Population-based comparison of prognostic factors in invasive micropapillary and invasive ductal carcinoma of the breast.

BACKGROUND: Invasive micropapillary carcinoma (IMPC) is a variant of breast carcinoma with a higher propensity for lymph node metastases compared with invasive ductal carcinoma (IDC). METHODS: Retrospective analysis of 636 IMPC and 297 735 IDC cases in the Surveillance, Epidemiology and End RESULTS database comparing disease-specific survival (DSS) and overall survival (OS) between IMPC and IDC. RESULTS: A higher percentage of IMPC cases (52.0%) had nodal metastases compared with IDC cases (34.6%). The 5-year DSS and OS for IMPC was 91.8% and 82.9%, respectively compared with 88.6% and 80.5% for IDC, respectively. For both IMPC and IDC, oestrogen-receptor positivity was associated with better survival, while having four or more positive lymph nodes or larger tumour size correlated with worse survival. Radiotherapy provided a survival benefit for both histological types. CONCLUSIONS: Despite IMPC's higher propensity for lymph node metastasis, IMPC has DSS and OS that compare favourably with IDC.

The local field in infratentorial ependymoma: does the entire posterior fossa need to be treated?

INTRODUCTION: In the past decade, there have been multiple reports indicating that the predominant problem in the curative treatment of intracranial ependymoma is local failure. As a result, many have recommended local field radiotherapy. For infratentorial ependymoma, there is controversy regarding what constitutes the local field. Some radiation oncologists advocate coverage of the entire posterior fossa, whereas others recommend radiotherapy to the tumor bed and a safety margin. METHODS AND MATERIALS: From 1984 to 1998, 28 patients with posterior fossa ependymoma were diagnosed at our institution. There were 18 males and 10 females with a median age of 12 years (range, 2-81 years). Four patients (14%) had high-grade ependymoma and 3 (11%) had M+ disease at initial diagnosis. Gross total resection was achieved in 17 (61%) and postoperative radiotherapy (RT) was given to 22 (77%). Radiotherapy fields were craniospinal in 10, whole brain in 1, posterior fossa in 2, and tumor bed with a 2-cm. margin in 9. Median dose to the primary site was 54 Gy (range, 45-55 Gy). All 4 patients with high-grade ependymoma received craniospinal RT. Six patients did not receive RT after surgery. Magnetic resonance imaging (MRI) or computed tomography (CT) scans of the brain at initial diagnosis were compared to MRI or CT scans of patients at relapse to determine if the local relapse was in the tumor bed or nontumor bed posterior fossa. Median follow-up was 127 months (range, 14-188 months). RESULTS: Six patients have relapsed. For the 11 patients who had craniospinal or whole brain radiotherapy (RT), 3 recurred (tumor bed 1, spine 1, nontumor bed posterior fossa + spine 1). Both patients who failed in the spine had high-grade tumors. Neither of the 2 treated with posterior fossa fields relapsed. For the 9 patients who had tumor bed RT alone and the 6 who did not receive RT, there were 3 relapses; all were in the tumor bed. There were no relapses in the nontumor bed posterior fossa. CONCLUSION: For nondisseminated, low-grade infratentorial ependymoma, the radiotherapy volume does not need to include the entire posterior fossa. This information can be used to minimize late effects of RT in the era of three-dimensional (conformal) radiotherapy. No conclusion can be reached regarding the appropriate local field for high-grade infratentorial ependymoma because of the small number of patients.

The significance of radiotherapy treatment duration in intracranial ependymoma.

PURPOSE: To determine if radiation therapy treatment duration has an impact on the outcome of intracranial ependymoma. METHODS AND MATERIALS: From 1965 to 1997, 34 patients with intracranial ependymoma were seen and treated with postoperative radiotherapy (RT). There were 26 male and 8 female patients with a median age of 14 years (range, 18 months to 60 years). Tumor location was infratentorial in 23 (67.6%) and supratentorial in 11 (32.4%). Pathology was low-grade in 25 (73.5%) and high-grade in 9 (26.5%). Two (5.9%) patients had M+ disease at initial diagnosis. Gross total resection was achieved in 12 (35.3%), subtotal resection in 19 (55. 9%), and biopsy alone in 3 (8.8%). Median RT dose to the primary site was 5400 cGy (range, 4500 cGy to 6600 cGy). The entire neuraxis was treated in 14 (41.2%), whole brain in 9 (26.5%), and the primary site alone in 11 (32.4%). Median radiotherapy treatment duration was 43 days (range, 26 to 122 days). Patients treated with craniospinal RT were more likely to have treatment duration of > or =50 days. Median follow-up was 108 months (range, 24 to 252 months). RESULTS: The 5-, 10-, and 15-year overall survival rates were 71.5%, 63.3%, and 55.4%. Local control rates at 5, 10, and 15 years were 61.8%, 57.7%, and 57.7%. For patients with RT treatment duration <50 days, the 5-, 10-, and 15-year overall survivals were 85.5%, 78.9%, and 65.7% whereas for those with treatment duration > or = 50 days, the corresponding rates were 45.5%, 36.4%, and 36.4% (p = 0.01, log-rank test). Local control rate at 5, 10, and 15 years was 70.6% for patients with RT treatment duration <50 days and 45.5%, 36.4%, and 36.4% for patients with RT treatment duration > or =50 days (p = 0.05, log-rank test). Age, gender, tumor location, tumor grade, degree of surgical resection, RT volume, and RT dose (4500-5399 cGy vs. 5400-6600 cGy) were not found to have a statistically significant impact on overall survival, disease-free survival, or local control. The most common site of failure was local, found in 13 (38.2%) patients. Three patients also had spinal failure in addition to a primary site failure. There were no isolated nonprimary site brain or spine failures. CONCLUSIONS: Radiation therapy treatment duration is an important prognostic factor for survival and local control with patients having treatment duration <50 days having a better outcome than those with treatment duration > or =50 days. Because the overwhelming site of failure was local, we recommend local field irradiation in patients with M0 disease.

The cranial-spinal junction in medulloblastoma: does it matter?

PURPOSE: Late effects of treatment in children and young adults with medulloblastoma can be influenced by the technique employed in radiating the craniospinal axis. The purpose of this study is to determine whether the placement of the cranial-spinal junction has an impact on dose to the cervical spinal cord and surrounding organs. METHODS AND MATERIALS: Five patients underwent computed tomography (CT) simulation in the prone position for craniospinal irradiation. A dose of 36 Gy was prescribed to the entire neuraxis. The doses to the cervical spinal cord and surrounding organs were calculated using a cranial-spinal junction at the C1-C2 vertebral interspace (high junction) or at the lowest point in the neck, with exclusion of the shoulders in the lateral cranial fields (low junction). The volume of critical organs at risk, as well as dose to these structures using the cranial and spinal field(s) were outlined and calculated using the CMS FOCUS 3-dimensional treatment planning system. RESULTS: The average dose to the cervical spinal cord was 11.9% higher than the prescribed dose with the low junction, and 6.7% higher with the high junction. However, doses to the thyroid gland, mandible, pharynx, and larynx were increased by an average of 29.6%, 75.8%, 70.6%, and 227.7%, respectively, by the use of the high junction compared to the low junction. CONCLUSION: A higher dose to the cervical spinal cord can be minimized by using a high junction. However, this would be at the cost of substantially increased doses to surrounding organs such as the thyroid gland, mandible, pharynx, and larynx. This can be critical in children and young adults, where hypothyroidism, mandibular hypoplasia, and development of second malignancies may be a late sequela of radiation therapy.

Posterior fossa boost in medulloblastoma: an analysis of dose to surrounding structures using 3-dimensional (conformal) radiotherapy.

PURPOSE: The purpose of this study was to compare three different techniques of delivering the posterior fossa boost in patients with medulloblastoma. METHODS AND MATERIALS: Five patients underwent CT simulation for treatment planning of the posterior fossa boost. For each slice, the posterior fossa was contoured in addition to the cochlea, non-posterior fossa brain, pituitary gland, mandible, parotid glands, thyroid gland, pharynx, and cervical spinal cord. For each patient, plans for three techniques of delivering the posterior fossa boost were compared. Technique A utilized parallel-opposed lateral fields using bony landmarks (2-dimensional radiotherapy); the other two techniques were planned using 3-dimensional radiotherapy. Technique B utilized a pair of coplanar wedged posterior oblique beams, whereas Technique C utilized a pair of posterior oblique fields and a vertex field. Dose-volume histograms (DVH) were obtained for each of the organs contoured and for each technique and patient. The maximum, minimum, and mean dose to each organ were determined using the DVH program in our treatment planning system. RESULTS: In 3 of 5 patients, the planning target volume (PTV) was not included in the treatment field with Technique A. The cochlea received 100%, 50%, and 42% of the prescribed posterior fossa dose using Techniques A, B, and C respectively. The mean dose to the non-posterior fossa brain was highest with Technique C, intermediate with Technique A, and lowest for Technique B. The mean dose to the non-posterior fossa brain with Technique B was comparable to the mean non-posterior fossa brain dose delivered using parallel-opposed lateral fields based on the CT definition of the PTV. Likewise, mean dose to the pituitary gland was lowest for Technique B. Both Techniques B and C were associated with a higher mean dose to the thyroid gland, mandible, parotid glands, and pharynx. CONCLUSIONS: The use of Technique B minimized the radiotherapy dose to the cochlea, pituitary gland, and non-posterior fossa brain. Contrary to what one may expect, conformal radiotherapy using Technique B did not deliver a higher dose to the non-posterior fossa brain over standard parallel-opposed lateral fields. Other advantages of conformal techniques B and C over 2-dimensional radiotherapy are the inclusion of the PTV in all patients and a lower mean dose to the pituitary gland. The main disadvantage of conformal Techniques B and C employed in our patients is a higher mean dose to the thyroid gland and other tissues in the neck.

Determination of the inferior border of the thecal sac using magnetic resonance imaging: implications on radiation therapy treatment planning.

PURPOSE: To determine whether the traditional teaching of placing the caudal border of the spinal field at the S2-S3 interspace in children receiving craniospinal irradiation (CSI) is appropriate. METHODS AND MATERIALS: Twenty-three children had magnetic resonance imaging (MRI) of the spine with gadolinium prior to craniospinal irradiation at one institution. Thecal sac termination using MRI was determined by drawing a perpendicular line from the point of convergence of dural margins to the corresponding vertebral body. RESULTS: Location of thecal sac termination varied from mid-S1 to low S3 vertebral body, with the most frequent site at the upper S2 vertebral level. Only 2 of 23 (8.7%) children had thecal sac terminations below the S2-S3 interspace. For the nine patients with neuraxis disease, none had thecal sac terminations below the S2-S3 interspace. In seven of the nine patients who had neuraxis seeding at initial presentation, MRI of the spine after CSI was performed and showed that thecal sac termination was lower after radiation therapy in two children, higher in one, and the same in four. CONCLUSIONS: In 2 of 23 children (8.7%), placement of the inferior border at the bottom of the S2 vertebral body would have missed the entire thecal sac. Treatment to the entire neuraxis with adequate coverage of distal spinal theca can be achieved by using MRI. Individualized spinal fields using the MRI may help minimize radiation scatter to the gonads while adequately covering the target volume.

Is prophylactic neck irradiation indicated in patients with squamous cell carcinoma of the maxillary sinus?

PURPOSE: To determine the proportion of patients with squamous cell carcinoma of the maxillary sinus who will fail in regional nodes without elective neck treatment and to identify any prognostic factors that may influence neck control. METHODS AND MATERIALS: From 1971-1995, 42 consecutive patients with squamous cell carcinoma of the maxillary sinus were seen at our department for curative treatment. There were 35 males and 7 females, with a median age at diagnosis of 63.5 years (range, 42-77 years). One tumor was classified as T1, 5 had T2, 15 had T3, and 21 had T4 disease. Four of 42 patients (9.5%) had cervical lymphadenopathy at initial presentation. Thirty-three patients had surgical resection and radiotherapy and nine had radiotherapy alone. None of the 38 patients with clinical N0 necks received elective treatment to the cervical nodes. RESULTS: Median overall survival was 30 months for all patients. Of the 38 patients with N0 disease, 11 (28.9%) had neck recurrence. Of the 11 neck failures, 9 were ipsilateral only, 1 was contralateral, and 1 had bilateral neck recurrence. The most common site of neck failure was in the upper neck (submandibular and jugulodigastric lymph nodes). Four of the 38 patients (10.5%) had isolated neck failure. Only tumor stage was found to be significant for neck relapse, with T1 and T2 doing worse compared to T3 and T4 tumors. Location of tumor (infrastructure vs. suprastructure), involvement of the oral cavity/oropharynx, nasal cavity, nasopharynx or orbit did not predict for cervical node relapse. Local control at the primary site was likewise not prognostic. The median overall survival for patients who remained N0 was 80 months and for those with initial cervical involvement or recurred in the neck without elective neck irradiation was 25 months (p = 0.05). CONCLUSION: Based on the 28.9% rate of neck recurrence and the poor median survival of patients who recur in the neck, we recommend prophylactic ipsilateral neck irradiation in patients with T1-T4 squamous cell carcinoma of the maxillary sinus.

Long-term effects in children treated with radiotherapy for head and neck rhabdomyosarcoma.

PURPOSE: To examine the long-term effects of treatment in children receiving radiotherapy for head and neck rhabdomyosarcoma. METHODS: From 1967 to 1994, a total of 30 children with head and neck rhabdomyosarcoma received megavoltage radiotherapy at one institution. Seventeen patients (57%) have survived and have at least a 5-year follow-up. There were 11 males and 6 females, with a median age of 5.7 years (range 2.2-11.6) at the time of radiotherapy. Tumor location was orbit in 6 patients, infratemporal fossa in 4, paranasal sinuses in 2, and supraglottic larynx in 2; the nasopharynx, pterygopalatine fossa, and parotid gland were sites for the remaining children. All but 2 patients had tumors of embryonal histology. The Intergroup Rhabdomyosarcoma Study (IRS) Group was I in 2, II in 3, and III in 11 children; 1 patient had a recurrent tumor after surgery alone. Radiotherapy volume was the primary tumor or tumor bed in 13, tumor and whole brain in 3, and tumor and craniospinal axis in 1. Median radiotherapy dose to the primary site was 5,040 cGy (range 4,140-6,500) and to the whole brain was 3,000 cGy. All but 1 were treated with 150-200-cGy fractions; 1 patient received 250-cGy fractions for a tumor in the larynx. Chemotherapy was vincristine (V), actinomycin-D (A), and cyclophosphamide (C) in 10 patients, VAC + adriamycin in 2, VA in 1, VA + ifosfamide in 1, VC + adriamycin in 1, and none in 2. One patient had salvage chemotherapy consisting of cisplatin and etoposide. Median follow-up time was 20 years (range 7.5-33). RESULTS: Late effects of treatment were seen in all patients and included facial growth retardation in 11, neuroendocrine dysfunction in 9, visual/orbital problems in 9, dental abnormalities in 7, hearing loss in 6, and hypothyroidism in 3. Intellectual and academic delays were documented in 3 patients who had received whole brain radiotherapy. While neuroendocrine, thyroid, dental, and cognitive sequelae were primarily attributed to radiotherapy, hearing loss was thought to be a direct result of tumor destruction and, in 1 case, cisplatin chemotherapy. Late effects at or beyond 10 years from radiotherapy were few, but severe, and included chondronecrosis, esophageal stenosis, second malignancy, and brain hemorrhage. CONCLUSION: Late effects of treatment in children receiving radiotherapy for head and neck rhabdomyosarcoma are frequent. Although radiotherapy is a significant contributor of neuroendocrine, dental, thyroid, and cognitive toxicity, it is not usually implicated with hearing loss. Late toxicity of treatment beyond 10 years is not as frequent as those occurring within 10 years of therapy.

Local control in synchronous bilateral Wilms tumor.

PURPOSE: To evaluate the role of radiation therapy (RT), chemotherapy (CT), and surgery in the local control of synchronous bilateral Wilms Tumor (WT). METHODS AND MATERIALS: Between 1962 to 1993, 45 children were treated for bilateral WT; 38 patients with synchronous tumors were reviewed. Initial surgery depended on the era of treatment and included unilateral nephrectomy (N)/partial nephrectomy (PN) and contralateral PN in 6, unilateral N/PN alone in 7, and biopsy only in 25. Chemotherapy (CT) consisted of vincristine, actinomycin-D, and adriamycin in 32 and vincristine/ actinomycin-D in 6. Radiation therapy (RT) was given to 32 patients. Treatment included both kidneys in 20, unilateral kidney plus contralateral renal bed in 9, unilateral kidney in 2, and unilateral renal bed in 1. Follow-up was 16 months to 25 years (median: 6.3 years). RESULTS: Local control (LC) has been maintained in 66 out of 76 sites (87%). For Stage I-II disease with initial N/PN, LC was 10 out of 12 with RT and 11 out of 11 without RT; for Stage III with initial N/PN, LC was 8 out of 9 with RT and 1 out of 1 without RT. Initial CT and RT was followed by delayed N/PN for 20 sites; LC was 15 out of 17 in post induction Stage I-II and 1 out of 3 in postinduction Stage III. In 23 sites undergoing biopsy and chemotherapy, LC was 19 out of 20 with RT and 1 out of 3 without RT. Seven of 23 sites had a complete response (CR) after induction CT, and LC was maintained in four out of four with RT and one out of three without RT. Univariate Cox Regression analysis demonstrated that sites receiving two drugs had a statistically significant increase in loco-regional relapse when compared to sites receiving three drugs (p = 0.004). Major morbidities related to multimodality therapy have included renal failure in one patient and small bowel obstruction requiring lysis of adhesions in two patients. CONCLUSIONS: Local control does not seem to be compromised by renal conservation therapy. Local control is excellent in sites treated with radiation therapy in combination with three drug chemotherapy.

Late effects in children treated with radiation therapy for Wilms' tumor.

PURPOSE: To determine the frequency and types of late effects in children receiving radiation therapy (RT) for Wilms' tumor. MATERIALS AND METHODS: From 1968 to 1994, 55 children received megavoltage RT at our institution as part of treatment for Wilms' tumor. A total of 42 (76.4%) have survived and have a minimum follow-up of 5 years. There were 25 female and 17 male patients with a median age at diagnosis of 48 months (range, 7-126 months). There were 12 Stage I, eight Stage II, 15 Stage III, six Stage IV, and one Stage V patient. RT was delivered to the hemiabdomen in 36 and whole abdomen in six patients. RT dose was 1000-1200 cGy (Group A) in 12, 1201-2399 cGy (Group B) in 11, and 2400-4000 cGy (Group C) in 19. Whole-lung RT was delivered to 13 patients either at diagnosis or pulmonary relapse. All patients received chemotherapy; the most common agents were actinomycin-D/vincristine/adriamycin in 13 and actinomycin-D/vincristine in 18. Median follow-up was 181 months (range, 60-306 months). RESULTS: Of 42 patients, 13 (31.0%) did not have late effects of treatment. The number of patients who developed muscular hypoplasia, limb length inequality, kyphosis, and iliac wing hypoplasia were seven (16.7%), five (11.9%), three (7.1%), and three (7.1%), respectively. Scoliosis was seen in 18 (42.9%) with only one patient requiring orthopedic intervention. Median time to development of scoliosis was 102 months, with a range of 16-146 months. The actuarial incidence of scoliosis at 5, 10, and 15 years after RT was 4.8 +/- 3.3%, 51.8 +/- 9.0%, and 56.7 +/- 9.3%, respectively. Only one of 12 Group A patients developed scoliosis. The 10- and 15-year actuarial incidences of scoliosis for Group A and B patients were 37.7 +/- 12.4% and 37.7 +/- 12.4%, whereas for Group C patients the incidences were 65.8 +/- 12.0% and 74.4 +/- 11. 7% (p = 0.03, log rank test). The actuarial incidence of bowel obstruction at 5, 10, and 15 years was 9.5 +/- 4.5%, 13.0 +/- 5.6%, and 17.0 +/- 6.5%. Of 23 patients, five irradiated within 10 days of surgery and one of 19 irradiated after 10 days developed bowel obstruction (p = 0.09, log rank test). Three patients developed hypertension with normal blood urea nitrogen (BUN) and creatinine levels; another patient had chronic renal insufficiency in a nonirradiated kidney. One patient developed diffuse interstitial pneumonitis. Of the 19 female patients who have reached puberty, three have given birth, and 15 have regular and one has irregular menstrual periods. Four patients developed benign neoplasms; three were in the RT field (two osteochondroma, one lipoma) and one outside (cervical intraepithelial neoplasia II). There were three second malignancies (chronic myelogenous leukemia at 9 years, osteosarcoma at 11 years, and breast cancer at 25 years after initial diagnosis of nephroblastoma); both solid malignancies occurred in the RT field. CONCLUSIONS: Late effects of therapy were seen in more than two thirds of children treated for Wilms' tumor. Children treated with lower doses (<2400 cGy) had a lower incidence of scoliosis compared with those who received more than 2400 cGy. There is also a suggestion that the incidence is lower in patients who received 1000-1200 cGy. Severe physical and functional deformity from RT was uncommon.

Radiotherapeutic management of medulloblastoma.

Although craniospinal irradiation has been employed in children with medulloblastoma for the past 40 years, many issues concerning its use have been raised and examined, and some continue to be debated. Careful radiation technique includes adequate irradiation of the neuraxis with special attention to the cribriform plate region and termination of the thecal sac. Conventional-dose craniospinal radiation therapy, in combination with chemotherapy, is currently recommended for patients with high-risk medulloblastoma. The appropriate dose of radiation to the craniospinal axis when this modality is combined with chemotherapy for low-risk medulloblastoma remains to be defined. Long-term results of hyperfractionated radiation therapy are likewise awaited. In an effort to decrease late toxicity to the immature central nervous system, radiation therapy can be delayed in a proportion of infants by administering chemotherapy after maximal tumor debulking.

Current issues in the diagnosis and management of Wilms' tumor.

Significant advances have been made in the treatment of children with Wilms' tumor. Whereas 50 years ago overall survival was less than 10%, current survival estimates approach 90%. This progress has been made possible by the cooperation of pediatric oncologists, surgeons, pathologists, radiation oncologists, and support personnel in their joint quest to find a cure for this intriguing childhood neoplasm. The next goals are to determine better treatment for patients with unfavorable histology tumors or recurrent disease and to identify patients who may not require adjuvant therapy. Also, the role of renal parenchymal-sparing resections needs to be defined further.

Controversies in the management of intracranial germinomas.

Intracranial germinomas are uncommon tumors. In the past, patients have traditionally been diagnosed with a trial of focal radiotherapy without biopsy. If the tumor was radiosensitive, it was presumed to be a germinoma. Because of the minimal morbidity and mortality associated with microsurgical techniques, almost all patients now have a histologic diagnosis and can be treated appropriately. Radiation therapy has been employed for the past three decades in the curative management of this disease. The long-term results have been excellent; however, current approaches have explored the possibility of reducing or even eliminating radiation therapy. This article provides a brief overview of intracranial germinomas and explores the controversies in the surgical, radiotherapeutic, and chemotherapeutic management of this tumor.

Postoperative irradiation of patients with malignant tumors of skull base.

To better understand how to irradiate patients after skull-base surgery, the authors of this study analyzed the tumor recurrence patterns in 22 patients with various malignancies. Of these patients, 13 underwent craniofacial resection, 6 had infratemporal fossa resection, and 3 had temporal bone resection. The entire operative field was irradiated in 8 patients (total-field group), and part of the operative field was irradiated in 14 patients (partial-field group). Tumor regrew within the irradiated area in 25% of the total-field group and 14% of the partial-field group. Tumor appeared outside the operative field in 13% of the total-field group and 57% of the partial-field group. Recurrences appeared in contiguous sites in 5 patients, in the unirradiated scar in 5 patients, and in the lymph nodes in 4 patients. Based on the study findings, the authors state that irradiation of the entire operative field, including scars and potential areas of contiguous spread, may be necessary to maximize local control after skull-base surgery.

Esthesioneuroblastoma: the University of Iowa experience 1978-1998.

OBJECTIVE: Esthesioneuroblastoma is rare and the best treatment has yet to be defined. The purpose of this study is to analyze the natural history, treatment, and patterns of failure of esthesioneuroblastoma treated at one institution. METHODS: Between 1978 and 1998, 13 patients with esthesioneuroblastoma were identified using the University of Iowa Tumor Registry. All patients were staged according to Kadish criteria. Mean follow-up was 6.3 years. Six patients had 5 or more years of follow-up and four had follow-up exceeding 9.5 years. One patient was lost to follow-up at 36 months. RESULTS: No patients had Kadish stage A disease, five were stage B, and eight stage C. Overall actuarial 5- and 10-year survival rate was 61% and 24%, respectively. Disease-free survival rate at 5 and 10 years was 56% and 42%, respectively. Seven patients have died, three of intercurrent disease and three of disease progression, one with an unknown disease status. Six patients remain alive, three without evidence of disease and three have experienced a local or regional recurrence. Five patients who were initially controlled developed recurrence, three local only, one locoregional, and one regional and distant. Median time to failure was 96 months. All patients with follow-up exceeding 12 years have experienced either a local or regional recurrence. Survival after salvage therapy in these patients ranged from 3 to 12 years. CONCLUSION: Esthesioneuroblastoma has a long natural history characterized by frequent local or regional recurrence after conventional treatment. Successful retreatment can lead to prolonged survival.

The location of the prostatic apex on retrograde urethrography and its relationship to the bottom of the ischial tuberosities.

To determine the proportion of patients undertreated if the inferior border of the prostate field is set at the bottom of the ischial tuberosities, we reviewed the ports of 80 patients with prostate cancer who had retrograde urethrography as part of simulation for radiation therapy. For the 75 evaluable urethrograms, the mean distance from the top of the urethrogram cone to the bottom of ischial tuberosities was 1.38 cm (range, -0.48-2.90 cm). A comparison of the inferior border defined by the bottom of the ischial tuberosities and retrograde urethrography showed that 47 of 75 (62.7%) patients would have been undertreated if a margin of 1.5 cm was employed, and the prostatic apex was thought to be directly above the urethrogram cone. If the apex was thought to be 1 cm above the cone, six of 75 (8.0%) patients would have been undertreated, using a margin of 1.5 cm. Although previously published reports have established that using the bottom of the ischial tuberosities as the inferior border of the prostate field results in 10-45% undertreatment, our findings, when adjusted for the variability of prostatic apex location and margin of normal tissue employed, indicate that only 8% may actually be undertreated.

Relapsed Wilms tumor: is there a role for radiation therapy?

From 1968 to 1996, 21 of 97 (22%) patients with Wilms tumor treated at the University of Iowa Hospitals and Clinics relapsed after initial therapy. The most common sites of initial recurrence were the lungs in 12 (57%), abdomen in 3 (14%), bone in 3 (14%), and liver in 2 (10%). Treatment for recurrence included chemotherapy in 16, radiotherapy in 15 and surgery in 8. Radiotherapy was given to bilateral lungs of 8 previously untreated chests; total dose was 1,200 cGy to 1,500 cGy, with 2 patients receiving 1,000-cGy boost. One patient had abdominal radiotherapy after surgery for relapse; six patients had palliative radiotherapy to distant sites. The 1-, 3-, and 5-year postrelapse survivals were 62 +/- 11%, 38 +/- 11%, and 33 +/- 10%, respectively. Using Cox proportional-hazard regression models, abdominal recurrence (p = 0.0002; hazard ratio, 39.5) and initial stage IV disease (p = 0.019; hazard ratio, 6.27) were found to have a negative impact on postrelapse survival. Seven patients with previously nonirradiated chests are alive at a median follow-up of 123 months after relapse (range, 55--272 months); all had whole lung irradiation with or without chemotherapy at relapse for lung metastases. Radiotherapy was also found to be an effective treatment in the palliation of painful metastases.

Resected pancreatic cancer treated with adjuvant radiotherapy with or without 5-fluorouracil: treatment results and patterns of failure.

There are relatively little data regarding patterns of recurrence after curative resection and postoperative radiotherapy with or without 5-fluorouracil (5-FU) for patients with adenocarcinonima of the pancreas. Between 1978 and 1997, 41 patients underwent postoperative radiotherapy (RT) at Loyola-Hines Department of Radiotherapy. Of the 38 evaluable patients, 30 had RT + 5-FU and 8 had RT alone. Twenty-nine patients (76.3%) had a Whipple's resection, seven (18.4%) had distal pancreatectomy, and two (5.2%) had total pancreatectomy. Thirty-three (86.8%) of the 38 patients received > or =4,500 cGy to the tumor bed. Median survival for all patients was 21 months. The median survivals for patients who received RT + 5-FU and RT alone were 26 months and 5.5 months (p = 0.004). The most common site of failure was the liver, as seen in 79.2% of all recurrences. The peritoneum, other distant sites (lungs, bone, distant lymph nodes), and locoregional tumor bed were components of failure in 33.3%, 29.2%, and 25.0%, respectively. Locoregional failure alone was found in only one patient. Our median survival with postoperative RT + 5-FU is consistent with results reported by the Gastrointestinal Tumor Study Group and Mayo Clinic. Although patients who had RT + 5-FU had a better median survival than those who received RT alone, our RT-alone group had an inferior survival outcome compared to other published reports and may represent patient selection bias. Efforts in controlling this disease should be directed to prevention of intraabdominal relapse.

Diffuse malignant leptomeningeal gliomatosis in a child: a case report and review of the literature.

Diffuse leptomeningeal gliomatosis is a rare condition characterized by glioma in the leptomeninges without a dominant mass lesion. The difficulty in diagnosis of this condition, its rarity. and its extensive nature have hampered its successful treatment. Most cases of primary diffuse leptomeningeal gliomatosis have occurred in adults. Reported here is a case of this condition in a 9-year-old girl; to the authors' knowledge, she is the youngest patient with diffuse leptomeningeal gliomatosis and the longest survivor of the malignant variety.

Comparison of posterior fossa and tumor bed boost in medulloblastoma.

To quantify the difference between the area of brain irradiated using the posterior fossa boost (PFB) and tumor bed boost (TBB) in medulloblastoma, we studied 15 simulation radiographs of patients treated in our institution from 1990 and 1999. The PFB was compared with the TBB, which was defined as the tumor bed plus 2-cm margin as demonstrated by postoperative magnetic resonance imaging. The PFB field treated a mean area of 9.43 cm2 more brain than the TBB. In 3 patients (20%), the area of the brain in the TBB was larger than the PFB. In 11 patients (73.3%), the PFB field had more than 10% more brain than the TBB. The cochlea was in the PFB and TBB field in all patients. In more than two thirds of patients, the area of brain irradiated with the PFB was at least 10% greater than the TBB. Future studies are needed to determine whether the TBB can replace the PFB in patients with medulloblastoma.