Steven J. Skates

Multi-site assessment of the precision and reproducibility of multiple reaction monitoring-based measurements of proteins in plasma.

Verification of candidate biomarkers relies upon specific, quantitative assays optimized for selective detection of target proteins, and is increasingly viewed as a critical step in the discovery pipeline that bridges unbiased biomarker discovery to preclinical validation. Although individual laboratories have demonstrated that multiple reaction monitoring (MRM) coupled with isotope dilution mass spectrometry can quantify candidate protein biomarkers in plasma, reproducibility and transferability of these assays between laboratories have not been demonstrated. We describe a multilaboratory study to assess reproducibility, recovery, linear dynamic range and limits of detection and quantification of multiplexed, MRM-based assays, conducted by NCI-CPTAC. Using common materials and standardized protocols, we demonstrate that these assays can be highly reproducible within and across laboratories and instrument platforms, and are sensitive to low μg/ml protein concentrations in unfractionated plasma. We provide data and benchmarks against which individual laboratories can compare their performance and evaluate new technologies for biomarker verification in plasma.

AN ANALYSIS OF THE LOWEST EFFECTIVE INTENSITY OF PROPHYLACTIC ANTICOAGULATION FOR PATIENTS WITH NONRHEUMATIC ATRIAL FIBRILLATION

BACKGROUND To avert major hemorrhage, physicians need to know the lowest intensity of anticoagulation that is effective in preventing stroke in patients with atrial fibrillation. Since the low rate of stroke has made it difficult to perform prospective studies to resolve this issue, we conducted a case-control study. METHODS We studied 74 consecutive patients with atrial fibrillation who were admitted to our hospital from 1989 through 1994 after having an ischemic stroke while taking warfarin. For each patient with stroke, three controls with nonrheumatic atrial fibrillation who were treated as outpatients were randomly selected from the 1994 registry of the anticoagulant-therapy unit (222 controls). We used the international normalized ratio (INR) to measure the intensity of anticoagulation. For the patients with stroke, we used INR at admission; for the controls, we selected the INR that was measured closest to the month and day of the matched case patients hospital admission. RESULTS The risk of stroke rose steeply at INRs below 2.0. At an INR of 1.7, the adjusted odds ratio for stroke, as compared with the risk at an INR of 2.0, was 2.0 (95 percent confidence interval, 1.6 to 2.4); at an INR of 1.5, it was 3.3 (95 percent confidence interval, 2.4 to 4.6); and at an INR of 1.3, it was 6.0 (95 percent confidence interval, 3.6 to 9.8). Other independent risk factors were previous stroke (odds ratio, 10.4; 95 percent confidence interval, 4.4 to 24.5), diabetes mellitus (odds ratio, 2.95; 95 percent confidence interval, 1.3 to 6.5), hypertension (odds ratio, 2.5; 95 percent confidence interval, 1.1 to 5.7), and current smoking (odds ratio, 5.7; 95 percent confidence interval, 1.4 to 24.0). CONCLUSIONS Among patients with atrial fibrillation, anticoagulant prophylaxis is effective at INRs of 2.0 or greater. Since previous studies have indicated that the risk of hemorrhage rises rapidly at INRs greater than 4.0 to 5.0, tight control of anticoagulant therapy to maintain the INR between 2.0 and 3.0 is a better strategy than targeting lower, less effective levels of anticoagulation.

Screening for ovarian cancer: a pilot randomised controlled trial

BACKGROUND The value of screening for ovarian cancer is uncertain. We did a pilot randomised trial to assess multimodal screening with sequential CA 125 antigen and ultrasonography. METHODS Postmenopausal women aged 45 years or older were randomised to a control group (n=10,977) or screened group (n=10,958). Women randomised to screening were offered three annual screens that involved measurement of serum CA 125, pelvic ultrasonography if CA 125 was 30 U/mL or more, and referral for gynaecological opinion if ovarian volume was 8.8 mL or more on ultrasonography. All women were followed up to see whether they developed invasive epithelial cancers of the ovary or fallopian tube (index cancers). FINDINGS Of 468 women in the screened group with a raised CA 125, 29 were referred for a gynaecological opinion; screening detected an index cancer in six and 23 had false-positive screening results. The positive predictive value was 20.7%. During 7-year follow-up, ten further women with index cancers were identified in the screened group and 20 in the control group. Median survival of women with index cancers in the screened group was 72.9 months and in the control group was 41.8 months (p=0.0112). The number of deaths from an index cancer did not differ significantly between the control and screened groups (18 of 10,977 vs nine of 10,958, relative risk 2.0 [95% CI 0.78-5.13]). INTERPRETATION These results show that a multimodal approach to ovarian cancer screening in a randomised trial is feasible and justify a larger randomised trial to see whether screening affects mortality.

Proteogenomic characterization of human colon and rectal cancer

Extensive genomic characterization of human cancers presents the problem of inference from genomic abnormalities to cancer phenotypes. To address this problem, we analysed proteomes of colon and rectal tumours characterized previously by The Cancer Genome Atlas (TCGA) and perform integrated proteogenomic analyses. Somatic variants displayed reduced protein abundance compared to germline variants. Messenger RNA transcript abundance did not reliably predict protein abundance differences between tumours. Proteomics identified five proteomic subtypes in the TCGA cohort, two of which overlapped with the TCGA ‘microsatellite instability/CpG island methylation phenotype’ transcriptomic subtype, but had distinct mutation, methylation and protein expression patterns associated with different clinical outcomes. Although copy number alterations showed strong cis- and trans-effects on mRNA abundance, relatively few of these extend to the protein level. Thus, proteomics data enabled prioritization of candidate driver genes. The chromosome 20q amplicon was associated with the largest global changes at both mRNA and protein levels; proteomics data highlighted potential 20q candidates, including HNF4A (hepatocyte nuclear factor 4, alpha), TOMM34 (translocase of outer mitochondrial membrane 34) and SRC (SRC proto-oncogene, non-receptor tyrosine kinase). Integrated proteogenomic analysis provides functional context to interpret genomic abnormalities and affords a new paradigm for understanding cancer biology.

A novel multiple marker bioassay utilizing HE4 and CA125 for the prediction of ovarian cancer in patients with a pelvic mass

INTRODUCTION Patients diagnosed with epithelial ovarian cancer (EOC) have improved outcomes when cared for at centers experienced in the management of EOC. The objective of this trial was to validate a predictive model to assess the risk for EOC in women with a pelvic mass. METHODS Women diagnosed with a pelvic mass and scheduled to have surgery were enrolled on a multicenter prospective study. Preoperative serum levels of HE4 and CA125 were measured. Separate logistic regression algorithms for premenopausal and postmenopausal women were utilized to categorize patients into low and high risk groups for EOC. RESULTS Twelve sites enrolled 531 evaluable patients with 352 benign tumors, 129 EOC, 22 LMP tumors, 6 non EOC and 22 non ovarian cancers. The postmenopausal group contained 150 benign cases of which 112 were classified as low risk giving a specificity of 75.0% (95% CI 66.9-81.4), and 111 EOC and 6 LMP tumors of which 108 were classified as high risk giving a sensitivity of 92.3% (95% CI=85.9-96.4). The premenopausal group had 202 benign cases of which 151 were classified as low risk providing a specificity of 74.8% (95% CI=68.2-80.6), and 18 EOC and 16 LMP tumors of which 26 were classified as high risk, providing a sensitivity of 76.5% (95% CI=58.8-89.3). CONCLUSION An algorithm utilizing HE4 and CA125 successfully classified patients into high and low risk groups with 93.8% of EOC correctly classified as high risk. This model can be used to effectively triage patients to centers of excellence.

Proteogenomics connects somatic mutations to signalling in breast cancer

Summary Somatic mutations have been extensively characterized in breast cancer, but the effects of these genetic alterations on the proteomic landscape remain poorly understood. We describe quantitative mass spectrometry-based proteomic and phosphoproteomic analyses of 105 genomically annotated breast cancers of which 77 provided high-quality data. Integrated analyses allowed insights into the somatic cancer genome including the consequences of chromosomal loss, such as the 5q deletion characteristic of basal-like breast cancer. The 5q trans effects were interrogated against the Library of Integrated Network-based Cellular Signatures, thereby connecting CETN3 and SKP1 loss to elevated expression of EGFR, and SKP1 loss also to increased SRC. Global proteomic data confirmed a stromal-enriched group in addition to basal and luminal clusters and pathway analysis of the phosphoproteome identified a G Protein-coupled receptor cluster that was not readily identified at the mRNA level. Besides ERBB2, other amplicon-associated, highly phosphorylated kinases were identified, including CDK12, PAK1, PTK2, RIPK2 and TLK2. We demonstrate that proteogenomic analysis of breast cancer elucidates functional consequences of somatic mutations, narrows candidate nominations for driver genes within large deletions and amplified regions, and identifies therapeutic targets.

Ovarian cancer screening and mortality in the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS): a randomised controlled trial

Summary Background Ovarian cancer has a poor prognosis, with just 40% of patients surviving 5 years. We designed this trial to establish the effect of early detection by screening on ovarian cancer mortality. Methods In this randomised controlled trial, we recruited postmenopausal women aged 50–74 years from 13 centres in National Health Service Trusts in England, Wales, and Northern Ireland. Exclusion criteria were previous bilateral oophorectomy or ovarian malignancy, increased risk of familial ovarian cancer, and active non-ovarian malignancy. The trial management system confirmed eligibility and randomly allocated participants in blocks of 32 using computer-generated random numbers to annual multimodal screening (MMS) with serum CA125 interpreted with use of the risk of ovarian cancer algorithm, annual transvaginal ultrasound screening (USS), or no screening, in a 1:1:2 ratio. The primary outcome was death due to ovarian cancer by Dec 31, 2014, comparing MMS and USS separately with no screening, ascertained by an outcomes committee masked to randomisation group. All analyses were by modified intention to screen, excluding the small number of women we discovered after randomisation to have a bilateral oophorectomy, have ovarian cancer, or had exited the registry before recruitment. Investigators and participants were aware of screening type. This trial is registered with ClinicalTrials.gov, number NCT00058032. Findings Between June 1, 2001, and Oct 21, 2005, we randomly allocated 202 638 women: 50 640 (25·0%) to MMS, 50 639 (25·0%) to USS, and 101 359 (50·0%) to no screening. 202 546 (>99·9%) women were eligible for analysis: 50 624 (>99·9%) women in the MMS group, 50 623 (>99·9%) in the USS group, and 101 299 (>99·9%) in the no screening group. Screening ended on Dec 31, 2011, and included 345 570 MMS and 327 775 USS annual screening episodes. At a median follow-up of 11·1 years (IQR 10·0–12·0), we diagnosed ovarian cancer in 1282 (0·6%) women: 338 (0·7%) in the MMS group, 314 (0·6%) in the USS group, and 630 (0·6%) in the no screening group. Of these women, 148 (0·29%) women in the MMS group, 154 (0·30%) in the USS group, and 347 (0·34%) in the no screening group had died of ovarian cancer. The primary analysis using a Cox proportional hazards model gave a mortality reduction over years 0–14 of 15% (95% CI −3 to 30; p=0·10) with MMS and 11% (−7 to 27; p=0·21) with USS. The Royston-Parmar flexible parametric model showed that in the MMS group, this mortality effect was made up of 8% (−20 to 31) in years 0–7 and 23% (1–46) in years 7–14, and in the USS group, of 2% (−27 to 26) in years 0–7 and 21% (−2 to 42) in years 7–14. A prespecified analysis of death from ovarian cancer of MMS versus no screening with exclusion of prevalent cases showed significantly different death rates (p=0·021), with an overall average mortality reduction of 20% (−2 to 40) and a reduction of 8% (−27 to 43) in years 0–7 and 28% (−3 to 49) in years 7–14 in favour of MMS. Interpretation Although the mortality reduction was not significant in the primary analysis, we noted a significant mortality reduction with MMS when prevalent cases were excluded. We noted encouraging evidence of a mortality reduction in years 7–14, but further follow-up is needed before firm conclusions can be reached on the efficacy and cost-effectiveness of ovarian cancer screening. Funding Medical Research Council, Cancer Research UK, Department of Health, The Eve Appeal.

Calculation of the risk of ovarian cancer from serial CA-125 values for preclinical detection in postmenopausal women.

PURPOSE Previous studies of CA-125 levels from screening trials for ovarian cancer have indicated that serial CA-125 levels may identify cases better than a fixed CA-125 cutoff. We conducted a study to assess the screening performance of the risk of ovarian cancer calculation based on serial CA-125 levels from prospectively collected serum samples compared with a fixed CA-125 cutoff. PATIENTS AND METHODS The calculation was applied to data from a prospective trial of screening for ovarian cancer involving 22,000 postmenopausal women older than 45 years. The analysis was performed using 33,621 CA-125 results from 9,233 women for whom two or more serial samples were available. All serum samples from the patients with ovarian cancer were obtained before clinical detection. Sensitivity and specificity levels for preclinical detection of index cancers were calculated for various cutoffs for the risk and a single CA-125 measurement, and receiver operator curves were constructed. RESULTS The risk calculation significantly improved the area under the curve from 84% to 93% compared with a fixed cutoff for CA-125 (P =.01). For a target specificity of 98%, the risk achieved a sensitivity of 86% for preclinical detection of ovarian cancer, whereas CA-125 achieved a sensitivity of 62%. The estimates of performance are unbiased, because the risk calculation was derived independent of the data from this trial. CONCLUSION These results provide the first evidence that preclinical detection of ovarian cancer using serial CA-125 levels interpreted with the risk calculation significantly improves screening performance compared with a fixed cutoff for CA-125. The results justify the incorporation of the risk calculation in a prospective, randomized, controlled trial.

Prospective Study Using the Risk of Ovarian Cancer Algorithm to Screen for Ovarian Cancer

PURPOSE To evaluate prevalence screening in the first prospective trial of a new ovarian cancer screening (OCS) strategy (risk of ovarian cancer or ROC algorithm) on the basis of age and CA125 profile. PATIENTS AND METHODS Postmenopausal women, > or = 50 years were randomly assigned to a control group or screen group. Screening involved serum CA125, interpreted using the ROC algorithm. Participants with normal results returned to annual screening; those with intermediate results had repeat CA125 testing; and those with elevated values underwent transvaginal ultrasound (TVS). Women with abnormal or persistently equivocal TVS were referred for a gynecologic opinion. RESULTS Thirteen thousand five hundred eighty-two women were recruited. Of 6,682 women randomly assigned to screening, 6,532 women underwent the first screen. After the initial CA125, 5,213 women were classified as normal risk, 91 women elevated, and 1,228 women intermediate. On repeat CA125 testing of the latter, a further 53 women were classified as elevated risk. All 144 women with elevated risk had TVS. Sixteen women underwent surgery. Eleven women had benign pathology; one woman had ovarian recurrence of breast cancer; one woman had borderline; and three women had primary invasive epithelial ovarian cancer (EOC). The specificity and positive predictive value (PPV) for primary invasive EOC were 99.8% (95% CI, 99.7 to 99.9) and 19% (95% CI, 4.1 to 45.6), respectively. CONCLUSION An OCS strategy using the ROC algorithm is feasible and can achieve high specificity and PPV in postmenopausal women. It is being used in the United Kingdom Collaborative Trial of Ovarian Cancer Screening and in the United States in both the Cancer Genetics Network and the Gynecology Oncology Group trials of high-risk women.

Prevalence of Antineutrophil Cytoplasmic Antibodies in a Large Inception Cohort of Patients with Connective Tissue Disease

Antineutrophil cytoplasmic antibodies (ANCA) are strongly associated with the spectrum of vasculitis that includes Wegener granulomatosis, microscopic polyangiitis, the Churg-Strauss syndrome, idiopathic necrotizing and crescentic glomerulonephritis, and related or overlapping forms of vasculitis [1-3]. Other forms of vasculitis, including Takayasu arteritis, Henoch-Schonlein purpura, and cryoglobulinemia, are not associated with the presence of ANCA. Different assays have been used to test for ANCA, including indirect immunofluorescence and immunoassays that use either crude or highly purified preparations of specific antigens. Although several ANCA antigens have been described [2, 4], only antiproteinase 3 antibodies (anti-PR3) and antimyeloperoxidase antibodies (anti-MPO) have been shown to be of value in the diagnosis of vasculitis [1-3]. When used to stain ethanol-fixed, cytocentrifuged, normal human neutrophils by indirect immunofluorescence, anti-PR3 produce a cytoplasmic pattern of staining (C-ANCA) and anti-MPO produce a perinuclear or nuclear pattern (P-ANCA). At presentation, the clinical features of patients with Wegener granulomatosis, microscopic polyangiitis, and the Churg-Strauss syndrome may include glomerulonephritis, alveolar hemorrhage, tracheobronchitis, sinusitis, palpable purpura, arthritis, ocular inflammation, and neuropathy. Patients with connective tissue diseases may also display many of these features. Therefore, testing for ANCA, if highly specific, could be of great importance in the initial diagnostic evaluation of patients with a differential diagnosis that includes both connective tissue disease and vasculitis. Determination of the specificity of tests for ANCA in the diagnosis of vasculitis is crucial because the decision of whether to pursue biopsies or initiate potentially toxic immunosuppressive therapy may be made on the basis of results of such testing. We report the results of a blinded, controlled study to determine the prevalence of ANCA in a unique group of patients with various connective tissue diseases who were followed for as long as 5 years. A standard testing system, including indirect immunofluorescence and enzyme-linked immunosorbent assay (ELISA) for anti-PR3 and anti-MPO, was used to determine the prevalence of ANCA. Methods Patients The Early Undifferentiated Connective Tissue Disease project, a multicenter study funded by the National Institutes of Health through the Cooperative Systematic Studies of the Rheumatic Diseases Program, was designed to develop and prospectively follow a large cohort of patients with rheumatologic disease early in their clinical course. All patients were enrolled within 1 year of the onset of signs, symptoms, or serologic abnormalities that suggested connective tissue disease. Patients were evaluated at study entry and at years 1, 3, and 5. More than 800 clinical and laboratory variables were recorded for each patient according to a standardized protocol. Details of the original project and other study results have been published elsewhere [5-8]. Enrollment began in 1982 and was completed in June 1987. Patients with systemic lupus erythematosus, rheumatoid arthritis, inflammatory myositis, polymyositis or dermatomyositis, or scleroderma had to meet standardized criteria for the diagnosis of these diseases [9-12]. Early undifferentiated connective tissue disease (EUCTD) was diagnosed if patients did not meet criteria for the other connective tissue diseases and met specific criteria that have been described elsewhere [5]. We used serum samples that had been collected from the study patients at baseline. The original study enrolled 410 patients; for 386 (94%) of these, enough serum was available so that the patients could be included in our study. Final diagnoses were determined at the last visit and were therefore based on the cumulative data that had been collected. All analyses and results were based on the final diagnosis; as a result, patients were separated into the following diagnostic groups: systemic lupus erythematosus (n = 70), rheumatoid arthritis (n = 70), scleroderma (n = 45), polymyositis (n = 36), and EUCTD (n = 165). Within the original group, a subgroup of patients who had the Sjogren syndrome was identified. The Sjogren syndrome was defined by the presence, at any time during the study, of xerophthalmia (as determined by positive results on a Schirmer test); xerostomia; and positive results for any one of the following tests: antinuclear antibodies, rheumatoid factor, anti-Ro (anti-SS-A) antibody, or anti-La (anti-SS-B) antibody. All patients in the subgroup with the Sjogren syndrome also had a diagnosis of a primary connective tissue disease as outlined above. Forty-four patients met our definition for the Sjogren syndrome; these patients were drawn from all five diagnostic groups: systemic lupus erythematosus (n = 6), rheumatoid arthritis (n = 9), scleroderma (n = 5), polymyositis (n = 1), and EUCTD (n = 23). Serum specimens from 33 patients who were known to have the antiphospholipid syndrome [13, 14] with medium-to-high titers of IgG or IgM anticardiolipin antibodies (provided by EN Harris) were also studied. Serum samples from 200 random blood donors were collected through the Massachusetts General Hospital Blood Transfusion Service; these donors served as a control group. Serum samples were also collected from 52 patients with Wegener granulomatosis, microscopic polyangiitis, or related forms of vasculitis who had positive results on tests for ANCA; these patients were selected as positive controls for the ANCA assays. This control group of patients with vasculitis included 26 patients with anti-PR3 and 26 patients with anti-MPO; patients with high, low, and intermediate antibody titers were included. Serum Serum samples were both stored and shipped at 20C. All 671 samples, each of which had a unique identifier based on its original source, were assigned new, randomized, study identification numbers and were redivided and relabeled. The laboratory investigators who did the ANCA assays were thus blinded to the diagnosis for each patients sample. The code for the serum samples was not revealed until all data were collected and the analysis was ready to begin. Indirect Immunofluorescence for Antineutrophil Cytoplasmic Antibodies Indirect immunofluorescence was done as described elsewhere [15]. The results of staining were classified as having one of four patterns: C-ANCA (cytoplasmic), P-ANCA (perinuclear), atypical (neither cytoplasmic nor perinuclear), or negative. Because of the subjective nature of scoring the results of immunofluorescence for ANCA, each sample was stained twice and interpreted independently. Results of the first round of staining were interpreted by one observer, and results of the second round were interpreted by this observer and a second observer; both observers had considerable experience in interpreting the results of immunofluorescence staining of ANCA. If all three readings were the same, the interpretation was considered final. If the interpretations differed, a third slide was prepared and reexamined by the two observers. If at least three of the five interpretations matched, the results were considered final; if not, the staining results were considered to be atypical ANCA. Both observers were blinded to the previous results of immunofluorescence and ELISAs. Testing for Antineutrophil Cytoplasmic Antibodies by Enzyme-Linked Immunosorbent Assay We used direct antigen-specific ELISAs to detect anti-PR3 and anti-MPO, as described elsewhere [15-17]. A sandwich ELISA was also done on each sample. In the sandwich ELISA, monoclonal antibody 1E8 [18] was adhered to the wells of microtiter plates and used to bind proteinase 3. Subsequent steps were the same as those of the direct ELISA. An additional control in the sandwich ELISA for anti-PR3 was performed with selected serum specimens. To control for antibodies to the monoclonal catching antibody, additional wells were coated with monoclonal anti-PR3 catching antibody 1E8 but were not subsequently incubated with cytoplasmic extract of granulocytes. The reactivity of the serum to the monoclonal 1E8 alone was then subtracted from the reactivity to the 1E8-PR3 complex. The result is the titer for a revised sandwich ELISA for anti-PR3. Final Interpretation of Results of Testing for Antineutrophil Cytoplasmic Antibodies A final interpretation of ANCA test results was determined for each patient by using the results of immunofluorescence and ELISA. The set of decision rules used for the final interpretation of testing for anti-PR3 is outlined in Figure 1. A final interpretation for the presence of anti-MPO was considered positive only if samples were positive on immunofluorescence for P-ANCA or atypical ANCA patterns and on direct ELISA for anti-MPO. This is the same system that we use to provide a final interpretation for clinical samples submitted to our laboratory. Figure 1. Testing algorithm used for the final determination of the presence of antiproteinase 3 antibodies (anti-PR3). Statistical Analysis Comparisons between groups were analyzed by the Fisher exact test for categorical variables using a two-tailed significance level of 0.05. All data were stored on a SUN SPARC-5 workstation (Sun Microsystems, Mountain View, California) and analyzed using SAS software (SAS Institute, Cary, North Carolina) for UNIX. The 95% CIs for test specificity were determined using the methods described by Collett [19]. Results Indirect Immunofluorescence for Antineutrophil Cytoplasmic Antibodies The final results of immunofluorescence are shown in the (Table 1). None of the study patients or controls had C-ANCA by immunofluorescence staining. The rate of P-ANCA positivity by immunofluorescence was low for all study groups except patients with systemic lupus erythematosus, who had a rate of 31%. However, atypical patterns of ANCA immunofluorescence we