Overview of Heartworm Disease

Mosquito vector species acquire microfilariae (a neonatal larval stage) while feeding on an infected host. Once ingested by the mosquito, microfilariae develop into the first larval stage (L₁). They then actively molt into the second larval stage (L₂) and again to the infective third stage (L₃) within the mosquito in ~1–4 wk, depending on environmental temperatures. This development phase requires the shortest time (10–14 days) when the average ambient temperature is >81°F (27°C) and the relative humidity is 80%. When mature, the infective larvae migrate to the labium of the mosquito. As the mosquito feeds, the infective larvae erupt through the tip of the labium with a small amount of hemolymph onto the host’s skin. The larvae migrate into the bite wound, beginning the intramammalian phase of the life cycle. A typical Aedes mosquito is capable of surviving the complete development of only small numbers of HW larvae, usually <10 larvae per mosquito.

In canids and other susceptible hosts, infective larvae (L₃) molt into a fourth stage (L₄) in 3–12 days. After remaining in the subcutaneous tissue, abdomen, and thorax for ~2 mo, L₄ undergo their final molt at day 50–70 into young adults, arriving in the heart and pulmonary arteries ~70–120 days after initial infection. Only 2.5–4 cm in length on arrival, worms rapidly grow in the pulmonary vasculature to adult worms (males ~15 cm long, females ~25 cm). When juvenile heartworms first reach the lungs, blood flow forces them into the more distal small pulmonary arteries of the caudal lung lobes; as the parasites grow, they occupy larger and larger pulmonary arteries, moving into the right ventricle and atrium when the worm burden is high. Gravid females produce microfilariae as early as 6 mo after infection but more typically at 7–9 mo after infection.

Microfilariae are detectable in most infected canids (~80%) not receiving macrolide prophylaxis and occasionally in those dogs placed on macrolide preventives when a HW infection was already present. The number of circulating microfilariae does not correlate well to the adult female HW burden. Adult worms typically live 3–5 yr, whereas microfilariae may survive for up to 2 yr in the dog, while awaiting arrival of a mosquito intermediate host.

Most dogs are highly susceptible to HW infection, and most (an average of 56%) experimentally administered infective larvae (L₃) develop into adults. Ferrets and cats are susceptible hosts, but the rate of infective larvae developing into adults is low (an average of 6% in cats and 40% in ferrets). In cats, the adult burden is often only one to three worms. It appears that early death of juvenile worms on arrival at the pulmonary vasculature is largely responsible for the heartworm-associated respiratory disease (HARD) syndrome in cats. HARD does not require maturation of heartworms but is due to the body’s response to the dying/dead heartworms. When maturation does occur, adult worm survival in cats is typically not longer than 2–3 yr. In all animals capable of being infected, aberrant larval migration may occur, resulting in parasitic lesions in the CNS, systemic arterial system, and in visceral and subcutaneous sites.

The severity of cardiopulmonary pathology in dogs is determined by worm numbers, host immune response, duration of infection, and host activity level. Live adult heartworms cause direct mechanical trauma, and other suspected factors (eg, antigens and excretions) are thought to directly irritate or to stimulate the hosts’ immune system to damage vessel intima, leading to proliferative endarteritis and perivascular cuffing with inflammatory cells, including infiltration of high numbers of eosinophils. Live worms seem to have an immunosuppressive effect; however, the presence of dead worms leads to more severe vascular reactions and subsequent lung pathology, even in areas of the lung not directly contacting the dead heartworms. Longterm infections, due to all of the factors noted (ie, direct irritation, worm death, and immune response) result in chronic lesions and subsequent scarring. Active dogs tend to more often develop pulmonary hypertension than inactive dogs for any given worm burden. Frequent exertion increases pulmonary arterial pathology and pulmonary artery resistance (with resultant pulmonary hypertension) and thereby may precipitate overt clinical signs, including congestive heart failure (CHF). High worm burdens are most often the result of infections acquired from numerous mosquito exposures. High exposures in young, naive dogs in temperate climates can result in severe infections, possibly precipitating vena cava syndrome the year after. In general, because of the worm size and smaller dimensions of the pulmonary vasculature, small dogs do not tolerate infections and treatment as well as large dogs.

The role of the endosymbiotic bacteria Wolbachia pipiens, which live intracellularly within the filarid parasite, is still being determined. However, these bacteria have been implicated as playing a role in the pathogenesis of filarial diseases, possibly through endotoxin production. Furthermore, studies have demonstrated that a primary surface protein of Wolbachia (WSP) induces a specific IgG response in hosts infected by D immitis. For veterinarians, the most important aspect of Wolbachia is its symbiotic relation with D immitis. This bacterium is necessary for normal maturation, reproduction, and infectivity of the heartworm. If Wolbachia are eradicated, the heartworm gradually dies, after first becoming sterile. This can be accomplished with doxycycline therapy.

HW-associated inflammatory mediators that induce immune responses in the lungs and kidneys (eg, immune complex glomerulonephritis) cause vasoconstriction and possibly bronchoconstriction. Leakage of plasma and inflammatory mediators from small vessels and capillaries causes parenchymal lung inflammation and mild, noncardiogenic edema formation. Pulmonary artery disease compromises vascular compliance, and this, with reduced ability to adequately vasodilate, results in increased flow velocity, especially with exertion, and resultant shear stresses further damage the endothelium. The process of endothelial damage, vascular dysfunction, increased flow velocity, and local ischemia is a vicious cycle. Inflammation with ischemia can result in irreversible interstitial fibrosis.

Pulmonary arterial pathology in cats and ferrets is similar to that in dogs, although the small arteries develop more severe muscular hypertrophy. Despite this, pulmonary hypertension with CHF is less common in cats than in dogs or ferrets. Arterial thrombi, thromboemboli, and living or dead worms become lodged within pulmonary arteries or arterioles, resulting in complete or partial obstruction. In cats, parenchymal changes associated with dead heartworms differ from those observed in dogs and ferrets. Rather than type I alveolar cell damage, as found in dogs, cats develop type II alveolar cell hyperplasia, which can act as a significant barrier to oxygenation. Most significantly, because of restricted pulmonary vascular capacity and subsequent pathology, ferrets and cats are more likely than dogs to die as a result of HW infection.

In dogs, infection is ideally identified by serologic testing before onset of clinical signs; however, at the earliest, HW antigenemia and microfilaremia do not appear until ~5 and 6.5 mo after infection, respectively. When dogs do not receive preventive medication and are not appropriately tested, infection and disease progress undetected. Clinical signs of HW infection, such as coughing, exercise intolerance, unthriftiness, dyspnea, cyanosis, hemoptysis, syncope, epistaxis, and ascites (right-side CHF) may develop. The frequency and severity of clinical signs correlate to lung pathology and level of animal activity. Signs are often not observed in sedentary dogs, even though the worm burden may be relatively high. Infected dogs experiencing a dramatic increase in activity, such as during hunting seasons, may develop overt clinical signs. Likewise, worm death and thromboemboli precipitate clinical signs.

A dog may be classified as low- or high-risk for developing clinical signs, based on assessment of potential worm burden, the health and age of the dog, and its lifestyle. There is also a more complex classification system in which dogs are classified from I to IV, based on severity of signs. Class I dogs are minimally affected clinically. Class II dogs exhibit cough. Class III dogs are severely affected and variably present with cough, hemoptysis, weight loss, lethargy, exercise intolerance, dyspnea, heart failure (ascites), and radiographic findings suggestive of HW disease (large main pulmonary artery and lobar pulmonary arteries, truncated and tortuous pulmonary arteries, pulmonary infiltrate, and hilar lymphadenopathy). Class IV includes dogs with caval syndrome. Dogs 5–7 yr old are at higher risk of having a heavy worm burden, presumably because of increased time of exposure and for disease development. Other concurrent health factors (eg, concurrent cardiopulmonary or other organ system disease) affect risk assessment. The degree to which exercise can be restricted during the recovery period is another important consideration.

Infected cats may be asymptomatic or exhibit intermittent coughing, dyspnea, heart failure, vomiting, lethargy, anorexia, or weight loss. When evident, signs usually develop during two phases of the HW life cycle: 1) the arrival of juvenile worms in the pulmonary vasculature ~3–4 mo after infection, and 2) death of adult heartworms. The early signs are associated with an acute vascular and parenchymal inflammatory response to the newly arriving young worms and the subsequent death of many or all of these juveniles. This initial phase is often misdiagnosed as asthma or allergic bronchitis. However, this is now considered to be part of HARD. Antigen tests in such cats are negative (measured antigens are associated with mature female worms) during the early eosinophilic pneumonitis syndrome, although antibody tests typically are positive. Although not yet well characterized, it is believed that clinical signs often resolve and may not reappear for months. HARD has been postulated to contribute to longterm lung damage. Cats harboring mature worms may exhibit intermittent vomiting, lethargy, coughing, or episodic dyspnea. Death of even one adult heartworm can lead to acute respiratory distress and shock, which may be fatal and appears to be the consequence of pulmonary thrombosis and/or anaphylactic-like shock.

Ferrets, more so than cats, mimic canine HW infection in terms of clinical signs. The large parasite:host body weight ratio dictates that ferrets (and cats) develop clinical signs with relatively small worm burdens. Ferrets with HW disease may demonstrate one or more of the following: weight loss, fatigue, rapid and/or labored breathing, heart murmur, distended and pulsatile jugular veins, cough, grey and cold mucous membranes, ascites, pleural effusion, fainting, and sudden death. See table: Diagnostic Tests, Clinical Signs, and Treatment for Heartworms in Dogs, Cats, and Ferrets.

The antigen detection test is the preferred diagnostic method for routine screening of asymptomatic dogs or when seeking verification of a suspected HW infection. Antigen testing is the most sensitive and specific diagnostic method available to veterinary practitioners. Even in areas where the prevalence of HW infection is high, ~20% of infected dogs are not microfilaremic, which diminishes the utility of screening by testing for microfilariae. This figure is even higher for dogs infected with adult heartworms and that are consistently administered monthly macrolide prophylaxis, because this kills microfilariae and induces embryo stasis in mature female dirofilariae.

Timing of antigen testing is critical. A pre-detection period must be considered, because these tests detect only adult, female worms. This takes into account the time from exposure to seroconversion to a positive antigen test. A reasonable interval is 7 mo after last possible exposure. There is no value in testing a dog for antigen or microfilariae before ~7 mo of age. To ensure that a previously acquired infection does not exist in these young dogs, they should be tested 6–7 mo after beginning HW prophylaxis. For dogs >7 mo old, testing should be performed when preventive therapy is started and 7–12 mo later. Subsequently, annual antigen detection tests are recommended.

The terminology for HW antigen tests has changed, with the word negative being replaced by “below detectable limits” to underscore the possibility of HW-infected pets being antigen negative and that a negative test may become positive as worms mature.

The level of antigenemia is directly related to the number of mature female worms present. Most dogs harboring more than two adult female worms will test positive with most available tests. For low-burden suspects, commercial laboratory–based microwell titer tests are the most sensitive. There is, however, no test that can determine worm burden. Testing for microfilariae may be useful as an adjunctive test in suspect cases that have negative antigen test results.

In dogs, echocardiography is relatively unimportant as a diagnostic tool. Visualization of worms in the right heart and vena cava is associated with high-burden infection with or without caval syndrome. Severe, chronic pulmonary hypertension causes right ventricular hypertrophy, septal flattening, underloading of the left heart, and high-velocity tricuspid and pulmonic regurgitation. The ECG of infected dogs is usually normal. However, right ventricular hypertrophy patterns are seen when there is severe, chronic pulmonary hypertension, often associated with overt or impending right-side CHF (ascites). Cardiac rhythm disturbances are usually absent or mild, but atrial fibrillation is an occasional complication in dogs.

The diagnosis of HW disease in cats is based on historical and physical findings, index of suspicion, thoracic radiographs, echocardiography, and serologic test results. Cats may develop a positive antigen test 7–8 mo after L₃ inoculation. However, antigen tests alone are considered too unreliable (insensitive, missing 25%–50% of mature infections) as the initial screening test for cats. This occurs with unisex (all male) infections, infections with insufficient numbers of mature females to be detectable, and in cats with HARD. Cats with HARD may remain antigen negative if no adults develop, in all-male infections, or when only one adult female matures. These cats can also be only temporarily negative if tested before detectable antigenemia develops. The antigen test is strongly recommended in cats in which HW infection is suspected.

Antibodies to heartworms, produced by 90% of infected cats, often appear by 2–3 mo after L₃ infection and are generally present by 5 mo. However, antibodies can persist for several months after worm death. Also, antibodies induced by larvae can persist in aborted infections and after macrolide prophylaxis has been instituted, killing the early larval stages. Thus, a positive antibody test indicates infection by HW larval stages, and possibly HARD, but not necessarily of a mature infection. In conjunction with other provocative findings, antibody seropositivity is useful in making a clinical diagnosis of HW disease in cats, and it certainly identifies cats at risk. False-positive results from cross-reactivity with other parasites have not been seen. A negative antibody test indicates ≥90% probability of the absence of mature infection. Microfilariae are rarely detected by modified Knott’s tests (<10%) in cats. Annual screening of cats is not necessary but may yield information for concerned cat owners. For this purpose, the antibody test is preferred in that it detects cats with heartworms and those at risk. The antigen test is not appropriate for screening in cats because of its low sensitivity.

In cats, worms can often be imaged using echocardiography. This is because of the relative sizes of the heartworm(s) and the right heart and pulmonary arterial system of cats. Heartworms, particularly the females, are long enough to occupy the pulmonary arteries as well as the right heart, where they can be easily imaged. Parallel hyperechoic lines, produced by the HW cuticle, may be seen in the right heart and pulmonary arteries. Echocardiography is more important in cats than in dogs because of the increased difficulty of diagnosis in cats (low antigen test sensitivity and low antibody test specificity for mature infection) and the relatively high sensitivity of the test in experienced hands.

In addition to special diagnostic tests in cats and dogs, a CBC, chemistry profile, urinalysis, and particularly thoracic radiographs are indicated. Laboratory data are often normal. Eosinophilia and basophilia alone or together may occur in dirofilariasis. Eosinophilia is most often seen at the time that stage 5 (young adult) larvae arrive in the pulmonary arteries. Subsequently, eosinophil counts vary but are usually high in dogs with immune-mediated occult infections, especially if eosinophilic pneumonitis develops (<10% of total infections).

Hyperglobulinemia due to antigenic stimulation may be present in dogs and cats. Hypoalbuminemia in dogs can be associated with proteinuria in severe immune-complex glomerulonephritis or with severe emaciation/cardiac cachexia. Serum ALT and alkaline phosphatase are occasionally increased but do not correlate well with abnormal liver function, efficacy of adulticide treatment, or risk of drug toxicity. Urinalysis may reveal proteinuria that can be quantitated by a urine protein:creatinine ratio. Occasionally, severe glomerulonephritis can lead to hypoalbuminemia and nephrotic syndrome. Dogs with hypoalbuminemia, secondary to glomerular disease, also lose antithrombin III and are at risk of thromboembolic disease. Hemoglobinuria is associated with caval syndrome and occurs when RBCs are lysed in the circulation.

In dogs, thoracic radiography provides the most information on disease severity and is a necessary screening tool to assess the clinical status of dogs with dirofilariasis, particularly when symptomatic. High-risk infections are characterized by a large main pulmonary artery segment and dilated, tortuous caudal lobar pulmonary arteries. Right ventricular enlargement may also be seen and, along with enlarged pulmonary arteries, is indicative of pulmonary hypertension. With pulmonary thromboembolism and pulmonary infiltrate with eosinophils (pneumonitis), ill-defined parenchymal infiltrates surround the caudal lobar arteries, typically most severe in the right caudal lobe.

In cats, cardiac changes and pulmonary hypertension are less common. In ~50% of infected cats, caudal lobar arteries are larger than the corresponding vein and >1.6 times the diameter of the ninth rib at the ninth intercostal space. Patchy parenchymal infiltrates may also be present in cats with respiratory signs. The main pulmonary artery segment usually is not visible because of its relatively midline location.

In ferrets, the diagnosis is less readily made with thoracic radiographs, because only the right ventricle tends to be enlarged. However, the commercial antigen tests have detected HW antigen experimentally, as early as 5 mo after infection, and have been shown to be effective in clinical situations. False-negative results may occur, especially in species that harbor lower worm burdens (cats and ferrets). Furthermore, although microfilaria testing is only rarely helpful, adult worms can often be seen with echocardiography and nonselective angiography.

The extent of the preadulticide evaluation varies, depending on the clinical status of the dog, the likelihood of coexisting diseases that may affect the outcome of treatment, the owner's ability to restrict the dog's exercise, and cost considerations. Clinical laboratory data should be collected selectively to complement information obtained from a thorough history, physical examination, antigen test, and thoracic radiography.

Two important variables known to directly influence the probability of thromboembolic complications after adulticide treatment and the outcome of treatment are the extent of concurrent pulmonary vascular disease and the current worm burden. Assessment of cardiopulmonary status is indispensable for evaluating a dog's prognosis. Pulmonary thromboembolic complications after adulticide treatment are most likely to occur in heavily infected dogs already exhibiting clinical and radiographic signs of severe pulmonary vascular disease, especially when severe pulmonary hypertension and CHF are present.

Before adulticide therapy, HW-infected dogs are assessed and rated for risk of postadulticide thromboembolism. Dogs can be categorized as follows: 1) low risk of thromboembolic complications, light worm burden, and no evidence of parenchymal and/or pulmonary vascular lesions; or 2) high risk of thromboembolic complications. Dogs in the low-risk category would ideally fulfill the following conditions: no clinical signs, normal thoracic radiographs, a low level of circulating antigen or a negative antigen test with circulating microfilariae, no worms visualized by echocardiography, no concurrent disease, and with owners capable of completely restricting exercise. The low-risk group would also include dogs having previously undergone adulticidal therapy but that remain antigen positive. Dogs with near-normal thoracic radiographs may develop severe thromboembolic disease, occurring most often when exercise is not restricted. Dogs at high risk of thromboembolic complications include those with signs related to HW infection (eg, coughing, dyspnea, ascites), abnormal thoracic radiographs, high level of circulating antigen, worms visualized by echocardiography, concurrent disease, and little or no possibility that the owners will restrict exercise.

The only approved heartworm adulticide is melarsomine dihydrochloride, which is variably effective against mature (adult) and immature heartworms of both sexes, with male worms being more susceptible. Melarsomine is given at 2.5 mg/kg, deep IM in the belly of the epaxial (lumbar) musculature in the area of the third to fifth lumbar vertebrae, using a 22-gauge needle (1 in. long for dogs <10 kg or 1.5 in. for dogs >10 kg). Pressure at the injection site is applied and maintained for 5 min to prevent drug migration. Approximately one-third of dogs will exhibit local pain, swelling, soreness with movement, or sterile abscessation at the injection site. Local fibrosis is not uncommon (and is the reason for targeting the belly of the epaxial musculature). In standard use, the procedure is repeated on the opposite side 24 hr later for dogs at low risk of treatment complications. However, to reduce the danger of thromboembolism, a two-phase (also termed “split-dose” or "three-dose” method) treatment is highly recommended for at-risk dogs and, indeed, for all patients, unless cost considerations prohibit this approach. Using this protocol, a single injection of melarsomine is given, followed by two injections 24 hr apart, after an interval of at least 30 days. The American Heartworm Society recommends this three-dose alternative regimen, regardless of the stage of disease or risk category. Exercise restriction is essential once treatment is started to minimize the risk of pulmonary thromboembolism due to dead and dying adult worms.

An approach to adulticidal therapy in which preventive is started at time of diagnosis is doxycycline (10 mg/kg, bid for 30 days) and monthly HW preventive, at the standard preventive dosage. After 2 mo, adulticidal injections (melarsomine at 2.5 mg/kg, IM) are initiated, as the dog's condition allows. Daily corticosteroids, using a tapering dosage, may also be administered during this period to reduce pulmonary inflammatory lesions from dying worms and from melarsomine. Although exercise is minimized from the day of diagnosis, cage rest must be enforced from the day of each initial injection for 4–6 wk. If the dog's condition allows, melarsomine injections are repeated in 1 mo (2 injections 24 hr apart), with the same regimen of prescribed exercise restriction. If, after the first injection, the dog has suffered significant pulmonary damage from the resultant worm death, the second and third injections can be withheld indefinitely.

Dogs with high worm burdens are at risk of severe respiratory complications. Because only ~50% of heartworms are destroyed after the first injection, the cumulative impact of worm emboli on severely diseased pulmonary arteries and lungs is reduced. Furthermore, if serious thromboembolism develops, the second two-dose part of the regimen can be delayed, allowing the lungs to heal from the first insult. Lastly, this approach destroys a higher percentage of adult heartworms than the standard two-dose protocol. For the utility and advisability of various therapeutic protocols, see Table: Guide to Choosing Heartworm Therapeutic Protocol.

Doxycycline has become an important part of treatment of HW infection in dogs. Through its negative action on Wolbachia, it provides benefits to the canid host and works to the detriment of D immitis. Doxycycline is indicated in preadulticide therapy (at 10 mg/kg, bid, for 30 days) in HW-infected dogs. It is given in conjunction with ivermectin at the preventive dosage (6–12 mcg/kg/mo). This combination reduces the severity of lung injury after adulticidal therapy, probably through reducing the amount of Wolbachia antigen and the proteins released from the HW uterus as the bacteria die and the uterus degenerates. Doxycycline at this dosage hastens worm death when the “slow-kill” approach is used, thereby presumably reducing the negative impact of worms on the host. Doxycycline with ivermectin also clears the host of microfilariae. Therefore, in dogs undergoing slow-kill treatment, this combination decreases risk of macrolide resistance, which is a concern in the slow-kill method using ivermectin alone. Doxycycline is advocated in treating dogs with HW infection regardless of severity classification or protocol.

The American Heartworm Society recommends administration of prophylactic doses of macrolides for 2 mo before administration of melarsomine, with the first dose given concurrently with the first dose of doxycycline (day 1) and a second dose given after the end of the doxycycline treatment (day 30). A third dose is then given concurrently with the first dose of melarsomine (day 60). Macrolide administration is continued monthly thereafter at the preventive dosage. The rationale for this approach is to eliminate susceptible migrating D immitis larvae and to allow nonsusceptible 2–4 mo old larvae to age to a point at which they are more susceptible to melarsomine. This approach of a 2-mo pretreatment with macrolides has become less compelling with the recent knowledge that doxycycline kills developing larvae (L₃>L₄>young adults) when administered at 10 mg/kg, bid, for 30 days, thereby closing the gap during which developing larvae are not susceptible to melarsomine treatment.

High-risk dogs should be stabilized before melarsomine administration. Stabilizing treatment includes cage confinement, oxygen, corticosteroids, and possibly heparin (75–100 U/kg, SC, tid) for 1 wk before the alternative (split-dosage) melarsomine treatment protocol.

Dogs with right-side CHF should be treated with furosemide (1–2 mg/kg, bid), an angiotensin-converting enzyme (ACE) inhibitor such as enalapril (0.5 mg/kg/day, increased to 0.5 mg/kg, bid, after 1 wk pending renal function test results), moderate dietary sodium restriction, and abdominal paracentesis, as needed.

The inodilator, pimobendan, is also indicated at 0.25 mg/kg, bid, to support myocardial function and reduce cardiac work. Sildenafil can be used initially at 1 mg/kg, tid, as a pulmonary vasodilator. Caution is warranted with this and other vasodilators to avoid the adverse effect of systemic hypotension.

After melarsomine injection(s), exercise must be severely restricted for 4–6 wk to minimize pulmonary thromboembolic complications. Adverse effects of melarsomine are otherwise limited to local inflammation, cough, brief low-grade fever, and salivation. Hepatic and renal toxicity are seldom, if ever, seen.

Laboratory findings associated with adulticidal therapy may include an inflammatory leukogram, thrombocytopenia, and prolonged activated clotting time or prothrombin time. A postinjection increase in serum CK may be noted. Local or disseminated intravascular coagulopathy may occur when platelet counts are <100,000/μL. Treatment for severe thromboembolism should include oxygen, cage confinement, a corticosteroid at an anti-inflammatory dosage (eg, prednisone at 1 mg/kg/day, PO), and low-dose heparin (75–100 U/kg, SC, tid) for several days to 1 wk. Severe lung injury is likely present if, after 24 hr of oxygen therapy, no improvement is noted and partial pressures of oxygen remain <70 mmHg.

The standard melarsomine protocol (two-dose, 24-hr treatment regimen) kills most adult worms, clearing 50%–85% of dogs. Antigen testing should be performed 8–12 mo after the third dose of the split-dose (alternative) protocol. If a positive test result is obtained at this time, consideration can be given to abbreviated retreatment (two injections, 24 hr apart). A “slow-kill” approach with ivermectin (0.6 mcg/kg, every 2 wk for 6 mo) can be substituted for repeated melarsomine injection but should definitely be preceded by 30 days of doxycycline therapy (10 mg/kg, bid) because this minimizes the reaction to dead and dying worms, enhances the kill rate to ~1 yr (vs 2.5 yr with ivermectin alone) versus the standard slow-kill approach, and is thought to decrease the risk of resistance (see above). The standard "slow-kill" approach with ivermectin alone is against the current recommendations of the American Heartworm Society. Longterm use of macrolides, rather than melarsomine, to kill adult worms allows pulmonary pathology to progress during the lengthy period in which worms are dying and being processed.

Caval syndrome results from worms migrating retrograde to the right atrium and great veins and is usually the result of a precipitous fall in cardiac output, as might occur with pulmonary thrombosis. Severe pulmonary hypertension is then complicated by worm-induced tricuspid valve leakage, hemolysis, and damage to liver and kidneys. In caval syndrome, removal of worms from the right atrium and orifice of the tricuspid valve is typically necessary to save the life of the dog. This may be accomplished by using light sedation, local anesthesia, and either a rigid or flexible alligator forceps, or an intravascular retrieval snare, introduced preferentially via the right external jugular vein. With fluoroscopic guidance, if available, the instrument should continue to be passed until worms can no longer be retrieved. Immediately after a successful operation, the clinical signs should lessen or disappear. Fluid therapy may be necessary in critically ill, hypovolemic dogs to restore hemodynamic and renal function. After full recovery from surgery, adulticidal therapy is undertaken to eliminate remaining worms. Particular care should be taken if many worms are still visible echocardiographically.

There is currently no satisfactory treatment approach for heartworm infection in cats. Infection often is lethal. Thus, all cats in regions endemic for canine HW disease should receive drug prophylaxis. The lifespan of adult heartworms in cats has traditionally been thought to be 2 yr, so spontaneous recovery is possible. Cats may remain asymptomatic, experience episodic vomiting and/or episodic dyspnea (resembling asthma), may die suddenly from pulmonary thromboembolism or an anaphylactoid reaction, or rarely, develop CHF.

Because there is no safe or approved adulticide for cats, many are managed conservatively with restricted activity and corticosteroid therapy, such as prednisolone (1–2 mg/kg, PO, every 24–48 hr). Steroids reduce the severity of vomiting and respiratory signs. The hope is that episodes of pulmonary complications will not prove fatal as the worms die. Barring consecutive, additional infection, 25%–50% of cats may survive with this approach. Serial antigen and antibody testing (at intervals of 6–12 mo) can be used to monitor status. Although there are no supportive data, administration of doxycycline (10 mg/kg, bid for 30 days) and ivermectin (24 mcg/kg/mo) to an infected cat could be theorized to cause worm degradation and contracture, thereby lessening the potential for catastrophic consequences when the worms die. Of course, the macrocyclic lactone would also protect the cat from a new infection, if more exposure is encountered.

Surgical retrieval of worms from the right atrium, right ventricle, and vena cavae via jugular venotomy can be attempted in cats with high worm burdens detected by echocardiography. An endoscopic basket, snare, or horsehair brush can also be advanced via the right jugular vein under fluoroscopy. Cats in CHF have been cured by worm removal.

Treatment in ferrets is, likewise, difficult, because there is no approved agent for this purpose. Adulticidal therapies (thiacetarsemide and melarsomine) have resulted in ~50% mortality in ferrets. Moxidectin (injectable and topical formulations) has been widely thought to be adulticidal for heartworms in ferrets and is given at the same dosage and frequency as in dogs. Moxidectin and imidacloprid (combination) is approved by the FDA for use in ferrets to prevent HW infection and to prevent and treat flea infestations.

Heartworm infection is generally completely preventable with macrolide prophylaxis. Year-round prevention is advised. Preventive therapy in dogs is recommended beginning at 6–8 wk of age. No testing is necessary at this age, because the presence of mature female heartworms is required to produce a positive heartworm test (antigen or microfilaria). When prophylaxis is started after 7 mo of age, an antigen test and a test for presence of microfilariae is recommended, followed by another antigen test 6–7 mo later. This series of tests will help to avoid unnecessary delay in detecting subclinical infections, as well as potential confusion concerning effectiveness of the preventive program, because it cannot be determined until the second test whether infection existed before beginning chemoprophylaxis.

Formulations of the macrolide preventives ivermectin, milbemycin oxime, moxidectin, and selamectin are safe and effective, as prescribed, for all breeds of dogs. Currently marketed products have additional chemicals and parasite spectra, including GI and ectoparasites: ivermectin (hookworms), ivermectin/pyrantel pamoate (hookworms and roundworms), ivermectin/pyrantel pamoate/praziquantel (hookworms, roundworms, and tapeworms), milbemycin/lufenuron (hookworms, roundworms, whipworms, and sterilizes fleas), milbemycin/lufenuron/praziquantel (hookworms, roundworms, whipworms, tapeworms, and sterilizes fleas), selamectin (fleas, ticks, ear mites, sarcoptic mites), moxidectin injectable (hookworms), moxidectin/imidacloprid (roundworms, hookworms, whipworms, adult fleas, microfilariae), milbemycin/spinosad (hookworms, roundworms, whipworms, fleas).

At the approved dosage, milbemycin kills microfilariae quickly, and in the face of high microfilarial concentrations a shock reaction may occur. Thus, milbemycin should not be administered without close monitoring and/or prophylactic pretreatment (steroids and/or antihistamine) as a preventive in dogs with high numbers of microfilariae.

HW prevention is also recommended for all cats in endemic regions, regardless of housing status, because of the potential for severe consequences with infection. Performing microfilaria testing in cats before starting preventive therapy is not required, because cats have small microfilarial numbers and the presence of microfilaria is typically transient. Ivermectin for cats is safe and effective at 24 mcg/kg, PO, once monthly. At this dosage, the formulation is also effective against Ancylostoma tubaeforme and A braziliense. Preventive treatment should be started in kittens at 6 wk of age and continued lifelong. There is currently no milbemycin product marketed for use in cats in the USA.

Formulations of selamectin and a combination of imidacloprid/moxidectin are labeled for both dogs and cats. Selamectin is administered topically at a monthly dosage of ~6 mg/kg and also kills adult fleas and prevents flea eggs from hatching for 1 mo. It also is indicated for treatment and control of Otodectes cynotis in dogs and cats, sarcoptic mange, Dermacentor variabilis infestations in dogs, Ancylostoma tubaeforme, and Toxocara cati in cats. A topical combined formulation of imidacloprid and moxidectin administered at dosages of 10 mg/kg for imidacloprid and 1 mg/kg for moxidectin is also effective against HW infection and flea infestations. Although all currently marketed preventives are likely effective in ferrets, only imidacloprid with moxidectin is approved by the FDA. Importantly, the preventive dosage for ferrets is the same as that for dogs (not cats).

Sporadic resistance of heartworms to the macrocyclic preventive class has been recognized since 2013. All the current molecules used to prevent heartworm disease have been implicated. However, some formulations appear to be more effective against resistant isolates than others. There have been isolates from six dogs with varying degrees of resistance. There is little evidence of spread out of the Mississippi Delta region, where resistance was first recognized. It is important to realize that the current preventives are effective in the vast majority of cases and should not be abandoned. Emphasis should be placed on owner compliance and year-round preventive therapy, as well as on alternative methods of HW prevention, including topical and oral mosquito repellants, indoor/screened housing, especially at night, and mosquito abatement programs. The role of slow-kill macrolide adulticidal therapy has been questioned in the development of resistance and should be avoided. If such therapy is unavoidable, it should absolutely be accompanied by 30 days of doxycycline treatment at the outset, with assurance that microfilariae are eradicated.