upright can sit on a bucket and try to drink as much ORS [oral rehydration solution] as possible. Children quickly develop convulsions and coma. It is important that hypoglycaemia should be considered. For an adult 50 ml of a 50% glucose solution is given IV, for a child 2-4 ml/kg 25% glucose or 10 ml/kg of a 10% glucose solution.
Antibiotics may be useful because they reduce the duration and thus the total volume of the diarrhoea and may therefore reduce the need for rehydration fluids. On the other hand, they are not essential (given the non-invasive character of the infection) and resistance often occurs. At present a single dose of azithomycin 1 gram (child 20 mg/kg) or a single dose of doxycycline 300 mg (child 4 mg/kg) are possible treatments. Doxcycline is usually contraindicated in pregnant women and children under 8 years. However, the administration of a single dose should not provoke major adverse effects. Single-dose ciprofloxacin may also be effective. V. cholerae O139 is often resistant to cotrimoxazole (sulphamethoxazole-trimethoprim). There is insufficient data examining the effect of antibiotics on secondary transmission of cholera. However in published studies to date antibiotics have not been shown to decrease secondary transmission of cholera within households. Anti-peristaltic drugs such as loperamide may cause accumulation of fluid in the intestinal lumen with unfavourable consequences and should be avoided.
Prevention
In the industrialised world a patient with cholera will remain a sporadic case. In developing countries one case can lead to several secondary cases. Therefore ‘enteric contact precautions’ are essential in health care settings, focusing on very strict hand hygiene and thorough environmental cleaning and disinfection. The contamination of clothing and bedding is unavoidable. Boiling in water for five minutes is sufficient for disinfection. Mattresses and blankets can be dried in the sun. It is better to do this before washing them, to prevent infection of the washing area.
After surviving cholera a patient is probably immune for homologous biotypes for more than 3 years. There is some controversy: infection with the classic biotype seems to protect against recurrent infection by either biotype, but El Tor does not. No cross-immunity between V. cholerae O1 and V. cholerae O139 is seen, although they produce the same toxin. Immunity relies on antibodies in the intestinal lumen (the bacteria are not invasive). Systemic vibriocidal as well as anti-toxin antibodies develop during illness. Babies which are being breast-fed receive protective antibodies in their mother’s milk.
Vaccination
Former parenteral vaccination with dead V. cholerae bacteria (IM administration) did not lead to sufficient formation of protective antibodies in the intestinal lumen. Only about 50-65% of people living in endemic areas were protected for 3-6 months. The IM vaccine was associated with local reactions in 50% and systemic reactions (fever, malaise) in 10-30%. Advice to vaccinate with this type of vaccine was discontinued in 1972 by the WHO [World Health Organisation]. Parenteral vaccination, mass chemoprophylaxis and “cordon sanitaire” (= restrictions on travel and trade) are not effective in preventing or limiting outbreaks. A newly developed oral cholera vaccine is based on a killed whole cell cholera vaccine combined with the recombinant B subunit of cholera toxin (Dukoral®). The vaccine contains 1 mg of recombinant B subunit, as well as 25 x 109 bacteria each of V. cholerae O1 classic Inaba, V. cholerae O1 classic Ogawa, V. cholerae O1 El Tor Inaba (heat-inactivated), V. cholerae O1 El Tor Inaba (formaline inactivated). Dukoral does not contain the A subunit of cholera toxin and therefore, no pathogenic toxin is present. Two to three doses need to be given. Two recent studies showed an effectiveness of 86% and 40% respectively; the latter study indicating a 63% protection against severely dehydrating cholera episodes. Lower levels of protection continue for 3 years. Protection wanes rapidly in young children. A herd immunity effect is expected in areas where vaccine coverage is more than 50%. Because the risk of cholera for most travellers is extremely low, vaccination should be considered only for those working in relief or refugee settings or for those who will be travelling in cholera-epidemic areas and who will be unable to obtain prompt medical care. WHO recommends that current available cholera vaccines be used as complements to traditional control and preventive measures in areas where the disease is endemic and should be considered in areas at risk for outbreaks.
Mass (antibiotic) chemoprophylaxis is not effective because (1) the infection spreads faster than the organisation of drug distribution, (2) the effect of a drug only lasts 2 days, after which re-infection may occur, (3) the whole population needs to be treated simultaneously and people should then be isolated and (4) it is difficult to convince asymptomatic people to take a drug. In addition, the selection of highly resistant strains has been observed in settings using mass-administration of antibiotics.
Correct eating and drinking habits, safe stool habits and personal hygiene are the most effective means for individuals to limit their risk of cholera. Improved sanitation is the pre-eminent method of eliminating cholera and many other faecal-orally transmitted infections. This is directly linked to the degree of poverty in a region. Boiling drinking water is often difficult since fuel may be scarce and expensive. Since a significant proportion of Vibrio cholerae can adhere to plankton, the drinking water can be filtered through a fine cloth, which removes both plankton and a lot of bacteria in a single operation. This is of course less effective than obtaining water from a clean pipe or pump but it is cheaper. Chlorination of drinking water may be important (piped water or via water trucks). This is difficult to accomplish in rural areas. Chlorination is much less effective if the water is turbid due to organic debris.
Eating raw fish, shellfish (e.g. oysters, mussels) and crustaceans (such as crabs, shrimps) should be avoided. Washing hands is important for transmission control within a household. Infected faeces should not be disposed of in a poorly functioning drain (hospital: e.g. in pit with unslaked lime = CaO). When large groups of people come together (funerals, festivals, etc.) there should be latrines with facilities for washing hands and plenty of soap.
An attempt must be made to trace the source of small, local outbreaks (see Historical note on John Snow). Contaminated water is the chief suspect in a sudden local epidemic, while in isolated cases the cause should be sought in contaminated food. This is of course not an absolute rule. Food cooked by street vendors and in restaurants poses specific problems. Flies probably play an underestimated part in transmission, but their numbers also reflect the sanitary conditions in a region.
The following points should be emphasised during information campaigns:
Drink only clean water (boiled or chlorinated)
Cook food completely and eat it while it is hot
Avoid uncooked food, unless it can be peeled
Wash hands after a bowel movement
Wash hands before preparing food
Wash hands before eating
Correct use of a good latrine (also for children)
With correct treatment cholera is rarely fatal
If cholera is suspected medical help should be sought immediately
In diarrhoea, give plenty of fluids (e.g. ORS)
Cholera vaccines should be used as complements to traditional control and preventive measures in areas where the disease is endemic and should be considered in areas at risk for outbreaks
In case of an epidemic, it is important to have a large stock of IV rehydration fluid available as well as the means of preparing large amounts of oral rehydration fluid. Normally such buffer stocks should be stored at various strategic points. The stocks for cholera treatment should not be segregated in storage, but should be rotated during normal use to avoid stock expiring. As soon as an epidemic is suspected, use as much oral rehydration as possible first so that stocks of the IV solutions last as long as possible. Cholera beds should be made ready. In a normal epidemic an attack rate of 0.2% can be taken as a rule of thumb (i.e. 200 cases can be expected in a population of 100,000). This is useful for estimating the size of stocks that will be needed. Sometimes the attack ratios are higher (e.g. the Rwanda-Zaire border in 1994).
