Abstract
Objective
The cancer anorexia/weight loss syndrome commonly occurs in patients with non-small cell lung cancer (NSCLC) and is characterized by loss of weight and appetite as well as diminished survival. The current study explored whether any of 22 single nucleotide polymorphisms (SNPs) of certain previously implicated inflammatory cytokines (interleukin-1 beta, interleukin-1RN, interleukin-6, and tumor necrosis factor) are associated with this syndrome.
Patients and Methods
All NSCLC patients who had been enrolled in the Mayo Clinic Lung Cancer Cohort, had completed a health-related questionnaire approximately 6 months after enrollment, and had blood drawn were included in this study, thus yielding a sample size of 471 patients.
Results
Sixty-six (14%) patients manifested weight loss shortly after diagnosis, and 152 (32%) reported appetite loss. Only tumor necrosis factor alpha rs800629 was associated with anorexia (odds ratio: 0.46; 95% confidence interval: 0.29, 0.72; p<0.001); patients who were heterozygous and minor homozygous were less likely to suffer anorexia. Otherwise, there were no statistically significant associations between any of the other 21 SNPs and weight loss and/or anorexia. In univariate analyses, weight loss, anorexia, more advanced cancer stage, and interleukin-1 beta rs1143627 were associated with a worse survival, and interleukin-6 rs2069835 was associated with better survival. However, in multivariate analyses, cancer stage and patient age were the only statistically significant predictors of worse survival.
Conclusion
No specific SNP was associated with all aspects of the cancer anorexia/weight loss syndrome, but rs800629 may merit further study in cancer-associated anorexia.
Keywords: Cancer-associated anorexia, Single nucleotide polymorphism, Non-small cell lung cancer
The cancer anorexia/weight loss syndrome is characterized by loss of weight, diminished appetite, inferior quality of life, and a shortened survival. The latter is a salient aspect of this syndrome, as indicated by the four studies described below.
First, evaluating 116 patients with relatively early stage lung cancer, Jeremic and others observed that patients with weight loss died early and that this sign was one of the most robust predictors of early demise [1]. Second, in a landmark study, Dewys and others evaluated 3,047 patients, with a variety of malignancies, including non-small cell lung cancer [2]. These investigators found once again that weight loss predicted early death and that this predictive capability occurred independently of performance status and tumor burden. Third, this prognostic effect is not confined to weight loss that occurs immediately prior to a diagnosis of cancer. Chuang and others evaluated 356 late-stage cancer patients, all of whom were candidates for hospice and many of whom had received prior cancer therapy [3]. Lung cancer was the most common cancer type within this group. Evaluating patient-reported weight loss in the preceding 3 months, these investigators observed that even under these challenging circumstances and well after a patient's original cancer diagnosis, weight loss continued to be a powerful predictor of early death. Finally, the symptom of anorexia also predicts early death. In a 1,115-patient North Central Cancer Treatment Group study, in which approximately half the cohort had lung cancer, patients were assessed for anorexia in a prospective fashion. These investigators observed that this one symptom carried a statistically significant negative prognostic effect [4]. Hence, these four studies, as well as others, underscore the fact that weight loss and/or anorexia in cancer patients signals a poor outcome.
What mediates this weight loss and anorexia and thereby leads to early death? Although ongoing preclinical research suggests that a complex series of factors are at work, previous studies point to a cascade of inflammatory cytokines that ultimately lead to loss of appetite, wasting of functional lean tissue, weight loss, and shortened survival [5–10]. These cytokines include interleukin-1 beta (IL-1 beta), interleukin-1RN (IL-1RN), interleukin-6 (IL-6), and tumor necrosis factor (TNF). Therefore, the current study was conducted to focus specifically on functional single nucleotide polymorphisms (SNPs) of these inflammatory cytokines in an effort to explore whether associations exist between these potential mediators and the symptom complex of anorexia, weight loss, and poor survival in patients with non-small cell lung cancer.
Methods
Overview
This study was undertaken within the Mayo Clinic Lung Cancer Cohort, a prospectively conducted study that has attempted to recruit consecutively all lung cancer patients seen at the Mayo Clinic in Rochester, Minnesota since January 1, 1997. The Mayo Clinic Lung Cancer Cohort is reviewed at least annually by the Mayo Clinic Institutional Review Board, and previously published studies provide details of the cohort. All patients had provided written informed consent prior to enrollment.
Ascertainment of clinical data
After enrollment, each patient's medical record was reviewed by trained personnel, and extensive demographic information on the patient's health, cancer status, and subsequent treatment were entered into the database. The Revised Tumor–Node–Metastasis staging system for non-small cell lung cancer was used to assign a cancer stage to each patient [11]. Thereafter, all patients were asked to complete a series of health-related surveys. The first of these were mailed to patients within 6 months of enrollment. Subsequently, surveys were mailed to patients on an annual basis in an effort to acquire follow-up information and, as particularly relevant to the current report, to ascertain vital status.
The survey, which had been sent within 6 months of enrollment, asked patients to record their current weight. This weight was then compared to the weight in the medical record recorded in the peri-diagnostic period; the percent change in weight over this time interval was then calculated. Greater than 10% weight loss during this period indicated weight loss of clinical concern. This threshold was chosen because of its negative prognostic significance and because it is indicative of nutritional compromise that often merits nutritional support [12].
To gain a broader spectrum of symptomatology relevant to weight loss, patients were asked at approximately the 6-month time point about their appetite. A previously validated questionnaire item [13], “How would you compare your appetite now to what it was before your present illness?” provided patients with five response options, “increased,” “same,” “slightly reduced (about 75% of normal),” “moderately reduced (about 50% of normal),” and markedly reduced (about 25% of normal or less).”
Vital status had been ascertained by means of follow-up surveys. Moreover, within the Mayo Clinic Lung Cancer Cohort, vital status is continuously verified every year by means of the following sources: the Mayo Clinic registration clinical database, other next-of-kin reports, death certificates, documentation of death in the clinical record, and/or documentation of vital status within the Mayo Clinic Tumor Registry.
Single nucleotide polymorphism analyses
Blood was drawn at baseline and utilized for genotyping the SNPs for the following inflammatory cytokines: IL-1 beta, IL-1RN, IL-6, and TNF. These cytokines were chosen because of prior research which has implicated them in cancer-associated weight loss and anorexia.
A deliberate decision was made not to assay and assess serum cytokine concentrations as such levels can be highly variable, often do not correlate with clinical endpoints, and are not reflective of cytokine concentrations at the tissue level where wasting occurs [14]. Thus, the current study focused exclusively on the SNPs of these inflammatory cytokines.
Genotyping was performed in the Mayo Clinic Genomic Shared Resources. Two genotype approaches were used, SNPstream and TaqMan. Each was run according to the manufactures' protocol, as described previously [15, 16]. For SNPstream panels and Taqman assays primer sets were designed using software specific to each platform. Control CEPH DNAs (n=4–12) and no template controls (n=8) were included in each 384 well plate to assure accuracy and reproducibility across the project. In addition, sample duplicates (2%) were included in the plate design. Genotyping was successful for all samples with an average SNP call rate of 99.8%.
Statistical Analyses
Demographic data and frequencies of each SNP are presented descriptively. Patient-reported anorexia was categorized as a dichotomous variable. Patients were considered to have anorexia if they responded that their appetite was “slightly,” “moderately,” or “markedly” reduced. Similarly, patients were categorized as not having anorexia if they had reported that their appetite was “increased” or the “same.” As mentioned earlier, weight loss was defined as a 10% drop in weight at 6 months after diagnosis.
The polymorphisms of interest were assessed and, in this case, confirmed for Hardy–Weinberg equilibrium. Then cytokine polymorphisms were categorized based on whether they were major homozygous, heterozygous, and minor homozygous SNP alleles. For the analyses, the minor homozygous SNP alleles with frequencies of <5% were grouped together with the heterozygous SNP alleles. SNPs with minor allele frequencies of <5% were excluded from the analyses. This approach was necessary to ensure the stability of the models described below.
Logistic regression models were used to assess associations between the various genetic polymorphisms and weight loss and anorexia. Odds ratios and 95% confidence intervals are reported as appropriate. The Bonferroni method for p value adjustment was used for multiple comparisons. In view of the fact that 22 SNPs were examined, a p value of <0.002 was deemed statistically significant in the logistic regression models that assessed associations between these clinical variables of weight loss and anorexia and these polymorphisms. Gene–gene interactions and other two-way interactions of factors were not investigated in view of the relatively limited patient sample size and the exploratory nature of these analyses.
This study utilized a landmark survival analysis that included the postdiagnosis, baseline survey completion data as the baseline time-point. Overall survival was defined as the time from this baseline time-point to death from any cause. The distribution of overall survival was estimated by the Kaplan–Meier method. Univariate Cox proportional hazard models were then used to explore associations between each polymorphism within the context of the clinical factors of weight loss, anorexia, cancer stage, age, gender, and overall survival. Factors with a p value of <0.1 in the univariate analysis were then examined in a multivariate model. Cancer treatment was not utilized in this model because of collinearity with stage. Hazard ratios and 95% confidence intervals are reported as appropriate. A p value of <0.05 is considered statistically significant. In these two, separate multivariate models for weight loss and anorexia, a p value of <0.025 was considered statistically significant.
A sample size of 471 patients provided greater than 90% power to detect an association assessed by an odds ratio of 0.48 for a two-level SNP polymorphism with a prevalence, for example, of 34% versus 66% (two-sided chi-square test, alpha = 0.05). A sample size of 471 patients also provided approximately 90% power to detect an effect assessed by a hazard ratio of 1.5 for this two-level factor with a prevalence of 34% versus 66% (two-sided logrank test, alpha level = 0.05) using the actual enrollment rates for the present data and assuming an exponential distribution for survival with a minimum of 2-years follow up for each patient.
Results
Demographics
A total of 471 patients with non-small cell lung cancer completed the 6-month health-related questionnaire on weight loss and anorexia and had blood available for genotyping. These patients are the focus of this report. Patient demographics are shown in Table 1.
Table 1.
Baseline characteristics; N=471a
| Age at diagnosis, median (range) | 67 (18, 89) |
|---|---|
| Gender | |
| Female | 241 (51) |
| Male | 230 (49) |
| Tumor histology | |
| Adenocarcinoma | 244 (52) |
| Squamous | 119 (25) |
| Bronchioalveolar | 23 (5) |
| Large cell | 12 (2) |
| Adenosquamous | 9(2) |
| Other, including non-small cell, unspecified | 64 (13) |
| Cancer stageb | |
| I | 224 (48) |
| II | 45 (10) |
| III | 110 (24) |
| IV | 89 (19) |
| Cancer therapy reported on baseline Questionnaireb | |
| Surgery | 337 (73) |
| Radiation | 143 (31) |
| Chemotherapy | 223 (48) |
Numbers in parentheses refer to the percentage of the cohort unless otherwise specified
Not all percentages sum to 100% either because of rounding, as seen with cancer stage, or patients' falling into more than one category, as seen with cancer therapy
Weight loss, anorexia, and associations with SNPs
Sixty-six patients (14%) had lost weight between the time of their diagnosis and their 6-month postdiagnosis assessment, and 152 (32%) described appetite loss (Table 2). Of incidental note, a total of 12 patients (2.5%) manifested weight gain of 10% or greater. SNP polymorphism frequencies are shown (Table 3).
Table 2.
Baseline weight loss and anorexia data; N=471a
| Weight loss of >10% since diagnosis | |
| Yes | 66 (14) |
| How would you compare your appetite now to what it was before your present illness? | |
| Increased | 73 (15) |
| Same | 246 (52) |
| Slightly reduced (about 75% of normal) | 100 (21) |
| Moderately reduced (about 50% of normal) | 26(6) |
| Markedly reduced (about 25% of normal or less) | 26 (6) |
Numbers in parentheses denote the percentage of the entire cohort unless otherwise specified
Table 3.
SNP frequencies
| Gene | SNP | Homozygous common allele | Heterozygotes N(%) | Homozygous rare allele N(%) | Harding-Weinberg equilibrium (pvalue) N(%) |
|---|---|---|---|---|---|
| IL1B | rs1143627 | 216(45.9) | 205(43.5) | 50(10.6) | 0.90 |
| rs1143630a | 417(88.5) | 51(10.8) | 3(0.6) | 0.30 | |
| rs1143633 | 191(40.6) | 219(46.5) | 61(13) | 0.89 | |
| rs1143634 | 269(57.2) | 168(35.7) | 33(7) | 0.34 | |
| rs2853550a | 398(84.5) | 68(14.4) | 5(1.1) | 0.28 | |
| IL1RN | rs3087263a | 381(81.1) | 87(18.5) | 2(0.4) | 0.20 |
| rs3087266a | 345(73.3) | 118(25.1) | 8(1.7) | 0.56 | |
| rs315952 | 236(50.2) | 197(41.9) | 37(7.9) | 0.64 | |
| rs380092 | 214(45.4) | 208(44.2) | 49(10.4) | 0.88 | |
| rs397211 | 221(46.9) | 203(43.1) | 47(10) | 0.97 | |
| rs4252041 | 434(92.3) | 36(7.7) | - | 1.00 | |
| IL6 | rs1800795 | 144(30.6) | 243(51.6) | 84(17.8) | 0.29 |
| rs2066992 | 425(90.2) | 46(9.8) | - | 0.62 | |
| rs2069835a | 421(89.4) | 49(10.4) | 1(0.2) | 0.73 | |
| rs2069840 | 205(43.6) | 218(46.4) | 47(10) | 0.32 | |
| rs2069843a | 443(94.3) | 25(5.3) | 2(0.4) | 0.07 | |
| rs2069852 | 441(93.8) | 29(6.2) | - | 1.00 | |
| rs2069857b | 458(99.1) | 4(0.9) | - | 1.00 | |
| rs2069860b | 465(98.7) | 6(1.3) | - | 1.00 | |
| rs2069861a | 375(79.8) | 91(19.4) | 4(0.9) | 0.55 | |
| TNF | rs1800629a | 302(66.2) | 133(29.2) | 21(4.6) | 0.20 |
| rs1800630a | 334(70.9) | 124(26.3) | 13(2.8) | 0.72 | |
| rs3093661 | 433(92.1) | 37(7.9) | - | 1.00 | |
| rs3093662 | 410(87.1) | 61(13) | - | 0.13 | |
| rs3093665b | 455(96.6) | 16(3.4) | - | 1.00 | |
| rs3093671b | 448(95.3) | 22(4.7) | - | 1.00 | |
| rs4645843b | 470(100) | - | - | 1.00 |
Due to low frequencies (<5%), the homozygous rare alleles of these SNPs were grouped together with heterozygotes for the stability of statistical models
Due to low frequencies (<5%) of the heterozygotes and absent of homozygous rare alleles, these SNPs were not included the analyses
With one exception, there were no statistically significant associations between any of these 22 SNPs and weight loss and/or anorexia. Only TNF polymorphism rs800629 was associated with anorexia (odds ratio: 0.46; 95% confidence interval: 0.29, 0.72; p<0.001). Patients who were heterozygous and minor homozygous were less likely to suffer anorexia compared to those who were major homozygous.
Survival
At the time of these analyses, 237 patients (50%) had died. The median follow-up time for the living patients is 5.1 years (range: 1.0, 8.2 years) with 95% of patients followed beyond 2 years.
In univariate analyses, weight loss, anorexia, and advanced cancer stage were associated with a worse survival (Table 4). The IL-1 beta rs1143627 SNP was also associated with worse survival. In contrast, IL-6 rs2069835 was associated with an improved survival. However, in multivariate analyses, cancer stage and age were the only statistically significant factors that predicted a worse survival.
Table 4.
Relationships between relevant variables and survival
| Univariate models | Multivariate models for weight loss | Multivariate models for anorexia | |||||
|---|---|---|---|---|---|---|---|
|
|
|
|
|
||||
| Factor | Categories | Hazard ratio (95% confidence interval) | pvalue | Hazard ratio (95% confidence interval) | pvalue | Hazard ratio (95% confidence interval) | pvalue |
| >10% weight loss | Yes versus no | 1.48 (1.05, 2.08) | 0.02 | 1.36 (0.97, 1.92) | 0.08 | - | - |
| Anorexia | Yes versus no | 1.56 (1.20, 2.02) | <0.001 | - | - | 1.37 (1.05, 1.78) | 0.02 |
| Cancer stage | II versus I | 1.78 (1.09, 2.90) | 0.02 | 1.89 (1.16, 3.09) | 0.01 | 1.82 (1.11, 2.98) | 0.02 |
| III versus I | 2.54 (1.81, 3.57) | <0.001 | 2.84 (2.00, 4.03) | 0.001 | 2.80 (1.97, 3.97) | <0.001 | |
| IV versus I | 6.77 (4.87, 9.43) | <0.001 | 7.37 (5.23, 10.4) | <0.001 | 7.14 (5.07, 10.1) | <0.001 | |
| Age | 10 year increments | 1.14 (1.00, 1.30) | 0.046 | 1.28 (1.13, 1.46) | <0.001 | 1.28 (1.13, 1.45) | <0.001 |
| rs1143627 (IL1B) | Heterozygous vs major homozygous | 1.50 (1.14, 1.96) | <0.001 | 1.14 (0.86, 1.51) | 0.36 | 1.14 (0.86, 1.50) | 0.36 |
| Minor homozygous vs major homozygous | 1.25 (0.80, 1.96) | 0.33 | 1.08 (0.68, 1.69) | 0.75 | 1.07 (0.68, 1.69) | 0.76 | |
| rs2069840 (IL6) | Heterozygous vs major homozygous | 1.19 (0.92, 1.56) | 0.19 | 1.15 (0.88, 1.50) | 0.32 | 1.14 (0.87, 1.49) | 0.35 |
| Minor homozygous vs major homozygous | 0.56 (0.33, 0.97) | 0.04 | 0.59 (0.34, 1.01) | 0.05 | 0.56 (0.33, 0.97) | 0.04 | |
Discussion
The current study explored whether SNPs from the inflammatory cytokines, IL-1 beta, IL-1RN, IL-6, and TNF are associated with weight loss, anorexia, and diminished survival. One statistically significant association emerged, that between TNF rs800629 and the presence of anorexia. This observation merits further investigation. However, the current study did not find that a single SNP or group of SNPs is associated with all aspects of this cancer anorexia/weight loss syndrome; and, therefore, this study did not generate compelling clinical evidence to suggest that these SNPs act alone as key mediators of the cancer anorexia/weight loss syndrome.
Do such negative findings demonstrate that SNPs of IL-1 beta, IL-1RN, IL-6, and TNF do not in any way mediate this syndrome? We believe the answer to this question is “no,” and cite the following two explanations. First, a growing consensus suggests that SNPs in general may exert their effects by means of interactions with multiple other environmental and genetic factors [17]. This opinion has been invoked to explain why other previous studies that have focused on other SNPs have also often yielded negative results. In the cancer anorexia/weight loss syndrome, the SNPs studied in this investigation may in fact give rise to this syndrome, but other factors, which remain undiscovered, must perhaps act in concert for them to do so. Thus, we cannot conclude with certainty that these SNPs are not mediators of the cancer anorexia/weight loss syndrome; at best, we can only conclude that, when acting alone, these SNPs do not appear to be robust key mediators, and that further study of their role is warranted.
Second, from a logistic standpoint, this study examined weight loss that occurred from the time of cancer diagnosis to approximately 6 months afterwards. Many of these patients might have been suffering weight loss as a result of cancer treatment or other comorbid conditions, not necessarily from the cancer itself. Although previous studies have shown that the inflammatory cytokines are implicated in cancer-associated weight loss, it remains unknown if these inflammatory cytokines mediate weight loss that occurs in cancer patients who may have other non-cancer-related causes of weight loss. Again, further study of weight loss and its pathophysiology in cancer patients is warranted.
Acknowledgments
This project was funded by R01 CA115857 and R01 84354 to Dr. Yang and by K24CA131099 to Dr. Jatoi.
References
- 1.Jeremic B, Milicic B, Dagovic A, Acimovic L, Milisavljevic S. Pretreatment prognostic factors in patients with early stage (I/II) non-small-cell lung cancer treated with hyperfractionated radiation therapy alone. Int J Radiat Oncol Biol Phys. 2006;65(4):1112–1119. doi: 10.1016/j.ijrobp.2006.01.044. [DOI] [PubMed] [Google Scholar]
- 2.Dewys WD, Begg C, Lavin PT, Band PR, Bennett JM, Bertino JR, Cohen MH, Douglass HO, Jr, Engstrom PF, Ezdinli EZ, Horton J, Johnson GJ, Moertel CG, Oken MM, Perlia C, Rosenbaum C, Silverstein MN, Skeel RT, Sponzo RW, Tormey DC. Prognostic effect of weight loss prior to chemotherapy in cancer patients. Eastern Cooperative Oncology Group. Am J Med. 1980;69(4):491–497. doi: 10.1016/s0149-2918(05)80001-3. [DOI] [PubMed] [Google Scholar]
- 3.Chuang RB, Hu WY, Chiu TY, Chen CY. Prediction of survival in terminal cancer patients in Taiwan: constructing a prognostic scale. J Pain Symptom Manage. 2004;28(2):115–122. doi: 10.1016/j.jpainsymman.2003.11.008. [DOI] [PubMed] [Google Scholar]
- 4.Loprinzi CL, Laurie JA, Wieand HS, et al. Prospective evaluation of prognostic variables from patient-completed questionnaires. J Clin Oncol. 1994;12:601–7. doi: 10.1200/JCO.1994.12.3.601. [DOI] [PubMed] [Google Scholar]
- 5.Seruga B, Zhang H, Bernstein LJ, Tannock IF. Cytokines and their relationship to the symptoms and outcome of cancer. Nat Rev Cancer. 2008;8(11):887–899. doi: 10.1038/nrc2507. [DOI] [PubMed] [Google Scholar]
- 6.Grim-Stieger M, Keilani M, Mader RM, Marosi C, Schmidinger M, Zielinski CC, Fialka-Moser V, Crevenna R. Serum levels of tumour necrosis factor-alpha and interleukin-6 and their correlation with body mass index, weight loss, appetite and survival rate—preliminary data of Viennese outpatients with metastatic cancer during palliative chemotherapy. Eur J Cancer Care. 2008;17(5):454–462. doi: 10.1111/j.1365-2354.2007.00874.x. [DOI] [PubMed] [Google Scholar]
- 7.Bennani-Baiti N, Davis MP. Cytokines and cancer anorexia cachexia syndrome. Am J Hosp Palliat Care. 2008;25(5):407–411. doi: 10.1177/1049909108315518. [DOI] [PubMed] [Google Scholar]
- 8.Zhang D, Zhou Y, Wu L, Wang S, Zheng H, Yu B, Li J. Association of IL-6 gene polymorphisms with cachexia susceptibility and survival time of patients with pancreatic cancer. Ann Clin Lab Sci. 2008;38(2):113–119. [PubMed] [Google Scholar]
- 9.Tan BH, Deans DA, Skipworth RJ, Ross JA, Fearon KC. Biomarkers for cancer cachexia: is there also a genetic component to cachexia? Support Care Cancer. 2008;16(3):229–234. doi: 10.1007/s00520-007-0367-z. [DOI] [PubMed] [Google Scholar]
- 10.Zhang D, Zheng H, Zhou Y, Tang X, Yu B, Li J. Association of IL-1beta gene polymorphism with cachexia from locally advanced gastric cancer. BMC Cancer. 2007;7:45. doi: 10.1186/1471-2407-7-45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Rami-Porta R, Crowley JJ, Goldstraw P. The revised TNM stating system for lung cancer. Ann Thorac Cardiovasc Surg. 2009;15(1):4–9. [PubMed] [Google Scholar]
- 12.Bozetti F, Gavazzi C, Micelli R, et al. Perioperative total parenteral nutrition in malnourished, gastrointestinal patients. JPEN. 2000;24(1):7–14. doi: 10.1177/014860710002400107. [DOI] [PubMed] [Google Scholar]
- 13.Loprinzi CL, Ellison NM, Schaid DJ, Krook JE, Athmann LM, Dose AM, Mailliard JA, Johnson PS, Ebbert LP, Geeraerts LH. Controlled trial of megestrol acetate for the treatment of cancer anorexia and cachexia. J Natl Cancer Inst. 1990;82(13):1127–1132. doi: 10.1093/jnci/82.13.1127. [DOI] [PubMed] [Google Scholar]
- 14.Jatoi A, Egner J, Loprinzi CL, et al. Investigating the utility of serum cytokine measurements in a multi-institutional cancer anorexia/weight loss trial. Support Care Cancer. 2004;12(9):640–4. doi: 10.1007/s00520-004-0624-3. [DOI] [PubMed] [Google Scholar]
- 15.Dhiman N, Poland GA, Cunningham JM, et al. Variations in measles vaccine-specific humoral immunity by polymorphisms in SLAM and CD46 measles virus receptors. J Allerg Clin Immunol. 2007;120:666–672. doi: 10.1016/j.jaci.2007.04.036. [DOI] [PubMed] [Google Scholar]
- 16.Roledo R, Beggs W, Bender P. A simple and effective method for rapid genotyping of insertion/deletion polymorphisms. Genomics. 2003;82:580–582. doi: 10.1016/s0888-7543(03)00172-1. [DOI] [PubMed] [Google Scholar]
- 17.Schmidt C. SNPS not living up to promise; experts suggest new approach to disease ID. J National Cancer Institute. 2007;99(3):189–9. doi: 10.1093/jnci/djk076. [DOI] [PubMed] [Google Scholar]