0360-3016/92 $5.00 + .I0 Copyright Q 1992 Pergamon Press Ltd.
In1 J Radiarron Oncology BID/. Phys Vol. 23. PP. 525-531 Printed in the U.S.A. All rights reserved.
??Clinical Original Contribution
IS THERE A PLACE HEPATOBLASTOMAS
FOR RADIATION THERAPY IN THE MANAGEMENT OF AND HEPATOCELLULAR CARCINOMAS IN CHILDREN?
J.-L. HABRAND, M.D.,’ D. NEHME, M.D.,’ C. KALIFA, M.D.,2 F. GAUTHIER, M.D.,4 M. GRUNER, M.D.,’ D. SARRAZIN, M.D.,’ M.-J. TERRIER-LACOMBE, M.D.3 AND J. LEMERLE, M.D.2 ‘Pediatric Unit, Department of Radiation Medicine, ‘Department of Pediatrics, ‘Department of Pathology, Institut GustaveRoussy, 94805 Villejuif Cedex; 4Department of Pediatric Surgery, Hopital BicCtre, 94275 Le Kremlin-Bicetre Cedex; and ‘Department of Pediatric Surgery, Hopital Trousseau, 75571 Paris Cedex, France
From May 1978 to August 1988, 15 childrenwith a primarymalignantliver tumor receivedradiationtherapy as part of their managementat the Institnt Gustave-Roussy. Age ranged from 4 months to 13 years. The male to female ratio was 1.5. Eleven patients had a histologically proven hepatoblastoma, two a hepatocellular carcinoma, and histology was not documented in two. Resection of the primary liver tumor was performed in nine cases, and all patients also received sequential chemotherapy, generally preoperative and alternating vincrlstine, doxorubicin, cyclophosphamide with vincristine, cyclophosphamide, and cis-platinum. Radiotherapy was performed postoperatively in eight incompletely resected patients. Six of eight are alive and free of disease 4-83 months following treatment (median 39 months) and 11-98 months since diagnosis (median 45 months). All but one were treated to limited fields to a total dose of 25-45 Gy (median 40 Gy). One patient became resectable by a combination of 24 Gy to the whole liver and concomitant 5FU and Cis-Platinum and remains with no evidence of disease 68 months following radiation therapy. Of four unresectable primaries, only one was controlled by radiotherapy. Neither of two children with pulmonary metastases were controlled by whole lung irradiation to a dose of 18 and 20 Gy, respectively, and one still remains stable 41 months after resection of a residual metastatic nodule. Neither of two hepatocellular carcinomas were controlled by doses up to 40 Gy. This small series suggests that in hepatoblastoma, radiotherapy to a total of 2545 Gy fractionated doses, combined with chemotherapy, can play a role in selected inoperable children and also in those with minimal postoperative residues below 2 cm. It also indicates that in bepatocellular carcinoma, radiotherapy is ineffective in this dose-range. Radiotherapy, Children, Hepatoblastoma, Hepatocellular carcinoma. INTRODUCTION Primary malignant liver tumors (PMLTs) are extremely rare in children and account for approximately 2% of all malignant tumors observed (6, 16). As intraabdominal neoplasms, they occur less frequently than nephroblastomas and neuroblastomas (8). Most of them are hepatoblastomas (Hbs) and less frequently hepatocellular carcinomas (HcCs), a form similar to that observed in adults. In one series, they represented 52 and 39% of PMLTs, respectively (14). The remaining 9% of cases consisted of unusual histologic types such as rhabdomyosarcoma and angiosarcoma. Although both Hb and HcC are of similar origin, various epidemiological factors differentiate them (7). Microscopically the Hb may be subclassified as pure epithelial Hb and mixed Hb, which also contains mesenchymal tissue (osteoid, cartilaginous, or rhabdomyo-
b&tic). The epithelial component is represented by either fetal or embryonal cells. It can be separated into predominantly fetal or embryo& types, depending which pattern occupies more than 50% of cross-sections. An additional anaplastic form is composed mainly of poorly differentiated small to medium size cells without any glandular structure visible. With respect to outcome, the HcC is generally regarded as the worst form (13) with the exception of the fibrolamellar form, which carries an excellent prognosis (27). As far as Hb is concerned, it is generally regarded that the fetal type carries a more favorable prognosis and that the anaplastic form is invariably lethal. For decades, surgical resection has been the mainstay of treatment. At the time of diagnosis only 1560% tumors are operable (8, 14, 20, 26, 27, 28). But even complete surgical resection achieves long-term survival only in 30-
Reprint requests to: J. L. Habrand, M.D., Department of Radiation Medicine, Institut Gustave-Roussy, 39, rue Camille Desmoulins, 94805 Villejuif Cedex, France.
Acknowledgements-We are grateful to Catherine Joelle GuCry for skillful technical assistance. Accepted for publication 12 December 199 1. 525
Loge and
526
I. J. Radiation Oncology 0 Biology 0 Physics
60% (7,8, 14,26,28). Most relapses occur within the first year following diagnosis (14). These results have stimulated the interest for chemotherapy, especially in inoperable patients. Various series (2,32,35) have reported that most patients had a dramatic tumor shrinkage following polychemotherapy. On the other hand, very sparse data are available on the efficacy and role of radiation therapy, but it is generally regarded as an “ancillary” treatment ( 14). These issues were addressed in this review of 15 children irradiated at the Institut Gustave-Roussy (IGR) of Villejuif between 1978 and 1988. METHODS
AND MATERIALS
Patients From May 1978 to August 1988, 15 children with a diagnosis of Hb or HcC were treated at the IGR with radiation therapy. All of them also received chemotherapy k surgery. The group requiring radiotherapy (see treatment protocol below) represented 22% (g) of all children managed at our institution with this type of tumor in the same interval. Age ranged from 4 months to 13 years with 3 patients being below the age of 3 years. The male to female ratio was 1.5 (E). All patients presented with a rapidly growing abdominal mass associated with an elevated alpha-feto-protein (AFP) serum level: it ranged from 0.4 to 370 X lo3 rig/ml (normal level: < 30 ng/ml, above the age of 3 months). The imaging work-up included abdominal ultrasound in 12 patients, Tc 99 liver scan in 4, CT scan in 6, preoperative angiography in 12, and chest x-ray in 15. The right hepatic lobe was involved in all patients and was associated with a left involvement in eight. The histology was available in 13 of them (surgical specimen: 12; biopsy alone: 1) and was reviewed for the purpose of this study. Eleven children had an Hb, purely or predominantly fetal in seven, embryonal in two, and mixed in two. Among them, one patient with a fetal type and one with an embryonal one had an osteoid component, whereas one patient with a mixed form had both osteoid and anaplastic components. Two children presented with an HcC. In two other inoperable cases, the diagnosis of malignancy was made exclusively clinically and on a persistently high level of AFP. Treatment protocol The primary emphasis of therapy is gross total surgical resection of tumor. We sought to achieve this goal short of total hepatectomy. To increase the chance of tumors being resectable, 12 of 13 cases without metastatic disease received initial cytoreductive chemotherapy. It alternated usually vincristine (VCR = 1.5 mg/m2, day l), Doxorubicin (ADR = 40 mg/m2, day l-2), and Cyclophos-
Volume 23, Number 3, 1992
phamide (CYC = 600 mg/m2, day 2) with a combination of Vincristine (1.5 mg/m2, day 22), Cyclophosphamide (600 mg/m2, day 22), and Cis-Platinum (CDDP = 100 mg/m2, day 23). Its efficacy was assessed clinically, radiologically, and on the AFP serum profile. This protocol was conducted until maximum tumor reduction. If the tumor was still deemed inoperable, a salvage therapy, generally Actinomycin D (ACD = 0.45 mg/m2, per day for 5 days), was instituted. One patient received the IVA regimen (ifosfamide, vincristine, actinomycin D) and carboplatin. Curative radiotherapy was applied to patients with grossly or microscopically incomplete resection. Radiotherapy was started 4 to 6 weeks following surgery so as not to interfere with hepatic regeneration (34). A dose of 25 to 45 Gy was administered, targeted to the residual post-operative disease only, in 5 daily sessions of 1.5 to 2 Gy per fraction. Multiple field arrangements were employed to avoid the right kidney as much as possible. Co 60 or high energy x-rays from linear accelerators were used in all cases. Palliative irradiation of the entire liver or whole lung coverage in pulmonary metastases were also contemplated in selected situations. Postoperative chemotherapy was given in 8 of 9 resected patients, combining CDDP, VCR, and CYC. Overall treatment duration totaled 1 year or so. RESULTS At the time of evaluation, E (66%) patients were alive and free of disease. Total follow-up since initial symptoms ranged from 4 to 98 months (median 37) and disease-free survival upon completion of radiotherapy, from 0 to 83 months (median 23). Actuarial 3 and 5 year-survival were, respectively, 65 and 53% and disease-free survival, 57% and 57%. Resected patients (9 cases) (patientsI-IX, Table I) Eight out of nine children received postoperative irradiation after incomplete resection (following preoperative chemotherapy in seven). Four of them had gross residuals (Group 1) mostly encased along the inferior vena cava, the portal, or the supra hepatic veins. Tumors were barely detected by post-operative imaging techniques and never exceeded more than 2 cm in greatest dimension. Four had microscopic positive margins (Group 2). Pathological examination of the surgical specimen in the seven patients managed with preoperative chemotherapy, displayed a considerable amount of necrosis, attributed to the efficacy of chemotherapy. But tumor foci were still visible. Total dose ranged from 25 Gy to 45 Gy (median 40 Gy) with the largest dose given to gross residues. The whole liver was irradiated only in one case due to disseminated tumor foci.
RT
in the management of childhood hepatoblastomas and hepatocellular carcinomas 0 J.-L. HABRAND etal.
Table 1. Children treated with resection, radiotherapy and chemotherapy:
521
Status post initial symptoms
Radiotherapy Status (in months since Dg)
Patient (sex, age in months) pathology
Chemotherapy pre/ post op.
Timing
Target
Technique (TD in Gy)
I (F, 120) Fetal Hb + 0 II (M, 8) Fetal Hb III (M, 15) Fetal Hb IV (F, 156) HcC V (M, 4) Pure fetal Hb VI (F, 9) Embryo HB + 0 VII (F, 5) Pure fetal Hb VIII (M, 24) Fetal +embryoHb+O+A IX (M, 168) Fetal Hb
VADC, CDDP +/+ VAC, CDDP +/+ VADC, 5FU -/+ VADC, CDDP +/+ Id VADC, CDDP +/VADC, CDDP +/+
Pre op. Post op. Id Id Id Id Id
Macro Micro Id Id Id Macro Id
Whole, 25 + 5FU, CDDP Focal, 25 Focal, 30 Whole, 40 Focal. 45 Id Focal, 35
NED, 76 NED, 43 NED, 98 D, 48, LF NED, 22 NED, 81 NED, 51
Id Id
Id Id
Id Id
Focal, 42 Focal, 40
NED, 37 D, ll,LF+PM
TD = Total dose; NED = No evidence of disease; D = Dead; LF = Local failure; PM = Pulmonary metastases; Id = Identical to the above; Fetal Hb = Hb, fetal type predominantly (50%); Embryonal Hb = Hb, embryonal type predominantly (50%); Fetal + embryonal Hb = Mixed fetal (50%) and embryonal (50%) type of Hb, 0 = Osteoid component; A = Focal anaplasia; Micro = Microscopic invasion; Macro = Macroscopic invasion; Dg = Diagnosis; other abbreviations: see text.
Six of eight children (75%) have remained free of disease 4 to 83 months following completion of radiotherapy (median 32 months). One patient in group 2, with an HcC failed locally and one in group 1 both locally and distantly. One out of nine children (patient I) received preoperative radiotherapy potentiated by cytotoxic agents used as radiosensitizers, 5 fluorouracil(5 FU), and CDDP: this 10 year-old girl presented with a massive right Hb, 12 X 15 cm large. A partial clinical and biological response had been obtained after four cycles of chemotherapy (VCR, ADR, CYC, CDDP). Nevertheless, at laparotomy, extensive involvement of both lobes and hepatic convergence precluded any resection. The patient was started on a pilot study combining each month, one week of radiotherapy (3 Gy X 2, day 3 and 4), along with 5 PU (350 mg/m2, day 2, 3,4, 6) + CDDP (15 mg/m2, days 1, 5, 7). After 4 cycles of 6 Gy with 5 FU and CDDP, she was operated on again. A complete resection was made possible at that time and the tumor was completely necrotic at microscopic examination. This patient is still alive and doing well, 68 months post irradiation. None of the untoward late side effects that have been reported after the use of high doses per fraction, have been recorded at the last follow-up.
Unresectablepatients (5 cases) (patients X-XIII. Table 2) Radiotherapy was tested in three patients with massive tumors that progressed under chemotherapy (patients XIXIII). Doses of 30 Gy in large volumes + 15 Gy boost in one of them, were administered. All died at 4, 8, and 12 months, respectively, with only one transient tumor stabilization. Another 10 month-old-child (patient X), presented with a limited recurrence two months after complete resection. A new resection was ruled out given the tumor’s vascular connection and he was managed by chemotherapy (Ifosfamide, Doxorubicin, and Carboplatin) for 9 months. Radiotherapy was targeted at small ultrasonically defined abnormalities. A total dose of 45 Gy was delivered. This patient has not failed with 15 months follow-up.
Metastatic patients (2 cases) (patients XIV, XV, Table 3) Two children were irradiated for pulmonary metastases. One of them received 20 Gy with correction for air-transmission in the lungs using Co60, and the second one 18 Gy with 18 MV x-rays, concomitant with Actinomycin D. In the first child, radiation therapy was administered
Table 2. Unresectable children treated with chemotherapy and radiotherapy Radiotherapy Patient (sex, age in months) pathology
Chemotherapy
Target
Technique (TD in Gy)
X (M, 10) Fetal + embryo Hb XI (F, 30) XII (M, 36) HcC XIII (F, 102) -
VADC, CDDP IVA, Carboplatin VADC Id VADC, CDDP
Macro Massive Id Id
Focal, 45 Focal, 30 Whole, 30 Whole, 30 + boost, 15
Abbreviations:
see Table 1.
status (in months since Dg) NED, 33 D, 12, LF D, 8, LF D. 4. LF
528
I. J. Radiation Oncology 0 Biology0 Physics
Volume 23, Number 3, 1992
Table 3. Radiation therapy of pulmonary
metastases
Radiotherapy Patient (sex, age in months) pathology
Chemotherapy
Target
Technique (Td in Gy)
Salvage surgery
XIV (M, 13) Fetal Hb XV (M, 114) Embryo Hb + 0
VADC, 5FU VADC, CDDP
Macro Micro
Whole lungs, 20 Whole lungs, 18
+ -
Abbreviations:
Status (in months since Dg) NED, 58 Alive, 20, PM
see Table 1.
for multiple pulmonary metastases, discovered shortly after an extended hepatectomy, and which progressed under chemotherapy. After a 4 month-free interval, multiple nodules were again visible and motivated a triple wedge resection, through a double thoracotomy. Pathologically, only one nodule still contained tumor. This patient was still doing well 4 1 months later. In the second child, multiple pulmonary metastases were also discovered under maintenance chemotherapy, and completely resected through a double thoracotomy. Although radiation therapy was only administered as a consolidation, this patient failed in the lungs 8 months later and was placed on a phase 1 interleukin 2 trial. None of both patients have relapsed at the site of the primary. Late toxicity (patients VI and X) Two patients presented a late complication in relation with radiotherapy. One (patient VI) had a scoliosis 6 years following 45 Gy delivered to a wide clipped area through AP-PA fields, abutting but not covering the entire spine, in order to spare as much normal liver as possible. Consequently, the right part of the vertebral bodies from T8 to L2 received a dose gradient between 50 and 20% of the prescribed dose; another one (patient X) was operated on for a bowel injury 3 years after postoperative irradiation that delivered 42 Gy. DISCUSSION Surgery So far surgery has remained the only way of curing a child with an Hb or an HcC. The most extensive survey, largely based on surgery, was published in 1975 by Exelby et al. (14). They analyzed 129 Hbs and 98 HcCs. Ten percent of the patients had metastatic disease at the time of diagnosis. One third involved both lobes. Eighty-six of 129 Hbs (67%) and $$ HcCs (34%) underwent resection: perioperative mortality was 11%. Forty-five of 129 Hbs (35%) and $$HcCs (12%) remained alive. No patient with unresectable tumor survived. Forty-five of 78 Hbs (58%) that benefited from complete resection survived. Equivalent survival rates have been brought out by Lack and colleagues ($$,45%) (27), Clatworthy et al. ($) (8), Mahour et al. (37.5%) (28) and Campbell (less than 40%) (7). Although surgery has greatly benefited from recent technical innovations like vascular reconstruction ( 18), ultrasonic aspiration, or dissection under vascular exclu-
sion (19), a complete resection may still not be feasible in case of involvement of both lobes, of the porta hepatis, of the vena cava, or the supra hepatic veins. Orthotopic liver transplantation can represent an alternative in highly selected situations, like in the fibrolamellar variant of HcC or in small tumors arising in a liver affected by a nonmalignant disease. The experience in this pediatric field remains almost anecdotal, but indicates that local recurrences have not been circumvented (5). Chemotherapy The outcome of resected and non-resected patients is severely compromised both by local evolution and distant metastases (mainly pulmonary) (26), so that 35% at most eventually survive (see above). These results have paved new avenues for chemotherapy. Preoperative chemotherapy has the reputation of greatly facilitating the surgical act: the child comes to surgery in much better condition, with a smaller tumor, generally presenting with a pseudo capsule that facilitates a closer dissection. The most effective drug-combinations include ADR and CDDP ( 12, 17,3 1,32). Of 20 children managed with preoperative chemotherapy by our group and evaluated in 1985 (24) 15 (70%) presented a good tumor response. Similar observations have been reported by Mahour et al. (28), Weinblatt and colleagues (36), Quinn et al. (32) and Andrassy et al. (2) in tumors frequently inoperable at initial laparotomy. In the current series, only three of our 13 patients without metastatic disease (23%), remained inoperable following chemotherapy. On the other hand, there is some evidence that postoperative chemotherapy improves long-term survival ( 13, 16, 20), even after complete surgical resection of disease ( 13, 17). Moreover, rare patients with pulmonary metastases have enjoyed long-term survival after chemotherapy + wedge resection (15). Radiation therapy Radiotherapy has gained limited acceptance in the therapeutic armamentarium. For example, in a survey conducted by the International Society of Pediatric Oncology (SIOP), they could only collect 36 cases including the 15 of the Villejuif series presented here (22). It is generally accepted that normal liver tolerance is not compatible with the curative doses required when large volumes are involved (37): severe hepatic toxicity has been described at doses above 25 Gy to the whole liver in adults
RT in the management of childhood hepatoblastomas and hepatocellular carcinomas 0 J.-L.
(25). There have been also severe complications with 18 Gy whole hepatic irradiations followed by a 12 Gy boost to one-third of the gland (3). Furthermore, PMLTs have gained a reputation of radioresistance from the poor efficacy reported in the adult hepatoma (9). But several series mention the use of radiotherapy. Radiotherapy has been used with various dose, fraction, and volume schemes and is almost always combined with chemotherapy. Such disparate literature cannot clarify the role of irradiation. The available retrospective reviews, published in the medical literature, do not offer a clear cut indication for preoperative irradiation. Some of the literature might be construed as arguing for the use of total or sub-total hepatic irradiation on the order of 10 to 40 Gy. Most of them indicate a relative responsiveness of Hbs (13, 16, 28, 33) and a constant lack of response of HcCs (9, 14, 27): Exelby et al.‘s survey ( 14) reports that 32/ 129 Hbs and 35/98 HcCs received radiotherapy. Doses ranged from 10 to 40 Gy in the liver, lo-40 Gy in the whole abdomen, and 7-20 Gy in the lungs. Using radiotherapy + the VAC regimen, only three inoperable Hbs became operable. The three of them enjoyed long-term survival. None of the inoperable children were cured. None of the HcCs responded to radiation, in spite of doses up to 40 Gy. Clatworthy et al. (8) reported that only one of seven children with inoperable Hbs responded to 30 Gy alone. In Shafer’s experience, three Hbs became operable following concomitant administration of ACD and 12-30 Gy irradiation (33). Tumor shrinkage was equal or superior to 50%. Two of them survived 8 and 13 months following surgery. Regarding progressive disease under the most effective drugs, our own data clearly indicate that this is almost a hopeless situation, granting pilot investigations: O/3 children survived more than a year without any clear palliative benefit from radiation. But in an additional patient (patient I), we elected to test a sequential combination of 5 FU + CDDP and low dose irradiation. An experimental study was designed by the Clermont-Ferrand group in France (4), in which the well-known radiosensitization of these drugs was optimized, administering 5 FU prior and CDDP upon radiation. The documented complete necrosis and durable remission 68 months post irradiation and 76 months since diagnosis of this child following 4 cycles of 6 Gy, substantiates other promising clinical observations (10, 11). Alternative experiments could be the intra-arterial administration of chemotherapy (2 l), of lymphokine-activated killer cells (29), or the use of Iodine 13 1 anti-ferritin (30). Postoperative residual tumor constitutes, in our protocol, an indication for irradiation: of our eight patients so treated, six (75%) are alive and free of disease 4 to 83 months post treatment (mean 44.6). One of our patients suffered a local failure at 83 months post treatment and one failed locally and in the lungs at 5 months post treatment. We acknowledge that our experience on minimal residues treated with irradiation, although very attractive,
HABRAND
et al.
529
has not yet been long enough and suffers the lack of comparison with a group treated without it. No convincing evidence either shows up from the literature. On one hand, some authors have reported a poor outcome after incomplete resection followed by no adjuvant therapy: None of 78 children survived in the Exelby et aZ.‘s survey, which mixed incomplete excisions and simple biopsies. On the other hand, no clear difference in survival between children completely and incompletely resected was established on 101 Hbs by the “committee of malignant tumors of the Japanese Society of Pediatric Surgeons,” for the subgroup with one segment involvement (Stage I: 60% both groups); but a definite difference was recorded for higher stages. Some of these patients also received various chemotherapeutic drugs (23). Two authors have reported on the outcome of microscopic residues after irradiation: Weinblatt described the cases of two children, one of whom received 15 Gy postoperatively and did well 2 years later and the other one who didn’t and failed locally and distantly within 3 years (36). Giacomantonio found only one out of six survivors who received post-operative irradiation and/or chemotherapy; this patient had been submitted to both agents (20). Only one multi-institutional study conducted by the CCSG-POG group in the United States introduced postoperative radiotherapy in a non-randomized prospective fashion in case of microscopic residues, at the dose of 45 Gy limited to the area at risks plus a minimum 2 cm margin ( 1, 13). Updated in 1988, 3-year patients free survival was 60% and slightly inferior to that of completely resected ones (67%) (1). Advanced cases received 30 Gy to the whole liver if they didn’t respond to two cycles of a quadruple chemotherapy (vincristine, cyclophosphamide, doxorubicin, bleomycin). Poor results (22%) were observed in such situations. As far as technical considerations, our group has also adapted the target volume to the residual and to the anatomical structure concerned as defined by surgical report, clips as well as post-operative imagings. Only one patient (patient IV) required whole liver coverage due to multifocal microscopic residues. With the limitations mentioned above, it appears that a durable control was obtained by 25-45 Gy in microscopic targets of 4/5 Hbs and by 3 l-45 Gy in gross residual disease, not exceeding 2 cm, of 4/5 Hbs; no failure was observed in four patients after 40-45 Gy. Local control was not achieved by 40 Gy in microscopic residues of an HcC (patient IV). Massive progressive disease (three cases) was not controlled in the range 20-40 Gy delivered to large volumes. Current IGR protocol Our current treatment strategy is to administer preoperative chemotherapy to all patients. We are evaluating epirubicin and carboplatin as first line chemotherapy. These agents have been selected for their likely efficacy
530
I. J. Radiation Oncology 0 Biology 0 Physics
and their more limited toxicity compared to other agents. Our current protocol administers a maximum of six courses to increase the probability of tumor shrinkage and resectability. A child with progressive disease is eligible for second line chemotherapy. Radiation therapy is indicated in case of incomplete resection according to surgical and pathological reports. Tailor-fit irradiation is applied with a doseescalation policy: 35-40 Gy in Hbs with postoperative microscopic residual tumor and 45 Gy to PMLTs with gross one. In HcC it looks reasonable to recommend 40 to 45 Gy in case of microscopic invasion and even more whether feasible in gross invasion, given their reputation of lower radiosensitivity.
CONCLUSION The role of radiation therapy in the management of Hb and HcC in children remains controversial. From the literature, surgery is still essential to cure the patients and has largely benefited from pre-operative polychemotherapy, which appears a safe and reliable procedure. As far
Volume 23, Number 3, 1992
as radiotherapy, there are clinical evidences that Hb is relatively radio-responsive to 20-40 Gy, and sometimes lower doses (lo-30 Gy) when combined with chemotherapeutic agents, like actinomycin D. HcC appears generally resistant to doses tolerable by the liver. From our experience based on 15 children and adolescents irradiated between 1978 and 1988 there is a possible role for focal irradiation combined with chemotherapy in the management of minimal residual Hbs (i.e., 2 cm) left behind by the surgeon: a minimum 25 Gy administered, in five daily sessions of 1.5 to 2 Gy, in case of limited microscopic invasion, and 35 Gy in case of macroscopic one, is well-tolerated and potentially curative in four of five patients. On the other hand, radiotherapy has little place but for palliation of symptoms in case of inoperable tumors refractory to chemotherapy. In any event, most reviews on the value of radiotherapy have little chances to be definitive: PMLTs are scarce tumors in the pediatric age and so the indications for radiotherapy are extremely rare and make virtually impossible any matched comparison between patients treated with and without, both in retrospect and prospectively.
REFERENCES 1. Ablin, A.; Krailo, M.; Hass, J.; Rogers, P.; Campbell,
2.
3.
4.
5.
6. 7.
8.
9.
J.; Tech, M.; Land, V.; Vietti, T.; Hammond, D. Hepatoblastoma and hepatocellular carcinoma in children: a report from the childrens’ cancer study group (CCG) and the pediatric oncology group (POG) (Abstr.). Med. Ped. Oncol. 16: 417; 1988. Andrassy, R. J.; Brennan, L. P.; Siegel, M. M.; Weitsman, J. J.; Siegel, S. E.; Stanley, P.; Mahour, G. M. Preoperative chemotherapy for hepatoblastoma in children: report of six cases. J. Ped. Surg. 15: 5 17-522; 1980. Austin-Seymour, M. M.; Chen, G. T. Y.; Castro, J. R.; Saunders, N. M.; Pitluck, S.; Woodruff, K. H.; Kessler, M. Dose volume histogram analysis of liver radiation tolerance. Int. J. Radiat. Oncol. Biol. Phys. 12: 31-35; 1986. Beaupain, R.; Dionet, C. Effects of combined treatments of cis-diamminedichloroplatinum (II), 5-fluorouracil, and Xrays on growth of human cancer nodules maintained in continuous organotypic culture. Cancer Res. 45: 3 150-3 154, 1985. Bismuth, H.; Ericzon, B. G.; Rolles, K.; Castaing, D.; Otte, J. B.; Ringe, B. Hepatic transplantation in Europe. First report of the European liver transplant registry. Lancet 2: 674-676; 1987. Boles Jr, E. T. Tumors of the abdomen in children. Pediatr. Clin. N. Am. 9: 467-484; 1962. Campbell, J. R. Malignant hepatic tumors: hepatoblastoma and hepatocellular carcinoma. In: Hays, D. M., ed. Pediatric surgery and oncology. Orlando: Grune and Stratton Pub.; 1986: 153-158. Clatworthy Jr., H. W.; Schiller, M.; Grosfeld, J. L. Primary liver tumors in infancy and childhood. 41 cases variously treated. Arch. Surg. 109: 143-146; 1974. Cochrane, A. M. G.; Murray-Lyon, I. M.; Brinkley, D. M.; Williams, R. Quadruple chemotherapy versus radiotherapy in treatment of primary hepatocellular carcinoma. Cancer 40: 609-614; 1977.
10. Dionet, Y.; Dumesnil, Y.; Achard, J. L.; Rozan, R.; Gauthier, J-P.; Feniere, J-P.; Chollet, P.; Plagne, R. Traitement des sarcomes des parties molles par une association de radiotherapie et de chimiotherapie. Presse Med. 11: 3351; 1982. 11. Dionet, C.; Rozan, R.; Achard, J. L.; Verrelle, P.; Dumesnil, Y.; Russier, M.; Lemesle, P.; Theron, H.; De Latour, M.; Mondie, J. M.; Lafaye, M.; Bolla, M.; Chollet, P. Sequential combination of 5Fluorouraci1, Cis-platinum and irradiation: 1. Advanced head and neck cancers. Radiat. Oncol. 11: 123-131; 1988. 12. Douglass, E. C.; Green, A. A.; Wrenn, E.; Champion, J.; Shipp, M.; Pratt, C. B. Effective cisplatin (DDP) based chemotherapy in the treatment of hepatoblastoma. Med. Ped. Oncol. 13: 187-190; 1985. 13. Evans, A. E.; Land, V. J.; Newton, W. A.; Randolph, J. G.; Sather, H. N.; Tell?, M. Combination chemotherapy (vinctistine, adriamycin, cyclophosphamide and 5-fluorouracil) in the treatment of children with malignant hepatoma. Cancer 50: 821-826; 1982. 14. Exelby, P. R.; Filler, R. M.; Grosfeld, J. L. Liver tumors in children in the particular reference to hepatoblastoma and hepatocellular carcinoma: American Academy of pediatrics surgical section survey. 1974. J. Ped. Surg. 10: 329-337; 1975. 15. Feusner, J.; Beach, B.; O’Leary, M.; Free, E.; Johnson, L. M.; Betts, J. Pulmonary metastatic-disease does not preclude survival in children with hepatoblastoma (Abstr.). Proc. ASCO. 212: 833; 1986. 16. Filler, R. M.; Hagen, J. Liver tumors. Surg. Clin. N. Am. 61: 1209-1217; 1981. 17. Forouhar, F. A.; Quinn, J. J.; Looke, R.; Foster, J. H. The effect of chemotherapy in hepatoblastoma. Arch. Pathol. Lab. Med. 108: 311-314; 1984. 18. Fortner, J. G.; Kinne, D. W.; Kim, D. K.; Castro, E. B.;
RT in the management of childhood hepatoblastomas and hepatocellular carcinomas 0 J.-L. HABRAND
Shiu, M. H.; Beattie, E. J. Vascular problems in upper abdominal cancer surgery. Arch. Surg. 109: 148-153; 1974.
19. Gauthier, F.; Saliou, C.; VaIayer, J.; Montupet, P. Surgery of hepatoblastoma and hepatocarcinoma in children in the era of preoperative chemotherapy. Current progress and limitations. Chir. Pediatr. 29: 307-312; 1988. 20. Giacomantonio, M.; Ein, S. H.; Mancer, K.; Stephens, C. A. Thirty years of experience with pediatric primary malignant liver tumors. J. Ped. Surg. 19: 523-526; 1984. 21. Golladay, E. S.; Mollitt, D. L.; Osteen, P. K.; Lang, D. H.; Neuberg, B. R.; KIetzel, M. Conversion to resectability by intraarterial infusion chemotherapy after failure of systemic chemotherapy. J. Paediatr. Surg. 20: 715-7 17; 1985. 22. Habrand, J. L.; Pritchard, J. Letter to the editor: role of radiotherapy in hepatoblastoma and hepatocellular carcinoma in children and adolescents: results of a survey conducted by the SIOP liver tumour study group. Med. Pediatr. Oncol. 19: 208; 1991. 23. Hata, Y. The clinical features and prognosis of hepatoblastoma: follow-up studies done on pediatric tumors enrolled in the Japanese Pediatric Tumor Registry between 197 1 and 1980. Part I. Jpn. J. Surg. 20: 492-502; 1990. 24. Kalifa, C.; Hartmann, 0.; Flamant, F.; Geoffray, A.: Lemerle, J. An attempt to improve the prognosis of hepatoblastoma (Hb) with chemotherapy (CT) (Abstr.). Proc. XVIIth SIOP Meeting, Venice, 30 Sept.-4 Oct., 1985. 25. Kaplan, H. S.; Bagshaw, M. A. Radiation hepatitis: possible prevention by combined isotopic and external radiation therapy. Radiology 91: 12 14-1220; 1968. 26. Lack, E. E.; Neave, C.; Vawter, G. F. Hepatoblastoma. A clinical and pathologic study of 54 cases. Am. J. Surg. Pathol. 6: 693-705; 1982. 27. Lack, E. E.; Neave, C.; Vawter, G. F. Hepatocellular carcinoma. Review of 32 cases in childhood and adolescence. Cancer 52: 1510-1515; 1983. 28. Mahour, G. H.; Wogu, G. H.; Siegel, S. E.; Isaacs, H. Im-
29.
30.
31.
32.
33.
34.
35. 36.
37.
etal.
531
proved survival in infants and children with primary malignant liver tumors. Am. J. Surg. 146: 236-240; 1983. Okuno, K.; Takagi, H.; Nakamura, T. ; Nakamura, Y. ; Iwasa, Z.; Yasutomi, M. Treatment for unresectable hepatoma via selective hepatic arterial infusion of lymphokine activated killer cells generated from autologous spleen cells. Cancer 58: 1001-1006; 1986. Order, E. S.; Stillwagon, G. B.; Klein, J. L.; Leichner, P. K.; Seigelman, S. S.; Fishman, E. K.; Ettinger, D. S.; Haulk, T.; Kopher, K.; Finney, K.; Surdyke, M.; Self, S.; Leibel, S. Iodine 13 1 antiferritin, a new treatment modality in hepatoma: a radiation therapy oncology group study. J. Clin. Oncol. 3: 1573-1582; 1985. Pritchard, J.; Plowman, P. N.; Broadbent, V. A.; Spitz, L.; Howard, E. R.; Mowat, A. P.; Mieli-Vergani, G. High rate of complete surgical resection (CSR) in advanced hepatoblastoma (HBL) after Cisplatinum-Doxorubicin (“Plado”) chemotherapy (Abstr.). Med. Ped. Oncol. 16: 417; 1988. Quinn, J. J.; Altman, A. J.; Robinson, H. T.; Cooke, R. W.; Hight, D. W.; Foster, J. H. Adriamycin and cisplatin for hepatoblastoma. Cancer 56: 1926-1929; 1985. Shafer, A. D.; Selinkoff, P. M. Preoperative irradiation and chemotherapy for initially unresectable hepatoblastoma. J. Pediatr. Surg. 12: 1001-1006; 1977. Tefft, M.; Mitus, A.; Das, L.; Vawter, G. F.; Filler, R. M. Irradiation of the liver in children and in the infant: review of experience in the acute and chronic phases. Am. J. Roentgen. 108: 365-385; 1970. Weinberg, A. G.; Finegord, M. J. Primary hepatic tumors of childhood. Hum. Pathol. 14: 512-537; 1983. Weinblatt, M. E.; Siegel, S. E.; Siegel, M. M.; Stanley, P.; Weitzman, J. J. Preoperative chemotherapy for unresectable primary hepatic malignancies in children. Cancer 50: 106 I1064; 1982. Weinel, P.; Freddara, R.; Burger, D. Hepatoblastoma and hepatocellular carcinoma. In: Monogram paediatric, Vol. 18. Basel: Karger Pub.; 1986: 359-367.
In1 J Radiarron Oncology BID/. Phys Vol. 23. PP. 525-531 Printed in the U.S.A. All rights reserved.
??Clinical Original Contribution
IS THERE A PLACE HEPATOBLASTOMAS
FOR RADIATION THERAPY IN THE MANAGEMENT OF AND HEPATOCELLULAR CARCINOMAS IN CHILDREN?
J.-L. HABRAND, M.D.,’ D. NEHME, M.D.,’ C. KALIFA, M.D.,2 F. GAUTHIER, M.D.,4 M. GRUNER, M.D.,’ D. SARRAZIN, M.D.,’ M.-J. TERRIER-LACOMBE, M.D.3 AND J. LEMERLE, M.D.2 ‘Pediatric Unit, Department of Radiation Medicine, ‘Department of Pediatrics, ‘Department of Pathology, Institut GustaveRoussy, 94805 Villejuif Cedex; 4Department of Pediatric Surgery, Hopital BicCtre, 94275 Le Kremlin-Bicetre Cedex; and ‘Department of Pediatric Surgery, Hopital Trousseau, 75571 Paris Cedex, France
From May 1978 to August 1988, 15 childrenwith a primarymalignantliver tumor receivedradiationtherapy as part of their managementat the Institnt Gustave-Roussy. Age ranged from 4 months to 13 years. The male to female ratio was 1.5. Eleven patients had a histologically proven hepatoblastoma, two a hepatocellular carcinoma, and histology was not documented in two. Resection of the primary liver tumor was performed in nine cases, and all patients also received sequential chemotherapy, generally preoperative and alternating vincrlstine, doxorubicin, cyclophosphamide with vincristine, cyclophosphamide, and cis-platinum. Radiotherapy was performed postoperatively in eight incompletely resected patients. Six of eight are alive and free of disease 4-83 months following treatment (median 39 months) and 11-98 months since diagnosis (median 45 months). All but one were treated to limited fields to a total dose of 25-45 Gy (median 40 Gy). One patient became resectable by a combination of 24 Gy to the whole liver and concomitant 5FU and Cis-Platinum and remains with no evidence of disease 68 months following radiation therapy. Of four unresectable primaries, only one was controlled by radiotherapy. Neither of two children with pulmonary metastases were controlled by whole lung irradiation to a dose of 18 and 20 Gy, respectively, and one still remains stable 41 months after resection of a residual metastatic nodule. Neither of two hepatocellular carcinomas were controlled by doses up to 40 Gy. This small series suggests that in hepatoblastoma, radiotherapy to a total of 2545 Gy fractionated doses, combined with chemotherapy, can play a role in selected inoperable children and also in those with minimal postoperative residues below 2 cm. It also indicates that in bepatocellular carcinoma, radiotherapy is ineffective in this dose-range. Radiotherapy, Children, Hepatoblastoma, Hepatocellular carcinoma. INTRODUCTION Primary malignant liver tumors (PMLTs) are extremely rare in children and account for approximately 2% of all malignant tumors observed (6, 16). As intraabdominal neoplasms, they occur less frequently than nephroblastomas and neuroblastomas (8). Most of them are hepatoblastomas (Hbs) and less frequently hepatocellular carcinomas (HcCs), a form similar to that observed in adults. In one series, they represented 52 and 39% of PMLTs, respectively (14). The remaining 9% of cases consisted of unusual histologic types such as rhabdomyosarcoma and angiosarcoma. Although both Hb and HcC are of similar origin, various epidemiological factors differentiate them (7). Microscopically the Hb may be subclassified as pure epithelial Hb and mixed Hb, which also contains mesenchymal tissue (osteoid, cartilaginous, or rhabdomyo-
b&tic). The epithelial component is represented by either fetal or embryonal cells. It can be separated into predominantly fetal or embryo& types, depending which pattern occupies more than 50% of cross-sections. An additional anaplastic form is composed mainly of poorly differentiated small to medium size cells without any glandular structure visible. With respect to outcome, the HcC is generally regarded as the worst form (13) with the exception of the fibrolamellar form, which carries an excellent prognosis (27). As far as Hb is concerned, it is generally regarded that the fetal type carries a more favorable prognosis and that the anaplastic form is invariably lethal. For decades, surgical resection has been the mainstay of treatment. At the time of diagnosis only 1560% tumors are operable (8, 14, 20, 26, 27, 28). But even complete surgical resection achieves long-term survival only in 30-
Reprint requests to: J. L. Habrand, M.D., Department of Radiation Medicine, Institut Gustave-Roussy, 39, rue Camille Desmoulins, 94805 Villejuif Cedex, France.
Acknowledgements-We are grateful to Catherine Joelle GuCry for skillful technical assistance. Accepted for publication 12 December 199 1. 525
Loge and
526
I. J. Radiation Oncology 0 Biology 0 Physics
60% (7,8, 14,26,28). Most relapses occur within the first year following diagnosis (14). These results have stimulated the interest for chemotherapy, especially in inoperable patients. Various series (2,32,35) have reported that most patients had a dramatic tumor shrinkage following polychemotherapy. On the other hand, very sparse data are available on the efficacy and role of radiation therapy, but it is generally regarded as an “ancillary” treatment ( 14). These issues were addressed in this review of 15 children irradiated at the Institut Gustave-Roussy (IGR) of Villejuif between 1978 and 1988. METHODS
AND MATERIALS
Patients From May 1978 to August 1988, 15 children with a diagnosis of Hb or HcC were treated at the IGR with radiation therapy. All of them also received chemotherapy k surgery. The group requiring radiotherapy (see treatment protocol below) represented 22% (g) of all children managed at our institution with this type of tumor in the same interval. Age ranged from 4 months to 13 years with 3 patients being below the age of 3 years. The male to female ratio was 1.5 (E). All patients presented with a rapidly growing abdominal mass associated with an elevated alpha-feto-protein (AFP) serum level: it ranged from 0.4 to 370 X lo3 rig/ml (normal level: < 30 ng/ml, above the age of 3 months). The imaging work-up included abdominal ultrasound in 12 patients, Tc 99 liver scan in 4, CT scan in 6, preoperative angiography in 12, and chest x-ray in 15. The right hepatic lobe was involved in all patients and was associated with a left involvement in eight. The histology was available in 13 of them (surgical specimen: 12; biopsy alone: 1) and was reviewed for the purpose of this study. Eleven children had an Hb, purely or predominantly fetal in seven, embryonal in two, and mixed in two. Among them, one patient with a fetal type and one with an embryonal one had an osteoid component, whereas one patient with a mixed form had both osteoid and anaplastic components. Two children presented with an HcC. In two other inoperable cases, the diagnosis of malignancy was made exclusively clinically and on a persistently high level of AFP. Treatment protocol The primary emphasis of therapy is gross total surgical resection of tumor. We sought to achieve this goal short of total hepatectomy. To increase the chance of tumors being resectable, 12 of 13 cases without metastatic disease received initial cytoreductive chemotherapy. It alternated usually vincristine (VCR = 1.5 mg/m2, day l), Doxorubicin (ADR = 40 mg/m2, day l-2), and Cyclophos-
Volume 23, Number 3, 1992
phamide (CYC = 600 mg/m2, day 2) with a combination of Vincristine (1.5 mg/m2, day 22), Cyclophosphamide (600 mg/m2, day 22), and Cis-Platinum (CDDP = 100 mg/m2, day 23). Its efficacy was assessed clinically, radiologically, and on the AFP serum profile. This protocol was conducted until maximum tumor reduction. If the tumor was still deemed inoperable, a salvage therapy, generally Actinomycin D (ACD = 0.45 mg/m2, per day for 5 days), was instituted. One patient received the IVA regimen (ifosfamide, vincristine, actinomycin D) and carboplatin. Curative radiotherapy was applied to patients with grossly or microscopically incomplete resection. Radiotherapy was started 4 to 6 weeks following surgery so as not to interfere with hepatic regeneration (34). A dose of 25 to 45 Gy was administered, targeted to the residual post-operative disease only, in 5 daily sessions of 1.5 to 2 Gy per fraction. Multiple field arrangements were employed to avoid the right kidney as much as possible. Co 60 or high energy x-rays from linear accelerators were used in all cases. Palliative irradiation of the entire liver or whole lung coverage in pulmonary metastases were also contemplated in selected situations. Postoperative chemotherapy was given in 8 of 9 resected patients, combining CDDP, VCR, and CYC. Overall treatment duration totaled 1 year or so. RESULTS At the time of evaluation, E (66%) patients were alive and free of disease. Total follow-up since initial symptoms ranged from 4 to 98 months (median 37) and disease-free survival upon completion of radiotherapy, from 0 to 83 months (median 23). Actuarial 3 and 5 year-survival were, respectively, 65 and 53% and disease-free survival, 57% and 57%. Resected patients (9 cases) (patientsI-IX, Table I) Eight out of nine children received postoperative irradiation after incomplete resection (following preoperative chemotherapy in seven). Four of them had gross residuals (Group 1) mostly encased along the inferior vena cava, the portal, or the supra hepatic veins. Tumors were barely detected by post-operative imaging techniques and never exceeded more than 2 cm in greatest dimension. Four had microscopic positive margins (Group 2). Pathological examination of the surgical specimen in the seven patients managed with preoperative chemotherapy, displayed a considerable amount of necrosis, attributed to the efficacy of chemotherapy. But tumor foci were still visible. Total dose ranged from 25 Gy to 45 Gy (median 40 Gy) with the largest dose given to gross residues. The whole liver was irradiated only in one case due to disseminated tumor foci.
RT
in the management of childhood hepatoblastomas and hepatocellular carcinomas 0 J.-L. HABRAND etal.
Table 1. Children treated with resection, radiotherapy and chemotherapy:
521
Status post initial symptoms
Radiotherapy Status (in months since Dg)
Patient (sex, age in months) pathology
Chemotherapy pre/ post op.
Timing
Target
Technique (TD in Gy)
I (F, 120) Fetal Hb + 0 II (M, 8) Fetal Hb III (M, 15) Fetal Hb IV (F, 156) HcC V (M, 4) Pure fetal Hb VI (F, 9) Embryo HB + 0 VII (F, 5) Pure fetal Hb VIII (M, 24) Fetal +embryoHb+O+A IX (M, 168) Fetal Hb
VADC, CDDP +/+ VAC, CDDP +/+ VADC, 5FU -/+ VADC, CDDP +/+ Id VADC, CDDP +/VADC, CDDP +/+
Pre op. Post op. Id Id Id Id Id
Macro Micro Id Id Id Macro Id
Whole, 25 + 5FU, CDDP Focal, 25 Focal, 30 Whole, 40 Focal. 45 Id Focal, 35
NED, 76 NED, 43 NED, 98 D, 48, LF NED, 22 NED, 81 NED, 51
Id Id
Id Id
Id Id
Focal, 42 Focal, 40
NED, 37 D, ll,LF+PM
TD = Total dose; NED = No evidence of disease; D = Dead; LF = Local failure; PM = Pulmonary metastases; Id = Identical to the above; Fetal Hb = Hb, fetal type predominantly (50%); Embryonal Hb = Hb, embryonal type predominantly (50%); Fetal + embryonal Hb = Mixed fetal (50%) and embryonal (50%) type of Hb, 0 = Osteoid component; A = Focal anaplasia; Micro = Microscopic invasion; Macro = Macroscopic invasion; Dg = Diagnosis; other abbreviations: see text.
Six of eight children (75%) have remained free of disease 4 to 83 months following completion of radiotherapy (median 32 months). One patient in group 2, with an HcC failed locally and one in group 1 both locally and distantly. One out of nine children (patient I) received preoperative radiotherapy potentiated by cytotoxic agents used as radiosensitizers, 5 fluorouracil(5 FU), and CDDP: this 10 year-old girl presented with a massive right Hb, 12 X 15 cm large. A partial clinical and biological response had been obtained after four cycles of chemotherapy (VCR, ADR, CYC, CDDP). Nevertheless, at laparotomy, extensive involvement of both lobes and hepatic convergence precluded any resection. The patient was started on a pilot study combining each month, one week of radiotherapy (3 Gy X 2, day 3 and 4), along with 5 PU (350 mg/m2, day 2, 3,4, 6) + CDDP (15 mg/m2, days 1, 5, 7). After 4 cycles of 6 Gy with 5 FU and CDDP, she was operated on again. A complete resection was made possible at that time and the tumor was completely necrotic at microscopic examination. This patient is still alive and doing well, 68 months post irradiation. None of the untoward late side effects that have been reported after the use of high doses per fraction, have been recorded at the last follow-up.
Unresectablepatients (5 cases) (patients X-XIII. Table 2) Radiotherapy was tested in three patients with massive tumors that progressed under chemotherapy (patients XIXIII). Doses of 30 Gy in large volumes + 15 Gy boost in one of them, were administered. All died at 4, 8, and 12 months, respectively, with only one transient tumor stabilization. Another 10 month-old-child (patient X), presented with a limited recurrence two months after complete resection. A new resection was ruled out given the tumor’s vascular connection and he was managed by chemotherapy (Ifosfamide, Doxorubicin, and Carboplatin) for 9 months. Radiotherapy was targeted at small ultrasonically defined abnormalities. A total dose of 45 Gy was delivered. This patient has not failed with 15 months follow-up.
Metastatic patients (2 cases) (patients XIV, XV, Table 3) Two children were irradiated for pulmonary metastases. One of them received 20 Gy with correction for air-transmission in the lungs using Co60, and the second one 18 Gy with 18 MV x-rays, concomitant with Actinomycin D. In the first child, radiation therapy was administered
Table 2. Unresectable children treated with chemotherapy and radiotherapy Radiotherapy Patient (sex, age in months) pathology
Chemotherapy
Target
Technique (TD in Gy)
X (M, 10) Fetal + embryo Hb XI (F, 30) XII (M, 36) HcC XIII (F, 102) -
VADC, CDDP IVA, Carboplatin VADC Id VADC, CDDP
Macro Massive Id Id
Focal, 45 Focal, 30 Whole, 30 Whole, 30 + boost, 15
Abbreviations:
see Table 1.
status (in months since Dg) NED, 33 D, 12, LF D, 8, LF D. 4. LF
528
I. J. Radiation Oncology 0 Biology0 Physics
Volume 23, Number 3, 1992
Table 3. Radiation therapy of pulmonary
metastases
Radiotherapy Patient (sex, age in months) pathology
Chemotherapy
Target
Technique (Td in Gy)
Salvage surgery
XIV (M, 13) Fetal Hb XV (M, 114) Embryo Hb + 0
VADC, 5FU VADC, CDDP
Macro Micro
Whole lungs, 20 Whole lungs, 18
+ -
Abbreviations:
Status (in months since Dg) NED, 58 Alive, 20, PM
see Table 1.
for multiple pulmonary metastases, discovered shortly after an extended hepatectomy, and which progressed under chemotherapy. After a 4 month-free interval, multiple nodules were again visible and motivated a triple wedge resection, through a double thoracotomy. Pathologically, only one nodule still contained tumor. This patient was still doing well 4 1 months later. In the second child, multiple pulmonary metastases were also discovered under maintenance chemotherapy, and completely resected through a double thoracotomy. Although radiation therapy was only administered as a consolidation, this patient failed in the lungs 8 months later and was placed on a phase 1 interleukin 2 trial. None of both patients have relapsed at the site of the primary. Late toxicity (patients VI and X) Two patients presented a late complication in relation with radiotherapy. One (patient VI) had a scoliosis 6 years following 45 Gy delivered to a wide clipped area through AP-PA fields, abutting but not covering the entire spine, in order to spare as much normal liver as possible. Consequently, the right part of the vertebral bodies from T8 to L2 received a dose gradient between 50 and 20% of the prescribed dose; another one (patient X) was operated on for a bowel injury 3 years after postoperative irradiation that delivered 42 Gy. DISCUSSION Surgery So far surgery has remained the only way of curing a child with an Hb or an HcC. The most extensive survey, largely based on surgery, was published in 1975 by Exelby et al. (14). They analyzed 129 Hbs and 98 HcCs. Ten percent of the patients had metastatic disease at the time of diagnosis. One third involved both lobes. Eighty-six of 129 Hbs (67%) and $$ HcCs (34%) underwent resection: perioperative mortality was 11%. Forty-five of 129 Hbs (35%) and $$HcCs (12%) remained alive. No patient with unresectable tumor survived. Forty-five of 78 Hbs (58%) that benefited from complete resection survived. Equivalent survival rates have been brought out by Lack and colleagues ($$,45%) (27), Clatworthy et al. ($) (8), Mahour et al. (37.5%) (28) and Campbell (less than 40%) (7). Although surgery has greatly benefited from recent technical innovations like vascular reconstruction ( 18), ultrasonic aspiration, or dissection under vascular exclu-
sion (19), a complete resection may still not be feasible in case of involvement of both lobes, of the porta hepatis, of the vena cava, or the supra hepatic veins. Orthotopic liver transplantation can represent an alternative in highly selected situations, like in the fibrolamellar variant of HcC or in small tumors arising in a liver affected by a nonmalignant disease. The experience in this pediatric field remains almost anecdotal, but indicates that local recurrences have not been circumvented (5). Chemotherapy The outcome of resected and non-resected patients is severely compromised both by local evolution and distant metastases (mainly pulmonary) (26), so that 35% at most eventually survive (see above). These results have paved new avenues for chemotherapy. Preoperative chemotherapy has the reputation of greatly facilitating the surgical act: the child comes to surgery in much better condition, with a smaller tumor, generally presenting with a pseudo capsule that facilitates a closer dissection. The most effective drug-combinations include ADR and CDDP ( 12, 17,3 1,32). Of 20 children managed with preoperative chemotherapy by our group and evaluated in 1985 (24) 15 (70%) presented a good tumor response. Similar observations have been reported by Mahour et al. (28), Weinblatt and colleagues (36), Quinn et al. (32) and Andrassy et al. (2) in tumors frequently inoperable at initial laparotomy. In the current series, only three of our 13 patients without metastatic disease (23%), remained inoperable following chemotherapy. On the other hand, there is some evidence that postoperative chemotherapy improves long-term survival ( 13, 16, 20), even after complete surgical resection of disease ( 13, 17). Moreover, rare patients with pulmonary metastases have enjoyed long-term survival after chemotherapy + wedge resection (15). Radiation therapy Radiotherapy has gained limited acceptance in the therapeutic armamentarium. For example, in a survey conducted by the International Society of Pediatric Oncology (SIOP), they could only collect 36 cases including the 15 of the Villejuif series presented here (22). It is generally accepted that normal liver tolerance is not compatible with the curative doses required when large volumes are involved (37): severe hepatic toxicity has been described at doses above 25 Gy to the whole liver in adults
RT in the management of childhood hepatoblastomas and hepatocellular carcinomas 0 J.-L.
(25). There have been also severe complications with 18 Gy whole hepatic irradiations followed by a 12 Gy boost to one-third of the gland (3). Furthermore, PMLTs have gained a reputation of radioresistance from the poor efficacy reported in the adult hepatoma (9). But several series mention the use of radiotherapy. Radiotherapy has been used with various dose, fraction, and volume schemes and is almost always combined with chemotherapy. Such disparate literature cannot clarify the role of irradiation. The available retrospective reviews, published in the medical literature, do not offer a clear cut indication for preoperative irradiation. Some of the literature might be construed as arguing for the use of total or sub-total hepatic irradiation on the order of 10 to 40 Gy. Most of them indicate a relative responsiveness of Hbs (13, 16, 28, 33) and a constant lack of response of HcCs (9, 14, 27): Exelby et al.‘s survey ( 14) reports that 32/ 129 Hbs and 35/98 HcCs received radiotherapy. Doses ranged from 10 to 40 Gy in the liver, lo-40 Gy in the whole abdomen, and 7-20 Gy in the lungs. Using radiotherapy + the VAC regimen, only three inoperable Hbs became operable. The three of them enjoyed long-term survival. None of the inoperable children were cured. None of the HcCs responded to radiation, in spite of doses up to 40 Gy. Clatworthy et al. (8) reported that only one of seven children with inoperable Hbs responded to 30 Gy alone. In Shafer’s experience, three Hbs became operable following concomitant administration of ACD and 12-30 Gy irradiation (33). Tumor shrinkage was equal or superior to 50%. Two of them survived 8 and 13 months following surgery. Regarding progressive disease under the most effective drugs, our own data clearly indicate that this is almost a hopeless situation, granting pilot investigations: O/3 children survived more than a year without any clear palliative benefit from radiation. But in an additional patient (patient I), we elected to test a sequential combination of 5 FU + CDDP and low dose irradiation. An experimental study was designed by the Clermont-Ferrand group in France (4), in which the well-known radiosensitization of these drugs was optimized, administering 5 FU prior and CDDP upon radiation. The documented complete necrosis and durable remission 68 months post irradiation and 76 months since diagnosis of this child following 4 cycles of 6 Gy, substantiates other promising clinical observations (10, 11). Alternative experiments could be the intra-arterial administration of chemotherapy (2 l), of lymphokine-activated killer cells (29), or the use of Iodine 13 1 anti-ferritin (30). Postoperative residual tumor constitutes, in our protocol, an indication for irradiation: of our eight patients so treated, six (75%) are alive and free of disease 4 to 83 months post treatment (mean 44.6). One of our patients suffered a local failure at 83 months post treatment and one failed locally and in the lungs at 5 months post treatment. We acknowledge that our experience on minimal residues treated with irradiation, although very attractive,
HABRAND
et al.
529
has not yet been long enough and suffers the lack of comparison with a group treated without it. No convincing evidence either shows up from the literature. On one hand, some authors have reported a poor outcome after incomplete resection followed by no adjuvant therapy: None of 78 children survived in the Exelby et aZ.‘s survey, which mixed incomplete excisions and simple biopsies. On the other hand, no clear difference in survival between children completely and incompletely resected was established on 101 Hbs by the “committee of malignant tumors of the Japanese Society of Pediatric Surgeons,” for the subgroup with one segment involvement (Stage I: 60% both groups); but a definite difference was recorded for higher stages. Some of these patients also received various chemotherapeutic drugs (23). Two authors have reported on the outcome of microscopic residues after irradiation: Weinblatt described the cases of two children, one of whom received 15 Gy postoperatively and did well 2 years later and the other one who didn’t and failed locally and distantly within 3 years (36). Giacomantonio found only one out of six survivors who received post-operative irradiation and/or chemotherapy; this patient had been submitted to both agents (20). Only one multi-institutional study conducted by the CCSG-POG group in the United States introduced postoperative radiotherapy in a non-randomized prospective fashion in case of microscopic residues, at the dose of 45 Gy limited to the area at risks plus a minimum 2 cm margin ( 1, 13). Updated in 1988, 3-year patients free survival was 60% and slightly inferior to that of completely resected ones (67%) (1). Advanced cases received 30 Gy to the whole liver if they didn’t respond to two cycles of a quadruple chemotherapy (vincristine, cyclophosphamide, doxorubicin, bleomycin). Poor results (22%) were observed in such situations. As far as technical considerations, our group has also adapted the target volume to the residual and to the anatomical structure concerned as defined by surgical report, clips as well as post-operative imagings. Only one patient (patient IV) required whole liver coverage due to multifocal microscopic residues. With the limitations mentioned above, it appears that a durable control was obtained by 25-45 Gy in microscopic targets of 4/5 Hbs and by 3 l-45 Gy in gross residual disease, not exceeding 2 cm, of 4/5 Hbs; no failure was observed in four patients after 40-45 Gy. Local control was not achieved by 40 Gy in microscopic residues of an HcC (patient IV). Massive progressive disease (three cases) was not controlled in the range 20-40 Gy delivered to large volumes. Current IGR protocol Our current treatment strategy is to administer preoperative chemotherapy to all patients. We are evaluating epirubicin and carboplatin as first line chemotherapy. These agents have been selected for their likely efficacy
530
I. J. Radiation Oncology 0 Biology 0 Physics
and their more limited toxicity compared to other agents. Our current protocol administers a maximum of six courses to increase the probability of tumor shrinkage and resectability. A child with progressive disease is eligible for second line chemotherapy. Radiation therapy is indicated in case of incomplete resection according to surgical and pathological reports. Tailor-fit irradiation is applied with a doseescalation policy: 35-40 Gy in Hbs with postoperative microscopic residual tumor and 45 Gy to PMLTs with gross one. In HcC it looks reasonable to recommend 40 to 45 Gy in case of microscopic invasion and even more whether feasible in gross invasion, given their reputation of lower radiosensitivity.
CONCLUSION The role of radiation therapy in the management of Hb and HcC in children remains controversial. From the literature, surgery is still essential to cure the patients and has largely benefited from pre-operative polychemotherapy, which appears a safe and reliable procedure. As far
Volume 23, Number 3, 1992
as radiotherapy, there are clinical evidences that Hb is relatively radio-responsive to 20-40 Gy, and sometimes lower doses (lo-30 Gy) when combined with chemotherapeutic agents, like actinomycin D. HcC appears generally resistant to doses tolerable by the liver. From our experience based on 15 children and adolescents irradiated between 1978 and 1988 there is a possible role for focal irradiation combined with chemotherapy in the management of minimal residual Hbs (i.e., 2 cm) left behind by the surgeon: a minimum 25 Gy administered, in five daily sessions of 1.5 to 2 Gy, in case of limited microscopic invasion, and 35 Gy in case of macroscopic one, is well-tolerated and potentially curative in four of five patients. On the other hand, radiotherapy has little place but for palliation of symptoms in case of inoperable tumors refractory to chemotherapy. In any event, most reviews on the value of radiotherapy have little chances to be definitive: PMLTs are scarce tumors in the pediatric age and so the indications for radiotherapy are extremely rare and make virtually impossible any matched comparison between patients treated with and without, both in retrospect and prospectively.
REFERENCES 1. Ablin, A.; Krailo, M.; Hass, J.; Rogers, P.; Campbell,
2.
3.
4.
5.
6. 7.
8.
9.
J.; Tech, M.; Land, V.; Vietti, T.; Hammond, D. Hepatoblastoma and hepatocellular carcinoma in children: a report from the childrens’ cancer study group (CCG) and the pediatric oncology group (POG) (Abstr.). Med. Ped. Oncol. 16: 417; 1988. Andrassy, R. J.; Brennan, L. P.; Siegel, M. M.; Weitsman, J. J.; Siegel, S. E.; Stanley, P.; Mahour, G. M. Preoperative chemotherapy for hepatoblastoma in children: report of six cases. J. Ped. Surg. 15: 5 17-522; 1980. Austin-Seymour, M. M.; Chen, G. T. Y.; Castro, J. R.; Saunders, N. M.; Pitluck, S.; Woodruff, K. H.; Kessler, M. Dose volume histogram analysis of liver radiation tolerance. Int. J. Radiat. Oncol. Biol. Phys. 12: 31-35; 1986. Beaupain, R.; Dionet, C. Effects of combined treatments of cis-diamminedichloroplatinum (II), 5-fluorouracil, and Xrays on growth of human cancer nodules maintained in continuous organotypic culture. Cancer Res. 45: 3 150-3 154, 1985. Bismuth, H.; Ericzon, B. G.; Rolles, K.; Castaing, D.; Otte, J. B.; Ringe, B. Hepatic transplantation in Europe. First report of the European liver transplant registry. Lancet 2: 674-676; 1987. Boles Jr, E. T. Tumors of the abdomen in children. Pediatr. Clin. N. Am. 9: 467-484; 1962. Campbell, J. R. Malignant hepatic tumors: hepatoblastoma and hepatocellular carcinoma. In: Hays, D. M., ed. Pediatric surgery and oncology. Orlando: Grune and Stratton Pub.; 1986: 153-158. Clatworthy Jr., H. W.; Schiller, M.; Grosfeld, J. L. Primary liver tumors in infancy and childhood. 41 cases variously treated. Arch. Surg. 109: 143-146; 1974. Cochrane, A. M. G.; Murray-Lyon, I. M.; Brinkley, D. M.; Williams, R. Quadruple chemotherapy versus radiotherapy in treatment of primary hepatocellular carcinoma. Cancer 40: 609-614; 1977.
10. Dionet, Y.; Dumesnil, Y.; Achard, J. L.; Rozan, R.; Gauthier, J-P.; Feniere, J-P.; Chollet, P.; Plagne, R. Traitement des sarcomes des parties molles par une association de radiotherapie et de chimiotherapie. Presse Med. 11: 3351; 1982. 11. Dionet, C.; Rozan, R.; Achard, J. L.; Verrelle, P.; Dumesnil, Y.; Russier, M.; Lemesle, P.; Theron, H.; De Latour, M.; Mondie, J. M.; Lafaye, M.; Bolla, M.; Chollet, P. Sequential combination of 5Fluorouraci1, Cis-platinum and irradiation: 1. Advanced head and neck cancers. Radiat. Oncol. 11: 123-131; 1988. 12. Douglass, E. C.; Green, A. A.; Wrenn, E.; Champion, J.; Shipp, M.; Pratt, C. B. Effective cisplatin (DDP) based chemotherapy in the treatment of hepatoblastoma. Med. Ped. Oncol. 13: 187-190; 1985. 13. Evans, A. E.; Land, V. J.; Newton, W. A.; Randolph, J. G.; Sather, H. N.; Tell?, M. Combination chemotherapy (vinctistine, adriamycin, cyclophosphamide and 5-fluorouracil) in the treatment of children with malignant hepatoma. Cancer 50: 821-826; 1982. 14. Exelby, P. R.; Filler, R. M.; Grosfeld, J. L. Liver tumors in children in the particular reference to hepatoblastoma and hepatocellular carcinoma: American Academy of pediatrics surgical section survey. 1974. J. Ped. Surg. 10: 329-337; 1975. 15. Feusner, J.; Beach, B.; O’Leary, M.; Free, E.; Johnson, L. M.; Betts, J. Pulmonary metastatic-disease does not preclude survival in children with hepatoblastoma (Abstr.). Proc. ASCO. 212: 833; 1986. 16. Filler, R. M.; Hagen, J. Liver tumors. Surg. Clin. N. Am. 61: 1209-1217; 1981. 17. Forouhar, F. A.; Quinn, J. J.; Looke, R.; Foster, J. H. The effect of chemotherapy in hepatoblastoma. Arch. Pathol. Lab. Med. 108: 311-314; 1984. 18. Fortner, J. G.; Kinne, D. W.; Kim, D. K.; Castro, E. B.;
RT in the management of childhood hepatoblastomas and hepatocellular carcinomas 0 J.-L. HABRAND
Shiu, M. H.; Beattie, E. J. Vascular problems in upper abdominal cancer surgery. Arch. Surg. 109: 148-153; 1974.
19. Gauthier, F.; Saliou, C.; VaIayer, J.; Montupet, P. Surgery of hepatoblastoma and hepatocarcinoma in children in the era of preoperative chemotherapy. Current progress and limitations. Chir. Pediatr. 29: 307-312; 1988. 20. Giacomantonio, M.; Ein, S. H.; Mancer, K.; Stephens, C. A. Thirty years of experience with pediatric primary malignant liver tumors. J. Ped. Surg. 19: 523-526; 1984. 21. Golladay, E. S.; Mollitt, D. L.; Osteen, P. K.; Lang, D. H.; Neuberg, B. R.; KIetzel, M. Conversion to resectability by intraarterial infusion chemotherapy after failure of systemic chemotherapy. J. Paediatr. Surg. 20: 715-7 17; 1985. 22. Habrand, J. L.; Pritchard, J. Letter to the editor: role of radiotherapy in hepatoblastoma and hepatocellular carcinoma in children and adolescents: results of a survey conducted by the SIOP liver tumour study group. Med. Pediatr. Oncol. 19: 208; 1991. 23. Hata, Y. The clinical features and prognosis of hepatoblastoma: follow-up studies done on pediatric tumors enrolled in the Japanese Pediatric Tumor Registry between 197 1 and 1980. Part I. Jpn. J. Surg. 20: 492-502; 1990. 24. Kalifa, C.; Hartmann, 0.; Flamant, F.; Geoffray, A.: Lemerle, J. An attempt to improve the prognosis of hepatoblastoma (Hb) with chemotherapy (CT) (Abstr.). Proc. XVIIth SIOP Meeting, Venice, 30 Sept.-4 Oct., 1985. 25. Kaplan, H. S.; Bagshaw, M. A. Radiation hepatitis: possible prevention by combined isotopic and external radiation therapy. Radiology 91: 12 14-1220; 1968. 26. Lack, E. E.; Neave, C.; Vawter, G. F. Hepatoblastoma. A clinical and pathologic study of 54 cases. Am. J. Surg. Pathol. 6: 693-705; 1982. 27. Lack, E. E.; Neave, C.; Vawter, G. F. Hepatocellular carcinoma. Review of 32 cases in childhood and adolescence. Cancer 52: 1510-1515; 1983. 28. Mahour, G. H.; Wogu, G. H.; Siegel, S. E.; Isaacs, H. Im-
29.
30.
31.
32.
33.
34.
35. 36.
37.
etal.
531
proved survival in infants and children with primary malignant liver tumors. Am. J. Surg. 146: 236-240; 1983. Okuno, K.; Takagi, H.; Nakamura, T. ; Nakamura, Y. ; Iwasa, Z.; Yasutomi, M. Treatment for unresectable hepatoma via selective hepatic arterial infusion of lymphokine activated killer cells generated from autologous spleen cells. Cancer 58: 1001-1006; 1986. Order, E. S.; Stillwagon, G. B.; Klein, J. L.; Leichner, P. K.; Seigelman, S. S.; Fishman, E. K.; Ettinger, D. S.; Haulk, T.; Kopher, K.; Finney, K.; Surdyke, M.; Self, S.; Leibel, S. Iodine 13 1 antiferritin, a new treatment modality in hepatoma: a radiation therapy oncology group study. J. Clin. Oncol. 3: 1573-1582; 1985. Pritchard, J.; Plowman, P. N.; Broadbent, V. A.; Spitz, L.; Howard, E. R.; Mowat, A. P.; Mieli-Vergani, G. High rate of complete surgical resection (CSR) in advanced hepatoblastoma (HBL) after Cisplatinum-Doxorubicin (“Plado”) chemotherapy (Abstr.). Med. Ped. Oncol. 16: 417; 1988. Quinn, J. J.; Altman, A. J.; Robinson, H. T.; Cooke, R. W.; Hight, D. W.; Foster, J. H. Adriamycin and cisplatin for hepatoblastoma. Cancer 56: 1926-1929; 1985. Shafer, A. D.; Selinkoff, P. M. Preoperative irradiation and chemotherapy for initially unresectable hepatoblastoma. J. Pediatr. Surg. 12: 1001-1006; 1977. Tefft, M.; Mitus, A.; Das, L.; Vawter, G. F.; Filler, R. M. Irradiation of the liver in children and in the infant: review of experience in the acute and chronic phases. Am. J. Roentgen. 108: 365-385; 1970. Weinberg, A. G.; Finegord, M. J. Primary hepatic tumors of childhood. Hum. Pathol. 14: 512-537; 1983. Weinblatt, M. E.; Siegel, S. E.; Siegel, M. M.; Stanley, P.; Weitzman, J. J. Preoperative chemotherapy for unresectable primary hepatic malignancies in children. Cancer 50: 106 I1064; 1982. Weinel, P.; Freddara, R.; Burger, D. Hepatoblastoma and hepatocellular carcinoma. In: Monogram paediatric, Vol. 18. Basel: Karger Pub.; 1986: 359-367.
