Index of Suspicion

[Alon]

Постараюсь публиковать здесь интересные клинические случаи по детским болезням.
К участию и обсуждению приглашаются все желающие.
------------------------------------------------------------
Case 1
Presentation

A 21-month-old boy known to have chronic gastroesophageal reflux is brought to your office for a second opinion regarding poor growth. Born at term, he was diagnosed as having reflux at 3 weeks of age due to arching and crying with feedings. He was noticed to be failing to thrive initially at 12 months of age when, despite adequate caloric intake, his weight percentile had fallen from the 75th to the 5th percentile over 6 months. His height and head circumference have remained between the 10th and 25th percentiles. He has experienced intermittent episodes of constipation and frequent colds and ear infections. He walked at 15 months of age and says about 150 words.

His evaluation has included multiple laboratory tests, including cystic fibrosis and celiac disease screening, thyroid function tests, CBC, measurement of electrolytes, and urinalysis. He has been evaluated by pediatric gastroenterology, and despite a formula change and multiple antireflux medications, his weight has remained in the 5th percentile. A pH probe study and endoscopy performed at 20 months of age demonstrated only mild reflux. He has had negative skin testing for food allergies.

The diagnosis is revealed after additional history and a laboratory test are obtained.
--------------------------
Case 1 Discussion

On further questioning, the child’s mother mentioned that he drinks about 70 to 80 oz of fluid a day! Generally, he drinks 40 oz of water, 16 oz of rice milk, and two to three cans of liquid dietary supplement. This excessive thirst started at about 6 months of age. He awakens three to five times a night asking for water, and his wet diapers are noted to be incredibly heavy.

Measurement of urine electrolytes demonstrated concentrations of sodium below 5 mEq/L (5 mmol/L) and chloride below 20 mEq/L (20 mmol/L), with urine osmolality below 50 mOsm/L while the serum osmolality was high-normal at 303 mOsm/L. In retrospect, previous urinalysis results had been abnormal because of a specific gravity of 1.005, although the other indices had been normal. His serum chloride value had been mildly elevated at 109 mEq/L (109 mmol/L), but his sodium value was normal at 144 mEq/L (144 mmol/L).

The boy was admitted to the hospital for a water deprivation test, which showed an increase of serum sodium (from 144 to 152 mEq/L [144 to 152 mmol/L]), increase in serum osmolality (from 297 to 310 mOsm/L), no change in urine osmolality (range, 64 to 77 mOsm/L), and continued passage of dilute urine (specific gravity below 1.003). His weight decreased 5% (9.4 kg to 8.95 kg), and his urine volume averaged 16 mL/kg per hour.

After an injection of arginine vasopressin (AVP), his urine osmolality and specific gravity remained low at 70 mOsm/L and 1.002, respectively. His urine output was more than 30 mL/kg per hour during the subsequent hour after injection. The diagnosis of congenital nephrogenic diabetes insipidus (CNDI) was confirmed.

The Signs
Causes of polyuria and polydipsia can be differentiated based on laboratory values. Solute diuresis can be identified by finding a urine-to-plasma osmolality ratio greater than 0.7. Examples of solute diuresis are diabetes mellitus and intrinsic renal disease, which can be determined by routine chemistries. Water diuresis, in addition to a urine-to-plasma osmolality of less than 0.7, demonstrates a urine specific gravity below 1.010. Three common causes of water diuresis are compulsive water drinking, central diabetes insipidus (CDI), and nephrogenic diabetes insipidus (NDI).

Compulsive water drinking can be differentiated from diabetes insipidus (DI) by laboratory testing, if the history is not forthcoming. Interestingly, compulsive water drinkers often have a preference for ice water and are less likely to have nocturia. Low serum osmolality combined with low urine osmolality suggests compulsive water drinking.

CDI often can be suspected on the basis of history, as when the patient manifests brain tumor, head trauma, or CNS infection. In approximately 25% of cases, no cause can be determined.

It is important to differentiate congenital from acquired causes of NDI, such as drug effects, electrolyte disorders, urinary tract obstruction, and systemic disorders such as sickle cell anemia. It is helpful to recall that acquired NDI can be caused by primary renal disease, systemic disease with renal involvement, medication effects, and miscellaneous conditions.

Making the Diagnosis
If the patient can drink freely, the baseline serum sodium concentration generally normalizes, and there is no evidence of dehydration. A first step to distinguish NDI from other causes of polydipsia is to collect a 24-hour urine sample. For a small child, a shorter collection period may be necessary. The finding of polyuria in a patient who is dehydrated and has an elevated serum sodium concentration as well as hypo-osmolar urine implies a renal concentration defect. The diagnosis of NDI is confirmed by a water deprivation test, with results showing the inability to concentrate the urine despite the release of vasopressin.

The water deprivation test involves dehydrating the patient and testing for changes in specific laboratory values. Generally, the criteria for dehydration are 3% to 5% weight loss, orthostasis, a plateau in urine osmolality, or hypernatremia. Urine volume, osmolality of plasma and urine, and electrolyte values are followed closely. A plasma vasopressin level is measured at the start and completion of dehydration.

Both types of DI are characterized by the inability to concentrate urine after water deprivation. The urine osmolality usually remains below 200 mOsm/kg H2O, and the urine specific gravity remains below 1.010. On the other hand, compulsive water drinkers pass concentrated urine (urine osmolality >500 mOsm/kg H2O) after being deprived of water. Patients who have CDI demonstrate inadequate concentrations of vasopressin after a water deprivation test, but concentrations are elevated in those having NDI. To distinguish CDI further from NDI, vasopressin is administered. Individuals who have CDI have a significant increase (usually >50%) in urine osmolality after vasopressin is administered either intranasally, subcutaneously, or intravenously. Patients having NDI maintain a low urinary osmolality (usually <200 mOsm/kg).

The Condition
CNDI is a genetically inherited error, with 90% of cases being X-linked and 10% involving an autosomal recessive or autosomal dominant mechanism. The distal nephron is insensitive to the antidiuretic effects of AVP, which leads to the production of large amounts of hypotonic urine. Normally, in the collecting duct, AVP from the posterior pituitary gland binds with the vasopressin type-2 receptor (V2R), which activates adenylate cyclase to increase cyclic AMP. This reaction stimulates protein kinase A, which redistributes aquaporin-2 water channels (AQP2), permitting the apical plasma membrane to become permeable to water. As a result, water flows from the tubular lumen to the medullary interstitium, which produces concentrated urine. In the X-linked form, the mutation exists in the V2R gene; in the autosomally inherited forms, the mutation is in the AQP2 gene.

Clinical Features
Classic features of DI are polyuria and polydipsia, which can lead to hypernatremic dehydration. The excessive fluid intake also may blunt the appetite. Nonspecific signs and symptoms include failure to thrive, recurrent fevers due to dehydration, vomiting or poor feeding, and constipation. More severe manifestations are seizures and mental retardation, which occurs rarely because awareness of the disorder has increased. Physical findings may be normal or the patient may show signs of dehydration and irritability.

Treatment
CNDI is treated by dietary restrictions and medication. Sodium intake is reduced to 0.7 mEq/kg per day, and protein intake is decreased to 8% of total caloric intake to decrease renal solute load and decrease the volume of urine required for solute excretion. Thiazide diuretics, often coupled with potassium supplements, or potassium-sparing diuretics with dietary changes paradoxically can reduce urine volume by 20% to 50%. Amiloride and prostaglandin synthetase inhibitors have been used in combination with thiazide diuretics. Nonsteroidal anti-inflammatory drugs are useful in the acute setting to reverse hyperosmolality. It is important to remember that an increase in urine osmolality to more than 200 mOsm/kg H2O has a dramatic effect on urinary losses.

Lessons for the Clinician
A comprehensive history is essential in determining a cause for failure to thrive. Growth failure can be a presenting sign of CNDI. It is important to ask not only about inadequate urination and oral intake, but also about polyuria and polydipsia.


Laurie Liang, MD
Kaiser Santa Clara Hospital, Santa Clara, Calif

[Alon]

Case2
Presentation

A 3-year-old boy is seen in the ED because of a "bloodshot eye." Several weeks ago he was hit in the eye with a ball. Two days ago his eye became red but has not bothered him. He has had no ocular discharge, fever, or respiratory symptoms. He comes from a disrupted family complicated by parental drug use and has siblings in foster care. He has not received routine care, although he was seen regularly in infancy and is fully immunized.

Physical examination reveals an active, well-nourished toddler who has conjunctival injection and a blood layer in the anterior chamber of his right eye (hyphema). There appears to be no vision in his affected eye, and a red reflex cannot be obtained. There are no signs of facial trauma, strabismus, or pigmented yellow lesions around the eye. Examination of the skin shows no petechiae or bruising except on the anterior shins. The child is admitted for management of hyphema.

Laboratory findings include a WBC count of 8.5x103/mcL (8.5x109/L), Hgb concentration of 12 g/dL (120 g/L), platelet count of 275x103/mcL (275x109/L), international normalized ratio of 1, and partial thromboplastin time of 41 seconds. A skeletal survey shows no fractures. Imaging confirms the cause of the hyphema.
-----------------
Case 2 Discussion

Further history revealed that the child’s right eye has glowed like a "cat’s eye" when viewed from across the room since he was 24 months old. CT scan revealed a calcified mass within the right globe and no abnormalities of the left globe. The findings of a calcified mass within the globe on CT scan were consistent with retinoblastoma.

The Condition
Hyphema in children most commonly occurs after blunt or lacerating trauma, but the differential diagnosis is broad. Spontaneous hyphema can arise from juvenile xanthogranuloma, leukemia, hemophilia, thrombocytopenia, aspirin or warfarin ingestion, and intraocular tumors such as retinoblastoma.

Although hyphema in children usually resolves without sequelae, complications can ensue and include increased intraocular pressure, optic atrophy, synechiae between the iris and the lens, and corneal blood staining. Because complications are more common with rebleeding, initial management is aimed at preventing further bleeding. Placement of a hard metal shield to protect the eye (Fox shield) is an important preventive step. Bed rest or quiet ambulation should be instituted, along with elevation of the head of the bed. Admission should be strongly considered for all children, especially if there is any question of compliance. Because patients who have sickle cell hemoglobinopathy are at higher risk for complications, Hgb electrophoresis should be obtained if the diagnosis is in question.

Ophthalmologic consultation is required for any child presenting with hyphema to prevent late sequelae. The ophthalmologist administers medicines that lower intraocular pressure, decrease inflammation, and prevent synechiae. Surgical intervention is considered if intraocular pressure cannot be managed with maximal medical therapy or if corneal blood staining occurs.

The Underlying Condition
Retinoblastoma, although rare, is the most common malignant intraocular tumor in children (incidence of 1 in 17,000) and must be considered in any case of spontaneous hyphema. The tumor typically is diagnosed in the first years after birth. The initial sign usually is leukocoria, and family members often describe a glow or cat’s eye appearance. Retinoblastoma also can present with strabismus, hyphema, and periocular inflammation resembling orbital cellulitis. The disease can be familial or sporadic and can present unilaterally or bilaterally. Bilateral disease almost always is familial. A mutation of the retinoblastoma gene (a tumor suppressor gene) on chromosome 13 is responsible for tumor development. Metastasis is uncommon. Spread can occur by direct extension along the optic nerve into the brain or the subarachnoid space or through the bloodstream to bone, bone marrow, lung, and lymph nodes. When retinoblastoma is diagnosed, both parents and all siblings should be examined for retinoblastoma.

The management of retinoblastoma involves a multidisciplinary team using multiple modalities that prioritize tumor eradication. Careful and judicious attempts should be made to preserve vision. Treatments include chemotherapy, external beam radiation, laser photocoagulation, enucleation, or combinations of these approaches. Although enucleation remains the most common treatment for unilateral disease, the trend in more recent years has been toward less frequent use of this approach, probably because of the more effective use of other approaches such as radiation, chemotherapy, and laser therapy. The prognosis for patients who have unilateral retinoblastoma without optic nerve invasion or metastasis is excellent, with more than 90% survival among those using current therapy. However, if tumor is disseminated, the mortality increases substantially. Surgical washout, which sometimes is performed for bleeding in the eye, carries the potential for dissemination, which could have proved a lethal choice for this child.

Lessons for the Clinician
The possibility of retinoblastoma should be entertained for any child who develops a spontaneous hyphema. In this case, the child’s complicated social situation might have led the clinician to suspect nonaccidental trauma as the cause of the child’s hyphema, delaying the diagnosis. When the retina cannot be visualized, it is imperative to obtain an imaging study to rule out disease in the posterior pole.


Carrie Phillipi, MD, PhD
Laurie Christensen, MD
John Samples, MD
Oregon Health and Science University, Portland, Or

[Alon]

The knotty problem in an infant girl's groin
Jul 1, 2005
By: Doran L. Fink, MD, PhD, Janet R. Serwint, MD

Mid-afternoon, and you're working at the pediatric clinic of your university hospital. With the beautiful spring weather outside, you've seen few acute care visits today; itchy eyes and runny noses in otherwise healthy children have established an unusually relaxing pace! Only one patient remains to be seen: a nearly 3-month-old girl whose mother is concerned about a "knot in the baby's groin."

Mom greets you with an uneasy smile as you enter the examination room. "It hasn't gone away," she says. "I think it's gotten bigger."

Instantly, you remember that the last time you saw the infant was four weeks ago—at her 2-month-old health maintenance visit. She had a diaper rash.

Reviewing your notes from that encounter, you recall that the rash had appeared one week before that visit. Apart from that, your patient was growing well, without complaints of fever, cough, congestion, diarrhea, or emesis. She was afebrile; vital signs were normal for age. You noted a pink, macular, anterior inguinal rash with satellite lesions that suggested diaper candidiasis. You also noted a firm, nontender, 1 x 1-cm mass in the right inguinal area that was freely mobile. Otherwise, the exam was unremarkable with no other rash or palpable mass.

Having decided that the inguinal rash was most consistent with a candidal infection and the mass most consistent with a reactive lymph node, you prescribed nystatin cream to be applied until three days after the rash resolved. You instructed the mother to return to the clinic if either the rash or the lymph node persisted for longer than several weeks. Now, four weeks later, she informs you that the rash did indeed improve within a week after the nystatin was started. The mass in the right groin has not disappeared as expected, however; instead, she reports that its size has nearly doubled!

The patient takes her lumps On further questioning, you learn that the mass has been present continuously over the past four weeks. It has remained nontender and does not appear to bother the baby when it is touched or when it is manipulated during a diaper change. The patient has regular, soft, yellow-to-brown bowel movements several times a day; no blood is obvious in the stool. She has had no emesis and has been taking formula well—four ounces every four or five hours. She has still not had a fever.

Today, rectal temperature is 36.4° C; heart rate, 150/min; and respirations, 42/min. Weight is 3.91 kg (between 5th and 10th percentiles for postconceptual age, tracking along her expected growth velocity). Overall, she is alert and active and appears well. The anterior fontanelle is soft; sclerae are clear; and a pupillary red reflex is present bilaterally. Tympanic membranes are unremarkable. Oropharynx is moist with no enanthem. Neck is supple without lymphadenopathy. Examination of the heart and lungs is unremarkable. The abdomen is soft and nontender, with normal bowel sounds and no organomegaly.

When you remove the girl's diaper, you discern the firm, freely mobile subcutaneous mass in the right inguinal region, midline along the path of the inguinal canal. The mass is 2 x 1.5 cm, nontender on palpation and manipulation, nonpulsatile, and lacking overlying skin changes. Turning to the left inguinal region, you palpate two firm, mobile, nontender masses, each 0.5 x 0.5 cm. You find no other palpable masses. Surveying the baby's skin carefully, you find no rashes anywhere. There are several small areas of hypopigmentation along the inguinal creases and inner thighs.

You're intrigued by the persistence of these inguinal masses. As a first step, the mother—20 years old, with negative serologic tests for syphilis, hepatitis B, and HIV infection—recounts her daughter's medical history for you.

The patient was born prematurely, at 35 weeks' gestation, product of an identical twin pregnancy. Vaginal delivery was complicated by breech presentation; the baby required positive pressure ventilation for several minutes for respiratory distress. Apgar score was 3 at one minute, 7 at five minutes, and 8 at 10 minutes. She was admitted to the neonatal intensive care unit, where she remained stable on room air. A workup for sepsis was negative. After two days in the NICU, she was transferred to the well-baby nursery. Three days later, she was discharged with her twin sister.

The patient had unremarkable health maintenance visits at 1 and 3 weeks of age. At 7 weeks, she was evaluated by your partner for an erythematous, maculopapular, scaling rash on the face and shoulders. Hydrocortisone 1% cream was prescribed for presumed seborrheic dermatitis, and the rash resolved after one week. She had no further medical concerns until that 2-month-old visit. At each visit, weight was at the 10th percentile for postconceptual age.

Differential diagnosis of inguinal mass in an infant

The patient takes no medications at the moment, has no known drug allergies or exposures to animals or insects, and has not traveled outside her home city. Her mother confirms a negative family history of inguinal hernia, malignancy, immunodeficiency, and lymphoproliferative disorders. The girl's twin sister has had no medical problems.

You suspect that the inguinal mass may be something other than a simple reactive lymph node that arose from candidal dermatitis. Although it has the feel of an enlarged lymph node, its persistence and gradual growth over the past four weeks—despite resolution of the rash—is somewhat unusual.

You consider possible causes of isolated inguinal lymphadenopathy:
reaction to infection of the inguinal area or lower extremities by various pathogens
bacterial infection of the lymph node itself
malignancy, including lymphoproliferative disorders
an immunodeficient state
an infiltrative process

You are reassured that many of these potentially serious conditions are quite rare in young infants and most often present with generalized, rather than localized, lymphadenopathy. You also wonder whether the mass is truly a lymph node. Its proximity to the inguinal canal suggests herniated bowel, omentum, or, even, an ovary. Femoral artery aneurysm is also a consideration, although an unlikely one because the mass is nonpulsatile. Last, you keep in mind other soft-tissue masses, such as fibroma and neuroma.

Pictures don't lie
Faced with a large differential diagnosis, you decide that imaging may narrow your focus or even identify the problem definitively. You discuss the case with the medical center's pediatric radiologist and together decide that ultrasonography (US) would most likely provide useful information, and would also have the benefits of quick performance and low risk.2 Good fortune! There's time in today's radiology schedule to perform the study.

The baby heads off to the radiology suite with her mother. Twenty minutes later, the radiologist calls you back. The sonogram revealed an indeterminate, well-defined hypervascular mass in the region of the inguinal canal. The right ovary could not be visualized, but the mass doesn't resemble normal or enlarged ovary, bowel, or adenopathic tissue. The radiologist recommends that you strongly consider herniated omentum—even though a lymph node or herniated ovary cannot be excluded.

As you wait for your patient to return, you contemplate what to do next. The sonogram suggests a hernia but she has no history of intestinal symptoms, no findings of acute abdomen on exam, and no tenderness in the mass. This may not be a surgical emergency, but you're uneasy about the persistence and gradual growth of the mass over almost a month's time.

You reason that, because US could not definitively identify the mass, evaluation by a surgeon may be warranted sooner rather than later. By the time she returns, you've arranged an appointment in the pediatric surgery clinic later that week.

Time to go inside Two days pass. You take a call from the surgeon, who is as perplexed as you are. He has decided to admit the girl that evening for exploratory surgery. His thinking is similar to yours: The preoperative differential diagnosis favors hernia with entrapped ovary versus lymphadenopathy. Preoperative lab tests, including a complete blood count, comprehensive metabolic profile, and coagulation panel, have already been performed, and all results are within normal limits.

Later that day, the surgeon pages you at home. Exploration of the right inguinal area revealed no hernia, and he removed a single large lymph node. Your patient tolerated the procedure well and was discharged home from the recovery unit. The excised node has been sent for histopathologic analysis.

[Alon]

Bumpy ride ahead?
Ten days later, you receive the pathology report. Analysis of the lymph node revealed a paracortical histiocytic infiltrate with antigen markers indicative of an immature dendritic cell tumor, or early-stage Langerhan cell histiocytosis. The mass did hide something sinister after all! Although the report confirms your suspicions, you still have questions about the disease and the prognosis.

You discuss the patient's history with the pediatric oncologist who participated in the analysis of the node. He proposes, in retrospect, that the earlier facial and inguinal rashes may have been transient cutaneous manifestations of histiocytic disease. He further explains that excision of the lymph node may have been curative and that chemotherapy isn't indicated. Full-body imaging is necessary, however, to detect any occult tumors or skeletal lesions. He also recommends periodic close follow-up at your office and with the staff of the oncology clinic.

Langerhan cell histiocytosis (LCH) is a disorder of abnormal dendritic cell proliferation characterized by the presence of pathologic Langerhan cells (a type of dendritic cell, or histiocyte) within the centers of proliferation. Once known as histiocytosis X, LCH is a disease of diverse clinical presentations that encompass several conditions once thought to be distinct pathologic entities, including:
eosinophilic granuloma (localized bone lesions)
Hand-Schuller-Christian disease (multiple organ involvement characterized by a classic triad of skull defects, diabetes insipidus, and exophthalmos)
Letterer-Siwe disease (visceral lesions involving multiple organs)
Hashimoto Pritzker variant (congenital skin lesions that may resolve by 2 or 3 months of age but can recur).

Other histiocytic disorders feature proliferation of cell types other than the Langerhan cell, such as hemophagocytic lymphohistiocytosis (HLH), histiocytoma with macrophage phenotype, and Rosai-Dorfman disease (benign painless cervical lymphadenopathy with occasional extra-nodal involvement).3 Although LCH is the most common form of childhood histiocytosis, it is rare, affecting only one in 25,000 children annually.

The initial clinical presentation of LCH is varied and depends on the organ systems involved. In single-system LCH, only one organ or tissue is affected. For example, skin rash is the only manifestation in approximately 10% of cases. The rash often appears on the scalp and in flexural areas as reddish papules that progress to vesicles, pustules, or ulcers, followed by depigmentation and, eventually, healing. Often, the rash is mistaken for seborrheic dermatitis in newborns and infants.

LCH can also present as disseminated disease that affects multiple organs and tissues. Osteolytic bone lesions, manifesting as painful swelling, are present in more than 80% of cases. Other affected organs can include bone marrow, liver, spleen, lung, the central nervous system, and, more rarely, the gastrointestinal and genitourinary systems. Lymph node enlargement occurs in only 10% of cases, usually at initial presentation.

When LCH—in any form—is suspected, guidelines recommend a comprehensive workup of chest radiography and a skeletal survey, CBC, coagulation studies, and a liver function panel. A bone scan or body computed tomography scan may be helpful in certain cases. Definitive diagnosis requires biopsy and histopathologic analysis of a suspected lesion. Hallmarks of LCH include positive staining for the histiocytic markers CD-1a and S-100, as well as the presence of the Langerhan cell organelle, or Birbeck granule, within pathologic histiocytes.

Management of LCH depends on the extent of organ system involvement. Mild cutaneous disease may require only close observation or treatment with local corticosteroids. Severe cutaneous disease or multisystem disease may require extensive chemotherapy. Outcome correlates well with extent of disease involvement: Patients whose disease is more disseminated at presentation are at greater risk of relapse and death.

When do you watch the big picture? Your patient's case illustrates the challenge faced by a primary care provider when addressing common complaints with a wide differential diagnosis. Although careful history-taking and physical examination are often helpful in narrowing possibilities, potentially lethal rare diseases such as LCH may, in an early stage, offer few, if any, clues to help distinguish them from more common, less serious conditions. Our recommendation is not to be overzealous in pursuing zebras but to give thoughtful attention to the big picture when a constellation of symptoms evolves over time, and to take every opportunity to re-evaluate the differential diagnosis.

How quickly should workup of a suspected unusual condition proceed? In the case of your patient, the decision to send her for prompt surgical evaluation arose from two critical factors: persistence and gradual enlargement of the mass over more than one month despite resolution of the perineal rash and the inability of US to definitively identify its nature or visualize the right ovary.

Plan of observation You call the baby's mother to deliver the news. She is alarmed to hear "tumor" but relieved that the disease may have been cured by the surgery. You prepare the family for her upcoming visit to the oncology clinic. Information offered by the oncologist proves helpful as you carefully explain the details of the diagnosis and address her mother's concerns.

The family returns two weeks later for the twins' 4-month-old health maintenance visit. Your patient has done well since her operation and has no complaints. Her sister remains healthy too. The physical exam is unremarkable—you find neither a rash nor lymphadenopathy. She has had her first oncology clinic appointment and has undergone several imaging studies since you saw her last. A skeletal survey showed no bony lesions, and a full-body CT scan revealed only five small pulmonary nodules on the right side, suggesting asymptomatic pulmonary histiocytic disease. There is no remaining evidence on imaging of the left-sided inguinal nodes that you observed. Repeat lab tests were normal. The baby will alternate seeing you and then the oncologist at least once a month, and she will undergo serial imaging study and lab testing every three months. If she remains symptom-free from her knotty problem—if things go smoothly now!—the frequency of visits and testing may, gradually, decrease over time. Although the oncologist does not believe that the twin sister is at elevated risk of LCH, you will monitor her for clinical signs of disease at her well-child visits.

DR. FINK is a third-year resident in pediatrics, and DR. SERWINT is director of pediatric resident education and associate professor of pediatrics, both at Johns Hopkins University School of Medicine, Baltimore.

[Alon]

"That ever this should be!" The cellulitis that would not go away
The next patient waiting to be seen at your office today is a previously healthy 2-year-old girl who, according to her father when he called for an appointment, developed a "pimple-like" swelling on the lateral aspect of her right middle finger three days ago. Now, he tells you that he expressed "green fluid" from the lesion, and that erythema developed around it over the following two days. You examine the lesion, offer a diagnosis of cellulitis, and promptly begin treatment with amoxicillin-clavulanic acid.
"Full many shapes ... in crimson colours came"
Within one week of that office visit, however, the family notes progression of the process: Red streaks extend to the base of the middle finger, where a firm nodule develops. Despite multiple antibiotic changes that you institute—cephalexin, ceftriaxone, cefadroxil—the girl's condition does not improve over the ensuing four weeks.
You decide that the best course now is to consult an infectious disease specialist. A radiograph of the hand is taken and read as normal, excluding the possibility of chronic osteomyelitis. The consultant is concerned about the presence of an unusual, or resistant, organism. She institutes a trial of clindamycin but does not obtain material for culture, while you arrange referral to a plastic surgeon to have a biopsy performed.
"I fear thy skinny hand!"
At admission for biopsy, the parents confirm the absence of fever throughout the girl's illness. She has been acting well, they report, and has not had other skin lesions on the affected finger, including vesicles or puncture wounds by a thorn, or any musculoskeletal problems. The medical history includes infrequent episodes of otitis media and well-controlled atopic dermatitis. She does not take medications other than the antibiotics you recently prescribed. Immunizations are current. She is not allergic to medications.
The family reports that they have a pet cat but deny that the girl has been bitten or scratched by the animal. The family, including your patient, has not traveled outside the northeastern region of the United States or visited a farm in recent months. The girl has no known exposure to plants or moss, or to a person with tuberculosis. The family history is negative for chronic infection, immunodeficiency, and recurrent skin infection. One parent and one sibling have atopic dermatitis and food allergy. The patient is developmentally normal.
When you examine the girl, who is sitting on her mother's lap, she is somewhat apprehensive but does not appear to be in acute distress. She easily engages in conversation and play, and uses her right hand with little evidence of discomfort. You record vital signs: temperature, 36.2°C; heart rate, 112/min; respirations, 22/min; and blood pressure, 85/55 mm Hg. She is at the 25th percentile for height and weight for her age.
The physical exam is unremarkable except for the known findings on the right hand. The middle finger of that hand is swollen and erythematous but nontender. There is also as a pea-sized nodule in mid-dorsum on the right hand.
"Now wherefore stopp'st thou me?"
You consider the diagnostic possibilities—infectious and noninfectious. An osteoid osteoma of the finger, you recall, may not be painful and is not associated with fever. Ewing sarcoma and osteosarcoma typically do not occur in toddlers, and it is the long bones that are most often involved. Because of the girl's multiple lesions and the nonworrisome radiographic findings, you conclude that neither a benign nor a malignant tumor of bone is likely.
More and more, you are convinced that this patient has an unusual infection. (More and more, you feel like that Ancient Mariner of epic poetry—your search for a diagnosis like his voyage across a sea of unusual and inexplicable creatures and unexpected occurrences in "The Rime of the Ancient Mariner." Would the hero of Samuel Taylor Coleridge's puzzling 18th-century work know just how you feel as you confront this challenging investigation?)
You begin with a more mundane undertaking, however, by considering the most common organisms associated with cellulitis in an immunocompetent host. Staphylococcus aureus and Streptococcus pyogenes are the most common bacterial causes of cellulitis and osteomyelitis in children; methicillin-resistant S aureus should be considered, given its recent increase in incidence in children. Streptococcus pneumoniae and Haemophilus influenzae can cause facial and buccal cellulitis but probably not a hand cellulitis. Pasteurella multocida and Bartonella henselae are associated with cat bites and scratches. You even find cases in the literature of tularemia related to a cat bite. All these pathogens will need to be considered.
Now, what about infection with a more unusual organism? Cutaneous anthrax is commonly characterized by an eschar, absent here. As you think about the pattern of a line of nodules with erythema, you consider that the infection may fit the category of nodular or sporotrichoid lymphangitis—except that the nodules do not extend very far up her hand. Nocardia species are present in soil, and traumatic inoculation can cause this type of nodular cellulitis in a normal host. Sporotrichosis (caused by Sporothrix schenckii) is transmitted from plants (especially those with thorns), moss, and hay through small breaks in the skin; ulcerations of painless bumps are typically noted within three weeks of the onset of symptoms. Mycobacterium marinum can produce this picture after exposure to an aquatic environment.
As you are about to begin your investigation of possible uncommon sources of infection, the ID specialist heads out of the patient's room and stops you. She is smiling.

[Alon]

"As I was leaving the room," she says, "I said to her parents, 'I have one more question. Do you have a fish tank?' Mom replied, 'No. But we have a fish bowl. She helps us clean it every weekend.'"
"Yea, slimy things did crawl with legs"
The consultant orders a battery of laboratory tests, including a complete blood count. The white blood cell count is 11.4 X 103/mL, with a differential count of 39% polymorphonuclear neutrophils, 50% lymphocytes, 9% monocytes, and 2% eosinophils; hemoglobin, 12.3 g/dL; platelets, 343 X 103/mL. The C-reactive protein level is 0.1 mg/dL.
Biopsy of the skin in the affected area reveals a focal area of necrosis that is consistent with acute and chronic inflammation. A Gram stain is negative for bacteria, with 3+ polymorphonuclear neutrophils. Acid-fast staining is weakly positive for bacilli. Specimens are taken for bacterial and fungal cultures.
Those cultures eventually return negative. However, the acid-fast culture grows Mycobacterium marinum—and that jibes with the parents' comment about their fish bowl! The organism is sensitive to clarithromycin, rifampin, trimethoprim-sulfamethoxazole, and minocycline. The ID specialist is still smiling!
A course of clarithromycin and rifampin is started and, initially, the finger appears to improve. Approximately one month later, however, the nodule at the base of the finger enlarges and continues to do so. A three-month course of antibiotics ensues, but does not succeed in eliminating the nodule. The girl undergoes incision and drainage. Staining of material from the nodule reveals rare acid-fast organisms. A culture no longer grows M marinum, however.
She continues antibiotic therapy for a total of six months. One year later, your patient continues to do well and has no problem with recurrence.
"I fear thee, ancient Mariner!"
Like all Mycobacterium species, M marinum resides in macrophages. Hosts sequester the pathogen in granulomas, leading to long-term infection. M marinum infects fish and frogs, and was first isolated in saltwater fish in a Philadelphia aquarium in 1926. In 1951, the organism was identified in humans. The great majority of diagnoses are associated with fish tank, lake, pond, and swimming pool exposures. The annual incidence is 0.27/100,000 persons.
M marinum infection is most often limited to skin, where manifestations develop eight to 30 days after exposure at a site of minor skin disruption. Erythematous papules, nodules, or plaques erupt with or without purulent drainage. Infection may extend proximally in a nodular pattern. Although complications are unlikely, tenosynovitis, arthritis, bursitis, and osteomyelitis have been reported; disseminated infection is rare in an immunocompetent host. Treatment is based on the extent of disease. Generally, a prolonged course of an antimicrobial agent is administered. Surgical debridement may also be required for treatment to be effective.
In fish, M marinum infects numerous organs, resulting in a wasting syndrome, nodular lesions, ulceration and loss of scales, necrotic fins, discoloration, and rapid breathing. Because 25% of cases of M marinum infection occur in people who work with aquariums, a recent survey examined fish sales representatives' knowledge of M marinum infection (so-called fish tank granuloma) and their strategies for preventing infection. Of 40 salesmen, only six reported that they "knew the disease well"; 30 said they "knew about it, but ignored it." Thirty-three percent were introduced to the condition during their training to become an aquarium worker. Only eight were "concerned about human disease"; three of those had acquired an infection and five knew someone who had. Importantly, the majority of sales representatives immersed their ungloved hands in fish tanks daily, admitting that they discount the importance of fish tank granuloma.
"Water, water, everywhere"
When you see a patient with an erythematous skin lesion and fever, think of a staphylococcal or streptococcal infection first, of course, and treat the patient appropriately. In this time of increasingly common MRSA infection, remember to obtain material for culture whenever possible (see "Fighting a rising tide of MRSA infection in the young"). But if your patient fails to improve as expected, or if fever is absent, move to clarify the history and consider unusual organisms. That might help you swim toward the correct diagnosis—"alone on a wide wide sea!"

DR. YOUTH is director of the pediatric residency program at The Barbara Bush Children's Hospital at Maine Medical Center, Portland, Maine, where DR. JEWELL is a pediatric hospitalist and DR. MCCARTHY is director of the division of pediatric infectious diseases.

[Nancy]

ruptv,

Надо полагать, пациент не кардиологический?) Диафрагмальная грыжа?

[ruptv]

Сообщение от Nancy

ruptv,
Диафрагмальная грыжа?

Нет.
-----

[Nikulin Denis]

Первое о чем подумал: стеноз аорты

[ruptv]

Сообщение от Nikulin Denis

стеноз аорты

На основании чего?

[Nikulin Denis]

1. У ребенка имеются симптомы которые можно расценить как синдром малого сердечного выброса (среднее давление 29, циркуляторные нарушения акроцианоз, наполняемость капиллров 3 сек, форсирует дыхание)
2. Шум на основании сердца справа в проекции аорты громкий систолический
3. Ухудшение состояния после рождения возможно связано с закрытием ОАП

Первое впечатление в дифференциальной диагностике о ВПС с обструкцией выносящего тракта левых отделов с дуктус-зависисмым системным кровотоком. Возможны ли другие варианты - надо подумать.

[Nancy]

Жаль, что не угадала
Уважаемый Денис, мне тоже аортальный стеноз первым пришел в голову. Поэтому я о нем и...не написала. Показалось(?), что задачка с подвохом.

[ruptv]

Нет, это не стеноз.

Сообщение от Nikulin Denis

3. Ухудшение состояния после рождения

Возможно, я не совсем четко выразилась. Состояние ребенка было стабильным.

-----
На ЭКГ - подъем ST.
Может быть есть еще мнения? Интересно, что скажут педиатры...

[Nancy]

Мать ребенка здорова? Каков уровень ТТГ у матери и ребенка?

[Nikulin Denis]

Нужны дополнительные данные, чтобы дифференцировать заболевания
с поражением перикарда и плевры

Общий анализ крови с лейкоцитарной формулой, СОЭ
Рентген грудной клетки
ЭХО-КГ