Role of HER2 status in the treatment of brain metastases arising from breast cancer by stereotactic radiosurgery

117

Background: Overexpression of human epidermal growth factor receptor-2 (HER2) in breast cancer has been shown to be an independent risk factor of development of brain metastases (BM). However, the role of HER2 on the radiosensitivity of BM remains controversial. We investigated the efficacy of single fraction stereotactic radiosurgery (SRS) in the treatment of limited BM from breast cancer, based on the HER2 status.

Methods: From 2004 to 2010, 57 patients with limited BM were treated with SRS using Leksell Gamma-Knife to a median dose of 20 Gy (range 12-20 Gy) prescribed to the 50% isodose line. The median number of lesions treated were 2 (range 1-7) and the median tumor volume was 92.8 mm3 (range 3.4 -793.7 mm3). There were 27 HER2 positive patients, with 32 ER positive and 22 PR positive patients. Six patients were triple negative. ER/PR information was not available for 4 patients. Overexpression of HER2 was defined as HER2 +++ by immunohistochemistry using the DAKO HercepTest or fluorescent in situ hybridization (FISH) analysis. Time to local recurrence, time to development of new brain metastases, progression-free survival (PFS), and overall survival (OS) were assessed from the date of SRS.

Results: With a median follow-up of 11 months (range 1-66 months), 12 patients developed recurrence at the previously treated site (9 HER2 positive patients) and 30 developed new brain metastases (16 HER2 positive patients) The median PFS and OS for all patients were 8 and 16 months respectively. The median PFS for HER2 positive patients compared to HER2 negative patients were 7 and 11 months, respectively (p=0.352). HER2 positive patients had a median OS of 20 months, compared to 14 months in HER2 negative patients (p=0.086). Five patients went on to develop leptomeningeal disease, all of whom were HER2 positive.

Conclusions: HER2 overexpression does not increase the risk of recurrence at the site of treatment or appearance of new BM. It does not affect PFS or OS following SRS. However, it remains a risk factor for the development of leptomeningeal disease.

p27-p16 fusion gene inhibits angioplasty-induced neointimal hyperplasia and coronary artery occlusion

Inhibition of proliferative neointima formed by vascular smooth muscle cells is a potential target in preventing angioplasty-induced restenosis. We have created a potent antiproliferative by fusing the active regions of the p27 and p16 cell cycle inhibitors. Intravascular delivery of a replication-deficient adenoviral vector (AV) encoding this p27-p16 fusion protein, named W9, inhibited balloon injury-induced neointimal hyperplasia in rabbit carotid arteries. In a therapeutically more relevant model, AV-W9 was delivered to balloon-injured porcine coronary arteries in vivo using an infusion catheter. Of the three coronary arteries, two were injured with a 15-mm balloon catheter and either were left untreated or were treated with 10(12) viral particles of either AV-W9 or a control null virus. AV-W9 treatment significantly inhibited neointimal hyperplasia in this porcine arterial balloon injury model compared with untreated or control virus-treated vessels. The average intimal area of the AV-W9-treated group 10 days after balloon injury and treatment was 0.42+/-0.36 mm(2), whereas the AV-null group demonstrated an intimal area of 0.70+/-0.52 mm(2). At day 10 the average intimal thickness of the AV-W9-treated vessels was 9.1 microm (n=5, x 20 magnification) compared with 21.2 microm (n=5, x 20 magnification) in control virus-treated vessels. This trend was also observed at 28 days after balloon injury and gene transfer during which AV-W9-treated vessels demonstrated an average intimal thickness of 4.7 microm (n=8, x 20 magnification) compared with 13.3 microm (n=3, x 20 magnification) in control virus-treated vessels and 7.3 microm (n=5, x 20 magnification) in the sham-treated vessels. The AV-W9 treatment was safe and well tolerated. These data suggest that AV-W9 gene therapy may be useful in preventing angioplasty-induced intimal hyperplasia in the coronary artery.

Strong correlation between imaging response and survival among patients with primary central nervous system lymphoma: a secondary analysis of RTOG studies 83-15 and 88-06

PURPOSE

Primary central nervous system lymphoma (PCNSL) is the brain tumor with the most rapidly increasing incidence, yet little is known about its radiographic response to cranial irradiation. If traditionally used doses of radiotherapy ( approximately 60 Gy) are associated with low rates of complete response and poor survival, then an argument can be made to consider dose escalation of radiotherapy. Alternatively, if poor survival rates are associated with high rates of complete response, there would be no reason to subject patients to higher radiation doses with increased risks of treatment-related morbidity. The purpose of this analysis is to provide a detailed review of response following cranial irradiation. Based on these findings, recommendations are offered for future protocol design.

METHODS AND MATERIALS

Patients were treated on either RTOG 83-15 (whole brain irradiation to 40 Gy followed by a 20 Gy boost to the tumor plus 2-cm margin) or RTOG 88-06 (induction cyclophosphamide, doxorubicin, vincristine, dexamethasone [CHOD] followed by whole brain irradiation to 41.4 Gy and an 18 Gy tumor boost). Imaging surveillance (CT, MR) was required following surgery, prior to the initiation of RT and following completion of RT. Complete response referred to the absence of enhancement on follow-up scans in comparison to the pretherapy study. A tumor size reduction of at least 50% in the product of the largest cross-sectional diameter and its largest perpendicular diameter was scored as a partial response.

RESULTS

Seventy-nine patients had scans available for central review. Complete response was achieved in 83% and 85% of patients treated on RTOG 83-15 and 88-06, respectively. The rates of partial response (14%, 11%) and radiographic progression (3%, 4%) also were comparable between the studies. For survival analyses, data were aggregated from the two studies. The 4-year survival rates were 24% for complete responders versus 11% for other patients (p = 0.0006). In multivariate analysis, only complete radiographic response (p < 0.0007), and high Karnofsky performance status (KPS >/= 70) (p < 0.005) were independently associated with increased rates of 4-year survival.

CONCLUSION

A high rate of complete radiographic response was observed following moderate doses of cranial irradiation (alone or in combination with CHOD chemotherapy). Although complete responders had a statistically significant survival advantage at 4 years when compared with partial responders and nonresponders, the majority of patients who achieved complete response were dead of disease by 4 years following treatment. Based on this analysis of the RTOG database, there is no rationale for radiation dose escalation as a therapeutic strategy to combat PCNSL. Consequently, the radiotherapy component of the current RTOG Phase II trial (RTOG 93-10) now includes relatively low total doses of hyperfractionated irradiation for patients without residual disease (36 Gy/1.2 Gy, twice a day) as well as a more aggressive chemotherapy regimen.

A phase I/II study of carboplatin combined with hyperfractionated radiotherapy for brainstem gliomas

BACKGROUND

Brainstem gliomas often respond to radiotherapy but long term disease control is exceptional. The concomitant administration of a chemotherapy agent with radiosensitizing properties such as carboplatin may increase the efficacy of radiotherapy.

METHODS

A dose escalation schedule of carboplatin was devised to determine the maximum tolerated dose (MTD) of intravenous carboplatin when given on a twice-weekly schedule during a course of hyperfractionated, involved field radiotherapy (100 centigrays [cGy] twice daily to 7200 cGy). The starting dose was 20 mg/m(2) and the dose was increased by 15 mg/m(2) after every 3 patients provided no Grade 3 or 4 (according to the National Institutes of Health Common Toxicity Criteria) toxicity occurred. Magnetic resonance imaging (MRI) scans (brain and spine) were obtained before treatment and at the time of disease progression. Clinical entry criteria included an MRI scan demonstrating a diffuse intrinsic pontine tumor and a typical 2-3-month history of evolving cranial neuropathies and a gait disorder. Biopsy-confirmed evidence of a high grade glioma was required for nonpontine brain stem tumors.

RESULTS

A total of 34 patients were enrolled. The median age of the patients was 7.8 years (range, 3.6-15.4 years) and the median prodrome duration was 1.5 months (range, 0.25-36 months). The MTD was 110 mg/m(2) or a total cumulative dose of 1540 mg/m(2) over 7 weeks. The dose-limiting toxicity was hematologic. The median progression free survival was 8 months (range, 0-104+ months) and the overall survival was 12 months (range, 5-104+ months). At last follow-up there were 5 long term survivors (15%) who remained in continuous remission after a mean follow-up period of 79 months (range, 46-104 months). Fifteen of the 29 patients (52%) with recurrence and or disease progression developed leptomeningeal/intraaxial tumor spread beyond the local radiation field.

CONCLUSIONS

The cumulative MTD for carboplatin is 1540 mg/m(2) when administered concomitantly with involved field, hyperfractionated radiotherapy in a twice-weekly schedule for 7 weeks. Subsequent Phase II and III clinical trials can be conducted safely at this level.

Intracranial ependymoma in the adult patient: successful treatment with surgery and radiotherapy

PURPOSE

Ependymoma is an uncommon intracerebral tumor in adults. Since the site of origin frequently prevents complete surgical removal, adjunctive radiotherapy is used to destroy residual disease. We present our experience in treating 10 adults with intracranial ependymoma.

METHODS

Five men and 5 women were treated in the past 16 years. The median age was 38 (range 24-69). All had contrast enhanced CT or MRI showing the extent of the tumor. One patient had total excision while the remainder had subtotal removal. Radiation therapy was delivered to the tumor bed with a 1-2 cm margin of normal tissue generally at 180-200 centiGray (cGy) per treatment once a day. Total dose ranged from 5400 to 7200 cGy. Two patients received experimental treatment with 100 cGy delivered twice a day for total of 6800 and 7200 cGy respectively. Four patients received initial treatment to a large field with a subsequent boost to the tumor bed. One patient received his entire course of treatment via this large field.

RESULTS

With a median follow-up of 64 months, 7 patients are alive and free of disease while 2 died of intercurrent disease, without evidence of tumor, at 7 and 9 years following treatment. Another patient died 1 1/2 years after treatment of unknown causes.

CONCLUSION

We conclude that postoperative radiotherapy is effective in preventing regrowth of intracranial ependymoma following subtotal resection in adults. Treatment fields should cover the initial tumor bed with a 1-2 cm margin to avoid long term radiation damage.

Patterns of recurrence in brain stem gliomas: evidence for craniospinal dissemination

PURPOSE

The 3-year survival rate of pediatric patients with infiltrating brain stem gliomas (BSG) is < 10%. Treatment involves local field radiation, and local failure has been the hallmark of recurrence. With therapeutic advances and improved radiographic monitoring, perceived and actual patterns of failure may change. We report patterns of recurrence in a group of patients with close follow-up, treated on an institutional protocol incorporating hyperfractionated involved-field radiation therapy and concomitant carboplatin, who have been uniformly staged and treated and have undergone MRI surveillance.

METHODS AND MATERIALS

From 1990-1995, 18 pediatric patients with BSG were treated on a Phase I-II trial of concurrent carboplatin and hyperfractionated radiotherapy. Eight had surgical procedures to document histology. Nine had hydrocephalus prior to death. All had pretreatment brain and spine MRIs, with and without gadolinium, that showed no other evidence of disease. Treatment consisted of 72.00 Gy involved-field hyperfractionated radiation therapy and dose-escalating concomitant carboplatin.

RESULTS

Fifteen children have had progression of disease (median PFS = 9 months); and 13 have died (median OS = 14 months). Fourteen of the 15 children with progression had local failures, 8 of whom had evidence of noncontiguous spinal (4) or intracranial (7) disease documented by MRI or autopsy. One child with local control developed an intracranial metastasis. None had clinical manifestations of leptomeningeal disease.

CONCLUSION

Leptomeningeal dissemination occurred within 1 month of local progression in nearly 30% of our patients and, overall, occurred in 50% prior to death. This high incidence may reflect close MRI surveillance or a changing pattern of recurrence. Because the majority of leptomeningeal disease occurs in the setting of local progression, treatment efforts must be directed primarily toward local control. However, management of leptomeningeal dissemination at recurrence is of increasing concern.

Multiple myeloma of an extremity: must the entire bone be treated?

PURPOSE/OBJECTIVE

Radiation of the entire shaft of a long bone affected by multiple myeloma (MM) is often advocated to prevent recurrent disease in the bone remote from the symptomatic site. Our standard of care has been to irradiate only the symptomatic area. We investigated the pattern of recurrence in patients treated in this manner.

METHODS AND MATERIALS

163 patient with MM were treated between 1971 and 1994. Twenty-seven patients received treatment to a long bone with 41 sites irradiated (17 humeri, 22 femurs, 1 radius, 1 ulna). The most common long bone treated was the femur. All patients were treated with megavoltage therapy. The symptomatic lesion, plus a margin of 1-2 cm was treated with no attempt to treat the entire shaft. Mean radiation dose was 27.82 Gy (range 6.00-44.80 Gy). The length of the field was measured in centimeters and expressed as both an absolute (AL) and relative (RL) length (i.e., percentage of total length of bone).

RESULTS

The mean total AL and RL for femur fields was 18 cm and 42%, respectively. For the humerus, the AL and RL were 20 cm and 68%, respectively. Only four patients developed progressive disease in the same bone but outside the previously irradiated field. In three of the four patients the RL was between 20 and 30%. The dose of radiation given to these patients was 12.50, 21.00, 30.00, and 35.00 Gy. In all of these four cases, treatment of progressive disease in adjacent sites provided effective palliation of symptoms.

CONCLUSION

Radiation therapy to the symptomatic portion of a long bone affected by MM is effective for palliation. Symptomatic recurrence out of the irradiated field is uncommon and can be effectively treated. Potential benefits of this approach include irradiation of less normal marrow and elimination of use of pairs of fields or extended distance treatment to cover the entire femur.

Tc-99m LL-2 Fab' monoclonal antibody imaging in acquired immune deficiency syndrome-related lymphoma

BACKGROUND

Both systemic and primary central nervous system (CNS) non-Hodgkin's lymphomas (NHL) occur in people with acquired immune deficiency syndrome (AIDS). The radiographic manifestations may be similar to other neoplasms and opportunistic infections that are also found frequently in AIDS. Furthermore, these diseases may coexist with NHL in the AIDS patient.

METHODS

To evaluate the use of Tc-99m Lymphoscan (the Fab' fragment of the anti-CD-22 antibody LL-2; Immunomedics, Inc., Morris Plains, NJ) in patients with suspected AIDS lymphoma, we studied 7 patients with 35 sites of suspected disease. Six had CNS lesions suspicious for parenchymal brain lymphoma. Each patient underwent planar and single photon emission computed tomography imaging at 3-5 and 18-24 hours after administration of Lymphoscan. Scintigraphic results were compared with results of conventional diagnostic modalities.

RESULTS

Overall, the sensitivity of Lymphoscan was 92% and the specificity was 86%. In brain lesions, there was 100% sensitivity and 100% specificity. Lymphoscan also had 100% sensitivity for sites of lymphomatous lymphadenopathy and for liver involvement. Although less specific in extracranial sites, Lymphoscan was correctly negative in sites of coexisting adenocarcinoma and pneumonia. Two patients had both parenchymal CNS and systemic lymphoma proven by biopsy.

CONCLUSIONS

Lymphoscan appears to be a sensitive and specific method for diagnosing CNS lymphoma in AIDS patients. Although slightly less specific in extracranial sites, it may be helpful in differentiating lymphoma from other etiologies in these patients at risk for multiple neoplasms and opportunistic infections.

Thyroid dysfunction as a late effect in survivors of pediatric medulloblastoma/primitive neuroectodermal tumors: a comparison of hyperfractionated versus conventional radiotherapy

BACKGROUND

Primary hypothyroidism is a common sequela of craniospinal radiotherapy in the treatment of pediatric brain tumors.

METHODS

The authors compared the incidence of primary hypothyroidism after hyperfractionated radiotherapy (HFRT) (n = 14 patients) versus conventionally fractionated radiotherapy (CRT) (n = 34 patients) in a group of pediatric patients with medulloblastoma/primitive neuroectodermal tumors (MB/PNET).

RESULTS

The mean age at the time of tumor diagnosis was 7.9 years in the HFRT group and 8.4 years in the CRT group. The patients were followed for a mean of 4.6 years (HFRT) and 8.3 years (CRT) after diagnosis. Mean radiation doses to the thyroid were similar in both radiotherapy groups (29 gray [Gy] [HFRT] vs. 24 Gy [CRT]). Approximately 14% of the HFRT and 62% of the CRT patients developed primary hypothyroidism within a similar period after irradiation (3.2 years [HFRT] vs. 3.0 years [CRT]). Analysis by cumulative incidence function demonstrated a significant difference in the risk of developing thyroid dysfunction between these two groups of patients (P = 0.02).

CONCLUSIONS

The current study findings suggest that the use of HFRT in the treatment of pediatric patients with MB/PNET is associated with a lower risk of these patients developing primary hypothyroidism.

Influence of an oligodendroglial component on the survival of patients with anaplastic astrocytomas: a report of Radiation Therapy Oncology Group 83-02

PURPOSE

Seven percent of patients with high grade gliomas enrolled in RTOG 83-02 had mixed astrocytoma/oligodenroglial elements on central pathology review. It has often been assumed that the most aggressive histologic component of a tumor determines biologic behavior; however in this trial, the survival of patients who had mixed glioblastomas/oligodenrogliomas was significantly longer than that of patients with pure glioblastomas (GBM). We therefore evaluated the effect of an oligodendroglial component on the survival of patients who had anaplastic astrocytomas (AAF) treated in the same trial.

METHODS AND MATERIALS

One hundred nine patients who had AAF and 24 patients with mixed AAF/oligodendrogliomas (AAF/OL) were enrolled in a Phase I/II trial of randomized dose-escalation hyperfractioned radiotherapy plus BCNU. AAF/OL patients were older and more likely to have had more aggressive surgery than AAF patients. Other pretreatment characteristics were balanced between groups, as was assigned treatment.

RESULTS

The median survival time for AAF was 3.0 years versus 7.3 years for AAF/OL (p = 0.019). In a multivariate analysis, adjusting for extent of surgical resection and age, an oligodendroglial component was an independent prognostic factor for survival.

CONCLUSION

The results support the concept that AAFs with an oligodendroglial component have a better prognosis than pure AAF tumors, similar to the effect seen among patients with glioblastoma multiforme tumors. This better survival outcome should be taken into consideration in the design and stratification of future trials. Additionally, in contrast to patients with GBMs, patients who have AAF/OL have the potential for prolonged survival; therefore, late sequelae of treatment (both radiation and chemotherapy) must be weighed more heavily in the benefits to risks analysis.

The effect of M-stage on patterns of failure in posterior fossa primitive neuroectodermal tumors treated on CCG-921: a phase III study in a high-risk patient population

PURPOSE

To analyze patterns of failure in patients (pts) with high-risk posterior fossa primitive neuroectodermal tumors (PF-PNETs) treated with combined modality therapy on a large, randomized multiinstitutional study.

METHODS AND MATERIALS

One hundred eighty-eight prospectively staged pts with PF-PNET confirmed by central pathology review, with high-risk features, were treated on Children's Cancer Group Study 921 (CCG-921), comparing two chemoradiotherapy regimens. Patterns of initial sites of failure were analyzed, specifically evaluating the impact of Chang M-stage.

RESULTS

Progression-free survival (PFS) correlated with the presence or absence of metastatic disease (p < 0.001), with 5-year PFS of 68 +/- 5.8% for M0 vs. 43 +/- 6.8% for M+ pts. The cumulative incidence functions (CIF) of recurrence were different (p = 0.005) and at 5 years were 29 +/- 4.7% for M0 pts and 48 +/- 5.5% for M+ pts. Involvement of the PF at time of initial failure as measured by CIF correlated with M-stage (p = 0.047) and occurred in 18 +/- 3.9% of M0 pts and 8 +/- 2.9% of M+ pts overall; PF as the only site of relapse also correlated with M-stage (p = 0.019) and was seen in 6 +/- 2.5 and 0% of M0 and M+ pts, respectively, at 5 years. Relapse in the spine and/or cerebrospinal fluid (CSF) at initial recurrence was correlated with M-stage (p < 0.002), with 5-year cumulative incidences of 14 +/- 3.7%, 26 +/- 8.2%, 40 +/- 15%, and 40 +/- 7.7% for M0, M1, M2, and M3 pts, respectively. Isolated spine/CSF recurrence correlated with M-stage (p = 0.034) and occurred in 2 +/- 1.5% of M0 and 9 +/- 3.2% of M+ pts by 5 years. The median time to relapse for pts who failed was 1.2 years (range 0.2-5.3). Ninety percent of all relapses occurred by 3 years.

CONCLUSIONS

Original sites of disease are at the highest risk for relapse, but the entire neuraxis remains at significant risk, despite combined-modality treatment. M-Stage was prognostic for spine/CSF relapse as well as PFS and may be an important tool in guiding therapy. A more aggressive approach to local control in the neuraxis is warranted, especially in M+ patients.

Hypothalamic-pituitary-adrenal function following cranial irradiation

We assessed the effect of cranial irradiation on hypothalamic-pituitary (HP)-adrenal function in 17 patients (12 females, 5 males) treated with cranial/ craniospinal irradiation for acute leukemia (2 patients) or tumors distant from the hypothalamus and pituitary (8 medulloblastoma, 3 astrocytoma, 3 rhabdomyosarcoma, 1 ependymoma). Estimated doses of radiation (RT) to the HP region ranged from 18 to 72 Gy. Thirteen of seventeen patients were also treated with chemotherapy. Patients were a median of 3.75 years of age (1.5-19 years) at diagnosis and were studied at a median of 5 years (0.1-20 years) after RT. Patients received corticotropin-releasing factor (oCRF, 1 microgram/kg i.v.), and sampling for cortisol and ACTH levels was performed at -15, 0, 15, 30, 60, 90 and 120 min. The-5- and 0-min levels were combined for a standardized baseline value (Base). Cortisol levels at 0, Base, 30 and 120 min, as well as the peak cortisol response, were significantly lower in the patients. Twelve of seventeen patients' peak cortisol levels fell below the normal range. The patients' mean integrated values for cortisol (area under the curve) were not, however, different from controls. The ACTH responses to oCRF did not differ between patients and controls. No relationship was observed between ACTH or cortisol responses and the time elapsed from treatment or dose of HP RT. Further, in 10 of 12 patients, 0-min dehydroepiandrosterone sulfate levels were lower than the expected normal mean levels for age, sex and pubertal status, and in 4 of these 10 patients the values were below the normal range. These data suggest that some patients treated with HP RT may be at risk for adrenal insufficiency.

Hyperfractionated craniospinal radiotherapy and adjuvant chemotherapy for children with newly diagnosed medulloblastoma and other primitive neuroectodermal tumors

PURPOSE

This single-institution Phase III study conducted from 1989 to 1995 evaluates the feasibility of a multimodality protocol combining hyperfractionated craniospinal radiotherapy (HFRT) followed by adjuvant chemotherapy in 23 patients with newly diagnosed primitive neuroectodermal tumors (PNET) arising in the central nervous system.

METHODS AND MATERIALS

All 23 patients had a histologically confirmed PNET and were over 3 years of age at diagnosis. The eligibility criteria for PNET patients with cerebellar primaries (medulloblastoma) included either a high T stage (T3b or 4) or high M stage (M1-3). All patients with noncerebellar primaries were eligible regardless of T or M stage. The median age of the 23 patients was 9 years (mean 3-25); 11 were female. The primary tumor arose in the cerebellum in 19. Of these medulloblastoma patients, 15 had high T stages (T3b or T4) with large locally invasive tumors and no evidence of metastases (M0), constituting Group 1. Thirteen (86%) of these patients had gross total resections. Four other medulloblastoma patients had both high T and high M stages, constituting Group 2. Group 3 consisted of four other patients with exocerebellar primaries (two brain, one brain stem, and one cauda equina), three of whom were M3. Hyperfractionated radiotherapy was administered within 4 weeks of surgery. Twice-daily 1-Gy fractions were administered separated by 4-6 h. The total dose to the primary intracranial tumor and other areas of measurable intracranial disease was 72 Gy. The prophylactic craniospinal axis dose was 36 Gy, and boosts of 44-56 Gy were administered to metastatic spinal deposits. Following radiotherapy, monthly courses of multiagent chemotherapy were administered sequentially (cyclophosphamide-vincristine followed by cisplatin-etoposide followed by carboplatin-vincristine) for a total of 9 months.

RESULTS

All patients completed radiotherapy as planned. Only three patients lost >10% of their body weight. One patient had clinically apparent radiation-induced esophagitis. The mean white blood count (WBC) nadir was 2.5/dl, and hematologic recovery occurred in all within 4 weeks of completing HFRT without the need of granulocyte-colony-stimulating factor. Two patients refused adjuvant chemotherapy, 3 patients experienced tumor progression during chemotherapy, and 2 of 18 remaining patients could not tolerate the full 9 months owing to hematologic toxicity. Of the 15 patients (93%) in Group 1, 14 remain in continuous remission for a median of 78 months, and none have died. Two of four patients in Group 2 are in continuous remission at 67 and 35 months, and two died at 18 and 30 months. One of the two patients in Group 2 who died refused adjuvant chemotherapy and developed tumor progression in the bone marrow. None of the three patients in Group 3 with evaluable disease (M3) had a complete response to therapy, and eventually all four died of progressive or recurrent disease.

CONCLUSION

This multimodality protocol is feasible in the short term, and long-term monitoring of neurocognitive and neuroendocrine effects are in progress. Excellent long-term disease control has been achieved for medulloblastoma patients with high T stages who were M0 at diagnosis (Group 1), the majority of whom had gross total resections. This group has a progression-free survival of 95% after a median period of follow-up of 6.5 years. Alternative treatment strategies must be developed for patients with high M stages, as five of seven patients died of progressive or recurrent disease.

Preirradiation chemotherapy with cyclophosphamide, doxorubicin, vincristine, and dexamethasone for primary CNS lymphomas: initial report of radiation therapy oncology group protocol 88-06

PURPOSE

This study was a prospective phase I/II trial performed by the Radiation Therapy Oncology Group (RTOG) to test the tolerance and efficacy of preirradiation cyclophosphamide, doxorubicin, vincristine, and dexamethasone (CHOD) chemotherapy followed by large-volume, high-dose brain radiation therapy (RT) for patients with primary CNS lymphoma (PCNSL).

PATIENTS AND METHODS

Fifty-four (52 assessable) human immunodeficiency virus (HIV)-negative patients with PCNSL were entered on study and received two (n = 20) or three (n = 32) cycles of CHOD (six patients with positive CSF cytology received intrathecal methotrexate in addition to CHOD). Whole-brain RT to 41.4 Gy and tumor boost to 18 Gy (total dose, 59.4 Gy) followed chemotherapy.

RESULTS

As of July 1994, with a minimum potential follow-up time of 20 months, 12 of 52 assessable patients remain alive without evidence of progression. The median survival time for the entire group is 16.1 months, with a 2-year survival rate of 42%. By univariate analysis, patient age was found to be a significant prognostic factor with respect to survival (P = .005) in favor of age less than 60 years. Karnofsky performance status (KPS) was of borderline significance (P = .057). Survival for patients treated on RTOG 88-06 was compared with that of patients treated on RTOG 83-15, which tested RT alone. No difference in overall survival was found (P = .53). Grade 4 neutropenia developed in 29 of 51 patients during chemotherapy. There were two deaths during chemotherapy: one as a result of sepsis and one of a pulmonary embolus. The worst toxicity during RT was < or = grade 2 in 50 of 52 patients.

CONCLUSION

Preirradiation CHOD chemotherapy does not significantly improve survival over RT alone for patients with PCNSL. Age remains a powerful prognostic factor independent of therapy and must be considered in testing alternative combined approaches.

Delay in the diagnosis of glioblastoma multiforme: is age a factor?

Discrimination in the delivery of health care, based on the age of the patient, is attracting increased attention. We investigated this problem by studying possible age-related delay in diagnosis in patients treated for glioblastoma multiforme (GBM). A total of 213 charts of patients with GBM seen from 1972 to 1992 were evaluated for type and duration of symptom, with 204 charts having sufficient data to be analyzed. The mean and median duration of symptoms for the entire group was 48.9 and 28 days, respectively. No age-related difference in duration of symptoms was noted. In light of the increasing incidence of GBM in patients above 60 years of age, further investigation of other possible areas of age discrimination in patients with GBM is warranted.

A phase II trial of preirradiation carboplatin in newly diagnosed germinoma of the central nervous system

BACKGROUND

Central nervous system (CNS) germinomas respond readily to both radiotherapy and chemotherapy. This study was designed to selectively reduce the dose of radiotherapy in those patients expressing a complete response (CR) to neoadjuvant carboplatin.

METHODS

A Phase II trial with carboplatin was conducted in 11 newly diagnosed patients with histologically confirmed, radiologically evaluable CNS germinomas before they received radiotherapy. All patients had normal cerebrospinal fluid and serum tumor markers (i.e., human chorionic gonadotropin [HCG] and alpha fetoprotein [AFP]). Seven patients had localized tumors (three pineal, three suprasellar, and one thalamus), and four had multifocal disease. Their median age at diagnosis was 13 years (range, 7-31). One course of carboplatin consisted of 150 mg/m2 weekly for 4 consecutive weeks followed by a 2-week break. Response was evaluated after two courses. If a patient had a CR to chemotherapy, the radiotherapy doses to the involved field and the craniospinal axis were lowered from 50 Gy to 30 Gy and from 36 Gy to 21 Gy, respectively. If less than a CR was observed, two additional courses of chemotherapy were administered, after which the patient's response was reevaluated. Less than a CR required full radiotherapy doses. The radiotherapy volume was determined by the extent of disease at diagnosis (i.e., localized disease was treated with an involved field and craniospinal therapy was used for disseminated disease).

RESULTS

Seven patients had a CR to carboplatin (five patients after two courses and two patients after four courses). Three patients had a partial response (one after four courses and two after two courses). The investigators of the latter two patients chose not to give additional chemotherapy. Another patient opted for radiotherapy after receiving only one course of chemotherapy and was not evaluable for response. Ten of 11 patients remain in continuous remission for a median of 25 months. One patient had a recurrence. He presented with a localized pineal germinoma and had a CR after two courses of carboplatin. He received 30 Gy of involved field radiotherapy and suffered a relapse 5 months later in multiple CNS sites. He died 23 months after diagnosis with diffuse CNS and peritoneal metastases. His serum AFP and HCG levels were elevated, consistent with a nongerminoma germ cell tumor.

CONCLUSIONS

Carboplatin was highly active in treating newly diagnosed CNS germinomas. Further chemotherapy studies eventually may permit additional dose reductions and/or elimination of radiotherapy for patients with CNS germinomas.

Short- and long-term complications of radiation therapy for pediatric brain tumors

Brain tumors are the second most common malignancy of childhood after acute lymphocytic leukemia. Improvements in therapy have led to increased survival. It is estimated that by the year 2000 there will be approximately 200,000 survivors of childhood cancer. A proportion of these will have survived a central nervous system malignancy. As more children survive, complications of treatment are increasingly recognized. This paper reviews the spectrum of radiation-induced complications, both short- and long-term. Their frequency and characteristics will be reviewed as well as suggestions made to decrease their incidence.

Prolactin reduction after combined therapy for prolactin macroadenomas

The ability of surgery or bromocriptine to produce endocrine control of a prolactin macroadenoma decreases as the prolactin level increases. Guidelines for the use of multimodality therapy have not been developed for tumors associated with markedly elevated prolactin levels. We reviewed the records of 21 patients with prolactin levels greater than 200 ng/ml treated by transsphenoidal surgery and postoperative radiotherapy with or without a dopamine agonist. Values before and after treatment were available for 19 patients (13 men and 6 women). The mean basal prolactin level before treatment for the entire group was 2410 ng/ml. Surgery and radiotherapy resulted in a 90% reduction and serum prolactin levels within normal limits in 0 of 7 patients, versus the combination of surgery, radiotherapy, and dopamine agonist, which resulted in a 99.5% reduction and values within the normal range in 12 of 12 patients. Spontaneous physiological improvement was not often observed. One woman and two men were able subsequently to have children. A plan for these patients is discussed.

Treatment of aggressive keratoacanthomas by radiotherapy

Keratoacanthomas infrequently are treated by radiotherapy. However, keratoacanthomas that are recurrent after surgical excision or whose resection would result in cosmetic deformity may benefit from radiotherapy. Between January 1970 and June 1988, 29 such keratoacanthomas in 18 patients were irradiated. Doses ranged from 3500 cGy in 15 fractions to 5600 cGy in 28 fractions. Measured end points of therapy were (1) initial response, (2) freedom from recurrence, and (3) quality of the subsequent cosmetic appearance (scored as good, fair, or poor). No lesion progressed and all eventually regressed completely. Cosmetic results generally were considered good by both the patient and the referring dermatologist; none of the results was considered poor. Our results demonstrate that radiation is an effective means of treating keratoacanthomas.