Discussion 299 



every respect similar to that following the rapid intravenous injection of 

 saline. These facts point to a dependence of the sodium and water output 

 on blood pressure or on blood flow ; there is no direct relation to volume. 

 A fact worth remarking is that the diuresis may continue several hours 

 after the tachycardia stops. This suggests that the effect of the abnormal 

 circulation on the renal tubules is mediated by a slowly acting mecha- 

 nism, possibly a renal hormone. Experiments in animals in which the 

 functions of the two kidneys have been compared also prove that the 

 adrenal is not essential and that the receptor must be in the kidney. 

 When one renal artery is gradually narrowed the sodium and water 

 excretion of the corresponding kidney may fall sharply before a fall in 

 PAH and creatinine clearance can be demonstrated. Probably the kidney 

 responds even to the slightest reduction in intrarenal blood pressure by 

 an increased tubular reabsorption of sodium chloride and water. 



Fejfar: I quite agree with you in all points. It is also my personal 

 view that this reaction might start in the heart itself. In all these types 

 of circulatory disturbances (mitral stenosis, pericarditis, acute heart 

 failure, hypoxaemia, anaemia), and in muscular effort, the only common 

 factor is a very low oxygen content in the central venous blood. When 

 we gave oxygen to patients with normal cardiac output, the cardiac 

 output did not change. When oxygen was given to patients with a 

 lowered cardiac output, the output increased ; there is, therefore, indirect 

 evidence that if more oxygen is given to the heart muscle in congestive 

 failure the performance of the heart improves. 



Milne: Have you any observations on a similar correlation, or the 

 reverse, in other conditions besides congestive heart failure associated 

 with nocturnal diuresis? In starvation and cortisone overdosage, parti- 

 cularly, a similar reversal of the normal diurnal rhythm may be seen. 



Fejfar: No, I have no comments to make. 



Milne : I would agree with all the points you make regarding the diag- 

 nosis of potassium deficiency in heart failure, but I think that most 

 clinicians are now using a very useful clinical method of diagnosis — excess 

 sensitivity to digitalis. Of course, as you say, it can be checked by 

 balance or exchangeable potassium if necessary. 



Fejfar : You are right about digitalis, but of course this usually occurs 

 in advanced stages of potassium deficiency. When patients are in potas- 

 sium deficiency it takes weeks and weeks to restore the balance. This is 

 not just an academic question because we had three deaths due to these 

 metabolic changes shortly after mitral valvulotomy. When a patient 

 already has a negative balance with loss of potassium, the added opera- 

 tive trauma and hypotension will easily lead to so-called metabolic death. 



Olesen: I have had the opportunity of studying the problem of the 

 diagnosis of potassium depletion in congestive heart failure with the 

 dilution methods used in Boston (McMurrey et al. (1956). Metabolism, 

 5, 447). I would say first that the diagnosis is not very easy; in fact it is 

 probably impossible to make it by the dilution methods alone. We 

 found, however, that there were very marked changes in the body 

 composition of these patients with congestive failure. There was a rela- 

 tive decrease in the total intracellular mass, as expressed either by total 



