NEUROBLASTOMA IN CHILDHOOD--,4N EVALUATION OF SURGICAL MANAGEMENT BY C. EVERETT KOOP, M.D., ScD., F.A.C.S., W. B. KIESEWETTER, M.D., F.A.C.S. AND ROBERT C. HORN, M.D. Reprinted from Pediatrics, Vol. 16. No. 5 November, 19.55 NEUROBLASTOMA IN CHILDHOOD An Evaluation of Surgical Management By C. Everett Koop, M.D., Sc.D., F.A.C.S.,* W. B. Kiesewetter, M.D., F.A.C.S., and Robert C. Horn, h.0. E PER since Cushing and Wolbachl re- ported, in 1927, the transformation of a malignant neuroblastoma (sympathico- blastoma) into a benign ganglioneuroma there has been much interest in, and specu- lation concerning, the unique behavior of the neuroblastoma. Because of our observation of the sur- viva1 of children in `spite of management contrary to the usual principles of cancer surgery and because neuroblastoma is the most common malignant tumor we have encountered in childhood, it seemed worthwhile to record our experience with this tumor and its management. The pa- tients under our observation also serve as a "control" for those reported series where radiation therapy was used in therapeutic doses and is logically assumed to have played a large part in the survival of pa- tients. In the present series only 4 of 15 patients, surviving long enough to report here, received radiation in therapeutic dosage. After the unique behavior of neuroblas- toma with its tendency to change spon- taneously from a malignant to a benign form was mentioned repeatedly in the lit- erature, Farber,2 in 1940, recorded the fol-' lowing: Neuroblastomas may necrose and disappear after only a biopsy, they may mature into benign ganglioneuromas and they may be successfully treated by irradia- tion even after local metastasis. He re- ported 10 of 40 patients surviving 3 to 8 From the Surgical Clinic, the Children's Hospital of Philadelphia and the Harrison Department of Surgical Research, School of Medicine, University of Pennsylvania. Presented before the Centennial Medical Convo- cation of the Children's Hospital of Philadelphia, June 2-4,1%5. ' ADDRESS: 1740 BAnbridge Street, Philadelphia 46, Pennsylvania. years after histologic diagnosis, Wittenborg3 later reported on 73 patients including the 40 reported by Farber; he noted that 22 of these were alive 3 or more years and em- phasized the importance of irradiation ther- apy although indicating 2 of the survivals seemed to be "spontaneous cures." Phil- lips4 after an extensive review of over 600 cases in the literature, concluded that neuroblastoma could be brought under ul- timate control, if fatal termination could be prevented during the acutely malignant phase, because of the natural tendency of neuroblastoma to spontaneous remission. Snyder5 in 1951 recorded the known sur- viving patients with neuroblastoma re- ported to that time as numbering 37. Since then 45 additional survivals have been re- ported and to that number we now add 15. SUMMARY OF THE CLINICAL I.`./?.TERIAL A review of our clinical material indi- cates that of 45 cases, 19 are alive, but only 15 have survived 16 months or longer (Table I). Fourteen months is used as :j crucial time because in our experience children with neuroblastoma who have SIU- vived 14 months without x-ray therap! have not died at a later date with 1 excel'- tion, a child whose tumor was only biopsied 11 years before his death. It is also true that a review of all of our experience wit11 malignant tumors in childhood indicates that a 1Cmonth period of survival is al) parently equivalent to the 5year perio(' used to denote a "cure" in adult cancel statistics. All of the patients in this series ]l;ll! histological diagnoses made at operatio" Of the 41 patients reviewed 14 rnontll' after such diagnosis and/or treatment ? are dead, a survival rate of 36.6 per cellt Table II compares this survival rate witi other reported series. 652 ORIGINAL ARTICLES 653 TABLE I PATIENTS SURV~VINQ OPERATION FOR NEWOBLASTOMA FOURTEEN MONTHS OR LONQER patient Primwy sit8 - Izmrwval X-ray Iwmpleta Metastasea Therapy Survival Neuroblastoma Spinal Cord, Abdomen Abdomen Abdomen Abdomen Abdomen Abdomen + Media&mm, Neck nodes Liver No Liver No Liver, Bone Marrow Nodes No No No 12 yr. JT BU LW LM MG SW Neuroblastoma Neuroblastoma Neuroblastoma Neuroblastoma Neuroblastoma No Yes No Ye3 No 11 yr. 7 yr. 11 mo. 6yr.8mo. 6 ;yr. 5 yr. + ? + - + Neuroblastoma Neuroblastoma Malignant Ganglio- neuroma Malignant Ganglio- neuroma Malignant Ganglio- neuroma Malignant Ganglio- neuroma Malignant Ganglio- neuroma Malignant Ganglio- neuroma Malignant Ganglio- neuroma Chest, Neck Neck Chest CM DB HL + + + NO Yes No Q yr. 3 mo. 1 yr. 2 mo. 4yr.Smo. Chest cw + No Yes 4 yr. 1 mo. JM Abdomen + No No 4 yr. 1 mo. Abdomen, Pelvis JS No No 4 yr. - VA Abdomen + * Nodes No Q yr. 11 mo. EB Abdomen No YeS 1 yr. 4 mo. - MB Abdomen + Kidney, Liver Yes 1 yr. 4 mo. The tumors in this series ranged from completely undifferentiated neuroblastomas to ganglioneuromas showing fair, but not complete, differentiation. Most of the tu- mors showed some differentiation, at times fairly uniform throughout the tumor, biit at other times varying greatly from one area to another. Most of the surviving pa- tients had tumors which showed at least partial differentiation but this was not in- variably the case and, furthermore, many of the neuroblastomas showing appreciable differentiation proved fatal. A detailed study of the pathology is in preparation! The primary tumor was found in the Wroperitoneal space in 35 patients. Four tumors were in the neck, 4 in the retro- pleural space, and 2 in the spinal canal. The metastatic sites in the surviving pa- tients are indicated in Table I. Three con- clusions may be drawn from these data on metastases: All patients who were free of metastases at the time of operation have survived; no patient who had metastasis to the lungs or skeleton is alive; metastases in liver, lymph nodes or bone marrow do not necessarily indicate a fatal outcome. Bone marrow aspiration may be a valu- able diagnostic and prognostic aid. Marrow aspirations were obtained from 30 patients; 9 showed neuroblastoma cells, only 1 of these survived. At the titne of diagnosis the average age of those patients who have survived was 4.4 months, and for the fatal cases was 34 months. In reviewing the operative procedures carried out in the surviving patients as compared with those who succumbed, it is obvious that the attack made upon the tumor in those patients whd survived was far more radical than in the fatal cases. If one considers maximal surgical insult to be either extirpation of the tumor in toto, or as nearly complete removal as possible, 63 per cent of the surviving patients received such a surgical insult to the tumor, while 654 KOOP - NEUROBLASTOMA IN CHILDHOOD TABLE LT Snnvrvti PIREPORTEDSERIEBOF NEUROBLASTOMA casea survivai-s WittenborgS 73 2Q (30%0) Sn yde+ 24 3 w%`o) Phillips' 58 10 07Ya'o) Authors 41 15 @f-m only 29 per cent of the fatal cases received similar therapy. Case I CASE REPORTS A 17-month-old white female child was ad- mitted February 14, 1952, because of a mass in the abdomen, discovered on routine physical examination the previous day. The child had been asymptomatic. November 15, 1951, dur- ing hospitalization for traumatic cataract, it had been noted that the spleen was enlarged. The patient was a well-developed, well- nourished child in no acute distress. A large, hard mass extended 4 fingerbreadths below the left costal margin and the liver edge was palpa- ble. The hemoglobin was 12.1 gm., the leuko- cyte count 8,500/mm.3 with a normal differ- ential.' Urinalysis was negative. Intravenous urograms revealed a huge tumor in the left upper quadrant. No evidence of metastases was seen on chest films or in a survey of the long bones. Operation was performed on February 15, 1952. A large tumor mass was found in the lesser peritoneal sac surrounding the celiac axis, and extending into stomach, spleen, colon, and pancreas. Complete removal was imprac- tical but as much of the tumor as possible was removed by blunt and sharp dissection; about half of the mass was extirpated. The liver and a mesenteric lymph node were biopsied. The report of the pathology was: malignant ganglio- neuroma of the retroperitoneal region with metastasis to retroperitoneal lymph node; no histologic change in the liver tissue. The mass was palpable after the wound healed and seemed to grow larger until it was palpable as low as the umbilicus in April, 1952. There was little change thereafter until June, 1954 when the mass was barely palpable at the left costal margin. Throughout this tune the child grew and developed normally and re- mains well at the present time. Case II - An 8-month-old white female was admitted to the Children's Hospital on January 18, 1942, presenting an enlarged liver. Exploratory lapa. rotomy at another hospital had revealed an enlarged liver with nodules diagnosed on bi- opsy as metastatic neuroblastoma. However, the primary tumor was beneath the liver and no attempt was made to remove it. Her post- operative course was so good that she was re- ferred for confirmation of the diagnosis. The child was pale and well-developed with essentially no abnormal physical findings other than a hard, non-tender, nodular liver extend- ing 4 cm. below the right costal margin. The hemoglobin was 13.0 gm., leukocyte count 13,400/mm.s with 71 per cent lymphocytes. The urine was negative. Roentgenograms of the chest and long bones showed no evidence of metastases. Bone marrow was normal, with- out evidence of neuroblastoma cells. Operation was performed on January 20. A 6 cm. mass, later reported as a malignant ganglioneuroma, was removed from above the right kidney and multiple metastatic nodules were noted in the liver. X-ray therapy was started `on the second postoperative day and amino-an-fol, * 40 mg. by mouth daily, was started on the fifth postoperative day and continued at home through October, 1952. The child received vitamin B complex and injections of crude liver extract. During the first postoperative year, the liver became smooth and finally just palapable at the right costal margin. The child's growth and development were within normal limits and she is well at the present time. Case III A 2-month-old white male was admitted to the Children's Hospital on May 24, 1948, be- cause of anemia and a mass in the abdomen noted 2 weeks previously. He was a pale, wall- nourished infant with a smooth, irregular non- tender mass in the left upper quadrant of the abdomen and an enlarged liver palpable " fingerbreadths below the right costal rnargill, The hemoglobin was 11 gm., leukocyte count 19,2OO/mm.3 with 75 per cent normal lyn$o cytes and 20 per cent neutrophiles. The uril"' was negative. Roentgenogram of the chest did not reveal any abnormality, an intravenn"' pyelogram showed diminished secretion but "' * A folic acid antagonist. ORIGINAL definite tumor. The blood NPN was 29 mg./ 100 ml. Exploratory laparotomy revealed a mass in- ,.elving the area of both renal hili. The liver ,"as the site of hundreds of small, white nod- ules. The mass was broken into, its contents scooped and sucked out and a liver biopsy tnken. Both the primary tumor and the meta- stases proved to be neuroblastoma. The patient !vas discharged June 6, 1948. Six weeks after operation, the child devel- aped subcutaneous nodules which were as- rumed to be metastases and not biopsied. Three months after operation the mass in the ab- domen, which was palpable in June, 1948, was getting smaller. The child did well at home but was admitted on June 28, 1949, because of polyuria. An intravenous pyelogram revealed moderate hy- honephrosis with dilatation of the left pelvis and calyces. Neither ureter was well visualized. Retrograde pyelography demonstrated a normal right urinary tract and obstruction in the left ureter about 5 cm. above the uretero-vesicle junction. Calcification within a tumor in the abdomen in the region of Ll to L3 was noted. It was felt that further abdominal surgery was contraindicated and the boy was discharged. Repeat studies 2 years later revealed no change in the kidneys and no calcification. The third hospital admission was on Sep- tember 20, 1953, for orchiopexy and inguinal heroiorrhaphy. No tumor could be felt under anesthesia and the child was asymptomatic. He remains so at the present time. DISCUSSION There are a number of unanswered ques- tions concerning neuroblastomas. Do ma-. lignant tumors actually "mature" when they become ganglioneuromas? What histo- logic changes are evident in those tumors which remain in spite of the patients' sur- vival? Does any therapy, whether surgical, radiation, or chemical, merely hasten a nat- ural phenomenon, or do these materially affect the mortality? Some of the unique characteristics of tumors in children must be recognized be- fore answering any of these questions. Most tumors in childhood have a period of ~~ptomatology as well as a life expect- ancy for the patient in fatal cases which is foreshortened, in contrast to experience ARTICLES 655 with cancer in adults. Death in a fatal case is early in contrast to experience with adults and survival for 14 months after definitive diagnosis appears to approximate the conventional 5-year "cure" used in eval- uating therapy of cancer in adults. The age at which the tumor first makes itself evi- dent clinically seems to have some rela- tionship with survival as is noted above. The answers to the first 2 questions would seem to be impossible except by conjecture until studies are available on tumors, diagnosed as malignant by histo- logic examination, which have persisted in a patient who has survived in spite of appar- ently playing host to a neoplasm. In answering the question concerning the effectiveness of therapy, one must con- fess ignorance concerning the mechanism of any type of therapy. It would seem that 2 observations can be made from our data which do not seem to be evident in the writings of other authors. First, most of the patients in the present series seem to have survived in spite of the fact that they were not treated with radiation. In the 15 sur- viving patients considered here, only 4 re- ceived x-ray therapy in what could be con- sidered a therapeutic dose. Therefore, it seems obvious that we cannot assign any major role to radiation therapy in promot- ing survival in this series. This is in con- trast to the assumption of most authors that radiation therapy is responsible for the control of neuroblastomas, if not the ultimate cure. Second, surgical attack which might be considered "radical" does seem to bear some positive relationship to survival in this series. In a tumor, known to regress spontaneously on some occasions, one does not know how to approach the tnmor mass if it is not easily removable in toto. It has been the custom in many clinics when a child has a huge retroperitoneal tumor and gross metastases, to biopsy the primary tumor and perhaps a metastatic site but not to attempt complete excision of the tumor. There is certainly a sound basis for this type of approach in other tumors, but on the basis of our experience we can question whether this is proper KOOP - NEUROBLASTOMA IN CHILDHOOD management in a patient with a neuro- blastoma. Patients so treated by us have survived only rarely, 3 of 22 patients, a mortality of 87 per cent. When we have approached a tumor of large size, even in the presence of metas- tases, with the intent of removing it com- pletely or if such is impossible, with the aim of removing as much as can be done without danger to the patient, we have had survival of 12 of 19 patients or 63 per cent. In view of the survival rate in patients treated in this fashion as compared with the survival rates of those treated chiefly by x-ray therapy after. histologic diagnosis, it seems valid to assume that major surgical insult might have something to do with the regression of the neuroblastoma and/ or the survival of the patient. In our experience, when a tumor can be removed in toto without difficulty, this course should be followed. If complete ex- tirpation is not possible, then we remove as much of the tumor as is practicable, guided by its attachments. If the tumor is fixed to an adjacent structure, such as kidney, adrenal gland or spleen, where re- moval of the tumor and the attached solid viscus can be undertaken without endan- gering the child's future, this type of pro- cedure is carried out without question. When a tumor mass surrounds vital areas such as both renal hili, the celiac axis or the porta of the liver, we have carried out partial extirpation of the tumor even though there may be obvious evidence of metastases. Such removal may consist of making an opening into the tumor and scooping out its central portion or separa- tion, by blunt dissection, of the adjacent structures and the tumor. If fixation to an adjacent viscus gives way to infiltration, the tumor is cut across at that point. It is sometimes possible in a necrotic tumor to remove the bulk of the mass by vigorous suction. Such a procedure is almost always followed by rather brisk bleeding, but this has never been beyond control. Ligation of blood vessels not supporting the life of an adjacent viscus has also been carried out. In spite of our desire to remove as much tumor as possible, we have not made it a practice to resect en bloc stomach, colon, or duodenum when the tumor has been tightly fixed to these structures, We feel that, because the survival rate is as good as it is by the above management and spontaneous regression is also possible, it does not seem justifiable to remove large segments of adjacent hollow viscera in an effort to remove gross evidence of invasion, when microscopic invasion probably al- ready exists beyond that point. The reason the patients in this series did not receive x-ray therapy was that in early cases treated by us the infiltration of adjacent structures with tumor, in adition to tumor cells spilled at operation, led us to believe x-ray therapy would be of little material benefit. As this series grew, and survival seemed to be a relatively frequent occurence, we contin- ued along the line of not administering x-ray therapy. If our thesis concerning the effect of major surgical insult is valid, radiation should be of additional value and we are currently using it in therapeutic dosage after surgery. It is interesting that all of the patients in this series who died had metastases, but almost 50 per cent of the survivors also had metastases. No patient who was free of metastases succumbed to the primar! tumor. Others have called attention to the poor prognosis associated with metastases to the skull and long bones. 3~ `, * Our es- perience certainly confirms this. No patieilt in this series with metastases to the lu$s I or skeleton survived. SUMMARY AND CONCLUSIONS Forty-five patients with neuroblastcmj are reviewed, 41 more than 16 months post- operative. Fifteen of the forty-one are aliJ'r and well, a survival of 36.6 per cent. There appears to be some correlation he tween survival of patients with neuroblas- toma and a major surgical insult to tllc' tumor. In 22 patients where biopsy alclly was carried out, there was a mortality of f;' per cent. In 19 patients treated by a sur? cal attempt at radical removal of the tuill"`, ORIGINAL ,,-itbout subsequent radiation therapy, 12 ,,,,.vived, a rate of 63 per cent. The age of the surviving patients at the tirlle of diagnosis was 4.4 months as con- tri,sted with an average of 34 months for t\,ese who died. The survival of 11 of 15 patients for ,yriods ranging from 16 months to 12 years ,fter diagnosis is unrelated to x-ray iher- No patient in this series, free of gross ,,,ctastases at the time of operation, has (lied up to the time of this report. Forty- ,line per cent of the surviving patients had gross metastases, but no patient with me- tastases to lung or bone survived. REFERENCES ARTICLES 657 neuroblastoma ha essite ininterrumpitemente Ie object0 de multe interesse e speculation. A causa de no&e observation que juveniles con neuroblastoma superviveva in despecto de tractamento contrari al acceptate principios de cancerochirurgia e proque le neuroblastoma es le plus frequente tumor de character maligne que nos ha incontrate in juveniles, nos ha con- cludite que publicar nostre experientias con iste tumor e su tractamento esserea un interprisa ben valente le pena. Le patientes observate per nos servi simultaneemente de "controlo" pro ille previemente reportate series de cases in que radiation in doses therapeutic esseva usate e logicamente considerate coma un factor im- portante in le superviventia de1 patientes. In le presente serie solmente 4 de1 15 patientes supervivente usque al tempore de1 redaction de1 . report0 ha recipite radiation in doses thera- peutic. 1. Gushing, H., and Wolbach, S. B.: Trans- formation of malignant paravertebral sympathicoblastoma into a benign gan- glioneuroma. Am. J. Path., 3:203, 1927. 2. Farber, S.: Neuroblastoma. (Transactions American Pediatric Society.) Am. J. Dis. Child., 60:749, 1940. ,j. Wittenborg, M. H.: Roentgen therapy in neuroblastoma. Radiology, 54:679, 1950. 1. Phillips, R.: Neuroblastoma. Ann. Roy. Coll. Surgeons England, 12:29, 1953. 5. Snyder, W. H., Jr., Kruse, C. A., Greaney, E. M., and Chaffin, L.: Retroperitoneal tumors in infants and children. Arch. Surg., 63:26, 1951. 6. Horn, R. C., and Koop, C. E.: In prepara- tion. 7. Silverstone, S. M., and Harris W.: Treat- ment of neuroblastoma. J. Mt. Sinai Hosp., 17:1083, 1150. 9. Uhlmann, E. M., and von Essen, C.: Neuro- blastoma (neuroblastoma sympatheticurn). PEDIATRICS, 15:402, 1955. INTERLINGUA ABSTRACT Neuroblastoma in Infantia: Un Evaluta- tion de1 Tractamento Chirurgic Depost que Cushing e Wolbach reportava in 1927 le transformation de maligne neuro- blastoma (sympathicoblastoma) in benigne ganglioneuroma, le conducta extraordinari de1 Es revidite le cases de 45 patientes con neuroblastoma. In 41 cases de1 serie le opera- tion esseva executate plus que 16 menses retro. In 15 de iste 41 cases le patientes vive e se trova ben. Isto es un superviventia de 36 pro cento. II existe apparentemente un certe relation inter le superviventia de1 patientes e un major insulto chirurgic de1 tumor. Inter 22 patientes in qui solmente un examine bioptic esseva executate le mortalitate attingeva 87 pro cento. Inter 19 patientes in qui le essay0 esseva interprendite de obliterar le tumor radicalmente per medios chirurgic e in qui nuIle therapia a radiation sequeva le intervention chirurgic, le superviventia attingeva 63 pro cento. Le etate median al tempore de1 diagnose esseva 4,4 menses pro le gruppo de1 patientes supervi- vente; pro le patientes qui moriva le etate median al tempore de1 diagnose esseva 34 menses. Le facto que 11 inter 15 patientes superviveva pro periodos de inter 16 menses e 12 annos post Ie diagnose initia1 esseva sin relation al roentgenotherapia. Nulle de1 patientes qui esseva Iibere de grossier meta- stases al tempore de1 operation ha morite usque al tempore de1 presente reporto. Quaranta- nove pro cent0 de1 patientes supervivente habeva grossier metastases, sed nulle patiente ha supervivite in que le metastase habeva at- tingite p&nones 0 osso.