MIETFASTASIZING INTRACRANIAL TUMORSKENNETH H. ABBOTT, M.D.

FELLOW IN NEUROSURGERYMAYO FOUNDATION

AND

J. GRAFTON LOVE, M.D.SECTION ON NEUROLOGY AND NEUROSUJRGERY, MAYO CLINIC

ROCHESTER, MNI-N.

I'HE OCCURRE.NCE. OF METASTASIS fromii anly primiiary intracranial neoplasmlias been the subject of great interest to both clinicians and pathologists foriany decades. Particularly has this been true when metastasis seemed tolhave occurred outside the central nervous system and its envelopes. To besure, there have been a number of reports of such metastasis, but consideringthe mnany thousands of intracranial tumors that have been studied in manylarge clinics, these few reports speak well for the rarity of such a metastasizinglesion. In view of this, we wish to report one case in which an intracranialtulmor appears to have metastasized to the lungs and hilus nodes.

Case Report.-A male, age 32, registered at the Mayo Clinic, May 20, I938,because of headaches and vomiting. In January, I936, he first had noticed a lump inhis forehead, very slightly to the left of the midline. He had persuaded his familyphysician to lance this, but only blood had been obtained. During the next two monthsthe swelling had pulsated and about this time he had found that he could pushthe lump back into his head, although this had been somewhat painful. In April,1936, he had been examined elsewhere, and roentgenotherapy had been employed. Thishad given relief from headaches for about two weeks, they then gradually had returned.In August, 1936, he had received another course of deep roentgenotherapy, withsimilar re,sults; however, he had been able to get along fairly well until September,1937, when the headaches had become so severe that he had requested, and hadreceived, another course of treatmnent. In February and March, 1938, his headachesagain had become severe, and at this time they had been associated with frequentvomiting.

When he was first examined at the clinic, neurologic and general physicalexaminations disclosed the presence of a small mass in the forehead, with obviouserosion of the frontal bone. The optic disks were choked four diopters, about equallyin both eyes. The roentgenograms of the skull confirmed the clinical impressionof erosion of the frontal bone and, in addition, there was evidence of prolongedincreased intracranial pressure.

On May 28, 1938, one of us (J. G. L.) removed a large right parasagittal frontaltumor, which had completely surrounded and invaded the longitudinal sinus. It wasnecessary to resect the sinus to the entrance of the rolandic veins and to remove apart of the invaded and eroded frontal bone. The tumor weighed I40 Gm., andmeasured 8 x 7 x 6 cm (Fig. i). The patient made an uneventful recovery, andwhen he was dismissed, July i8, 1938, the wound had healed.

He returned to the clinic in November, 1938, because of the presence of a peduncu-lated growth, two centimeters in diameter, in the anterior end of the operative incision.Radium was applied to this at that time and again in February, I939, because ofrecurrence. In spite of this, the patient was able to work from August, I938, toJune, I939, when headaches again began to be severe. In July, he had a generalized

343

ABBOTTAND LON-E ~~~~~~Annials of SurgeryABBOTT AND LOV-E September, 1 9 4 3

ConIvilsioni, lollowed fiv,e days later by mnuil)ness assuciate(l with paralysis in the lefthalf of his face. He again returned to the clinic, August i, 1939, because of thepersistence of the facial weakness and of dysarthria. On August 3, 1939, with theuse of local anesthesia, a tumor the size of a hen's egg (4o Gm.) was removed fromthe original site of the neoplasm. He did only fairly well after this procedure and,September I, I939, a plastic closure of a cerebrospinal fistula, which had developedsoon after the seconid operation, was performed. He had a very stormy course afterthis operationi, and (lied, in hyperthermia, September i l, 1939, approximately threevears and( nine miotitlhs after he first had nioticed the lumllp in his forelhead. A conmpleteniecropsy (lisclosedl the defect of the receilt remioval of tlle fronital tumlior alnd, in]addition, another large residual andl recurrent ttumor ini the right fronitotemporallobes (Fig. 2). There also was evidence of terminal ependvmitis anld imild mellingitis.

I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .FIG. i.-Ttumors removed at the two operations.

Gross examiniationi disclosed nloduiles of fir-m grav tissue scattere(d througlhoutall lobes of both lungs (Figs. 3 and 4). The hilus no(les also appeare(d to be ilnfiltratedby smiiall tumor masses. These niodules varied in size up to four millimleters in diameter.T'liere was also pulmonary inifarction fromii arterial emiboli.

Histologic examiniationi of the braini ttumior (lisclosed a malignialnt hemalngioblastotmla(Grade 3), anid all three specimens (two opeerative anid onie postmortem) were foundto be idenitical (Figs. j, 6 ancl ). Sectionis of the miliary nodules in the lungs anilhilus nodes disclosed a tumior of identical structure (Fig. 8) to that founld in the braini.There was nio doubt as to the identitv of the tumor.

Sectionis of the removed tumlors alnd of the resi(lual tumors were examinie(l. Thefollowing stains were used: Hiematoxylin anld eosini, Mallory's phosphotunigstic acidstaiin, Perdrau',s silver staini, Cajal's anid Hortega's stainis, and cresyl violet. Thehematoxylin and eosini stain (lisclose(l a tuimor malde up of small blood vessels, capillaries.sheets, strands and islands of endothelial-like cells. Thin-walled vascular chanlnels linedwith endothelial cells frequiently were founid filled with fat, oval an doccasionally some-what spindle-shaped, hyperclhromiiatic and foalm-i-like endothelial cells. The nuclei wereusuially smnall anid the cytoplasm was misty and( even foam-like. 1Mitoses were every-where abunidant anid were tvpically presenlt in the intimiia of the .small proliferatinigblood vessels (pseudocapillar i, ). The larger vessels occasiolnally conitainied erythrocytes,but many (lid not.

344

Volumue 115'Nunuber 3 INTRACRANIAL TUMORS

The tiny tumor nodules found in the lungs, pleura and hilus nodes were discrete,circumscribed masses of cells, with exactly the same characteristics as seen in theintracranial tumor. Examination of sections of the metastatic tumors that had beenstained with Perdrau's silver stain revealed the typical vascular reticulin pattern ofa hemangioblastoma (Fig. 9) just as it did in the case of sections of the intracranialtuimor obtained at operation and at necropsy. Mlallory's phosphotungstic acid stain,

FIG. 3FIC. 2.-Recurrent hemanigioblastonia in coronal section at the level of the optic chiasm.

FIG. 3.-AIetastatic ttumors in the lung (arrows) from primary intracranialhemangioblastoma.

345

ABBOTT ANiD LOVE Annals of SurgeryABBOTTAND LOVE ~~~~~~~~September, 1943

Cajal's stain and Hortega's stain were only of minor additional value in proving thatthe tumor was not of glial origin but definitely of vascular origin.

COMMENT

The problem of metastasis encountered any place in the body is notalways easily solved, mainly because it is not always possible to be surewhich neoplasmn is the primary one. This is particularly the situation whenone is dealing with metastasis of intracranial tumors. As was, facetiously butcorrectly, remnarked by one pathologist: "The tumor cells do not oblige uswith arrows on them showing which way they are going"-certainly onewishes they did.

FIG. 4.-Metastatic tumors in the lungs. (x6)

We do not wish to discuss in detail the natural barriers which seem toexist against the process of metastasis of intracranial tumors as these arebeing extensively investigated by other investigators at the clinic; however,it does seem pertinent to point out the possible r6le which surgical inter-vention may play in the production of such metastasis. Undoubtedly, in thecourse of the surgical removal of a neoplasm, venous channels are opened andthe intraluminal negative pressure favors the aspiration into their luminaof viable tumor cells. Thus, the tumor cells gain access to the blood streamand are carried to the lungs where they may be "filtered out" in the capillarybed, or the tumor cells, theoretically, could pass on into the arterial bloodfor dissemination throughout the body. This seems to have been thesituation in the metastasizing tumor we have here detailed, although, to besure, we have not excluded the lymphatics of the scalp as the portal of entryand route of metastasis nor does it seem possible to do so in this instance.Just why in most instances of operative removal of a neoplasm these sup-

346

Volume 118Number 3 INTRACRANIAL TUMORS

posedly "viable" tumor cells do not form metastatic lesions while in rareinstances they lodge and grow to reproduce the identical tissue of the primaryneoplasm is not known. However, at some future day, this may be explainedby certain cytochemical, cyto-electric and physical phenomena which as yet

A

BFIG. 5.-Sections of hemangioblastoma removed at first operation;

A, stained wVith hematoxylin and eosin (X285); B, stained with Perdrau'ssilver stain for reticulin. (XI50)

remain obscure to investigations. Certainly, nothing is to be gained at thistime by pure speculation.

In reviewing the cases reported as cases of metastasizing intracranialneoplasms, we have collected about sixteen. A study of these reports hasemphasized in our minds the difficulties one encounters in being certain thata given tumor is primary in the brain or elsewhere. Consequently, we have

347

A11B)OTTAND LOVE ~~~~~Annial rf Surge~ryABBOTI' AND\5iL) ,OV L September, 194 ;

divided these cases illto three groups: (i) Those cases we do iiot believehave been showni to be cases of primary tumor of the brain, or we cannotbe sure the metastatic tumnors, if suclh they were, came from the braintumor; (2) cases in whiclh metastasis was questionable, but could possiblyhave arisen from tumors of the brain; and (3) cases in which metastasis wasprobable but not entirely certain.

Group 1 .-Although it does not seeml- of mtlch value to review thosereports which occurred prior to the beginning of the era of modlern methodsof staining anid histologic (liaguiosis, it has become necessary to do so, sincecertain cases have beeni reported by otlher authors as examples of this typeof neoplastic activity. Among such cases are those of Klebs6 (I889), and

_rE~~~~~~~~~~~~~~~

Fic- 6.-Sections of hemangioblastoma removed at second operation, stainiedwith hematoxvlin and eosin. (X285)

Lindner7 (1902). Klebs' report seemis to us to hiave beeni extremnely brief,to have lacked details anid photomicrographs or othier illuistrationis. Liniduer'scase, althoughi reported in miore detail, miay well have beeni a case of highlyaniaplastic primiaryr tumior of the uriniary bladder with mietastasis to themieninges (carcinomiatosis ?).

In more recent years, certainrieports based ot clinical observations cer-tainly cannot be accepted as reports of proved cases. Pendergrass andWilbur' (I928) failed to report any hiistologic examination of the metastases,and in the report by Rogers, et al.1' ('193-), and by Sachs, et al.'13 (936),there were ndo postmortem stuidies migadle to confirm their observations. Al-though Foot and Zeek4 reported that they were unable to find ana melanotict aor niodules othier thai those intheluniigs anhd meningesu, it is well knownthat sich tumrsimetastasize early when the primary neoplasm may beever so small. Athough the patieidts ini these cases mav haveiad a primarny

348

Volume 118Number 3 INTRACRANIAL TUMORS

melanomiia in the leptomeninges, in view of the imaiy- possible primary lociit is certainly open to considerable doubt.

Group 2.-In this group we include those cases in which we believe thereis still a reasonable doubt that they fulfill all of the criteria for primaryintracrainial mietastasiziing tumiiors.

It w-ouldl seem to tus that Davis8 (1928) lhad fairly well l)roved his pointwere it tnot that his plhotoilmicrograplhs were not too distinct, and it is alsonoted that Bailey has niot accepte(d this case as a typical case. Mohlwills15(1930) photomicrographs are niot at all convincing, although his description\-would conifirmli hlis conitention.

AIt -

Ii c

W n, . .~~~~~~~~~~~~~~~~~AN'FIG. 7.-Sections of hemangioblastonia obtained at niecropsy and stained with hematoxylin anid

Cosin; A. showving sheets and strands of cells wsith nialiv mitotic figtires (x285); B, showing vascui-lar and psetidovascular chaninels (s250) ; C, shosing tumor tissue partially filling a vascular chaninel.(x55)

Brandt'st (1934) report was given without details and withotut photo-graphic evidence to support his case. In the case reported by MIittelbach8(I935), the patient may well have had a highly- malignant primary carcinomaof the lung, as suggested by Fischer-Wasels. For this reason, we are notat all sure he has proven that this is a case of metastasizing intracranial tumor.

Nelson's" (1936) report is a little more convincing than some of theothers, but certain inaccuracies ikake tus hesitate in accepting his case asa case of inedtilloblastonia. First of all, the tisstue obtained at the time of

349

ABBOTT AND LOVE Annials of SurgerySeptember, 19 4 3

the first operation performed at the clinic was examined by Dr. Kernohan,and reported as inadequate for diagnosis. For that reason, he did not com-mit himself any further than to state that the lesion appeared to be amalignant tumor. From the photomicrographs shown in Nelson's article,we agree with Dr. Bailey that it would be difficult to distinguish this tumorfrom a sarcoma. The question then arises as to the site of the primaryneoplasm.

A

B

FrG. 8.-Sections of hemangioblastoma in lung; stained with hematoxy-lin and eosin (X285); A, arrangement of cells in sheets and strands; B, for-mation of vascular channels.

The case recently reported by Jurow5 (1941) is most interesting in thatthe tumor was not a malignant but apparently a benign meningioma thatmetastasized to the lung. It seems regrettable that the photomicrographicevidence of the pulmonary lesion was not more convincing.

350

Volume 118 INTRACRANIAL TUMORSNumber 3

Russell and Sachs12 (Case 2) is open to some question since there wasan interval of I5 years between the removal of the malignant meningiomaand the appearance of the metastases in the lungs, pleura, abdominal lymphnodes and vertebra. Might the patient not have had another primaryextracranial tumor which produced these metastases ? In spite of theexcellent description given by these authors we could not be certain ofthe nature of the pulmonary lesions from their photomicrographs.

Group 3.-In the third group we have included the report of metastasizingtumors which we feel are, in all probability, excellent examples of primaryintracranial tumors which have metastasized outside the central nervoussystem and its envelopes. In this group we would include the cases of

S - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ IA

FIG. g.-Metastatic tumor of lung showing reticulum pattern; stained withPerdrau's reticulin stain. (XI5O)

(i) Cushing and Eisenhardt2 (malignant meningioma), Cases i and 3 ofRussell and Sachs ("fibrosarcoma"-malignant meningioma), and thecase which we have just reported (hemangioblastoma).

Another case, that reported by Towne,'4 might well belong in this groupwere it not for a more strict conception of metastasis. His patient had amalignant meningioma which invaded the inferior longitudinal sinus withextension into the straight, superior, right and left lateral sinuses and downinto the internal jugular and innominate veins as far as the upper part ofthe superior vena cava. It truly was a most remarkable neoplasm. Inaddition, there was a report of the removal, one year previous to the patient'sdeath, of an "endothelioma of the carotid body." This may or may not havebeen a metastatic nodule from the primary tumor in the brain or mighthave been an extension of the tumor through the wall of the left internaljugular vein !

351

ABBOTTAND LONTE ~~~~~~~Annals of SurgeryABBOTT AND LOVE September, 1 94 3

SUMMARYIn the case of metastasizing intracraniial hemangioblastoma which we

have reported, the patient lived three years and nine months after thefirst appearance of symptoms, and one year and three months after thefirst surgical procedure. Sixteen cases of primary intracranial tumors whichhave been reported as metastasizinig outside the central nervous systemand its envelopes have been reviewed. Six are niot accepted and sevenothers are accepted as possible cases of metastasizing intracranial tumors.but are open to question. Another three cases are in all probability excellentexamples of this type of lesioni. To these three cases we have added another.

REFERENCESI Brandt, M.: Fiinf FJalle mehrfacher Gliome im Grosshirnl. 'Verhandl. d. deutsch.

path. Gesellsch., 27, 39-42, 1934.2 Cushing, Harvey, and Eisenhardt, Louise: 'Meningiomas: Their Classification,

Regional Behaviour, Life History, and Surgical End-Results. Springfield, Illinois,Charles C. Thomas, I938, P. 7I6.

3 Davis, Loyal: Spongioblastoma Multiforme of the Brain. ANNALS OF SURGERY,87, 8-14, January, I928.

4 Foot, N. C., and Zeek, Pearl: Two Cases of M\elanoma of the MIeninges, withAutopsy. Am. J. Path., 7, 6o5-6i8, November, 193I.

5 Jurow, H. N,: Psammomatous Dural Endothelioma (NIeningioma), with PulmonaryMetastasis. Arch. Path., 32, 222-226, August, 194I.

6 Klebs, Edwin: Die allgemeine Pathologie. 'Vienna, Gustav Fischer, 18&9, vol 2,p. 628.

Lindner, Edwin: Ein Fall voni Endotheliom der Dura mater mit Metastase in derHirnblase. Beitrag zur Histogenese und Differentialdiagnose der Endotheliome.Inaug.-Dissert., Munchen, No. 59, Jule, 1902.

8 Mittelbach, Maria: Uber Gliome mit NIetastasen. Beitr. z. path. A.nat. u. z. allg.Path., 95, 538-572, I935.

Nelson, A. A.: Metastases of Intracraniial Tumors. Anm. J. Cancer-, 28 1-12.September, 1936.

10 Pendergrass, E. P., and( Wilbur, D. L.: Tumor of the Brain with WidespreadMetastases: Report of Two Cases. Arch. Neurol. & Psychiat., 19, 437-445,March, I928.

1 Roger, Henri, Alliez, Joseph, and Jouve, Andre: Syndrome d'adiposite du tronc etde la tee (a type d'adenome basophile hypophysaire de Cushing) avec metastasesneoplasiques pulmonaires et rachidiennes. Rev. neurol., I, 969-976, 1935.

2 Russell, V>. O., and Sachs, Ernest: Fibrosarcoma of Araclioidal Origin, witlhMetastases: Report of Four Cases, with Necropsy. A\rch. Path., 34, 240-261,JulY, 1942.

SSachs, Ernest, Rubinsteini, J. E.. and Arnieson, A. N.: Results of Roenitgeni Treat-menit of a Series of i i9 Gliomas. Arch. Neuirol. & Psychiat., 35, 597-6I5,March, I936.

4 Towne, E. B.: Invasion of the Intracranial Venous Sinuses by Meningiomia (DuralEndothelioma). ANNALS OF SURGERY, 83, 32 1-327, March, 1926.

13 Wohlwill, F.: Zur pathologischen Anatomie der malignen medianen Kleinhirntumorender Kinder (sog. "Medulloblastome"). Ztschr. f. d. ges. Neurol. u. Psychiat.. 128,587-614. 1930.

352