CHAPTER 188d DENGUE AND DENGUE HEMORRHAGIC FEVER Scott B.

Halstead Dengue fever is an acute febrile illness syndrome caused by several arthropod-borne viruses and characterized by biphasic fever, myalgia or arthralgia, rash, leukopenia, and lymphadenopathy. Synonyms are dengue and breakbone fever. Dengue hemorrhagic fever, a febrile disease caused by dengue viruses, is characterized by abnormalities in hemostasis and by leakage of fluid and protein from capillaries, which in severe cases results in shock (dengue shock syndrome). It is thought to have an immunopathologic basis. Synonyms are hemorrhagic dengue; acute infectious thrombocytopenic purpura; and Philippine, Thai, and Singapore hemorrhagic fever. The dengue subgroup is composed of four antigenically distinct members.[21,][48,][56,][96] From 1956, according to reports received by the World Health Organization (WHO), dengue viruses were thought to be responsible for more than 5,500,000 hospital admissions and 74,000 deaths in Southeast Asia, southern China, India, Sri Lanka, Pakistan, Cuba, Venezuela, Colombia, Guyana, Brazil, Puerto Rico, and Central America, mostly in vital, healthy children. In tropical Asian countries, dengue is among the 10 leading causes of death in children ranging in age from 1 to 15 years.[46] The first outbreak that resembled a disease now recognized as dengue fever was described by Benjamin Rush in Philadelphia, Pennsylvania, in 1780.[11,][103] Epidemics probably caused by dengue were common from the 18th to the 20th centuries in inhabitants of the Atlantic coastal areas of the United States and South America, the Caribbean islands, and the Mississippi basin. [103] Dengue viruses almost certainly were the cause of the 5- and 7-day fevers that occurred in European colonists in tropical Asia.[11] Similar epidemics occurred commonly in settlers in tropical Australia, where in 1905 Aedes aegypti was identified as a dengue vector by Bancroft. [72] Ashburn and Craig found the etiologic agent in human blood and showed that it could pass through a diatomaceous earth filter.[103] An intrinsic incubation period of 3 to 8 days in humans, an extrinsic incubation period of 8 to 11 days in mosquitoes, immunity in people and monkeys, and the nonsusceptibility of most domestic animals were demonstrated in the classic studies of Siler and Simmons and their coworkers[103,][104] between 1924 and 1930. When dengue viruses were isolated in laboratory mice in 1943 and 1944, the modern era of dengue research began.[57,][98] Two strains from Hawaii and Papua New Guinea failed to cross-protect humans. From this experiment, researchers recognized the existence of at least two different dengue viruses; they were named dengue virus type 1 and type 2.[97,][98] During most of the pre-virologic era, dengue viruses were thought to be the cause of a generally benign, self-limited febrile exanthem. However, death, shock, and severe hemorrhagic manifestations accompanied the classic dengue fever outbreaks in Australia in 1897 and thereafter for 15 years. Similar phenomena were recorded in Greece in 1928 and in Formosa in 1931.[35] This new syndrome was recognized again in Manila in 1954. It was called Philippine hemorrhagic fever because of a resemblance to the epidemic hemorrhagic fever then occurring in United Nations troops in the Korean Peninsula.[48] In 1956, Philippine hemorrhagic fever was associated with dengue when types 3 and 4 were recovered.[48] It now has become endemic throughout tropical Asia.[35,][43,][46] Since 1967, the terms dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS) have come into general use.[18] In 1981, Cuba reported a severe outbreak involving more than 116,000 patients hospitalized within 3 months, 10,000 of whom had DHF/DSS.[63] In 1986, an epidemic of DHF/DSS occurred on Hainan Island, China[80]; in 1988, the Maldive Islands, Sri Lanka, and India were involved. [105,][114] From about 1987, DHF/DSS outbreaks have occurred in Guyana, Venezuela, Brazil, Colombia, Central America, and, on a lesser scale, Puerto Rico.[46,][83] By epidemiologic criteria, dengue viruses are arthropodborne (arboviruses) because they are transmitted biologically by various members of the genus Stegomyia.[72,][75,][103,][104] Gene

structure, replicative strategy, and antigenic relatedness place the dengue viruses in the family Flaviviridae.[56,][115] At present, 68 members of the flavivirus family have been identified; 29 of them are established as human pathogens.[56] Cross-comparisons by plaque reduction neutralization tests have shown dengue viruses to be an antigenic subgroup with little relationship to other flaviviruses.[21] In addition to their antigenic relatedness and their ability to be transmitted by Stegomyia, each type of dengue virus produces a closely similar clinical syndrome in susceptible human beings.[36,][97,][103,][104] Dengue virions are spherical particles approximately 50 nm in diameter. The central core, approximately 25 nm in diameter, has icosahedral symmetry and contains a single plus strand of RNA. Dengue RNA consists of approximately 11,000 nucleotides coding from the 5 end for core, premembrane, envelope, and five nonstructural proteins.[34] Large amounts of nonstructural protein 1 (NS1) are released from infected cells.[69] The envelope, which is studded with poorly resolved projections, is composed of many replicates of the envelope protein (54 kd) embedded in a lipid bilayer. When it is assembled on the virion, the envelope protein bears some epitopes unique to serotypes. Antibodies to these epitopes neutralize by hindering viral attachment or entry into cells. Other epitopes are shared between dengue viruses (dengue subgroup antigens) and other flaviviruses (group antigens). Four clearly defined types exist, as determined by plaque reduction neutralization tests using antibodies raised by infection of monkeys or fluorescent antibody tests using monoclonal antibodies raised in mice.[21,][94] Presumably, the cell attachment receptor differs for each dengue serotype and is blocked by neutralizing antibody.[49] The four types have distinctive genetic structures.[68,][69,][86,][87,][115,][118] Phylogenetic studies suggest that human dengue viruses diverged from four zoonotic dengue types relatively recently, whereas the four zoonotic types evolved from a common ancestor in the more remote past.[115] Different dengue strains cluster in groups that differ genetically (genotypes).[15,][16,][68,][69,][86,][110] Genotypes consist of viruses of similar genetic structure that usually circulate within one geographic area.[16,][69,] [110] Genotyping can be used to trace the movement of dengue viruses between the continents. Of particular interest, a strain of dengue type 2 isolated in Jamaica in 1981 and the dengue 2 viruses recovered from 1981 Cuban DHF patients belong to a Southeast Asian genotype.[86,] [100] The most sharply divergent genetic differences are among human viruses and strains from Asian and African zoonotic cycles.[86,][115] Dengue virus can be grown in 1- to 2-day-old mice or hamsters by intracerebral inoculation or in various mosquitoes by oral or parenteral inoculation. High mouse-passaged virus grows and produces deaths in weanling mice. Various tissue cultures of vertebrate and invertebrate origin support dengue virus growth in vitro, as reviewed in the following section. TRANSMISSION A. aegypti, a crepuscular daytime-biting mosquito, is the principal vector. All four virus types have been recovered from naturally infected A. aegypti.[35,][42] In most tropical areas, A. aegypti is highly domesticated and breeds in water stored for drinking, washing, or bathing or in any container collecting fresh water. Dengue viruses also have been recovered from naturally infected Aedes albopictus, which breeds outdoors in vegetation.[29,][35,][43,][75] Outbreaks in the Pacific area have been attributed to Aedes scutellaris and Aedes polynesiensis. In urban areas, transmission of dengue may be explosive and involve as much as 70 to 80 percent of the population.[103] Because A. aegypti has a limited flight range, spread of virus is mainly by mobile viremic human beings. Dengue viruses replicate in the gut, brain, and salivary glands of infected mosquitoes without apparent harm to adult mosquitoes.[90] Mosquitoes are infectious for a lifetime and as long as 70 days in experimental circumstances.[104] Because female mosquitoes take repeated blood meals, long-lived female mosquitoes have great potency as vectors. Several species of Stegomyia and Toxorhynchites are infected readily by intrathoracic inoculation, although the

threshold of infection by oral feeding is higher.[89,][90] A. aegypti and Culex quinquefasciatus can transmit dengue mechanically by interrupted feeding.[46,][104] The contribution of mechanical feeding to the spread of dengue virus during epidemics has never been measured. Because of the skittishness of A. aegypti and its habit of feeding during the day when its intended victim is awake and often moving, interrupted feeding with simultaneous transmission to multiple hosts within a household must be a common occurrence. A. aegypti preferentially feeds on people and hence is most abundant in and around human habitations. The mosquito breeds in clean water. Biting activity is reduced at temperatures below a wet bulb temperature of 14.[23] Transmission of dengue in temperate countries is interrupted during winter weather, and dengue has not established itself endemically at latitudes above 25 degrees north or south. Breeding sites may be provided by humans through living habits, as in Thailand, where water is stored in and around homes in large earthenware jars.[35,][42] In contrast, A. aegypti is not abundant in some parts of India because only small amounts of water are brought to homes from village wells for immediate use. Water in flower vases, household offerings, ant traps, coconut husks, tin cans, and rubber tires may supply breeding sites for A. aegypti.[29,][42] In the tropics, outbreaks of dengue generally coincide with the monsoon season. Eggs, which resist desiccation, are deposited inside water containers above the water line.[42] With the beginning of monsoon rains, a large number of eggs laid outdoors are hatched. Indoor populations do not show seasonal change. Temperature is important in controlling viral transmission. Evidence indicates that the extrinsic incubation period shortens with increasing mean temperatures; mosquito biting rates increase with increased temperature and relative humidity.[102] In sylvan settings, dengue virus has been isolated from three subgenera of Aedes, namely, Stegomyia, Diceromyia, and Finlaya, some in circumstances suggesting the occurrence of transovarial transmission. This phenomenon has been demonstrated experimentally, but its contribution to maintenance of virus in a habitat is unknown.[91] EPIDEMIOLOGY HOST RANGE Inoculation of strains of dengue with known human pathogenicity does not produce demonstrable infection in adult chickens, lizards, guinea pigs, rabbits, hamsters, or cotton rats.[101,][104] Subhuman primates generally are susceptible to infection by dengue viruses. Numerous species belonging to Macaca, Cercopithecus, Cercocebus, Papio, Hylobates, and Pan can be infected by the bite of virus-infected mosquitoes or by injection of infectious virus preparations.[38,][101,] [104] Infection is essentially asymptomatic. Viremia occurs at levels sufficient to infect mosquitoes. Simmons and colleagues[104] were the first to note that wild-caught Macaca philippinensis resisted dengue infection whereas Macaca fuscatus (Japanese macaque) was susceptible. Work by Rudnick[93] in Malaysia has revealed a jungle cycle of dengue transmission involving canopy-feeding monkeys and Aedes niveus, a species that feeds on both monkeys and humans. Although the existence of a jungle dengue cycle in the Malaysian rain forest has been documented, the full geographic range of the subhuman primate zoonotic reservoir is not known. In the 1980s and 1990s, extensive epizootics of dengue virus type 2 involved subhuman primates over wide areas of West Africa.[88] Genetic and epidemiologic studies have shown that urban human dengue and jungle monkey dengue are relatively compartmentalized.[86,][115] Urban dengue is vectored by anthropophilic mosquitoes, and the virus travels along routes of transportation. A. aegypti and susceptible humans are so abundant and so widespread that should dengue viruses be exchanged between humans and monkeys, detection would be extremely difficult. If, in the future, urban dengue is eliminated but vector mosquito populations are unabated, dengue viruses may once again be introduced from jungle cycles. GEOGRAPHIC DISTRIBUTION

Outbreaks of dengue fever have been documented on every continent except Antarctica.[35] Evidence suggests that human dengue may have originated from enzootic or endemic foci in tropical Asia.[37,][75] The probable spread of A. aegypti during historical times from Africa throughout the world provided an ecologic niche quickly occupied by several human viral pathogens: yellow fever, chikungunya, and the dengue viruses. During the 18th and 19th centuries, epidemics occurred in newly settled lands, largely because of the necessity for storage of domestic water in frontier areas. Isolated shipboard or garrison outbreaks often confined to nonindigenous settlers or visitors were reported in Africa, the Indian subcontinent, and Southeast Asia.[11,][72,][103] During World War II, dengue virus infections occurred commonly in combatants of the Pacific War and spread to staging areas not normally infected: Japan, Hawaii, and Polynesia.[38,][46,][97] DHF-like disease was described clinically in Thailand beginning in 1950 and in the Philippines from 1953; cases were confirmed etiologically as dengue in 1958 and 1956, respectively.[48] DHF first was described in Singapore and Malaysia in 1962, Vietnam in 1963, India in 1963, Ceylon (Sri Lanka) in 1965, Indonesia in 1969, Burma (Myanmar) in 1970, China in 1985, and Kampuchea and Laos from about 1985; major outbreaks have occurred in Sri Lanka and India since 1988, in French Polynesia since 1990, in Pakistan since 1998, and in Bangladesh since 1999.[1,][26,][37,][38,][80,][105,][114,][119] DHF has occurred at consistently high endemicity in Thailand, Burma, Vietnam, and Indonesia.[43] In Thailand, it is the third ranking cause of hospitalization and death in children. Intermittent epidemics have involved Malaysia and the Philippines. The largest epidemic in history occurred in Southeast Asia in 1998, when more than 490,000 hospitalizations and 4000 deaths were reported to the WHO. During the past 20 years, major epidemics of all four dengue serotypes have occurred on several Pacific islands.[29,][37,][38,][46] American genotype dengue 2 virus circulated in the American hemisphere for hundreds of years and was the only virus to survive an intensive period of control of A. aegypti designed to eradicate urban yellow fever.[29] In 1963, an Asian strain dengue 3 was introduced into the Caribbean, dying out in the 1980s. Dengue viruses became fully established in the hemisphere after the introduction of a Southeast Asian dengue 1 in 1977.[22,][83,][86] Dengue 1 spread rapidly and now is endemic on the larger Caribbean islands and in Mexico, coastal Central America, and the tropical areas of Guyana, Venezuela, Colombia, Ecuador, Peru, and Brazil.[37,][46,][116] A sharp dengue virus type 2 epidemic in Cuba in 1981 involving 116,000 hospitalizations in a 3-month period led to island-wide control of A. aegypti and apparent eradication of the virus.[63,][82] In 1981, an Asian dengue 4 was introduced into the Caribbean, spread widely, and now is endemic throughout the Caribbean basin. In 1986 and 1987, dengue virus type 1 spread through most of coastal Brazil and from there to Paraguay and to Peru and Ecuador.[22,][29,][83] In 1990, more than 9000 dengue cases were reported from Venezuela; 2600 of them were classified as DHF, and 74 deaths were associated with the epidemic.[83] Dengue virus types 1, 2, and 4 were isolated.[83] Shortly thereafter, DHF/DSS caused by dengue type 2 was reported from Brazil and French Guiana.[79,][84] In 1994, dengue virus type 3 was introduced into the region.[53] In 1997, dengue virus type 2 with a Southeast Asian genotype was introduced into Santiago de Cuba and caused a sharp DHF/DSS outbreak observed only in individuals 20 years and older.[33] Dengue virus types 1 and 2 have been recovered from humans with mild clinical illness in Nigeria in the absence of epidemic disease.[37] In 1983, dengue virus type 3 was isolated in Mozambique.[43,][46] DHF/DSS has not been reported, and even dengue fever outbreaks are rare events. In this respect, Africa resembles the situation in Haiti, where multiple dengue serotypes are transmitted at high rates among a predominantly black population but severe disease is not recognized.[47] CLINICAL MANIFESTATIONS DENGUE FEVER In classic dengue fever (seen most frequently in adults), after an incubation period of 2 to 7 days,

patients experience a sudden onset of fever, which rapidly rises to 39.5 C to 41.4 C (103 F to 106 F) and usually is accompanied by frontal or retro-orbital headache. On occasion, back pain precedes the fever. A transient, macular, generalized rash that blanches under pressure may be seen during the first 24 to 48 hours of fever. The pulse rate may be slow in proportion to the degree of fever. Myalgia or bone pain occurs soon after onset and increases in severity. During the second to sixth day of fever, nausea and vomiting are likely to occur, and generalized lymphadenopathy, cutaneous hyperesthesia or hyperalgesia, aberrations in taste, and pronounced anorexia may develop. One or 2 days after defervescence, a generalized, morbilliform, maculopapular rash appears, with sparing of the palms and soles. It disappears in 1 to 5 days. In some cases, edema of the palms and soles may be noted, and desquamation may occur. About the time of appearance of this second rash, the body temperature, which has fallen to normal, may become elevated slightly and establish the biphasic temperature curve. Epistaxis, petechiae, and purpuric lesions are uncommon manifestations but may occur at any stage of the disease. Swallowed blood from epistaxis may be passed per rectum or be vomited and could be interpreted as bleeding of gastrointestinal origin. Gastrointestinal bleeding, menorrhagia, and bleeding from other organs have been observed in some dengue fever outbreaks.[85,][111,][117] Very clear evidence argues that peptic ulcer predisposes to gastrointestinal hemorrhage; in some cases, patients may exsanguinate during an otherwise normal dengue fever.[111] This syndrome is confused with DSS (see later) and contributes to a misunderstanding of the pathogenesis of severe dengue disease. The mechanism of the hemorrhagic diathesis that commonly occurs with dengue virus infection is not known, but speculation centers on platelet abnormalities. The DHF syndrome is differentiated from dengue fever by its association with thrombocytopenia and capillary leakage.[123] After the febrile stage, prolonged asthenia, mental depression, bradycardia, and ventricular extrasystoles are noted commonly in adults.[72] Primary infections with dengue virus types 2 and 4 are thought to be largely inapparent, particularly in children.[33,][55,][112,][113] Primary infections with dengue virus types 1 and 3 in adults produce biphasic fever and rash as the most characteristic features of the dengue fever syndrome.[10,][20,][36,][72,][97,][104] Manifestations vary with age and among patients. In infants and young children, the disease may be undifferentiated or characterized by a 1- to 5-day fever, pharyngeal inflammation, rhinitis, and mild cough. A distinctive mean incubation period, duration of illness, or spectrum of clinical findings could characterize disease with different dengue types, although these factors have not been studied carefully. Differences in mild dengue syndromes, hospitalized dengue cases (predominantly during secondary dengue virus infections), and chikungunya illnesses are illustrated in the section on chikun-gunya (see Tables 187-4187-51876187-7 to 187-8). Chikungunya virus infections produce a dengue fever syndrome. Because chikungunya and the dengue viruses are transmitted by the same vector mosquitoes, the diseases frequently are observed in the same patient populations. Chikungunya illnesses begin more abruptly than does dengue and are of shorter duration. Maculopapular rash, conjunctival injection, and myalgia or arthralgia occur more frequently in chikungunya than in dengue illnesses, but other features associated with both viruses are remarkably similar.[11,][77] DENGUE HEMORRHAGIC FEVER/DENGUE SHOCK SYNDROME DHF/DSS is an acute vascular permeability syndrome accompanied by abnormal hemostasis. The incubation period of DHF/DSS is unknown but is presumed to be the same as that of dengue fever. In children, progression of the illness is characteristic.[18,][76,][77,][123] A relatively mild first phase with an abrupt onset of fever, malaise, vomiting, headache, anorexia, and cough may be followed after 2 to 5 days by rapid deterioration and physical collapse. In Thailand, the median day of admission to the hospital after the onset of fever is day 4. In this second phase, the patient

usually has cold and clammy extremities, a warm trunk, a flushed face, and diaphoresis. Patients are restless and irritable and complain of midepigastric pain. Frequently, scattered petechiae appear on the forehead and extremities, spontaneous ecchymoses may develop, and easy bruisability and bleeding at sites of venipuncture are common findings. Circumoral and peripheral cyanosis may occur. Respirations are rapid and often labored. The pulse is weak, rapid, and thready, and the heart sounds are faint. The pulse pressure frequently is narrow (20 mm Hg); systolic and diastolic pressure may be low or unobtainable. The liver may become palpable two or three fingerbreadths below the costal margin and usually is firm and nontender. Chest radiographs show unilateral (right) or bilateral pleural effusions. Approximately 10 percent of patients have gross ecchymosis or gastrointestinal bleeding. After a 24- or 36-hour period of crisis, convalescence is fairly rapid in children who recover. The temperature may return to normal before or during the stage of shock. PATHOGENESIS AND PATHOLOGY In rhesus monkeys experimentally infected with dengue virus by subcutaneous inoculation, virus replicates initially in the skin. Virus disseminates rapidly to regional lymph nodes and then to lymphatic tissue throughout the body.[73] Early in the viremic period, virus can be recovered only from the skin inoculation site and lymph nodes, whereas 2 to 3 days later, evidence of general dissemination to skin and other tissues is found. Virus is recovered from the skin, lymph nodes, and several leukocyte-rich tissues for up to 3 days after termination of viremia. Virus can be recovered from circulating leukocytes and from the skin only at the end of the viremic period. The number of sites from which virus can be recovered increases as the infection progresses. Intracellular infection is terminated abruptly 2 to 3 days after viremia ceases. In humans, dengue viruses infect and replicate efficiently in intracutaneous Langerhans cells in vitro and in tissue explants.[124] Virus ultimately targets liver parenchymal cells, where infection produces apoptosis, but such cells may not serve as replicative hosts.[51] Late in infection, viral antigen is found associated with circulating B lymphocytes.[58] Fluorescent antibody, virus isolation, electron-microscopic, and in situ hybridization studies suggest mononuclear phagocytes as major infection hosts.[5,][3741,][43,][53a,][78] Animals infected initially with dengue virus type 1, 3, or 4 and then with dengue type 2 virus had higher viremia than when the same strain was inoculated into susceptible animals.[38,][43] This phenomenon, in vivo antibody-dependent enhancement of dengue virus infection, provides an explanatory hypothesis of the immunopathogenesis of dengue in humans. Epidemiologic, clinical, and virologic studies of DHF/DSS in humans have shown a significant association between severe illness and infection in the presence of circulating dengue antibody, whether it is passively acquired from the mother or actively acquired from previous infection.[9,][31,][32,][44,][59,][60,] [95,][99] This circulating antibody has two biologic activities: neutralization of virus and enhancement of infection.[39,][41] In Thailand, DHF/DSS developed in infants during dengue virus type 2 infection only when maternal neutralizing antibody had catabolized to low titer and infection-enhancing antibodies were left in circulation.[59] Similarly, in a prospective study of dengue virus infection in Thai children, DHF/DSS occurred in children who had circulating enhancing antibodies from a previous single dengue virus infection, but it did not occur in children whose first infection left them with low levels of cross-reactive dengue virus type 2 neutralizing antibody at the time of the second dengue virus infection.[60] A similar mechanism explains the failure of secondary infections to produce DHF/DSS with the American genotype dengue type 2. [116] American genotype dengue 2 viruses are significantly neutralized by human antidengue 1, whereas Southeast Asian dengue 2 viruses are not.[62] The full-length sequences of the American and Southeast Asian dengue 2 genomes reveal limited amino acid differences.[68] In vitro studies of dengue virus type 2 demonstrated enhanced growth in cultures of human mononuclear phagocytes that were supplemented with very small quantities of dengue antibodies.[45] Investigators have proposed that the number of infected mononuclear phagocytes in humans with naturally or passively acquired antibody may exceed that in nonimmune individuals.[38,][39,][43] In serial blood samples taken early in the illness in children experiencing

secondary dengue infections, enhanced viremia levels or enhanced levels of dengue NS1 predicted severe disease.[69,][113] Vascular permeability is thought to result when infected cells are attacked by activated T lymphocytes, with the subsequent release of vasoactive cytokines.[3,] [27,][28,][3841,][64,][74,][92] Cytokine production should be quantitatively related to the number of infected target cells. The reduced risk for DHF/DSS in protein-calorie malnourished children[109] and the increased risk for DHF/DSS in girls versus boys are consistent with the hypothesis that a competent immune elimination system must be available to generate the cytokines that produce DHF/DSS.[38,][40,][46] Evidence indicates the existence of a human dengue resistance gene. Epidemiologic studies of the 1981 Cuban outbreak demonstrated a higher risk for DHF/DSS in white than in black individuals.[31,][63] A search for DHF/DSS in black children in Haiti revealed no cases, despite the presence of high dengue type 1, 2, and 4 infection rates and circulation of the Southeast Asian genotype dengue 2 viruses.[47] Several HLA antigens have shown differing frequencies in DHF/DSS cases and controls.[13] Early in the acute stage of secondary dengue virus infection, rapid activation of the complement system occurs.[8,][81] During shock, blood levels of C1q, C3, C4, C5, C6, C7, C8, and C3 proactivator are depressed and C3 catabolic rates are elevated. The blood clotting and fibrinolytic systems are activated.[40,][120] As yet, neither the mediator of vascular permeability nor the complete mechanism of bleeding has been identified unequivocally. The kinin system apparently is not involved. Recent studies suggest a role for tumor necrosis factor, interleukin-2, and interferon-.[64,][92] Capillary damage allows fluid, electrolytes, protein, and, in some instances, red blood cells to leak into intravascular spaces. [121] This internal redistribution of fluid, together with deficits caused by fasting, thirsting, and vomiting, results in hemoconcentration, hypovolemia, increased cardiac work, tissue hypoxia, metabolic acidosis, and hyponatremia. A mild degree of disseminated intravascular coagulation, plus liver damage and thrombocytopenia, could contribute additively to produce hemorrhage.[120] If tissue cultures or suckling mice are used for recovery of virus, dengue virus usually is absent in tissues at the time of death.[78] If patients experienced a second dengue infection, their tissue suspensions contain large quantities of dengue-neutralizing antibodies. The use of mosquito inoculation techniques improves viral isolation rates. [89,][90,][107] Genetic probes increase viral detection sensitivity still further.[51] On pathologic examination, usually no gross or microscopic lesions are found that might account for death.[5] In rare instances, death may be caused by gastrointestinal or intracranial hemorrhage. Minimal to moderate hemorrhage is seen in the upper gastrointestinal tract, and petechial hemorrhage occurs frequently in the intraventricular septum of the heart, on the pericardium, and on the subserosal surfaces of major viscera. Focal hemorrhaging is seen occasionally in the lungs, liver, adrenals, and subarachnoid space. The liver usually is enlarged, often with fatty changes. Yellow, watery, at times blood-tinged effusions are present in serous cavities in approximately three fourths of patients. Retroperitoneal tissues are markedly edematous. On microscopic examination, perivascular edema in soft tissues and widespread diapedesis of red blood cells can be seen. Maturational arrest of megakaryocytes may be noted in the bone marrow,[66] and increased numbers of such megakaryocytes are seen in the capillaries of the lungs, in the renal glomeruli, and in the sinusoids of the liver and spleen. Proliferation of lymphocy-toid and plasmacytoid cells, lymphocytolysis, and lymphophagocytosis occur in the spleen and lymph nodes.[5] In the spleen, malpighian corpuscle germinal centers are necrotic. Depletion of lymphocytes occurs in the thymus. In the liver, varying degrees of fatty metamorphosis, focal midzonal necrosis, and hyperplasia of Kupffer cells are present. [5] Non-nucleated cells with vacuolated acidophilic cytoplasm resembling Councilman

bodies (apoptotic hepatocytes[51]) are seen in the sinusoids. A mild, proliferative glomerulonephritis is present. Biopsy specimens of the rash reveal swelling and minimal necrosis of endothelial cells, subcutaneous deposits of fibrinogen, and, in a few cases, dengue antigen in extravascular mononuclear cells and on blood vessel walls.[40,][43] DIAGNOSIS DENGUE FEVER A clinical diagnosis can be made by having a high index of suspicion and knowledge of the geographic distribution and ecology of dengue viruses. Activities of the patient during the period preceding the onset of illness may give important clues to the possibility of infection. The differential diagnosis includes many viral, respiratory, and influenza-like diseases and the early stages of malaria, typhoid fever, scrub typhus, hepatitis, and leptospirosis. Abortive forms of these diseases may never evolve beyond a denguelike stage. Four arbovirus diseases are dengue-like: chikungunya and o'nyong-nyong fever (togaviruses), West Nile fever (flavivirus), and Oropouche (bunyavirus). Four other diseases are dengue-like but without rash: Colorado tick fever, sandfly fever, Ross River fever, and the mild form of Rift Valley fever. Because of the variation in clinical findings and the multiplicity of possible causative agents, the descriptive term dengue-like disease should be used until a specific etiologic diagnosis is provided by the laboratory. DENGUE HEMORRHAGIC FEVER/DENGUE SHOCK SYNDROME According to WHO criteria, DHF is a dengue illness accompanied by thrombocytopenia (<100,000/mm3) and hemoconcentration (hematocrit >20% of the recovery value). Early detection of vascular permeability remains a diagnostic problem; however, the use of strain-gauge plethysmography documents up to 50 percent higher microvascular permeability in DHF/DSS patients than in controls.[4] Pleural or peritoneal effusions observed by ultrasonography or radiography are virtually pathognomonic. DSS is diagnosed when these manifestations are accompanied by hypotension or narrow pulse pressure (20 mm Hg). In areas endemic for dengue, hemorrhagic fever should be suspected in children with a febrile illness who exhibit shock and hemoconcentration with thrombocytopenia. Hypoproteinemia, hemorrhagic manifestations, and hepatic enlargement are frequent accompanying findings. Because many rickettsial diseases, meningococcemia, and other severe illnesses caused by a variety of agents may produce a similar clinical picture, the diagnosis should be made only when epidemiologic or serologic evidence suggests the possibility of dengue. Hemorrhagic manifestations have been described in other diseases of viral origin, including the arenavirus hemorrhagic fevers of Argentina, Bolivia, and West Africa (Lassa fever); the tick-borne hemorrhagic fevers of India and the former Soviet Union; hemorrhagic fever with renal syndrome, which occurs across northern Eurasia, specifically from Scandinavia to Korea; and Marburg and Ebola virus infections in central Africa.[54] LABORATORY STUDIES An etiologic diagnosis can be made by serologic study of properly collected serum samples, by isolation of the virus, and by identification of viral RNA or the nonstructural protein NS1 in acute-phase sera.[2,][52,][123] The acute-phase serum or plasma collected for isolation of virus should be stored optimally at 65 C or colder. Serologic diagnosis depends on a fourfold or greater increase in antibody titer by hemagglutination inhibition, complement fixation, radioimmunoassay, enzyme-linked immunosorbent assay (ELISA), or neutralization. IgM capture ELISA has revolutionized dengue

serology, and commercial kits are available.[19,][52] Primary and sequential (secondary) dengue virus infections result in the production of dengue-reactive IgM antibodies, which appear during the acute phase and disappear within 60 days of infection.[52] Secondary or primary dengue virus infections can be confirmed in a single serum specimen by quantitating IgM-IgG antibody ratios. IgG antibody concentrations are abundant in secondary but minimal in primary dengue virus infection. Dengue NS1 proteins can be detected in blood during the acute illness phase by dengue groupspecific antibodies, and these have been formatted into commercial ELISA or rapid immunochromatographic tests.[2a,][25a] By combining IgM-capture ELISA and NS1 detection, point-of-care tests may soon be available to diagnose acute dengue infections. Numerous techniques are available for the recovery and identification of dengue viruses.[101,] [123] Recommendations for general use have been made by a WHO expert committee.[123] Acute-phase serum, mosquito suspensions, or other materials thought to contain dengue virus may be inoculated into suckling mice, which may be examined for sickness or subtle neurologic signs or challenged at 14 days with a neurovirulent dengue virus. Repeated subpassage markedly increases the neurovirulence of dengue virus. Alternatively, materials may be inoculated into any of several tissue cultures of mammalian or mosquito origin and examined for plaques under agar or methylcellulose overlay, for cytopathic effect or resistance to a challenge cytopathic virus by use of a fluid overlay, or for fluorescence or other markers with the use of an appropriate detection system. Intrathoracic inoculation of A. albopictus, A. aegypti, or Toxorhynchites spp. is a highly sensitive dengue virus recovery system.[89,][90] TREATMENT DENGUE FEVER Treatment is supportive. Bed rest is advised during the febrile period. Antipyretics or cold sponging should be used to keep the body temperature below 40 C (104 F). Paracetamol (1015 mg/kg every 4-6 hours) is the preferred antipyretic agent. Analgesics or mild sedation may be required to control pain. Fluid and electrolyte replacement therapy is required when deficits caused by sweating, fasting, thirsting, vomiting, or diarrhea are present. Because of the risk of Reye syndrome and the dengue hemorrhagic diathesis, aspirin should not be given to reduce fever or to control pain. DENGUE HEMORRHAGIC FEVER/DENGUE SHOCK SYNDROME Explicit recommendations for management of DSS have been made by a WHO expert committee.[120] These and earlier recommendations by Cohen and Halstead[18] and recent studies by Dung and colleagues[25] and Wills and associates[122] are the basis of this section. No specific antiviral treatment exists, but in DHF/DSS, symptomatic and supportive measures are effective. The major pathophysiologic abnormality seen in DHF/DSS is an acute increase in vascular permeability that leads to leakage of plasma. Plasma volume studies revealed a reduction of more than 20 percent in severe cases. Supporting evidence of plasma leakage (and consequent hypovolemia) includes a rapid, weak pulse; diaphoresis; cool, pale skin of the extremities; decreased urine output; and direct measurement by strain-gauge plethysmography,[4] pleural effusion on chest radiography or ultrasonography, hemoconcentration, and hypoproteinemia. Pleural effusion may not be evident until after fluid resuscitation is started. In the absence of increased vascular permeability, clinically significant hemoconcentration may result from thirst, dehydration, fever, anorexia, and vomiting. Fluid intake by mouth should be as ample as tolerated. Electrolyte and dextrose solution (as used in diarrheal disease), fruit juice, or both are preferable to plain water. With high fever, a risk of convulsions exists, so antipyretic drugs may be indicated. Salicylates should be avoided because they are known to cause bleeding and acidosis. Acetaminophen is preferable at the following doses: younger than 1 year, 60 mg per dose; 1 to 3 years of age, 60 to 120 mg per dose; 3 to 6 years of age, 120 mg per

dose; and 6 to 12 years of age, 240 mg per dose. Children should be observed closely for early signs of shock. The critical period is the transition from the febrile to the afebrile phase. Frequent hematocrit determinations are essential because they reflect the degree of plasma leakage and the need for administration of intravenous fluid. Hemoconcentration usually precedes changes in blood pressure and pulse. The hematocrit should be determined daily from the third day until the temperature becomes normal for 1 or 2 days. Oral or parenteral fluid therapy can be administered in an outpatient rehydration unit for correction of dehydration or acidosis or when signs of hemoconcentration are present. The volume of fluid and its composition are similar to the fluids used for the treatment of diarrhea with moderate dehydration. The fluids should consist of the following: One third to half of the total fluid as physiologic saline solution. Half to two thirds of the remainder as 5 percent glucose in water. For acidosis: one fourth of the total fluids should be one-sixth molar sodium bicarbonate. Solution for fluid therapy in DHF: lactated Ringer solution, 5 percent glucose in one half normal physiologic saline solution, 5 percent glucose in one half lactated Ringer solution, 5 percent glucose in one third physiologic saline solution. Fluids as listed are calculated to be given during a 24-hour period. If the child seems severely dehydrated, half the calculated fluid is given in the first 8 hours and the second half in the next 16 hours. During rapid administration of fluids, watching for signs of cardiac failure is especially important. Written orders should be explicit about the type of solution and the rate of administration. A rough estimate of flow may be derived from the formula SHOCK Patients should be hospitalized and immediately treated when they have any of the following signs and symptoms of shock: restlessness or lethargy, cold extremities and circumoral cyanosis, rapid and feeble pulses, narrowing of pulse pressure (20 mm Hg) or hypotension, and sudden rise in hematocrit or continuously elevated hematocrit despite the administration of intravenous fluid. Shock is a medical emergency. Immediate administration of intravenous fluid to expand plasma volume is essential. In children, shock may develop or subside during the course of a 48-hour period, so close observation 24 hours a day is imperative. Patients with similar degrees of severity should be grouped together. Those with shock require intensive 24-hour care by nurses and physicians. Paramedical workers or parents can assist in provision of oral fluid therapy or in surveillance of the rate of intravenous fluid administration and general status of the patient. Initial fluid therapy with lactated Ringer or isotonic saline solution (20 mL/kg intravenously) infused as rapidly as possible may be required. Positive pressure may be necessary. In continued or profound shock, plasma expanders (6% dextran 70 or 6% hydroxyethyl starch [MW 200,000]) may be given to replace the initial fluid and administered at a rate of 10 to 15 mL/kg/hr or more until improvement in vital signs is apparent. In most cases, not more than 20 to 30 mL/kg of plasma is needed. Intravenous fluids (5% dextrose, half-normal lactated Ringer, or half-normal saline solution) are continued, even after improvement in vital signs and a declining hematocrit. The rate of fluid replacement should be adjusted as judged by the rate of plasma loss. Plasma loss may continue for 24 to 48 hours. Microhematocrit determination is a simple and reliable index for estimating plasma leakage. Monitoring of central venous pressure may be necessary in the management of severe cases of shock that are not easily reversible.

Administration of intravenous fluids should be discontinued when the hematocrit drops to approximately 40 percent and the patient's appetite improves. Good urine flow indicates sufficient circulating volume. In general, fluid therapy is not needed beyond 48 hours after termination of the shock. Extravasated plasma is reabsorbed and causes a further drop in hematocrit after the administration of intravenous fluid is stopped; if more fluid is given, hypervolemia, pulmonary edema, or heart failure may result. Of importance is that a drop in hematocrit at this stage is not viewed as a sign of internal hemorrhage. A strong pulse and blood pressure along with a wide pulse pressure and diuresis are good vital signs at this resorption phase. They rule out the likelihood of gastrointestinal hemorrhage, which occurs most frequently during the shock stage. Hyponatremia and, commonly, metabolic acidosis occur. Electrolyte and blood gas determinations should be performed periodically in severely ill patients as well as in those who do not seem to respond as promptly as expected. These determinations will provide an estimate of the sodium deficit and help determine the presence and degree of acidosis. Acidosis, in particular, may lead to disseminated intravascular coagulation if it is uncorrected. Heparin may be indicated in some of these patients, but extreme caution should be exercised in its use. In general, early volume replacement and early correction of acidosis with sodium bicarbonate result in a favorable outcome, and heparin is not required. Heparin should be reserved for patients with laboratory evidence of consumptive coagulopathy (disseminated intravascular coagulation) or intractable bleeding. Sedatives are needed in some cases because of marked agitation. Hepatotoxic drugs should be avoided. Chloral hydrate administered orally or rectally is recommended in a dose of 30 to 50 mg/kg as a single hypnotic dose (maximal dose, 1 g). In patients without pulmonary complications, paraldehyde, 0.1 mL/kg intramuscularly (maximal dose, 10 mL), also may be used. Oxygen therapy should be given to all patients in shock, but an oxygen mask or tent may increase apprehension. Blood transfusion is indicated only in patients with severe bleeding (e.g., gastrointestinal bleeding, hematemesis, melena). Fresh whole blood is preferable. Blood grouping and matching for prompt treatment should be carried out as a routine precaution for every patient in shock. In general, steroids do not shorten the duration of disease or improve the prognosis in children receiving careful supportive therapy.[108] Frequent recording of vital signs and determination of hematocrit are important in evaluating the results of treatment. If patients show any signs of shock, vigorous antishock therapy should be instituted promptly. Patients should be monitored constantly until it is reasonably certain that the danger has passed. In practice, the following should be carried out: 1. Pulse, blood pressure, respiratory rate, and temperature should be taken every 15 to 30 minutes or more often, until the shock resolves. 2. Hematocrit or hemoglobin studies should be performed every 2 hours for the first 6 hours and then every 4 hours thereafter until the patient is stable. 3. An accurate record of intake and output, including the type of fluid given, should be made. The frequency and volume of urine output should be recorded. Having a pro forma sheet for recording symptoms, signs, and treatment of DHF and DSS cases is useful. A blinded comparison of four intravenous fluids by Dung and associates[25] provided evidence that the relatively more expensive lactated Ringer solution provides no greater benefit than 0.9 percent saline does. Dextran could contribute to altered hemostasis.[46] Patients with DHF/DSS are resuscitated as though they have diarrhea. A more apt therapeutic analogy may be burn injury or hypovolemia from third-space loss in surgery. Of interest would be a trial of smallvolume hypertonic saline with or without colloid.[46] The widespread and unstudied use of blood

products to treat hemorrhage or simple thrombocytopenia[14] suggests a need for many more careful studies of DHF/DSS resuscitation.[46] In this regard, placebo-controlled or blinded studies are the ideal.[25,][108] EPIDEMIC DENGUE HEMORRHAGIC FEVER During epidemics, outpatient and inpatient facilities may be overwhelmed. Under these conditions, only children requiring hospital care should be admitted. A recently elevated body temperature and positive tourniquet test result are sufficient to suggest DHF; when possible, a microhematocrit and platelet count should be performed in the outpatient department. Patients with thrombocytopenia and an elevated hematocrit should be sent to a rehydration ward or, if the hematocrit does not fall or rises in the face of fluid therapy, admitted to a hospital. If a patient lives a long distance from the hospital and nearby accommodations are not available, admission for observation may be necessary. Triage can be performed by properly instructed paramedical workers. Competent laboratory assistance is an essential factor. Cool extremities, skin congestion, circumoral cyanosis, and rapid pulse are signs that suggest the need for hospitalization. Patients should be hospitalized until 2 days after the fever terminates. REGULATORY MEASURES Dengue diseases are not subject to international quarantine or surveillance regulations. An intensive and effective voluntary reporting system has been devised by the regional offices of the WHO. PROGNOSIS Not all patients suspected of having DHF need to be hospitalized because circulatory failure and shock may develop in only approximately a third of patients. Mild and moderate cases may be treated on an outpatient basis. For the purpose of early recognition of shock, parents should be advised to bring the patient back if evidence of clinical deterioration is noted or such warning signs as restlessness with or without lethargy, severe abdominal pain, cold extremities, and skin congestion occur on or after the third day following the onset of fever. In most cases, early and effective replacement of lost plasma with plasma, plasma expanders, or fluid and electrolyte solutions (or any combination of these products) results in a favorable outcome. The acute onset of shock and the rapid, often dramatic clinical recovery, together with the fact that no destructive or inflammatory vascular lesions are observed, suggest that the disease is produced by transient functional vascular changes caused by short-acting pharmacologic mediators. Sequelae in dengue or in DHF have not been studied systematically. Common sequelae of mild and uncomplicated dengue virus infection include bradycardia and ventricular extrasystoles during the convalescent stage, often persisting for several weeks. Profound asthenia with or without mental depression has been described. In patients with DHF/DSS, great care must be taken to reduce use of invasive procedures for managing shock. Nosocomial infections such as gram-negative sepsis can masquerade as DHF/DSS. Overhydration during the shock resuscitation phase may lead to heart failure and a complicated, stormy post-shock stage. Infrequently, residual brain damage occurs, apparently as a result of either prolonged shock or, occasionally, intracranial hemorrhage. Children in whom profound shock develops rapidly with no detectable diastolic pressure or with unobtainable blood pressure, children in shock with delayed admission to the hospital, or children in shock with gastrointestinal hemorrhage have a poor prognosis. Mortality rates may exceed 50 percent in these groups. PREVENTION Tissue culturebased vaccines for dengue virus types 1, 2, 3, and 4 are immunogenic but not yet licensed for use.[6,][7] Numerous multivalent dengue vaccines using a variety of approaches are in various stages of development.[30,][50,][61,][65] At present, prophylaxis depends on the use of insecticides, repellents, protective body clothing, and screens on houses to avoid mosquito bites.

Destruction of A. aegypti breeding sites also is effective.[16] If water storage is mandatory, a tight-fitting lid or a thin layer of oil may prevent eggs from being deposited or hatching. A larvicide such as temefos (Abate), which is available as a 1 percent sand granule formulation and effective at a concentration of 1 ppm, may be added safely to drinking water. EPIDEMIC MEASURES The WHO recommendations are as follows. On the basis of epidemiologic and entomologic information, the size of the area that requires adult mosquito abatement should be determined. With technical-grade malathion or fenitrothion at 438 mL/hectare, two adulticide treatments at a 10-day interval should be made with the use of a vehicle-mounted or portable ultra-low-volume aerosol generator or mist blower.[71,][123] Cities of moderate size should stockpile at least one vehicle-mounted aerosol generator, five mist blowers, 10 swing fog machines, and 1000 L of ultra-low-volume insecticide to be prepared to perform adulticide operations over a 20-km2 area rapidly. With limited funds, such equipment and insecticides can be stockpiled centrally for rapid transportation when required. Priority areas for launching ground applications are those that have a concentration of cases. Special attention should be focused on areas where people congregate during daylight hours, such as hospitals and schools. During the early stages of epidemics, an ultra-low-volume spray of 4 percent malathion in diesel oil or kerosene may be used to spray all houses within a 100-m radius of the residence of patients with DHF. ERADICATION AND CONTROL A. aegypti was eradicated successfully from countries and whole continents with use of the techniques pioneered by the Rockefeller Foundation in its worldwide program to control urban yellow fever.[106] With time, the species successfully re-established itself in much of its former range. An eradication campaign in the United States was abandoned and replaced by a program of disease surveillance and containment of introduced virus. Mosquito control or eradication programs require the simultaneous use of two approaches: a reduction in breeding sites and the application of larvicides. Alternatively, a significant reduction in population may be effected by closely spaced applications of adulticide.[82] Source reduction requires public support either by legal sanctions or by voluntary actions (see the following section). Source reduction campaigns should be well organized, supervised, and evaluated. Proper disposal of discarded cans, bottles, tires, and other potential breeding sites not used for storage of drinking or bathing water should be performed. Sides of water storage containers should be scrubbed to remove eggs when the water level is low. Water storage containers for drinking and bathing and flower vases should be emptied completely once weekly. Water containers that cannot be emptied should be treated with Abate 1 percent sand granules at a dosage of 1 ppm (e.g., 10 g of sand to 100 L of water). Treatments should be repeated at intervals of 2 to 3 months. Vehicle-mounted or portable ultra-low-volume aerosol generators or mist blowers can be used to apply technical-grade malathion or fenitrothion at 438 mL/hectare. Three applications made at 1week intervals can suppress A. aegypti populations for approximately 2 months. In dengue-endemic countries, little effort has been made to adopt building codes or waste collection methods to reduce the number of mosquito breeding sites.[67] Furthermore, almost no way has been found to use the private sector to implement vector control despite ample evidence of long-term successful programs in the United States.[12,][17] HEALTH EDUCATION Control of A. aegypti has been maintained effectively in some tropical areas through the simple expedient of emptying water containers once a week. During the yellow fever campaigns, strong sanitary laws made the breeding of mosquitoes on premises a crime punishable by fine or jail.

[106] In the modern era, Singapore and Cuba have adopted these measures successfully. Health education through mass media or through the schools has been attempted in Burma, Thailand, Malaysia, and Indonesia, but without spectacular success.[24] The goals of health education and community participation approaches are to make the population aware of the identity of the vector of DHF, to describe its biting habits (daytime feeding) and its breeding habits (containers holding clean water), and to motivate people to reduce breeding sources by emptying water from containers on a regular basis.[42] The use of piped water rather than water storage should be encouraged. Studies in Malaysia after the 1973 epidemic of DHF indicated a very low level of functional knowledge among the inhabitants of Kuala Lumpur, Malaysia, about the vector of DHF. [24] Discouragingly, persons who were informed correctly, in most instances, took no action to protect themselves against mosquito breeding in their homes. Extensive effort is being made to apply social science methods to gain the voluntary participation of the population in sustained mosquito control programs.[70]