Infrared spectra of primary melanomas can predict response to chemotherapy: The example of dacarbazine.

Metastatic melanomas are highly aggressive and median survival is 6-9months for stage IV patients in the absence of treatment with anti-tumor activity. Dacarbazine is an alkylating agent that has been widely used in the treatment of metastatic melanomas and that could be still used in combination with targeted or immune therapies. Indeed, therapeutic benefits of these treatments in monotherapy are poor and one option to improve them is to combine drugs and/or to better anticipate the individual response to a defined treatment. To our best knowledge and to date, there is no test available to predict the response of a patient to dacarbazine. We show here that examination of melanoma histological sections by infrared micro-spectroscopy reveals the sensitivity of the cancer to dacarbazine. Unsupervised analysis of the FTIR spectra evidences spontaneous and significant clustering of infrared spectra into two groups that match the clinical responsiveness of the patients to dacarbazine used as a first-line treatment. A supervised model resulted in 83% of the patient status (responder/non-responder) being correctly identified. The spectra revealed a key modification in the nature and quantity of lipids in the cells of both groups.

Identification of melanoma cells and lymphocyte subpopulations in lymph node metastases by FTIR imaging histopathology.

While early stages of melanoma are usually cured by surgery, metastatic melanomas are difficult to treat because the widely available options have low response rates. Careful and precise diagnosis and staging are essential to determine patient's risk and to select appropriate treatments. Fortunately, the recent progress in immunotherapy is very encouraging. In this context, it is important to characterize the intratumoral infiltration of immune cells in each patient, which is however not done routinely due to the lack of standardized methods. In this study, we used Fourier transform infrared (FTIR) imaging combined with multivariate statistical analyses to investigate non-metastatic and metastatic lymph nodes from melanoma patients. Our results show that the different cell types have different infrared spectral features allowing automated identification of these cell types. High recognition rates were obtained using a supervised partial least square discriminant analysis (PLS-DA) model. Melanoma cells were recognized with 87.1% sensitivity and 85.7% specificity, showing that FTIR spectroscopy has similar detection power as immunohistochemistry. Besides, FTIR imaging could also distinguish lymphocyte subpopulations (B and T cells). Finally, we investigated the changes in lymphocytes due to the presence of metastases. Interestingly, specific features of spectra of lymphocytes present in metastatic or tumor-free lymph nodes could be evidenced by PCA. A PLS-DA model was capable of predicting whether lymphocytes originated from invaded or non-invaded lymph nodes. These data demonstrate that FTIR imaging is capable to distinguish known and also novel biological features in human tissues, with potential practical relevance for histopathological diagnosis and biomarker assessment.

UK Lung Cancer RCT Pilot Screening Trial: baseline findings from the screening arm provide evidence for the potential implementation of lung cancer screening.

BACKGROUND: Lung cancer screening using low-dose CT (LDCT) was shown to reduce lung cancer mortality by 20% in the National Lung Screening Trial. METHODS: The pilot UK Lung Cancer Screening (UKLS) is a randomised controlled trial of LDCT screening for lung cancer versus usual care. A population-based questionnaire was used to identify high-risk individuals. CT screen-detected nodules were managed by a pre-specified protocol. Cost effectiveness was modelled with reference to the National Lung Cancer Screening Trial mortality reduction. RESULTS: 247 354 individuals aged 50-75 years were approached; 30.7% expressed an interest, 8729 (11.5%) were eligible and 4055 were randomised, 2028 into the CT arm (1994 underwent a CT). Forty-two participants (2.1%) had confirmed lung cancer, 34 (1.7%) at baseline and 8 (0.4%) at the 12-month scan. 28/42 (66.7%) had stage I disease, 36/42 (85.7%) had stage I or II disease. 35/42 (83.3%) had surgical resection. 536 subjects had nodules greater than 50 mm(3) or 5 mm diameter and 41/536 were found to have lung cancer. One further cancer was detected by follow-up of nodules between 15 and 50 mm(3) at 12 months. The baseline estimate for the incremental cost-effectiveness ratio of once-only CT screening, under the UKLS protocol, was £8466 per quality adjusted life year gained (CI £5542 to £12 569). CONCLUSIONS: The UKLS pilot trial demonstrated that it is possible to detect lung cancer at an early stage and deliver potentially curative treatment in over 80% of cases. Health economic analysis suggests that the intervention would be cost effective-this needs to be confirmed using data on observed lung cancer mortality reduction. TRIAL REGISTRATION: ISRCTN 78513845.

Prevention of neural tube defects in the UK: a missed opportunity.

OBJECTIVE: In 1991, the Medical Research Council (MRC) Vitamin Study demonstrated that folic acid taken before pregnancy and in early pregnancy reduced the risk of a neural tube defect (NTD). We aimed to estimate the number of NTD pregnancies that would have been prevented if flour had been fortified with folic acid in the UK from 1998 as it had been in the USA. DESIGN: Estimates of NTD prevalence, the preventive effect of folic acid and the proportion of women taking folic acid supplements before pregnancy were used to predict the number of NTD pregnancies that would have been prevented if folic acid fortification had been implemented. SETTING: Eight congenital anomaly registers in England and Wales. MAIN OUTCOME MEASURES: The prevalence of pregnancies with an NTD in the UK and the number of these pregnancies that would have been prevented if folic acid fortification had been implemented. RESULTS: From 1991 to 2012, the prevalence of NTD pregnancies was 1.28 (95% CI 1.24 to 1.31) per 1000 total births (19% live births, 81% terminations and 0.5% stillbirths and fetal deaths ≥20 weeks' gestation). If the USA levels of folic acid fortification from 1998 onwards had been adopted in the UK, an estimated 2014 fewer NTD pregnancies would have occurred. CONCLUSIONS: Failure to implement folic acid fortification in the UK has caused, and continues to cause, avoidable terminations of pregnancy, stillbirths, neonatal deaths and permanent serious disability in surviving children.

An infrared spectral signature of human lymphocyte subpopulations from peripheral blood.

Metastatic melanomas are frequently refractory to most adjuvant therapies such as chemotherapies and radiotherapies. Recently, immunotherapies have shown good results in the treatment of some metastatic melanomas. Immune cell infiltration in the tumor has been associated with successful immunotherapy. More generally, tumor infiltrating lymphocytes (TILs) in the primary tumor and in metastases of melanoma patients have been demonstrated to correlate positively with favorable clinical outcomes. Altogether, these findings suggest the importance of being able to identify, quantify and characterize immune infiltration at the tumor site for a better diagnostic and treatment choice. In this paper, we used Fourier Transform Infrared (FTIR) imaging to identify and quantify different subpopulations of T cells: the cytotoxic T cells (CD8+), the helper T cells (CD4+) and the regulatory T cells (T reg). As a proof of concept, we investigated pure populations isolated from human peripheral blood from 6 healthy donors. These subpopulations were isolated from blood samples by magnetic labeling and purities were assessed by Fluorescence Activated Cell Sorting (FACS). The results presented here show that Fourier Transform Infrared (FTIR) imaging followed by supervised Partial Least Square Discriminant Analysis (PLS-DA) allows an accurate identification of CD4+ T cells and CD8+ T cells (>86%). We then developed a PLS regression allowing the quantification of T reg in a different mix of immune cells (e.g. Peripheral Blood Mononuclear Cells (PBMCs)). Altogether, these results demonstrate the sensitivity of infrared imaging to detect the low biological variability observed in T cell subpopulations.

Infrared imaging of primary melanomas reveals hints of regional and distant metastases.

Melanoma is the deadliest form of skin cancer. Metastatic melanomas are resistant to almost all existing adjuvant therapies such as chemotherapy and radiotherapy, so detection of metastases remains a challenge, and no biomarkers are currently available to detect primary tumors with the highest risk of metastasis. Results presented in this paper show that Fourier Transform Infrared (FTIR) imaging of histological sections followed by supervised partial least squares discriminant analysis (PLS-DA) can accurately (>90%) identify the main cell types commonly found in melanoma tumors. Here we define six cell types: melanoma cells, erythrocytes, connective tissue (includes blood vessel walls, dermis and collagen regions), keratinocytes, lymphocytes and necrotic cells. Interestingly, more than 98% of the melanoma cells are correctly identified. The spectra of the cells identified as melanomas were then further analyzed. First, we compared melanoma cells in primary tumors (from 26 patients) with melanoma cells from metastases (from 25 patients). Neither supervised nor unsupervised analyses revealed any significant difference. Similarly, we found no significant correlation between the infrared spectra of melanoma cells and the number of proliferative cells assessed by Ki67 immunostaining. Finally, we compared the infrared spectra of primary tumors from patients diagnosed at different stages of the disease. Infrared spectroscopy was capable of pointing out differences between primary tumors of patients at stage I or II and patients at stage III or IV, even by unsupervised analysis. We then developed a supervised PLS-DA model capable of predicting whether tumor cells belonged to one of the two aggregated disease stage groups. The model predicted a high rate of true positives (sensitivity of 88.9%) and a good rate of true negatives (specificity of 70.6%) in external validation. These results demonstrate that infrared spectroscopy can be used to help identify melanoma characteristics related to the cells' invasive capability.

Is the area under an ROC curve a valid measure of the performance of a screening or diagnostic test?

OBJECTIVES: The area under a receiver operating characteristic (ROC) curve (the AUC) is used as a measure of the performance of a screening or diagnostic test. We here assess the validity of the AUC. METHODS: Assuming the test results follow Gaussian distributions in affected and unaffected individuals, standard mathematical formulae were used to describe the relationship between the detection rate (DR) (or sensitivity) and the false-positive rate (FPR) of a test with the AUC. These formulae were used to calculate the screening performance (DR for a given FPR, or FPR for a given DR) for different AUC values according to different standard deviations of the test result in affected and unaffected individuals. RESULTS: The DR for a given FPR is strongly dependent on relative differences in the standard deviation of the test variable in affected and unaffected individuals. Consequently, two tests with the same AUC can have a different DR for the same FPR. For example, an AUC of 0.75 has a DR of 24% for a 5% FPR if the standard deviations are the same in affected and unaffected individuals, but 39% for the same 5% FPR if the standard deviation in affected individuals is 1.5 times that in unaffected individuals. CONCLUSION: The AUC is an unreliable measure of screening performance because in practice the standard deviation of a screening or diagnostic test in affected and unaffected individuals can differ. The problem is avoided by not using AUC at all, and instead specifying DRs for given FPRs or FPRs for given DRs.

Breast cancer and melanoma cell line identification by FTIR imaging after formalin-fixation and paraffin-embedding.

Over the past few decades, Fourier transform infrared (FTIR) spectroscopy coupled to microscopy has been recognized as an emerging and potentially powerful tool in cancer research and diagnosis. For this purpose, histological analyses performed by pathologists are mostly carried out on biopsied tissue that undergoes the formalin-fixation and paraffin-embedding (FFPE) procedure. This processing method ensures an optimal and permanent preservation of the samples, making FFPE-archived tissue an extremely valuable source for retrospective studies. Nevertheless, as highlighted by previous studies, this fixation procedure significantly changes the principal constituents of cells, resulting in important effects on their infrared (IR) spectrum. Despite the chemical and spectral influence of FFPE processing, some studies demonstrate that FTIR imaging allows precise identification of the different cell types present in biopsied tissue, indicating that the FFPE process preserves spectral differences between distinct cell types. In this study, we investigated whether this is also the case for closely related cell lines. We analyzed spectra from 8 cancerous epithelial cell lines: 4 breast cancer cell lines and 4 melanoma cell lines. For each cell line, we harvested cells at subconfluence and divided them into two sets. We first tested the "original" capability of FTIR imaging to identify these closely related cell lines on cells just dried on BaF2 slides. We then repeated the test after submitting the cells to the FFPE procedure. Our results show that the IR spectra of FFPE processed cancerous cell lines undergo small but significant changes due to the treatment. The spectral modifications were interpreted as a potential decrease in the phospholipid content and protein denaturation, in line with the scientific literature on the topic. Nevertheless, unsupervised analyses showed that spectral proximities and distances between closely related cell lines were mostly, but not entirely, conserved after FFPE processing. Finally, PLS-DA statistical analyses highlighted that closely related cell lines are still successfully identified and efficiently distinguished by FTIR spectroscopy after FFPE treatment. This last result paves the way towards identification and characterization of cellular subtypes on FFPE tissue sections by FTIR imaging, indicating that this analysis technique could become a potential useful tool in cancer research.

Direct and indirect tasks on assessment of dose and time distributions and thresholds of acute radiation exposure.

Mathematical methods were developed to construct dose and time distributions and their associated risks and threshold values for lethal and non-lethal effects of acute radiation exposure to include mortality and incidence, prodromal vomiting, and agranulocytosis. A new distribution (T-model) was obtained to describe time parameters of acute radiation syndrome such as the latency period, time to onset of vomiting, and time to initiation of agranulocytosis. Based on the dose and time distributions, the parameter translation method was defined using an orthogonal regression, which allows one to solve for these distributions in the case of acute radiation exposure. The assessment of threshold doses was performed for some effects of acute radiation syndrome: for the latency period, ∼6-8 Gy absorbed dose and ∼0.7-0.9 h time to onset of vomiting; and for incidence (agranulocytosis), ∼2-3 Gy absorbed dose and ∼2-3 h time to onset of vomiting. The obtained new formula for assessment of radiation risk is applicable to the time parameters of acute radiation syndrome.

UK Lung Screen (UKLS) nodule management protocol: modelling of a single screen randomised controlled trial of low-dose CT screening for lung cancer.

The UK Lung Screen (UKLS) is a randomised controlled trial of the use of low-dose multidetector CT for lung cancer screening. It completed the Health Technology Appraisal (HTA)-funded feasibility stage in October 2009 and the pilot UKLS will be initiated in early 2011. The pilot will randomise 4000 subjects to either low-dose CT screening or no screening. The full study, due to start in September 2012, if progression criteria are met, will randomise a further 28,000 subjects from seven centres in the UK. Subjects will be selected if they have sufficient risk of developing lung cancer according to the Liverpool Lung Project risk model. The UKLS employs the 'Wald Single Screen Design', which was modelled in the UKLS feasibility study. This paper describes the modelling of nodule management in UKLS by using volumetric analysis with a single initial screen design and follow-up period of 10 years. This modelling has resulted in the development and adoption of the UKLS care pathway, which will be implemented in the planned CT screening trial in the UK.

The Radiation Injury Severity Classification system: an early injury assessment tool for the frontline health-care provider.

Our goal was to adapt current diagnostic methods for radiation overexposure patients into a practical system that can be implemented rapidly and reliably by responders unfamiliar with the effects of radiation. Our Radiation Injury Severity Classification (RISC) system uses clinical and haematological parameters from the prodromal phase of the acute radiation syndrome (ARS) to classify acute radiation injury for purposes of managing treatment disposition. Data from well-documented ARS cases were used to test the RISC system. Three-day summaries were generated for each case. These were individually reviewed by the three physicians most involved with the development of the system to establish both a consensus case score (CCS) and disposition category ranges. 30 volunteer raters from varying health disciplines using the RISC system then each independently rated a random selection of 12 cases for injury severity in a self-trained field-simulation exercise. The CCS identified discrete cut-off ranges for the three disposition categories in both manageable and mass casualty events. The group of raters, after a modest period of self-training, achieved overall levels of pairwise agreement with the CCS category of 0.944 for manageable events and 0.947 for mass casualty situations. In conclusion, an early assessment of the severity of the ARS injury is required for an appropriate disposition determination. The RISC system should produce reasonably accurate and reliable assessments of radiation injury severity within 6-12 hours post exposure despite the probable absence of physical dosimetric data.

The effect of correlations between screening markers on screening performance.

OBJECTIVES: It is widely thought that correlations between screening markers will tend to degrade screening performance. We performed a computer simulation study to investigate the quantitative effect of correlations between two markers on screening performance, using prenatal screening for Down's syndrome as an example, although the results apply generally. METHODS: Monte Carlo simulation was used to generate values of two hypothetical markers, A and B, in 1000 affected and 1000 unaffected pregnancies. The means, standard deviations and correlations of A and B were varied in five different examples. RESULTS: If markers A and B are, on average, higher in affected than unaffected pregnancies and each marker, individually, has the same detection rate for a given false-positive rate (i.e. the same screening performance), then the screening performance of A and B together tends to decrease as A and B become more positively correlated with each other (within affected or unaffected categories) and tends to increase as A and B become more negatively correlated. If A is, on average, higher in affected pregnancies and B is, on average, lower in affected pregnancies (but again each marker has the same screening performance), the opposite pattern is observed; screening performance increases as A and B become more positively correlated and screening performance decreases as they become more negatively correlated. If A and B have unequal screening performances, modest correlations between A and B have little effect on the screening performance of A and B together, but when the correlations are strong whether positive or negative (with r values greater than about 0.45 or less than -0.45) screening performance progressively increases. CONCLUSION: Correlations between screening markers considered separately in affected and unaffected pregnancies can either decrease or increase screening performance. In practice, these effects are usually modest, because most screening markers are not highly correlated with each other and the effects become important only with strong correlations, whether positive or negative.

Prevalence of neural tube defect pregnancies in England and Wales from 1964 to 2004.

OBJECTIVES: To determine the prevalence of pregnancies with a neural tube defect (NTD) in England and Wales between 1964 and 2004 and to estimate the relative impact of antenatal screening and a change in the incidence of these defects on the prevalence of births with NTDs. SETTINGS: Use of data published by the Office for National Statistics (ONS) on terminations of pregnancies with an NTD and births with an NTD from 1964 to 2004. METHODS: Estimates were made of the total number of terminations of pregnancies and births with an NTD by taking account of the under-reporting of these terminations and births using a previously described method. In 1995 ONS started to report the number of terminations with an NTD rather than the total numbers of terminations with a central nervous system (CNS) malformation that had previously been used to estimate the number of NTD terminations. Adjustment was made for this and new estimates of the total number of NTD pregnancies were produced to 2004. RESULTS: There were an estimated 969 pregnancies with NTDs (168 (17%) births and 801(83%) terminations) in England and Wales in 2004. An estimated 44% of NTD terminations and 32% of births were not reported as such. The birth prevalence per 1000 decreased fallen 93% from 3.6 in 1964 to 0.3 in 2004, 59% due to an underlying decrease in the prevalence of NTDs and 34% due to screening diagnosis and subsequent termination of affected pregnancies. CONCLUSION: The prevalence of NTD pregnancies decreased by around two per 1000 from 1964 to 1990 and thereafter remained fairly constant. The prevalence of NTD pregnancies is substantially underestimated if it is based only on reported NTD births (by 88%) and also if it is based on reported NTD births and terminations (by 52%), because most NTD pregnancies in England and Wales are terminated following antenatal screening and most of these terminations are not reported.

Insulin-like growth factors and cancer: no role in screening. Evidence from the BUPA study and meta-analysis of prospective epidemiological studies.

Insulin-like growth factor-1 (IGF-1), insulin-like growth factor-2 (IGF-2), and insulin-like growth factor binding protein-3 (IGFBP-3) were measured in frozen serum samples from 1051 men with cancer and 3142 controls in a nested case-control study from the British United Provident Association (BUPA) study cohort and associations with 14 cancers were examined, including prostate, colorectal, and lung. A meta-analysis of studies on these three cancer sites was also conducted. In the meta-analysis the odds ratio between the highest quartile IGF-1 group and the lowest quartile group was 1.31 (95% confidence interval (CI): 1.03-1.67) for prostate, 1.37 (1.05-1.78) for colorectal and 1.02 (0.80-1.31) for lung cancer, and for IGF-2 it was 0.72 (0.36-1.44) for prostate and 1.95 (1.26-3.00) for colorectal cancer. Results from the BUPA study were consistent with the estimates from the other studies. There were no statistically significant associations with IGFBP-3 and any of the cancer sites considered. Our results suggest that IGF-1, IGF-2, and IGFBP-3 measurements have no value in cancer screening, although IGF-1 and IGF-2 may be of aetiological significance in relation to colorectal and prostate cancer.

Graphical presentation of distributions of risk in screening.

OBJECTIVE: The screening performance of tests involving multiple markers is usually presented visually as two Gaussian relative frequency distributions of risk, one curve relating to affected and the other to unaffected individuals. If the distribution of the underlying screening markers is approximately Gaussian, risk estimates based on the same markers will usually also be approximately Gaussian. However, this approximation sometimes fails. Here we examine the circumstances when this occurs. SETTING: A theoretical statistical analysis. METHODS: Hypothetical log Gaussian relative distributions of affected and unaffected individuals were generated for three antenatal screening markers for Down's syndrome. Log likelihood ratios were calculated for each marker value and plots of the relative frequency distributions were compared with plots of Gaussian distributions based on the means and standard deviations of these log likelihood ratios. RESULTS: When the standard deviations of the distributions of a perfectly Gaussian screening marker are similar in affected and unaffected individuals, the distributions of risk estimates are also approximately Gaussian. If the standard deviations differ materially, incorrectly assuming that the distributions of the risk estimates are Gaussian creates a graphical anomaly in which the distributions of risk in affected and unaffected individuals plotted on a continuous risk scale intersect in two places. This is theoretically impossible. Plotting the risk distributions empirically reveals that all individuals have an estimated risk above a specified value. For individuals with more extreme marker values, the risk estimates reverse and increase instead of continuing to decrease. CONCLUSION: It is useful to check whether a Gaussian approximation for the distribution of risk estimates based on a screening marker is valid. If the value of the marker level at which risk reversal occurs lies within the set truncation limits, these may need to be reset, and a Gaussian model may be inappropriate to illustrate the risk distributions.