1. Interesting news about mechanical kidneys currently being developed in the Economist's technology quarterly-- one model, which is basically a tiny wearable dialysis machine, runs diasylate in a coutercurrent flow across a membrane from blood, and then transfers the waste to a solid sorbent, which can be disposed of. In this way requiring much less fluid than a traditional dialysis machine. Another model being developed utilizes the body's heart rate instead of an internal pump, and involves constructing a filter out of layers of silicon alternating with kidney tubular epithelial cells (harvested from kidneys unsuitable for transplant). The advantage of this system is the ability to recover glucose and electrolytes from the filtered blood. Currently, it is not yet in trials in people, and when it is probably will require immunosuppressants still. But it's easy to imagine seeding these silicon filters with harvested stem or iPS cells from the recipient. (Economist)
2. Using an inhaler without a spacer decreases the efficacy by 40-60%
3. Immediately after an albuterol nebulizer treatment, the user's sats will drop due to a V/Q mismatch. Blood had been shunted away from previously hypoxic/non-aerated parts of the lung, and albuterol affects the b-2 receptors in blood vessels before the b-2 receptors in bronchial smooth muscle-- the reperfusion of blood happens before re-aeration of the same part of the lung.
4. The NNT to prevent acute mastoiditis in uncomplicated AOM (i.e. no fevers, otorrhea) is 4800. Antibiotics do not improve pain at 24 hours; the NNT to improve pain at 2-7 days is 20. Antibiotics do prevent tympanic membrane perforations (NNTB: 33) and contralateral AOM (NNTB: 11). However, the number needed to harm for side effects of antibiotics that otherwise wouldn't have happened (diarrhea, vomiting, rash) was 14. The antibiotics were most beneficial for kids under 2 years with bilateral AOM, or with both b/l AOM and otorrhea. Other kids do not benefit. (cochrane review)
5. If you suspect CHF in a child-- look for S3, S4, hepatic enlargement, edema (can be hard to visualize-- look for recent acute weight changes). CHF due to congenital heart disease will often manifest as failure to thrive, poor weight gain, stressful feeding (sweating, panting, increased work of breathing)
6. For MRSA pneumonia: clinda has much better lung penetration than vanc. And it's delicious! Like drinking sugar water. Linezolid works but is really, really expensive.
7. Hematopoietic stem cell transplant is one of the few treatments that have shown to be of benefit in Krabbe's disease; transplants are from umbilical cord blood, either autologous or matched donors. Transplants do not cure the disease, but they do slow the progression/onset of the neurodegeneration. Of note, transplants are generally regarded as most effective before the onset of symptoms. A 2005 study in the NEJM of 22 infants, half who received transplant and half who didn't, showed that at 3-4 years postop, among those who got a transplant before the onset of symptoms, survival was 100% and nearly all had age-appropriate development of cognitive and language skills, although about half showed delays in motor skills. Among those who got a transplant after the onset of symptoms, survival and outcomes were comparable with untreated controls. (NEJM)
8. Steroids do not affect renal outcomes in patients with HSP. 
9. In the winter, bronchiolitis is usually RSV. In the summer, it's more likely to be parainfluenza, rhino/corona, adeno, or influenza.
10. Someone who is very tachypneic may not be able to take PO meds (i.e. steroids)-- they may aspirate.

1. Exudative pharyngitis in a child, think: group A strep, strep pneumo, EBV, adenovirus. Adenovirus also causes conjunctivitis, otitis, gastroenteritis, and hemorrhagic cystitis. If a child does not have any URI symptoms, its more likely to be strep, but if they have any URI symptoms, its much more likely to be viral, and hence antibiotics would not be indicated. Testing for ASO is not particularly useful, since there is a high carrier rate in the general population.
2. The reason for giving high-dose amoxicillin for acute otitis media is to overcome whatever resistance the bacterial strains may carry.
3. The treatment for scarlet fever is bicilin (pencillin g benzathine) a long-acting injectable penicillin. IM, in the gluteus. If you give this, watch them for 30 minutes for an adverse drug reaction: rigors, nausea. The pills (penicillin V) are large, hard to swallow, and taste bad.
4. Otitis externa is associated with pain upon movement of the pinna, and is common after swimming. Cover for pseudomonas. If you're sure its not a ruptured OM, treat with ciprodex drops: ciprofloaxcin + dexamethasone. You don't want steroids in the inner ear.
5. Ruptured tympanic membrane after AOM is not that uncommon: treat with ofloxacin.
6. For nasal congestion in a child, nasal steroids (flonase) are first-line, oral antihistamines (zyrtec) are second-line. You can also run saline through their nose, and then suction it out with a bulb-- the saline loosens up the mucus.
7. Acute liver failure in pediatric populations: think viral hepatitis, wilson's disease, toxins. To track liver synthetic function, follow coags. For this reason, do not transfuse FFP/cryo or platelets in someone who's coags are down unless they have clinically significant bleeding. Use vitamin K as your main/only treatment. In liver failure, gluconeogenesis is the last synthetic function to go.
8. You cannot diagnose (or rule out) Wilson's disease based on serum Cu or ceruloplasmin levels. Serum Cu is not always elevated, and ceruloplasmin is not always low-- in fact, it is an acute phase reactant, so it may be high or normal. Cu+ is toxic to RBCs, so you will get hemolytic anemia in wilson's.
9. Kayser-Fleischer rings are the best way to diagnose wilson's disease: it will be present in 50% of kids with abdominal signs of wilson's, and nearly 100% of kids with neurological or psychiatric symptoms of wilson's.
10. Serum VMA is not a sensitive test for pheochromocytoma. You want to use serum and urine metanephrines and normetanephrines, which have sensitivities in the 95-100% range. Normetanephrines are more likely to be elevated in extra-adrenal pheo.

1. Insulin requirements in a:
-prepubertal child: 0.5-0.7 units/kg/day: half given as long-acting "baseline" insulin such as lantus, and half given as a short-acting "bolus" insulin before meals such as novalog.
-pubertal child: 1-1.2 u/kg/day. Puberty is a time of growth, and sex-hormones also create a degree of insulin resistance.
2. Goal blood sugar: 80-140 in kids (even though non-diabetic kids run 60-100), 200s in babies. You really don't want kids running hypoglycemic, since its such a critical time for brain development/growth, you'd rather run high sugars than low.
3. 1800 rule: 1800/(total insulin daily dose units) = number of glucose units one unit of insulin will drop your blood sugar, if your sugar is over 150. For example, if you are aiming for a goal sugar of 125, and you are currently at 225, and your normal daily dose is 36 units/day, then 1800/36=50, so one unit of insulin will drop you 50 points, so you need 2 units of insulin-- assuming you are not about to eat a meal, in which case you will also need to correct for the sugars you plan on ingesting. This rule breaks down a little at very high blood sugar levels-- insulin is less effective then, and you will need more.
4. If you have ketones in your blood, that means you do not have enough insulin. 
-small amount of ketones: bolus +5% of your total daily dose of insulin on top of whatever correction
-medium amount of ketones: +10%
-large amount: +15%
5. When correcting blood sugars, assuming you have calculated the correct dose of insulin to give, wait 2 hours before checking your sugars again and giving more correction, if needed. Do not give lots of insulin in a short period of time. With insulin, its always better to start lower and titrate up.
6. For emergency fluid resuscitation, it's good to use a short, fat line-- i.e. a jugular venous line, rather than a longer line like a femoral.
7. Diagnosing wheezing in a child:
-Asthma: +viral trigger, +responds to albuterol, +history of atopy
-VIW/RAD: +viral trigger, +responds to albuterol, no history of atopy
-Bronchiolitis: +viral trigger, doesn't respond to albuterol, no history of atopy
8. Usually asthma is not diagnosed until someone is 3-4 years old, because of the frequency of RAD/VIW in children. Generally, kids who just have VIW will grow out of it by age 4-5 That being said, if a young kid has been hospitalized more than 4 times in the last year for wheezing/difficulty breathing and they have a personal of atopy and a family history of asthma, they need BID inhaled corticosteroids whether you want to call it asthma or not.
9. Most common causes of headache in pediatric populations: migraine/tension headaches, ICP (2/2 tumor/bleed), pesudotumor cerebri
10. Sunscreen needs to be applied 15-60 minutes before sun exposure because it needs time to form a protective film across your skin.

1. 2012 IgNobel prize for neuroscience went to Craig Bennett, Abigail Baird, Michael Miller, and George Wolford [USA], "for demonstrating that brain researchers, by using complicated instruments and simple statistics, can see meaningful brain activity anywhere — even in a dead salmon." From the poster:
Subject. One mature Atlantic Salmon (Salmo salar) participated in the fMRI study. The salmon was approximately 18 inches long, weighed 3.8 lbs, and was not alive at the time of scanning.
Task. The task administered to the salmon involved completing an open-ended mentalizing task. The salmon was shown a series of photographs depicting human individuals in social situations with a specified emotional valence. The salmon was asked to determine what emotion the individual in the photo must have been experiencing.
Analysis. Voxelwise statistics on the salmon data were calculated through an ordinary least-squares estimation of the general linear model (GLM). Predictors of the hemodynamic response were modeled by a boxcar function convolved with a canonical hemodynamic response. A temporal high pass filter of 128 seconds was include to account for low frequency drift. No autocorrelation correction was applied.
Results: A t-contrast was used to test for regions with significant BOLD signal change during the photo condition compared to rest.... Several active voxels were discovered in a cluster located within the salmon’s brain cavity (Figure 1, see above). The size of this cluster was 81 mm3 with a cluster-level significance of p = 0.001.... Out of a search volume of 8064 voxels a total of 16 voxels were significant.... Identical t-contrasts controlling the false discovery rate (FDR) and familywise error rate (FWER) were completed. These contrasts indicated no active voxels, even at relaxed statistical thresholds (p = 0.25).
Moral of the story: when working with fMRI data, generously apply robust statistical methods to reduce false positive findings!
2. There is not enough work for vets-- the current estimate is that supply exceeds demand by the equivalent of over 11,000 full time jobs. (NPR)
3. A meta-analysis in the british medical journal of 72 RCTs (total n>10,000) trying to determine the effects of IV iron vs oral iron/placebo. IV iron more effective in increasing Hgb and reducing risk of transfusions (esp when given with epo or in patients with baseline low ferritin), but also associated with an increased risk of infections. Overall, there was no difference in mortality. I guess anemia of chronic disease really does serve a purpose... evolution wins again. (BMJ)
4. In a dog bite in a child: recently published RCT data shows that given good quality debridement, irrigation, and antibiotic treatment, there is no difference in outcomes between suturing the wound or leaving it to heal by second intention. Cosmetic outcomes are better with primary closing, thus facial wound should be sutured. Time to treatment of <8 hours from bite was associated with a deceased risk of infection. (Injury) Other studies have shown that location on the hand is also associated with an increased risk of infection. Organisms to cover: Pasturella (clinda does not cover, amp does), anaerobes, gram positives from skin flora of victim. First line antibiotic: penicillin+b-lactamase inhibitor (unasyn, zosyn, augmentin). Other options: 3rd gen cephalosporin plus flagyl, floroquinolone plus flagyl, carbapenem monotherapy. Amp/clinda is a good option if you're worried about MRSA, amp covers pasturella and some anaerobes, clinda hits anaerobes gram+ incl MRSA. Don't forget to get clinda sensitivities before you send someone home on it!
5. Cat/human bites in a child: should both be debrided to remove all nidus of infection, irrigated, and left to heal by second intention. Human bites: eikenella is one of the more common organisms that you'll find.
6. Catfact: 89% of cat bites are provoked.
7. Erysipelas vs Cellulitis: Same organisms cause both, treat with amox. Erysipelas are more superficial infections: red, more acute-onset with systemic symptoms. Lesion is less warm, more raised/well-demarcated/clear borders. Often on face, although you can get cellulitis on your face. Cellulitis is deeper, more chronic; in the mid-face area can progress to sinusitis/orbital cellulitis and finally to (rare) cavernous sinus thromboses. Skin infection in an immunocompromised person: think pseudomonas.
8. Neonatal hgb/hct reaches a nadir at 6-8 weeks of life, as lysis of fetal hemoglobin cells begin. At birth, 2/3 of neonatal blood is HgF and 1/3 is HgA. By 4 mos of age, the adult variant predominates.
9. Mammalian bites: 60-90% dogs, 5-20% cats, rest is other animals (raccoon, bats, etc)
10. Increased ICP 2/2 pancytopenia 2/2 lupus: possibly due to microclots decreasing CSF drainage (a-cardiolipin associated with increased risk of clots), or perhaps an increased CSF production in response to severe anemia.

1. LVAD increases overall survival as well as quality of life in comparison to medical management (MM) in patients with NYHA class IV heart failure (REMATCH trial, n=129). LVAD was associated with a 48% reduction in risk of death from any cause; survival at 1 year was 52% in LVAD vs 25% in MM, at 2 years was 23% and 8%. (NEJM)  Another paper doing sub-group analysis of REMATCH trial data showed clear survival benefit in patients receiving inotropic therapy. In patients not receiving inotropic therapy (n=38), survival for LVAD/MM were 57%/40% at 1 year and 22%/16% at 2 years, but the survival differences were not statistically significant (probably underpowered). (circulation) My interpretation is that the benefits of LVAD are clear among people who are very sick (i.e. NYHA IV and inotrope dependent). Among those who are less sick (NYHA class IV not inotrope dependent), LVAD probably still offers a survival benefit, but it is less clear, and must be weighed against the costs (emotional and financial) of going through a big, painful open heart surgery. Note: all-cause mortality includes mortality from complications of LVAD.
2. ECMO use in neonates is associated with sensorineural hearing loss, with a mean incidence of 7.5% (range 3-21%), compared to an overall rate of 1-3% in NICU survivors. This hearing loss is frequently progressive (70%), and delayed-onset (50%). One study (n=111) found the factors associated with an increased risk of developing SNHL after ECMO are diagnosis of congenital diaphragmatic hernia, extended time on ECMO, and extended use of aminoglycosides. (Pediatrics) My interpretation is that ECMO is probably associated with an overall hypoxia, relative to not being on ECMO (ie. having heart and lungs that work), and perhaps the nerve cells of the cochlea are more sensitive to hypoxia than other cells. CDH is associated with lung hypoplasia and sometimes severe hypoxia, so that fits in well, and aminoglycosides are known ototoxins, so that also fits.
3. Factors associated with congenital/neonatal SNHL: 
-Infection: prenatal CMV infection (most common overall cause of congenital SNHL), toxo, rubella, syphillis, bacterial meningitis (s.pneumo, hib).
-trauma: head trauma, temporal bone fracture
-toxin: aminoglycoside, loop diuretic, lead, arsenic, radiation.
-low birth weight (<1500g), apgar scores less than 4 at 1 min and 6 at 5 min, hyperbilirubinemia requiring exchange transfusion
-hypoxia: mechanical ventilation >5 days, ECMO
-stigmata of conditions associated with SNHL: renal abnormalities, craniofacial abnormalities
4. Waardenburg syndrome: partial albinism (white forelock), SNHL, broad mandible.
5. Sandifer syndrome: GERD, hiatal hernia, hypotonia, spasmodic torticollis and dystonia: spasms last 1-3 minutes and occur multiple times a day, sometimes associated with feeding. Can be mistaken for infantile spasms or epilepsy, esp if the GI symptoms are not clear.
6. Mucopruluent drainage from one eye in a neonate, in the absence of conjuncitivitis, with a history of increased tears in one eye: think nasolacrimal duct obstruction, rather than GC/chlamydia (must have conjunctivitis). These things can also progress to infected cysts that require IV antibiotics.
7. Isolated throbocytopenia in a child with a prodome of viral illness (esp VZV, EBV, CMV), think ITP. CBC with diff will reveal normal morphology of platelets and RBCs, and normal levels of other blood cells. If there is a decrease in WBC/RBCs as well, or LAD/splenomegaly, do a BM biopsy to r/o aplastic anemia or cancer. Among people with ITP, 50% resolve in 1 month, another 30% in 6 months. Treatment is to watch them carefully to prevent injury. Pharmacologic treatment is not shown to improve outcomes, and is only indicated if platelets are <20,000, there are severe bleeding symptoms (intracranial- occurs in 1% of ITP, massive GI bleed) or if a safe home environment cannot be guaranteed. Tx options: IVIg, systemic steroids, splenectomy (need lifelong vaccines against strep pneumo). Platelet transfusion only if there is life-threatening bleeding.
8. Drugs that cause decreased platelets: Heparin (HIT), bactrim, quinine, quinidine, cimetidine, benzos, penicillin.
9. Deep brain stimulators are now incorporating recording as well as stimulating functionality, moving closer to the ultimate goal of a closed-loop, self-adjusting stimulator system. (medgadget)
10. Closed-loop insulin management is finally here! A team at the University of Virginia have linked a blood glucose monitor to an insulin pump via smartphone (bluetooth). The first trial enrolled 20 people with type 1 diabetes for 42 hours, and found the system had a 97% uptime. Paving the way for new trials. (c/o medgadget) (diabetes care) 

1. Human sleep cycles are affected by the phase of the moon, even when they cannot see the light: researchers at the University of Basel re-analyzed old data attained during a sleep study where people were kept inside for long stretches of time, and found that people slept worse on nights when there was a full moon, even when they couldn't have possibly been affected by the light since they were inside for days on end. Quote: "Electroencephalography showed that the volunteers slept, on average, 20 minutes less around the time of the full Moon. It also took them five minutes longer to get to sleep, their delta activity (a measure of how deeply they were sleeping) was 30% lower than at other times, their level of melatonin, a sleep-related hormone, was reduced, and they reported, subjectively, that they had not slept as well as usual. Nor was any of this connected, in female volunteers, with their menstrual cycles." (economist)
2. Hep B immune globulin is expensive, so you don't want to administer it to a neonate unless you have positive Heb B serologies from mom during pregnancy. You don't want to give it to every infant of a mom with unknown Hep B status.
3. To determine whether blood is maternal or fetal, you can use the following tests:
-Apt test: generally a test of secreted blood to see if it is maternal or fetal, i.e. 3rd trimester vaginal bleeding to r/o vasa previa (although if you really suspect previa, don't take the time to do this test, run to surgery) or spit-up from a neonate to figure out if its swallowed maternal blood or infant GI bleed (usually brighter red in color). Technique: put blood into tap water to lyse cells. Spin down, take supernatant (where Hb is), add strong base (KOH, NaOH). Fetal hemoglobin will remain red/pink, but maternal hemoglobin will denature to hematin (turn brown-yellow).
-Kleihauer-Betke: a test of maternal blood to ascertain presence of fetal blood cells. Good to determine extent of maternal-fetal transfusion. Strong acid denatures maternal Hb, but not fetal. So smear the blood on a slide, put it in an acid bath, look under the microscope. Maternal cells will be "ghosted" pale, while fetal red cells will be bright red and normal.
4. Vascular lesions on back/lower back of neonate: blanching suggests vascular malformation (hemangioma). The presence of 2 or more abnormalities on the lower back of a neonate (hair tuft, abnormal/asymmetrical cleft, distorting mass/lipoma, stain, skin tag, dimple) is more suggestive of an occult spinal dysraphism and should be worked up with MRI; single abnormality is overwhelmingly unlikely to be associated with spinal findings and should be worked up with u/s.
5. Caput succedaneum: "boggy" area on baby's head, common finding after delivery, happens from the pressure of baby's head against the dilating cervix (tourniquet effect). Subcutaenous/extraperiosteal collection of serosanguinous fluid, crosses sutures/midline. Resolves spontaenously after a few days.
6. Subgaleal hemorrhage: bleeding under the periosteum, above the skull. Associated with vacuum operative deliveries. Does not cross suture lines. Can lead to increased risk of jaundice. Feels fluctuant upon palpation. Will absorb slowly over weeks-- do not aspirate as it carries a risk of infection/abscess formation. Get imaging if new neuro sx appear. After it absorbs, it may calcify, leaving a relatively softer center and giving the impression of a depressed fracture.
7. The three most common causes of ataxia in a child:
--Postinfectious acute cerebellar ataxia (usu s/p varicella, GI or respiratory infection.)
--Ingestion: barbituates, alcohol, antifreeze, lead, CO, benzos, benadryl
--Guillain-Barre: acute symmetrical ascending demyelinating autoimmune/inflammatory disease. Associated with campylobacter infection.
8. Can't-miss causes of ataxia in a child: 
--Infection (meningitis/encephalitis/sepsis)
--Brain abscess
--Tumor
--Bleed (2/2 trauma, NAT)
--Post-infectious autoimmune (ADEM)
--ICP
--Ischemia (clot, moya moya)
9. Other auses of ataxia in a child:
--Tick paralysis/lyme disease
--Neuromuscular disease (muscular dystrophy, injury, myasthenia)
--Vestibular system disease (labrynthitis- often s/p URI, menieres)
--Basilar migraine
--Seizure, post-ictal
10. Posterior fossa mutism-- happens in 8-38% of children who receive resection of posterior fossa tumors, specifically medulloblastoma. Usually transient, speech returns in 1-15 weeks. Almost always accompanied by cerebellar ataxia.

1. Flocks of migratory birds avoid crashing into each other upon landing by always landing in alignment with earth's magnetic field. Quote: "The latest research suggests that birds detect magnetic fields in two ways. One relies on small pieces of magnetite (a magnetic iron oxide) lodged in their beaks, or inner ears, or both. The other employs a magnetism-sensitive chemical reaction in their eyes, allowing them to “see” the Earth’s magnetic field, probably as bright and dark spots superimposed on their visual fields" (Economist)
2. Fascinating show from radiolab about Emergence, the property of order emerging spontaneously from a large group of disorganized individuals (i.e. acting without central leadership). Emergence is addressed in the context of swarm animals (ants, fireflies) and in neurons in the brain. A theory is proposed that brain cells of different senses firing at the same time/rate is the foundation for coherent, integrated ideas-- i.e. the smell, appearance, feel, taste of coffee are integrated to make the concept of "coffee" (Radiolab)
3. GDM leads to organomegaly as well as macrosomia: hepatosplenomegaly and enlargement of the interventricular septum. The latter reverses upon reduction of insulin levels. At birth, IDM will have hypoglycemia, as well as certain electrolyte abnormalities: hypomagnesemia and hypocalcemia (present 10-50% of the time, depending on the study, often associated with hyperphosphatemia-- possibly a function of polycythemia and cell lysis? possibly a function of decreased PTH from increased Ca in utero?). Finally, the polycythemia (hi glucose -> hi metabolism -> hypoxia -> polycythemia) leads to paradoxical iron deficiency elsewhere in the body as iron is shunted into RBC synthesis.
4. Ceftriaxone displaces bilirubin from albumin, increasing blood [UCB] and leading to gallstones and biliary sludging.
5. Red cell factors that will lead to increased hemolysis in a neonate, and thus more jaundice: 
-Structural: spherocytosis
-Enzyme: G6PD, pyruvate dehydrogenase def.
-Immune: ABO/Rh
6. GI factors that will lead to decreased excretion of UCB: meconium ileus, duodenal atresia
7. Physiological (i.e. nonpathological) causes of jaundice: 
-Race (see previous post)
-Physiologic: infants have b-glucoronidase, their RBCs have shorter lifespans, more ineffective hematopoesis takes place, liver function is not as good, UGT1A1 activity is not as good.
-Breastfeeding (early jaundice): the low amount of breastmilk/colostrum made in the first days post-partum mean that exclusively breastfed babies will be slightly food/drink deprived in the early days of life. Less albumin/protein, and increased hepato-GI cycling of GI contents, leading to increased b-glucorindase activity.
-Breastmilk: (later jaundice) some substance in the breast milk increases UCB in the blood. Perhaps there's something that inhibits UGT1A1, or has b-glucorinidase activity (or is b-glucoronidase)
7. Jaundice due to increased CB is rarer than UCB, and merits an extensive workup for biliary obstruction (choledochal cyst, progressive familial intrahepatic cholestasis, biliary atresia) or genetic diseases (dubin-johnson, rotor)
8. Catfact: cats can jump up to 5x their height in a leap.
9. Insects must beat their wings hundreds of times a second in order to stay aloft, which is far too fast to be explained by calcium-regulated myosin-actin mechanisms. A new paper out in Nature examining bumblebees flying by x-ray scattering found data to support a theory that bee muscles act via oscillatory stretch activation, where stretching of one muscle by a counterpoint/yoked muscle flexing leads to increased myosin head affinity, leading to reflexive contraction. This process is Ca independent. (Nature)
10. Significant steps were made in figuring out why anti-EGFR drugs cause those awful pruritic rashes, published this week in Science. From the abstract: "We established a parallel mouse model by ablating EGFR in the epidermis. These mice developed skin lesions similar to the human rash. Before lesion development, we detected increased mRNA expression of chemokines in the skin associated with early infiltration of macrophages and mast cells and later infiltration of eosinophils, T cells, and neutrophils. As the skin phenotype evolved, changes in blood counts and circulating chemokines reproduced those seen in the gefitinib-treated patients. Crossing the mutant mice with mice deficient for tumor necrosis factor–α (TNF-α) receptors, MyD88, NOS2, CCR2, T cells, or B cells failed to reverse the skin phenotype. However, local depletion of macrophages provided partial resolution..." (Science)