Author reply

Abstract

I appreciate the comments provided by Moll et al. regarding my article.1 I agree with their comment that we truly will never know whether screening is beneficial without a prospective trial. However, I disagree with certain points they have made. First, Moll et al. state that screening is beneficial only if the screened disease is treatable, implying that trilateral retinoblastoma is not treatable. I would like to point out that there are at least 6 children reported in the literature who have survived ≥ 2 years after the diagnosis of trilateral retinoblastoma; in fact, 1 patient was alive 8 years after the diagnosis of the pineoblastoma.1, 2 Whereas in the past pineoblastomas and other supratentorial primitive neuroectodermal tumors (PNETs) had a dismal prognosis, current treatment strategies involving surgical resection of the primary tumor, chemotherapy, and craniospinal radiotherapy have yielded 3-year overall and progression free survivals of 73% and 61% respectively, for pineoblastoma and 57% and 45%, respectively, for other supratentorial PNETs.3 Because of these new data and major advances in treatment, the prognosis of patients who develop a PNET after bilateral or hereditary retinoblastoma should improve within the next decade. The pessimism shared by Moll et al. dates back to the time when PNET largely was incurable. Second, Moll et al. stated that for screening to be effective, the rate of incidence of the screened disease should be high. It should be noted that screening for an intracranial tumor should not be performed for all patients with retinoblastoma but only for the subset with hereditary or bilateral retinoblastoma. In this population of children, approximately 6–10% will develop an intracranial PNET.4 Third, Moll et al. state that obtaining magnetic resonance imaging (MRI) scans will detect “irrelevant abnormalities that may lead to inadvertent ‘treatment’.” This statement opens another Pandora's box in the management of trilateral retinoblastoma. Is a biopsy of a pineal or suprasellar mass necessary when a patient with hereditary or bilateral retinoblastoma develops intracranial disease? Moll et al. describe a case with clinically benign disease in a patient they have been following with brain MRI. We know that in isolated pineal tumors biopsy is recommended because of the variety of neoplasms that can occur in this region, including 10–15% that are benign.5 The histology of the tumor dictates the type of treatment and does not necessarily translate into the use of craniospinal irradiation. Fourth, Moll et al. state that craniospinal irradiation results in extensive neuropsychologic sequelae. When examining the neuropsychologic sequela of craniospinal irradiation, three factors appear to be most important: the age of the patient, dose of radiation, and volume of tissue irradiated. The young age of patients with retinoblastoma obviously is an important factor in the development of cognitive problems. Recent advances in treatment have lowered the craniospinal dose in PNET of the posterior fossa and have resulted in an improvement in neuropsychologic outcome.6 The use of conformal radiotherapy techniques also will limit the high dose region in patients treated in the current era.7, 8 Finally, as with the majority of tumors, the size of the lesion usually correlates with outcome. Screening of patients with bilateral or hereditary retinoblastoma will detect smaller tumors. In a meta-analysis by Kivelä, patients with intracranial tumors < 15 mm had the best prognosis.9 In analyzing all the available data, I believe that the gloomy picture for children with trilateral retinoblastoma painted by Moll et al. is not justified in a climate of modern technology and advances in the treatment of intracranial PNET. Arnold C. Paulino M.D.*, * Division of Radiation Oncology, The University of Iowa, Hospitals and Clinics, Iowa City, Iowa