Randomized phase 2 evaluation of preoperative radiation therapy and preoperative chemotherapy with mitomycin, vinblastine, and cisplatin in patients with technically unresectable stage IIIA and IIIB non-small cell cancer of the lung. LCSG 881

Between June 1988 and January 1980, 67 patients with pathologic stage III non-small cell lung cancer were randomized to receive either preoperative mitomycin, vinblastine, and cisplatin (MVP) chemotherapy (cisplatin 120 mg/m2, and mitomycin, 8 mg/m2 day 1 + 29, and vinblastine, 4.5 mg/m2 on day 1, 15, 22, and 29 and 2.0 mg/m2 day 8), or preoperative radiotherapy (44 Gy in 22 fractions to the primary tumor and mediastinum). The purpose of this study was to identify a treatment approach that showed sufficient effectiveness and acceptable toxicity to warrant testing by prospective randomized trial against "standard" nonsurgical treatment. All patients had surgical staging of the mediastinum and had either unresectable N2 disease or T4 disease with proximal extension of disease along the pulmonary artery. Response to preoperative therapy was evaluated 8 weeks after beginning treatment and patients with complete or partial radiographic response were to undergo surgical exploration and resection if possible. Fifty-seven patients were eligible and evaluable for response. Of the 67 total patients, 3 were unavailable for follow-up, 4 were ineligible, 1 was canceled, and 2 refused all treatment after having been randomized. Of the eligible and evaluable patients, 49 had stage IIIA and 8 had stage IIIB disease. Randomization was to MVP in 26 cases and to radiotherapy (XRT) in 31. Radiographic response to treatment was virtually identical for the two approaches, with 29 of the 57 evaluable patients achieving objective responses. In patients achieving radiographic response, 24 underwent surgical exploration and 20 underwent resection, of which 18 were complete. The mediastinum was free of tumor in seven patients but only two pathologic complete responses were seen (one each to XRT and MVP). In addition, ten nonresponders underwent surgery; seven underwent resection. Median survival for the entire group is 12 months, with a 27% actuarial survival at 4 years. Two patients died of treatment toxicity during preoperative therapy. Overall toxicity included 2 preoperative toxic deaths and 6 postoperative deaths in 34 patients who underwent surgical exploration (3 each with XRT and MVP) due to adult respiratory distress syndrome (3), myocardial infarction (1), pulmonary edema (1), and esophageal fistula (1), for an overall death rate 8 of 57 (14%) and a perioperative death rate in surgically explored patients of 6/34 (18%). These preoperative regimens, in the population studied herein, were of modest efficacy and substantial toxicity.(ABSTRACT TRUNCATED AT 400 WORDS)

Intrapleural chemotherapy without pleurodesis for malignant pleural effusions. LCSG Trial 861

Malignant pleural effusions are a common and significant problem in patients with advanced malignancies. In contrast to traditional sclerosing agents, intrapleural chemotherapy has the potential advantage of treating the underlying malignancy, in addition to treating the effusion. The Lung Cancer Study Group evaluated intrapleural cisplatin and cytarabine in patients with malignant pleural effusions from a variety of solid tumors. Forty-six patients with cytologically proven symptomatic and previously untreated malignant pleural effusions were entered. Cisplatin, as a single dose of 100 mg/m2, plus cytarabine 1,200 mg, were instilled into the pleural space via a chest tube that was then immediately removed. The overall response rate, complete plus partial at 3 weeks, was 49% (18/37 patients). One patient experienced reversible grade 3 renal toxic reactions, four patients had grade 3 hematologic toxic reactions, and five patients had grade 3 cardiopulmonary toxic reactions. Median length of response was 9 months for a complete remission and 5.1 months for a partial remission. Although chemotherapy has the potential advantage of treating the underlying malignancy in addition to controlling the malignant effusion, intracavitary cisplatin and cytarabine therapy as administered in this trial appears inferior to existing sclerosing agents for the control of malignant pleural effusions. Although administration is safe, it cannot be recommended for the standard control of malignant pleural effusions, but it may have a role incorporated into combination modality therapies for diseases such as malignant pleural mesothelioma.

Preoperative chemotherapy (cisplatin and fluorouracil) and radiation therapy in stage III non-small cell lung cancer. A phase 2 study of the LCSG

STUDY OBJECTIVE

To determine the feasibility, toxicity, and potential efficacy of neoadjuvant chemoradiotherapy before surgery in patients with non-small cell lung cancer limited to the chest.

DESIGN

Phase 2 pilot study.

SETTING

Multi-institutional, multimodality cooperative group.

PATIENTS

Eight-five patients with advanced stage III-A or minimal stage III-B non-small cell lung cancer in whom attempted resection would have been likely to leave residual disease.

INTERVENTION

Cisplatin, 75 mg/m2, was given on days 1 and 29; fluorouracil, 1 g/m2 for 24 h, was given as a continuous infusion on days 1 to 4 and 29 to 32; thoracic radiation, 30 Gy in 15 fractions, was administered on days 1 to 19. Thoracotomy with tumor resection was planned for day 57.

MEASUREMENTS AND RESULTS

Two patients achieved a complete and 46 achieved a partial response after the neoadjuvant chemoradiotherapy for an overall response rate of 56%. Toxicity was moderate but acceptable. Fifty-four patients underwent thoracotomy and tumor resection was attempted in 44; 29 (34%) had complete and 15 (18%) had incomplete resections. There was no apparent increase in postoperative complications. In eight patients (9%), no viable tumor was detected pathologically in the resection specimen. Of the 18 patients whose tumors were completely resected and had disease recurrence, none had recurrence only in the chest, 15 (83%) had recurrence in distal sites, and 3 (17%) developed second primary tumors. Median survival of all patients was 13 months.

CONCLUSIONS

This neoadjuvant regimen did not appear to provide major benefit in patients with advanced but potentially resectable non-small cell lung cancer. Further studies are needed to better define the relative roles of preoperative radiotherapy and chemotherapy.

The adverse effect of perioperative blood transfusion in lung cancer

Perioperative blood transfusion appears to increase the risk of recurrence and death in patients with surgically resected lung cancer. This finding is consistent with that in other cancers and several studies in lung cancer report similar risk elevations. We have reanalyzed the Lung Cancer Study Group data relevant to this question, assessing the potential confounding effects of some prognostic factors not examined previously. The results are nearly identical to those reported earlier, suggesting that increased risk is attributable to blood transfusion and not to confounding by known prognostic factors.

A prospective randomized trial to determine the benefit of surgical resection of residual disease following response of small cell lung cancer to combination chemotherapy

Three hundred twenty-eight patients with limited stage small cell lung cancer were enrolled in a trial to evaluate surgical treatment for such patients responding to chemotherapy. Cyclophosphamide, doxorubicin, and vincristine were administered every 21 days for five cycles. Patients achieving at least partial response who had confirmation of pure small cell histologic features by pathology review and who were fit enough for thoracotomy were randomized to undergo or not to undergo pulmonary resection. All randomized patients received radiotherapy to the chest and brain. Two hundred seventeen (66%) of the patients achieved objective response (90 complete response; 127 partial response). One hundred forty-six patients were randomized (66% of responders, 44% of all patients): 70 to surgery and 76 to no surgery. Results of surgery were 83% resection rate, 19% pathologic complete remission rate, and 9% with residual non-small cell histologic features only, for a total of 28% eradication of small cell lung cancer. The survival curves for the two arms are not different (log rank p = 0.78). Median survivals were 12 months for all enrolled patients and 16 months for those who were randomized. Actuarial 2-year survival is 20%. The results of this trial do not support the addition of pulmonary resection to the multimodality treatment of small cell lung cancer.

Quality of life in lung cancer surgical adjuvant trials

The Functional Living Index-Cancer (FLIC) was administered to 438 patients in the Lung Cancer Study Group on whom long-term follow-up was available in 1993. Across all trials, the total FLIC score was predictive for survival even when corrected for extent of disease, although individual items on the FLIC were not. There was no significant impact of a short course of chemotherapy on quality of life. The FLIC is a reliable means of assessing quality of life in lung cancer surgical adjuvant trials.

Impact of neuroendocrine differentiation in non-small cell lung cancer. The LCSG experience

Non-small cell lung cancers with neuroendocrine differentiation (NSCLC-NE) may demonstrate biologic behavior intermediate between non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) with impact on prognosis. We studied the expression of four well-defined neuroendocrine (NE) markers: neuron-specific enolase (NSE), chromogranin A, Leu-7, gastrin-releasing peptide, and a panel of three non-NE markers, including vimentin, and the epithelial markers carcino-embryonic antigen (CEA) by immunohistochemistry, and mucin by histochemistry in 237 resected NSCLCs from patients on six LCSG protocols. Twenty-nine (12%) tumors were positive for 2 or more NE markers. An NE differentiation score was calculated but failed to correlate with recurrence as did other combinations of markers. However, the presence of tissue staining for CEA was strongly associated with improved survival (p = 0.011), whereas the presence of mucin was associated with a worse outcome (p < 0.001). Individually, CEA and mucin remained prognostic even when corrected for stage, histologic features, and performance status. We conclude that NE differentiation is not predictive of recurrence in patients with resected NSCLC but data on patterns of CEA and mucin expression may improve prognostication and permit rational design of new therapeutic approaches.

p53 immunostaining positivity is associated with reduced survival and is imperfectly correlated with gene mutations in resected non-small cell lung cancer. A preliminary report of LCSG 871

We investigated the correlation of p53 abnormalities with survival in 85 patients with non-small cell lung cancer (NSCLC) who had undergone resection with curative intent as part of Lung Cancer Study Group (LCSG) 871. Our previous studies showed that only a subset of p53 mutations in lung cancers result in overexpression. In addition, protein overexpression has been described in the absence of mutation. Therefore, we determined both p53 protein overexpression (by immunostaining) and p53 and ras gene mutations (by single-strand conformation polymorphism and DNA sequencing) in this set of resected tumor specimens. Clinical follow-up data were available for 75 cases. Of the studied patients, 64% showed p53 overexpression and 51% had mutant p53 sequences; however, the concordance rate was only 67%. There was a negative survival correlation with positive p53 immunostaining (p = 0.05), but not with the presence of gene mutations (p = 0.62) in this group of patients. Overexpression of p53 protein determined by immunostaining may contribute to adverse outcome due to the ability of p53 to act as a dominant oncogene, or alternatively, overexpression may reflect ongoing DNA damage in the tumor as a marker for a more aggressive behavior. When adjusted for stage, age, and gender by multivariate analysis, however, there was no independent impact of p53 overexpression on survival.

The early detection of second primary lung cancers by sputum immunostaining. LCEWDG Investigators. Lung Cancer Early Detection Group

STUDY OBJECTIVE

To determine whether monoclonal antibody (Mab) detection of tumor-associated antigen expressed on sputum epithelial cells precedes clinical presentation of second primary lung cancer. DESIGN SETTING/PARTICIPANTS: Eleven oncology centers collaborate in the accrual of 1,000 patients with stage I non-small cell lung cancer (NSCLC) who had undergone resection. The Mabs examined in this study (624H12, 703D4) detect two promising oncofetal/differentiation markers (ie, a difucosylated Lewis X and a 31-Kd glycoprotein antigen).

INTERVENTIONS

Induced sputum specimens are evaluated for quality, then are Papanicolaou and immunostained by independent central laboratories at enrollment and annually thereafter. The predictive value of Mab markers is compared with routine morphologic study for detection of second primary lung cancer during an anticipated 3 years of accrual and 1 year of follow-up.

MEASUREMENTS AND RESULTS

Five hundred eighty of an anticipated 1,000 patients have been accrued on schedule. Patients are primarily white (88.6%), former smokers (75.9%), men (55.6%), with a median age of 66.7, and joined the study at an average of 3.7 years following resection of a stage 1 NSCLC (34.4% squamous, 43.6% adenocarcinoma). Central laboratories found less dysplasia and more unsatisfactory specimens (27.3%) than do the accrual institution laboratories. Immunostaining identifies more suspicious cells than does morphologic study. However, only two second primary lung cancers (eight total deaths) have occurred to date.

CONCLUSIONS

Halfway through the accrual, we describe the study design and preliminary observations. This study illustrates rational selection of carcinogenesis markers by linkage of marker expression on preneoplastic specimens with subsequent expression on tumor tissue.

Prospective assessment of 30-day operative morbidity for surgical resections in lung cancer

Prospective morbidity and mortality rates associated with resection of lung cancer that are reflective of the current trend toward preoperative therapy are not readily available in the current literature. To determine their prevalence, we prospectively analyzed the results of 783 resections performed within contributing Lung Cancer Study Group (LCSG) centers. There were 543 men and 240 women with a mean age of 63.44 years. Of the 783 resections, there were 411 lobectomies, 135 pneumonectomies, and 237 other procedures. Thirty patients died postoperatively (mortality, 3.8%) and 211 had a major complication (27%). Complications occurred more commonly in men (34.3%, p = 0.001), in patients age 60 or older (34.0%, p = 0.001), and in patients with a Karnofsky index < 9 (44%, p < 0.001). There was no significant difference between mortality, significant morbidity rates for lobectomy (28.2%), and pneumonectomy (31.9%), or for simple (28.3%) and extended resection (31.9%). The seemingly higher incidence of major postoperative events reported in this series not only reflects the prospective nature of this analysis but also the fact that over 25% of patients were in other therapeutic trials involving neoadjuvant or postoperative adjuvant regimens. Within that context, these data appear to be a reasonable estimate of modern surgical morbidity rates in the treatment of lung cancer.

Are informed consent forms that describe clinical oncology research protocols readable by most patients and their families?

PURPOSE

This study was conducted to assess the readability of informed consent forms that describe clinical oncology protocols.

METHODS

One hundred thirty-seven consent forms from 88 protocols that accrued patients at The Johns Hopkins Oncology Center were quantitatively analyzed. These included 58 of 99 (59%) institutional protocols approved by The Johns Hopkins Oncology Center's Clinical Research Committee and the Institutional Review Board (IRB) over a 2-year period, and 30 active Eastern Cooperative Oncology Group (ECOG), Radiation Therapy Oncology Group (RTOG), and Pediatric Oncology Group (POG) trials. The consent forms described phase I (17%), phase I/II (36%), phase III (29%), and nontherapeutic (18%) studies. Each was optically scanned, checked for accuracy, and analyzed using readability software. The following three readability indices were obtained for each consent form: the Flesch Reading Ease Score, and grade level readability as determined by the Flesch-Kincaid Formula and the Gunning Fog Index.

RESULTS

The mean +/- SD Flesch Reading Ease Score for the consent forms was 52.6 +/- 8.7 (range, 33 to 78). The mean grade level was 11.1 +/- 1.67 (range, 6 to 14) using the Flesch-Kincaid Formula and 14.1 +/- 1.8 (range, 8 to 17) using the Gunning Fog Index. Readability at or below an eighth-grade level was found in 6% of the consent forms using the Flesch-Kincaid Formula and in 1% using the Gunning Fog Index. Readability was similar for consent forms that described institutional, cooperative group, and phase I, II, and III protocols.

CONCLUSION

Consent forms from clinical oncology protocols are written at a level that is difficult for most patients to read, despite national, cooperative group, institutional, and departmental review. The consent process, which is crucial to clinical research, should be strengthened by improving the readability of the consent forms.

Predictors of acute respiratory failure after bone marrow transplantation in children

OBJECTIVE

To determine factors associated with acute respiratory failure after bone marrow transplantation which can be identified before the onset of lung disease.

DESIGN

Population-based, retrospective study.

SETTING

A referral-based pediatric intensive care unit and bone marrow transplant center.

PATIENTS

Thirty-nine patients with lung disease (abnormal chest radiograph or a need for supplemental oxygen) were identified from a group of 318 pediatric bone marrow transplant patients from 1978 to 1988. Thirty-four of 39 patients with complete data were further classified into patients with mild lung disease (recovery without needing endotracheal intubation, n = 16) and patients with acute respiratory failure (requirement for endotracheal intubation, n = 18).

INTERVENTIONS

Regression analyses were performed to define risk factors for development of respiratory failure (multivariate logistic regression) and for a shortened interval between the identification of lung disease and respiratory failure (Cox proportional hazards analysis).

MEASUREMENTS AND MAIN RESULTS

Ninety-three percent (15/16) of patients with mild lung disease survived. Conversely, only 9% (2/23) of patients with respiratory failure survived. Predictors of respiratory failure included graft vs. host disease (odds ratio 28.3, 95% confidence interval 1.9-421, p = .015), a prelung disease (baseline) circulating creatinine concentration of > 1.5 mg/dL (> 132.6 mumol/L) (odds ratio 28.4, 95% confidence interval 1.4-577, p = .029), and male gender (odds ratio 14.6, 95% confidence interval 1-210, p = .049). Predictors of a shortened time to onset of respiratory failure included baseline serum creatinine value of > 1.5 mg/dL (> 132.6 mumol/L) (hazard ratio 6.2, 95% confidence interval 1.5-26.5, p = .013) and baseline total bilirubin concentration > 1.4 mg/dL (> 23.9 mumol/L) (hazard ratio 4.5, 95% confidence interval 0.98-20.7, p = .053). The median time to onset of respiratory failure was 4 days in patients with baseline creatinine values > or = 1.5 mg/dL (> 132.6 mumol/L) and 5 days in patients with baseline bilirubin concentrations > or = 1.4 mg/dL (> 23.9 mumol/L) vs. > 26 days in patients with creatinine < 1.5 mg/dL (< 132.6 mumol/L) and > 29 days in patients with bilirubin < 1.4 mg/dL (< 23.9 mumol/L) (Kaplan-Meier analysis).

CONCLUSIONS

Renal and liver dysfunction preceded clinical evidence of lung disease in bone marrow transplant patients who developed respiratory failure. Lung disease leading to respiratory failure and adult respiratory distress syndrome appears to develop as one component of the multiple organ failure syndrome in pediatric bone marrow transplant patients.

Allelic loss of chromosome 18q and prognosis in colorectal cancer

BACKGROUND

Colorectal cancer occurs in approximately 150,000 people each year in the United States. Prognostic assessment influences the treatment of patients with colorectal cancer, including decisions about adjuvant therapy. We evaluated chromosome 18q allelic loss, a genetic event associated with tumor progression, as a prognostic marker for this disease.

METHODS

We developed procedures to examine the status of chromosome 18q with microsatellite markers and DNA from formalin-fixed, paraffin-embedded tumors. Allelic loss of chromosome 18q was assessed in 145 consecutively resected stage II or III colorectal carcinomas.

RESULTS

Among patients with stage II disease, the five-year survival rate was 93 percent in those whose tumor had no evidence of allelic loss of chromosome 18q and 54 percent in those with allelic loss; among patients with stage III disease, survival was 52 and 38 percent, respectively. The overall estimated hazard ratio for death in patients whose tumor had chromosome 18q allelic loss was 2.83 (P = 0.008) according to univariate analysis. Furthermore, chromosome 18q allelic loss remained a strong predictive factor (hazard ratio for death, 2.46; 95 percent confidence interval, 1.06 to 5.71; P = 0.036) after adjustment for all other evaluated factors, including tumor differentiation, vein invasion, and TNM stage.

CONCLUSIONS

The status of chromosome 18q has strong prognostic value in patients with stage II colorectal cancer. The prognosis in patients with stage II cancer and chromosome 18q allelic loss is similar to that in patients with stage III cancer, who are thought to benefit from adjuvant therapy. In contrast, patients with stage II disease who do not have chromosome 18q allelic loss in their tumor have a survival rate similar to that of patients with stage I disease and may not require additional therapy.

CD4+ lymphocyte response to zidovudine as a predictor of AIDS-free time and survival time

It is unknown whether the early rise in CD4+ lymphocyte count seen in zidovudine-treated patients is associated with increased AIDS-free time and survival time. To determine the association of this and other changes in immunologic and hematologic markers with prognosis for time to AIDS and survival, we followed 747 AIDS-free patients from initiation of zidovudine therapy in the Multicenter AIDS Cohort Study (MACS). Participants were seen semiannually and had data collected on medication use, immunologic and hematologic variables, and clinical outcomes. AIDS was diagnosed in 216 participants and 165 died during the median follow-up period of 2.0 years. Duration of zidovudine use was categorized into 1-6 months (after 6 months of follow-up) and 7-12 months (12 months of follow-up). During the 6-month follow-up period in which zidovudine was first used, CD4+ lymphocyte levels rose by 17 cells/microliters compared with a mean decrease of 30 cells/microliters/6 months in untreated individuals (after adjusting for baseline CD4+ lymphocyte count). Baseline levels of CD4+ lymphocytes, platelets, and hemoglobin were significant predictors of AIDS and death. After 1-6 months with patients taking zidovudine and controlling for baseline (pretreatment) variables, change in markers after 6 months of follow-up and use of Pneumocystis carinii pneumonia prophylaxis and other antivirals, change in CD4+ lymphocyte count significantly predicted progression to AIDS [relative hazard (RH) = 0.71 for each 100 cell/microliters increase, p = 0.0001] and survival (RH = 0.78 per 100 cell/microliters increase; p = 0.004).(ABSTRACT TRUNCATED AT 250 WORDS)

Guidelines for analysis and reporting of clinical trials in oncology

When analyzing and reporting the results of clinical trials, investigators should follow a simple approach. The purpose of a trial is to estimate an effect or treatment difference, which if present would have clinical utility when treating new patients. Procedures or methods that do not facilitate precisely and impartially estimating and reporting the treatment effect are likely to mislead investigators. Most often in clinical trials, investigators are interested in estimates of risk ratios (specifically odds or hazard ratios) between the treatment groups or levels of a prognostic factor. These simple ideas suggest that the most useful results from clinical trials will be estimated risk ratios and their confidence limits. Especially in cancer, where disease progression, recurrence, and death are common events following treatment, estimates of risk difference are very relevant. Hypothesis tests and associated P-values, although often (or exclusively) reported, are of lesser utility because they do not fully summarize the data. These recommendations may be seen by some investigators to be contrary to accepted practice. It is true that they are somewhat contrary to common practice but their general acceptance is evident in many journals and presentations by clinical trial methodologists. Despite some disagreement among statisticians regarding the need for adjustment of analyses for imbalanced prognostic factors, it is helpful to see if treatment effects change after accounting for imbalances. When this occurs, it may be of clinical interest. Although we discourage analyses that exclude any patients who meet the eligibility criteria, some circumstances will require that this be done (e.g., when a patient refuses to participate after randomization). Investigators should report, and emphasize as primary, those analyses that include all eligible patients. It is our hope and belief that analysis and reporting of trial results along the guidelines suggested here will result in impartial and useful information for journal readers.

A comparison of the power of two tests for qualitative interactions

'Qualitative' or 'crossover' interactions arise when a new treatment, compared with a control treatment, is beneficial in some subsets of patients and harmful in other subsets. We present a new range test for crossover interactions and compare it with the likelihood ratio test developed by Gail and Simon. The range test has greater power when the new treatment is harmful in only a few subsets, whereas the likelihood ratio test has greater power when the new treatment is harmful in several subsets. We provide power tables for both tests to facilitate sample size calculations for designing experiments to detect qualitative interactions and for interpreting the results of clinical trials.