In appendix A a brief outline is given of the six tests (W. R. on blood serum and spinal fluid, cell count, globulin test, albumin test, gold sol test). This is not intended as a complete working manual but rather as indicating the methods used in diagnosis in the cases presented herein. For more complete details the reader may be referred to textbooks on the subject of serology, among which may be mentioned Kaplan: “Serology of the Nervous System”; Plaut, Rehm and SchottmÜller: “Leitfaden zur Untersuchungen der ZerebrospinalflÜssigkeit”; Kolmer: “Infection, Immunity and Specific Therapy,” and, for the Wassermann technique, an article by Dr. W. A. Hinton in M. J. Rosenau’s “Preventive Medicine and Hygiene.”

Our own W. R’s. have been performed at the Wassermann laboratory of the Massachusetts State Board of Health (formerly the Neuropathological Testing Laboratory, Harvard Medical School), under the supervision of Dr. W. A. Hinton. The other tests are performed at the Psychopathic Hospital. It is very important that a close relationship should exist between the clinician and the Wassermann laboratory if the most is to be obtained from the reactions. This relationship has been effectively close between the authors and the above-mentioned laboratory; and has enabled us to get very much clearer ideas about certain cases than could otherwise have been obtained.

Cell Count. In order to obtain the number of cells per cmm., the examination should be made of the fresh fluid as soon as possible after this is withdrawn. The most convenient counting chamber for this purpose is the so-called Fuchs-Rosenthal counting chamber, the ruled spaces of which contain slightly over 3 cmm. (an ordinary blood cell counting chamber may be used). According to the method used by us the cells are stained in a pipette with Unna’s polychrome methylene blue. Using a white-counting pipette, stain is drawn up to the first or second marking and the remainder of the pipette filled with spinal fluid. This makes no change in the dilution for practical purposes. After two or three minutes the staining is satisfactory and the counting may be done. With this stain a differential count may be made. Plasma cells stain a lavender as contrasted to the blue of the lymphocytes. The characteristic halo surrounding the eccentric nucleus is visible. The blood cells do not assume color with this stain; hence it is unnecessary to add any acetic acid.

For permanent preparations, and more accurate differential counts of the spinal fluid, the Alzheimer method may be used. The technique is given in a paper by H. A. Cotton and J. B. Ayer as follows:^[151]

1. Lumbar puncture in the usual manner.

2. 96% alcohol, in proportion to twice the amount of cerebrospinal fluid, is added drop by drop and well mixed.

3. Centrifuge the mixture for one hour at high speed in a glass tube with conical end. (An ordinary electric urinary centrifuge apparatus can be employed, the tube to be well stoppered to prevent evaporation.)

4. The supernatant fluid is poured off, leaving a small coagulum in the bottom of the tube.

5. Add absolute alcohol—alcohol and ether—ether, each separately for one hour, to dehydrate and harden coagulum.

6. The coagulum can now be gently loosened from the bottom of the tube by a long needle. The tube is then inverted, and the coagulum allowed to fall into the hand by a quick tap on the end of the tube. Care must be taken not to squeeze or handle the coagulum. The hand is placed over a small homeopathic vial, containing thin celloidin, and the coagulum allowed to drop into the celloidin, where it remains over night (twelve hours usually).

7. Coagulum is placed in thick celloidin which is allowed to evaporate slowly. It is then mounted on blocks and sections cut 14µ in thickness.

8. The sections are stained and mounted according to the following procedure:

(a) Remove celloidin by absolute alcohol and ether.

(b) 80% alcohol.

(c) Water.

(d) Sections are carried on glass or platinum needle into a dish of Pappenheim’s pyronin-methyl green stain and kept in a water bath at 40° C. five to seven minutes.

(e) Quickly cool dish in running water.

(f) Wash off superfluous stain in plain water.

(g) Absolute alcohol to differentiate—until no more stain comes away from section.

(h) Clear in Bergamot oil.

(i) Mount in balsam.

The normal cell count may be stated as being up to 6 cells per cmm.; from 6 to 12 cells may be considered as suggestive of pathological condition and more than 12 cells per cmm. as definitely pathological. The type of cell in syphilitic diseases is preponderantly the small lymphocyte. A low percentage, that is, very rarely over 20%, of large lymphocytes, endothelial phagocytic cells, polymorphonuclear leucocytes and plasma cells may also be found. The finding of plasma cells in any number in the spinal fluid is suggestive although not conclusive evidence for the diagnosis of paretic neurosyphilis.

Globulin is an albumin which is precipitated by half saturation with a salt. A very simple and satisfactory test is known as the Nonne-Appelt test, which has been modified by Ross-Jones. Into a test tube of small diameter, run 1 cc. of spinal fluid. Place under this fluid with a pipette, 1 cc. of a saturated solution of ammonium sulphate ((NH₄)₂SO₄). If any globulin is present a white, sharply-defined ring will form at the junction of the two fluids. According to our readings, a ring that is just visible with the aid of a black background is called 1+, a ring that is just visible without the black background, 2+; a ring easily perceptible, 3+ and a relatively very heavy ring, 4+. On shaking the tube, if globulin is present, the fluid will show turbescence.

Another simple globulin test used in our laboratory as a check on the Nonne-Appelt test is the Pandy test. A few cc. of a clarified 10% solution of phenol are placed in a watch glass. One drop of spinal fluid is run into this solution. A milky turbescence indicates globulin.

The presence of globulin in the spinal fluid is always an indication of abnormality of the cerebrospinal axis. There is nothing differential in this finding as it occurs in all inflammatory processes. However, it is characteristically present in most cases of neurosyphilis (exception to the rule: the pure vascular type does not show globulin in a very high per cent).

Albumin Test. Albumin in small quantities is present in all spinal fluids. Increase over the normal amount occurs in pathological conditions such as most cases of neurosyphilis, especially in those in which globulin is found. Any albumin precipitant may be used for rough clinical calculation, comparing the amount of precipitate with that from the normal fluid. Our method is to place 1 cc. of spinal fluid in a small test tube of about 5 mm. diameter and to precipitate the albumin by the addition of 3 drops of 33?% of trichloracetic acid. This test has its chief value as confirmatory of the globulin test, since in the vast majority of instances where globulin is found there will also be found an increase in albumin.

The Gold Sol Reaction is an empirical test discovered by Carl Lange in the utilization of the work of Zsigmondi with solutions of colloidal gold and albumins. Briefly the details of the test are as follows:

Ten tubes are set up in a rack. To the first tube 1.8 cc. of a 0.4% of salt solution is added and to each of the following tubes 1 cc. of this solution. Then to the first tube containing 1.8 cc. of salt solution one adds 0.2 cc. of the spinal fluid to be tested. This gives a dilution of 1 to 10. From this tube 1 cc. is pipetted into the second tube and this process continued through the ten tubes. This gives dilutions of spinal fluid of 1 to 10, 1 to 20, 1 to 40, etc., to 1 to 5120 in the last tube. Then 5 cc. of colloidal gold solution is added to each tube. A positive reaction is indicated by the precipitation or throwing down of the colloidal gold into its metallic form. This produces a change in color. This precipitation may be partial or complete and the amount of precipitation is indicated by the color and is read as follows:

The unchanged fluid is called 0; a slight change giving a red-blue as 1; a further change giving a blue-red as 2; a straight blue as 3; a lavender or violet as 4; and the colorless fluid representing complete precipitation as 5. The numbers are placed in a row, indicating the tube in which the color occurs. The fluid from a case of paretic neurosyphilis will give a complete precipitation beginning in the first tube and running through a number of tubes and then grading off. It may be indicated 5 5 5 5 4 3 1 0 0 0. The characteristic reaction of fluids from tabetic and diffuse neurosyphilis is less strong than from the paretic. The greater part of the reaction will take place, however, in the first five tubes, but as a rule it will not begin very strongly in the first two. A characteristic reaction is 1 2 3 3 2 1 0 0 0 0. Another reaction that may be considered characteristic of the tabetic or diffuse form is 3 3 3 2 1 0 0 0 0 0. Fluids from non-syphilitic cases as a rule give a reaction having its greatest intensity beyond the fifth tube, that is, in the high dilutions.

A reaction characteristic of brain tumor or tuberculous meningitis is 0 0 0 0 1 3 3 2 1 0.

The conclusions that may be drawn from the gold sol reaction have been summarized by one of the authors as follows:

1. Fluids from cases of general paresis will give a strong and fairly characteristic reaction, especially if more than one sample is tested, in the vast majority of cases.

2. Very rarely a general paresis fluid will give a reaction weaker than the characteristic one.

3. Fluids from cases of syphilitic involvement of the central nervous system other than general paresis often give a weaker reaction than the paretic, but in a fairly high percentage of cases give the same reaction as the paretics.

4. Non-syphilitic cases may give the same reaction as the paretics; these cases are usually chronic inflammatory conditions of the central nervous system.

5. When a syphilitic fluid does not give the strong “paretic reaction,” it is good presumptive evidence that the case is not general paresis; and this test offers a very valuable differential diagnostic aid between general paresis, tabes and cerebrospinal syphilis.

6. The term “syphilitic zone” is a misnomer, as non-syphilitic as well as syphilitic cases give reactions in this zone; but no fluid of a case with syphilitic central nervous system disease has given a reaction out of this zone (test thus valuable negatively). Any fluid giving a reaction outside of this zone may be considered non-syphilitic.

7. Light reactions may occur without any evident significance, while a reaction of no greater strength may mean marked inflammatory reaction.

8. Tuberculous meningitis, brain tumor and purulent meningitis fluids characteristically, though not invariably, give reactions in higher dilutions than syphilitic fluids.

9. The unsupplemented gold sol test is insufficient evidence on which to make any diagnosis, but used in conjunction with the Wassermann reaction, chemical and cytological examinations, it offers much information looking toward the differential diagnosis of general paresis, cerebrospinal syphilis, tabes dorsalis, brain tumor, tuberculous meningitis, purulent meningitis.

10. We believe that no cerebrospinal fluid examination is complete for clinical purposes without the gold sol test.

The Wassermann reaction as carried out in the Wassermann Laboratory is based on the principles of the original method—the only essential modification consists in the employment of cholesterinized alcoholic extracts of human hearts as antigen instead of aqueous extracts of foetal livers from cases of congenital syphilis. Experience has shown that properly standardized antigens made from human hearts are much more sensitive in the detection of true cases of syphilis.

Antigens. Three antigens are used, each being an alcoholic extract of human heart which is saturated at room temperature with cholesterin. These antigens differ slightly in their sensitiveness. Before the test is made each antigen is diluted with 0.85% salt solution in the proportion of four parts of the cholesterinized antigen extract to sixteen parts of 0.85% salt solution. The amount to be used, the dosage, is carefully determined by testing each antigen against a large number of known positive and known negative specimens of blood. The dosage of the antigens employed is less than one-half the amount which inhibits hemolysis when the antigen is incubated for one hour with the hemolytic system which consists of complement, amboceptor and cells in the proper proportions. These antigens are designated as A, B, and C. Antigen A is the most sensitive. B and C are very similar to each other quantitatively and qualitatively.

Specimens to be tested. The serum which separates from the clot is withdrawn, centrifugalized if necessary, and then heated at 55 degrees for thirty minutes. 0.1 cc. of serum is used in the test and 0.2 cc. of each specimen is used as a control to exclude the presence of anti-complementary substances. Spinal fluids are tested in two ways. As a routine 0.5 cc. of the spinal fluid is used in the test and 1.0 cc. is used in the control; or when especially requested spinal fluids are titrated by using respectively 1.0, 0.7, 0.5, 0.3, and 0.1 cc. of the spinal fluid for each test and 1.0 cc. of spinal fluid for the control. Spinal fluids are not inactivated.

Complement. The complement is obtained from the serum of guinea pig’s blood. No complement is used when older than eighteen hours. A 10% solution and 0.85% salt solution is used in the test. The amount used is twice the minimum quantity necessary to hemolyze the sensitized cells.

Sheep’s Corpuscles. A 5% suspension of sheep’s corpuscles in 0.85% salt solution is prepared from defibrinated sheep’s blood. The corpuscles are washed three times and for each washing four to five times as much 0.85% salt solution is used as the original volume of the defibrinated blood.

Amboceptor. The amboceptor is prepared by injecting sheep’s corpuscles into a rabbit. The serum of this rabbit which contains amboceptor is diluted with 0.85% salt solution so that 0.25 cc. will hemolyze 0.5 cc. of a 5% suspension of sheep’s corpuscles. In the test twice the quantity or 0.5 cc. of amboceptor is used.

Sensitized Cells. The sensitized cells consist of equal parts of washed sheep’s corpuscles and diluted amboceptor. This mixture is incubated in a water bath at 37° C. for a half hour to effect the sensitization of the cells.

Technique of the Wassermann Test. One-tenth cubic centimeter of each inactivated specimen of serum and 0.5 cc. of each uninactivated specimen of spinal fluid is pipetted into a separate tube. A mixture is freshly prepared in salt solution, each cubic centimeter of which contains the proper amount of antigen A (the most sensitive antigen), and two units of a 10% solution of guinea pig serum (complement). One cubic centimeter of this mixture is pipetted into each test tube. These tubes are then incubated for forty minutes in a water bath at 37° C. At the end of this period, sensitized cells are added, and the tubes are again incubated in a water bath at 37° C. for one hour. Each specimen which shows any degree of inhibition of hemolysis is retested in the afternoon. For this second test antigen A is again used and in addition antigens B and C. A control is also made for each specimen retested to eliminate any possibility of the inhibition of hemolysis being due to anti-complementary substances in the serum or spinal fluid tested. The technique of the second test differs in no wise from that of the first, except for the use of a control in each retested specimen and the employment of three antigens instead of one. The degree of positiveness is noted for each retested specimen and compared with the degree of positiveness obtained for the corresponding specimen with the same antigen-complement-salt solution mixture in the morning’s test. The specimen is retested on the next day when discrepancies occur between the morning reading for antigen A and the afternoon reading for antigen A. From the above description it will be noted that the negative specimens have but a single test with one antigen only, while the positive specimens are retested, thus permitting a confirmation of any positive reaction. In this way attention is focalized on the positive specimens.

Interpretation of Results. Antigen C (the weakest of the three antigens) is used entirely for diagnostic purposes and any specimen showing the slightest degree of inhibition with this antigen and stronger degrees of inhibition with the other antigens is reported as positive. The specimens which are strongly or moderately positive with antigens A and B and negative with antigen C are reported as doubtful. In testing spinal fluids by the titration method, antigen C is used and the readings are based upon the degree of inhibition of hemolysis noted. The intensity of this inhibition is indicated by Arabic numerals: “5” indicates complete inhibition, while “1” means a faint cloudiness, hence a weak reaction. Intermediate numbers show relative intensity varying between complete inhibition “5” (strong positive) and slight inhibition “1” (weak positive); “—” equals no inhibition (negative).

Although it is commonly believed that the recent administration of antisyphilitic treatment will affect the reaction by making it negative, this is not our experience, and it is, therefore, not necessary that treatment be withdrawn for a short period before the specimen is submitted for examination.

The reaction as carried out in this laboratory has the following diagnostic significance: Positive indicates syphilis, except very rarely in acute febrile conditions such as malaria and pneumonia. Negative does not exclude syphilis. In obscure conditions a series of less than three negatives has little diagnostic significance. Doubtful suggests syphilis. It is therefore advisable to submit three or more specimens in such a case, and interpret a persistently or predominatingly doubtful reaction as indicative of syphilitic infection.

Bruck Test. A new serum test for syphilis has recently been described by C. Bruck.^[152] Following are recent results in our laboratory with this test.^[153]

This new test for the diagnosis of syphilis by C. Bruck has aroused much interest. The scientific standing of Bruck and the simplicity of the technique led us to overcome our prejudice, that has been the offspring of the numerous tests that have been offered of late. Bruck states that since the discovery of the complement fixation test for syphilis by Wassermann, Neisser and himself in 1906, he has been trying to find a simple chemical reaction that would take the place of the complicated technique of the Wassermann reaction. This method, as he has published it, was worked out and is being used at the front, in the present war, where complete laboratory equipment is not available.

Commencing our experiments with a great deal of scepticism, we were much surprised at the results obtained, which are given below. Whatever may be the final status of the test in the determination of syphilis, we feel that there is a great deal of interest in the fact that this simple chemical reaction does pick out certain differences in the composition of blood sera and that apparently a large number of syphilitic sera differ in their chemical composition percentage from the majority of non-syphilitic sera.

The technique, while exceedingly simple, offers many chances for errors and individual variations so that we have thought it well to give directions and cautions at some length.

Bruck’s^[154] technique is described as follows: “The test is made with 0.5 cc. clear serum in a test tube, to which is added 2 cc. of distilled water, and the whole shaken. Then, with a precision pipette, 0.3 cc. of the ac. nitr. purum of the German pharmacopeia is added and the whole thoroughly shaken and then set aside at room temperature for ten minutes. Then 16 cc. of distilled water at room temperature is added, and closing the tube with the finger, it is shaken up and down three times carefully, not vigorously enough to make it foam. This is repeated ten minutes later, and the tube is then set aside for half an hour. By this time the precipitate is entirely dissolved in the tube with the normal serum, while the syphilitic serum shows a distinct, flocculent turbidity. In two or three hours, or better still, in twelve hours, the gelatinous and characteristic precipitate is piled up on the floor of the test tube.”

The acid is prepared by diluting the Acidum nitricum of the U. S. P. (Sp. gr. 1.403) with distilled water until the hydrometer shows the specific gravity 1.149, which corresponds to the nitric acid of the German pharmacopeia, but since this requires a special hydrometer, a simpler method is to make a 25 per cent solution of the Acidum nitricum, which will give about the proper specific gravity.

The serum is obtained by allowing 10 cc. of blood to stand at room temperature for an hour, and then centrifuging. Serum that has stood for some time may be used as well as the fresh, and even bloody serum does not seem to confuse the results to any great degree. The serum gives the same results with or without inactivation. Post mortem blood gave results as constant as that obtained during life, in the few cases that we had in this series. But the reaction may be influenced markedly by the size of the test tubes. We have found that the 13×1.9 cm. is the most favorable size.

When one first thinks of this test it appears very simple and probably somewhat crude as a chemical reaction, but there are certain precautions that must be observed, and several hundred normal and syphilitic sera should be tried before the investigator can feel that he has a refined routine technique. There is the personal equation which must be watched, for here is probably the greatest source of error, and readily explains why two different persons get widely varying results with the same sera if they have done only a few dozen tests. We must take it for granted that the reaction is a quantitative one, where some positive reactions may differ only slightly from the normal non-syphilitic, and, furthermore, any normal serum may be made to give a positive reaction, and almost any positive serum be made to give a negative by improper manipulation at some point in the test. There are as many places for error to creep in as there are steps in the process. Bruck has omitted many details in his publication, which allow personal variations, and so we have tried to develop a routine process that will eliminate as many of these as possible.

We shall here attempt to explain the methods which we have found most satisfactory and at the same time indicate the places where error is likely to occur. The 0.5 cc. of serum is added to 2 cc. of distilled water, and shaken thoroughly. Now add slowly exactly 0.3 cc. of acid from a precision pipette, care being taken it does not flow down the side of the tube. The tube should be shaken gently while the acid is being added, for this prevents the formation of a flocculent precipitate in normal serum which is difficult to dissolve later. After the acid is added shake each tube gently to make sure that these flakes do not persist. It is difficult to shake each tube in exactly the same manner, as must be done if we expect uniform results.

The first 250 tests of this series were made by allowing the tubes to stand for ten minutes as Bruck advocates. Then we found that practically all sera gave a positive reaction if allowed to stand 15–20 minutes, and so in the other tests of the series an attempt was made to make the reaction more sensitive by allowing the tubes to stand only 6–7 minutes. During this time the tubes should be shaken gently once or twice. The manner in which the 16 cc. of water is added also influences the reaction. If allowed to flow freely in upon the precipitate, the positive may be forced into solution as well as the negative. Both pipette and tube should be slanted and the water allowed to flow down the side of the tube without disturbing the precipitate. If all has gone well up to this point, we may see a marked difference between the normal and syphilitic precipitates, in that the normal will begin to go into solution at once, thus clouding the water, while a positive precipitate will be composed of large flakes which show little or no tendency to go into solution or cloud the water above. It must be remembered that the most flocculent positive precipitate will go into solution if the fluid is splashed or shaken too hard while the tube is being inverted. If any doubt as to the character of the precipitate now exists, it may be allowed to stand ten minutes longer, and again inverted as before, or even repeated several times during the next hour or two. We see no reason why the tubes should be left to stand over night, for during this time a precipitate usually settles in the normal tubes. This, however, differs from the syphilitic precipitate in that it is still finely granular and goes back into solution readily when the tubes are inverted.

In view of these possible grounds for error, it is only logical to run controls of known positive and known negative sera along with each group of unknown bloods, and even then certain tubes will seem doubtful, in which event the test should be repeated with added precaution to see if a definite positive or negative reaction may be obtained.

In the last tests of this series we seemed to aid the reaction by rendering the serum-water solution alkaline by one or two drops of 10 per cent potassium hydroxide before the acid was added. The positive sera have a larger precipitate, while the normal seem to dissolve more readily.

The tests here reported were made on blood sera obtained from patients admitted to the Psychopathic Hospital and its Out-Patient Department. As a routine Wassermann test is made on each patient who enters the hospital, it was only necessary to take another tube of blood from each patient, and check the results in each instance with the Wassermann reaction. As it takes several days to get the report from the Wassermann laboratory of the State Board of Health, there was no chance of being prejudiced by a previous knowledge of the Wassermann reaction. The cases for the most part were those of mental disease; the majority in good general physical health.

A comparison of the total number with the Wassermann reaction shows that there was a general agreement of 304 of the 405 cases tested, or a percentage agreement of practically 75%. In considering the cases of syphilis of the central nervous system in a group by themselves, we find that the agreement is closer, since 69 of the 79 cases tested, or 87% agreed without any question of doubt. It will be noted that in several cases of general paresis, the Wassermann reaction, which was repeated at intervals, was negative, and in most of these cases the Bruck test was negative also. Our few cases of congenital and latent syphilis also checked very closely with the Wassermann test. In the various groups of mental cases in this series, no factor of interference was discovered. It is also of interest that in the cases where the blood was obtained post mortem, the Bruck test agreed with the Wassermann result obtained on ante mortem blood serum. Further work on post mortem sera will be reported. Some of the patients not included in the syphilitic groups that have a negative Wassermann and no clinical signs of syphilis, give a history of previous infection at some time, which might partly account for the variations in the two tests.

CONCLUSIONS

1. We present results of the Bruck sero-chemical test in 405 cases. In 101 of these cases there were definite clinical manifestations of syphilis, in which the Wassermann and Bruck tests agreed positively in 74 or 75%. The two tests agreed negatively in 12 instances, and were at variance in 15.

2. In the group which showed syphilis of the nervous system we had 64 cases of clinically certain general paresis, of which the Wassermann and Bruck tests agreed in 54 instances, or practically 85%. In other forms of central nervous system involvement the agreement was 100% in the 15 cases tested.

3. In the cases with no apparent involvement of the nervous system the agreement was somewhat less, being 76%. This may be in keeping with the fact that the Wassermann test was not so strongly positive in these cases.

4. The advantages of the test are: (1) the short time required to do the test; (2) the limited amount of apparatus necessary, and (3) the simplicity of the technique.

5. The disadvantages of the test seem, for the most part, to be bound up in the personal variations that are apt to occur.

6. We are here dealing, most probably, with a quantitative chemical difference in the protein content of syphilitic and non-syphilitic sera, the nature of which is not understood by us. It is our hope that this may be brought to light in the near future in the field of chemistry.