to the diet of a high mammary cancer strain of mice, Strong '-noted

THE EFFECT OF HEPBISUL (HEPTYL ALDEHYDE-SODIUM BISULFITEADDITION COMPOUND) AND THYROXIN ON WALKER

RAT CARCINOMA 256 *PETER A. HERBUT, M.D., WTTIA H. KRAEMER, M.D., and JOHN JACKSEN

(From the Elizabeth Storck Kraemer Memorial Foundation and Department of Pathology,Jefferson Medical CoUege and Hospital, Philadelphia, Pa.)

During the investigation of the effects of flavoring oils when addedto the diet of a high mammary cancer strain of mice, Strong '-noted thattrue oil of gaultheria not only delayed the appearance of the tumors butalso prolonged the life of the animals after the growths became apparent.Although true oil of gaultheria is more than 95 per cent methyl salicy-late, synthetic oil of wintergreen (methyl salicylate) had very littleeffect on the tumors.2'3 Serial distillation of true oil of gaultheria yieldeda high boiling point fraction that produced no changes in the tumors,and a low boiling point fraction (distilling over below the boiling pointof methyl salicylate) that contained the active principle.4'5 The latterwas more effective than true oil of gaultheria, for when tested on ani-mals it showed an increase in survival time, an increase in softening,liquefaction, and necrosis of the growths, and complete regression of 4of 34 tumors.

Assuming that heptyl aldehyde was an ingredient of the low boilingpoint fraction, it was tested and was found to be effective, but to a lesserdegree than the active principle itself.6'7 Heptyl aldehyde, however, hadtwo unfavorable properties-it deteriorated rapidly and it producedulceration at the site of the injection. The addition of synthetic methylsalicylate protected heptyl aldehyde from oxidation and the combina-tion was actually more effective than was heptyl aldehyde alone, pro-ducing complete regression of the tumors in I2 of 50 animals used.6'8The new mixture, however, did not decrease the local irritative proper-ties upon injection. Subsequently, Strong discovered that the additionof sodium bisulfite to heptyl aldehyde protected the latter from deterio-ration, practically neutralized its harmful action at the site of injection,and had a definite salutary effect upon the tumors.9" 0

Following the work of Strong, several experimenters tested the tumorinhibitory property of heptyl aldehyde with mixed, but generally speak-ing, unfavorable results. Thus Baumann, Kline, and Rusch," in I938,administered heptyl aldehyde in food and subcutaneously to mice with(I) spontaneous mammary adenocarcinoma (strain A mice), (2) pri-

* Received for publication, January 6, I950.Presented at the Forty-seventh Annual Meeting of the American Association of Path-

ologists and Bacteriologists, Madison, April I3, 1950.

59

HERBUT, KRAEMER, AND JACKSEN

mary aural tumor induced by ultraviolet light, (3) primary epithelialtumor induced by painting with benzpyrene, (4) primary sarcoma in-duced by the subcutaneous injection of benzpyrene, and (5) transplantedspindle cell sarcoma originally induced by benzpyrene. They confirmedthe local irritative effect of heptyl aldehyde when used subcutaneouslyand stated that feeding the drug failed not only to prolong the life ofthe tumor-bearing animals but also to alter the character of the tumor.They concluded that heptyl aldehyde was not a universal tumor in-hibitor.

Clark,"2 in I939, administered heptyl aldehyde by stomach tube, infood, drinking water, and by injection directly into the tumors of ratsbearing a subcutaneously transplanted spontaneous sarcoma that orig-inated in the liver of a female rat. The treated animals showed nogreater regression, no greater liquefaction, and no increase in survivaltime over the controls. Willmer and Wallersteiner,"3 in I939, found thatheptyl aldehyde inhibited the growth of chick periosteal fibroblastsgrowing in vitro. Boyland,'4 in I940, reported no inhibition of theCrocker sarcoma i8o by feeding 50 mg. of heptyl aldehyde per day. Hedid, however, obtain I4 per cent inhibition of tumors induced by theinjection of i mg. of methylcholanthrene in stock mice, and in spon-taneous mammary tumors of mice. Garai,'5 in I94I, administered heptylaldehyde-sodium bisulfite intraperitoneally and subcutaneously tomammary-tumor-bearing mice of the C3H, Dba and Paris R3 strains.He noted that the drug exerted a destructive effect on the kidneys andliver which far outweighed the partial retardation of the growth of thespontaneous tumors. He concluded, therefore, that it was not a suitablechemotherapeutic agent for cancer.

In a general review of cancer therapy, Woglom,'6 in I947, stated thatat first heptyl aldehyde seemed to cause the regression of 2 5 per cent ofmammary carcinomas in mice, but that further observations did not con-firm these findings, and that subsequently interest in the compoundappeared to have been lost. In a discussion of Woglom's remarks,Strong stated that his work with heptyl aldehyde had been verified byregressions of tumors in both Switzerland and in England, that the effectof liquefaction lies in a very narrow range of dosage, and that the ani-mals must be put on a strictly quantitative basis. He further stated thatheptyl aldehyde of itself would not do enough but that some otherchemical combined with heptyl aldehyde would continue the process ofcontrol farther than any compound then known to science.Our interest in heptyl aldehyde and heptyl aldehyde-sodium bisulfite

addition compound dates to the work of Strong. We have worked ex-clusively with transplantable tumors. Formerly we used the Crocker

60

HEPBISUL AND THYROXIN ON CARCINOMA 256

mouse sarcoma i8o, rat sarcoma 39, and Walker rat carcinoma 256.More recently we have been using only the Walker rat carcinoma 256because (i) the tumor takes in almost Ioo per cent of cases, (2) it isattended by less than o.6 per cent spontaneous regressions, (3) it killsthe host in almost ioo per cent of cases, (4) it is vigorous in its growthand is not readily influenced by factors that often retard growth of othertransplantable tumors, and (5) the viable portion of the tumor in con-trol animals remains remarkably uniform, thus making alterations in-duced by chemicals easily recognizable. Although we have had somesuccess with the heptyl aldehyde-sodium bisulfite addition compoundalone, its effect upon the tumors was slight, and we therefore begancasting about for a possible catalytic agent. This we found in naturalthyroxin.* By administering a combination of the two compounds toSprague-Dawley rats bearing the Walker carcinoma 256 we have beenable to produce complete regression of the tumors in one-fourth of theanimals treated. These results, we feel, justify the following report.

MATERIALS AND METHODChemicals

The chemicals used in this experiment, except as otherwise indicated,were prepared at the Chemical Unit of the Elizabeth Storck KraemerMemorial Foundation located at the du Pont Experimental Station onthe Brandywine, Delaware. Since the term heptyl aldehyde-sodiumbisulfite addition compound and the alternative term sodium alpha-hydroxy heptane sulfonate are cumbersome, the word "hepbisul" hasbeen coined to designate this agent and will be used henceforth in thispresentation. The empirical formula for this compound is C7H1504SNaand its structural formula is CH3CH2CH2CH2CH2CH2CH(OH) SO3Na.It was prepared by adding freshly distilled heptyl aldehyde (obtainedfrom the Baker Castor Oil Company) to a hot concentrated aqueoussolution of sodium bisulfite, with stirring. After the reaction masscooled, alcohol was added to complete the precipitation. The solid wasfiltered, washed with alcohol, and dried in air. Aqueous solutions of thepowder in desired strengths were put up in ampules, sealed, and thensterilized using I2 lbs. of pressure for 25 minutes. The material wasthus suitable for parenteral use.Two types of thyroxin were used: synthetic and natural. The syn-

thetic compound (Roche-Organon) was obtained already prepared inampules for parenteral administration and was used unaltered. Natural

* The theoretic basis for selecting thyroxin was the possibility of stimulating the neo-plastic cells to more rapid proliferation and thus rendering them more vulnerable to heptylaldehyde-sodium bisulfite.


thyroxin was obtained in the form of crystals from E. R. Squibb andSons. The crystals were dissolved in a small amount of o.i N sodiumhydroxide, diluted with distilled water to the desired concentration, putin ampules, and then sterilized. An attempt was made to combine hep-bisul and thyroxin in a single ampule, but the effort was doomed to fail-ure because the degree of alkalinity required to keep thyroxin in solutionis sufficient to cause destruction of hepbisul. The latter is unstable ina pH much below or above 7.0.

Animals and TumorsThe animals used in this experiment were the Sprague-Dawley strain

of rats. Originally one set of animals was obtained from Memorial Hos-pital, New York City, while 4 years ago another set was obtained fromSprague-Dawley, Inc., Madison, Wisconsin. Since then they have beeninbred (brother-sister mating) in our own laboratory. The diet con-sisted of purina dog chow with a small amount of buckwheat added.Lettuce was given once a week and the animals had access to water atall times.The transplantable tumor-Walker rat mammary carcinoma 256-

was obtained from the Crocker Laboratory, Columbia University,New York City, about I936. The technic of transplantation into theaxilla was carried out under aseptic conditions and was standard exceptthat forceps were used instead of a trochar. Animals of IY2 to 2 monthsof age and weighing about I50 gm. served as hosts and the donor ratcontaining the transplanted tumor was killed on the seventh to the tenthday after transplantation.

ExperimentIn testing the chemicals for tumor inhibitory activity the animals

were used i week after the tumor was transplanted. They were pairedso that the tumors in the control group were of approximately the samesize as those in the treated group. Animals bearing tumors that measuredless than i5 by io by 5 mm. were not used. Subcutaneous injections ofhepbisul and thyroxin were administered simultaneously (but notmixed) in the flank opposite the tumor on the seventh and each subse-quent day after transplantation except Sunday. The animals wereweighed every other day and the dosage was calculated accordingly.The tumors were measured in three dimensions at weekly intervals.Animals bearing tumors that showed no response to the chemicals (andtheir corresponding controls) were killed 2 Idays after transplantation,but those in which the tumor was small and regressing were allowed tolive until the growth completely disappeared or until it was apparent

62

HEPBISUL AND THYROXIN ON CARCINOMA 256

that it would regress no further. At necropsy the major organs, par-ticularly the liver and kidneys, were carefully inspected and, from ani-mals selected at random, liver and kidney were taken for histologicexamination. All tumors were inspected grossly and examined histo-logically. Tissues were fixed in io per cent formalin and stained rou-tinely with hematoxylin and eosin. In selected cases Masson's stain andsudan II were used also.

After preliminary trials the optimum dosage arrived at and thereafterused throughout the experiment, was 30 mg. of hepbisul per ioo gm. ofbody weight and 0.052 mg. of thyroxin per ioo gm. of body weight.The control animals received an equivalent amount (in cc.) of normalsaline solution; In the experiment, io8 animals were given both hep-bisul and natural thyroxin and io8 animals were used as controls. Inaddition, 5o animals received hepbisul alone; 50, natural thyroxin alone;50, hepbisul and synthetic thyroxin; and 50, normal saline solution.Each of the groups of io8 and 5o animals, both treated and controls,was composed of an equal number of males and females.

RESULTSThe results of the experiment are outlined in Table I. Under the head-

ing of "favorable response" are listed those growths that showed vary-ing degrees of degeneration, necrosis, and fibrosis of the neoplastic tis-

TABLE ISummary of Experimental Data Using Hepbisul, Natural Thyroxin, and SyntheticThyroxin, Alone and in Combination, on Rats Bearing the Walker Carcinoma 256

Number of Complete FavorableTreatment animals regression response Total

Hepbisul and natural thyroxin I08 27 12 39 or 36%Saline solution Io8 I O I or o.g%Hepbisul 5o 4 0 4 or 8%Natural thyroxin 50 I O or 2%Hepbisul and synthetic thyroxin 50 2 0 2 or 4%Saline solution 5o o o o

sue but in which viable neoplastic cells were still apparent. Of 'the ani-mals that received hepbisul and natural thyroxin, histologic examinationof the tumor area disclosed no tumor cells in 27 cases (I4 males and 13females) and a favorable response in 12 (8 males and 4 females), mak-ing a total of 39 animals or 36 per cent in which the drug had an effectupon the tumor. Of the corresponding Io8 animals that received normalsaline solution, one showed complete regression and no others showed afavorable response. Of the 50 animals that received hepbisul alone, 4showed complete regression of the tumor and none showed a favorableresponse otherwise, making a total of 4, or 8 per cent, favorable reaction.

63


Of the 50 animals that received natural thyroxin, only one showed com-plete regression of the tumor and no others showed a favorable response,making a total of one, or 2 per cent, that showed a favorable effect. Ofthe 50 animals that received hepbisul and synthetic thyroxin, in 2 therewas complete regression of the tumor and no others gave a favorableresponse, making a total of 2, or 4 per cent, in which the reaction wasfavorable. Of the corresponding 50 animals that received saline solu-tion, none showed complete regression of the tumor and none revealeda favorable response.

PATHOLOGIC FINDINGS

Gross ExaminationIn control animals the tumors grew steadily and progressively until

by the end of 3 weeks the average diameter was 4I.34 mm. During thefirst I 2 weeks after transplantation the skin overlying the tumor wasentirely normal but toward the end of the second week it exhibited "in-flammation" (redness) in 53.27 per cent, scab formation in IO.27 percent, and focal breakdown with discharge of necrotic tumor in 4.67per cent of the animals (Table II).

TABLE IIPathologic Changes in Skin Over Tumor

Two weeks Three weeks

Hepbisul and Hepbisul andControl natural thyroxin Control natural thyroxin

Inflammation 53.27% 43.79% 1093% 4.76%Scab 10.27% I 7.24% 56.25% 50.78%Discharge 4.67% 8.04% 20.31I% 19.04%

By the end of the third week the "inflammatory" reaction decreasedto IO.93 per cent but scab formation soared to 56.25 per cent and focalbreakdown with discharge increased to 20.3I per cent of the animals.Grossly, the tumors were always sharply circumscribed and did not in-vade the adjacent tissues. A definite capsule did not surround the massbut sometimes the neighboring connective tissue disclosed considerableedema. The tumors ordinarily were solid throughout until IO to 12days after transplantation, when they began to show central necrosis.As the breakdown of neoplastic tissue progressed it was sometimes ac-companied by moderate degrees of liquefaction, but in none of the ani-mals did the entire tumor liquefy (Figs. i and 2). The fluid was usuallystraw-colored and clear or opaque, but occasionally it was somewhathemorrhagic. The peripheral portions of even the largest tumors showedgrayish white, moderately firm, somewhat gelatinous, viable neoplastictissue.

64

HEPBISUL AND THYROXIN ON CARCINOMA 2 56

In animals treated with hepbisul and natural thyroxin there was defi-nite retardation of growth in practically all of the tumors during thefirst 3 or 4 days. Following this initial depression the tumors (i) re-mained stationary in size or continued to lag far behind the controlsthroughout the experiment, or at the end of the third or fourth week(after transplantation) they either regressed completely or suddenlybegan to grow vigorously, (2) continued to regress and by the end ofthe week completely disappeared, or (3) began to grow but still laggedbehind the controls (Figs. i and 2).By the end of the second week (after transplantation) the average

diameter of the tumors in control animals was 30.o6 mm., whereas inanimals treated with hepbisul and the natural thyroxin it was 25.20 mm.(Table III). By the end of the third week the average diameters were4I.34 mm. and 34.o6 mm., respectively. These figures do not includethe tumors that completely regressed. When they are incorporated, theaverage diameter of the growths in the treated animals dropped to20.88 mm. at the end of the second week and to 25.55 mm. at the endof the third week. "Inflammation," scab formation, and focal break-down with discharge of necrotic neoplastic tissue started sooner and

TABLE IIIAverage Diameters of Tumors

Treatment One week Two weeks Three weeks

mm. mm. mm.Control I1.50 30.o6 4I-34Hepbisul and natural thyroxin II.82 25.20 34.o6Hepbisul and synthetic thyroxin 10.94 29.93 41.96Hepbisul I1.44 28.42 39.26Natural thyroxin II.37 30°I5 43.08

was of greater severity in treated than in control animals. By the endof the second week "inflammation" was already on the wane in treatedanimals (43.79 per cent), while in control animals it was reaching itspeak (53.27 per cent) (Table II). Scab formation, however, in treatedanimals exceeded that in controls by over one-third, and discharge intreated animals exceeded that in controls by approximately one-half.By the end of the third week, control animals with an "inflammatory"response still outnumbered treated animals with a similar response, butscab formation and discharge were essentially the same in both groups.The chain of events in animals treated with hepbisul and synthetic

thyroxin, hepbisul alone, and natural thyroxin alone, was similar tothat in control animals (Table III).

Microscopic ExaminationIn order to understand better the histologic changes induced by hep-


bisul and thyroxin, it seems desirable first to outline briefly the normalmicroscopic appearance of our particular sample of the Walker rat car-cinoma 256 and to point out the changes which may occur as the resultof post-mortem autolysis.

In control animals, the peripheral portions of the tumor in 2I daytransplants and the entire tumor in 7 to I2 day transplants were com-posed of almost solid viable neoplastic tissue. The connective tissuestroma was so scanty that for practical purposes it was non-existent.Capillaries were not numerous and those present were inconspicuous.In the majority of tumors the cells were moderate to large in size. Asa rule, the borders were indistinct and the cytoplasm of one cell mergedwith that of another to give diffuse syncytial sheets (Fig. 3). Less fre-quently the borders were more distinct and the cells then appeared poly-hedral. As the outlines became sharper and the cells enlarged, theircontours became distinctly rounded. In the smaller cells, the cytoplasmwas moderate in amount, homogeneous, finely granular, and lightlyeosinophilic (Fig. 4). In the larger cells it became more abundant andlighter staining, until definite small and later coalescing vacuoles ap-peared (Fig. 5). Some of these took the sudan stain and therefore rep-resented accumulations of fat, while others remained -clear and presum-ably represented hydropic degeneration. The nuclei were sharplydefined, round, oval, or slightly irregular and light staining. The chro-matin was coarse and clumped, and nucleoli were quite prominent.Mitotic figures were both regular and irregular and their number variedfrom 6 to 25 per high-power field. Occasionally, particularly in trans-plants up to 2 weeks of age, some of the tumor presented a glandulararrangement (Fig. 3). The glands were round or elongated, and oneor several cell layers thick. The cell borders usually were indistinctand the cytoplasm formed syncytial masses. Occasionally the cells weremore sharply delineated and the outlines were cuboidal. The cytoplasmwas lightly eosinophilic and granular to reticulated but contained novacuoles. The nuclei were similar to those encountered in cells present-ing a diffuse non-glandular arrangement. Rarely the cells were long,spindle-shaped, and sarcoma-like (Fig. 6). They were fairly sharplydemarcated and the cytoplasm was moderate in amount and denselyeosinophilic. The nuclei were round, oval, elongated or irregular, andpossessed a sharp delineating membrane and dense granular chromatin.At other times the nucleoplasm was more solid and distinctly hyper-chromatic. Mitotic figures were less frequent than they were in themore rounded cells.

In control animals, killed 3 weeks after transplantation, the centralportion of the tumor usually showed complete and diffuse, but some-

66


times partial and focal, areas of necrosis. Complete necrosis was mani-fested by an absolute disappearance of the normal architectural pat-tern and a replacement with an abundant amount of amorphous, pink-staining material. When the disintegration of cells was not quitecomplete, necrosis was represented by faint outlines of former neoplasticcells, nuclear fragments, and, occasionally, shadows of nuclei. In eithercase the vessels were inconspicuous. In the areas of more completenecrosis, calcification, in the form of bluish staining granular materialor small spherical bodies, sometimes was present in the focal areas, butfibrosis was never seen. In tumors showing focal necrosis the dead tis-sue appeared similar. Interspersed between these foci, however, therewere collars of viable or degenerating neoplastic cells surrounding en-gorged capillaries. The viable cells were similar to those seen in themore solid areas at the periphery, whereas the degenerating cells usuallywere distinct, rounded, separated from their neighbors, and smallerthan is normal (Fig. 7). The cytoplasm was densely eosinophilic andmoderate in amount. The nuclei were small, round, oval or irregular,and showed varying degrees of pyknosis, karyorrhexis, and karyolysis.Progression of these changes led to complete disintegration of the cells.

Sections of tumors from control animals that had died showed vary-ing degrees of post-mortem autolysis. Frequently the capillaries werelarge, prominent, and engorged. Old foci of complete necrosis weresimilar to those already described. The tumor cells that were viableat the time of death were present in sheets at the periphery or wereaggregated around vessels and disclosed varying, but in a given tumoruniform, degrees of change. They usually were shrunken, small, irreg-ular, and might contain elongated cytoplasmic processes. The cyto-plasm was moderate in amount, still somewhat granular, and eitherlightly eosinophilic or basophilic. The nuclei showed beginning con-densation. They were round, oval, or irregular, deeply stained, and con-tained large blue granules. Mitotic figures were absent. With progres-sion of these changes the cells became rounded off and smaller. Thecytoplasm was moderate in amount and densely eosinophilic, and thenuclei became darker, smaller, and more pyknotic. Further progressiondisclosed liquefaction or coagulation necrosis. The former was mani-fest by disintegration and disappearance of some of the cells, greatershrinkage and pyknosis of others, and partial disintegration with fray-ing of the peripheral portions of the cytoplasm and karyorrhexis of thenuclei in still others. In coagulation necrosis the cell outlines were in-distinctly discernible; the nuclei were completely, or almost completely,faded, and the chromatin granules from disintegrated nuclei were irreg-ularly and diffusely scattered. Beyond this there was complete disin-

67


tegration of the cells, leaving a mass of amorphous, pink-stainingdebris speckled with fine, powdery, nuclear remains.The pattern of changes observed in animals treated with hepbisul

and thyroxin was distinctly different from that observed in saline-treated animals or from animals that had died and showed post-mortemautolysis. In order to obtain an understanding of the pathogenesis,some of the animals used in preliminary tests were killed at varyingintervals from the 8th to the 2 ist day after transplantation. The pic-ture on and beyond the 2 ist day was obtained from animals used in theexperiment proper.

After two injections of hepbisul and thyroxin there were no foci ofnecrosis but the entire tumor nevertheless was affected. With low-powermicroscopic examination the uniformity seen in tumors from controlanimals was lacking and was replaced with irregular light and darkstaining patches, streaks, and foci. The capillaries were inconspicuous.With higher magnification the most striking feature was a disruptionof the cells. This occurred in ill defined focal areas that merged withadjacent neoplastic tissue. In less affected areas the cells showed aslight degree of over-all distortion (Fig. 8). They were, however, irreg-ular, ill defined, smaller than usual, and contained scanty, homogeneous,or somewhat vacuolated cytoplasm. The nuclei were less prominent,as a rule more lightly stained, but some showed also varying degrees ofpyknosis. In more severely affected areas the cells were fragmented,separated, and of irregular shapes and sizes (Figs. 9 and io). Theborders were broken and the cytoplasm was scanty and peripherallyfrayed. The nuclei were extremely irregular and pyknotic.

Following the initial disruption of the over-all pattern and cell struc-ture, the destruction proceeded along one of two lines. On the one hand,necrosis, both liquefactive and coagulative, progressed unabated. Thecells showed further distortion, disintegration, and liquefaction. Con-comitantly, the products of disintegration were absorbed into the bloodstream and the tumor gradually decreased in size and was ultimatelycompletely absorbed. In such instances there was either no surround-ing connective tissue reaction and the skin and subcutaneous tissuewere returned to normal, or the resultant fibrosis might be minimalleaving only a small scar in the dermis (Fig. ii). On the other hand,the degeneration and disintegration of the tumor cells were accompa-nied by a rather marked fibroblastic proliferation. The fibroblasts orig-inated throughout the tumor area from the inconspicuous supportingstroma and did not invade from the adjacent capsular region. Morefrequently the proliferating connective tissue formed branching andinterlacing bands that separated and surrounded foci of tumor cells

68


(Figs. I2, I3, and I4). At first, the strands were composed of long, illdefined, spindle-shaped fibroblasts with a relatively large amount oflight staining, somewhat granular, eosinophilic cytoplasm and oval,elongated or spindle-shaped, evenly and lightly stained, nuclei. Scat-tered throughout these areas there were varying numbers of neutrophils,eosinophils, and lymphocytes. Capillaries were inconspicuous. Laterthe leukocytic cells tended to disappear; the capillaries became evenless prominent, and the fibroblasts became transformed into fibrocytes.The foci of tumor cells that seemed to be entrapped in the meshes ofthe proliferating tissue continued to exhibit progressive degrees of ne-crosis, liquefaction, and disintegration. Ultimately they showed com-plete disintegration, and were either resorbed and replaced by fibroustissue, or they remained as masses of amorphous debris in which cal-cium salts were precipitated (Fig. I5). The latter existed as minute,fine granules or as larger, irregular clumps of intensely bluish stainingmaterial. About the periphery of the necrotic and calcified foci, giantcells of foreign body type sometimes became quite numerous. Through-out the connective tissue that had replaced most of the original tumorthere also were frequently present numerous large, polyhedral orrounded, sharply defined foam cells containing fat droplets in their re-ticulated cytoplasm, and exhibiting small, round, somewhat eccentric,evenly stained, uniform appearing nuclei (Fig. I6). In addition, fattymaterial was sometimes crystallized in the form of long, spindle-shapedneedles which, in sections stained with hematoxylin and eosin, were rep-resented as empty spaces (Fig. I7). Leukocytic infiltration throughoutthe connective tissue was always minimal or absent.

Less frequently, fibrous tissue appeared to spring up diffusely, uni-formly, and heavily throughout the tumor rather than in scatteredbranching strands (Fig. i8). In such cases the tumor cells were sepa-rated and existed singly or in clumps of two or three. They showedvarying degrees of degeneration followed by necrosis and then completedisappearance. Upon their absorption the connective tissue simplycondensed and ultimately existed as a solid mass of fibrous tissue.

DIsCUSSIONThe results of this experiment clearly indicate that hepbisul combined

with natural thyroxin is an effective inhibitor of the Walker rat carci-noma 256. Strong's statement in I947,16 that heptyl aldehyde of itselfwas not enough but that other chemicals in combination with heptylaldehyde would produce a salutory effect upon the tumor, is thus borneout.The histologic appearance of the tumor we used is similar to that

69


used by Earle,17 who also obtained a sample from Columbia University.While the regressive changes that usually occur in the central portionof the tumor consist of degeneration, necrosis, liquefaction and absorp-tion, and appear similar when due to post-mortem autolysis and wheninduced by hepbisul and thyroxin, it is a combination of factors whichallows a distinct differentiation among the three processes. Spontane-ous ante-mortem changes, for example, occur centrally. They are focalwhen the tumor is young but diffuse when it is older. In none of ouranimals, however (and we have examined several thousand already),are these spontaneous changes associated with fibroblastic prolifera-tion. Also in the tumors that show spontaneous retrogressive changesthe neoplastic tissue at the periphery is always vigorously viable. As arule it consists of solid sheets of neoplastic cells with an extremelyscanty or imperceptible stroma. Foci of necrosis, when they do occur,are likewise unaccompanied by fibrous tissue reaction. Occasionally,there is an increase of fibrous tissue about the circumference of thetumor, but this does not invade the neoplastic tissue and the two arereadily separable. The autolytic changes, too, present a definite patternand are easily distinguishable from the changes induced by hepbisuland thyroxin. The degeneration, necrosis, and liquefaction are similar,but they affect the entire mass, peripherally as well as centrally, andare not present in focal areas only. There are, therefore, no viable andvigorous tumor cells present and, of course, there is no fibrosis. Thechanges induced by hepbisul and natural thyroxin early in the process,to be sure, also consist of degeneration, necrosis, liquefaction, and ab-sorption of neoplastic cells throughout the tumor, but, in addition,there is usually a diffuse increase of connective tissue that doubtlesslysprings from the existing stroma. In our experience the odd tumorsthat regress spontaneously almost always show a complete resorptionof the neoplastic tissue and complete restoration of the adjacent tissuesto normal. While this occurs also in animals that have been treated withhepbisul and thyroxin, more often the tumor cells, instead of beingcompletely liquified and resorbed, become calcified and the calcific fociare surrounded by foreign body giant cells and fibroblasts. In someonly the fibrous tissue remains.Although it is well known that an inadequate or starvation diet will

result in retrogressive changes in tumors, we believe that inadequacyof food intake played no part in our experiment. Both the controlandthe treated animals were kept under similar experimental conditions,were fed the same diet, consumed approximately similar portions, andmaintained essentially parallel weights (Text-Figs. i and 2). Animalsin both groups upheld their weights until approximately the 2 ist day, at

70


which time, if the tumors were large and showed considerable centralnecrosis and liquefaction, they began to lose weight. In other words theloss in weight was terminal, was parallel in both control and treatedanimals, and probably resulted from absorption of toxic material fromthe tumor rather than from a decrease in consumption of food. In addi-tion, none of the large tumors in either group revealed more than cen-

160

~o

7 9 *11 15 17 19Dars after tunor trsuantio

TSrcGJ uitiL pSivsul MAn 1*in (YIAW&tv).~~~~~ " ~nomal salim so. (control).

Text-Figure i. Curves presenting the average weights of ioo tumor-bearing rats ofeach of the two groups indicated.

tral regressive changes histologically. Fibrosis was not present. As faras we could determine, infection likewise played no part in either thecontrol or the treated animals.At present we have no facts regarding the mechanism of action.

Since the retrogressive changes in the tumor cells are far in advance of

r7o

160

-15

: _'~~~~~~~~~~~~~~~~~~~~~~~~O d"""',_ .*., ..~~~~~~~~~~~~~~~~~~~~~-

w~~7 9 11 13 15 17 19Dakler tumo r*~z Witlon

-- - kt.a r v)thTh, .........(u.a&...t)...58b sol. (Cols).

Text-Figure 2. Curves presenting the average weights of 50 tumor-bearing rats ofeach of the four groups indicated.

71

-ft .ft- "W .

am

..- M. - M .. .

.0


the connective tissue proliferation, there seems to be no doubt that thedrugs act upon and destroy the tumor cells directly rather than by aroundabout process of fibrosis. We have no explanation for the factthat natural thyroxin in combination with hepbisul is effective whilesynthetic thyroxin is not. With the doses employed, neither productresulted in a clinically visible metabolic change in any of the animals.Because there is this distinct difference in the response induced by thenatural and synthetic products, the first assumption is that thyroxinacts not by increasing the metabolism and multiplication of the tumorcells but by some other synergistic means. From a chemical point ofview, however, there is another possible explanation which we have notas yet pursued. Natural thyroxin is levorotatory, while synthetic thy-roxin consists of equal parts of both the levorotatory and dextrorotatoryfractions. Since the levorotatory fraction alone is responsible for theincrease in metabolism in the body, it would appear that to obtain thedesired effect the dose of synthetic thyroxin should be twice that of thenatural product. The amount of natural and synthetic thyroxin usedin our experiment, however, was the same.

Statements in the literature regarding the toxicity of heptyl aldehydesodium bisulfite are not in agreement. Strong 16 was of the opinion thatthe preparation was relatively innocuous, whereas Garai 15 stated thatthe compound was not a suitable chemotherapeutic agent for cancerbecause its destructive effects on the kidneys and liver far outweighedany beneficial action. Our experience supports the observations madeby Strong. When the amount of hepbisul used is small it may be injectedsubcutaneously, intraperitoneally, or intravenously with impunity.When the amount used is relatively large (as in our experiment), re-peated subcutaneous injections in the same area will produce local irri-tation and moderate necrosis in approximately one-half the animals.Histologic examination of portions of the liver and kidneys from bothtumor-bearing and normal animals, given similar amounts of hepbisul,revealed no untoward histologic changes. Some of the kidneys showedslight tubular degeneration, but at no time was necrosis or other damageencountered. Furthermore, Gruber has just completed full toxicitystudies on hepbisul (to be published separately) and states that thecompound is quite innocuous. He reports that the LD 50 in rabbits is0.45 gm. intravenously per kg. of body weight and in rats I.3 gm. intra-peritoneally per kg. of body weight. Histologic examination of the liversand kidneys of some of the animals used by Dr. Gruber, that receivedlarge amounts of hepbisul, likewise revealed no pathologic changes.

72

HEPBISUL AND THYROXIN ON CARCINOMA 2 56 73

SUMMARY

Hepbisul and natural thyroxin were administered subcutaneously toSprague-Dawley rats bearing the Walker rat carcinoma 256. Of theio8 animals treated, 27 showed complete regression of the tumors andI2 others showed a favorable histologic response. A salutary effect,therefore, occurred in 39 animals or 36 per cent of the total.The retrogressive changes consisted of degeneration, necrosis, lique-

faction, and fibrosis. The chemicals apparently act directly upon thetumor cells and the fibrosis is compensatory and secondary.

Hepbisul and synthetic thyroxin resulted in a favorable response in2 of 5o animals treated or a total of 4 per cent. The cause of this dis-crepancy is unknown.

Despite the relatively high doses used, local irritation and ulcerationat the point of injection occurred in less than 50 per cent of the animals,while general toxic effects, as manifested by morphologic changes in theliver and kidneys, were not encountered.

It is concluded that hepbisul and natural thyroxin when administeredsubcutaneously are effective inhibitory agents for the Walker rat carci-noma 256 in rats of the Sprague-Dawley strain.

REFERENCESi. Strong, L. C. Possible effect of oil of gaultheria in diet of mice susceptible to

spontaneous carcinoma of the breast. I. A suggestion. Proc. Soc. Exper. Biol.& Med., I932-33, 30, 386-390-

2. Strong, L. C. Possible effect of oil of gaultheria in diet of mice susceptible tospontaneous carcinoma of the mammary gland. V. Growth rate and certainretrogressive changes of tumors after onset of malignancy. Am. J. Cancer,I936, 28, 550-558.

3. Strong, L. C. The effect of oil of wintergreen on the incidence of spontaneouscarcinoma in mice. IV. Effect on growth rate and survival time after onsetof malignancy. Am. J. M. Sc., I936, 192, 546-553.

4. Strong, L. C. The liquefaction of spontaneous tumors of the mammary glandin mice by heptyl aldehyde. Science, I938, 87, I44-I45.

5. Strong, L. C. Effect of oil of wintergreen on spontaneous tumors of the mam-mary gland in mice. VI. The different effect of two fractions obtained by thedistillation of the true oil. Am. J. Cancer, I938, 32, 227-239.

6. Strong, L. C. The synergistic effect of heptyl aldehyde and methyl salicylateon spontaneous tumors of the mammary gland in mice. Yale J. Biol. & Med.,I938-39, II, 207-2I8.

7. Strong, L. C., and Whitney, L. F. The treatment of spontaneous tumors in dogsby the injection of heptyl aldehyde. Science, I938, 88, III-II2.

8. Strong, L. C. Effect of oil of wintergreen on spontaneous tumors of the mam-mary gland in mice. VII. The liquefaction of spontaneous tumors of the mam-mary gland by heptyl aldehyde. Am. J. Cancer, 1939, 35, 40I-407.

74 HERBUT, KRAEMER, AND JACKSEN

9. Strong, L. C. Further approach toward control of spontaneous cancer of mam-mary gland in mice by heptyl aldehyde-sodium bisulphite. Proc. Soc. Exper.Biol. & Med., I940, 43, 634-637.

io. Strong, L. C. The effect of heptyl aldehyde-sodium bisulphite on spontaneoustumors of the mammary gland in mice. Cancer Research, I94I, I, 473-476.

ii. Baumann, C. A., Kline, B. E., and Rusch, H. P. Heptaldehyde as a tumor in-hibitor. Proc. Soc. Exper. Biol. & Med., I938, 39, 354-356.

12. Clark, W. G. Effect of heptyl aldehyde on a spindle-cell sarcoma. Proc.Soc. Exper. Biol. & Med., I939, 40, 562-564.

13. Wilhmer, E. N., and Wallersteiner, K. The effects of certain aldehydes uponthe growth of cells in vitro. J. Physiol., I939, 96, i6P-i7P.

I4. Boyland, E. Experiments on the chemotherapy of cancer. 4. Further experi-ments with aldehydes and their derivatives. Biochem. J., 1940, 34, II96-I20I.

i5. Garai, F. Heptyl aldehyde-sodium bisulfite: toxicity and effect on spontaneousmammary carcinoma in mice. Cancer Research, 1941, I, I44-I45.

i6. Woglom, W. H. General Review of Cancer Therapy. In: Moulton, F. R. (ed.)Approaches to Tumor Chemotherapy. American Association for the Advance-ment of Science, Washington, D.C., I947, PP. I-I2.

I7. Earle, W. R. A study of the Walker rat mammary carcinoma 256, in vivo andin vitro. Am. J. Cancer, I935, 24, 566-6I2.

DESCRIPTION OF PLATES

PLATE I2

FIG. I. Gross appearance of three hemisected tumors in situ. The growth in thecontrol animal (C) is large and shows central necrosis and liquefaction; thatin the animal treated with hepbisul alone (H) is small and shows central ne-crosis but no liquefaction, while the tumor in the animal treated with hepbisuland natural thyroxin (HT) has completely disappeared.

FIG. 2. Cross sections of two tumors from control (C) animals both male (M) andfemale (F) showing central necrosis without liquefaction. The remainingspecimens (T) are from treated male (M) and female (F) animals in whichthe tumors had entirely disappeared or were replaced by yellowish foci ofcomplete necrosis surrounded by fibrous tissue.

AMERICAN JOURNAL OF PATIIOLOGY. VOL. XXVII

........................ux...............T,

1

2

Herbut, Kraemer, and Jacksen Hepbisul and Thyroxin on Carcinoma 256

75

PLATE I2

.I- I J,,,

..; .! .j.:.:,kM- ! >w1)

PLATE I3

FiG. 3. Tumor from control animal showing diffuse sheets of large, ill defined cellsand one well formed gland. X 380.

FIG. 4 Tumor from control animal showing more d stinct. round.d or polyhedralcells. some of which contain light-staining cytoplasm. Mitotic figures are nu-merous. X 380.

FiG. 5. Tumor from control animal showing large, round. often vacuolated neo-plastic cells. X 380.

FiG. 6. Tumor from control animal exhibiting sarcoma-like cells. X 380.

76

PLATE 13AMERICAN JOURNAL OF PATHOLOGY. VOL. XXVII

._

i5F..t~~~~~,0 .

_*~_. i* . _.

~~ ~ ~ ~ ~ ~ ~

Iqa~~~~~~~i~~~u;N~~~:01

5 6Herbut, Kraemer, and Jacksen Hepbisul and Thyroxin on Carcinoma 256

77

PLATE I4

Fic. 7. Tumor from control animal illustrating spontaneous degeneration andliquefaction. X 375.

FIG. 8. Tumor from animal that received one injection of hepbisul and naturalthyroxin, showing an over-all disruption of architecture and beginning dis-integration of the neoplastic cells. X 3j5.

FIG. 9. Tumor from treated animal illustrating progression of the degenerativechanges seen in Figure 8. X 375.

FIG. io. Tumor showing further progression of the changes illustrated in Figures8 and 9. Pvknosis. karyorrhexis. and liquefaction are apparent. X 375.

78

PLATE I4AMERICAN JOURNAL OF PATHOLOGY. VOL. XXVII

7 a

9 10Herbut, Kraemer, and Jacksen Hepbisul and Thyroxin on Carcinoma 256

79

",'"7Ii

V.a'.

la

PLATE I5

FIG. ii. Tumor site from treated animal showing only a scar in the dermis. Thetumor has completely disappeared. X 95.

FIGS. I2, I3, and I4. Tumors from treated animals showing varying degrees offibrosis and of necrosis and liquefaction of neoplastic cells. X 375.

80

It

Pk*_ S

i.;

A

.

_

co:I-

8i

N*I-

*3-

cJ

0

F.0

H

G-

i- .;: A

U-U-

GACG

%O

,i ll.:",.T

rw'.; '"",

Wl%,:-..--,A,"`-4

.. :..Ii ..

j 41:

w 4

-*.A.MAW

(L)

m

PLATE i 6

FiG. I5. Tumor site from treated animal showing fibrous tissue, foam cells, andgiant cells surrounding calcific foci. X I85.

FIG. i6. Fatty material deposited at the site of a previous tumor in a treated ani-mal. Sudan II stain. X 95.

FI. I 7. Tumor site from treated animal showing only foam cells and crystals.X I85.

FIC. i8. Tumor from treated animal illustrating a diffuse increase of fibrous tissueenmeshing single neoplastic cells in various stages of necrosis. X 375.

82

PLATE i6AMERICAN JOURNAL OF PATHOLOGY. VOL. XXVII

15 16

17 18

Herbut, Kraemer, and Jacksen Hepbisul and Thyroxin on Carcinoma 256

83

Documents

to the diet of a high mammary cancer strain of mice, Strong '-noted