Granulocyte colony-stimulating factor stimulates recovery of granulocytes in patients receiving dose-intensive chemotherapy without bone marrow transplantation.

Bone marrow colony-stimulating factors (CSF) ameliorate hematologic toxicity of standard chemotherapy regimens and may allow relatively safe use of intensive and more efficacious doses of anticancer drugs. Twenty-four patients with cancers for which no standard regimens were likely to be effective received repeated courses of a combination of cisplatin (150 mg/m2), etoposide (1,500 mg/m2), and cyclophosphamide (5,000 mg/m2) at doses for which bone marrow transplantation is usually used. A total of 10 patients received escalating doses of recombinant human granulocyte CSF (rhG-CSF); 11 patients receiving identical chemotherapy and supportive therapy without rhG-CSF served as controls for the first cycle of therapy. Five of these patients and 3 additional patients also served as their own controls, receiving rhG-CSF for all cycles after the first. No patient received bone marrow transplantation. rhG-CSF shortened the median duration of severe granulocytopenia (less than or equal to 100/mm3) in a dose-related fashion (P less than .03; Kruskal-Wallis test). Patients not receiving rhG-CSF had a median of 8.5 days of granulocytopenia. Those receiving 40 micrograms/kg of rhG-CSF for approximately 20 days from the third day after chemotherapy had a median of 7.0 days (P less than .23) and those receiving 60 micrograms/kg had a median of 5.5 days (P less than .007) of granulocytopenia. An rhG-CSF dose of 20 micrograms/kg had no effect. Recovery to a granulocyte count of at least 500/mm3 took a median of 12 days in the control group and 8 days (P less than .03) in patients receiving rhG-CSF at a dose of 60 mg/kg. The duration of antibiotic therapy (a median, 9.0 days v 5.0 days) was shortened with the two higher and effective doses of rhG-CSF compared with control patients. The duration of hospitalization (median of 20 days v 19 days) was not shortened. These findings that rhG-CSF decreases the risk of granulocytopenia associated with this particular dose-intensive chemotherapy regimen therapy administered without bone marrow transplantation.

Phase I study of repeated cycles of high-dose cyclophosphamide, etoposide, and cisplatin administered without bone marrow transplantation.

Forty-two patients with advanced malignancy judged unlikely to respond to standard treatment received high-dose combination chemotherapy with cyclophosphamide, etoposide, and cisplatin in a phase I trial. Twenty-two of these patients who had at least a partial response (PR) to the first cycle of therapy received a second cycle, and eight patients received three or more cycles of therapy. Bone marrow replacement was not used. The maximum-tolerated doses (MTDs) were cyclophosphamide 2.5 g/m2 on days 1 and 2; etoposide 500 mg/m2 on days 1, 2, and 3; and cisplatin 50 mg/m2 on days 1, 2, and 3. Hematologic toxicity was not dose-limiting by study design. Recovery to an absolute granulocyte count above 100/microL occurred at a median of 9 days from onset (range, 3 to 23 days) at the MTD. Recovery was delayed after the third cycle. Only one patient on his third cycle failed to recover peripheral blood counts and died of sepsis an day 43. Hematologic toxicity was not dose-dependent. Nonhematologic toxicities included emesis, fatigue, alopecia, diarrhea, and anorexia and were generally well tolerated. The dose-limiting toxicities were fatal pulmonary or cardiac toxicities in five of nine patients treated at the highest dose level. Patients likely to do well can be selected by tumor type, response to prior therapy, and performance status. Nine of 36 assessable patients had a complete response (CR) and 13 a PR for a response rate of 61%. Five patients (12%) remain alive and free of disease at 15 to 32 months. Repeated cycles of dose-intensive combination therapy can produce long-term disease-free remissions in patients with refractory tumor types. The toxicity of the regimen is acceptable if patients are carefully selected.

Dosing regimen of granulocyte-macrophage colony-stimulating factor to support dose-intensive chemotherapy.

PURPOSE: This trial evaluated the optimum dosing regimen for recombinant human granulocyte-macrophage colony-stimulating factor (rhu GM-CSF) to support a dose-intensive chemotherapy regimen given without progenitor cell replacement. PATIENTS AND METHODS: Fifty-one patients with refractory malignancy received cyclophosphamide 2,500 mg/m2 on days 1 and 2, etoposide 500 mg/m2 on days 1, 2, and 3, and cisplatin 50 mg/m2 on days 1, 2, and 3. Patients were hospitalized from cycle days 1 to 4 for chemotherapy and readmitted for cytopenic temperatures above 38.5 degrees C. Cycles were repeated every 35 days in patients who responded to a total of three cycles. GM-CSF was given at doses of 250 to 1,000 micrograms/m2 by continuous intravenous infusion (CIV) or subcutaneously starting on cycle days 3 to 6. Two nonrandomized control groups are used. RESULTS: The optimum regimen of GM-CSF for shortening the duration of leukopenia (WBC count less than 300/microL) was 500 micrograms/m2 given CIV. Duration of leukopenia was 5.9 days compared with 13.2 and 10.2 days in the controls (P less than .05). The optimum regimens for shortening duration of hospitalization, however, were 500 and 750 micrograms/m2/d given as divided (twice daily) subcutaneous injections. Durations of hospitalization were 9.6 and 9.8 days compared with 15.7 and 22.2 days in the controls (P less than .08). At the higher GM-CSF dose, only 36% of patients required readmission for cytopenic fever. Toxicities of GM-CSF at clinically useful doses were minimal. Twelve patients had complete response (24%) and 22 partial response (43%). CONCLUSIONS: This dose-intensive regimen can be given safely without progenitor replacement. rhu GM-CSF decreases the duration of severe leukopenia and decreases the need for hospitalization and antibiotic therapy.

Z-2 microsatellite allele is linked to increased expression of the aldose reductase gene in diabetic nephropathy.

Epidemiological studies support the hypothesis that genetic factors modulate the risk for diabetic nephropathy (DN). Aldose reductase (ALDR1), the rate-limiting enzyme in the polyol pathway, is a potential candidate gene. The present study explores the hypothesis that polymorphisms of the (A-C)n dinucleotide repeat sequence, located 2.1 kb upstream of the transcription start site, modulate ALDR1 gene expression and the risk for DN. We conducted studies at two different institutions, the University of New Mexico Health Sciences Center (UNMHSC), and the Istituto Scientifico H San Raffaele (HSR). There were four groups of volunteers at UNMHSC: group I, normal subjects; group II, patients with insulin-dependent diabetes mellitus (IDDM) without DN; group III, IDDM with DN; and group IV, nondiabetics with kidney disease. At HSR we studied volunteers in groups I, II, and III. ALDR1 genotype was assessed by PCR and fluorescent sequencing of the (A-C)n repeat locus, and ALDR1 messenger ribonucleic acid (mRNA) was measured by ribonuclease protection assay in peripheral blood mononuclear cells. At UNMHSC we identified 10 alleles ranging from Z-10 to Z+8. The prevalence of the Z-2 allele among IDDM patients was increased in those with DN. Sixty percent of group III and 22% of group II were homozygous for Z-2. Moreover, 90% and 67% of groups III and II, respectively, had 1 or more copy of Z-2. In contrast, among nondiabetics, 19% of group IV and 3% of group I were homozygous for Z-2, and 69% and 32%, respectively, had 1 copy or more of Z-2. Among diabetics, homozygosity for the Z-2 allele was associated with renal disease [odds ratio (OR), 5.25; 95% confidence interval, 1.71-17.98; P = 0.005]. ALDR1 mRNA levels were higher in patients with DN (group III; 0.113 +/- 0.050) than in group I (0.068 +/- 0.025), group II (0.042 +/- 0.020), or group IV (0.015 +/- 0.011; P < 0.01). Among diabetics, ALDR1 mRNA levels were higher in Z-2 homozygotes (0.098 +/- 0.06) and Z-2 heterozygotes (0.080 +/- 0.04) than in patients with no Z-2 allele (0.043 +/- 0.02; P < 0.05). In contrast, among nondiabetics, ALDR1 mRNA levels in Z-2 homozygotes (0.034 +/- 0.04) and Z-2 heterozygotes (0.038 +/- 0.03) were similar to levels in patients without a Z-2 allele (0.047 +/- 0.03; P = NS). At HSR we identified eight alleles ranging from Z- 12 to Z+2. The prevalence of the Z-2 allele was higher in group III than in group II. In group III, 43% of the patients were homozygous for Z-2, and 81% had one copy or more of the Z-2 allele. In contrast, in group II, 4% were homozygous for Z-2, and 36% had one copy or more of the Z-2 allele. IDDM patients homozygous for Z-2 had an increased risk for DN compared with those lacking the Z-2 allele (OR, 18; 95% confidence interval, 2-159). IDDM patients who had one copy or more of Z-2 had increased risk (OR, 7.5; 95% confidence interval, 1.9-29.4) for DN compared with those without the Z-2 allele. These results support our hypothesis that environmental-genetic interactions modulate the risk for DN. Specifically, the Z 2 allele, in the presence of diabetes and/or hyperglycemia, is associated with increased ALDR1 expression. This interaction may explain the observed association between the Z-2 allele and DN.

Mammography screening and breast cancer tumor size in female members of a managed care organization.

A study of temporal trends in mammography screening and changes in stage of disease at diagnosis was conducted among Hispanic and non-Hispanic white female members of the Lovelace Health Plan, Flexcare Plan, and Lovelace Senior Plan/Senior Options (LHP), a managed care organization. Two-year screening rates for female members ages 50-74 years were calculated for 1989-1996. From 1989-1996, mammography screening rates for non-Hispanic white female members increased from 65.5 to 71.6%, although this was not a statistically significant increase. Screening rates for Hispanic female members also increased from 50.6 to 62.7%, but they were significantly lower than for non-Hispanic white women. All breast cancers occurring among LHP female members ages 40-74 years were also identified for this same time period. A logistic regression model adjusting for age, year of diagnosis, ethnicity, and duration of enrollment prior to diagnosis found that statistically significant predictors of more advanced stage of disease at diagnosis included young age, diagnosis after 1991 for non-Hispanic white women, and diagnosis prior to 1992 for Hispanic women. Longer duration of enrollment prior to diagnosis was predictive of lower stage of disease, but the odds ratio was not statistically significant. For the time period 1992-1996, Hispanic women with breast cancer were more than twice as likely to have advanced stage of breast cancer compared with non-Hispanic white women (odds ratio, 2.12).

Cigarette smoking and other risk factors for cervical dysplasia in southwestern Hispanic and non-Hispanic white women.

To assess smoking-related and other risk factors for high-grade cervical dysplasia in southwestern Hispanic and non-Hispanic white women in New Mexico, we conducted a clinic-based case-control study among attendees at university-affiliated gynecology clinics. We collected data on cigarette use, sexual behavior, past and current sexually transmitted diseases, hygienic practices, contraception, and diet. For both ethnic groups combined, after adjustment for the effects of human papillomavirus, sexual behavior, and other risk factors, cigarette smoking at the time of diagnosis was associated with high-grade dysplasia (odds ratio, 1.7; 95% confidence limits, 1.0-2.8). In contrast, former smoking was not associated with cervical dysplasia (odds ratio 0.9; 95% confidence limits, 0.5-1.5). Analyses showed dose-response relationships for the amount of cigarettes smoked per day and for cumulative exposure (pack-years of use) in association with cervical dysplasia. Although our study lacked the power to show statistically significant ethnic differences in smoking-related risks for dysplasia, smoking at the time of diagnosis, high pack-years of use, and smoking at the time of menarche were associated with dysplasia only for non-Hispanic white versus Hispanic women. Our data support hypotheses that implicate cigarette use as an etiological factor in the development of high-grade cervical dysplasia and suggest ethnic differences in risks for dysplasia among women attending the same clinics.

Transperineal percutaneous iridium-192 interstitial template implant of the prostate: results and complications in 321 patients.

A total of 321 patients with localized adenocarcinoma of the prostate treated by modified pelvic lymphadenectomy, Iridium-192 implant, and external beam iridium radiation were retrospectively reviewed. Analysis covered 8 years between 1981 and 1989 with a median population age of 72 (range 42 through 82 years). Disease-free survival for the entire group is 69% at 5 years with a median follow-up of 34 months (range 1.5 months to 98.5 months). As expected, both bulkier disease and positive nodal status adversely affected 5-year disease-free survival (p = 0.0001 for both). For tumors stage T1b (A2), T2a (B1), T2b (B2), T3 (C) the disease-free survival is 89.5%, 89.9%, 64.7%, and 48.8%, respectively; for NO disease 5-year disease-free survival is 76.5% versus N1/N2 disease with 5-year disease-free survival of 33.2%. Local control was excellent except for bulkier disease (p = 0.009). Tumors T1b, T2a, T2b, and T3 have 60-month local control rates of 95%, 93%, 83.6%, and 73.1%, respectively. Histologic grade also affected disease-free survival and local control with grade 1, grade 2, grade 3 showing 81.2%, 65.7%, and 45.1% disease-free survival at 5 years; and 93.6%, 82.2%, and 72.4% local control at 5 years. Estimates obtained using Kaplan-Meier method. Radiation induced morbidity was analyzed separately for all patients, there were 41 patients (13% of total) with 54 documented complications. There were no Grade 4 or 5 complications as per RTOG categories. Only 3 cases showed grade 3 complications (1%) and 51 cases showed grade 2 complications (15.9%). Grade 1 complications were not recorded. Of the grade 2 and grade 3 complications 30 were GU and 22 were rectal. The morbidity associated with combined interstitial implantation by transperineal percutaneous template and external beam iridium radiation for the localized prostate cancer is minimal with excellent local control and disease-free survival.

Contraceptive and reproductive risks for cervical dysplasia in southwestern Hispanic and non-Hispanic white women.

BACKGROUND: Various contraceptive practices and reproductive factors have been associated with cervical neoplasia in case-control studies worldwide. METHODS: To investigate contraceptive and reproductive risk factors associated with high-grade cervical dysplasia in southwestern Hispanic and non-Hispanic white women, we carried out a clinic-based case-control study among university-affiliated clinic attendees. RESULTS: Oral contraceptive use ever (odds ratio [OR] = 0.4, 95% confidence interval [CI]: 0.2-0.9) and past diaphragm use (OR = 0.3, 95% CI: 0.1-0.8) were protective for dysplasia in analyses adjusted for age, ethnicity, sexual behaviour, and for cervical papillomavirus (HPV) infection. After further adjustment for Pap smear screening interval, oral contraceptive use ever remained protective for dysplasia. Vaginal deliveries were strongly associated with dysplasia with > 2 vaginal deliveries associated with a 3.9-fold increase in risk after adjustment for age, ethnicity, sexual behaviour, and HPV infection. Using logistic regression models to simultaneously control for effects of multiple factors as potentially related to cervical dysplasia, we found low educational attainment, cervical HPV infection, cigarette smoking, history of any sexually transmitted disease, and having one or more vaginal deliveries to be associated with dysplasia; oral contraceptive use and past diaphragm use also remained protective for high-grade cervical dysplasia in these regression analyses. CONCLUSIONS: The data suggest that use of oral contraceptives (ever) and past diaphragm use are protective for high-grade cervical dysplasia among Hispanic and non-Hispanic white women in New Mexico. The clinic-based perspective of this research (versus population-based studies) may help explain some of these findings.

Metabolic and cognitive response to human traumatic brain injury: a quantitative proton magnetic resonance study.

Proton magnetic resonance spectroscopy (1H-MRS) offers a unique insight into brain cellular metabolism following traumatic brain injury (TBI). The aim of the present study was to assess change in neurometabolite markers of brain injury during the recovery period following TBI. We studied 19 TBI patients at 1.5, 3, and 6 months postinjury and 28 controls. We used 1H-MRS to quantify N-acetylaspartate (NAA), creatine (Cre), choline (Cho), and myoinositol (mIns) in occipitoparietal gray matter (GM) and white matter (WM) remote from the primary injury focus. Neuropsychological testing quantified cognitive impairment and recovery. At 1.5 months, we found cognitive impairment (mean z score = -1.36 vs. 0.18,p < 0.01), lower NAA (GM: 12.42 mM vs. 13.03, p = 0.01; WM: 11.75 vs. 12.81, p < 0.01), and elevated Cho (GM: 1.51 vs. 1.25, p < 0.01; WM: 1.98 vs. 1.79, p < 0.01) in TBI patients compared with controls. GM NAA at 1.5 months predicted cognitive function at outcome (6 months postinjury; r = 0.63, p = 0.04). GM NAA continued to fall by 0.46 mM between 1.5 and 3 months (p = 0.02) indicating continuing neuronal loss, metabolic dysfunction, or both. Between 3 and 6 months, WM NAA increased by 0.55 mM (p = 0.06) suggesting metabolic recovery. Patients with poorer outcomes had elevated mean GM Cho at 3 months postinjury, suggesting active inflammation, as compared to patients with better outcomes (p = 0.002). 1H-MRS offers a noninvasive approach to assessing neuronal injury and inflammation following TBI, and may provide unique data for patient management and assessment of therapeutic efficacy.

Medical students as patients: a pilot study of their health care needs, practices, and concerns.

BACKGROUND: The personal health experiences of medical students may contribute in important but previously unacknowledged ways to their well-being and education. This pilot study surveyed medical students about their health care needs, practices, insurance status, and concerns about seeking care. METHOD: A questionnaire was developed and distributed to 151 students at the University of New Mexico School of Medicine in 1993-94. Participant privacy was protected. Responses were compiled and analyzed using logistic regression models and odds ratios. RESULTS: A total of 112 students responded. Most reported health care needs and half routinely received care at their training institution. One-third had informally requested prescriptions or diagnostic tests from medical school faculty and housestaff; one-fourth used such informal consultation as their "usual" method of obtaining care. Eighteen students were uninsured. The students reported that they had not sought care for several reasons, and many had experienced difficulty in obtaining care. The students indicated concern about confidentiality and about the dual role as both student and patient at the training institution. They believed that their academic standing would be jeopardized if they developed certain health problems. When asked about hypothetical scenarios, a majority preferred to avoid the dual role of medical-student-patient. When asked about scenarios in which medical student peers exhibited suicidal depression or severe drug abuse, the students overwhelmingly preferred not to notify the medical school administration. Significant differences in responses were found with respect to gender and training level. CONCLUSION: This pilot study examined the health care needs, practices (including the use of informal consultation), insurance status, and concerns of students at one medical school. The findings highlight the students' perceptions of illness and vulnerability during medical school training. Constructive implications for academic medicine are discussed regarding initiatives in the areas of policy, research, and the resources and structure of student health care services.

The XRCC1 399 glutamine allele is a risk factor for adenocarcinoma of the lung.

Defects in the repair and maintenance of DNA increase risk for cancer. X-ray cross-complementing group 1 protein (XRCC1) is involved with the repair of DNA single-strand breaks. A nucleotide substitution of guanine to adenine leading to a non-conservative amino acid change was identified in the XRCC1 gene at codon 399 (Arg/Gln). This change is associated with higher levels of aflatoxin B1-adducts and glycophorin A somatic mutations. A case-control study was conducted to test the hypothesis that the 399Gln allele is positively associated with risk for adenocarcinoma of the lung. XRCC1 genotypes were assessed at codon 399 in 172 cases of lung adenocarcinoma and 143 cancer-free controls. Two ethnic populations were represented, non-Hispanic White and Hispanic. The distribution of XRCC1 genotypes differed between cases and controls. Among cases, 47.7% were Arg/Arg, 35.5% were Arg/Gln, and 16.9% were Gln/Gln. Among controls, XRCC1 allele frequencies were 45.5% for Arg/Arg, 44.8% for Arg/Gln, and 9.8% for Gln/Gln. Logistic regression analysis was used to assess the association between lung adenocarcinoma and the G/G genotype relative to the A/A or A/G genotypes. In non-Hispanic White participants, the lung cancer risk associated with the G/G genotype increased significantly after adjustment for age (OR=2.81; 95% CI, 1.2-7.9; P=0.03) and increased further after adjustment for smoking (OR=3.25; 95% CI, 1.2-10.7; P=0.03). Among all groups, a significant association was found between the G/G homozygote and lung cancer (OR=2.45; 95% CI, 1.1-5.8; P=0.03) after adjustment for age, ethnicity, and smoking. This study links a functional polymorphism in the critical repair gene XRCC1 to risk for adenocarcinoma of the lung.

Behavioral correlates of differences between a child's assessment and the parents' assessment of the child's self-esteem.

When, as commonly occurs, a parent's and child's reports about the child differ, the lack of agreement usually leads to consideration of which information is objective and valid. Little attention has been given to understanding what meaning these differences might have in the context of parent-child relationships or to whether the existence or absence of these differences might be causally related to the child's psychosocial status and functioning. Third- through fifth-grade children (N = 178) in behavior disorder and regular classrooms were asked to complete a self-concept measure; parents were asked to independently complete the same instrument as they thought their child would. Parents of children without behavior disorders were significantly more accurate in their descriptions of their children's perceptions than were parents of children with behavior problems. They also tended to expect their children to have more positive self-concepts than the children actually reported. Parents who are able to accurately report their children's feelings and who err toward more positive assessments may be preventing behavior problems.

Sources of variability in posterior rhinomanometry.

Sources of variability in nasal airway resistance measured by posterior rhinomanometry were studied in 5 subjects tested on 5 different days and 56 subjects tested on 2 different days. On each day, a questionnaire on upper airway health and nasal symptoms was completed. The mean individual difference in nasal airway resistance between the 2 test days in the group of 56 subjects was 5.3% (SD 52.7%). Between-subject variability accounted for 74.9% and 72.5% of the total variability in the group of 5 and the group of 56 subjects, respectively. For the 5 subjects, by accounting for a change in upper airway symptoms or upper respiratory tract infection that occurred over the 5 test days, there was a significant decrease in the between-subject variability. The difference in sources of variation due to a change in upper airway symptoms was not seen in the group of 56 subjects. We conclude that the largest source of variability in nasal airway resistance is due to between-subject differences.

Predictive value of flow cytometric analysis in DNA contents in patients with locally advanced head and neck carcinoma.

A retrospective study was performed on 61 eligible patients with stage III and IV (AJC/UICC Staging System) squamous carcinomas of the head and neck region who were treated with definitive radiotherapy with, or without, surgery. DNA contents were measured by flow cytometric analysis of archival paraffin blocks and were correlated with clinicopathological findings, tumour response and patient survival. Comparison of variables including treatment modality was performed for identification of significant prognostic factors. There were 28 diploid, 27 aneuploid tumours and the remaining six were questionable. All patients were followed-up for at least two years or until death. Aneuploid tumours had a significantly higher S-phase fraction (percentage S-phase) (p < 0.001). Neither ploidy nor percentage S-phase were found to have predictive value in tumour response or patient survival within the power of a sample size of 61. Twenty of the 27 (74 per cent) aneuploid tumours had a complete response (CR) whereas 19 out of 28 (68 per cent) diploid tumours achieved CR. Five-year survival by the Kaplan-Meier method was 33 per cent for both aneuploid and diploid tumours. However, nodal stage (N stage) was found to have significant predictive value in both tumour response and patient survival. The complete response for stage N0 patients was 96 per cent, N1 patients 61 per cent, N2 patients 60 per cent and 43 per cent for N3 patients (p < 0.002). Similarly, the five year survival for the N0 and N3 groups of patients was 53 per cent and 29 per cent respectively (p < 0.05).

A review of ecologic studies of lung cancer and indoor radon.

Although radon exposure is an established cause of lung cancer among underground miners, the lung cancer risk to the general population from indoor radon remains controversial. This controversy stems in part from the contradictory results of published studies of indoor radon and lung cancer, including 15 ecologic studies, seven of which found a positive association, six no association, and two a negative association. To address the misunderstanding of the indoor radon risk that has resulted from these ecologic studies, the authors discuss the general methodologic problems and limitations of ecologic studies, and the particular limitations of these 15 studies. The authors conclude that the shortcomings of the ecologic studies render them uninformative on the lung cancer risk associated with indoor radon.

Oronasal distribution of ventilation at different ages.

The route of breathing, oral or nasal, is a determinant of the doses of inhaled pollutants delivered to target sites in the upper and lower respiratory tracts. We measured partitioning of ventilation, using a divided oronasal mask during a submaximal exercise test, in 37 male and female subjects who ranged in age from 7 to 72 y. The following four patterns of breathing were evident during exercise: (1) nasal only (13.5%), nasal shifting to oronasal (40.5%), oronasal only (40.5%), and oral only (5.4%). Children (i.e., 7-16 y of age) displayed more variability than adults with respect to their patterns of ventilation with exercise. Young adults (i.e., 17-30 y of age) who initially breathed nasally with exercise switched to oral ventilation at a lower percentage of the previously measured maximum ventilation (10.8%) than older subjects (31.8%). The partitioning of ventilation between the nasal and oral routes follows complex patterns that cannot be predicted readily by the age, gender, or nasal airway resistance of the subject.

Predicting gene promoter methylation in non-small-cell lung cancer by evaluating sputum and serum.

The use of 5-methylcytosine demethylating agents in conjunction with inhibitors of histone deacetylation may offer a new therapeutic strategy for lung cancer. Monitoring the efficacy of gene demethylating treatment directly within the tumour may be difficult due to tumour location. This study determined the positive and negative predictive values of sputum and serum for detecting gene methylation in primary lung cancer. A panel of eight genes was evaluated by comparing methylation detected in the primary tumour biopsy to serum and sputum obtained from 72 patients with Stage III lung cancer. The prevalence for methylation of the eight genes in sputum (21-43%) approximated to that seen in tumours, but was 0.7-4.3-fold greater than detected in serum. Sputum was superior to serum in classifying the methylation status of genes in the tumour biopsy. The positive predictive value of the top four genes (p16, DAPK, PAX5 beta, and GATA5) was 44-72% with a negative predictive value for these genes > or =70%. The highest specificity was seen for the p16 gene, and this was associated with a odds ratio of six for methylation in the tumour when this gene was methylated in sputum. In contrast, for serum, the individual sensitivity for all genes was 6-27%. Evaluating the combined effect of methylation of at least one of the four most significant genes in sputum increased the positive predictive value to 86%. These studies demonstrate that sputum can be used effectively as a surrogate for tumour tissue to predict the methylation status of advanced lung cancer where biopsy is not feasible.