Analysis of the probability and risk of cause-specific failure

PURPOSE

Kaplan-Meier curves are frequently misused in the analysis of nonsurvival endpoints, such as time to local failure or time to late complications. More appropriate analyses are available and described.

METHODS AND MATERIALS

Cumulative incidence is an unbiased estimate of probability of cause-specific failure. Cumulative conditional probability of cause-specific failure reflects risk to patients remaining at risk. Hazard rates also measure risk.

RESULTS

Kaplan-Meier curves overestimate the probability of late complications when there is a high mortality rate. Cumulative incidence and cumulative conditional probability accurately give the probability and risk of cause-specific failure.

CONCLUSION

Kaplan-Meier analysis of cause-specific failure should be avoided because of its misinterpretation as an estimate of probability, in favor of appropriate methods.

Thymoma. A retrospective study of 87 cases

BACKGROUND

The authors retrospectively analyzed 87 patients with malignant thymoma treated at M.D. Anderson Cancer Center between 1951 and 1990. The analysis examined the clinical stages, histologic types, and treatment modalities and attempted to determine if chemotherapy had an impact on survival.

METHODS

The patients were divided into three groups by their year of treatment and treatment modality. Patients treated between 1951 and 1975 were in Group I; patients treated between 1976 and 1980 were in Group II; and patients treated between 1981 and 1990 were in Group III. Most of the patients (18 [72%] in Group I; 16 [62%] in Group II; and 18 [50%] in Group III) had surgical resection alone or with radiotherapy. Patients with advanced-stage disease in Group I received single-agent chemotherapy, whereas patients with advanced-stage disease in Group II received a different, combination chemotherapy regimen, and those in Group III were treated primarily with cisplatin- and doxorubicin-based combination chemotherapy, e.g., the cyclophosphamide doxorubicin, and cisplatin with or without prednisone. The 17 patients treated with cisplatin with or without prednisone were separately evaluated for survival according to their response.

RESULTS

Twenty-eight patients (5 [20%] in Group I; 6 [23%] in Group II; and 17 [47%] in Group III) received chemotherapy alone or after surgery or radiotherapy. The cisplatin with or without prednisone regimen was used in 17 Group III patients for initial treatment or for relapse. The overall response rate among the patients receiving the cisplatin with or without prednisone regimen was 64%; 6 (35%) had a complete response, and 5 (29%) had a partial response. Thirty-one (36%) of the 87 total patients had 45 recurrent tumors; the lung (29%), pleura (22%), and mediastinum (18%) were the most common sites of recurrence, whereas bone was the most common distant metastatic site. The 5-year survival rate was 70% in patients with Stage I disease, 71% in patients with Stage II or III disease, and 46% in patients with Stage IV disease. The 10-year survival rate was 70% in patients with Stage I disease, 47% in patients with Stage II or III disease, and 21% in patients with Stage IV disease. Statistical analysis indicated a significant difference among the survival rates of patients with noninvasive Stage I, invasive Stage II plus III (P = 0.033), and Stage II plus III and IV tumors (P = 0.056), but not between patients with Stage II or III tumors. Patients with a major response to the cisplatin with or without prednisone regimen had a significant survival improvement compared to those with no response (P = 0.002, log-rank test).

CONCLUSIONS

Because thymoma is a chemosensitive tumor and frequently recurs in patients with Stage II or greater disease, chemotherapy carries a potential survival benefit and should be incorporated into the multimodality approach to prolong disease-free survival.

Dose intensity of radiation therapy in non-small cell carcinoma of the lung: a review of RTOG data and strategies

Prospective trials exploring questions of dose intensity for control of the local-regional tumor have been conducted by the Radiation Therapy Oncology Group for 20 years. Early studies established radiation therapy with a total dose of 60 Gy in 30 fractions as standard (STD) for unresectable non-small cell carcinomas of the lung (NSCCL). Hyperfractionated radiation therapy (HFX) with 1.2 Gy twice daily permitted higher total doses without increased normal tissue effects and led to comparison of STD vs. HFX (total dose 69.6 Gy in 58 fractions). Induction chemotherapy followed by STD, induction followed by concurrent chemotherapy and STD, and concurrent chemotherapy and HFX, have been explored in successive trials. Preliminary results of STD combined with chemotherapy suggest that increases in dose intensity increase local-regional control and survival. Confirmatory trials of increased dose intensity by combining chemotherapy with STD and HFX have been completed or are progressing.

Continuous split course irradiation for stage I and II Hodgkin's disease: 20 year experience at the Medical College of Wisconsin

Between January 1970 and December 1989, 153 consecutive patients with stage I or II Hodgkin's disease were treated at the Medical College of Wisconsin Affiliated Hospitals. Eighty-eight patients were clinically staged and 65 patients underwent staging laparotomy. Ninety-nine patients were treated using continuous split course irradiation (CSCI) technique. Overall survival (OS) at 5, 10, and 15 years was 89%, 79%, and 68%, respectively. Disease-free survival (DFS) at 5, 10, and 15 years was 75%, 72%, and 68%, respectively. No significant differences were found in OS or DFS when comparing pathologically and clinically staged patients or when comparing patients treated using CSCI with those treated with the standard technique. Patients treated using CSCI tended to complete therapy in a shorter time, without increased morbidity.

Hodgkin's disease with a mediastinal mass greater than 10 cm: results of four different treatment approaches

BACKGROUND

Management of Hodgkin's disease (HD) and large mediastinal adenopathy (LMA) usually includes intensive chemotherapy (CT) with or without radiation therapy (XT) regardless of stage.

PATIENTS AND METHODS

One hundred and eighteen evaluable patients received one of four treatment regimens: (1) 6 cycles of MOPP or similar CT and XT; (2) 2 of MOPP followed by XT; (3) 6 of CVPP/ABDIC (cyclophosphamide, vincristine, procarbazine, prednisone/doxorubicin, bleomycin, decarbazine, prednisone, lomustine) followed by XT; or (4) 3 of NOVP (mitoxantrone, vincristine, vinblastine, procarbazine) and XT. XT doses included 30-40 Gy to areas of nodal involvement noted prior to therapy.

RESULTS

Complete remission (CR) rates for groups 1, 2, 3, and 4 were 100%, 85%, 87%, and 96%. Respective 3-year freedom from progression (FFP) results were 88%, 66%, 82%, and 88%, and 3-year freedom from tumor mortality (FTM) results were 100%, 84%, 84%, and 100%. The presence of B symptoms and stage IV disease was correlated with lower CR and 3-year FFP rates but similar 3-year survival.

CONCLUSIONS

Results of this study suggest that patients with stage I-III Hodgkin's disease and LMA greater than 10 cm treated with 3 NOVP and XT have results similar to those obtained for a similar group of patients treated with 2 to 6 MOPP or 6 CVPP/ABDIC and XT. NOVP has also been reported to produce limited toxicity in this trial and should be considered as an alternative to MOPP or doxorubicin-containing regimens in treatment of patients with early-staged disease and LMA greater than 10 cm.

Interruptions of high-dose radiation therapy decrease long-term survival of favorable patients with unresectable non-small cell carcinoma of the lung: analysis of 1244 cases from 3 Radiation Therapy Oncology Group (RTOG) trials

PURPOSE

To determine if prolonged treatment time adversely affects survival for patients with inoperable non-small cell carcinoma of the lung.

METHODS AND MATERIALS

Patients enrolled on three randomized studies (RTOG 8311, 8321, 8403) between 1983-1989 formed the database. Previous analyses found that the addition of thymosin (8321) or prophylactic cranial irradiation (8403) failed to prolong survival: both studies used thoracic irradiation with standard fractionation to 55-60 Gy in 30 fractions. In 8311, patients were treated by hyperfractionated radiation therapy to randomly assigned total doses of 60.0 Gy, 64.8 Gy, 69.6 Gy, 74.4 Gy or 79.2 Gy, 1.2 Gy twice daily, 5 days per week. Patients analyzed received +/- 4% of the assigned total dose and lived > 90 days (to ensure that all patients would have completed treatment). Completion < 5 days beyond protocol specifications was classified as "per protocol." Elapsed treatment time exceeding specifications by 5-9 days was a minor deviation, 10-13 days was a major deviation-acceptable, and > or = 14 days was a major deviation-unacceptable. Absolute survival was the endpoint to evaluate the effect of delays. The log rank statistic was used to test for survival differences in the univariate setting, the Cox regression model was used in the multivariate setting.

RESULTS

Of 293 patients treated with standard fractionation, eight (2.7%) had deviations from the specified treatment time (six minor, two major-acceptable). With hyperfractionation, 90 (15%) patients had deviations (40 minor, 21 major-acceptable, 29 major-unacceptable). As the assigned dose increased, the deviation rate increased (9.7% for 60.0 Gy vs. 20.8% for 79.2 Gy). Survivals for hyperfractionation patients with any deviations in treatment time were significantly shorter than those treated "per protocol" (p = 0.16): estimated 2- and 5-years rates were 24% and 10% versus 13% and 3%, respectively. Multivariate analyses showed the delay effect to be entirely in patients treated with 69.6 Gy or higher; there was also dependence upon the patients' prognosis. In patients with favorable prognosis (KPS 90-100, weight loss < or = 5%, no N3), the difference in survival was pronounced (33% and 15% vs. 14% and 0% at 2- and 5-years, respectively). Such differences were not found in patients with unfavorable prognostic factors.

CONCLUSIONS

Interruptions delaying completion of planned radiation therapy were more frequent with higher total doses (> or = 69.6 Gy). Favorable patients (high KPS, little weight loss, < N3 nodal metastasis) had markedly adverse effects on long-term survival associated with delays to completion of the planned total dose.

The influence of field size and other treatment factors on pulmonary toxicity following hyperfractionated irradiation for inoperable non-small cell lung cancer (NSCLC)--analysis of a Radiation Therapy Oncology Group (RTOG) protocol

PURPOSE

The risk of pulmonary toxicity, observed in an Radiation Therapy Oncology Group Phase I/II randomized dose escalation trial of hyperfractionated irradiation for nonsmall cell lung cancer, was analyzed with regard to custom vs. hand blocking and compliance to protocol specified treatment field parameters.

MATERIALS AND METHODS

There were 832 evaluable cases analyzed. The protocol required field margins 2 cm beyond primary tumor and involved nodes. In 674, the field margins were considered "per protocol" or as having minor protocol variations. In 94, margins exceeded protocol specification ("excessive margin" group). In this group, the area (cm2) of the effective (blocked) field and the portion including lung was measured from simulator films with a computer scanning device. Based on size and location of the primary and nodal disease, "per protocol" fields were constructed and the area (cm2) of lung included beyond these margins was estimated. Patients from both groups who received less than 30 Gy to normal lung were excluded from analysis of pulmonary toxicity.

RESULTS

Grade 1 acute lung toxicity was higher (p = .009) in the "excessive margin" group compared to the "per protocol" group, whereas late lung toxicity was not significantly different (p = .94). The risk of Grade 2 or greater acute toxicity increased as area of excess irradiated lung increased. Overall lung toxicity, defined as the greater of either acute or late toxicity, was evaluated by multivariate analysis, in relation to assigned dose, effective treated field area, and type of shielding. Overall maximum lung toxicity (> or = Grade 2) was significantly greater in the "excessive margin" group, when lung treated beyond protocol margins exceeded an area of 35 cm2, than in the "per protocol" group, but only when the effective treated field size was > or = 180 cm2 (68% vs. 37%; p = .02). This effect was independent of assigned total dose or type of shielding.

CONCLUSION

For nonsmall cell lung cancer treated with hyperfractionated irradiation, the risk of overall pulmonary toxicity was increased for patients treated with field sizes in excess of protocol specified margins of tumor coverage in comparison to patients treated with protocol specified margins. This effect was seen only when the area of lung treated beyond protocol margins exceeded 35 cm2 and when the overall field size was below 180 cm2.

Central lymphatic irradiation for stage III nodular malignant lymphoma: long-term results

PURPOSE

To report the long-term results of central lymphatic irradiation for stage III nodular malignant lymphoma.

PATIENTS AND METHODS

Between 1969 and 1985, 34 patients (26 with nodular poorly differentiated lymphoma, four with nodular mixed lymphocytic/histiocytic lymphoma, and four with nodular histiocytic lymphoma) were treated with central lymphatic irradiation. Median age of the group was 51 years (range, 30 to 73). There were 15 men and 19 women. Staging work-up included a physical examination and bone marrow biopsy in all patients. Seventy-four percent had a lymphangiogram (LAG) and 44% a laparotomy (LAP). Eighty-two percent had either a LAP or a LAG. Thirty-two patients were Ann Arbor stage IIIA and two were stage IIIB. All patients received lymphatic irradiation that encompassed cervical, supraclavicular, axillary, mediastinal, paraaortic, mesenteric, pelvic, and femoral lymphatics to total doses ranging from 20 to 30 Gy in 1.0- to 1.8-Gy fractions. Waldeyer's ring was initially treated in 17 patients. Follow-up information is available on all 34 patients. Median follow-up is 9 years, 8 months (range, 15 to 244 months).

RESULTS

Life-table actuarial overall, disease-free, and cause-specific survival rates at 15 years are 28%, 40%, and 46%, respectively. Only one relapse was observed after 9 years. Disease-free survival was significantly improved in patients with five or fewer sites of involvement (P = .02). Age, sex, B symptoms, histology, and technique of irradiation were not prognostically significant. Salvage therapy, including further irradiation and/or chemotherapy, was delivered to 20 patients. Ten percent of these patients remain alive without evidence of disease. Toxicity data were available for the patients treated at the Medical College of Wisconsin (MCW). Radiation Therapy Oncology Group (RTOG) acute hematologic, gastrointestinal, and salivary toxicity scores were < or = 2 in 83% of patients. Late toxicity scores were < or = 2 in 96%. Persistent xerostomia was noted in 23% of patients who received initial treatment to Waldeyer's ring.

CONCLUSION

These results suggest that initial comprehensive central lymphatic irradiation may be the preferred approach to achieve a durable relapse-free interval for this group of patients.

Comparison of prognostic factors and survival among black patients and white patients treated with irradiation for non-small-cell lung cancer

BACKGROUND

Many studies have reported differences in cancer incidence and survival between populations of Blacks and Whites. A 45% higher death rate from lung cancer for Black men and a survival duration for Black patients with lung cancer that is generally shorter than that for White patients have also been reported.

PURPOSE

The purpose of this study was to evaluate whether race affects known prognostic factors for non-small-cell lung cancer in Black versus White patients. This analysis attempts to determine which prognostic factors may contribute to the reported differences in disease outcome.

METHODS

We used data from 1565 patients with non-small-cell lung cancer treated in four randomized prospective trials conducted by the Radiation Therapy Oncology Group (RTOG). The data were pooled for a retrospective analysis of survival and prognostic factors by race.

RESULTS

Univariate analysis showed significant differences between Blacks and Whites with regard to sex, weight loss, histology, and RTOG T stage (P < .05), but the only clinically significant difference (P < or = .01) was weight loss. Despite these findings, overall survival for Blacks and Whites did not differ significantly (P = .67). Median survival for Blacks and Whites with a Karnofsky performance status (KPS) of 90 or more was 12.1 and 11.3 months, respectively (P = .45). Survival for Blacks and Whites with a KPS of less than 90 was 7.8 and 6.8 months, respectively. Cause of death did not differ between the two races. For both races, KPS, age, sex, weight loss, and RTOG T and N stages were significant prognostic factors for survival (P < .01), but race was not a significant prognostic factor.

CONCLUSION

Further studies of the differential in cancer survival for Blacks and Whites may be indicated, but greater impact may be achieved by addressing socioeconomic factors, lifestyle and occupational risk factors, health education, and access to adequate health care.

Is long-term survival possible with external beam irradiation for stage D1 adenocarcinoma of the prostate?

From 1972 to 1986, 354 patients with local and locoregional adenocarcinoma of the prostate were treated with curative intent at the Medical College of Wisconsin. Fifty-six of these patients were found to have Stage D1 disease (evidence of pelvic lymph node involvement). Using external beam irradiation alone, these patients were treated aggressively to the pelvis followed by a boost to the prostate. The median dose to the prostate was 6800 cGy, and to the pelvis, it was 5040 cGy. The median period of observation after treatment was 9 years. Actuarial survival was 76% at 5 years, and disease-free survival was 61% at 5 years. Twenty-three patients had biopsy-proved pelvic lymph node involvement; the other 33 were considered to have Stage D1 disease based on abnormal computed tomographic scans, lymphangiograms, or both. Actuarial survival and disease-free survival were calculated for both groups separately, and there was no statistical difference in the results. Major complications occurred in 3.6% (two patients) of this group with Stage D1 disease. These results support the continued use of aggressive external beam irradiation in patients with locoregional adenocarcinoma of the prostate.

Interruptions adversely affect local control and survival with hyperfractionated radiation therapy of carcinomas of the upper respiratory and digestive tracts. New evidence for accelerated proliferation from Radiation Therapy Oncology Group Protocol 8313

Hyperfractionated radiation therapy (HFX) attempts to overcome tumor proliferation during treatment by permitting higher total doses in the same overall time as standard fractionation. Whereas interruptions, including splits, reduce local control with standard fractionation in carcinoma of the upper respiratory and digestive tracts, HFX might compensate for interruptions. Patients were randomized to receive total doses of 6720, 7200, 7680, and 8160 cGy, using 120 cGy twice daily, 5 days per week. Those analyzed received +/- 4% of assigned total dose and lived 90 days or more. Treatment was completed within 5 days of the time specified for each treatment arm in 233 patients; 48, 80, and 131 patients had delays 14, 10, and 5 days or more, respectively. Locoregional control and survival were significantly (P less than or equal to 0.03) reduced with delays of 5 days or more when corrected for prognostic factors. Late effects of radiation therapy were not affected by interruptions. These data support the hypothesis that proliferation (possibly accelerated) of tumor clonogens during treatment influences the outcome.

Thymoma. The prognostic significance of flow cytometric DNA analysis

The clinical course of patients with thymoma varies widely despite its histologically benign appearance. Treatment decisions are based on local invasion and the extent of resection. Because some patients have more aggressive tumors, the prognostic significance of flow cytometric (FCM) analysis of nuclear DNA content was examined. Adequate tissue from paraffin-embedded blocks was available for 25 patients. Using FCM, the percentage of cells in S-phase (%S) and the ploidy, based on the DNA index (DI), were determined. The mean patient age was 52 years, with a female-to-male ratio of 1.3:1 and a median follow-up of 64 months. Seventeen patients underwent total tumor resections, and 12 also received radiation therapy. Eight patients underwent subtotal resections, with five receiving radiation therapy (with or without chemotherapy) and three receiving chemotherapy alone. Based on invasion and intrathoracic dissemination, the tumors were classified into four stages. The mean %S was 5.6. There was no relationship observed between %S and patient outcome. The 5-year disease-free survival rate was 85% for the 16 patients with diploid (DI = 1) tumors and 33% for the 9 patients with aneuploid (DI more than 1) tumors (P less than 0.002). Similar significant differences were observed by stage and extent of surgery. For those who had total resection (n = 17), the disease-free survival rate was 89% when DI equaled 1 and 50% when DI was more than 1 (P = 0.01). Although the numbers studied were small, when stage, histologic findings, and type of surgery were subdivided by DI, a higher incidence of relapse was associated consistently with aneuploidy. The DI appears to be a useful prognostic parameter for identifying patients at high risk of relapse.

Thymoma: treatment and prognosis

Thirty-six patients with pathologically confirmed thymoma were treated at M.D. Anderson Cancer Center from 1962 to 1987. The tumors were staged based on invasion and intrathoracic dissemination. Twenty-one patients had total resection, five had subtotal resection, and 10 had biopsy alone. Twenty-two patients had definitive megavoltage radiation therapy with a median dose of 50 Gy. The 5-year, disease-free survival by stage was 74% for Stage I (n = 11), 71% for Stage II (n = 8), 50% for Stage III (n = 10), and 29% for Stage IVA (n = 7) (p less than 0.03). The 5-year, disease-free survival by extent of surgery was 74% for total resection, 60% for subtotal resection and 20% for biopsy only (p = 0.001). There were 15 patients with recurrences: two in Stage I, two in Stage II, five in Stage III, and six in Stage IVA. The median months to relapse, for those who failed treatment, were 46, 36, 2, and 13 for Stages I, II, III, and IVA respectively. Of the patients with recurrences four had a total resection, two subtotal resection, and nine biopsy only. Only one patient had distant metastases as the first site of relapse without intrathoracic relapse. For the eight patients who relapsed following radiation therapy, four were in the radiotherapy field. All four of the in-field failures were in patients who had a partial response. There were insufficient numbers of patients to determine a dose response to radiotherapy. For patients with invasive, incompletely resected disease, a multimodality approach may be necessary for long term, disease-free survival.

The effect of overall treatment time on the outcome of definitive radiotherapy for localized prostate carcinoma: the Radiation Therapy Oncology Group 75-06 and 77-06 experience

From 1976 to 1983, 1091 patients were entered into RTOG protocols 75-06 and 77-06. Of these, 780 patients complied with protocol requirements, received a minimum tumor dose of greater than or equal to 6500 cGy, and received no endocrine therapy. There were 78, 342, and 360 patients with localized prostate carcinoma, Stages T1b(A2), T2(B), and T3,4(C), respectively. The potential follow-up period ranges from 6 years 5 months to 13 years 3 months, with a median follow-up of 9 years. This study examines the influence of overall treatment time on the outcome of definitive radiotherapy for localized prostate carcinoma in this patient population. Within each stage, patients were divided into three groups according to the total number of elapsed days while on treatment: within 49 days (less than or equal to 7 weeks); 50 to 63 days (8 to 9 weeks); and greater than or equal to 64 days (greater than 9 weeks). Based on actuarial analysis, within each stage, the overall treatment time did not have any impact on the following: overall survival, NED survival, or local/regional control. When grouped under different histologic grades, that is, Gleason scores 2-5, 6-7, and 8-10, the actuarial local/regional control showed no statistical difference among the three groups. The actual local/regional failures were analyzed and stratified by stage and Gleason scores, and no statistical difference was noted among the three groups for each stratification. The range of local/regional failure rates among the three groups for T1b(A2), T2(B), and T3,4(C) disease were 0%-8%, 16%-23%, and 24%-27%, respectively. The corresponding range of local/regional failure rates for patients with Gleason scores of 2-5, 6-7, and 8-10 were 13%-14%, 18%-22%, and 22%-33%, respectively. The incidence of late complications was not related to the number of elapsed treatment days. Therefore, the overall treatment time does not have an impact on the outcome of definitive radiotherapy for localized prostate carcinoma. It is hypothesized that prostate carcinoma behaves as late-reacting tissue in which there is little, if any, accelerated repopulation of clonogenic tumor cells during the later half of a protracted course of radiotherapy. This observation is in direct contrast to that suggested for head and neck carcinoma and bears important implications in daily radiotherapeutic management of patients with prostate carcinoma.

Long-term treatment sequelae following external beam irradiation for adenocarcinoma of the prostate: analysis of RTOG studies 7506 and 7706

Significant late intestinal and urinary morbidity from external beam irradiation for adenocarcinoma of the prostate has been a constant concern of both the urologist and the radiation oncologist. We analyzed two large Radiation Therapy Oncology Group trials (7506 and 7706) using primary irradiation in the treatment of local or locoregional adenocarcinoma of the prostate to assess morbidity via the Radiation Therapy Oncology Group scoring scheme (grade 1-5). One thousand twenty patients were treated in total with a minimum follow-up of 7 years in the surviving patients. There was a 3.3% incidence of intestinal complications defined as grade 3 toxicity or more with .6% of patients experiencing bowel obstruction or perforation. Urinary complications defined as grade 3 toxicity or more were found in 7.7% of patients with only 0.5% experiencing morbidity that would require a major surgical intervention such as laparotomy, cystectomy, or prolonged hospitalization. Intestinal and urinary complications were evaluated in reference to several parameters that might have an impact on their incidence (i.e., previous laparotomy, stage of disease, hypertension, positive lymph nodes, previous transurethral resection, total dose, and energy of accelerator used). Only total dose (greater than 70 Gray) was found to have a significant impact on the incidence of the urinary complications. None of these factors had a significant impact on the incidence of intestinal complications. These data from two large multi-institutional trials represent a fair estimate of the actual incidence of major intestinal and urinary complications from external beam irradiation in the management of local and locoregional adenocarcinoma of the prostate. Since the incidence of these major complications remains very low, we believe that external beam irradiation remains an excellent alternative to radical prostatectomy in the management of these patients.

Randomized dose-searching phase ILE/II trials of fractionation in radiation therapy for cancer

This article describes a new design, phase ILE/II dose searching, used in four prospective, randomized, multicenter clinical trials of escalating total doses of hyperfractionated radiation. This design combines an experimental protocol with a statistical application of ranking and selection theory. Its purpose was to identify (within a certain margin of error) a dose that achieved the highest rate of clinical response from a set of doses that were tolerable in terms of both acute (within 90 days) and late (more than 90 days) toxic effects (LE). We calculated the number of patients required to reliably test toxicity under various assumptions. To determine the maximum tolerated total dose for hyperfractionated radiation, we randomly assigned patients with tumors that responded to radiation therapy in a dose-dependent manner from four body sites (lung, upper respiratory and digestive tract, bladder, and brain) to one of three regimens receiving total doses (D1, D2, or D3) differing by increments of 4.8 Gy. All patients received two fractions of 1.2 Gy each (separated by 4-6 hours) daily 5 days a week. The lowest total dose was set at the level considered tolerable with standard once-a-day radiation therapy. We tested tumor responses and late toxic effects of higher doses by assigning patients to these three regimens until acute effects and early estimates of late effects were found to be acceptable for the highest dose D3; thereafter, regimen D1 was closed, and additional patients were assigned to D2, D3, and D4 (an escalated total dose greater than D3 by an increment of 4.8 Gy). The assignment of patients was performed in a weighted manner (1:1:2), so that greater numbers were assigned to the highest dose regimen (whether D3 or D4) to allow rapid evaluation of the feasibility of the highest dose.

Five-year survival after hyperfractionated radiation therapy for non-small-cell carcinoma of the lung (NSCCL): results of RTOG protocol 81-08

RTOG Protocol 81-08, a feasibility study of hyperfractionated radiation therapy (HFX) with 1.2 Gy twice daily separated by 4-6 hours for non-small-cell cancer of the lung (NSCCL), was completed in 1983. Encouraging short-term results in a recently closed trial of HFX for NSCCL (RTOG 83-11) led to assessment of long-term outcome in the earlier trial. Of 120 evaluable patients who were assigned to total doses from 50.4 Gy to 74.4 Gy, all 5 of the 5-year survivors came from the 79 patients assigned to receive 69.6 Gy. The 5-year survival rates for the 79 patients were 14.3 +/- 9.4% for clinical RTOG Stage II, 5.9 +/- 4.0% for Stage III, and 3.2 +/- 3.2% for Stage IV. Combined Stage II and III 5-year survival rates were 8.3 +/- 4.0% for HFX 69.6 Gy compared to 5.6 +/- 1.5% for standard once-a-day irradiation in concurrent RTOG trials.

Comparison of the Radiation Therapy Oncology Group and American Joint Committee on Cancer staging systems among patients with non-small cell lung cancer receiving hyperfractionated radiation therapy. A report of the Radiation Therapy Oncology Group protocol 83-11

Since 1973, the Radiation Therapy Oncology Group (RTOG) has staged and stratified patients in non-small cell lung cancer (NSCLC) protocols according to the RTOG staging system. In 1985, the American Joint Committee on Cancer (AJCC) revised its lung cancer staging system, with the principle differences from the RTOG system being the staging of involvement of the chest wall and of contralateral mediastinal and hilar lymph nodes. To determine if the AJCC system discriminated outcome differently than the RTOG system in a nonoperative series, all 850 evaluable patients treated with hyperfractionated radiation therapy (RT) on the RTOG protocol 83-11 were restaged by the AJCC system. There was 67% agreement in patient distribution between the following comparable stages in each system: RTOG Stage II/AJCC Stage II; RTOG Stage III/AJCC Stage IIIA; and RTOG Stage IV/AJCC Stage IIIB. Both systems successfully predicted for survival (P less than 0.001), although the RTOG staging was more discriminating (relative risk ratios, 1.59 versus 1.38). Among the 507 favorable patients (those with less than or equal to 5% weight loss and Karnofsky performance status [KPS] of 70 to 100), the RTOG staging was also more predictive (P = 0.004 versus P = 0.01). When RTOG Stage III (462 patients) was divided into those without contralateral mediastinal or hilar adenopathy (AJCC Stage II/IIIA) and those with (AJCC Stage IIIB), a significant survival (P = 0.0001) was noted with 2-year survival rates of 26% versus 4%, respectively. When AJCC Stage IIIA (348 patients) was divided into the patients without chest wall invasion (RTOG Stage II/III) and those with (RTOG Stage IV), a difference in 2-year survival of 22% versus 10% was observed (P = 0.002). Although both staging systems independently predict for survival, a fusion of both staging systems is the most discriminating of outcome. Future nonoperative studies in locally advanced NSCLC should stratify for contralateral nodal involvement (per AJCC staging) and chest wall invasion (per RTOG staging).