Certain information about shock which has emerged from the study of a rare condition is worth noting here. When a patient with paroxysmal hemoglobinuria—also known as both hemoglobinuria a frigore and cold hemoglobinuria, immerses his hand in icy water, he experiences a chill a half hour later which is followed by the appearance of hemoglobin in the urine. Classically, this phenomenon was considered to result from the intervention of a hemoshock. We investigated such "attacks" of hemoglobinuria in three cases, inducing and studying the phenomenon several times in each subject. Usually, observations were carried out during a three hour period after the immersion of the patient's hands for ten minutes in icy water and included the following procedures:

1) Measurements of blood pressure and temperature every 5 minutes;

2) Determinations at ten minute intervals, of coagulation time, clot retraction, white cell count and differential; serum hemoglobin content, serum proteins, antitryptic power and esterase—all measurements being made on venous blood.

3) White cell count and differential measured on capillary blood obtained every 10 minutes by finger puncture at 5 minute intervals after withdrawals of venous samples.

4) Tests for the presence and amount of hemoglobin in the urine at 15-minute intervals.

Coagulation time was established in the centrifuge tube and was related to the moment when blood ceased to flow if the vertical position of the tube was changed. Clot retraction was determined by centrifuging the coagulated blood after 2 hours at room temperature and measuring the amount of serum obtained from 15 cc. of blood. The serum and urine hemoglobin content was determined photometrically. For total protein content, we used both the refractometric index of the serum and gravimetric measurements after adequate precipitation. For antitryptic power we determined the inhibitory effect of the serum upon the digestion of a solution of casein by trypsin. The quantity of esterase present was determined by the changes upon ethyl butyrate.

The data obtained were plotted as curves with time as common abscissa. Parallel variations were observed in all three patients during repeatedly induced attacks.

For almost all analyses, except for the presence of hemoglobin in urine, the variations indicated a diphasic phenomenon. (Fig. 266) The first phase was characterized clinically by hypotension and slight hypothermia. The characteristic analytical changes were leucopenia, prolonged coagulation time, reduction in clot retraction, lower refractometric serum value, lower antitryptic serum power and increased serum esterase. During this first phase of the diphasic phenomenon, hemoglobin also appeared in the serum and, when abundant in the serum, also was found in the urine. The first phase was followed by a second 5 to 10 minutes later. The clinical manifestations were a sensation of chill, varying from very slight to severe, followed by temperature elevation and slight hypertension. Analytical changes in the opposite direction from those noted during the first phase could be seen. Hyperleucocytosis, reduced coagulation time, higher retraction of the clots, elevated refractometric value and antitryptic power, and reduced esterase content were characteristic of the second phase. Hemoglobin present in the serum in the previous phase disappeared at this time.

The most interesting finding in paroxysmal hemoglobinuria was that two or three such distinct diphasic episodes followed each immersion. In all cases, the first diphasic complex appeared in about 10 minutes after immersion. It was relatively mild and lasted in general about 10 minutes, after which all values returned to pre attack.

About a half hour later, however, a second diphasic complex, much more intense in its manifestations, was noted. Hypothermia and hypotension were more marked. Leucopenia was more intense, the number of leucocytes falling to as low as 200 per cubic mm. The quantity of hemoglobin in the serum was very high, with hemoglobin spilling over into the urine in large amounts. Serum antitryptic power decreased to much lower levels than during the first diphasic complex. In some cases the coagulation time increased to 15 minutes and the clot almost failed to retract. These changes were followed by the second phase of the complex, with severe chill and manifest changes in the analyses in the opposite direction. It is the second phase of the diphasic complex that is recognized clinically as the "attack" of hemoglobinuria. The chill was often very intense, followed by a temperature above 39°C, and hypertension. Leucocytes increased to as much as 20,000 per cmm., blood coagulation time was abnormally shortened and clot retraction increased. Hemoglobin disappeared rapidly from the serum. The albumin content as well as the antitryptic power of the serum increased, while the esterase content fell. In about 30 minutes, however, all these changes were dampened and the blood slowly regained its normal characteristics. This period, with almost all manifestations slowly returning to normal often was followed by a third diphasic complex, not clinically evident but revealed by hematological findings. In most cases, it appeared about two hours after immersion. While it was much milder than the first two, its diphasic character was quite clear. Occasionally the patient reported a slight sensation of cold. The amount of hemoglobin in the serum was less than during the first complex, and hemoglobinuria was never seen. Figure 266 shows these findings in a typical succession of these complexes.

The clinical data and analytical changes of blood and urine

Fig. 266. The clinical data and analytical changes of blood and urine following the immersion in icy water of the hand of a subject with hemoglobinuria a frigore reveal three distinct diphasic phenomena corresponding respectively to three separate hemo shocks. Their intensity appears correlated to the degree of the occurring leucopenia.