Severe HTG may rarely develop as a result of autoimmune hyperlipidemia without the need for the presence of other aggravating factors. The original descriptions were in patients with myeloma sometimes associated with tuberous or planar xanthomas and in most cases the antibodies were found to react with VLDL and other lipoproteins or with heparin (16). Subsequently reports of autoimmune hyperlipidemia have been described in patients with lupus erythematosus and Sjögren’s syndrome and in several cases antibodies to LPL were identified. More recently 6 cases with CS were found to have antibodies to GPIHBP1, four of whom had either lupus or Sjögren’s syndrome (17) attesting to the importance of this protein for LPL activity. In addition among 33 patients with unexplained CS and no genetic mutations, one was found to have GPIHBP1 autoantibodies (18) and in another recent report a patient with newly acquired CS resistant to standard treatment was found to have GPIHBP1 antibodies that disappeared with normalization of HTG after treatment with the anti-B cell monoclonal antibody rituximab (19). These findings point to the “GPIHBP1 autoantibody syndrome” (17) as a novel form of autoimmune hyperlipidemia. Moreover, GPIHBP1 autoantibodies were identified as the cause of the CS in a patient treated with interferon for multiple sclerosis (20). Interferon is known to increase the risk of autoimmune disease and discontinuation of the interferon caused the antibody to disappear. Interferon does increase triglyceride levels modestly, attributable to increased hepatic lipogenesis and decreased LPL activity (21), but it is unknown whether the small percentage of individuals receiving interferon who develop severe HTG (22) do so because of the occurrence of GPIHBP1 antibodies.

Moderate to severe HTG is common in glycogen storage disease type 1 (glucose-6-phosphatase deficiency) and is attributed to shunting of the increased glucose-6-phosphate into hepatic lipogenesis and increased VLDL levels. CS is uncommon. Post-heparin LPL activity does not appear to be decreased (23). In a series of 230 cases with classical glycogen storage disease type 1, 72.7% had triglyceride levels >3 mmol/L (>270 mg/dl) but only 3 individuals had a history of pancreatitis (24).

Based on a representative US sample, 22.5% of people with diabetes have triglyceride levels >200 mg/dl and those with HbA1c >7.0% have higher triglyceride levels than those with HbA1c < 7.0% (38), although relatively few have triglyceride levels >500 mg/dl and even fewer >1,000 mg/dl. The HTG is thought to be due to TGL overproduction in type 2 diabetes as a result of increased release of free fatty acids from adipose tissue, enhanced hepatic lipogenesis and increased secretion of large TRL particles by the liver as well as the intestine as a consequence of insulin resistance and hyperglycemia (39). Although LPL activity is known to be decreased in insulin deficient states (40), recent developments in the understanding of LPL regulation have also shown that insulin resistance leads to increased APOC3 and ANGPTL4, which reduce LPL activity and TGR remnant particle clearance (41–43).

The prevalence of diabetes in reports of cohorts with severe HTG varies widely, ranging from 25% to 76% in reported studies (25–37). In a recent large US County Health System survey of risk factors for very severe HTG (triglyceride levels >2,000 mg/dl) in whom most of the identified patients were Hispanic, diabetes was present in 76% (78 out of 103 cases) and was also frequently present in those with other secondary factors, such as obesity, excessive alcohol intake and use of triglyceride-raising drugs (25). The patients with diabetes were said to be uncontrolled with a median HbA1c of 10.4% [interquartile range 8.2–12.6%], 3 out 78 had type 1 diabetes and 21 had diabetic ketoacidosis. Although the majority of patients with diabetes and severe HTG have type 2 diabetes, severe HTG may occur in poorly controlled insulin deficient diabetes (44), where the combination of LPL deficiency and continued dietary fat intake are key factors. In a survey of risk factors for severe HTG in which 75% had diabetes, 82% of these were receiving insulin treatment (37).

In an early report, the observation that index patients with triglycerides >2000 mg/dl had considerably higher triglyceride levels than their first- degree relatives with HTG, prompted a search for other potential causes of HTG in the index patients (45). Ninety-four percent were found to have a potential secondary cause, 63% of which was uncontrolled or untreated diabetes. Moreover, in the original report of the CS in diabetes, treatment of the diabetes led to a marked reduction in triglyceride levels, but in no cases did it fall into the normal range (46). This provides support for the notion that the pathophysiological disturbances in triglyceride metabolism that accompany diabetes can interact with polygenic forms of HTG to result in triglyceride levels that exceed the saturation point for triglyceride removal from plasma (47), thereby leading to severe HTG and the MFCS (9).

Overall, the findings support the contention that patients with diabetes who develop HTG-induced acute pancreatitis are typically in poor glycemic control despite being on insulin treatment. However, the extent to which poor glycemic control is a defining feature of concurrent diabetes and severe HTG has not been studied, nor is it known whether the genetic variants presumed to be present in these cases are any different from those in the general population with severe HTG. More detailed studies of the genetic, pathophysiologic and clinical characteristics of these individuals are needed in order to be able to better target at risk subgroups for more effective intervention. Complicating this area is evidence that genetic variants linked to triglyceride levels may be associated with lower glucose levels in the population (48). On the other hand, in a Japanese study the majority of those with severe HTG who developed pancreatitis did not have diabetes (32). It is likely that other risk factors for severe HTG as well as race/ethnicity and sociocultural differences may interact with less severe diabetes in many circumstances. Finally, the extent to which the CS contributes to the increased frequency of acute pancreatitis in diabetes is unclear (49).

Surveys of severe HTG indicate that up to 50% of patients are obese (25–37, 50). However many of these subjects have other risk factors for HTG such as diabetes or alcohol excess, and obesity is associated with an increased occurrence of biliary disease which is a common cause of pancreatitis, so that the exact role of obesity in the genesis of CS and risk of pancreatitis is unclear. The effect of obesity on triglyceride metabolism is thought to be more strongly related to visceral obesity than to total adiposity (50). Increased release of free fatty acids from visceral fat increase VLDL production especially in the presence of insulin resistance which is present in the majority of obese individuals (51). While LPL is increased in obesity its functional regulation is impaired, with recent evidence suggesting that increased APOC3 and ANGPTL3 in obese individuals may contribute to reduced LPL activity and reduced remnant particle clearance (52). These abnormalities at the least provide the substrate for aggravation of mild to moderate polygenic forms of HTG to exceed the saturation point for triglyceride removal and hence lead to MFCS.

Population studies demonstrate that alcohol consumption is linearly associated with plasma triglyceride levels and was 0.37 to 0.46 mmol/l (33–41 mg/dl) higher in subjects drinking more than 5 alcoholic drinks per day (53). In addition, alcohol consumption is more likely to cause HTG in overweight than in normal weight individuals (54). Alcohol has been shown to stimulate VLDL secretion possibly through enhanced assembly of VLDL, combined with elevated hepatic delivery of free fatty acids as a results of enhanced adipose tissue lipolysis (55, 56). In addition, there is evidence for a prolonged postprandial lipemic response due to decreased LPL activity although LPL activity has also been reported to be increased by chronic alcohol intake (57)

Excessive alcohol intake was found to be present in 9% to 50% of individuals with severe HTG and may coexist with obesity and diabetes (25–37) and in studies of the prevalence of secondary factors in cohorts with very severe HTG, excessive alcohol intake was the second most common secondary factor after diabetes (25, 45), accounting for 11% of potential causes of HTG in addition to a familial disorder (45). In a Dutch retrospective survey of subjects with severe HTG, excessive alcohol intake was more likely to be present in those with triglyceride levels >15.9 mmol/L (>1450 mg/dl) and especially >26.8 mmol/L (>2365 mg/dl) (58). Moreover, in patients with alcoholic HTG, triglyceride levels fell to within normal limits after alcohol withdrawal in only 1/5 subjects with triglyceride levels >1000 mg/dl at baseline (59), suggesting that they had an underlying cause of HTG in addition to excess alcohol intake. There is evidence to indicate that the effect of alcohol on triglycerides is influenced by genetic polymorphisms (60–62). These findings are again consistent with the notion that MFCS is a result of the co-existence of an underlying polygenic form of HTG and another potential cause of HTG, in this case, alcohol.

Chronic kidney disease is known to be associated with HTG in relation to the rise in serum creatinine and to albuminuria (63). The mechanism for HTG is still somewhat unclear but appears to be due mainly to reduced clearance of TRL and related to elevated APOC3 and the presence of insulin resistance known to accompany chronic renal insufficiency (64). There also is experimental evidence for reduced GPIHBP1 playing a role (65). The prevalence of severe HTG in patients with chronic kidney disease is unknown, although in many reports 5% to 10% of those with severe HTG or HTG-associated acute pancreatitis had chronic renal failure (25–37), possibly as a result of their also having background genetic HTG.

Severe HTG is known to occur in both children and adults with nephrotic syndrome and is thought to be due to reduced LPL activity due to decreased GPIHBP1 and increased ANGPTL4 and APOC3 levels (66). Since high-dose corticosteroids are often used in these patients, this may be a confounding factor, A case report of severe HTG in a child with the nephrotic syndrome was reported to be associated with an APOE mutation (67).

Although the typical lipid disorder in hypothyroidism is hypercholesterolemia, elevated triglyceride levels may also occur, and this is thought to result from an increase in ANGPTL3 and ANGPTL8 levels (68). In addition, hypothyroidism is associated with reduced hepatic lipase activity and slowed remnant clearance (69). Triglyceride levels are usually considerably lower than those seen in MFCS, but again the co-existence of hypothyroidism with a polygenic form of HTG could lead to saturation of triglyceride removal mechanisms, leading to severe HTG (9, 25).

Thiazide diuretics and beta blockers have both been shown to increase triglyceride levels by 0% to 50% in older studies (70, 71) although there is evidence that these effects wane with time (72). The mechanism for the thiazide effect on triglyceride levels is unknown. In the case of beta blockers there is evidence that non-selective beta blockers reduce LPL activity (73, 74). The use of indapamide instead of thiazides and of selective beta blockers with intrinsic sympathetic activity rather than non-selective beta blockers in potentially susceptible individuals are favored alternatives since they have little triglyceride raising potential (75). The relevance of these findings to the development of severe HTG and the CS in patients with HTG who use these agents is unclear particular since both agents have been reported to cause pancreatitis (73, 75). Whether pancreatitis associated with their use is due to severe HTG in these cases is unknown.

Oral estrogen administration increases triglyceride levels 10% to 50% (76). Importantly the effect varies by dose and type of estrogen preparation, is generally smaller with estrogen/progestin combinations (77) and not observed with transdermal estrogen. The increase in triglyceride level is thought to be due to VLDL overproduction resulting from activation of the estrogen receptor (78). It is unclear whether estrogen alters LPL in humans. Although estrogen administration or gestational HTG does not comprise a significant number of cases in surveys of factors causing CS or HTG pancreatitis, there are many case reports of estrogen-induced CS and of HTG pancreatitis and these presumably are due to exacerbation of underlying monogenic or polygenic HTG (79). Given the widespread use of estrogenic compounds in women and in transsexual men and the prevalence of mild to moderate HTG in the population at risk, individuals with baseline HTG should be identified when using these agents. Among estrogen receptor modulators, tamoxifen used in the management of breast cancer as well as clomiphene, used for infertility have both been reported to cause severe HTG and pancreatitis (80, 81), while raloxifene has minimal effects in normotriglyceridemic women (82), but may cause severe HTG in women with a history of estrogen-induced HTG (83). Importantly estrogen therapy is often not discontinued by clinicians in women with a history of HTG-induced pancreatitis (84). Endogenous hyperestrogenemia may cause severe gestational HTG and pancreatitis in women with HTG in the second or third trimesters of pregnancy, where it can be difficult to control, particularly in subjects with monogenic CS or with other aggravating factors such as diabetes (85, 86).

Although not widely appreciated, corticosteroids raise triglyceride levels modestly (87). Most studies reporting this finding have been in cohorts with chronic steroid-requiring conditions, such as asthma, rheumatoid arthritis and post-transplantation, where findings are often confounded by other factors. The most definitive study was in patients with Cushing’s syndrome who had mild hypertriglyceridemia compared to controls that resolved after curative surgery (88). Kinetic studies revealed that the elevated triglyceride level was due to VLDL overproduction likely due to insulin resistance rather than a removal defect since post-heparin LPL levels were normal. However reports of HTG-induced acute pancreatitis occurring in 2 patients with a history of fibrate-treated HTG receiving short courses of high-dose corticosteroids point to the likelihood that patients with polygenic HTG may be at risk for exacerbation of their HTG following this commonly used treatment in clinical practice (89). This is supported by data from a survey of severe HTG in Japanese subjects in whom 27% of cases were associated with corticosteroid use (32).

Treatment of human immunodeficiency virus with protease inhibitors has been shown to increase triglyceride levels by 22% in a 12 month study, although this does not apply to atazanavir (90). There is evidence for both VLDL overproduction as well as reduced LPL activity (91, 92). In addition, studies have demonstrated an interaction between the triglyceride raising effect and APOC3 gene polymorphisms (93). There have been case reports of chylomicronemia (94) and of HTG-induced pancreatitis attributable to protease inhibitors (95), but in surveys of drugs associated with severe HTG or HTG-induced pancreatitis, the use of protease inhibitors is not mentioned suggesting that these agents are uncommon causes of CS.

Both first and second generation antipsychotics may cause HTG (96). Among atypical antipsychotics, olanzapine and quetiapine increase triglycerides modestly in relation to weight gain and hyperglycemia, with risperidone, ziprasidone, and aripiprazole having milder effects (97). There is evidence that these agents increase hepatic synthesis and triglycerides as well as inhibit LPL (98). Olanzapine and quetiapine have both been reported to cause severe HTG and occasionally pancreatitis (99, 100), presumably when used in subjects with genetic forms of HTG.

Hypertriglyceridemia has also been described as occasionally occurring with antidepressant medications. Perhaps best described is the HTG in patients treated with the tetracyclic antidepressant mirtazapine, in which 6% of patients were reported to have triglyceride levels >500 mg/dl (101) and there are several reports of HTG-associated acute pancreatitis occurring in patients treated with mirtazapine (102).

Cyclosporine, a calcineurin inhibitor, and sirolimus and everolimus (also used as a chemotherapeutic and in drug-eluting coronary artery stents), both inhibitors of the mammalian target of rapamycin (mTOR), have all been shown to cause HTG. Tacrolimus has minimal effects on triglyceride levels. The effect of cyclosporine on triglyceride levels is relatively mild, increasing by 50 mg/dl in one study (103), but triglyceride levels were >400 mg/dl in 14% to 25% of patients with sirolimus in one report of renal transplant recipients (104) and in another in 28% of sirolimus and 35% of everolimus treated patients (105). Both sirolimus and everolimus have been associated with the development of severe HTG (106, 107). These agents increase APOC3 levels and inhibit LPL (108, 109).

Isotretinoin used in treatment for acne was found to cause mild to moderate and occasionally severe HTG and rarely HTG pancreatitis. As a result its use is largely limited to treatment of severe acne vulgaris. In one study isotretinoin increased mean triglyceride levels by 50% with 17% of individuals having triglyceride values >200 mg/dl. (110). Retinoids increase APOC3 levels but do not appear to directly inhibit LPL in humans (111). Occasional reports of isotretinoin-associated HTG-induced pancreatitis have been published, presumably in subjects with underlying genetic predisposition for HTG. In several cases patients were taking other medications that may cause HTG, and these reports are complicated by the fact that isotretinoin may also cause idiosyncratic pancreatitis (112). It is therefore prudent to check for baseline HTG before prescribing these agents.

Propofol is a hypnotic agent used to sedate ventilated patients in the intensive care unit. It is administered intravenously as a lipid emulsion and causes HTG in approximately 20% of patients (113). In one study, 18% of patients had triglyceride levels >400 mg/dl and in 20% of these the triglyceride levels were >1,000 mg/dl and half of this group developed acute pancreatitis (114). Since propofol infusions may deliver 24 to 48 g of fat/2 h this may be playing a significant role in the development of HTG, particularly in patients with acquired or underlying defects in triglyceride clearance (115). Propofol has also been shown to impair fatty acid oxidation (115).

In a retrospective analysis of records from a large tertiary care children’s hospital ~0.1% had triglyceride levels >2,000 mg/dl and 36% of this group developed acute pancreatitis (116). Of these 14% had FCS, 19% had type 2 diabetes, 11% type 1 diabetes, 28% had acute lymphoblastic leukemia, and 14% had undergone organ transplantation on immunosuppressive medications. The children with acute lymphoblastic leukemia all received treatment with L-asparaginase and high dose corticosteroids. The development of severe HTG occurs in 4% to 19% of these children (117) and is thought to be mainly due to L-asparaginase which decreases LPL activity (118).

If diabetes is present it is essential that optimal glycemic control should be achieved without delay. Many overweight or obese patients with severe HTG have marked insulin resistance and may require high doses of insulin to adequately control the hyperglycemia. In patients imbibing excessive alcohol it is essential that an effective alcohol cessation program be initiated. Medications that aggravate HTG should be discontinued if possible and replaced by alternatives that have less triglyceride raising properties, as discussed above. If not, they should be used at as low a dose as possible. FPLD should be ruled out by careful physical examination for lipodystrophic features. Patients without aggravating factors should be suspected of having FCS or unusual causes of CS such as autoimmune hyperlipidemia, particularly if they are relatively young.

For the rare patient with FCS, it is essential that both dietary animal and vegetable fat intake be permanently restricted to <5% to 10% of calories (approximately 10–20 g of fat) and medium chain triglyceride (MCT) may be used to further limit chylomicron formation. In MFCS, a similar degree of fat restriction together with management of aggravating factors should be used until triglyceride levels fall to <1,000 mg/dl, at which time the diet can be somewhat liberalized, since long-term adherence to an extremely low fat diet is very difficult. Patients who are overweight will benefit from weight reduction measures and most patients, especially those with diabetes, should benefit from a high fiber diet that is low in refined carbohydrate. Although the potential benefits of bariatric surgery for severe HTG have not been formally evaluated, bariatric surgery has been shown to lower triglyceride levels, and could be considered in obese individuals (124).

Pharmacologic triglyceride lowering in MFCS and FPLD begins with fibrate therapy. This should be initiated together with management of aggravating factors and dietary fat restriction. In the United States fenofibrate is preferred to gemfibrozil because of its once daily dosage and lower rate of rhabdomyolysis when combined with statin therapy. The dose should be reduced in renal failure. The goal is to lower and maintain the triglyceride level to <500 mg/dl and this usually requires long-term therapy with the fibrate. If this is not achieved with the fibrate alone, high dose (2 grams twice daily) omega 3 fatty acid preparations can be added. Triglyceride lowering doses of niacin (2 grams daily) is a third option although this may worsen diabetes control. Individuals with FCS typically do not respond to conventional pharmacotherapy. The use of the pancreatic lipase inhibitor orlistat to reduce fat absorption was recently shown to be effective in two patients with FCS (125) and the long-term use of the microsomal triglyceride lipase inhibitor lomitapide, was effective in preventing pancreatitis in a patient with FCS although it did result in cirrhosis (126). Although there are no controlled studies, plasmapheresis may have a role to play in patients with extreme HTG and recurrent pancreatitis (127) or in pregnant women with FCS (128).

The mainstay of CVD prevention in HTG patients is statin therapy. In addition, statins may have additive modest triglyceride-lowering effects in patients with mild to moderate HTG mainly by enhancing remnant clearance. While fibrate monotherapy may reduce CVD risk, fibrates have not been shown to add benefit to statins overall, although those with moderate HTG and low HDL-C may possibly benefit (129). A recent clinical trial with the synthetic omega 3 fatty acid icosapent ethyl was shown to further reduce CVD events in hypertriglyceridemic statin-treated patients with CVD and/or diabetes (130).

There has been considerable interest in developing new triglyceride lowering agents using approaches targeting LPL and its regulators because of the central role they play in TRL metabolism. None have yet reached routine clinical usage, but several appear to have considerable potential in future management of the CS and prevention of its complications.