1395 Connatal Localized Neuroblastoma The Case to Delay Treatment Reinhold Kerbl, M.D.’ Christian E. Urban, M.D.‘ Herwig Lackner, M.D.’ Gerald Hofler, M.D? lnge M. Ambros, M . D . ~ Manfred Ratschek, M.D? Peter F. Ambros, Ph.D.3 ’ Division of Hematology/Dncology, University Children’s Hospital, Graz. Austria. * Institute of Pathology, University of Graz, Graz, Austria. Children’s Cancer Research Institute, St. Anna Kinderspital, Vienna, Austria. BACKGROUND. Spontaneous regression is well documented for a subset of widespread neuroblastomas (Stage 4s) and for localized residual tumors after incomplete resection. Possible spontaneous regression of untreated localized neuroblastoma in infants is frequently discussed, but has very rarely been demonstrated clinically. METHODS. We report four patients with localized neuroblastoma detected early; all were tumors of the adrenal gland. One patient was detected antenatally by ultrasound, the other three tumors were detected incidentally by sonography at the ages of birth, 1 week, and 7 weeks, respectively. In three patients treatment was delayed in order to await a possible spontaneous regression, and in one patient treatment was delayed due to an uncertain diagnosis. RESULTS. Local tumor growth was observed in three patients, and the tumors were removed 7, 12, and 16 weeks, respectively, after the initial diagnosis of neuroblastoma. All three patients are free of disease. The fourth patient developed liver metastases 4 weeks after the first suspicion of neuroblastoma. Progressive disease ended in death at the age of 17 months. CONCLUSIONS. None of the four patients showed spontaneous tumor regression. Noninvasive examinations and invasive investigations (in two patients) were unable to predict the tumor’s behavior. Based on present knowledge, a general “wait and see” strategy cannot be recommended for early and incidentally detected neuroblastoma patients. Cancer 1996; 721395-1401. 0 1996 American Cancer Society. KEYWORDS: connatal neuroblastoma, incidental diagnosis, sonography, delayed treatment, spontaneous regression, prognostic factors. Laboratory investigations were supported by the Austrian Children’s Cancer Fund. The authors wish to thank Dr. Steiner from the Department of Radiology, University of Graz, for performing computed tomography guided needle biopsies in two patients and Professor Ebner from the Department of Radiology, University of Graz, for providing the magnetic resonance images. The authors also are very much obliged to Dr. Jillian Mann of Birmingham, England, and to Dr. Carlo Dominici of Rome, Italy, for their critical reading of the manuscript. Address for reprints: Reinhold Kerbl, M.D., Division of Hematology/Oncology, University Children’s Hospital, Auenbruggerplatz 30, A-8036 Graz, Austria. Received October 2, 1995; accepted December 11. 1995. 0 1996 American Cancer Society N euroblastoma is the neoplasm with the highest rate of spontaneous regression. Spontaneous regression has been well documented for widespread neuroblastoma in infants younger than 1 year of age (INSS Stage 4s patients) without any treatment.’-5 Spontaneous regression of Alresidual tumors after incomplete resection has also been rep~rted.~.’ though Knudson and Meadows’ hypothesis of regressing one-hit lesions and progressing two-hit lesions*has yet to be confirmed, the mechanisms leading to regression and maturation of neuroblastic cells have recently begun to be underst~od.~-’* There seem to be three different destinations or pathways for biologically favorable neuroblastomas: (1) involution (disappearance without clinical manifestation), (2) spontaneous regression (disappearance after clinical manifestation), and (3) maturation without disappearance of the tumor mass. However, genetic prerequisites for these different pathways still need to be elucidated. Current opinion suggests that many primarily untreated localized neuroblastomas may involute, regress, or mature, but only a few reports describe the spontaneous disappearance of localized neurob1astomas.’”l5 The in- 1396 CANCER April 1,1996 / Volume 77 / Number 7 creased incidence of neuroblastoma in screening areas has led to the notion of “overdiagnosis” in screened populations. It is argued that screening in patients younger than 1year of age mainly detects neuroblastomas with otherwise good prognoses, i.e., ones that would disappear or mature without any treatment.I6-’” This speculation, however, has not yet been proved by clinical observations. In this article, we describe the natural course of four incidentally detected and primarily untreated localized neuroblastomas and discuss possible benefits and pitfalls of a “wait and see” strategy. MATERIALS AND METHODS Between February 1991 and October 1993, four infants were admitted to our institution after neuroblastoma had been detected incidentally at other hospitals. In three patients with localized neuroblastoma, treatment was delayed to await a possible spontaneous regression. In the fourth patient, treatment was delayed due to an unclear initial diagnosis, and the infant was subsequently admitted after liver metastases developed. Urine catecholamines were monitored by use of high performance liquid chromatography equipment (Bio-Rad Laboratories, Richmond, CA) and also by an ELISA method (Yamasa, Choshi, Japan). Normal values for the relevant age group had been determined and were found to be less than 20 pg/mg creatinine for vanillylmandelic acid &M A) and less than 37 pg/mg creatinine for homovanillic acid (HVA).Neuronspecific enolase (NSE),lactate dehydrogenase (LDH) and ferritin were determined by standard routine laboratory methods. Upper limits were 15 p,glL for NSE, age-dependent between 360 and 430 UI L for LDH, and 300 pg/L for ferritin. Monitoring of tumor volume was performed by ultrasonography and magnetic resonance imaging. Informed consent was obtained from parents for delaying treatment in all patients and for obtaining needle biopsies in two patients. To perform cytogenetic analyses, in two patients a computed tomography-guided needle biopsy from the tumor was performed under short ketamine anesthesia. Tumor samples were resuspended in RPMI 1640 plus antibiotics and 10% fetal calf serum. R-banding was performed by employing a chromomycin/distamycin/ DAPI staining technique.*l Double fluorescence in situ hybridization (FISH) analyses were carried out on cytospin slides prepared from uncultured tumor cells using pUC1-77 (DlZ1) specific for the centromeric region of chromosome 1 and theVNTR probe pl-79 (DlZ2) specific for the subtelomeric region of l p (1~36.33). Hybridization conditions and detection of the hybridized probes were performed as reported previously.“ All tumors were later resected and staged according to the classifications of Evans’ and the International Neuroblastoma Staging System (INSS).23,24 Histologic grading FIGURE 1. Patient 1. Urine catecholamines, tumor volume, and homovanillic acid (HVA)/vanillylmandelic acid (VMA) ratio of an antenatally detected patient with neuroblastoma. The urine catecholamines [VMA + HVA] correlated well with tumor volume. The H V W M A ratio, however, showed an inversion during the postnatal obseNation period. was performed in accordance with Shimada et al. and Joshi et al.2s~zfi Cytogenetic and FISH analyses from resected tumors were performed using the above mentioned methods. In addition, the ploidy of tumor cells was determined by flow cytometry according to standard conditions using a FACStar flow cytometer (Becton Dickinson, Braintree, MA). N-myc amplification was analyzed from the resected tumors by using FISH and Southern blot analyses. Case Reports Patient 1 . A tumor of the right adrenal gland (measuring 45 mm x 35 mm x 40 mm) was detected antenatally at the gestational age of 36 weeks by routine ultrasonography. Tumor size was almost unchanged during the following weeks. At the gestational age of 40 weeks, the baby was delivered without any problems. Immediately after birth, the tumor was diagnosed as a neuroblastoma based on elevated urine catecholamines (VMA 87 pglmg creatinine; HVA 18 pg/mg creatinine; HVA/VMA ratio 0.2). NSE was slightly elevated (18 pg/L); LDH (398 U/L) and ferritin (294 pg/L) were in the upper normal range. Assuming that the tumor might possibly regress spontaneously, no intervention was performed. Tumor size and tumor markers were consequently monitored. However, increasing tumor volume, increasing urine catecholamines and an inversion of the HVA/VMA ratio (Fig. 11, led to the infant undergoing surgery at the age of 8 weeks. At that time, the tumor size was 75 mm x 48 mm x 55 mm, the VMA was 78 pg/mg creatinine and HVA was 108 pglmg creati- Delayed Treatment in Neuroblastoma/Kerbl et al. 1397 FIGURE 2. Patient 2. Incidentally detected neuroblastoma of the left FIGURE 3. Patient 2. At age 15 weeks, the patient’s tumor volume had adrenal gland. Magnetic resonance imaging was performed at the age ot 3 weeks. more than doubled. The tumor growth caused local compression of the left kidney. nine (HVA/VMA ratio 1.38). NSE was unchanged (18 p g / L), LDH (270 IJ/L) was in the normal range, and ferritin had increased to 422 p g l L . The tumor (Evans’ Stage I, INSS Stage 1) was completely removed. Histology showed a n undifferentiated stroma-poor neuroblastoma. The mitosis karyorrhexis index was less than 100, indicating a favorable prognosis. According to the study by Joshi et aL2” the tumor was classified as Grade 1 with low risk. Cytogenetic and molecular biologic findings pointed toward a favorable prognosis (triploid tumor and absence of N-rnyc amplification and lp36 deletion). The patient remained free of disease after a follow-up time of 39 nion t hs. Pcitient 2. A tumor of the left adrenal gland (measuring 24 mm x 22 m m x 25 mm) was diagnosed by routine ultrasonography of the urinary tract on the infant’s second day of life. The urine catecholamines were borderline in the first test and positive in the next test when the infant was aged 3 weeks WlA was 31 pg/mg creatinine and [IVA was 38 pg/mg creatinine), thus confirming the diagnosis “neuroblastoma.” NSE was 18 pg/L, LDH 340 U/L, and ferritin 379 pg/L. ‘To assess the biologic features of the tumor, a computerized tomography-guided needle biopsy was performed at the age of 5 weeks and tumor material was analyzed by the above mentioned methods. Hicolor FISH analysis of interphase cells from disaggregated fresh tumor material revealed three centromeric D1Z1 and three telomeric D1%2 signals in virtually all cells analyzed, thus indicating three intact chromosomes 1 (trisomy, absence of l p deletion). Because these findings are consistent with a favorable prognosis, no further treatment was performed for the next 11 weeks. However, periodic ultrasonography and magnetic resonance imaging showed slow tumor growth (Figs. 2 and 3). At the age of 15 weeks, the tumor volume had more than doubled (tumor size 33 m m x 30 m m x 34 mm) and the urine catecholamines had further increased (VMA was 63 pg/mg creatinine and IIVA was 53 pglmg creatinine). Other tumor markers, however, were almost unchanged (NSE was 20 pg/L, LDH was 283 (J/L, and ferritin was 312 pg/Lj. The tumor (Evans’ Stage 11, INSS Stage 2B due to positive lymph nodes) was resected when the infant was aged 16 weeks. I Iistologic investigations revealed an undifferentiated stroma-poor neuroblastoma. The mitosis karyorrhexis index was less than 100, indicating a favorable prognosis. According to the report by Joshi et al., the tumor was classified as Grade 2 with low risk. Analysis of tumor material revealed triploidy without N-myc amplification or l p deletion. The patient was free of disease after a follow-up time of 27 months. Patient 3. A 7-week-old infant was admitted to a collaborating hospital following repeated vomiting due to gastroesophageal reflux. Routine abdominal ultrasonography showed a tumor of the right adrenal gland as an incidental finding, and the infant was referred to our department for further treatment. Ultrasound examination confirmed a tumor of the right adrenal gland (measuring 35 m m x 40 m m x 30 mm). In addition, a smaller tumor of the left adrenal gland (measuring 18 m m x 12 m m x 20 mm) was detected and the urine catecholamines were slightly elevated ( V M A was 29 pg/mg creatinine and IIVA was 42 pg/mg creatinine). NSE was 23 p g / L , LDH was 224 U/L, and ferritin was 93 p.g/L. At the age of 9 weeks, a computerized tomography-guided needle biopsy was performed from the tumor of the right adrenal gland. Hicolor FISH analysis of interphase tumor cells showed trisorny of chromosome 1 and the absence of l p deletion. 1398 CANCER April 1, 1996 / Volume 77 / Number 7 Within the next 6 weeks, slow tumor growth of both affected sites was observed (the tumor of the right adrenal gland measured 35 mm x 43 mm x 33 mm and the tumor of the left adrenal gland measured 25 mm x 22 mm x 28 mm), and the urine catecholamines were also slowly increasing (VMA was 53 pg/mg creatinine and HVA was 79 pg/mg creatinine). NSE (18 pglL), LDH (270 UIL), and ferritin (112 pg/L) were almost unchanged. Both tumors (Evans' and INSS bilateral Stage 11) were resected when the infant was aged 15 weeks, and unaffected adrenal tissue was preserved on both sides to avoid adrenal insufficiency. Histologic investigations of both tumors revealed neuroblastoma, and according to the classification of Shimada et aLZ6the tumors were undifferentiated and stroma-poor. The mitosis karyorrhexis index was less than 100, indicating a favorable prognosis. According to the classification by Joshi et a1.,26both tumors were classified as Grade 2 with low risk. Cytogenetic and molecular biologic findings showed triploid tumors without N-myc amplification or 1p deletion. At a follow-up time of 22 months, the patient was free of disease and without symptoms of adrenal insufficiency. Patient 4. Five days after birth, a localized tumor (measuring 20 mm x 30 mm x 25 mm) of the right adrenal gland was detected by routine ultrasonography at an outlying hospital. Although neuroblastoma was suspected, adrenal bleeding could not be excluded. Normal values of urine catecholamines (VMA was 12 pglmg creatinine and HVA was 22 pglmg creatinine), LDH (272 U/L), and ferritin (172 pg/L) led to a delay of intervention. NSE was not determined. Weekly ultrasound examinations showed the tumor unchanged for the next 3 weeks. At the age of 4 weeks, however, liver metastases had developed and the urine catecholamines (VMA was 98 pg/mg creatinine and HVA was 314 pg/mg creatinine), NSE (258 pg/L), LDH (858 U/L), and ferritin (1058 pg/L) were extremely elevated. The patient was admitted to our hospital for further treatment. Thorough examination of the patient showed that the previously described tumor of the right adrenal gland remained unchanged, but multiple liver metastases were present. Bone marrow samples from four aspirates and four biopsies were free of tumor cells, and technetium 99m-diphosphonate and '"J-meta-iodobenzylguanidine scintigraphy revealed no further metastatic sites. Due to these findings, the tumor was classified as Stage IV-s (4s).A short course of chemotherapy (3x vincristine, 1x cyclophosphamide, and 2x doxorubicin) was administered because extreme hepatomegaly was causing mechanical problems and respiratory compromise. Because the liver infiltration showed good regression and the urine catecholamines returned to almost normal values (Fig. 4), chemo- FIGURE 4. Patient 4. Urine catecholamines during the course of a disseminated neuroblastoma. The hornovanillic acid/vanillylmandelic acid ratio was greater than 1 at all times. A negative result of the first test immediately after birth led to treatment delay. As the course continued, increasing urine catecholamines correlated well with tumor progression and relapse, respectively. therapy was stopped in anticipation of further spontaneous tumor regression.However, at the age of 5 months, progression of the primary tumor occurred and liver metastases reappeared. In addition, bone marrow infiltration was diagnosed from bone marrow aspirates and biopsies. To reduce tumor mass and assess biologic features, the primary tumor (right adrenal gland) was removed. Histologic examinations revealed an undifferentiated neuroblastoma. Biologic features were those of an unfavorable neuroblastoma (diploid tumor, lp36 deletion, 100-fold N-myc amplification). Repeated multiagent chemotherapy (dacarbazine, doxorubicin, mechlorethamine, vincristine, cyclophosphamide, ifosfamide, teniposide, etoposide, and cisplatin) led to a second remission. Further therapy consisted of melphalan administered monthly in combination with carboplatin or etoposide. However, at the age of 16 months, relapse (in the bone marrow, abdomen, bone) was diagnosed and the patient died at the age of 17 months. DISCUSSION Neuroblastoma is referred to as the tumor with the highest potential of spontaneous regression and Different biologic pathways leading to involution, regression, and maturation have been d i s ~ u s s e d . 8 - ' ~ ~ ~ ~ - ~ ~ The possible spontaneous regression or maturation of localized neuroblastomas in infants who do not receive any treatment is used as the major argument against early neuroblastoma ~ c r e e n i n g . ' ~In" ~fact, the incidence of localized neuroblastoma has increased in screening Delayed Treatment in NeuroblastomaKerbl et al. thus indicating that some of the diagnosed neuroblastomas would have disappeared or matured without treatment. Only a few neuroblastomas with unfavorable or intermediate biologic markers have been detected by mass screening."-39 Antenatal sonography has become routine in developed countries and some hospitals have started to perform ultrasound screening of the urinary tract in neonates. These examinations are leading to the detection of additional neuroblastoma patients who might otherwise never he diagnosed. We are therefore faced with a new therapeutic dilemma regarding how to treat such patients. [Jnder the speculative assumption that more than one-third of localized neuroblastomas detected incidentally or by screening in infancy might regress or mature spontancously,~",."it appears that a clinician may be justified in withholding therapy for early and incidentally detected localized neuroblastomas to await spontaneous regression or maturation. When considering adoption of a "wait and see" strategy in most patients, the following questions are of importance: ( I ) Is it possible to predict tumor behavior reliably? (2) For how long is it justified to await spontaneous regression or maturation? and (3) Do infants benefit from a "wait and see" strategy? Mathematical models applied to neuroblastomas detected by screening at 6 months of age have shown that regression is more likely in extraadrenal neuroblastomas than in adrenal tumors.""."' Several studies have discovered factors correlating with favorable or unfavorable prognoses.""'.' These factors, however, are of less importance for the individual patient.39As far as our four patients were concerned, noninvasive parameters (clinical presentation, NSE, LDM, and ferritin) were unable to predict tumor behavior in the early course of the disease. Analysis of tumor material provides more information and would probably have led to early intervention in Patient 4. Aside from histologic cla~~ification,~'~'~~."~' several markers (ploidy, lp36 deletion, N-myc amplification, expression of CD44, NGFII, etc.) discriminating between neuroblastoma patients with favorable and unfavorable prognoses can be d e ~ e r m i n e d , ~ , ~ - ~ ~ 2 , 2 ~ . ~ l ' . ~ : ~Recently, .:~9.4:~-4~ it has been possible for some of these biologic features to be assessed even from a small amount of tumor cells. 12,22:16 However, for determination of these parameters, a (needle) biopsy is required and misdiagnosis in cases of heterogeneous tumors (e.g., "composite ganglioneurohlastoma"'g) cannot be excluded. Although two out of our four patients underwent a needle biopsy and biologic features indicated favorable prognosis, tumor growth was observed during the following weeks. The question arises of whether we waited long enough to observe tumor regression or maturation. As areas,"l..",:lti 1399 we know from patients with Stage 45 disease, primary tumor growth is followed by secondary regression in many instances. In maturing neuroblastomas, there usually remains a residual mass made up of mature tissue without active neuroblastoma. We cannot exclude that tumor regression or maturation would have occurred later in our first three patients. However, in the absence of reliable predictive parameters, it is difficult for ethical reasons to watch a growing tumor, especially if tumor growth is associated with a worsening of tumor markers (Patient I ) , the risk of kidney impairment (Patient 2), or the risk of adrenal insufficiency (Patient 3). In Patient 4, delayed treatment might have even led to the fatal progression of the disease course, although it cannot he known whether earlier treatment would have changed the prognosis. As far as our four patients were concerned, no benefit from the "wait and see" strategy could he demonstrated. In contrast, repeated clinical examinations were necessary. The tumors were eventually removed from all patients because of tumor progression. In general, because the outcome for infants after surgical treatment of easily resectable neuroblastoma is excellent in all reported studies, the benefit of elective surgery after months of observation might be low when compared with early resection. However, in patients with a high risk of complications following surgical treatment"' and in patients with an already decreasing tumor volume (e.g., after prenatal detect i ~ n , ' ~ , . "delayed ~ . ~ ~ ) treatment under close observation may be justified. Improved prenatal and perinatal care, as well as the increasing frequency of routine ultrasonography and biochemical neuroblastoma screening, have led to a n increase in the incidence of neuroblastoma in infancy and early childhood and to the detection of neuroblastomas that might otherwise regress or mature spontaneously. 18.20.3fi.30.48-50 Future efforts must concentrate on avoiding unnecessary treatment in these early detected patients, as well as in later diagnosed patients with favorable prognostic markers. Recently, in several European countries, a more conservative approach that avoids chemotherapy has been introduced for incompletely resected neuroblastomas with favorable prognostic markers.51From these studies, more knowledge regarding factors that predict tumor behavior will be available. It remains to be seen whether reliable prediction will be possible for individual patients. Furthermore, it has yet to be shown whether biologic features are of the same relevance in incompletely resected and primarily untreated neuroblastomas. When considering adoption of a "wait and see" strategy in most patients, there are two conflicting interests: the scientist's desire to know more about the natural course of localized neuroblastoma and the physician's desire to cure the in- 1400 CANCER April 1,1996 / Volume 77 I Number 7 fant (or child] as soon as possible. Although the real proportion of spontaneously regressing or maturing neuroblastomas can be elucidated only by a “wait and see” strategy, in most instances of early and incidentally detected neuroblastoma, based on the present knowledge, this practice cannot be generally recommended and should be limited to selected patients. However, a needle biopsy should be considered in these patients to avoid missing those with unfavorable biologic features. 16. 17. 18. 19. REFERENCES 1. Carlsen NL. Neuroblastoma: epidemiology and pattern of regression. Problems in interpreting results of mass screening. Am J Pediatr Hematol Oncol 1992;14:103- 10. 2. D’Angio GJ, Evans AE, Koop CE. Special pattern of widesoread neuroblastoma with a favourable prognosis. Lancet 1’971;1:1046-9. 3. Shimotake T, Iwai N, Tokiwa K, Deguchi E, Sawada T, Fushiki s. Increased numbers of argyrophilic nucleolar organizer regions between primary and metastatic sites predict tumor progression in stage IV and IV-S neuroblastoma. Cancer 1994;73:3103-7. 4. Suarez A, Hartmann 0, Vassal G, Giron A, Habrand JL, Valteau D, et al. Treatment of stage IV-S neuroblastoma: a study of 34 cases treated between 1982 and 1987. Med Pediatr Oncol 1991;19:473-7. 5. Bourhis J, Dominici C, McDowell H, Raschella G, Wilson G, Castello MA, et al. N-myc genomic content and DNA ploidy in stage IV-S neuroblastoma. J Clin Oncol 1991;9:1371-5. 6. Kushner BH, LaQuaglia MP, Ambros PF, Bonilla MA, Gerald WL, Rosenfield NS, et al. Survival from locally invasive or metastatic neuroblastoma without cytotoxic therapy. Proc Am SOC CZin Oncol 1993;12:413. 7. Matthay KK, Sather HN, Seeger RC, Haase GM, Hammond GD. Excellent outcome of stage I1 neuroblastoma is independent of residual disease and radiation therapy. J Clin Oncol 1989;7:236-44. 8. Knudson AG, Meadows AT. Regression of neuroblastoma It-S: a genetic hypothesis. N Engl J Med 1980;302:1254-5. 9. Kogner P, Barbany G, Dominici C, Castello MA, Raschella G, Persson H. Coexpression of messenger RNA for TRK protooncogene and low affinity nerve growth factor receptor in neuroblastoma with favorable prognosis. Cancer Res 1993;53:2044-50. 10. Nakagawara A, Arima-Nakagawara M, Scavarda NJ, Azar CG, Cantor AB, Brodeur GM. Association between high levels of expression of the TRK gene and favorable outcome in human neuroblastoma. N-Engl J Med 1993;328:847-54. 11. Ambros IM, Ambros PF. Schwann cells in neuroblastoma. E u r J Cancer 1995;31A.429-34. 12. Ambros PF, Ambros IM, Strehl S, Bauer S, Luegmayr A, Kovar H, et al. Regression and progression in neuroblastoma. Does genetics predict tumour behaviour? Eur J Cancer 1995; 31A.510-5. 13. Huddart S, Mann J, Stevens MCG, Gornall P, Corkery JJ, Spooner D, et al. Neuroblastoma diagnosed antenatally: a treatment dilemma with implications for screening. Med Pediatr Oncol 1992;20:156- 6 1. 14. Matsumura M, Tsunoda A, Nishi T, Nishihira H, Sasaki Y. Spontaneous regression of neuroblastoma detected by mass screening [letter]. Lancet 1991;338:447-8. 15. Dominici C, Andersen RA, Berthold F, Brodeur GM, Castel 20. 21. - - 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. V, Castello MA, et al. Timing of delivery and early surgical management after prenatal ultrasonographic detection of neuroblastoma. Med Pediatr Oncol 1995;25:242. Carlsen NL. How frequent is spontaneous remission of neuroblastoma? Implications for screening. Br J Cancer 1990;61:441-6. Hayashi Y, Hanada R, Yamamoto K. Biology of neuroblastomas in Japan found by screening. Am J Pediatr Hematol Oncol 1992;14:342-7. Murphy SB, Cohn SL, Craft AW, Woods WG, Sawada T, Castleberry RP, et al. Do children benefit from mass screening for neuroblastoma? Lancet 1991;337:344-6. Parker L, Craft AW. Neuroblastoma screening: more questions than answers? [editorial]. Eur J Cancer 1991;27:682-3. Sawada T. Past and future of neuroblastoma screening in Japan. Am J Pediatr Hematol Oncol 1992;14:320-6. Schweizer D, Ambros PF. Chromosome banding. Stain combinations for specific regions. In: Gosden JR, editor. Methods in molecular biology. Volume 29. Chromosome analysis protocols. Totowa: Humana Press, 1994:97- 112. Strehl S, Ambros PF. Fluorescence in situ hybridisation combined with immunohistochemistry for highly sensitive detection of chromosome 1 aberrations in neuroblastoma. Cytogenet Cell Genet 1993;63:24-8. Evans AE, D’Angio GJ, Randolph J. A proposed staging for children with neuroblastoma. Cancer 1971;27:374-8. Brodeur GM, Pritchard 7, Berthold F, Carlsen NL, Castel V, Castleberry RP, et al. Revisions of the international criteria for neuroblastoma diagnosis, staging and response to treatment. J Clin Oncol 1993;11:1466-77. Shimada H, Chatten J, Newton WA, Sachs N, Hamondi AB, Chiba T, et al. Histopathologic prognostic factors in neuroblastic tumors: definition of subtypes of ganglioneuroblastoma and an age-linked classification of neuroblastoma. J Natl Cancer Inst 1984;405- 16. Joshi W, Cantor AB, Brodeur GM, Look AT, Shuster JJ, Altshuler G, et al. Correlation between morphologic and other prognostic markers of neuroblastoma. Cancer 1993; 71:3173-81. Beckwith JB, Perrin EV. In situ neuroblastomas: a contribution to the natural history of neural crest tumors. Am JPathol 1963;43~1089-104. Ikeda Y, Lister J, Bouton JM, Buyukpamukcu M. Congenital neuroblastoma, neuroblastoma in situ, and the normal fetal development of the adrenal. J Pediatr Surg 1981; 16:636-44. Brodeur GM, Castleberry RP. Neuroblastoma. In: Pizzo PA, Poplack DG, editors. Principles and practice of pediatric oncology. 2nd edition. Philadelphia: JB Lippincott, 1989:73967. Bolande RP. The spontaneous regression of neuroblastoma. Experimental evidence for a natural host immunity. Pathol A ~ T 1991;26:187-99. ~ u Brodeur GM, Azar C, Brother M, Hiemstra J, Kaufman B, Marshall H, et al. Neuroblastoma. Effect of genetic factors on prognosis and treatment. Cancer 1992;70:1685-94. Di Martino D, Ponzoni M, Cornaglia-Ferraris P, Tonini GP. Different regulation of mid-size neurofilament and N-myc mRNA expression during neuroblastoma cell differentiation induced by retinoic acid. Cell MoZ Neurobiol 1990;10:45970. Suzuki T, Bogenmann E, Shimada H, Stram D, Seeger RC. Lack of high-affinity nerve growth factor receptors in aggressive neuroblastomas. J Natl Cancer Inst 1993;85:377-84. Delayed Treatment in Neuroblastoma/Kerbl et al. 34. Pritchard J, Hickman JA. Why does 4s neurohlastoma regress spontaneously? Lancet 1994;344:869-70. 35. Goodman SN. Neuroblastoma screening data. An epidemiologic analysis. Am J Dis Child 1991;145:1415-22. 36. Woods WG, Tuchman M, Bernstein ML, Leclerc JM, Brisson L, Look T, et al. Screening for neuroblastoma in North America. 2-year results from the Quebec project. Am J Pediatr Hi.matol Oncol 1992;14:312-9. 37. Kerhl R, Urban C, Starz I, Amhros IM, Strehl S, Kovar H, et al. Neuroblastorna with N-myc amplification detected by urine mass screening in infants after the sixth month of life. Med Pediatr Oncol 1993;2 1:625-6. 38. Hachitanda Y, Ishimoto K, Jun-ichi H, Shimada H. One hundred neuroblastornas detected through a mass screening system in Japan. Cancer 1994;74:3223-6. 39. Kusafuka T, Nagahara N, Oue T, Imura K, Nakamura T, Kohayashi Y, et al. Unfavorable DNA ploidy and Ha-ras p21 findings in neuroblastoma detected through mass screening. Cancer 1995;76:695-9. 40. Nishi M, Miyake H, Takeda T, Yonemori H, Hanai J, Kikuchi Y, et al. Cases of spontaneous regression and true patients detected in mass screening for neuroblastoma. Int J Pediatr HematollOncol 1995; 1:557-63. 41. Nishi M, Miyake H, Takeda T, Yonemori H, Hanai J, Kikuchi Y, et al. A trial to discriminate spontaneous regression from non-regression cases during mass screening for neurohlastoma. Jpn I Clin Oncol 1994;24:247-51. 42. Berthold F, Kassenbohmer R, Zieschang J. Multivariate evaluation of prognostic factors in localized neuroblastoma. Am J Pediatr Hematol Oncol 1994; 16:107- 15. 43. Oppedal BR, Storm-Mathisen I, Lie SO, Brandtzaeg P. Prognostic factors in neuroblastoma. Clinical, histopathologic, 44. 45. 46. 47. 48. 49. 50. 51. 1401 and immunohistochemical features and DNA ploidy in relation to prognosis. Cancer 1988;62:772-80. Leone A, Seeger RC, Hong CM, Hu YY, Arholeda MJ, Brodeur GM, et al. Evidence for nm23 RNA overexpression, DNA amplification and mutation in aggressive childhood neurohlastorna. Oncogene 1993;8:8S5-66. Look TA, Hayes FA, Shuster JJ, Douglas EC, Castleberry RP, Bowman LC, et al. Clinical relevance of tumor cell ploidy and N-nzyc gene amplification in childhood neuroblastoma: a Pediatric Oncology Group study. 1Clin Oncol 1991;9:58191. Huddart SN, Mann JR, McGukin AG, Corhett R. MYCN amplification by differential PCR. Pediatr Hematol Oncol 1993;10:31-4. Losty P, @inn F, Breatnach F, O’Meara A, Fitzgerald RJ. Neuroblastoma-a surgical perspective. Eur J Surg Oncol 1993; 19:33-6. Dorninici C, Raschella G, Padula A, Cleric0 A, Fares C, Bartolozzi S, et al. Genornic characteristics in prenatally-detected neurohlastoma. Med Pediatr Oncol 1992;20:424. Dominici C, Berthold F, Brodeur GM, Caste1V, Castello MA, Castleberry RP, et al. Prenatal neurohlastoma: a report by the Prenatal Neuroblastoma Study Group. Proc Am Assoc Cancer Res 1994;35:239. Ho PTC, Estroff JA, Kozakewich H , Shamberger RC, Lillehei CW, Grier HE, et al. Prenatal detection of neurohlastoma: a ten-year experience from the Dana-Farher Cancer Institute and Children’s Hospital. Pediatrics 1993;92:358-64. Beck D, De Bernardi B, Michon J, Ambros PF. Phase I1 clinical trial of surgery as the only treatment for INSS stage 2 neuroblastorna (94.01trial and study of the Localized Neuroblastoma European Study Group; SIOP 95.01 study). Med Pediatr Oncol 1995:25:278.