Machine Learning to Allocate Palliative Care Consultations During Cancer Treatment

PURPOSE

For patients with advanced cancer, early consultations with palliative care (PC) specialists reduce costs, improve quality of life, and prolong survival. However, capacity limitations prevent all patients from receiving PC shortly after diagnosis. We evaluated whether a prognostic machine learning system could promote early PC, given existing capacity.

METHODS

Using population-level administrative data in Ontario, Canada, we assembled a cohort of patients with incurable cancer who received palliative-intent systemic therapy between July 1, 2014, and December 30, 2019. We developed a machine learning system that predicted death within 1 year of each treatment using demographics, cancer characteristics, treatments, symptoms, laboratory values, and history of acute care admissions. We trained the system in patients who started treatment before July 1, 2017, and evaluated the potential impact of the system on PC in subsequent patients.

RESULTS

Among 560,210 treatments received by 54,628 patients, death occurred within 1 year of 45.2% of treatments. The machine learning system recommended the same number of PC consultations observed with usual care at the 60.0% 1-year risk of death, with a first-alarm positive predictive value of 69.7% and an outcome-level sensitivity of 74.9%. Compared with usual care, system-guided care could increase early PC by 8.5% overall (95% CI, 7.5 to 9.5; P < .001) and by 15.3% (95% CI, 13.9 to 16.6; P < .001) among patients who live 6 months beyond their first treatment, without requiring more PC consultations in total or substantially increasing PC among patients with a prognosis exceeding 2 years.

CONCLUSION

Prognostic machine learning systems could increase early PC despite existing resource constraints. These results demonstrate an urgent need to deploy and evaluate prognostic systems in real-time clinical practice to increase access to early PC.

Postpartum Emergency Department Use Following Midwifery-Model vs Obstetrics-Model Care

Importance

Emergency department (ED) use postpartum is a common and often-preventable event. Unlike traditional obstetrics models, the Ontario midwifery model offers early care postpartum.

Objective

To assess whether postpartum ED use differs between women who received perinatal care in midwifery-model care vs in traditional obstetrics-model care.

Design, Setting, and Participants

This retrospective population-based cohort study took place in Ontario, Canada, where public health care is universally funded. Participants included women who were low risk and primiparous and gave birth to a live baby in an Ontario hospital between 2012 and 2018. Data were collected from April 2012 to March 2018 and analyzed from June 2022 to April 2023.

Exposures

Perinatal care clinician, namely, a midwife or obstetrician.

Main Outcome and Measures

: Any unscheduled ED visit 42 days postpartum or less. Poisson regression models compared ED use between women with midwifery-model care vs obstetrics-model care, weighting by propensity score-based overlap weights.

Results

Among 104 995 primiparous women aged 11 to 50 years, those in midwifery-model care received a median (IQR) of 7 (6-8) postpartum visits, compared with 0 (0-1) visits among those receiving obstetrics-model care. Unscheduled ED visits 42 days or less postpartum occurred for 1549 of 23 124 women (6.7%) with midwifery-model care compared with 6902 of 81 871 women (8.4%) with traditional obstetrics-model care (adjusted relative risks [aRR], 0.78; 95% CI, 0.73-0.83). Similar aRRs were seen in women with a spontaneous vaginal birth (aRR, 0.71; 95% CI, 0.65-0.78) or assisted vaginal birth (aRR, 0.70; 95% CI, 0.59-0.82) but not those with a cesarean birth (aRR, 0.94; 95% CI, 0.86-1.03) or those with intrapartum transfer of care between a midwife and obstetrician (aRR, 0.94; 95% CI, 0.87-1.04). ED use 7 days or less postpartum was also lower among women receiving midwifery model care (aRR, 0.70; 95% CI, 0.65-0.77).

Conclusions and Relevance

In this cohort study, midwifery-model care was associated with less postpartum ED use than traditional obstetrics-model care among women who had low risk and were primiparous, which may be due to early access to postpartum care provided by Ontario midwives.

The role of surgeon specialty in management and survival of malignant ovarian germ cell tumors: A population-based study

OBJECTIVES

The aim of this study is to describe management and survival in adult patients with malignant ovarian germ cell tumors (MOGCT) undergoing surgery by general gynecologists (GG) versus gynecologic oncologists (GO).

METHODS

This is a population-based retrospective cohort study, including patients (age ≥ 18 years old) with MOGCT identified in the provincial cancer registry of Ontario, (1996-2020). Baseline characteristics, surgical and chemotherapy treatment were compared between those with surgery by GG or GO. Cox proportional hazards (CPH) model was used to determine if surgeon specialty was associated with overall survival (OS).

RESULTS

Overall, 363 patients were included. One-hundred and sixty (44%) underwent surgery by GO and 203 (56%) by GG. There were higher rates of stage II-IV in the GO group (27.5% vs 3.9%, p < 0.001, and higher proportion of chemotherapy (64.4% vs 37.4%, p < 0.0001). Five-year OS was 90% and 93% in the GO vs GG groups, respectively (p = 0.39). CPH model showed factors associated with increased risk of death were older age at diagnosis (HR 1.09, 95% CI 1.07-1.12) and chemotherapy (HR 3.12, 95% CI 1.44-6.75). Surgeon specialty was not independently associated with all-cause death (HR 1.04, 95% 0.51-2.15, p = 0.91).

CONCLUSIONS

In this group of MOGCT, 5-year OS was not significantly different between patients having surgery by GO compared to GG. Nevertheless, survival rates were lower than expected in the GG group despite their low-risk features. Further exploration is warranted regarding the reasons for this and whether patients with suspected MOGCT may benefit from early assessment by GO for optimal management.

Machine Learning-Based Early Warning Systems for Acute Care Utilization During Systemic Therapy for Cancer

BACKGROUND

Emergency department visits and hospitalizations frequently occur during systemic therapy for cancer. We developed and evaluated a longitudinal warning system for acute care use.

METHODS

Using a retrospective population-based cohort of patients who started intravenous systemic therapy for nonhematologic cancers between July 1, 2014, and June 30, 2020, we randomly separated patients into cohorts for model training, hyperparameter tuning and model selection, and system testing. Predictive features included static features, such as demographics, cancer type, and treatment regimens, and dynamic features, such as patient-reported symptoms and laboratory values. The longitudinal warning system predicted the probability of acute care utilization within 30 days after each treatment session. Machine learning systems were developed in the training and tuning cohorts and evaluated in the testing cohort. Sensitivity analyses considered feature importance, other acute care endpoints, and performance within subgroups.

RESULTS

The cohort included 105,129 patients who received 1,216,385 treatment sessions. Acute care followed 182,444 (15.0%) treatments within 30 days. The ensemble model achieved an area under the receiver operating characteristic curve of 0.742 (95% CI, 0.739-0.745) and was well calibrated in the test cohort. Important predictive features included prior acute care use, treatment regimen, and laboratory tests. If the system was set to alarm approximately once every 15 treatments, 25.5% of acute care events would be preceded by an alarm, and 47.4% of patients would experience acute care after an alarm. The system underestimated risk for some treatment regimens and potentially underserved populations such as females and non-English speakers.

CONCLUSIONS

Machine learning warning systems can detect patients at risk for acute care utilization, which can aid in preventive intervention and facilitate tailored treatment. Future research should address potential biases and prospectively evaluate impact after system deployment.

Search for a light pseudoscalar particle in the decay K_{L};{0}-->pi;{0}pi;{0}X

We performed a search for a light pseudoscalar particle X in the decay K_{L};{0}-->pi;{0}pi;{0}X, X-->gammagamma with the E391a detector at KEK. Such a particle with a mass of 214.3 MeV/c;{2} was suggested by the HyperCP experiment. We found no evidence for X and set an upper limit on the product branching ratio for K_{L};{0}-->pi;{0}pi;{0}X, X-->gammagamma of 2.4x10;{-7} at the 90% confidence level. Upper limits on the branching ratios in the mass region of X from 194.3 to 219.3 MeV/c;{2} are also presented.

Search for the Decay K L0-->pi0nu nu[over]

We performed a search for the K L0-->pi0nu nu[over] decay at the KEK 12-GeV proton synchrotron. No candidate events were observed. An upper limit on the branching ratio for the decay was set to be 6.7 x 10(-8) at the 90% confidence level.

The relationship of conjunctival and arterial blood gas oxygen measurements

Conjunctival oxymetry (CjO2) measures peripheral tissue oxygen at the conjunctival level. CjO2 changes can indicate pulmonary or circulatory conditions leading to shock. Literature review does not define 'normal' CjO2/ABG PaO2 ratios. We designed a study to measure these ratios. Twenty-two healthy patients undergoing cardiac catheterization had simultaneous PcjO2 and PaO2 measurements completed. The range of conjunctival oxygen measurements was from 34 to 68 mmHg with a mean of 50.5 mmHg. The PaO2 readings ranged from 65 to 93 mmHg with a mean of 77.1 mmHg. The average PcjO2/PaO2 ratio was 0.656 with a range of 0.47-0.93. Thus the PcjO2 is on average 66% of the arterial blood gas PaO2. This ratio of 0.66 can serve as a base for further clinical studies in which PcjO2 is looked at in patients with pulmonary or circulatory illnesses or injuries.

Renin suppression by saline is blunted in nonmodulating essential hypertension

We have reported that 50% of subjects with normal renin essential hypertension have both delayed suppression of the renin-angiotensin-aldosterone axis following sodium infusion and a delayed rate of excretion of an acute salt load. In another study we have also described a subset of patients with essential hypertension (called nonmodulators) who have several abnormalities, including a pressor response to salt loading. To evaluate whether the abnormalities described in these different groups of patients actually occur in the same patient, we assessed the renin-angiotensin-aldosterone axis response to short-term saline loading in 38 hypertensive patients. Their ability to modulate was determined by their renal vascular response to infused angiotensin II on a high salt diet (200 mEq Na). In response to a 3-hour infusion of saline, 75 mEq/hr, the reduction in plasma renin activity at both 60 and 120 minutes was significantly greater (p less than 0.008) in patients with normal modulation than in the nonmodulators. Plasma aldosterone levels were also significantly lower (p less than 0.001) in those with intact modulation. Thus, nonmodulating essential hypertensive patients have abnormalities in several systems that influence sodium homeostasis, including altered adrenal and renal vascular response to angiotensin II, altered renal blood flow response to salt loading, and a delayed suppression of the renin-angiotensin-aldosterone system with short-term saline infusion.

A review of research methodology in an emergency medicine journal

All original articles published in JACEP and Annals of Emergency Medicine in 1972, 1975, 1978, and 1981 were reviewed and classified according to type of article, research design, specific methods utilized, and statistical analyses employed. The number of articles published annually increased during the 10-year period, from 34 in 1972 to 131 in 1981. The percentage of articles devoted to clinical research, basic science, animal studies, and health services research also increased: 1972, 8.8%; 1975, 39.6%; 1978, 41.2%; and 1981, 51.1%. In 1972, only two studies had defined research designs. In 1981, 47 studies had the following research designs: case control, 9; cohort, 9; cross-sectional, 13; uncontrolled clinical trial, 3; non-random clinical trial, 7; and randomized clinical trial, 6. Research methodologies were defined in only two studies published in 1972, but increased in each of the years reviewed to 41 in 1978. The most common methodology was the use of diagnostic categories. In 1972, statistical analysis was limited to descriptive enumeration and was included in only three articles. In 1981, 62 articles contained statistical analyses, including chi square in 11; t test in 6; Fischer's exact test in 2; ANOVA in 4; and P value (method not stated) in 10. This study reveals a major increase in both the quality and quantity of emergency medicine research published during the 10-year period reviewed.

Defect in the sodium-modulated tissue responsiveness to angiotensin II in essential hypertension

In normal subjects, dietary sodium intake modulates renovascular, adrenal, and pressor responses to infused angiotensin II (AII). To examine the hypothesis that this modulation is abnormal in some patients with essential hypertension, we studied 18 hypertensives and 9 normal subjects twice--during dietary sodium restriction and during loading. Paraaminohippurate (PAH) clearance was used to assess renal plasma flow. AII was infused in graded doses (0.3-3.0 ng/kg per min). Plasma aldosterone, cortisol, renin activity, AII, sodium, potassium, and PAH clearance were measured at the onset and end of each AII dose. During dietary sodium repletion, eight of the subjects with essential hypertension showed a normal renovascular response (greater than 125 ml/min per 1.73 m2) to AII infusion (3 ng/kg per min). The decrement in renal blood flow in these normal responders (NR) was 168 +/- 10, which was comparable to the range in normotensive subjects (206 +/- 25 ml/min per 1.73 m2). All of the remaining hypertensive patients, designated abnormal responders (AbR), had lower (less than 125) renal blood flow responses to the same dose of infused AII (mean decrement: 84 +/- 11 ml/min per 1.73 m2) compared with the NR and normotensive subjects. Renal blood flow responses to all AII doses were statistically greater on a high-vs.-low salt diet in the NR (P less than 0.001, chi-square) and normotensives (P = 0.004, chi-square) but sodium intake had no effect on this response in the AbR. Basal renal blood flow in NR increased significantly (P less than 0.001, paired t test) with dietary sodium repletion, from 491 +/- 36 (low salt) to 602 +/- 40 ml/min per 1.73 m2 (high salt), but was almost identical in the AbR on differing dietary sodium intakes (429 +/- 24 vs. 425 +/- 26 ml/min per 1.73 m2). The adrenal responses to sodium intake and infused AII also differed in the two subgroups. In the NR, the adrenal response to AII was significantly greater (P = 0.011, Wilcoxon signed rank test) after sodium restriction. In contrast, there was no significant difference in the aldosterone response to AII infusion between the low and high sodium diets in the AbR. Thus, a substantial subgroup of essential hypertensives has an abnormality in responsiveness to AII in two systems central to volume homeostasis: the kidney and adrenal. They fail to modulate their renal blood flow and aldosterone responses to AII with changes in dietary sodium intake. Moreover, basal renal blood flow does not increase appropriately with increased sodium intake. These abnormalities, which may be due to an increased local production of AII or a defect in the AII receptors in these three target tissues, could contribute to the elevated blood pressure.

Central venous catheterization in the emergency setting

To determine the appropriateness and rate of complications from central venous catheterization (CVC) in the emergency department of a university teaching hospital, criteria for justifiability were established and all such catheterizations during a 1-yr period were reviewed. Of 61 catheterizations in 57 patients, 41% could not be justified according to the criteria established. Eight (14%) patients had serious complications from catheter placement. This complication rate is higher than that in any previously published study, but no other study has examined only emergent catheter placement. These findings underline the importance of adhering to a limited set of indications for CVC in emergency departments.

Efficacy of cervical spine immobilization methods

Cervical spine immobilization devices are widely used to stabilize the cervical spine and prevent neurologic deficits associated with unstable fractures. In order to quantitate their efficacy we measured controlled cervical spine motion in three axes, using six different immobilization methods in 25 volunteers instructed to actively move their necks as much as possible in the directions of flexion, extension, rotation, and lateral bending while lying supine. Control measurements were made with no device and measurements were repeated following immobilization with: soft collar (SC), hard collar (HC), extrication collar (EC), Philadelphia collar (PC), bilateral sandbags joined with 3-inch cloth tape across the forehead (ST), and the combination of sandbags, tape, and the Philadelphia collar (ST/PC). Neck movements were reported in degrees recorded on a hand-held goniometer. There were no significant differences between control and SC measurements except in rotary movement. PC was not significantly better than the other two types of hard collars, except in limiting extension. ST immobilization was significantly better than any of the other four methods used alone, for all four movements. The addition of PC to ST was significantly more effective in reducing extension only.

Transcutaneous oxymetry measurements during acute blood loss

Transcutaneous PO2 (TcPO2) was measured in healthy adult blood donors to test the sensitivity of this method as a noninvasive means of diagnosing occult blood loss. TcPO2, the energy required to heat the electrode (MW), and postural changes in blood pressure and pulse, were measured before and after a 450-ml blood donation. There was a significant increase (P less than .005) in postural pulse of 4.8/min, but no significant change (P greater than .05) in postural blood pressure. There was no significant change (P greater than .05) in TcPO2, but there was a significant decrease (P less than .001) in MW. These MW changes probably reflect physiologic changes associated with blood loss.

Alterations in aldosterone biosynthesis in essential hypertensives

We studied hypertensives with decreased adrenal responsiveness to infused angiotensin II (AII) to assess their responsiveness to other aldosterone secretagogues, ACTH and potassium, which are thought to stimulate aldosterone synthesis in sites different from one another and from AII. All subjects, following sodium restriction, received an infusion of AII in increasing doses (0.1-3 ng/kg per min). The increment in aldosterone between control and the highest infusion dose divided by the increment in plasma AII was used as the index of adrenal responsiveness. All normotensive controls (NC) had a ratio greater than 0.5. Hypertensives with a normal ratio were designated normal responders (NR) and those with a lower ratio were abnormal responders (AbR). The slope of the regression line between aldosterone and AII was significantly less for the AbR (0.02 +/- 0.04) than for the NR (1.20 +/- 0.02, P less than 0.001) and the NC (1.00 +/- 0.03, P less than 0.001) groups. During infusion of cosyntropin in increasing doses (0.05-1.5 mIU/kg per 30 min), the aldosterone response of the AbR was significantly less than that of the NR (P less than 0.016) or the NC (P less than 0.05) groups. Similarly, after infusion of potassium (0.33 mEq/min), the increment in aldosterone in the AbR group (7.6 +/- 2.2 ng/dl) was significantly less than that in the NR (14.2 +/- 2.5 ng/dl, P less than 0.05) and the NC (18 +/- 5 ng/dl, P less than 0.05) groups. Thus hypertensives with decreased aldosterone responsiveness to infused AII also had decreased responsiveness to infused ACTH and potassium, suggesting that their defect lies in the intracellular aldosterone biosynthetic pathway.

Contribution of prostaglandins to the antihypertensive action of captopril in essential hypertension

To determine whether prostaglandins contribute to the depressor response to the converting enzyme inhibitor, captopril, we measured the plasma prostaglandin levels by radioimmunoassy before and after captopril administration, and then examined the effect of prostaglandin synthetase inhibition on captopril's antihypertensive effect. When a single oral captopril dose (25-100 mg) was given to 31 sodium-restricted patients with essential hypertension, the levels of the stable transformation product of prostacyclin remained unmeasurable and that of thromboxane A2 did not change, while the metabolite of PGE2 (PGE-M) increased by 53% (34 +/- 4pg/ml pre-captopril, 52 +/- 5 pg/ml after; p less than 0.001). As expected, blood pressure (BP) and angiotension II (AII levels fell, and kinin levels rose (all changes p less than 0.001). We then blocked prostaglandin synthesis in 18 of these subjects for 24 hours with either indomethacin (n = 10) or aspirin (n = 8) before repeating the captopril dose, to assess the importance of these PGE-M increments. The PGE-M responses to captopril were effectively blocked in nine of 10 subjects receiving indomethacin and four of eight receiving aspirin. In these 13 patients, the depressor response to captopril was significantly blunted (-20 +/- 3mm Hg pre-synthetase inhibition vs - 13 +/- 2 mm Hg post; p less than 0.05). When these agents did not block the PGE-M response to captopril, the BP response was also unchanged (-15 +/- 4mm Hg pre, -18 +/- 5mm Hg post). Neither indomethacin nor aspirin changed the AII or kinin responses to captopril. We conclude that the prostaglandins may be important mediators of captopril's antihypertensive effect in the sodium-restricted state.

Diabetic retinopathy: update on therapeutic advances

It is clear that real advances have occurred in the therapy of diabetic retinopathy. However, this complication of diabetes can and does lead to blindness in the older patient. The present treatment modalities offer a means of maintaining or, in some patients, improving the visual performance of the patient. Laser photocoagulation and vitrectomy are the major new therapies. The future will hopefully give us a better understanding of the cause of diabetic retinopathy so that it may be prevented and cured. However, in the interim, all patients with diabetes mellitus should be urged to see on ophthalmologist yearly in order to obtain the best available therapy.

Effects of long term sulfonylurea therapy on plasma insulin and fasting lipid levels

Insulin secretion was studied before and after the control of hyperglycemia in fourteen maturity onset male non-obese diabetics. Optimum control of hyperglycemia was achieved by the addition of the sulfonylurea chlorpropamide to dietary treatment. One patient was a primary treatment failure, but nine out of thirteen had excellent control of hyperglycemia. A standardized oral glucose tolerance test (GTT) was performed before and after eight months of individualized therapy with the sulfonylurea. The GTT was repeated with each patient taking his usual dose of chlorpropamide 90 min prior to the administrationo f the glucose load. In the baseline test glucose levels rose from 135.6 +/- 9.9 mg/dl to a peak level of 268.8 +/- 17.7 mg/dl at 120 min. After control of hyperglycemia glucose levels were significantly lower at 0, 30 and 60 min, and rose from 106.8 +/- 8.5 mg/dl to a maximum of 224.5 +/- 17.3 mg/dl at 120 min. Plasma insulin response was unchanged. Fasting serum cholesterol, triglyceride and total lipid levels changed only minimally during therapy. It is concluded that lowered serum glucose levels after long term treatment with chlorpropamide occured while plasma insulin response to glucose was no greater than before treatment. These findings may be explained by an extrapancreatic effect of the drug or by an indirect result of chlorpropamide induced insulin release which occured earlier in the course of therapy.

Hyperosmolar nonketotic coma in the elderly diabetic

The gravity of this syndrome of severe diabetic stupor without ketosis may not be recognized because patients are usually middle-aged or elderly with mild diabetes. A lack of urgency in treating these patients is probably the cause of the widely reported mortality of 40 to 70 per cent.

Absence of sleep-related growth hormone elevations in myotonic dystrophy

There is evidence that in myotonic dystrophy, the endocrine and central nervous systems are affected. To study a possible relationship between both defects, we investigated nocturnal sleep patterns and associated growth hormone secretion in two men and three women with myotonic dystrophy. In three patients who were clinically the most severely affected by myotonic dystrophy, plasma growth hormone elevations related to the slow-wave phase of sleep were absent. The two least severely affected patients had plasma growth hormone increases of low magnitude and brief duration (from 0.4 ng per milliliter to 13.0 ng per milliliter). These data suggest a failure of integration at a subcortical level of the slow-wave phase of sleep with the hypothalamic-pituitary mechanisms of growth hormone secretion. Thalamic neuronal lesions occurring in myotonic dystrophy could be responsible for such failure.

Abnormal glucose tolerance and arginine tolerance tests in Huntington's disease

Neuropathological studies of Huntington's disease reveal neuronal atrophy, lipofuscin accumulation and other findings characteristic of the aged brain, although the onset of disease is only the fourth decade. The pathology is limited to specific areas such as the caudate nucleus, cerebral cortex and hypothalamus. 14 patients with documented Huntington's disease (mean age of 44.4 years with a range of 27-79 years) were studied by oral glucose tolerance tests (GTT) and intravenous arginine tolerance tests performed under standardized metabolic conditions. Seven of the 14 patients had impaired carbohydrate tolerance. Mean plasma glucose level at 2h was 90.4+/-6.2 mg/100 ml in the patients with a normal GTT and 148.1+/-8.9 mg/100 ml in the patients with a diabetic type GTT. Mean peak insulin level in the nondiabetic group occurred at 1/2 h and was 60.2+/-10.1 muU/ml, but in the diabetic group the peak insulin level occurred at 2h and was 155.9+/-33.8 mgU/ml. There was failure of suppression of growth hormone during the GTT, with a rise to abnormally high levels at 5h (18.6+/-5.6 ng/ml). Arginine infusion resulted in normal glucose and insulin rise in the nondiabetic patients with Huntington's disease. However, arginine infusion provoked an elevated insulin response in those with a diabetic GTT, and an exaggerated growth hormone response in the majority of the patients. It is uncertain whether these observations are related to abnormal cerebral aging per se, direct hypothalamic neuronal degeneration, or perhaps a relative imbalance of intracerebral neurotransmitters including dopamine.

Improvement of growth hormone response to stimulation in primary aldosteronism with correction of potassium deficiency

Potassium depletion frequently occurs in primary aldosteronism and has been implicated as the cause of the impaired carbohydrate tolerance frequently associated with this syndrome. Glucose, insulin, and growth hormone regulation were studied in a 42-yr-old, male patient with an aldosterone-secreting adenoma when the patient was potassium-depleted and again after potassium repletion. Potassium repletion was documented by serial body potassium measurements, with an increase in body potassium from 2400 mEq to 2850 mEq after 400 mg spironolactone and 80 mEq supplemental potassium chloride were administered daily for 7 days. Potassium repletion resulted in improvement of the patient's glucose tolerance test, with a decrease in the peak glucose level from 184 mg/100ml to 130 mg/100ml and an increase in the peak insulin level from 46 muU/ml to 85 muU/ml. Intravenous administration of arginine resulted in a subnormal insulin response of 28 muU/ml in the base-line test and an increase to 59 muU/ml after potassium stores were repleted. Growth hormone response to arginine infusion was also initially minimal at 12.5 ng/ml, increasing markedly to 26 ng/ml after potassium replenishment. Insulin-induced hypoglycemia resulted in a depressed growth hormone response of 8 ng/ml when the patient was potassium-deficient, but a normal response of 30 ng/ml after potassium repletion. These observations demonstrate that impairment of both insulin and growth hormone responses to stimulation occur in primary aldosteronism with potassium depletion. These abnormalities may be reversed by potassium repletion.

Age, smoking inhalation, and pulmonary function

This study examined the relative effects of age and smoking on pulmonary function. Smoking was measured by six smoking variables, taken singly and as a composite. Subjects were 1,516 male participants in the Normative Aging Study. A stepwise multiple regression with vital capacity (VC) and forced expiratory volume at one second (FEV1.0) as the criteria accounted for 24.4% and 28.3% of the variance, respectively. Two-way analyses of variance showed that the age decline in pulmonary function was substantially greater for high inhalers than it was for low inhalers or nonsmokers. Age and the inhalation index were also noticeably and independently related to a decline in pulmonary function.

Exaggerated growth hormone response to arginine infusion in Huntington's disease

Growth hormone regulation was studied in 10 patients with Huntington's disease after intravenous administration of arginine. In 20 control subjects arginine infusion resulted in a rise of plasma growth hormone levels from a mean baseline value of 3.2+/-0.6 ng/ml to a peak level of 17.6+/-2.7 ng/ml at 60 min. Growth hormone rise in the majority of patients with Huntington's disease was clearly intact and significantly greater than normal in magnitude, increasing from the baseline level of 2.6+/-0.5 ng/ml to a peak level of 28.3+/-3.7 ng/ml at 60 min (P = less than 0.05). Carbohydrate tolerance of these patients was previously examined, and 4 with normal glucose tolerance and normal insulin responses to arginine infusion had growth hormone levels significantly higher than controls at 30 min. Six patients with impaired carbohydrate tolerance and exaggerated insulin responses to arginine had significantly higher growth hormone responses at 30 min and also at 60 min. Neuronal degeneration of several hypothalamic nuclei has been reported in Huntington's disease. The observations that growth hormone responds in an exaggerated fashion to stimulation by arginine infusion or falling glucose levels as previously described may be explained by intrahypothalamic dysfunction such as impairment of somatostatin secretion.