a Two separate ascertainment procedures in Quebec b Four specific areas in North America c Versus rest of Canada a Two separate ascertainment procedures in Quebec b Four specific areas in North America c Versus rest of Canada despite major efforts, early compliance in their trial was low, less than 50 %, demonstrating how difficult obtaining good compliance is. Although the ultimate compliance rates approached 65%, overall compliance was 61 %. The study was approved by a state ethics committee of the German Medical Association. They agreed that parents gave informed consent by mailing in urine-saturated filter papers for testing. The parents of each child in the screening area were offered screening once, at the time of the general checkup when the child was about 1 year of age (Schilling et al. 2002). Urine collected was analyzed for catecholamines by high-performance liquid chromatography. Similar to the Quebec trial, if a child had a positive assay, a second sample was requested. If that sample was positive, parents were contacted and asked to bring their child to a center for neuroblas-toma evaluation. The assay was purposely geared to be as sensitive as possible, knowing that such an approach might lead to lower specificity, thus generating a much larger number of false-positive cases than in the North American trial; hence, there were some very interesting and important differences between this and the Quebec trial (Table 2.2).
Compliance in the Quebec trial closely approximated that used to calculate sample size estimates, and further analyses of the Quebec cohort were done using the entire birth cohorts, i.e., children were included whether screened (overall 92%) or not. Because of the lower than expected compliance rate in the German trial, many analyses in their definitive paper were based on results in those individuals only screened (Schilling et al. 2002), as noted in Table 2.2. Almost 2.6 million children were born in the six states during the 5 years of the trial, with 1.5 million actually undergoing screening. On the other hand, both trials successfully utilized concurrent control groups with millions of infants born in those areas. Quebec investigators used two completely independent ascertainment procedures for identifying cases, and more importantly, neuroblastoma deaths. One procedure utilized resources set up by the pediatric oncologists in the various study and control areas noted above, with major input from the North American cooperative groups, the Pediatric Oncology Group and Children's Cancer Group. Collectively, these groups treated 95 % of all young children diagnosed with cancer in North America (Ross et al. 1996). The second ascertainment approach was performed independently by investigators at the Laboratory Center for Disease Control and Statistics Canada, part of Health and Welfare Canada, utilizing the whole of Canada without Quebec as the control. A remarkable congruence was found between the two procedures. In Germany, 10 of its 16 states in whom infants were not offered screening were used as the controls and included populations in the former East Germany. Fortunately,there were excellent childhood cancer registries in both East and West Germany before unification. Investigators were highly confident that these registries would be able to help them ascertain and follow patients (Schilling et al. 2002). Cas-
es were identified by the German Childhood Cancer Registry, "which receives information from all cases of childhood cancer in Germany, including all neuroblastomas. Follow-up of all cases was conducted in cooperation with the neuroblastoma treatment trial of the German Society Pediatric Hematology-Oncol-ogy" (Schilling et al. 2002). That the German ascertainment procedure was near complete was demonstrated by the fact that only two children diagnosed with neuroblastoma in Germany over a 5-year period were lost to follow-up within 5 years.
There were considerable differences in the sample sizes of the two studies. The much smaller Quebec trial was able to perform its study with fewer expected cases and deaths, in large part because it was investigating mortality from birth rather than from 1 year of age, as in the German study: about one-fifth of all neuroblastoma deaths occur in the first year of life. However, their study may have been underpowered (Esteve et al. 1995) had results been less conclusive. The much larger German trial still relied on sample-size estimates that required a 50% reduction in mortality to see "significant benefit for the study population" (Schilling et al. 2002). Finally, both studies nicely utilized uniform neuroblastoma evaluation, treatment, and follow-up in all study cases and in many of the control areas.
2.6.2 Studies' Results
Despite some substantial and interesting differences between the two trials, overall results were strikingly similar, and hence very revealing vis-à-vis neuroblas-toma behavior. Firstly, the Quebec trial nicely confirmed reports from smaller studies in Japan that highly sensitive assays measuring catecholamine metabolites would markedly raise the incidence of neuroblastoma in infants screened at 6 months of age or younger (Woods et al. 1996). The incidence of neu-roblastoma almost doubled over controls in the Quebec birth cohort. The potential for neuroblastomas to regress had been documented for over 30 years (d'Angio et al. 1971), but the magnitude of such regressing cases was never appreciated. The Quebec data suggest that in countries in which there is a very strong medical surveillance of infants, neuroblas-
toma incidence may rise, as previously noted in studies from Denmark (Carlsen 1986). The data also suggest that as newer perinatal technologies become widespread, such as intrauterine ultrasonography, neuroblastoma incidence will also rise. Furthermore, initial studies examining the incidence of neuroblas-toma in children born in Quebec during the 5 years immediately after screening was discontinued, 1 May 1994 to 30 April 1999, document that there has been a reduction in cases compared with the screened population, although not to baseline (WGW: personal observation).
Although many investigators may have predicted the marked rise in neuroblastoma incidence in Quebec, the German results vis-à-vis incidence were almost "shocking," they, too, found a doubling of the neuroblastoma incidence (Table 2.2), all over the age of 1 year. These data strongly suggest that tumors destined to regress may be present and excrete cate-cholamines for a much longer time than neuroblas-toma researchers previously would have hypothesized.
Both studies documented significant neuroblastoma deaths in the study population. In the Quebec trial, there were 22 deaths, with none in the screened detected cases (Woods et al. 2002); however, three infants diagnosed prior to screening at 3 weeks of age, all with extremely high catecholamine levels that would have been detected by screening, and all with stage 4-S disease, died. Two of these infants had classic stage 4-S disease with rapidly expanding liver masses leading to respiratory compromise, despite heroic surgical attempts at relief. The third infant, despite a clinical stage of 4-S, had unfavorable biologic features, including amplified MYCN gene. The patient responded initially to chemotherapy but ultimately relapsed and died; otherwise, only one child in the Quebec population who died was not screened. In the German trial, investigating only those individuals who were screened, there were 17 deaths, 14 in children missed by screening, and 3 who died after preclinical detection. Of these three, "two children died from complications from surgery (one with stage 2-B disease, and the other with stage 3 disease), and one died from complications of chemotherapy (for stage 2-B disease)" (Schilling et
Cumulative mortality due to neuroblastoma among children younger than 8 years of age al. 2002). Despite no deaths in the screened detected children in the Quebec trial, in 1 child with stage 2-B neuroblastoma that was detected by screening at 6 months and who was treated with doxorubicin and cyclophosphamide, a secondary leukemia with an abnormality in chromosome 11q23 subsequently developed. That child underwent bone marrow transplantation and is alive but has severe graft-vs-host disease. An additional child whose disease was detected by screening is in a persistent vegetative state as a result of complications of surgery for severe gastrointestinal obstruction and necrosis. The gastrointestinal problems were attributed to adhesions that resulted from the surgery to remove the neuroblastoma 7 years previously (Woods et al. 2002).
As an intermediate end point, both studies examined the incidence of advanced-stage neuroblastoma (INSS 3-4) in children over 1 year of age. Both showed, if anything, an increase in that incidence in the screened groups (Table 2.2). These results suggested that the screening procedure or the public health interventions instituted as part of the screening projects actually raised the incidence of advanced-stage disease in older infants, perhaps through clinical detection of cases that may have spontaneously regressed, the "halo effect." In the
Quebec trial, there was a significant increased incidence in neuroblastoma over 1 year of age (Woods et al. 1996). In the German trial, even the incidence of advanced-stage disease over the age of 2 years was not lowered by the screening procedure (Schilling et al. 2002).
Finally, and most importantly, cumulative mortality in the study populations in both the Quebec and German trials was not reduced compared with appropriate controls. Mortality was higher in the Quebec screened cohort than in the German, but included a 9-year analysis (Woods et al. 2002). Only preliminary mortality results were presented for the German trial, examining cumulative figures between 1 and 5 years of age (Schilling et al. 2002); however, in examining standardized mortality ratios (SMR) of neuroblastoma comparing study versus control groups, neither showed any reduction, with an SMR of 1.4 in Quebec and 1.1 in Germany (Table 2.2). In fact, in the Quebec trial, examining the rate of death due to neuroblastoma compared with the rest of Canada, as compiled by Statistics Canada, the overall SMR for Quebec was 1.39,with 95 % confidence intervals of 0.85-2.30. Figure 2.2 displays cumulative deaths in the Quebec population versus the four control populations whose deaths were ascertained by study investigators.
2.7 Biologic, Psychologic, Economic, and Clinical Aspects of Neuroblastoma Screening
Had neuroblastoma screening actually been associated with a reduction in mortality, one should have seen children with unfavorable biology detected pre-clinically with subsequent good outcomes. In gener-al,this was not the case. Even in early Japanese trials, virtually all children with neuroblastomas detected by screening demonstrated favorable biologic features, including histology, triploid DNA content, and lack of MYCN amplification (Kaneko et al. 1990). In only one international trial, that conducted in Austria, were any substantial number of patients found through screening with unfavorable biology (Kerbl et al. 1997): results from this study are a bit controversial because of various methodologic issues. Preliminary results from the Quebec trial documented that, similar to the Japanese uncontrolled studies, virtually all children detected clinically had favorable biologic features (Brodeur et al. 1998,2001); however, the vast majority of children who died after being missed by screening had unfavorable biologic features; for example, amplified MYCN oncogene identified in 11 of 19 patients studied (Woods et al. 2002). The German project is expected to publish biologic results in the future.
In summary, the current data overwhelmingly suggest that patients with favorable biology neuro-blastoma are able to be successfully detected preclin-ically; however, those with poor biologic characteristics are missed by screening at 3 weeks, 6 months, and 1 year of age. This suggests that such tumors are either in general not present at these ages, or small enough not to be excreting catecholamines in excess of normal urinary amounts, with subsequent other cellular events leading to a great expansion of the cancer, often with metastatic spread, and clinical detection at an advanced stage.
Unfortunately, very few studies have examined the potential psychological implications of screening infants for neuroblastoma (Bell et al. 1994). Investigators in the Quebec trial tried unsuccessfully to obtain funding for what they believed to be an important secondary aim of their trial. Austrian investigators fortunately were able to conduct interviews on parents of children who underwent neuroblastoma screening with negative results (Dobrovoljski et al. 2003). They found that a large portion of parents of infants who were referred to cancer centers because of elevated catecholamines and were found not to have neuroblastomas remained very concerned about their children, even years later. Hence, the screening procedure was felt to be very psychologically stressing with long-term consequences. The Quebec screening trial was geared toward a very high specificity, and in the end, less than 1 in 10,000 normal children were evaluated at medical centers for neuroblastoma and found not to have the cancer. The number and percent of such children who falsely tested positive was a log higher in the German study, and remains high in Japan today.
Very little has been written on the potential cost-effectiveness of neuroblastoma screening. Because screening has been found to be ineffective, one could argue that such studies would by necessity be negative. The Quebec investigators, however, did prospec-tively examine cost-effectiveness and the data have yet to be published; however, preliminary results lead to a very important and provocative conclusion: over $8 million USD were spent on the Quebec trial in funds provided through the peer-review grant mechanism of the National Cancer Institute. In addition, significant resources were provided by the Quebec Institute of Genetic Medicine for neuroblastoma screening, including costs associated with setting up the infrastructure for metabolic screening that enabled this study to be done as economically as possible. The German Trial cost more than $20 million USD; hence, at first glance these were highly expen sive "negative"studies. But over 4 million children are born in the U.S. every year, compared with 100,000 in Quebec. To put in place an effective infrastructure to screen a large portion of American newborns would have cost easily hundreds of millions of dollars. As importantly, such an infrastructure would have cost tens of millions of dollars to maintain on an annual basis. As noted above, major pediatric voices clamored for institution of neuroblastoma screening in the U.S. before definitive trials proving or disproving its efficacy were performed. Not only did the Quebec and German trials show that neuroblastoma screening was ineffective, but ultimately they saved the American, Canadian, German, and other health care system billions of dollars over a generation. The economic value of well-done research cannot be overestimated, even if results obtained are negative.
With the determination that a substantial number of preclinically detected neuroblastomas undergo spontaneous regression, it is highly likely that a substantial amount of favorable-biology neuroblastoma detected clinically would also spontaneously regress; hence, the results of the neuroblastoma screening studies may have practical implications for the care of infants with clinically detected disease. Yamamoto and colleagues have now defined criteria for observing patients with neuroblastomas detected by screening without incurring any untoward risk. The criteria include the identification of small masses on radiographic examinations but no invasion of the intraspinal canal or infiltration around the great vessels; relatively moderate catecholamines secretions; and parental consent (Yamamoto et al. 1998). Their initial results reveal that a substantial proportion of observed tumors regress, and even those infants that need subsequent treatment do well. It is therefore likely that a similar proportion of infants in whom neuroblastoma is detected clinically at less than 6 months of age can also be observed for potential regression of the tumor, rather than undergo major surgery.
The idea that one could detect childhood cancer pre-clinically by screening has been and remains an appealing prospect. In well-performed trials in the only childhood cancer in which proper studies could be performed at the end of the twentieth century, neu-roblastoma screening for elevated urinary cate-cholamines led to a marked increase in the incidence of the disease with no reduction in its mortality; hence, in 2004 using the markers studied, neuroblas-toma screening has been and should be abandoned throughout the world: in Japan, screening was finally halted in March of 2004 (Tsubono et al. 2004). In the future, there may be better opportunities as more selective markers for poor-biology neuroblastoma are discovered that can be utilized as screening tools. In the meantime, one needs to remember that even collecting urine from a wet diaper may have horrendous long-term consequences, as evidenced by the outcome of some infants screened in the Quebec, German, and Japanese trials. Physicians should always practice the "golden rule" of medicine: primum non nocere.
Ater JL, Gardner KL, Foxhall LE et al. (1998) Neuroblastoma screening in the United States: results of the Texas Outreach Program for neuroblastoma screening. Cancer 82:1593-1602 Bell S, Parker L, Cole M et al. (1994) Screening infants for neuroblastoma: the parents'perspective in false-positive cases. Pediatr Hematol Oncol 11:157-163 Bergeron C, Tafese T, Kerbl R et al. (1998) European experience with screening for neuroblastoma before the age of 12 months. Med Pediatr Oncol 31:442-449 Bernstein M, Leclerc J, Bunin G et al. (1992) A population-based study of neuroblastoma incidence, survival, and mortality in North America. J Clin Oncol 10:323-329 Bessho F, Hashizume K, Nakajo T, Kamoshita S (1991) Mass screening in Japan increased the detection of infants with neuroblastoma without a decrease in cases in older children. J Pediatr 119:237-241 Brodeur G, Ambros P, Favrot M (1998) Biological aspects of neuroblastoma screening. Med Pediatr Oncol 31:394-400 Brodeur GM, Look AT, Shimada H et al. (2001) Biological aspects of neuroblastomas identified by mass screening in Quebec. Med Pediatr Oncol 36:157-159
Brodeur GM, Nakagawara A (1992) Molecular basis of clinical heterogeneity in neuroblastoma. Am J Pediatr Hematol-Oncol 14:111-116 Carlsen NLT (1986) Epidemiological investigations on neuroblastomas in Denmark, 1943-1980. Br J Cancer 54:977-988 Craft AW, Dale G, McGill A et al. (1989) Biochemical screening for neuroblastoma in infants: a feasibility study. Med Pedi-atr Oncol 17:373-378 D'Angio G, Evans A, Koop C (1971) Special pattern of widespread neuroblastoma with a favourable prognosis. Lancet 1:1046-1049
Dobrovoljski G, Kerbl R, Strobl C et al. (2003) False-positive results in neuroblastoma screening: the parents' view. J Pedi-atr Hem Oncol 25:14-18 Esteve J, Parker L, Roy P et al. (1995) Is neuroblastoma screening evaluation needed and feasible? Br J Cancer 71:1125-1131 Hinterberger H, Bartholomew RJ (1969) Catecholamines and their acidic metabolites in urine and tumor tissue in neu-roblastoma, ganglioneuroma and pheochromocytoma. Clin Chim Acta 26:291-294 Kaneko Y, Kanda N, Maseki N et al. (1990) Current urinary mass screening for catecholamine metabolites at 6 months of age may be detecting only a small portion of high risk neuroblastomas: a chromosome and N-myc amplification study. J Clin Oncol 8:2005-2013 Kerbl R, Urban C, Ladenstein R et al. (1997) Neuroblastoma screening in infants postponed after the sixth month of age: a trial to reduce "overdiagnosis" and to detect cases with unfavourable" biological features. Med Pediatr Oncol 29: 1-10
LaBrosse EH (1968) Biochemical diagnosis of neuroblastoma;
use of the urine spot test. Proc Am Assoc Cancer Res 9:39 Mason GA, Hart-Mercer J, Miller ES et al. (1957) Adrenaline-secreting neuroblastoma in an infant. Lancet 2:322-325 Mathieu P, Frappaz D, Chauvin F et al. (1996) A 5-year (19901994) neuroblastoma screening feasibility study in France. Methodology and preliminary observations. Early Hum Dev 46:177-196
Matthay KK, Sather HN, Seeger RC et al. (1989) Excellent outcome of stage II neuroblastoma is independent of residual disease and radiation therapy. J Clin Oncol 7:236-244 Matthay KK, Villablanca JG, Seeger RC et al. (1999) Treatment of high-risk neuroblastoma with intensive chemotherapy, radiotherapy, autologous bone marrow transplantation, and 13-cis-retinoic acid. N Engl J Med 341:1165-1173 Nishi M, Miyake H, Takeda T et al. (1987) Effects of the mass screening of neuroblastoma in Sapporo City. Cancer 60:
Was this article helpful?
Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.