Graph diffusion distance: Properties and efficient computation.

We define a new family of similarity and distance measures on graphs, and explore their theoretical properties in comparison to conventional distance metrics. These measures are defined by the solution(s) to an optimization problem which attempts find a map minimizing the discrepancy between two graph Laplacian exponential matrices, under norm-preserving and sparsity constraints. Variants of the distance metric are introduced to consider such optimized maps under sparsity constraints as well as fixed time-scaling between the two Laplacians. The objective function of this optimization is multimodal and has discontinuous slope, and is hence difficult for univariate optimizers to solve. We demonstrate a novel procedure for efficiently calculating these optima for two of our distance measure variants. We present numerical experiments demonstrating that (a) upper bounds of our distance metrics can be used to distinguish between lineages of related graphs; (b) our procedure is faster at finding the required optima, by as much as a factor of 103; and (c) the upper bounds satisfy the triangle inequality exactly under some assumptions and approximately under others. We also derive an upper bound for the distance between two graph products, in terms of the distance between the two pairs of factors. Additionally, we present several possible applications, including the construction of infinite "graph limits" by means of Cauchy sequences of graphs related to one another by our distance measure.

In vitro studies of the rabbit immune system. V. Suppressor T cells activated by concanavalin A block the proliferation, not the induction of antierythrocyte plaque-forming cells.

The late B-cell proliferative phase of the in vitro antibody response by rabbit spleen cells is highly susceptible to suppression by activated T cells. The in vitro antisheep erythrocyte plaque-forming cell (PFC) response by spleen cells from normal or primed rabbits can be suppressed by adding concanavalin A (Con A), Con A-prestimulated peripheral blood or spleen lymphocytes, or supernates from Con A-prestimulated peripheral blood lymphocytes. The suppression is not mediated by a direct interaction of Con A with responding cells as shown by the effectiveness of prestimulated cells. Primed spleen cultures remain sensitive to Con A suppression as late as 72 h after initiation, and the addition of Con A after 24-72 h rapidly stops the increase in the number of PFC. T cells are required for Con A addition to be effective but the suppression can be induced at a time when T-helper cells are no longer necessary. Further, the suppressive effect of Con A addition is abrogated by specific antisera to rabbit T cells. We propose that Con A activates suppressor T cells which then exert their effects on proliferating PFC or their immediate precursor B cells. The early inductive or recruitment phase of the response is probably not blocked by suppressor cells. Also, there is an apparent relationship between the number of proliferating B cells and the number of suppressor cells required. Finally, the difficulties in inducing a stimulatory effect by Con A and the prolonged period that Con A addition is suppressive suggests that the rabbit has relatively more and/or longer-lived suppressor cells than the mouse and may be a particularly useful species for studying suppressive phenomena and their mechanisms.

An immunohistologic study of Shope fibroma virus in rabbits: tumor rejection by cellular reaction in adults and progressive systemic reticuloendothelial infection in neonates.

The development of lesions in adult and neonatal New Zealand White rabbits following intradermal inoculation of Shope fibroma virus was studied by immunofluorescence for viral antigens. T-cells, and immunoglobulin. In adults a self-limiting local fibroxanthosarcomatous tumor was rejected within 10-12 days in association with a dense infiltration of T-cells. In neonates expanding skin lesions were associated with systemic presence of virus in the reticuloendothelial system. In surviving infected neonates, granulomas formed at the site of infection after 3 weeks. These reactions may have limited further dissemination of the virus. These results support the hypothesis that the progressive disease produced by Shope fibroma virus in neonatal rabbits may be due to the inability of the reticuloendothelial system to clear infectious virus.

Cell-mediated immune response to Shope fibroma virus-induced tumors in adult rabbits.

The parameters of cell-mediated immune responses of adult rabbits infected with Shope fibroma virus (SFV) were characterized by measurement of the size of local draining nodes, number of cells per lymph node, mitogen responses of lymphocytes, and kinetics of virus-specific cell-mediated lymphocytotoxicity (CML). In addition, the cytolytic effector population was characterized. After intradermal injections, tumors appeared within 3-4 days, reached maximum size in 10-12 days, and then regressed completely with 24 days. The size of local popliteal lymph nodes, in particular the diffuse cortex (paracortex), and the number of cells per node increased during tumor growth but then declined as the tumor regressed. Maximum specific CML to SFV-infected kidney cell monolayers (RK-13) occurred 10 days after inoculation of SFV and correlated with the initiation of tumor regression. Adult cytotoxic lymphocytes passed through nylon wool, and most of their activity was removed by treatment with antithymocyte globulin plus complement. Cytotoxic T-cells from SFV tumor-bearing rabbits killed only targets infected with SFV and not targets uninfected or infected with vaccinia virus. Therefore, T-cell-mediated virus-specific CML appeared as a major immune effector mechanism that correlated with tumor regression. However, antibody-dependent cell-mediated and NK cytotoxicity were also demonstrable. The presence of different cell-mediated cytotoxic mechanisms suggested a heterogeneity of effector mechanism.

Recursive partitioning analysis of prognostic factors in three Radiation Therapy Oncology Group malignant glioma trials.

BACKGROUND: Despite notable technical advances in therapy for malignant gliomas during the past decade, improved patient survival has not been clearly documented, suggesting that pretreatment prognostic factors influence outcome more than minor modifications in therapy. Age, performance status, and tumor histopathology have been identified as the pretreatment variables most predictive of survival outcome. However, an analysis of the association of survival with both pretreatment characteristics and treatment-related variables is necessary to assure reliable evaluation of new approaches for treatment of malignant glioma. PURPOSE: This study of malignant glioma patients used a non-parametric statistical technique to examine the associations of both pretreatment patient and tumor characteristics and treatment-related variables with survival duration. This technique was used to identify subgroups with survival rates sufficiently different to create improvements in the design and stratification of clinical trials. METHODS: We used a recursive partitioning technique to analyze survival in 1578 patients entered in three Radiation Therapy Oncology Group malignant glioma trials from 1974 to 1989 that used several radiation therapy (RT) regimens with and without chemotherapy or a radiation sensitizer. This approach creates a regression tree according to prognostic variables that classifies patients into homogeneous subsets by survival. Twenty-six pretreatment characteristics and six treatment-related variables were analyzed. RESULTS: The years). Patients younger than 50 years old were categorized by histology (astrocytomas with anaplastic or atypical foci [AAF] versus glioblastoma multiforme [GBM]) and subsequently by normal or abnormal mental status for AAF patients and by performance status for those with GBM. For patients aged 50 years or older, performance status was the most important variable, with normal or abnormal mental status creating the only significant split in the poorer performance status group. Treatment-related variables produced a subgroup showing significant differences only for better performance status GBM patients over age 50 (by extent of surgery and RT dose). Median survival times were 4.7-58.6 months for the 12 subgroups resulting from this analysis, which ranged in size from 32 to 256 patients. CONCLUSIONS: This approach permits examination of the interaction between prognostic variables not possible with other forms of multivariate analysis. IMPLICATIONS: The recursive partitioning technique can be employed to refine the stratification and design of malignant glioma trials.

Fluorescent visualization of binding and internalization of the anticarcinogenic Bowman-Birk type protease inhibitors in transformed fibroblasts.

The Bowman-Birk protease inhibitors from soybeans and chick peas suppress in vitro malignant transformation. This fluorescent microscopy study is designed to visualize the cellular site of action of these protease inhibitors. Binding and internalization of active protease inhibitors occur over a time course of 2 h. The rate and character of fluorescent micrographs obtained are compared to insulin as a positive control. Internalization of both molecules is completely blocked at 4 degrees C. Interpretation of the fluorescent micrographs in conjunction with biochemical data suggests the anticarcinogenic action of Bowman-Birk type protease inhibitors may involve receptor-mediated endocytosis resulting in the internalization of both the protease inhibitors and a membrane-associated protease.

Survival comparison of radiosurgery-eligible and -ineligible malignant glioma patients treated with hyperfractionated radiation therapy and carmustine: a report of Radiation Therapy Oncology Group 83-02.

PURPOSE: The purpose is twofold: (1) to identify the malignant glioma patients treated in a trial of hyperfractionated radiotherapy (RT) and carmustine (BCNU) who may have been eligible for a stereotactic radiosurgery (SRS) boost; and (2) to compare survival of such patients with that of those considered SRS-ineligible. PATIENTS AND METHODS: From January 1983 to July 1989, 778 malignant glioma patients were enrolled on Radiation Therapy Oncology Group (RTOG) 83-02, a randomized phase I/II hyperfractionated RT dose-escalation trial with BCNU chemotherapy. The SRS criteria used in a single-institution trial were applied to these patients; they are: Karnofsky performance status (KPS) of greater than 60; well-circumscribed tumor less than 4.0 cm; no subependymal spread; and a location not adjacent to brainstem or optic chiasm. RESULTS: Eighty-nine patients (11.9%) were identified as potentially SRS-eligible. The median survival times (MST) and 18-month survival rates of the 89 eligible and 643 ineligible patients were 14.4 versus 11.7 months and 40% versus 27%, respectively (P = .047). The MST and 18-month survival rate of the 544 SRS-ineligible patients with KPS greater than 60 were 12.1 months and 29%, respectively, and were not statistically inferior to the survival of the SRS-eligible group (P = .21). Multivariate analysis revealed age, KPS, and histopathology to be strongly predictive of survival, and SRS eligibility was also significantly predictive (P = .047). CONCLUSION: SRS-eligible patients enrolled on RTOG 83-02 had survival superior to that of the SRS-ineligible group, and this advantage is mainly due to the selection of a subgroup with a high minimum KPS.

Issues in quality of life assessment during cancer therapy.

Current treatments for cancer do not differentiate between malignant and normal cells; this limitation results in the adverse effects associated with radiotherapy and chemotherapy. Adverse effects of treatment severely impact the cancer patient's quality of life. Quality of life assessment can help the clinician gauge the efficacy of interventions that reduce the adverse effects of radiotherapy and chemotherapy. Quality of life is an increasingly important outcome measure in the evaluation of cancer treatments, and a variety of tools have been developed for evaluating changes in quality of life. This report reviews several of these measurement tools, focusing primarily on those used in lung cancer trials.

Psychoeducational group increases vaginal dilation for younger women and reduces sexual fears for women of all ages with gynecological carcinoma treated with radiotherapy.

PURPOSE: The association between radiotherapy for gynecological carcinoma and sexual dysfunction is well established. Regular vaginal dilation is widely recommended to these women as a way for them to maintain vaginal health and good sexual functioning. However, the compliance rate with this recommendation is low. The purpose of this study was to test the effectiveness of a group psychoeducational program based on the "information-motivation-behavioral skills" model of behavior change in increasing the rate of compliance. METHODS AND MATERIALS: Thirty-two women with Stage I or II cervical or endometrial carcinoma who were being treated with radiotherapy were randomized and received either the experimental group program or the control intervention that consisted of written information and brief counseling. Outcome measures included global sexual health, knowledge about sexuality and cancer, fears about sexuality after cancer, and vaginal dilation compliance. RESULTS: Younger women attending the experimental program (44.4%) were significantly more likely to follow recommendations for vaginal dilation than those who received the control intervention (5.6%). Women, regardless of age, who received the experimental intervention reported less fear about sex after cancer treatment. The older women who received the experimental intervention gained more sexual knowledge. There was no evidence that the experimental intervention improved global sexual health. CONCLUSIONS: This is the first controlled study to provide evidence of an intervention's effectiveness 1. in increasing women's vaginal dilation following radiotherapy for gynecological carcinoma and 2. in reducing their fears about sex after cancer. Most women, particularly younger women, are unlikely to follow the recommendation to dilate unless they are given assistance in overcoming their fears and taught behavioral skills.

Cancer pain management by radiotherapists: a survey of radiation therapy oncology group physicians.

PURPOSE: Radiation Therapy Oncology Group (RTOG) physicians were surveyed to determine their approach to and attitudes toward cancer pain management. METHODS AND MATERIALS: Physicians completed a questionnaire assessing their estimates of the magnitude of pain as a specific problem for cancer patients, their perceptions of the adequacy of pain management, and their report of how they manage pain in their own practice setting. RESULTS: Eighty-three percent believed the majority of cancer patients with pain were undermedicated. Forty percent reported that pain relief in their own practice setting was poor or fair. Assessing a case scenario, 23% would wait until the patient's prognosis was 6 months or less before starting maximal analgesia. Adjuvants and prophylactic side effect management were underutilized in the treatment plan. Barriers to pain management included poor pain assessment (77%), patient reluctance to report pain (60%), patient reluctance to take analgesics (72%), and staff reluctance to prescribe opioids (41%). CONCLUSIONS: Physicians' perceptions of barriers to cancer pain management remain quite stable over time, and physicians continue to report inadequate pain treatment education. Future educational efforts should target radiation oncologists as an important resource for the treatment of cancer pain.

Results of a randomized comparison of radiotherapy and bromodeoxyuridine with radiotherapy alone for brain metastases: report of RTOG trial 89-05.

PURPOSE: To determine if the addition of bromodeoxyuridine (BrdUrd) to radiotherapy prolongs survival when compared to radiotherapy alone in patients with brain metastases. METHODS AND MATERIALS: Seventy-two patients with brain metastases were randomized to 37.5 Gy in 15 fractions of 2.5 Gy or to the same dose with BrdUrd 0.8 g/m2 per day for 4 days of each of 3 weeks. Drug treatment was begun on Thursday or Friday before the first week of radiotherapy. Patients had a Karnofsky performance score of at least 70, a neurological function classification of 1 or 2, and any primary tumor except central nervous system (CNS), leukemia, or lymphoma. The primary was absent, controlled, or under active radiotherapy. Patients were free of other metastases. They were stratified by primary site (breast, lung or other), number of metastases (single or multiple) and age (< 60 vs. > 60). RESULTS: There was no significant difference between the two treatment arms (p = 0.904). The study was open from October 1989 to March 1993 and accrued 72 patients. Only one patient in the RT only arm remains alive. The two treatment arms were balanced with respect to all stratification variables. Toxicity due to radiotherapy was similar in both arms. BrdUrd caused significant Grade 4 and 5 hematologic and skin toxicity in five patients. Two patients died due to hematologic toxicity and one from a Stevens-Johnson type skin reaction. Phenytoin played a role in the skin reactions and ranitidine was associated with the hematologic deaths. Ranitidine was eliminated, BrdUrd was discontinued after any hematologic toxicity, and no further Grade 4 or 5 toxicities were seen. The median survival was 6.12 months in the radiotherapy group and 4.3 in the BrdUrd group (p = 0.904). Patients with solitary brain metastases had significantly better survival (p = 0.031). CONCLUSIONS: BrdUrd did not enhance the efficacy of the radiotherapy regimen tested, in spite of the fact that brain metastases have shown high labeling indices. The toxicity of this schedule of BrdUrd administration was apparently increased by ranitidine and phenytoin.

The effect of local-regional control on distant metastatic dissemination in carcinoma of the head and neck: results of an analysis from the RTOG head and neck database.

A retrospective analysis of the effect of local control on the development of distant metastases was performed in 2648 patients with carcinoma of the head and neck selected from the RTOG database. The 5-year time-adjusted incidence of distant metastases was 21% for patients who were in local-regional control at 6 months after the start of treatment, compared to 38% for local-regional failure patients (p less than 0.001). The incidence of distant metastases detected between the interval of 6 months to 2.5 years after treatment was significantly increased in patients with tumors of the oral cavity, oropharynx, supraglottic larynx, and glottis who developed local-regional failure within this time period, compared to those who remained locally controlled (19% distant metastases for local-regional failure vs 7% for local-regional control (p less than 0.001)). In contrast, there as no difference in the incidence of distant metastases in patients with carcinoma of the nasopharynx or hypopharynx regardless of the local-regional disease status. A Cox proportional hazards regression analysis demonstrated that local-regional control was the most significant variable affecting the development of distant metastases, followed by tumor site, N-stage, and T-stage. For all tumor sites, except for the hypopharynx and nasopharynx, improvements in local-regional control are likely to improve survival. Tumors of the hypopharynx and nasopharynx have a higher probability of micro-metastatic dissemination at the time of initial diagnosis, and until effective methods to treat disseminated disease are developed, the effect of local control on survival will not be readily discerned.

Addition of chemotherapy to radiation therapy alters failure patterns by cell type within non-small cell carcinoma of lung (NSCCL): analysis of radiation therapy oncology group (RTOG) trials.

PURPOSE: To evaluate the influence of cell type within non-small cell carcinoma of lung (NSCCL) on failure patterns when chemotherapy (CT) is combined with radiation therapy (RT). METHODS AND MATERIALS: Data from 4 RTOG studies including 1415 patients treated with RT alone, and 5 RTOG studies including 350 patients also treated with chemotherapy (RT + CT) were analyzed. Patterns of progression were evaluated for squamous cell carcinoma (SQ) (n = 946), adenocarcinoma (AD) (n = 532) and large cell carcinoma (LC) (n = 287). RESULTS: When treated with RT alone, SQ was more likely to progress at the primary site than LC (26% vs. 20%, p = 0.05). AD and LC were more likely to progress in the brain than SQ (20% and 18% vs. 11%, p = 0.0001 and 0.011, respectively). No differences were found in intrathoracic and distant metastasis by cell type. When treated with RT + CT, AD was less likely to progress at the primary than either SQ or LC (23% vs. 34% and 40%, respectively; p = 0.057 and 0.035). AD was more likely than SQ to metastasize to the brain (16% vs. 8%, p = 0.03), and other distant sites (26% vs. 14%,p = 0.019). No differences were found in intrathoracic metastasis. LC progressed at the primary site more often with RT + CT than with RT alone (40% vs. 20%, p = 0.036). Death with no clinical progression was more likely with SQ than AD or LC for RT alone and RT + CT (p < 0.01). Brain metastasis was altered little by the addition of CT, but other distant metastases were significantly decreased (p < 0.001) in all cell types by the addition of CT. CONCLUSION: CT, although effective in reducing distant metastasis in all types of NSCCL, has different effects on the primary tumor by cell type, and has no effect on brain metastasis or death with no progression. Different treatment strategies should be considered for the different cell types to advance progress with RT + CT in NSCCL.

Validation and predictive power of Radiation Therapy Oncology Group (RTOG) recursive partitioning analysis classes for malignant glioma patients: a report using RTOG 90-06.

PURPOSE: The recursive partitioning analysis (RPA) classes for malignant glioma patients were previously established using data on over 1500 patients entered on Radiation Therapy Oncology Group (RTOG) clinical trials. The purpose of the current analysis was to validate the RPA classes with a new dataset (RTOG 90-06), determine the predictive power of the RPA classes, and establish the usefulness of the database norms for the RPA classes. PATIENTS AND METHODS: There are six RPA classes for malignant glioma patients that comprise distinct groups of patients with significantly different survival outcome. RTOG 90-06 is a randomized Phase III study of 712 patients accrued from 1990 to 1994. The minimum potential follow-up is 18 months. The treatment arms were combined for the purpose of this analysis. There were 84, 13, 105, 240, 150, and 23 patients in the RPA Classes I-VI from RTOG 90-06, respectively. RESULTS: The median survival times (MST) and 2-year survival rates for the six RPA classes in RTOG 90-06 are compared to those previously published. The MST and 2-year survival rates for the RTOG RPA classes were within 95% confidence intervals of the 90-06 estimates for Classes I, III, IV, and V. The RPA classes explained 43% of the variation (squared error loss). By comparison, a Cox model explains 30% of the variation. The RPA classes within RTOG 90-06 are statistically distinct with all comparisons exceeding 0.0001, except those involving Class II. A survival analysis from a prior RTOG study indicated that 72.0 Gy had superior outcome to literature controls; analysis of this data by RPA classes indicates the survival results were not superior to the RTOG database norms. CONCLUSION: The validity of the model is verified by the reliability of the RPA classes to define distinct groups with respect to survival. Further evidence is given by prediction of MST and 2-year survival for all classes except Class II. The RPA classes explained a good portion of the variation in survival outcome in the data. Lack of correlation in RPA Class II between datasets may be an artifact of the small sample size or an indication that this class is not distinct. The validation of the RPA classes attests to their usefulness as historical controls for the comparison of future Phase II results.

Failure patterns by prognostic group determined by recursive partitioning analysis (RPA) of 1547 patients on four radiation therapy oncology group (RTOG) studies in inoperable nonsmall-cell lung cancer (NSCLC).

PURPOSE: To identify groups of patients who might benefit from more aggressive systemic or local treatment, based on failure patterns when unresectable NSCLC was treated by radiation therapy (RT) alone. METHODS: From 4 RTOG trials, 1547 patients treated by RT alone were analyzed for patterns of first failure by RPA class defined by prognostic factors, including KPS, weight loss, nodal stage, pleural effusion, age and radiation therapy dose. All patients had NSCLC AJCC Stage II, IIIA, or IIIB, KPS > 50, with no previous RT or chemotherapy. Progressions in the primary (within irradiated fields), thorax (outside irradiated area, but within thorax), brain and distant metastasis other than brain were compared (2-sided) for each failure category by RPA. RESULTS: The RPA classes were 4 distinct subgroups that had significantly different median survivals of 12.6, 8.3, 6.3 and 3.3 months for Classes I, II, III and IV, respectively, (all groups, p = 0.0002). There were 583, 667, 249 and 48 patients in Classes I, II, III and IV, respectively. Primary failure was seen in 27%, 25%, 21% and 10% for Classes I, II, III, and IV, respectively (I vs. IV, p = 0.014; II vs. IV, p = 0.022). Distant metastasis, including brain metastasis, occurred at significantly higher rates among Classes I and II (58% and 54%) than in Classes III and IV (42% and 27%). A higher rate (58%) of death without an identifiable site of failure was found in Class IV than in Classes I, II and III (27%, 28% and 36%, respectively). CONCLUSIONS: The data suggest that physiologic compromise from the intrathoracic disease in Class IV patients is sufficient to cause death before specific sites of failure became evident. Clinical investigations using treatments directed at specific sites of failure could lead to improved outcome for Class I, II and, possibly, Class III patients. Inclusion of Class IV patients in clinical trials may obscure outcomes.

Is a surgical resection leaving positive margins of benefit to the patient with locally advanced squamous cell carcinoma of the head and neck: a comparative study using the intergroup study 0034 and the Radiation Therapy Oncology Group head and neck database.

PURPOSE: The purpose of this study was to determine whether or not for patients with squamous cell carcinomas of the head and neck, a surgical resection leaving positive margins followed by postoperative adjuvant therapy improves the outcome compared to a matched group of patients treated with definitive radiotherapy alone. METHODS AND MATERIALS: From January 1985 through January 1990 a consortium of national cooperative groups (Radiation Therapy Oncology Group, Cancer and Leukemia Group B, Eastern Cooperative Oncology Group, Northern California Oncology Group, Southeast Group, and Southwest Oncology Group) conducted a phase III clinical trial testing the efficacy of adjuvant chemotherapy for patients with resectable, squamous cell carcinomas of the head and neck. One hundred and nine patients were excluded from this study due to positive surgical margins. These patients have been followed prospectively with regards to local/regional tumor control, development of distant metastases, and survival. The postoperative treatment of these patients was not specified by the protocol but the majority of patients received postoperative radiotherapy +/- chemotherapy. These patients were compared with a matched group of patients from the Radiation Therapy Oncology Group head and neck database of patients treated with definitive radiotherapy alone using a standard fractionation schema. Matching parameters included primary tumor site, T-stage, N-stage, Karnofsky performance status, and age. RESULTS: Actuarial curves are presented for local/regional control and survival. At 4 years the local/regional control rate is 44% for the positive margin patients compared to 24% for the patients from the data base (p = 0.007). However, there is no significant difference between the survival curves (p = 0.76) with respective median survivals being 18.1 months vs. 17.9 months and 4-year survivals being 29% vs. 25%. CONCLUSION: While an incomplete excision followed by postoperative therapy does not seem to improve survival compared to treatment with radiotherapy alone, it appears to yield significantly better local/regional control. This would argue for its applicability in selected palliative settings. A follow-up, Phase III trial for patients with advanced tumors may be warranted to test traditional resectability criteria.

Adjuvant chemotherapy for resectable squamous cell carcinomas of the head and neck: report on Intergroup Study 0034.

To test the efficacy of sequential chemotherapy as an adjuvant to surgery and postoperative radiotherapy for patients with locally-advanced but operable squamous cell cancers of the head and neck region, a randomized clinical trial was conducted under the auspices of the Head and Neck Intergroup (Radiation Therapy Oncology Group, Southwest Oncology Group, Eastern Oncology Group, Cancer and Leukemia Group B, Northern California Oncology Group, and Southeast Group). Eligible patients had completely resected tumors of the oral cavity, oropharynx, hypopharynx, or larynx. They were then randomized to receive either three cycles of cis-platinum and 5-FU chemotherapy followed by postoperative radiotherapy (CT/RT) or postoperative radiotherapy alone (RT). Patients were categorized as having either "low-risk" or "high-risk" treatment volumes depending on whether the surgical margin was greater than or equal to 5 mm, there was extracapsular nodal extension, and/or there was carcinoma-in-situ at the surgical margins. Radiation doses of 50-54 Gy were given to "low-risk" volumes and 60 Gy were given to "high-risk" volumes. A total of 442 analyzable patients were entered into this study with the mean-time-at-risk being 45.7 months at the time of the present analysis. The 4-year actuarial survival rate was 44% on the RT arm and 48% on the CT/RT arm (p = n.s.). Disease-free survival at 4 years was 38% on the RT arm compared to 46% on the CT/RT arm (p = n.s.). At 4 years the local/regional failure rate was 29% vs. 26% for the RT and CT/RT arms, respectively (p = n.s.). The incidence of first failure in the neck nodes was 10% on the RT arm compared to 5% on the CT/RT arm (p = 0.03 without adjusting for multiple testing) and the overall incidence of distant metastases was 23% on the RT arm compared to 15% on the CT/RT arm (p = 0.03). Treatment related toxicity is discussed in detail, but, in general, the chemotherapy was satisfactorily tolerated and did not affect the ability to deliver the subsequent radiotherapy. Implications for future clinical trials are discussed.

Randomized study of preoperative versus postoperative radiation therapy in advanced head and neck carcinoma: long-term follow-up of RTOG study 73-03.

This is a report of a 10-year median follow-up of a randomized, prospective study investigating the optimal sequencing of radiation therapy (RT) in relation to surgery for operable advanced head and neck cancer. In May 1973, the Radiation Therapy Oncology Group (RTOG) began a Phase III study of preoperative radiation therapy (50.0 Gy) versus postoperative radiation therapy (60.0 Gy) for supraglottic larynx and hypopharynx primaries. Of 277 evaluable patients, duration of follow-up is 9-15 years, with 7.6% patients lost to follow-up before 7 years. Loco-regional control was significantly better for 141 postoperative radiation therapy patients than for 136 preoperative radiation therapy patients (p = 0.04), but absolute survival was not affected (p = 0.15). When the analysis was restricted to supraglottic larynx primaries (60 postoperative radiation therapy patients versus 58 preoperative radiation therapy patients), the difference for loco-regional control was highly significant (p = .007), but not for survival (p = 0.18). In considering only supraglottic larynx, 78% of loco-regional failures occurred in the first 2 years. Thirty-one percent (18/58) of preoperative patients failed locally within 2 years versus 18% (11/60) of postoperative patients. After 2 years, distant metastases and second primaries became the predominant failure pattern, especially in postoperative radiation therapy patients. This shift in the late failure pattern along with the increased number of unrelated deaths negated any advantage in absolute survival for postoperative radiation therapy patients. The rates of severe surgical and radiation therapy complications were similar between the two arms. Because of an increased incidence of late distant metastases and secondary primaries, additional therapeutic intervention is required beyond surgery and postoperative irradiation to impact significantly upon survival.

Induction cisplatin/vinblastine and irradiation vs. irradiation in unresectable squamous cell lung cancer: failure patterns by cell type in RTOG 88-08/ECOG 4588. Radiation Therapy Oncology Group. Eastern Cooperative Oncology Group.

PURPOSE: To analyze disease failure patterns by pretreatment characteristics and treatment groups in a prospective randomized trial. METHODS AND MATERIALS: Patients with medically inoperable Stage II, unresectable IIIA and IIIB nonsmall cell lung cancer with KPS > or =70 and weight loss < or =5% were randomized to one of three treatment groups: standard radiation therapy with 60 Gy at 2.0 Gy per day (STD RT), induction chemotherapy with cisplatin 100 mg/m2 days 1 and 29 with vinblastine 5 mg/m2 weekly for 5 weeks followed by 60 Gy at 2.0 Gy per day (CT + RT), or hyperfractionated radiation therapy with 69.6 Gy at 1.2 Gy b.i.d. (HFX RT). Of 490 patients enrolled, 458 were evaluable. Minimum and median periods of observation for this analysis were 4 years and 6 years, respectively. RESULTS: Pretreatment characteristics were equally distributed. Toxicities were previously reported. Median survival rates were 11.4, 13.6, and 12.3 months for STD RT, CT + RT, and HFX RT, respectively (log rank p = 0.05, Wilcoxon p = 0.04). Survivals were 20, 31, and 24% at 2 years, and 4, 11, and 9% at 4 years in the STD RT, CT + RT, and HFX RT groups, respectively. There were no differences in local tumor control rates among the treatments. Patterns of first failure showed less distant metastasis (DM) (other than brain) for CT + RT compared to the RT alone arms (p = 0.04). Within squamous cell carcinoma (SCC), DM (other than brain) rates were 43%, 16%, and 38% in SCC for STD RT, CT + RT, and HFX RT, respectively (p = 0.0015). Patients with peripheral/chest wall lesions were significantly more likely to fail first in the thorax when treated on STD RT compared to CT + RT and HFX RT (p = 0.009). Survival rates were similar among the treatment arms for patients with squamous cell carcinoma. Among patients with nonsquamous cell carcinoma, failure patterns did not differ by treatment group, but survival was significantly better in those who were treated by induction chemotherapy (p = 0.04). CONCLUSION: Patients with squamous cell carcinoma treated on the CT + RT arm had a significant reduction of first DM other than brain, but there was difference in survival. Survival favored CT + RT in nonsquamous carcinoma despite similar failure patterns. Reasons for improved survival with CT + RT in NSCLC are not yet available.

Influence of bromodeoxyuridine radiosensitization on malignant glioma patient survival: a retrospective comparison of survival data from the Northern California Oncology Group (NCOG) and Radiation Therapy Oncology Group trials (RTOG) for glioblastoma multiforme and anaplastic astrocytoma.

PURPOSE: To examine the effect of treatment using Bromodeoxyuridine (BrdU) during radiation therapy on malignant glioma patient survival by comparing historical survival data from several large clinical trials. METHODS: A retrospective analysis of patient data from Radiation Therapy Oncology Group (RTOG) trials 74-01, 79-18, and 83-02 and the Northern California Oncology Group (NCOG) study 6G-82-1 was conducted. Patient data was supplied by both groups, and analyzed by the RTOG. Pretreatment characteristics including age, extent of surgery, Karnofsky Performance Status (KPS), and histopathology were collected; the only treatment variable evaluated was the use of BrdU during radiation therapy. Radiation dose, dose-fractionation schedule, use of chemotherapy, and/or type of chemotherapy was not controlled for in the analyses. Univariate and multivariate analyses were conducted to examine the potential treatment effect of BrdU on patient survival. RESULTS: Data from 334 patients treated with BrdU on NCOG 6G-82-1 and 1743 patients treated without BrdU on 3 RTOG studies was received. Patients were excluded from the review if confirmation of eligibility could not be obtained, if the patient was ineligible for the study they entered, if central pathology review was not done, or if radiotherapy data was not available. Patients treated according to the RTOG studies had to start radiotherapy within 4 weeks of surgery; no such restriction existed for the NCOG studies. To ensure comparability between the studies, patients from the NCOG studies who began treatment longer than 40 days from surgery were also excluded. The final data set included 296 cases from the NCOG studies (89%) and 1478 cases from the RTOG studies (85%). For patients with glioblastoma multiforme (GBM) the median survival was 9.8 months in the RTOG studies and 13.0 months in the NCOG trial (p < 0.0001). For patients with AA the median survival was 35.1 months for the RTOG studies and 42.8 months in the NCOG trial (p = 0.126). Univariate results showed consistent results favoring BrdU among patients over 30 years of age, across the extent of surgery, and for GBM patients. A proportional hazards regression model that included treatment, histopathology, KPS, age, and extent of surgery demonstrated that treatment with BrdU was included in the best model only for the GBM group of patients (risk ratio 0.83). CONCLUSIONS: Because of the heterogeneity of the treatment groups, including potentially important differences in pathology reviewers assessment of nonglioblastoma cases, differences in radiation dose and schedules, and chemotherapy during or after radiation, these analyses cannot provide the definitive answer as to whether BrdU given during radiation therapy improves survival in patients with malignant glioma. There does appear to be a favorable treatment effect seen in patients with GBM, with a lesser effect in patients with AA.