Summary Objective Hyponatraemia is common in community‐acquired pneumonia (CAP) and is associated with increased mortality. The mechanism of hyponatraemia in CAP is not completely understood and treatment is therefore ill‐defined. We aimed to define the causation of hyponatraemia in CAP. Design Prospective, single‐centre, observational study of all patients with CAP and hyponatraemia (≤ 130 mmol/L) during a 9‐month period. Patients The prevalence of each subtype of hyponatraemia, and the associated mortality, was determined in 143 admissions with CAP (Study 1). A sub‐cohort of patients with SIAD (n = 10) was prospectively followed, to document the natural history of SIAD associated with CAP (Study 2). Measurements In Study 2, blood and urine were collected on day 1, 3, 5 and 7 following admission for measurement of plasma vasopressin, sodium, osmolality and urine osmolality. Results In study 1, 143/1723(8.3%) of CAP patients had hyponatraemia (≤130 mmol/L). About 66 had SIAD (46%), 60(42%) had hypovolaemic hyponatraemia (HON), 13(9%) had hypervolaemic hyponatraemia (HEN) and 4(3%) patients had hyponatraemia due to glucocorticoid hormone deficiency. Mortality was higher in the HEN than in the HON, SIAD or normonatraemic groups (P < 0.01). In Study 2, plasma sodium concentration normalized in 8/10 (80%) by day 7. Two patients with persistent hyponatraemia were discovered to have underlying bronchiectasis. Conclusions Hyponatraemia in CAP is most commonly secondary to SIAD or hypovolaemia. HEN is less common, but has worse prognosis. Prospective observation demonstrates that in SIAD, plasma AVP and sodium concentrations normalize with antimicrobials; failure of reversal of suggests underlying lung disease, such as bronchiectasis.

SummaryContext and ObjectiveNonfunctioning pituitary adenomas (NFPAs) are the most common subtype of pituitary tumour. Hypopituitarism is observed in NFPAs due to tumour‐ or treatment‐related factors and may increase mortality risk. Here, we analysed the associations of hypopituitarism, hormone replacement and mortality in a large NFPA cohort derived from two large European centres.Design, Setting and ParticipantsCase note review of all patients treated for NFPA in University Hospitals Birmingham and Beaumont Hospital Dublin between 1999 and 2014 was performed.Main Outcome MeasuresClinical presentation, treatment strategies, pituitary function and vitality status were recorded in each patient. A multivariate Cox regression model was used to examine the association between hypopituitarism, hormone replacement and premature mortality.ResultsA total of 519 patients were included in the analysis. Median duration of follow‐up was 7·0 years (0·5–43). A total of 81 deaths were recorded (15·6%). On multivariate analysis, adrenocorticotropic hormone (ACTH) and gonadotropin (Gn) deficiencies were associated with an increased relative risk of death (OR 2·26, 95% CI 1·15–4·47, P = 0·01 and OR 2·56, 95% CI 1·10–5·96, P = 0·01, respectively). Increased hydrocortisone (HC) (P‐trend = 0·02) and lower levothyroxine (LT4) doses (P‐trend = 0·03) were associated with increased risk of death. Mortality increased with the degree of pituitary failure observed (P‐trend = 0·04).ConclusionACTH and gonadotropin‐deficient patients have higher mortality rates compared to those with intact hormonal axes. Excessive HC and suboptimal LT4 replacement may also increase risk of death. Complex associations between hormone deficiency and replacement underpin the increased mortality risk in NFPA patients.

Aims To determine the fetal and maternal outcomes in pregnant women with Glucokinase‐Maturity onset diabetes of the young (GCK‐MODY). Methods We studied the obstetric and perinatal outcomes in 99 pregnancies of 34 women with GCK‐MODY. The mutation status of the offspring was known in 29 and presumed in 33. Clinical outcomes were determined and compared between affected (n = 39) and unaffected (n = 23) offspring. Results 59% of pregnancies were treated with diet alone and 41% received insulin. Birthweight, percentage of large for gestational age (LGA) and caesarean section (CS) in GCK‐unaffected offspring was significantly higher than in GCK‐affected offspring (4.0 ± 0.7 vs. 3.4 ± 0.4 kg, p = 0.001), 15 (65%) vs. 5(13%) (p = 0.00006) and 17 (74%) vs. 11 (28%) (p = 0.001), respectively. We observed an earlier gestational age at delivery on insulin in unaffected offspring (38.3 ± 1.0 vs. 39.5 ± 1.5 weeks, p = 0.03) with no significant change in LGA (9 (82%) vs. 6 (50%); p = 0.12), and a higher rate of CS (8 [73%] vs. 3 [11%]; p < 0.001), and no change in small for gestational age (0 [0%] vs. 4 [14%]; p = 0.30) in affected offspring. Conclusion Insulin therapy in unaffected offspring did not reduce LGA and was associated with earlier gestational age at delivery. Insulin treatment in GCK‐affected offspring was associated with an increased incidence of CS, but did not adversely affect fetal outcome. Fetal genotype determines birthweight rather than treatment. Pre‐pregnancy diagnosis of GCK‐MODY, use of continuous glucose monitoring and non‐invasive fetal genotyping may enable further investigation of targeted therapy in this condition.

Summary Thyroid dysfunction is among the most common immune-related adverse reactions associated with immune checkpoint inhibitors. It most commonly manifests as painless thyroiditis followed by permanent hypothyroidism. This usually causes mild toxicity that does not interfere with oncological treatment. In rare instances, however, a life-threatening form of decompensated hypothyroidism called myxoedema coma may develop. We present a case of myxoedema coma in a woman in her sixties who was treated with a combination of CTLA-4 and PD-1 immune checkpoint inhibitors; for stage four malignant melanoma. She became hypothyroid and required thyroxine replacement after an episode of painless thyroiditis. Six months after the initial diagnosis of malignant melanoma, she presented to the emergency department with abdominal pain, profuse diarrhoea, lethargy and confusion. She was drowsy, hypotensive with a BP of 60/40 mmHg, hyponatraemic and hypoglycaemic. Thyroid function tests (TFTs) indicated profound hypothyroidism with a TSH of 19 mIU/L, and undetectable fT3 and fT4, despite the patient being compliant with thyroxine. She was diagnosed with a myxoedema coma caused by immune-related enteritis and subsequent thyroxine malabsorption. The patient was treated with i.v. triiodothyronine (T3) and methylprednisolone in the ICU. While her clinical status improved with T3 replacement, her enteritis was refractory to steroid therapy. A thyroxine absorption test confirmed persistent malabsorption. Attempts to revert to oral thyroxine were unsuccessful. Unfortunately, the patient’s malignant melanoma progressed significantly and she passed away four months later. This is the first reported case of myxoedema coma that resulted from two distinct immune-related adverse reactions, namely painless thyroiditis and enterocolitis. Learning points Myxoedema coma, a severe form of decompensated hypothyroidism is a rare immunotherapy-related endocrinopathy. Myxedema coma should be treated with either i.v. triiodothyronine (T3) or i.v. thyroxine (T4). Intravenous glucocorticoids should be co-administered with thyroid hormone replacement to avoid precipitating an adrenal crisis. Thyroid function tests (TFTs) should be monitored closely in individuals with hypothyroidism and diarrhoea due to the risk of thyroxine malabsorption. A thyroxine absorption test can be used to confirm thyroxine malabsorption in individuals with persistent hypothyroidism.