B.U.K. Li

 New hope for children with cyclic vomiting syndrome 

By B U. K. Li, MD, and Jennifer C. Howard, RN 

 Research is shedding light on the mysteries of this distressing disorder, which isn't as rare as you might think. Once the diagnosis is made, antiemetic and antimigraine drugs provide reasonably effective relief. 

 Cyclic vomiting syndrome (CVS) is characterized by recurrent stereotypical spells of vomiting between which the child is completely well. For nearly two centuries, since the disorder was first described, its cause and pathophysiology have remained unknown.

 Over the past decade, however, some light has emerged from the darkness (see "

"). Pediatric gastroenterologists have doubled the number of relevant publications and assembled an interdisciplinary team to begin translational (bench-to-bedside) research into the mechanisms of and effective therapy for this disorder.

After reviewing this article the physician should be able to: 

 The 21st century dawns with new hope for affected children and adolescents. Collaborative research promises improved understanding and evidence-based treatment of CVS, and an international support group, the Cyclic Vomiting Syndrome Association (with 29 international chapters), has been formed to help families cope with this disruptive disorder (for 

 see below). We hope this article will help you diagnose CVS accurately and in a timely fashionthereby avoiding the delays that often occur when the condition is misdiagnosed as acute gastroenteritis, gastroesophageal reflux, food poisoning, or psychogenic vomitingso that your patients can get the treatment and support they need. 

 In 1882, Samuel Gee identified the three essential clinical features of CVS that form the core of the consensus diagnostic criteria established in 1994 (Table 1):

 Recurrent severe, discrete episodes of vomiting 

 Varying intervals of normal health between episodes 

 Duration of vomiting episodes from hours to days 

 No apparent cause of vomiting (negative laboratory, radiographic, and endoscopic testing) 

 Stereotypical: Each episode similar as to time of onset, intensity, duration, frequency, associated symptoms and signs within individuals 

 Self-limited: Episodes resolve spontaneously if left untreated 

 Nausea, abdominal pain, headache, motion sickness, photophobia (+ lethargy) 

 Fever, pallor, diarrhea, dehydration, excess salivation, social withdrawal 

 Source: The International Scientific Symposium on Cyclic Vomiting Syndrome

 To these features, we have added a fourth criterion: absence of laboratory findings that point to a specific cause of vomiting. 

 When seen in the office between episodes, the child with CVS is asymptomatic and appears so normal that the history of repetitive bouts of relentless vomiting and dehydration can be difficult to believe. Although the typical child spends 90% of the time well, we have documented substantial annual morbidity and medical costs: 20 missed days of school for each child, a high rate of IV rehydration (50% of children with CVS compared with fewer than 1% of patients with rotavirus), and more than $17,000 in costs for each family attributable to laboratory testing, imaging studies, endoscopic procedures, emergency room visits, hospitalizations, and missed work by parents.

 Although the prevalence of CVS in the United States has not been established, it can no longer be considered a rare disorder based on the available evidence. Two population surveys among Caucasian children in Scotland and Australia estimated the prevalence at 2%.

 In an informal survey of four-person pediatric practices, we found that between four and 12 cases per practice had been diagnosed. In our clinical experience, CVS causes one third of cases of recurrent vomiting in children older than 2 years. CVS occurs in all races, with a girl-to-boy ratio of 55:45. The median age of onset is 6 years, but the correct diagnosis is usually not made for 2.6 years. CVS can begin in infancy and is often difficult to differentiate from gastroesophageal (GE) reflux. We traced one case back to six days of age through hospital records that documented a recurrent nine-hour pattern of vomiting every 20 minutes. Although it appears to be primarily a disorder of childhood, CVS is more and more being diagnosed in adults.

 Prevalence in adults appears to be lower than in children, however. 

 We suspect that many cases of CVS are missed, judging from the fact that the apparent prevalence is much higher than the actual number of cases diagnosed. In our practice, we observed that the number of cases identified for each gastroenterologist rose from one to seven annually as awareness of the disorder increased. 

 In truth, any individual episode of CVS resembles a bout of acute gastroenteritis associated with dehydration. Episodes are also often labeled as food poisoning. Because different physiciansoffice-based pediatric partners, urgent care physicians, and emergency room doctorsmay evaluate individual vomiting episodes, the recurrent pattern often goes undetected. Once the pattern is recognized, labels such as GE reflux or psychogenic vomiting are often applied. It is usually the parents who first recognize, after multiple episodes, that the problem cannot be recurrent stomach flu. 

 An acute episode of CVS is commonly mistaken for a bout of viral gastroenteritis because of concomitant vomiting, low-grade fever (29%), and diarrhea (36%) (Table 2, available in the print issue, adapted from Li BUK: Cyclic vomiting: new understanding of an old disorder. 

 On closer examination, children with CVS appear substantially more ill than those with viral gastroenteritis. They vomit at a peak pace of every five to 10 minutes, an intensity unmatched by any other disorder.

 They can be so pale as to appear to be in shock and are often locked into a fetal position by unrelenting pain and nausea. Even before becoming dehydrated from fluid loss, they may become so listless that they cannot walk or talk and have to be carried everywhere. 

The temporal pattern of vomiting in CVS is distinct from the chronic vomiting of GE reflux. Our study of the patterns found that they could be quantitatively differentiated (Figure 1, available in the print issue, adapted from Li BUK: Cyclic vomiting: new understanding of an old disorder. 

 Children with a cyclic pattern vomited at a peak of six to 13 times an hour, but only twice a month (high intensity, low frequency). Those with the chronic pattern vomited a maximum of one or two times an hour but nearly every day (low intensity, high frequency). By using the criteria of more than four emeses an hour at peak and fewer than two episodes a week, we were able to differentiate CVS from recurrent vomiting with greater than 92% sensitivity and specificity. Although "cyclic" implies stable periodicity (usually every two or four weeks), that pattern is in fact found among only half of children with CVS; the other half experience vomiting that is episodic but unpredictable in timing. 

 Besides vomiting, children with CVS suffer from severe anorexia, nausea, retching, and midline abdominal pain. Despite the purported relationship between CVS and migraine headaches, our study found classic migraine symptoms of headache, photophobia, phonophobia, and vertigo in fewer than half of patients. Migraine-associated symptoms, however, were found in significantly higher percentages than among a comparison group with chronic vomiting or case controlsunderscoring the suspected relationship between CVS and migraine.

 CVS episodes can happen at any time but tend to occur consistently around the same time of day in a given patient, most often in the early morning hourseither between 2 a.m. and 4 a.m. or upon awakening between 6 a.m. and 7 a.m.

 Parents often describe the episodes as beginning and ending abruptly. When documented, the prodromal periodsome combination of pallor, anorexia, nausea, abdominal pain, and lethargy but not teichopsia (a sensation of luminosity before the eyes)and the recovery period (from the point of last emesis to eating and being playful) are both relatively brief: 30 minutes and 5 hours, respectively. 

 When asked, two thirds of parents can identify life events that appear to trigger the episodes. Two types of stressorsinfectious and psychologicalpredominate.

 The infections include common respiratory infections, the most frequent being confirmed chronic sinusitis. The psychological stresses are as likely to be positive excitement (birthdays, holidays, vacations) as negative stressors related to school (exams) or family (divorce). Other triggers include foods (chocolate, cheese, and monosodium glutamate), food or inhalant allergies, physical exhaustion, motion (car rides), anesthesia, and menses in postmenarchal girls. 

 Three questions are essential to make a tentative diagnosis of CVS: 

 Has the child had at least three episodes? (100% respond "Yes") 

Is the child completely asymptomatic between episodes? (100% respond "Yes"), and 

Are the episodes stereotypical? (98% respond "Yes") 

 If you ask simply whether the child has had previous similar bouts of "stomach flu" or vomiting, and the response is five or six episodes over the past few months, consider CVS. Other supporting evidence includes vomiting frequency greater than every 15 minutes at peak (71% of patients), associated symptoms of pallor (91%), lethargy (87%), abdominal pain (80%), retching (78%), anorexia (74%), and nausea (72%), as well as a positive family history of migraine (82%). 

 If the three main criteria are met, the child clearly has a cyclic vomiting pattern that is compatible with CVS. Laboratory testing is currently required, however, to exclude serious disorders that can present with recurrent episodic vomiting (see "

," below). Once testing is completed with negative results, the child can be given a diagnosis of idiopathic CVS. 

 CVS remains a diagnosis made by both historical criteria and absence of laboratory findings that support an alternate diagnosis. Yet how much testing is required to exclude the many potential underlying disorders remains open.

 One in eight patients (12%) with the cyclic vomiting pattern is found to have a specific underlying disorder that requires intervention. Costly lawsuits have resulted from missed diagnoses, especially malrotation with volvulus, which can lead to complications that include small bowel resection and long-term parenteral nutrition. 

 The complete evaluation includes standard blood work (glucose, electrolytes, alanine transaminase [ALT], 

-glutamyl transpeptidase [GGTP], amylase, lipase, urinalysis) and metabolic screening (lactate, ammonia, amino acids and urine organic acids, 

-ALA, and porphobilinogen) during the episode, small bowel radiography, abdominal ultrasonography, magnetic resonance imaging or computed tomography scan of the head and sinuses, and endoscopy (Table 3, available in the print edition, adapted from Li BUK, Balint JP and LiBuk: Cyclic Vomiting: New understanding of an old disorder. 

 1996;17(7):48). The main red flags in the initial evaluation that impel us to do additional testing include severe headaches, GI bleeding, unilateral abdominal pain, weight loss, failure to respond to any treatment, progressive worsening, prolonged episodes requiring repeated hospitalizations, and a change in pattern or symptoms.   

The fact that migraine-associated CVS (CVS with a family history of migraine) is found in 82% of patients and is likely to respond to antimigraine therapy (79%) suggests that a short-term empiric antimigraine trial could potentially precede diagnostic testing.

 A recent cost analysis found that an initial approach comprising small intestinal radiography to exclude malrotation plus an empiric two-month antimigraine trial was more beneficial than the standard approach of complete metabolic, radiographic, and endoscopic testing before any treatment.

 That strategy cannot be definitively recommended until tested prospectively, however. 

 The good news is that most children appear to outgrow CVS with advancing age, most often during the preteen or early teenage years.

 In our series, the median ages at onset, diagnosis, and resolution were 5, 8, and 10 years.

 The time from onset of symptoms to accurate diagnosis was 2.6 years. The average child experienced 22 vomiting episodes over three and a half years. The age at resolution among the children in our series, 10 years, is probably lower than the actual age because our series is biased toward children in whom CVS has ended early; it will likely rise with longer follow-up. 

 The bad news is that many children trade cyclic vomiting for migraine headaches at 10 years or older (Figure 2). So far, 28% of those in our study whose vomiting episodes have ended have developed migraine headaches, and we project that 75% will do so by 18 years of age. 

 The simple answer is "Yes." Strong circumstantial evidence points to substantial clinical overlap between CVS, abdominal migraine, and migraine headache.

 Some of the evidence is as follows. A positive family history of migraine is strongly associated with CVS, but not with chronic vomiting (82% vs. 14%).

 Nearly 75% of children with CVS, abdominal migraine, or migraine headache share associated symptoms of pallor, anorexia, and nausea. From the natural history, we know that CVS can evolve into migraine headaches. Perhaps most compelling, both CVS and abdominal migraine frequently respond to antimigraine headache therapy. It appears that CVS, abdominal migraine, and migraine headaches may represent three age-dependent phases of migraine. 

 How does one differentiate CVS and abdominal migraine when so many symptoms overlap (Table 4, available in the print edition, adapted from Li BUK, Balint JP: Cyclic vomiting syndrome: Evolution in understanding of a brain-gut disorder. 

 2000;47:117)? It depends on the predominant or most consistent primary symptom, whether vomiting, abdominal pain, or headachecorresponding to CVS, abdominal migraine, and migraine headache, respectively.

 Even then, nearly 80% of children with CVS can be given a diagnosis of abdominal migraine interchangeably, and 50% of those with abdominal migraine can be given a diagnosis of CVS concomitantly.  

 Although no established therapy for CVS has been proven efficacious in controlled trials, reasonably effective empiric treatments have been identified based on clinical experience and open trials.

 Our approach is to provide supportive care during acute episodes. This includes admitting the child to a quiet, dark hospital room to avoid overstimulation that can trigger further nausea and vomiting, giving IV fluids to replace losses, administering sedation to reduce nausea, and, occasionally, giving an analgesic.

 Pharmacologic treatment includes abortive medications, administered at the onset of the episode to break the attack, and prophylactic agents, given daily to prevent attacks (Table 5, available in the print edition, adapted from Li BUK, Balint JP: Cyclic vomiting syndrome: Evolution in understanding of a brain-gut disorder. 

 2000;47:117). The efficacy of these agents has been difficult to establish because of the absence of controlled trials and a high placebo responsegreater than 50%.

 Treatment is based on three variables: family history of migraine, frequency of episodes, and severity of episodes. If there is a family history of migraine, we start antimigraine therapy. If episodes occur more than once every four weeks or are severe or prolonged (requiring hospitalization lasting longer than two days), we give prophylactic therapy daily. If episodes occur less often than once a month or are relatively mild, we usually try abortive therapy at the onset of the episode. If abortive therapy fails, we resort to prophylactic therapy.   

No established hierarchy of medications exists for treating CVS. We therefore recommend medications whose dose titration and side effects you are familiar with. In the abortive category, 5-HT

 antagonists such as ondansetron and antimigraine triptans such as sumatriptan are effective in more than half of children when given parenterally or nasally, respectively. Oral doses are, as would be expected, ineffective because they cannot be kept down. 

 In the prophylactic category, antimigraine agents (propranolol, cyproheptadine, and amitriptyline) are most widely used, but anticonvulsant (phenobarbital) and prokinetic (erythromycin) medications have also proved successful.

 Medications that may become available for use include new antimigraine triptans, anticonvulsants (gabapentin and topamirate), tachykinin receptor antagonists (potential antiemetics in animal studies), and corticotropin-releasing factor antagonists (to truncate the stress response). 

 Once the cyclic vomiting pattern has been identified, the next step is to perform exclusionary radiographic and laboratory testing to either establish the diagnosis of CVS or identify a specific underlying disorder. Once screening tests are completed, the child can be started on an empiric trial of medication. Depending on your comfort level, some form of prophylactic antimigraine therapy is reasonable. Referral to a pediatric gastroenterologist, neurologist, or metabolic specialist who is familiar with the disorder is recommended: 

 for further testing, such as endoscopy and specific definitive metabolic tests 

when the history is atypical or red flags are raised (severe headaches or unilateral abdominal pain, for example), and 

when the child fails to respond to treatment within one or two months. 

Articles

Research is shedding light on the mysteries of this distressing disorder, which isn&t as rare as you might think. Once the diagnosis is made, antiemetic and antimigraine drugs provide reasonably effective relief.

B.U.K. Li

Articles

CME: New hope for children with cyclic vomiting syndrome