Lipoatrophic diabetes: Endocrine dysfunction and the response to control of hypertriglyceridemia

Lipoatrophic Diabetes: Endocrine Dysfunction and the Response to Control of Hypertriglyceridemia Norman G. Soler, Jacobo W o r t s m a n , and Inder J. Chopra Endocrine function was studied in a 24 year old female with lipoatrophic diabetes (LD). Baseline endocrine studies (serum triglycerides: 2600 mg/dl) demonstrated hyperprolactinemia (serum prolactin 51 ng/ml}, increased ACTH levels, absence of suppression of ACTH to a high dose of dexamethasone which suppressed serum cortisol normally and, hyperresponsiveness of TSH to stimulation with TRH. Thyroid hormone levels (total and free fraction) were essentially normal. Major metabolites of thyroid hormone (T 3, rT 3, 3, 3 ' - - T 2, and 3', 5 ' - - T 2) were also normal and exhibited a normal response to the administration of L-thyroxine and propylthiouracil. Exchange of 84% of the patient's plasma resulted in a decrease in serum triglycerides (700 mg/dl) which was followed by a rebound to the original level in seven days. After the sixth plasmapheresis serum triglycerides stabilized at less than 1000 mg/dl. Plasmapheresis was associated with the appearance of amenorrhea and galactorrhea, also hypertension and proliferative retinopathy developed during this therapy. Repeat endocrine function studies (serum triglycerides: 700 mg/dl) showed a further rise in serum prolactin ( > 1 6 0 ng/ml), persistence of abnormal ACTH secretion and normalization of TSH responsiveness. Lipoatrophic diabetes is associated with abnormal central endocrine function but appropriate peripheral target gland secretion. A course of plasmapheresis improves the hypertriglyceridemia but not the endocrine dysfunction. In this patient with LD the most important side effect of plasmapheresis was the development of cardiovascular complications.

I P O A T R O P H I C DIABETES is a rare condition characterized by typical somatic features and associated severe abnormalities of carbohydrate and lipid metabolism. 1 Patients with LD have absence of subcutaneous fat, hyperglycemia and hypertriglyceridemia and commonly demonstrate resistance to the action of insulin. In a large series of patients with this disease there was considerable heterogeneity in the mechanism for the insulin resistance. 2 It has been postulated that the primary cause of LD is a disturbance of the hypothalamo/pituitary axis. 3'4 An endocrine etiology is suggested by the demonstration of a lack of growth hormone (GH) release during slow wave sleep 5 and following stimulation with arginine 5'6 or insulin; 7 the detection in peripheral plasma of the releasing factors for corticotropin, follicle stimulating hormone and melanocyte stimulating hormone; 6 the detection of a peptide with lipid mobilizing activity in the urine; 8'9 and finally, the reversal of the metabolic abnormalities of LD with pimozide, a central dopaminergic blocking agent. 1° Moreover, some patients with LD have a history suggestive of encephalitis or abnormal pneumoencephalograms. 3 We are presenting the results of endocrine testing in a patient with lipoatrophic diabetes with special emphasis on the dynamics of thyroid function; in addition, the effect of hypertriglyceridemia upon endocrine function was specifically evaluated, by repeating the endocrine studies after a series of plasma exchanges had decreased the serum concentration of triglycerides.

L

CASE REPORT This white female patient first presented because of acute symptoms of diabetes mellitus in 1971, at 14 yr of age. She also reported heat intolerance, excessive perspiration and episodes of palpitations. Her menarche had occurred two years earlier. On examination she Metabolism, Vol. 31, No. 1 (January), 1982

had a generalized absence of subcutaneous fat and hepatosplenomegaly and investigations revealed hyperglycemia without ketonuria, hyperlipidemia and proteinuria. A diagnosis of lipoatrophic diabetes was made based on the clinical features and the histological finding of lack of subcutaneous fat. Photographs of the patient from infancy and from her childhood years indicated that the lipoatropbic diabetes was of the acquired type. Initial investigations carried out in 1971 included normal liver enzymes, a serum creatinine of 0.8 mg/dl, an endogenous creatinine clearance of 108 ml/min, proteinuria of 4 to 6 g daily, and a serum albumin of 3.5 g/dl with total serum proteins of 6.3 g/dl. Serum complement levels were normal. An intravenous pyelogram showed enlarged kidneys. On renal biopsy one segment demonstrated hyalinized glomeruli with the remaining tissue displaying normal architecture, lmmunofluorescent studies on the biopsy tissue were negative. A liver biopsy was essentially normal except for increased glycogen vacuoles. Rectal and gingival biopsies did not show evidence of amyloid. Following the diagnosis of diabetes the patient was started on treatment with a 2000 calorie diet and conventional insulin (beefpork). However, she developed localized skin allergy at the site of the insulin injections and after two months was switched to pork insulin. Clofibrate (2 g daily) was used unsuccessfully to control the hypertriglyeeridemia and was stopped within 12 mo. From the time of diagnosis in 1971 the patient was intermittently disabled by episodes of severe abdominal pain, associated with marked hypertriglyceridemia (>2500 mg/dl). During these episodes serum amylase levels increased up to 1500 units/dl. The patient was referred to us during one of these episodes in July 1977. We evaluated her after the acute episode had subsided. On physical examination, height was 160 cm and weight 54 kg. At rest and in the fasting state the pulse rate was 84/rain and

From the Department of Medicine, Division of Endocrinology, Southern Illinois University School of Medicine, Springfield, HI., and the Department of Medicine, Division of Endocrinology, UCLA Center for Health and Science, Los Angeles, Calif. Received for publication February 10, 198 I. Address reprint requests to N. G. Soler, M.D., S1U School of Medicine, Department of Medicine, P. O. Box 3926, Springfield, IlL 62708. © 1982 by Grune & Stratton, Inc. 0026-0495/82/3101-0005501.00/0 19

20

SOLER, WORTSMAN, AND CHOPRA

regular, blood pressure 150/80 mmHg and temperature 37°C. There was abgence of subcutaneous fat in th e face, limbs and trunk. The thyroid gland was smooth and soft, twice normal in ~size and without any bruits. The liver was diffusely enlarged with a total span of 18 cm in the midclavicular line. The spleen was palpable 4 cm below the costal margin. Neurologic examination demonstrated absence of deep tendon reflexes in the lower extremities and sensory impairment to both vibration and position. Retinoscopy showed normal fundi. There was no evidence of hirsutism, galaetorrhea, acanthosis nigricans, lymphadenopathy or cliteromegaly. Between July 1977 and December 1979 hypertriglyceridemia persisted, abdominal pain increased in frequency and severity, blood pressure increased to 160/90 mmHg and background diabetic retinopathy became apparent. These changes developed in spite of adherence to a 2000 calorie diet and strict control of the diabetes with up to 320 units of insulin daily (120 units before breakfast, 80 units before lunch and 120 units before the evening meal). Glycosylated hemoglobin (HbA~) levels were maintained consistently below 10% (normal 5.5%-8.5%). MATERIALS

AND

METHODS

Standard radioimmunoassays were used for measurement of thyrotropin stimulating hormone ( T S H ) , pro]aetin, growth hormone ( G H ) , adrenocortieotropin ( A C T H ) , luteinizing hormone ( L H ) and

follicle stimulating hormone (FSH). Measurements of serum thyroxine (T4) , free T4 (FT4), triiodothyronine (T~), reverse T 3 (rT3), 3,3'-T2, and 3',5'-T2 were carried out using methods that have been previously described,ll't2,~ Serum was carefully separated from the creamy layer for the determination of FT4 by equilibrium dialysis. Serum glucose, cholesterol and triglycerides were measured on a Technicon autoanalyzer and plasma aminoacids were determined on an aminoacid analyzer (Beckman, model 121-M). Blood samples were collected after a 12 hr overnight fast and before the morning insulin dose was administered. Aminoacid analyses were performed in duplicate and corrected for dilution due to lipemia. Plasmapheresis was carried out by intermittent flow using a cell separator (Haemonetics model 30); four liters of plasma were exchanged over a period of 4 hr. Albumin kinetics were studied with J25I labeled albumin (Malinckrodt, St. Louis, MO). Radiolabeled albumin was administered intravenously 30 min before the first plasmapheresis. Blood samples

for determination of radioactivity were obtained immediately before and after plasmapheresis was completed, and every 24 hr for a period of 8 .days thereafter. Urine was simultaneously collected, in 24 hr specimens, for 1251 measurements. Estimation of kinetic parameters was carried out according to the procedure of Weinstock et al. j4 Albumin kinetics were also determined in a 28-yr-old female patient with nephrotic syndrome, secondary to insulin dependent diabetes mellitus of 20 years standing.

RESULTS

Basal Studies In January 1978 the following results were obtained in blood collected in the fasting state: glucose 260 mg/dl, C-peptide 5.5 ng/dl (normal 0.9-4.2 ng/dl), triglycerides 2400 mg/dl (normal <150 mg/dl) and free fatty acids 0.6 meq/1 (normal 0.10-0.9 meq/1). Following a 50 g glucose load, peak serum glucose was 465 mg/dl at 90 minutes and peak C-peptide was 7.0 ng/dl at 120 min. Insulin binding by the patient's serum was normal: 12.5 and 11.7 mU/ml for beef and pork insulin, respectively (Lilly Laboratories, Indianapolis, IN). A tomogram of the pituitary fossa was normal. Serum LH, FSH, prolactin, GH and TSH and plasma ACTH were studied in October 1979, when the serum triglycerides were grossly elevated (2600 mg/dl), and the results are shown in Table 1. The basal GH level was normal but serum prolactin was increased. Serum cortisol was slightly elevated and ACTH levels were markedly increased; dexamethasone (2 mg q 6 hr × 16) suppressed the serum cortisol but plasma ACTH remained elevated. TSH response to thyrotropin releasing hormone (TRH 500 /zg/IV) was exaggerated. Thyroid status was initially evaluated in February 1978. The basal metabolic rate (BMR) was 14%

Table 1. Endocrine Studies Before and After Plasmapheresis in a Patient With Lipoatrophic Diabetes Before Plasmapheresis LH (IrnU/ml) FSH (ImU/ml) Prolactin (ng/ml) Cortisol (#g/d0 Basal After Dex:l: ACTH (pg/ml) Basal After Dex GH (ng/ml) TSH (l#U/ml) Basal 30 rain after TRH 60 min after TRH *Follicular phase. tAM. :l:Dexamethasone 2 rng q 6 hr × 4 days.

After Plasmapheresis

Normal

8.4 11.1 51

17.6 22.3 > 160

5-20* 5-20* <20

28.1 4.0

20.0 1.0

5-20 t

273 255 4.0

685 676 3.9

< 100

6.4 >40.0 32.6

4.2 14.2 12.6

< 10 <20 <20

< 10

21

LIPOATROPHIC DIABETES

Table 2. Protocol Baseline On T4 0.3 mg daily 7 days On T 4 0.3 mg daily PTU 1000 mg daily On T4 0.3 mg daily (1 day off PTU) On 1-, 0.3 mg daily (4 day off PTU) 3 days after T 4 stopped

Thyroid Hormone Studies Before and A f t e r Administration of L-Thyroxine (T 4) and Propylthiouracil (PTU) Day

T4 /zg/dl

%FT4

AFT~ ng/dl

T3 ng/dl

rT3 ng/dl

3,3'-T 2 ng/dl

3',5'-T 2 ng/dl

0

8,4

0.053

4.4

185

34

5,3

3.0

7

13.0

0.063

8.2

180

32

7,5

5.25

17

14.0

0.139

19.5

200

38

4.5

3.75

18

11.5

0.055

6.3

175

31

3.8

2.25

21

11.0

0.049

5.4

130

35

<2.0

<1.5

24

10.5

0.045

4.7

140

35

<2.0

1.8

(4.8-13.0)

(0.023-0.048)

(50-210)

(27-62)

(<1.25-10.0)

(2.7-13.0)

Normal range

(percentile derivation from normal standard BMR values) 14 and the 24 hr 131I uptake was 18% (normal 8%-30%). Table 2 shows the results of the baseline measurements of T4, FT4 (percentage and absolute level), T3, rT3, 3,3'-T2 and 3',5'-T2. The only abnormality observed was a slight elevation of FT 4. The sequential administration of thyroxine (0.3 mg/daily) and propylthiouracil (1000 mg/daily) led to a marked increase in FT4 which was reversed by withdrawal of both drugs. On the seventh day of L-thyroxine therapy, the 24 hr 131Iuptake had decreased to 5%. Further thyroid studies were carried out in July 1978 after the administration of thyroxine 0.6 mg daily for 21 days. FTa increased from 4.5 to 7.9 ng/dl and the BMR rose from 15% to 41%.

Metabolic Studies After Plasma Exchange Between December 1979 and April 1980 this patient was submitted to six sessions of plasmapheresis, which were initially performed every 2 to 3 wk and Table3.

lzSI-Albumin Kinetics in Lipoatrophic Diabetes and in Diabetes i e l l i t u s LD

DM

(1.8-4,2)

later at monthly intervals. Administration of radioactive albumin demonstrated that the first session of plasmapheresis had removed 84% of the original plasma volume. Plasmapheresis did not change either serum albumin (4.2 g/l) or the degree of proteinuria (8 g/24 hr). Calculation of albumin kinetic parameters (Table 3) demonstrated an increased fractional catabolic rate with a decreased biological half-life of 125I albumin. The volume of distribution of albumin was markedly decreased, with an increase in the intravascular fraction; these abnormalities were not observed in an insulin dependent diabetic patient with the nephrotic syndrome (serum albumin 3.5 g/dl and urine protein 3 g/24 hr). Serum triglyceride and cholesterol levels were measured immediately before and following the sessions of plasmapheresis (Fig. 1). After plasmapheresis, the serum triglycerides fell sharply but within seven to ten days had returned to pre-plasmapheresis levels. As seen in Figs. 1 and 2, this pattern was not altered by administration of prednisone or by tightening control of the diabetes. Prednisone, 40 mg/day, IF////////////~

Normal

PLASMAPHERES]S

I////////////A

PREDNISONE]

3000"

Equilibrium with Extravascular Space (days) Biological Half-life (days) Fractional Catabolic Rate (% Intravenous pool/24 hr) Plasma Volume (ml) Predicted* Determined Volume of Distribution (ml) Intravascular Volume/Volume of Distribution (%) *From weight and height.

2500-

5.2 13.8

2.8 55.5

3.7 20

2000.

~

3.6

0.9

2.5

15001000"

yTOLTos

500'

2,108 2,755 4,803

1,865 2,436 8,155

57

30

"

~

'

~

CHOLESTEROL

I

15

35-47

I 30

I 45

I 6o

Fig. 1. Effect of plasmapheresis on serum lipid levels. Prednisone (40 mg/daily p.o.) was given after the fourth plasma exchange.

22

SOLER, WORTSMAN, AND CHOPRA

(

INSULIN INFUSION |

|PLASMAPHERESlSl 2500-

- 500

~

2000

400

"~

1500

300

IO00-

-

"-o

o

_~

200

E

500

-100

o i 2

i 4

i 6

i 8

1

I0

Days Fig. 2. Effect of optimal diabetic control upon serum triglyceride response to plasmapheresis. Insulin was administered by pump as a continuous subcutaneous infusion.

was started immediately after the fourth session of plasmapheresis and continued for a period of 2 wk; control of diabetes was improved by administering the usual daily insulin dose (320 units) as a continuous subcutaneous infusion into the abdominal wall using a Mill Hill infuser. The basal insulin infusion rate was 5.2 units/hr and insulin boluses were administered before breakfast (72 units), before lunch (60 units) and before the evening meal (60 units.) Infusion of insulin was started three days before plasmapheresis, maintained during the procedure and for ten days afterwards. The effect of hypertriglyceridemia on plasma aminoacids was also studied. Plasma aminoacid levels were followed after the sixth plasma exchange (Table 4). Branched chain aminoacid levels changed in paral-

lel with serum triglyceride levels. Alanine levels were high before and remained elevated after plasmapharesis. After the sixth plasma exchange the predicted initial decrease of serum triglycerides was again followed by a rebound within seven days. However, unexpectedly, serum triglycerides declined steadily therafter. Twelve months after the last plasma exchange serum triglycerides remain stable between 500 and 1000 mg/dl. Review of dietary records does not indicate any significant change in this patient's caloric intake (2000 calories daily including 240 g of carbohydrate, 100 g of protein and 71 g of fat) before, during or after plasmapheresis.

Endocrine Studies After Plasmapheresis Following the first two sessions of plasmapheresis the patient developed galact0rrhea and amenorrhea which persisted throughout and after completion of treatment. The endocrine tests were repeated two months after the last plasma exchange, at a time when the serum triglycerides were 700 mg/dl (Table 1). The results showed that both serum prolactin and plasma ACTH were still abnormal and had increased to higher levels. However, the TSH response to TRH (500 #g) had returned to normal and serum gonadotropin levels had remained within the normal range. DISCUSSION

The diagnosis of lipoatrophic diabetes was made on the basis of absence of subcutaneous fat, hypertrigly-

Table 4. Effect of the Sixth Plasmapheresis Upon Fasting Plasma Aminoacids

Days From Plasmapheresis

Glucose (mg/dl) Triglycerides (mg/dl) Amino Acids/~M Aspartic Acid Threonine Serine Glutamic Acid Proline Glycine Alanine Cystine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine

Normal~

--1

+1

+3

+7

+12

263 2,500

200 628

250 980

270 2,150

290 1,500

30.0 75.5 506.8 154.4 200.3 173.6 604.0 70.5 242.3 22.7 111.5 152.3 41.6 44.3 70.3 203.4 100.1

34.6 71.4 629.6 244.1 170.9 153.1 614.7 83,9 238,3 24,5 119,5 169.9 37.9 52.5 77.1 249.8 117.5

42.8 69.7 498.3 210.8 221.5 161.6 600.3 72.2 278.6 25.9 141.6 202,1 47,5 58,1 83.1 226.5 122.6

60.5 114.9 552.2 357.0 271.8 202.9 670.0 85.3 410.2 31.8 181.0 276.0 66.2 87.7 91.8 279.8 163.0

39.2 68.4 584.2 270.9 165.1 151.4 575.7 89.9 245.5 22,3 123.6 180.2 39.6 55.3 85.1 21 I. 1 120.7

(mean ~+SD)

(15.2 (155.5 ( 122.6 (67.6 (146.8 (209.2 (321.0 (74.1 (204.5 (28.1 (59.9 (125.3 (65.4 (57.5 (108.3 (202.0 (90.8

_+ 5.7) _+ 39.7) _+ 25.4) _+ 25.4) _+ 48.0) _+ 37.3) _+ 51.0) _+ 27.0) _+ 27.0) _+ 6.3) _+ 8.8) _+ 22.8) _+ 6.2) _+ 5.9) _+ 36.3) _+ 47.7) _+ 22.7)

LIPOATROPHIC DIABETES

ceridemia, and hyperglycemia with insulin resistance. The biochemical finding of severe hypertriglyceridemia was, in turn, expressed clinically by repeated episodes of abdominal pain and pancreatitis. The absence of subcutaneous fat, evident on physical examination and confirmed histologically, was also reflected in the kinetic study with radioalbumin, which demonstrated a very low volume of distribution for albumin. The model used for the analysis of albumin kinetics requires the existence of a stea~iy state. We believe that this condition was met in our patient based on the following: (1) the concentration of serum albumin was identical before and after plasmapheresis (2) the patient's plasma was replaced by plasma containing an isoconcentration of albumin and (3) the curve of decline in the specific activity of 125I albumin had a shape similar to that of the control subject. If our assumptions are valid, it can be suggested that in lipoatrophic diabetes serum albumin concentration was maintained by a combination of two mechanisms: decreased extravascular catabolism of serum albumin and, increased albumin synthesis by the liver. The normal BMR found in this patient is not in accordance with Lawrence's original description, l However, there are other cases of lipoatrophic diabetes with a normal BMR 7 and in several studies the BMR has not even been measured, z: In view of the incomplete information regarding thyroid function available in this condition and since the patient had a goiter, we evaluated in detail the response of the thyroid gland to both stimulation and suppression. The normal results observed in this patient, with the classical picture of lipoatrophic diabetes, suggest that thyroid dysfunction is no t a universal component of the disease. The reported success with the use of pimozide, 1° a central antidopaminergic agent, led us to investigate the interactions between the endocrine system and the metabolic abnormalities found in our patient with LD. We observed that baseline hypothalamo/pituitary function was abnormal and that the response to both suppressive and stimulating factors was also inappropriate. In contrast, assessment of peripheral endocrine function gave normal results and our dynamic studies demonstrated the expected changes in the levels of cortisol and of thyroid hormone and its major metabolites. The dissociation between the responses of cortisol and A C T H to dexamethasone suggests either heterogeneity in the circulating A C T H molecules or the presence of an interfering substance affecting the A C T H assay. In view of the repeated episodes of abdominal pain, we decided to attempt the rapid reduction of serum triglyceride levels by means of plasma exchange. The

23

effect of plasmapheresis upon hypertriglyceridemia was observed immediately. A sharp decline in serum triglycerides and disappearance of the episodes of abdominal pain followed the first plasma exchange. This improvement in serum lipid levels was short-lived (7-10 days) and neither prednisone nor tighter control of the diabetes delayed the return of the serum triglycerides to preplasmapheresis levels. The effect of prednisone was investigated because of previous work 16 reporting that cortisol has a lipemic clearing action in lipoatrophic diabetes. We do not have a ready explanation for the spontaneous stabilization of serum triglycerides at a lower level, around 700 mg/dl, that was observed after the sixth plasma exchange. This new steady state has already persisted for 12 too, since the last plasmapheresis. Plasmapheresis appeared to have a selective action on serum lipids since we could not detect any effects on carbohydrate metabolism. Insulin resistance, not due to circulating antibodies, persisted and insulin requirements remained unchanged. Renal function and severe proteinuria were also unaffected by either the procedure or the changes in serum triglyceride levels. In some clinical respects our patient showed a definite deterioration with plasmapheresis. Amenorrhea and galactorrhea developed immediately after the first sessions of plasmapheresis. Later the patient became hypertensive (B.P. 180/100 mmHg), necessitating antihypertensive therapy with methyldopa, and she also developed florid proliferative retinopathy, requiring treatment with photocoagulation. The decrease in triglyceride levels by plasmapheresis was also associated with worsening of endocrine function. At a clinical level this deterioration included the development of galactorrhea with hyperprolactinemia and amenorrhea with normal serum gonadotropins (galactorrhea-amenorrhea syndrome). Specific testing also revealed that the preexisting abnormality in A C T H secretion became more exaggerated. The only exception to the overall deterioration in endocrine function was the return to normal of the T S H response to stimulation with TRH. In conclusion, our studies have confirmed the systemic nature of lipoatrophic diabetes which is characterized by derangements of both lipid and carbohydrate metabolism and by abnormalities of hypothalamic function. Plasmapheresis definitely has a place in the management of this condition, for it effectively reduces serum triglyceride levels and controls the episodes of abdominal pain. However, cardiovascular abnormalities and worsening of endocrine function may develop during the course of this therapy.

24

SOLER, WORTSMAN, AND CHOPRA

REFERENCES

1. Lawrence RD: Lipodystrophy and hepatomegaly, with diabetes, lipaemia and other metabolic disturbances. Lancet 1:724, 1946 2. Wachslicht-Rodbard H, Muggeo M, Kahn CR, et al: Heterogeneity of the Insulin-Receptor Interaction in Lipoatrophic Diabetes. J Clin Endocrinol Metab 52:416, 1981 3. Seip M: Lipodystrophy and gigantism with associated endocrine manifestations: A new diencephalic syndrome? Acta Paediat 48:555, 1959 4. Podolsky S: Lipoatrophic diabetes and miscellaneous conditions related to diabetes mellitus, in Marble A (ed): Joslin's Diabetes Melliuts, (ed 11). Philadelphia, Lea & Febiger, 1971, pp 722 5. Rossini AA, Aoki TT, Goldman RF, et al: Metabolic and endocrine studies in a case of lipoatrophic diabetes. Metabolism 26:637, 1977 6. Mabry CC, Hollingsworth DR, Upton GV, et al: Pituitaryhypothalamic dysfunction in generalized lipodystrophy. J Pediat 82:625, 1973 7. Samaan NA, Craig JW: Serum insulin and growth hormone in lipoatrophic diabetes. Metabolism 18:460, 1969 8. Louis CH, Conn JW, Minick MC: Lipoatrophic diabetes; isolation and characterization of an insulin antagonist from urine. Metabolism 12:867,1963 9. Hamwi G J, Kruger FA, Eymontt M J, et al: Lipoatrophic diabetes. Diabetes 15:262, 1966

10. Corbin A, Upton GV, Mabry CC, et al: Diencephalic involvement in generalized lipodystrophy: Rationale and treatment with the neuroleptic agent, Pimozide. Acta Endocrinologica 77:209, 1974. 11. Chopra I J, Chopra U, Smith SR, et ali Reciprocal changes in serum concentration of 3,3',5'-triiodothyronine (reverse T3) and 3,3',5-triiodothyronine (T3) in systemic illness. J Clin Endocrinol Metab 41:1043, 1975 12. Wu SY, Chopra IH, Nakamura Y, et al: A radioimmunoassay for measurement of 3,3'-L-diiodothyronine (T2). J Clin Endocrinol Metab 43:682, 1976 13. Chopra IH, Geola F, Solomon DH, et al:3'5'-Diiodothyronine in health and disease. Studies by a radioimmunoassay. J Clin Endocrinol Metab 47:1198, 1978 14. Weinstock JV, Kawaniski H, Sisson J: Morphologic, biochemical and physiologic alterations in a case of idiopathic hypoalbuminaemia (analbuminemia). Am J Med 67:132, 1979 15. Milsom JP, Morgan MY, Sherlock S: Factors affecting plasma aminoacid concentrations in control subjects. Metabolism 28:313, 1979 16. Ruvalcaba RH, Kelley VC: Lipoatrophic Diabetes II. Metabolic Studies Concerning the Mechanism of Lipemia. Amer J Dis Child 109:287, 1965