phosphatides,1 5 lipoids; by hydrolysis, co-workers,6 … any light upon the chemical differentiation between virulent and avirulent strains. Sabinet al. haveshown,however, that the

STUDIES OF A CARBOHYDRATE-LIPOID COMPLEXFROM THE HUMAN STRAIN TUBERCLE BACILLUS H37*

GEORGE V. KROPP AND CLEAVELAND FLOYD

The systematic study of cellular reactions to purified chemicalfractions of the tubercle bacillus was first demonstrated by Sabin andher co-workers.'5 16, 17, 18 These chemically purified derivatives wereprepared by Anderson of Yale. The above studies have contributedgreatly to our knowledge of the pathogenesis of tuberculosis. Through-out all of the above work, however, no one specific factor was isolatedto which the so-called "virulence" of the bacillus can be attributed. Therehas likewise as yet been no chemical evidence forthcoming which hasshed any light upon the chemical differentiation between virulent andavirulent strains. Sabin et al. have shown, however, that the lipoid sub-stances have exceeded all others in their ability to act as stimulantsand maturation factors in the life cycle of the epithelioid cells of thehost. This cellular phenomenon was demonstrated repeatedly with thephosphatide fraction of the bacillus. Later it was found that phthioicacid, a new fatty acid isolated by Anderson from the phosphatide frac-tion, was by itself capable of stimulating monocytes and epithelioidcells resulting in the formation of the tuberculous tissue. Although muchwork has been done with the phosphatide fraction, there has been noevidence assigning the latter to any specific role in the immunologicalmechanism of the disease. A great deal of investigation has been devotedto the antigenic properties of the lipoids, including fats, compoundlipoids such as the phosphatides,1 4 5 cerebrosides, and the derivedlipoids; also those substances obtained from the simple and compoundlipoids by hydrolysis, e. g., sterols and fatty acids. Recently, Gerstland his co-workers,6 using mycolic acids derived from the waxes ofhuman, bovine, and avian tubercle bacilli, and from the leprosybacillus, demonstrated that these hydroxy acids are responsible forthe presence and persistence of the necrotic lesions induced by the abovebacilli, but are not related to the elective pathogenicity of these differentstrains. The latter observations would tend to corroborate the findings

* From the Boston Health Department. This study was sponsored by the BostonTuberculosis Association.

YALE JOURNAL OF BIOLOGY AND MEDICINE

of other workers' that lipoid substances stripped of all bacillary debris,protein, carbohydrates, etc. are not related to virulence or native re-sistance. In combination with other chemical components, however,whether within the living organism or in isolated form from the latter,the subject of lipoid specificity becomes one which deserves much furtherstudy. Some light might be shed on the nature of lipoid complexes bytechniques which would permit isolation in pure condition of thevarious lipoids obtained by alcoholic or ether extractions of bacteria andtissues. Such complexes as the 0 and Vi antigens from the typhoidbacillus, the Forssmann body, the human blood groups, might serve aslogical starting points. The successful development of such techniquesmight well lead to the isolation of a "P" factor, as discussed by Middle-brook,'3 namely, "that factor or component within the organism havingthe capacity for causing progressive tuberculosis by bacterial multiplica-tion in vivo."

It was previously shown by one of the authors9 that a protein-freepolysaccharide fraction from the human strain bacillus H37, elicitedcertain cellular and immunological responses, when given repeatedlyand in increasing doses, in animals and human beings. In due course ofthese studies observations were made with other carbohydrate protein-free fractions isolated from the tubercle germ mass; some of these frac-tions contained lipoid substances in varying amounts. Some of the latterwere tested in small groups (6 to 12) of animals using an equal numberfor controls. It was found that the lipoid-containing fractions conferreda much greater degree of immunity against subsequent infection than didthe carbohydrate fractions free from lipoidal material. Even thoughthe number of animals used in these preliminary trials was small, wefelt the degree of protection conferred warranted further study. Conse-quently a lipoid-carbohydrate complex was isolated (as described be-low) and used in the present laboratory and clinical investigations.

Materials and methodsThe complex was isolated from alcohol-ether extracts of virulent H37

bacilli. The bacilli were grown on Long's synthetic medium for a period of 6to 8 weeks. The cultures were then filtered through Buchner funnels andplaced in a mixture of equal parts of alcohol and ether. This mixture wasallowed to stand at room temperature with intermittent shaking for a two-weekperiod. The bacilli were again filtered through Buchner funnels and the filtratesaved. The filtrate was passed through a Berkefeld filter and placed over asteam-bath and allowed to evaporate until a sticky residue remained. Theresidue was extracted with ether until no further amber color appeared in the

28

CARBOHYDRATE-LIPOID THERAPY OF TUBERCULOSIS

extract. To the pooled ether washings was added 95 per cent alcohol until nofurther precipitation occurred. The precipitate was separated by centrifugationand when dried formed a thin opalescent scale. In a mortar this scale wasground to a powder from which aliquots were taken to make up the desiredaqueous concentrations.

The product when ready for animal testing was free from protein andprotein-nitrogen according to the Biuret, ninhydrin, and sodium fusion tests.It gave an intense Molisch reaction and did not show reducing sugars untilhydrolyzed with 5 per cent sulfuric acid. It formed typical glucosazones withphenylhydrazine hydrochloride. The complex was slowly soluble in water andformed an opalescent solution. Further chemical study of the complex is nowbeing undertaken.

ToxicityThe tests for toxicity were carried out in tuberculous and in non-

tuberculous guinea-pigs and rabbits. It was found that tuberculousguinea-pigs succumbed to 5 mg. given intraperitoneally within 48hours. Non-tuberculous animals tolerated the above amount, but diedat the end of 72 hours from 20 mg. given by the same route. Normalrabbits tolerated 3 mg. intravenously without manifesting any untowardeffects. Repeated intravenous and subcutaneous injections of the complexwere given to rabbits and guinea-pigs, respectively, to rule out the pos-sibility of allergy and anaphylaxis. The rabbits received a total of 50mg. intravenously. They were then permitted to go untreated for twoweeks, at the end of which time they were given 3 mg. intravenously.There was no sign of any symptoms (rise in temperature, in breathing,etc.) in any of the rabbits or guinea-pigs so tested. Subcutaneous andintracutaneous injections were also made with the complex. It wasfound that 1 mg. given intracutaneously produced an induration witherythema in both the normal and the tuberculous guinea-pig. Whenthis amount was increased to 4 mg. abscess formation occurred inthe tuberculous animal within the following three or four days.

Post-mortem examinations performed on animals (tuberculous andnon-tuberculous) that succumbed to intraperitoneal injections of thecomplex, revealed no gross pathology in any of the organs otherthan a peritonitis. No microscopic study was made on the peritonealtissue of these animals. Temperatures (rectal) were seen to rise from1 to 2 degrees in both tuberculous and non-tuberculous animals follow-ing intraperitoneal injections of sub-lethal doses of the complex, butreturned to the original level after 24 hours. Temperatures rose from2 to 4 degrees following advent of lethal doses of the latter. There wasno fluid accompanying the peritonitis in any of the autopsied animals.

29


Experimental

It was decided to determine first the prophylactic effect of thisfraction on 24 animals; 12 of these were treated and 12 were used ascontrols. The number of animals and the variety of conditions have beenlimited due to the small yield of the fraction. The weight of the animalsvaried between 350 and 450 gm. All animals were weighed previous toinjections of any kind. All of the animals were skin-tested with 0.5per cent O.T. before treatment or injection, and no positive reactors wereseen. The 12 immunized animals received a total of 14 mg. of thefraction subcutaneously during an eight-week period. The injectionswere given on alternate days three times a week. A rise of 1 to 2 degreesin temperature was observed following each immunizing dose of thefraction during the entire period of immunization. We have as yet noexperimental evidence to account for this phenomenon, since all animalswere skin-test negative, and were apparently healthy in all other respects.A period of five days was allowed to elapse at the end of the three-weekimmunization period at which time all 24 animals were given aninfecting dose of 0.01 mg. of virulent H37 tubercle bacilli in the rightgroin. No further treatment was given to the immunized animals afterthe infecting dose. The results of the above experiment are recordedin table 1.

TABLE 1

Controls TreatedDuration Duratio.of experi- of experi-

Cause of Degree of ment. Cause of Degree of ment.No. death tuberculosis Days No. death tuberculosis Days

1 Tuberculosis + + + 40 1 Tuberculosis + + + 322 Tuberculosis + + + 46 2 Tuberculosis + + + 1573 Tuberculosis + + 80 3 Tuberculosis + + + 1854 Tuberculosis + + + 145 4 Sacrificed + 2975 Tuberculosis + + + 178 5 Sacrificed + 2976 Tuberculosis + + + 178 6 Sacrificed + + 2977 Tuberculosis + + 186 7 Sacrificed + 2978 Tuberculosis + + + 186 8 Sacrificed + + 2979 Tuberculosis + + + 190 9 Sacrificed + 29710 Tuberculosis + + + 198 10 Sacrificed + + 29711 Tuberculosis + 198 11 Sacrificed + + 29712 Sacrificed + 297 12 Sacrificed + 297

Ave. no. days-246 Ave. no. days-160

Degree of tuberculosis:+ + + General disease+ + Local site and liver and spleen+ Local site and lymph nodes

30


It will be seen in table 1 that all of the control animals died oftuberculosis with one exception. Sections of the liver, spleen, and lungof this animal were studied microscopically and no evidence of diseasewas seen in any of the above-mentioned organs. The inguinal andtracheo-bronchial lymph nodes, however, were enlarged and containedcaseous cores. Smears made from the caseous material of these glandsshowed acid-fast bacilli.

It is of interest to note that the animal which was the largest inthe control group at the beginning of the experiment (455 gm.) atthe time of death weighed 585 gm. Of the treated animals that diedof tuberculosis the gross and microscopic pathology was typical inall respects. In the immunized animals where the extent of disease was++, sections made for microscopic study of the infected organs showeda healed type of disease; that is, the tubercles seen in these sections werewithout central necrosis and were accompanied by a considerableamount of fibrosis. A few scattered acid-fast bacilli were demonstrable,however, in the lesions of these organs. All of the treated animals whichwere sacrificed had shown a progressive gain in weight throughout theexperiment. It would seem from the microscopic sections studied thatthe resistance in the treated animals (with three exceptions) had beenincreased to such a degree that the establishment of generalized diseasewas retarded. We are unable, as yet, to account for the extent of diseaseseen to develop in the three treated animals which succumbed to thedisease.

Following the above experiment it was decided to test the therapeuticproperties of the complex, as well as to repeat the prophylactic ex-periment, and to use a larger number of animals. For this purpose 58animals were selected, 20 for prophylaxis, 20 for therapeutic treatment,and 18 as controls. The infecting dose was 0.01 mg. of virulent strainH37 given subcutaneously in the right groin. An interval of 30 dayswas allowed to elapse before treatment with the complex was started onany of the animals. The amount of complex administered, the route ofinjections, and the period of treatment were the same for this pro-phylactic group as in the first experiment. The therapeutic group wastreated bi-weekly for a six-week period, during which time theseanimals received the same amount of the complex (6 mg.) as the pro-phylactic group. No further treatment of any kind was given to thisgroup. The results are presented in table 2.

Unfortunately an epidemic of spontaneous pneumonia, type 19,broke out in the animal colony forcing us to sacrifice all of the animals

31


r- e4 o000% 0%0 c V' -ve 0000001o% 00 o0 e 00 1 -oo - \0o V cnC%ON% O\%C00

-0+++++++++++++±+++++±±++±±±±+ ±0o++++++±+++++++++++

V- C\i4 cf~VV\\0 F--O 0%aN0 '-4 C~4 e VV' \3 F--C1 V~-4 V- V- V-4 -4--l -4 -4

1 \O0,_0F-- 00 F-- 00 (N 0 C4 F- F- 00000aa-00 0%N0%%ON%0%0%C0OsCt0%- ON0aGsON0ONCsON

:+++++++ + + +

~~~~~ ~ l% 00V 000(VN Ct' v 00 U )tj;Ib+++++++++++ooo++++++Z 0 --- --X- ---ese

o~ . n tVI~ 0GO_Net w|0

t1H

oX+++ ++±cX++ +++o§0 0 0 0 3

iEEvIiNg; UUUUUeee

Io-0Nenv0 0'r \%-0N vvZ --------- r- C%4

32

cl

H

U, U4

@ o

3 0

+++0o ++0

) +


in this experiment at the end of the third month. It will be noted fromtable 2, however, that there is a significant difference in the degree oftuberculosis present in the treated and in the control animals. A differ-ence is also seen in the prophylactic and therapeutic groups. The contrastin the latter instance may be due to the toxicity of the complex (rise intemperature in non-tubercular animals) in addition to the superimposedpneumonic infection. Microscopic sections made from the organs ofthe sacrificed animals, in both the prophylactic and therapeutic groupswhere the tuberculous process was not complicated by pneumonia,were similar to those observed in the first experiment. The degree offibrosis was not so great as that seen in the first experiment, however.Even though we were forced to terminate this experiment prematurely,we feel justified in concluding that the trend of this experiment par-alleled that of the first. The experiment is now being repeated.

Since there was a considerable degree of protection conferred onthe animals in the above experiment, we were interested in determiningif possible the nature of the mechanism elicited by injections of thecomplex in the animal body. For this purpose two avenues of approachwere considered, namely, the serological and hematological. Serologi-cally, tests for the presence of precipitin-type antibody were made witha variety of sera. The results of the tests appear in table 3.

TABLE 3

No. Sera Results

12 Non-T.B. 8 (+) 4 (-)persons__

62 T.B. patients62_____ before treatment 32 () 30(-T.B. Patients62 after treatment 46 (+) 16 ()

16 Normal rabbits 16 (+ )16 Complex-immune rabbits 16 (+ )

The technique employed in the above tests consisted in overlayingthe serum with an aqueous suspension of the complex. A test was con-sidered positive when after four or five minutes standing at room tem-perature a precipitate (white ring) developed at the interface of theserum and complex. Any precipitate forming after 15 minutes wasnot taken into consideration. The precipitate which formed usually fellto the bottom of the tube after standing at room temperature for 12

33


hours. These tests were so sharply defined that only two symbols (+)and (-) were used to designate positive or negative results. Theconcentration of the complex in the above tests was the same as thatused for injection purposes (1 mg. per cc.). The concentration of theserum was likewise unaltered. In some instances, however, where nega-tive tests were obtained variations in both serum and complex con-centrations were made, but without producing any effect upon thenegativity of the test. The fact that all the sera from normal and fromimmunized rabbits gave positive tests cannot be explained at thistime. We feel that the results of the serological tests are too inconsistentto permit association with any specific humoral (antibody) defensemechanism. Experiments are presently under way, however, which wefeel will yield more conclusive evidence as to the nature of the sero-logical reactivity of this carbohydrate-lipid complex.

The possibility that some one of the cellular elements of the bloodmight reflect the mechanism of protection stimulated by injections ofthe complex was also considered. For this purpose it was decidedto follow the trend of the differential blood picture for two weeks pre-vious to any injection of the complex in both the tuberculous patientsand animals. The counts were taken at 10 A. M. each day during thetwo-week period and were continued for one month after injection ofthe complex was started in both the human and animal groups. Thecounts were made 24 hours after each injection of the complex. It wasfound that the fluctuations in all of the cellular elements after each injec-tion were very similar to those seen in the series of counts taken beforeinjections were started. Wright's stain was used throughout and 100cells were counted in each instance.

Discussion of laboratory results

One can well be led into an unending and inconclusive discussionwhen attempting to isolate any specific factor in the chemotherapy andserology of tuberculosis. The subject of enzyme specificity in tuber-culosis has recently been given support by the work of Gerstl and hisco-workers5 at Yale. These workers feel that the splitting of the tuberclephosphatide by enzymes extracted from the lung, liver, and kidneyrepresents a cellular mechanism vitally concerned with resistance totuberculosis. In the present paper it is seen that no one specific factorwas isolated which could be intimately associated with the resistance ob-served in the treated animals. In view of these observations one can

34


logically look to a possible intracellular mechanism. This mechanism can-be seen in the light of two possibilities (a) an enzyme system,5 and (b)the intracellular antibody phenomenon as first suggested by Man-waring'1' 12 and Chase.3 The latter rendered guinea-pigs hypersensitiveto tuberculin and subcutaneous injection of killed tubercle bacilli.Exudates were then induced in the peritoneal cavities by intraperitonealinjections of liquid petrolatum. After 48 hours the peritoneal cavitieswere washed out with heparinized Tyrode solution containing 10 percent normal guinea-pig serum. The exudate thus obtained consistedmostly of large mononuclear cells. These cells were then immediatelyinjected (intraperitoneally or intravenously) into normal albino guinea-pigs. From 20 to 36 hours following intravenous injection, tuberculinhypersensitivity became established in the skin of the recipient. Thedegree of sensitivity in the recipient was correlated with the volume ofperitoneal cells and the degree of sensitivity of the tuberculous donor.Regardless of which of the two systems is involved in the resistancemechanism in tuberculosis, one or possibly both may function in theeffort of the fixed or the wandering cell to retard the multiplication oftubercle bacilli or greatly to limit the damaging effect of a specificcomponent in the tubercle bacillus to which virulence might be attrib-uted. We know also from the work of Lurie"0 that the activity of thephagocyte from immunized (tuberculous) animals is far greater thanthat seen in the phagocyte of the normal animal. This observation wasmade by .Kolmer.8 Sabin et al.15 16, 17 have repeatedly demonstrated thereaction of cells (fixed and wandering) to injections of lipoid derivativesfrom tubercle bacilli. In no instance during their studies, however, didthese workers ascribe any immunological specificity to any of thederivatives or the cellular elements reacting to injections of the former.Pinner,'4 on the other hand, working with a highly purified phosphatidefraction prepared by Anderson, has found that this fraction reacts invitro with the serum of tuberculous patients and animals in precipitinand complement fixation tests and is also capable of stimulating specificantibody production in animals. The latter work likewise confirms theearlier work of Boquet and Negre,2 who used a less purified preparationof the phosphatide. From our own observations with the complexmentioned here the humoral (antibody) factors as well as the fluctua-tions seen in the cellular elements following injections of the complexin both man and animals could not be correlated with either clinicalimprovement or decline.

35


Due to the absence of any known specific factor and until furtherexperiments now under consideration are completed, we feel justified inbelieving that some of the beneficial results seen in both man and animalsmight possibly be due to the intracellular mechanism mentioned above.

Clinical applicationThe clinical use of the products of the tubercle bacillus began with

Koch and his tuberculin O.T. The killed whole organism, its aqueousextract or emulsion together with the many other derivatives that havebeen used, have all contained the protein of the bacillus as a contami-nating and confusing factor.

While a vast amount of work has been done in the field of proteinsensitization, little has been recorded of the clinical results obtainedwith lipid or polysaccharide fractions, either alone or combined. Theproduction of tubercle and stimulation of the monocyte and fibroblastwith tuberculo-lipid and polysaccharide'5 has been clearly demonstratedexperimentally. The action of the polysaccharide as a hapten makes itof some importance in humoral activity.

With the background of experimental work, there is sufficientjustification for the therapeutic use of these fractions of the tuberclebacilli either alone or in combination. The low toxicity of thesederivatives, as contrasted with tuberculo-protein, permits of considerablefreedom in their use.

Even though bacteriolytic or bacteriostatic action may be negligible,the lipo-polysaccharide suspension may offer some aid to the body con-tending with a chronic destructive process. Furthermore, evidence isslowly accumulating that the intracellular (alkaline and acid phos-phatide) enzyme systems of the lung and liver actively respond to theparenteral use of this combination. The latter work is presently going onin our laboratory.

Our clinical study has been under way for more than two years.Such reports as are made are based upon cases where recovery has takenplace and appears likely to be permanent and upon others wheresteady sustained progress justifies optimism as to the final outcome.

Pulmonary tuberculosisOnly pulmonary tuberculosis has been under treatment, with

secondary infections in the bronchi in one case and with typical tuber-cular laryngitis in two others. All of the cases selected fall into one

36


group; that is, unfavorable or even hopeless. As subdivisions there arethose classed as unstable in the young adult; those with postoperativespread; the old fibroid; and the fibro-cavernous.

It is therefore apparent that the therapeutic test of the lipo-polysaccharide has been a rigid one, with little prospect, even at theoutset, of a high percentage of favorable results.

In 10 of a group of 14 cases, chosen because of their marked in-stability, the results have been favorable. All of this number hadpreviously had the benefit of routine sanatorium treatment. Thepulmonary process had been temporarily arrested only to show renewedactivity under normal living conditions. In half of this number this hadoccurred more than once. At the present time all 10 cases are wellarrested and serial x-ray films denote well-marked fibrosis. Some havebeen working for from six months to one year. In such patients whereroutine treatment has failed to give lasting results, the possibilitiesattendant upon the use of biological products should be thoroughlyexplored.

The four remaining patients, suffering from extensive, rapidly ad-vancing disease, failed to respond to treatment and no check upon theprogress of the infection was noted.

A series of 10 cases in which an active lesion developed or increasedin extent during treatment with pneumothorax or subsequent tothoracoplasty or pneumonectomy, have been given a course of inocula-tions for a period of three to five months.

To appraise the value of the fraction in these patients is difficult,since retrogression occurs from the effects of surgery alone. A dis-tinctively favorable impression has been gained, however, based partlyon the development of better appetite, increased sense of well-being,consistent gain in weight, as well as x-ray evidence of retrogression ofexudative lesions in the contralateral lung.

A third section of our series has consisted of 13 bilateral oradvanced fibro-cavernous cases. Of these, most had long-standing dis-ease and more than one cavity. A preliminary period of sanatoriumtreatment, in all but one case, of from six months to five years gavea control background. No rapid change in the x-ray picture could beexpected, but as the patient gradually built up increased tolerance tothe lipo-polysaccharide, the reaction of the pulmonary tissue was notice-able and progressive. In three cases in which the excavated area waslarger than 6 cm. in diameter, uniform narrowing proceeded for severalmonths but massive hemorrhage abruptly terminated the observations.

37


Hinshaw et al.7 speak of no-changes noted in the heavy-walled cavitywith the use of streptomycin. Some of our most striking results havebeen in this type of disease. Of a total of 13 patients, cavity closure hasoccurred in five; in five more there is every indication that this willshortly occur; the other three died of hemorrhage. Coincident with im-proved x-ray changes, there has been a gradual reduction in sputum anda conversion of positive tests to negative ones. Gastric analysis in thosewith complete cavity closure has been found negative.

Tuberculosis of larynx and bronchiTypical visual tuberculous lesions of the larynx and one of extensive

disease of the bronchi have been treated in patients with pulmonarydisease. Two cases of laryngeal ulceration have healed without localtreatment, and the structures about the chords have resumed a normalappearance. One patient with tuberculous granulations nearly filling thebronchi has healed with cicatrization and the complete relief of dyspneaand wheezing.

Fibroid tuberculosisOf the total of 51 cases that have been inoculated with the lipo-

polysaccharide fraction, 14 fall in this group. Extensive long-standingdisease in one or both lungs has not shown striking changes in serialx-ray films. Clinical symptoms, however, diminshed, sputum decreased,and the process did not advance.

Dosage and toxicity

The patients in this series have been given subcutaneous injectionsat bi-weekly intervals for a period of six months. This has been followedby weekly injections for another six months or longer if the case wasnot stable or the closure of the cavity incomplete. The initial injectionhas been 1/40 cc. of the saline suspension of the complex containing1 mg. per cc., and the amount has been increased as rapidly as possible,controlled by the extent of the local reaction.

A repetition of many of the smaller doses has been necessary inthe more acute cases. Those with more stable lesions tolerated asteadily rising scale, except where an occasional vigorous local effect hasunexpectedly appeared. The maximum amount given up to the presentin a single dose has been 0.5 cc. This has not been exceeded on account ofthe induration produced at the point of injection, but it has been re-

38


peated many times often with no marked inflammatory reaction butwith the constant production of areas of thickening in the subcutaneoustissue.

Local reactions

The local reaction which has followed the subcutaneous injectionhas been used as a guide in the use of the lipo-polysaccharide prepara-tion. In 18 to 24 hours an area of erythema has appeared varying insize from 2 to 10 cm., accompanied with heat and induration. In 48hours the redness has disappeared leaving a firm local injection point. Itsabsorption varied, often lasting a week. Occasionally subsequent in-jections gave rise to an Arthus' phenomenon. As the duration of thetreatment increased, the local reaction has become less marked and, withincreasing doses, often failed to appear even when full amounts wereinjected.

In the first preparation used, the lipid content was sufficiently con-centrated to cause such a marked local reaction that in three patients thearea broke down and yielded a serous secretion for from two days tofour weeks. Healing then took place leaving an indurated scar. Thesereactions led to a modification of the lipid content. In the past year onlyone such injection site softened and discharged.

General reactionsThe absence of tuberculo-protein in the fraction, as proven chemi-

cally and by experimental use, led us to believe that little if any generalreaction would occur. While this was true in the large majority of treatedcases a few have been seen. In four individuals having extensive activelesions, following the first or second inoculation (1/40 or 1/20 cc.)within 24 hours there was a sharp rise in temperature, accompaniedby chilliness, malaise, and slight nausea. At the end of 36 hours all ofthe symptoms disappeared and the patient felt little the worse for theexperience. The local reaction in these cases was very marked at this time.No increase of pulmonary symptoms, such as chest pain, hemoptysis, orincreased sputum, occurred.

On several occasions following 8 to 12 weeks of treatment ageneral reaction has occurred in a few patients. Temperature rose to99.2 to 1020 and lasted from one to two days without any lastingdetrimental effects. The reason for this irregular recurrence of thesesymptoms is not yet clear. When they have appeared and reappeared, acautious advance in dosage of the fraction has been carried out.

39

40 YALE JOURNAL OF BIOLOGY AND MEDICINE

CommentThe natural reaction of the phthisiologist to any new or partly tried

method of treating tuberculosis is that of extreme skepticism. Spectacularresults from bed rest alone are not infrequent, especially in the short-lived exudative lesion. Controlled observations are therefore desirable,but in a disease of so great diversity of form difficult of attainment.

The solution may lie in a selection of only the grave case fortreatment, or in instituting treatment only after the failure of a sub-stantial period of sanatorium care.

In the present series, one or both of these conditions pertain andon this account some optimism is justified at the recorded results. Thepossibilities of the use of combined fractions of the tubercle bacillusin the treatment of phthisis, either with more potent products orpossibly with type-specific carbohydrate are an unexplored field. Theyoffer promise of therapeutic results used alone or in conjunction withsome antibiotic.

SummaryA lipo-polysaccharide fraction of the tubercle bacillus has been

found to have some therapeutic value in the treatment of pulmonarytuberculosis. The clinical results are more suggestive because of theseverity of the type of case treated. The possibilities of this line of treat-ment justify further investigation.

REFERENCES

1 Boissevan, C. H., and C. T. Ryder: Am. Rev. Tuberc., 1931, 24, 751.2 Boquet, A., and L. Nigre: Ann. Inst. Pasteur, 1923, 37, 787.3 Chase, M. W.: Proc. Soc. Exper. Biol. & Med., 1945, 59, 134.4 Doan, C. A.: Proc. Soc. Exper. Biol. & Med., 1929, 26, 672.5 Gerstl, B., R. Tennant, and 0. Pelzman: Yale J. Biol. & Med., 1945, 17, 455.6 Gerstl, B., R. Tennant, and 0. Pelzman: Am. J. Path., 1945, 21, 1007.7 Hinshaw, H. C., W. H. Feldman, and K. H. Pfuetze: Am. Rev. Tuberc., 1946, 54, 91.8 Kolmer, J. F.: Personal communication.9 Kropp, G. V., and J. A. Foley: J. Lab. & Clin. Med., 1944, 29, 231.10 Lurie, M. B.: J. Exper. Med., 1939, 69, 579.11 Manwaring, W. H., and J. J. Bronfenbrenner: J. Exper. Med., 1913, 17, 409.12 Manwaring, W. H., and J. J. Bronfenbrenner: J. Exper. Med., 1913, 18, 601.13 Middlebrook, G.: Am. Rev. Tuberc., 1945, 51, 244-59.14 Pinner, M.: Am. Rev. Tuberc., 1928, 18, 497.15 Sabin, F. R.: Tubercle, 1932, 14, 324.16 Sabin, F. R., C. A. Doan, and C. E. Forkner: Am. Rev. Tuberc., 1930, 21, 290-94.17 Sabin, F. R., K. C. Smithburn, and R. M. Thomas: J. Exper. Med., 1935, 62, 751.18 Smithburn, K. C., and F. R. Sabin: J. Exper. Med., 1932, 56, 867.

The authors wish to thank Dr. H. A. Novack, Dr. C. G. Page, andMiss Domenica Paci for their valuable technical assistance.

JAN, 1945 SEPT. 1945

JAN.1946 APRIL

CASE 1: J. A. D., male, aged 40 years. Rightsided thorocoplasty 1939; cavity developed in theupper left lung in 1940. No change of cavity in nearly 5 )ears of sanatorium treatment. Use of fractionfor I I,' years. Cavity now closed:, sputum reverted from being constantlh positive to negative.

FEB. 1945 SEPT. 1945

JAN. 1946

CASE 2: L. B., female, aged 19 years. Sanatorium treatment for year 1942; relapse in 1944;cavitation in both upper lobes. Use of fraction for 1 1,4 years. Cavities closed during this period and haveremained so. Sputum converted from positive to negative and remainied so for six months, then dis-appeared. Well and working for about one year.

'II IF- -0000LID,5. ":

VP

w.0v,1B:: :'. ..................................................................................................... . ' .: ! .. .:.. ; . ;. .. .; . _ .DECv 1

DEC* 194N 1

CASE 3: S. P., male, aged 59 years. Disease advanced when first detected with bilateral upper lobecavitation. Treated with fraction for two years. Sputum reverted from positive to one consistently negative.Cavities on the right closed; that on the left almost so after two years of inoculations.

NOV. 194(;

,JUNE t 945

JAN. 1945

JUNE 1945 MAR. 1946

CASE 4: C. S., male, aged 24 years. Bad family history. Leftsided disease controlled bypneumothorax. Employed for three years; relapse, with cavity in upper right lung. Fraction givenand cavity closed; has remained so for 1 /2 years. Well and working.

Documents

phosphatides,1 5 lipoids; by hydrolysis, co-workers,6 … any light upon the chemical differentiation between virulent and avirulent strains. Sabinet al. haveshown,however, that the