Recent Diagnosis of Lupus and Severe Hypertriglyceridemia in a 12-year-old Girl

A 12-year-old girl with a recent diagnosis of systemic lupus erythematosus (SLE) presents to the emergency department (ED) with 1 week of central “stabbing” chest pain and severe fatigue. She denies orthopnea and shortness of breath but reports she must sit to rest after walking only short distances due to her fatigue. In the ED, her vital signs show a heart rate of 70 beats/min, normal oxygenation, and elevated blood pressures in the 160s/90s (mm Hg). Her mentation is appropriate. Her height is 62.2 in (158 cm), weight is 107.6 lb (48.8 kg), and BMI is 19.9, which are consistent with previous growth trends in the 75th percentile. Apart from 2 weeks of joint pain and fatigue before her new SLE diagnosis, she was generally healthy without other medical problems and actively engaged in sports. On examination her lungs are clear bilaterally with no increased work of breathing. Cardiac examination reveals a regular cardiac rhythm without murmurs or friction rub. Her abdomen is soft and nontender in all quadrants, with no abdominal distention or hepatosplenomegaly. She has 1+ pitting edema in the lower extremities but no joint swelling or rashes.During her previous admission for SLE diagnosis 4 weeks ago, she was found to have arthritis, pancytopenia, and class III lupus nephritis that was diagnosed on kidney biopsy. At that time, her creatinine level was 0.8 mg/dL (70.72 µmol/L), alanine aminotransferase (ALT) level was 143 U/L (2.39 µkat/L), and hemoglobin level was 9.3 g/dL (93 g/L), and urine studies showed 2+ protein, 3+ blood, and a urine protein/creatine ratio of 1.45 mg/dL (reference range, <0.2 mg/dL). Thyrotropin level was normal. Liver ultrasonography was performed due to an elevated ALT level and showed minimal liver echogenicity thought to be secondary to diffuse inflammation from SLE. An acute hepatitis panel was obtained to rule out infectious etiologies and was negative. An echocardiogram was also obtained as part of our local protocol for baseline evaluation for new diagnosis of SLE and was normal. After treatment with 3 days of 1,000 mg of intravenous methylprednisolone, she was discharged on mycophenolate mofetil (MMF), prednisone, and hydroxychloroquine with significant improvement in her symptoms.On this presentation to the ED, laboratory values include the following: sodium, 124 mEq/L (124 mmol/L); potassium, 4.2 mEq/L (4.2 mmol/L); chloride, 100 mEq/L (100 mmol/L); bicarbonate, 9 mEq/L (9 mmol/L); blood urea nitrogen, 22 mg/dL (7.85 mmol/L); creatinine, 0.5 mg/dL (44.20 µmol/L); and ALT, 17 U/L (0.28 µkat/L). Urinalysis shows 1+ protein, 3+ blood, and a urine protein/creatine ratio of 5.14 mg/dL.During the laboratory draw, the nursing team notes a pink, milky appearance to her blood (Fig 1), and lipid studies are ordered. They show a triglyceride (TG) level of 26,541 mg/dL (299.91 mmol/L) and a cholesterol level of 1,392 mg/dL (36 mmol/L). Her initial laboratory values are reported again with correction for lipemic interference and now show a sodium level of 134 mEq/L (134 mmol/L) and a bicarbonate level of 14 mEq/L (14 mmol/L). Lipase level is normal at 42 U/L (0.70 µkat/L). Hemoglobin level cannot be reported due to lipemic interference. Hematocrit concentration is 26% (0.26), and platelet count is 268 × 103/µL (268 × 109/L). Her electrocardiogram shows normal sinus rhythm, and chest radiography shows no evidence of cardiomegaly, consolidation, or pleural effusions. Cardiology is consulted and recommends no additional evaluation or interventions at this time. She is given hydralazine for elevated blood pressure.A thorough review of her medical record is performed, and it is noted that the patient had no evidence of lipemic interference in any of her specimens on the day of discharge from her admission last month. Because it is not part of the routine new SLE diagnosis protocol at our institution, TG and cholesterol studies were not performed during that admission. The patient had a follow-up visit 1 week after discharge. Laboratory values obtained at that time included a hemoglobin level that was unable to be reported by the laboratory. It was noted in the comments of the laboratory report that this was due to lipemic interference; however, this was not seen by any providers.In the ED, the patient is started on intravenous fluids for hyperhydration and is admitted to the rheumatology service for ongoing management and evaluation.After hyperhydration, the TG levels decrease to 23,134 mg/dL (261.41 mmol/L). She is started on gemfibrozil and transferred to the ICU for continuous insulin and dextrose administration (see Fig 2 for trends of TG and lipase levels during admission). It takes 29 hours to achieve the goal insulin rate of 0.1 U/kg per hour due to hypoglycemia. Despite TG level improvement, she develops new abdominal pain. Repeated lipase level is drawn and has now increased to 9,810 U/L (163.83 µkat/L). She is diagnosed as having acute pancreatitis. In the setting of capillary leak, she develops hypotension and pulmonary edema requiring bilevel positive airway pressure respiratory support. A central line is placed for centrifugal therapeutic plasma exchange (TPE). TPE is performed with 5% albumin as replacement fluid. The effluent is grossly lipemic (Fig 3). Her TG level decreases from 6,182 mg/dL (69.86 mmol/L) to 1,071 mg/dL (12.10 mmol/L) after the procedure. The insulin infusion is discontinued on day 4, when TG level decreases to 410 mg/dL (4.63 mmol/L). Lipase level is 2,720 U/L (45.42 µkat/L) at this time.Severe hypertriglyceridemia (TG levels >1,000 mg/dL [11.30 mmol/L]) is rare in children and often occurs due to an inability to break down triglyceride-rich lipoproteins (TRLs). (1) TRLs are synthesized in the liver as very low-density lipoproteins (VLDLs) or are absorbed via the intestines as chylomicrons. TRLs are then hydrolyzed by lipoprotein lipase (LPL) into free fatty acids. (2)(3)(4) Disruption of the process can lead to hypertriglyceridemia and/or chylomicronemia, both of which cause high serum TG levels. (5) The differential diagnosis for hypertriglyceridemia includes secondary etiologies such as diabetes mellitus, obesity, nephrotic syndrome, hypothyroidism, and medication toxicity. Primary genetic etiologies or antibody-mediated disruption of the TRL processing pathway are less common but are more frequently associated with severe hypertriglyceridemia.There are 2 main categories of genetic hypertriglyceridemia: monogenic disorders known as familial chylomicronemia syndrome and multifactorial chylomicronemia syndrome. (6)(7)(8) In this patient, primary genetic causes were unlikely because there was no family history of dyslipidemias, heart disease, or physical findings such as xanthomas.Patients with SLE treated with corticosteroids can have significantly higher TG levels compared with patients without exposure: 158 ± 11 mg/dL (1.79 ± 0.12 mmol/L) vs 87 ± 0.8 mg/dL (0.98 ± 0.01 mmol/L). (9) Given recent treatment with high-dose corticosteroids, this was considered in the differential diagnosis; however, her TG levels were much higher than previously reported with corticosteroids. (6)(9)(10)(11).Severe hypertriglyceridemia can be seen with anti-GPIHBP1 (glycosylphosphatidylinositol-anchored high-density lipoprotein binding protein 1) antibodies in SLE. (12)(13) The membrane protein GPIHBP1 anchors LPL to the capillary endothelium. (6) Inhibition of GPIHBP1 interaction with LPL, via anti-GPIHBP1 antibodies, causes severe hypertriglyceridemia by lowering LPL activity and reducing TG metabolism. (3)(12)(13) Our patient’s, anti-GPIHBP1 assay was strongly positive for anti-GPIHBP1 antibodies.SLE is an autoimmune disease characterized by antibody formation against self-proteins causing multiorgan damage. This may happen through the formation of antinuclear antibody or anti–double-stranded DNA antibody. Other, less routinely tested autoantibodies, such as anti-GPIHBP1 and anti-LPL, may also be seen. A recent review describes 22 patients with anti-GPIHBP1 antibodies and hypertriglyceridemia. The authors reported a maximum TG concentration of 7,971 mg/dL (90.07 mmol/L), three times less than our patient’s presentation. (13)Severe hypertriglyceridemia (TG level >1,000 mg/dL [>11.30 mmol/L]) can lead to pancreatitis and is associated with significant mortality. (14) This is presumed secondary to toxic damage to pancreatic endothelium and acinar cells from free fatty acids and due to ischemic damage from hyperviscosity and capillary obstruction. (15)Medical management for severe hypertriglyceridemia includes fasting, hydration, and continuous insulin. Insulin enhances LPL activity and decreases VLDL production in the liver. (14) TPE is considered when hypertriglyceridemia induces pancreatitis. According to the American Society for Apheresis guidelines, hypertriglyceridemia-induced pancreatitis is a category III indication to perform TPE. Category III indicates a need for individualized decision making because the optimum role of TPE is not established. (16) There are no recommendations regarding TPE for hypertriglyceridemia without pancreatitis, which influenced the decision not to immediately initiate TPE when the patient presented to the ED with chest pain and was first discovered to have hypertriglyceridemia but did not have pancreatitis. Although TPE can lower TG concentrations by 49% to 97% per session, its effect on outcomes compared with medical management is not clear. (16) In a randomized trial comparing TPE with continuous insulin in 22 adults with a Marshall Organ Dysfunction Score indicating “mild” hypertriglyceridemia-induced pancreatitis, the TPE group had a faster decrease in TG in the first 24 hours (67% drop vs 53% drop) and fewer days until the TG level was less than 1,000 mg/dL (<11.30 mmol/L) (1 day vs 2 days). However, the rate of progression to “severe” pancreatitis was similar between groups. (15)With TG levels 4 times higher than most other studies, the degree of extreme hypertriglyceridemia present in our patient is quite unusual. (14)(15) There are few data describing TPE versus medical management for prevention of pancreatitis with extreme hypertriglyceridemia.Medical management and TPE are temporary measures for managing hypertriglyceridemia. She was started on a strict, low-fat diet. However, definitive treatment requires addressing the underlying etiology. Reports of patients with anti-GPIHBP1 antibodies have shown success with immunosuppression, including rituximab, MMF, or corticosteroids. (13) This patient presented with severe hypertriglyceridemia despite immunosuppression with high-dose corticosteroids, hydroxychloroquine, and MMF. Immunosuppressive treatment was escalated such that she received an additional 3 days of 1,000 mg of intravenous methylprednisolone and rituximab, an anti-CD20 monoclonal antibody therapy that depletes B lymphocytes, which decreases autoantibody production. Although rituximab use in SLE is off-label, it is a common and appropriate option for refractory disease. Miyashita et al (13) reported normalization of serum TG levels in 5 patients with anti-GPIHBP1 antibodies after administration of rituximab. One of these patients had only transient improvement with pheresis but reported sustained improvement after rituximab therapy.After initial TPE and positive anti-GPIHBP1 results, additional glucocorticosteroids, 2 doses of rituximab, and the aggressive low-fat diet led to sustained normalization of TG levels. Her admission was complicated by severe pancreatitis with pseudocyst, which was managed conservatively and required no additional interventions. Additional complications included superficial and deep venous thrombosis requiring therapeutic low-molecular-weight heparin, anemia requiring recurrent transfusions, esophageal candidiasis, and persistent secondary hypogammaglobulinemia.Her SLE is now well controlled with hydroxychloroquine, mycophenolate sodium, and a prolonged prednisone taper with ongoing immunosuppression from rituximab as evidenced by ongoing B-cell depletion. She continues gemfibrozil use and a low-fat diet, with normal TG levels. Long-term prognosis remains unknown as cases of patients with anti-GPIHBP1 antibodies have been described only within the past 6 years. Ideally, with good SLE control, production of anti-GPIHBP1 antibodies will decrease, permitting an increased amount of fats in her diet in the future. Maintaining normal TG levels should decrease her risk of cardiovascular disease or metabolic syndrome.Our patient presented with an uncommon complication of SLE, which serves as a reminder that SLE can impact almost any organ system. In patients with lupus, new symptoms or abnormal laboratory findings should be thoughtfully explored. Early detection of hyperlipidemia may prevent complications. In our laboratory, the complete blood cell analyzer uses sodium lauryl sulphate to lyse red blood cells to measure hemoglobin. The sodium lauryl sulphate then binds to heme groups to form colored complexes, which are measured photometrically. Lipemia can interfere with this measurement due to increased turbidity, which causes light scatter. There is no known threshold at which lipemia interference occurs as different combinations of VLDLs, TGs, and chylomicrons lead to different amounts of turbidity. In adult populations, it is relatively common for lipemic interference to occur with laboratory draws given the higher prevalence of high TG levels (estimated to occur in 0.5%–2.5% of all laboratory samples). (17) However, in children, if hyperlipidemia or lipemic interference with laboratory results is unexpectedly discovered, this should be thoroughly investigated.Systemic lupus erythematosus can have a wide spectrum of clinical presentations due to autoantibodies against almost any organ system. Providers should be diligent in the evaluation of patient complaints even if symptoms are not typical of systemic lupus erythematosus.If unexpected lipemia or lipemic interference is discovered in a pediatric blood sample, further investigation is warranted.Secondary pancreatitis due to hypertriglyceridemia is associated with significant morbidity, and treatment should be individualized considering risks and benefits of medical management versus therapeutic plasma exchange.The authors thank Sahar Rooholamini, MD, Elena Griego, MD, Corrie McDaniel, DO, and Jimmy Beck, MD, for reviewing a draft of this manuscript and contributing a general pediatrician’s perspective.