Eye Tracking for Deep Learning Segmentation Using Convolutional Neural Networks.

Deep learning with convolutional neural networks (CNNs) has experienced tremendous growth in multiple healthcare applications and has been shown to have high accuracy in semantic segmentation of medical (e.g., radiology and pathology) images. However, a key barrier in the required training of CNNs is obtaining large-scale and precisely annotated imaging data. We sought to address the lack of annotated data with eye tracking technology. As a proof of principle, our hypothesis was that segmentation masks generated with the help of eye tracking (ET) would be very similar to those rendered by hand annotation (HA). Additionally, our goal was to show that a CNN trained on ET masks would be equivalent to one trained on HA masks, the latter being the current standard approach. Step 1: Screen captures of 19 publicly available radiologic images of assorted structures within various modalities were analyzed. ET and HA masks for all regions of interest (ROIs) were generated from these image datasets. Step 2: Utilizing a similar approach, ET and HA masks for 356 publicly available T1-weighted postcontrast meningioma images were generated. Three hundred six of these image + mask pairs were used to train a CNN with U-net-based architecture. The remaining 50 images were used as the independent test set. Step 1: ET and HA masks for the nonneurological images had an average Dice similarity coefficient (DSC) of 0.86 between each other. Step 2: Meningioma ET and HA masks had an average DSC of 0.85 between each other. After separate training using both approaches, the ET approach performed virtually identically to HA on the test set of 50 images. The former had an area under the curve (AUC) of 0.88, while the latter had AUC of 0.87. ET and HA predictions had trimmed mean DSCs compared to the original HA maps of 0.73 and 0.74, respectively. These trimmed DSCs between ET and HA were found to be statistically equivalent with a p value of 0.015. We have demonstrated that ET can create segmentation masks suitable for deep learning semantic segmentation. Future work will integrate ET to produce masks in a faster, more natural manner that distracts less from typical radiology clinical workflow.

International Working Group consensus response evaluation criteria in lymphoma (RECIL 2017).

In recent years, the number of approved and investigational agents that can be safely administered for the treatment of lymphoma patients for a prolonged period of time has substantially increased. Many of these novel agents are evaluated in early-phase clinical trials in patients with a wide range of malignancies, including solid tumors and lymphoma. Furthermore, with the advances in genome sequencing, new "basket" clinical trial designs have emerged that select patients based on the presence of specific genetic alterations across different types of solid tumors and lymphoma. The standard response criteria currently in use for lymphoma are the Lugano Criteria which are based on [18F]2-fluoro-2-deoxy-D-glucose positron emission tomography or bidimensional tumor measurements on computerized tomography scans. These differ from the RECIST criteria used in solid tumors, which use unidimensional measurements. The RECIL group hypothesized that single-dimension measurement could be used to assess response to therapy in lymphoma patients, producing results similar to the standard criteria. We tested this hypothesis by analyzing 47 828 imaging measurements from 2983 individual adult and pediatric lymphoma patients enrolled on 10 multicenter clinical trials and developed new lymphoma response criteria (RECIL 2017). We demonstrate that assessment of tumor burden in lymphoma clinical trials can use the sum of longest diameters of a maximum of three target lesions. Furthermore, we introduced a new provisional category of a minor response. We also clarified response assessment in patients receiving novel immune therapy and targeted agents that generate unique imaging situations.

A Multiparametric Model for Mapping Cellularity in Glioblastoma Using Radiographically Localized Biopsies.

BACKGROUND AND PURPOSE: The complex MR imaging appearance of glioblastoma is a function of underlying histopathologic heterogeneity. A better understanding of these correlations, particularly the influence of infiltrating glioma cells and vasogenic edema on T2 and diffusivity signal in nonenhancing areas, has important implications in the management of these patients. With localized biopsies, the objective of this study was to generate a model capable of predicting cellularity at each voxel within an entire tumor volume as a function of signal intensity, thus providing a means of quantifying tumor infiltration into surrounding brain tissue. MATERIALS AND METHODS: Ninety-one localized biopsies were obtained from 36 patients with glioblastoma. Signal intensities corresponding to these samples were derived from T1-postcontrast subtraction, T2-FLAIR, and ADC sequences by using an automated coregistration algorithm. Cell density was calculated for each specimen by using an automated cell-counting algorithm. Signal intensity was plotted against cell density for each MR image. RESULTS: T2-FLAIR (r = -0.61) and ADC (r = -0.63) sequences were inversely correlated with cell density. T1-postcontrast (r = 0.69) subtraction was directly correlated with cell density. Combining these relationships yielded a multiparametric model with improved correlation (r = 0.74), suggesting that each sequence offers different and complementary information. CONCLUSIONS: Using localized biopsies, we have generated a model that illustrates a quantitative and significant relationship between MR signal and cell density. Projecting this relationship over the entire tumor volume allows mapping of the intratumoral heterogeneity in both the contrast-enhancing tumor core and nonenhancing margins of glioblastoma and may be used to guide extended surgical resection, localized biopsies, and radiation field mapping.

Comparative Effects of CT Imaging Measurement on RECIST End Points and Tumor Growth Kinetics Modeling.

Quantitative assessments of tumor burden and modeling of longitudinal growth could improve phase II oncology trials. To identify obstacles to wider use of quantitative measures we obtained recorded linear tumor measurements from three published lung cancer trials. Model-based parameters of tumor burden change were estimated and compared with similarly sized samples from separate trials. Time-to-tumor growth (TTG) was computed from measurements recorded on case report forms and a second radiologist blinded to the form data. Response Evaluation Criteria in Solid Tumors (RECIST)-based progression-free survival (PFS) measures were perfectly concordant between the original forms data and the blinded radiologist re-evaluation (intraclass correlation coefficient = 1), but these routine interrater differences in the identification and measurement of target lesions were associated with an average 18-week delay (range, -20 to 55 weeks) in TTG (intraclass correlation coefficient = 0.32). To exploit computational metrics for improving statistical power in small clinical trials will require increased precision of tumor burden assessments.

Semiautomated volumetric measurement on postcontrast MR imaging for analysis of recurrent and residual disease in glioblastoma multiforme.

BACKGROUND AND PURPOSE: A limitation in postoperative monitoring of patients with glioblastoma is the lack of objective measures to quantify residual and recurrent disease. Automated computer-assisted volumetric analysis of contrast-enhancing tissue represents a potential tool to aid the radiologist in following these patients. In this study, we hypothesize that computer-assisted volumetry will show increased precision and speed over conventional 1D and 2D techniques in assessing residual and/or recurrent tumor. MATERIALS AND METHODS: This retrospective study included patients with native glioblastomas with MR imaging performed at 24-48 hours following resection and 2-4 months postoperatively. 1D and 2D measurements were performed by 2 neuroradiologists with Certificates of Added Qualification. Volumetry was performed by using manual segmentation and computer-assisted volumetry, which combines region-based active contours and a level set approach. Tumor response was assessed by using established 1D, 2D, and volumetric standards. Manual and computer-assisted volumetry segmentation times were compared. Interobserver correlation was determined among 1D, 2D, and volumetric techniques. RESULTS: Twenty-nine patients were analyzed. Discrepancy in disease status between 1D and 2D compared with computer-assisted volumetry was 10.3% (3/29) and 17.2% (5/29), respectively. The mean time for segmentation between manual and computer-assisted volumetry techniques was 9.7 minutes and <1 minute, respectively (P < .01). Interobserver correlation was highest for volumetric measurements (0.995; 95% CI, 0.990-0.997) compared with 1D (0.826; 95% CI, 0.695-0.904) and 2D (0.905; 95% CI, 0.828-0.948) measurements. CONCLUSIONS: Computer-assisted volumetry provides a reproducible and faster volumetric assessment of enhancing tumor burden, which has implications for monitoring disease progression and quantification of tumor burden in treatment trials.

Analysis of plasma cytokines and angiogenic factors in patients with pretreated urothelial cancer receiving Pazopanib: the role of circulating interleukin-8 to enhance the prognostic accuracy.

BACKGROUND: Pazopanib achieved the end point of clinical activity in pretreated patients with urothelial cancer in a single-group, phase 2 trial. The objective was to identify biological predictors of clinical benefit to pazopanib in these patients. METHODS: EDTA blood samples were collected at baseline (T0) and after 4 weeks (T1) of treatment, together with radiological imaging in all 41 patients to analyse plasma circulating angiogenic factor levels by multiplex ELISA plates. Changes from T0 to T1 in marker levels were matched with response with the covariance analysis. Univariable and multivariable analyses evaluated the association with overall survival (OS), adjusted for prespecified clinical variables. Net reclassification improvement (NRI) tested the performance of the recognised Cox model. RESULTS: Increasing IL8(T1) level associated with lower response probability at covariance analysis (P=0.010). Both IL8(T0) (P=0.019) and IL8(T1) (P=0.004) associated with OS and the prognostic model, including clinical variables and IL8(T1) best-predicted OS after backward selection. The NRI for this model was 39%.When analysed as a time-varying covariate, IL8(T1) level<80 pg ml(-1) portended significantly greater response (∼80%) and 6-month OS (∼60%) probability than level ≥ 80. CONCLUSION: IL8-level changes during pazopanib allowed for a prognostic improvement and were associated with response probability.

Prognostic aspects of DCE-MRI in recurrent rectal cancer.

OBJECTIVE: To explore whether pre-reoperative dynamic contrast-enhanced (DCE)-MRI findings correlate with clinical outcome in patients who undergo surgical treatment for recurrent rectal carcinoma. METHODS: A retrospective study of DCE-MRI in patients with recurrent rectal cancer was performed after obtaining an IRB waiver. We queried our PACS from 1998 to 2012 for examinations performed for recurrent disease. Two radiologists in consensus outlined tumour regions of interest on perfusion images. We explored the correlation between K(trans), Kep, Ve, AUC90 and AUC180 with time to re-recurrence of tumour, overall survival and resection margin status. Univariate Cox PH models were used for survival, while univariate logistic regression was used for margin status. RESULTS: Among 58 patients with pre-treatment DCE-MRI who underwent resection, 36 went directly to surgery and 18 had positive margins. K(trans) (0.55, P = 0.012) and Kep (0.93, P = 0.04) were inversely correlated with positive margins. No significant correlations were noted between K(trans), Kep, Ve, AUC90 and AUC180 and overall survival or time to re-recurrence of tumour. CONCLUSION: K(trans) and Kep were significantly associated with clear resection margins; however overall survival and time to re-recurrence were not predicted. Such information might be helpful for treatment individualisation and deserves further investigation.

Contrast-enhanced ultrasound monitoring of perfusion changes in hepatic neuroendocrine metastases after systemic versus selective arterial 177Lu/90Y-DOTATOC and 213Bi-DOTATOC radiopeptide therapy.

AIM: Radiopeptide therapy with beta emitter labeled (177)Lu/(90)Y- DOTA(0)-Phe(1)-Tyr(3)-octreotide (DOTATOC) and more recently also alpha emitting (213)Bi-DOTATOC are promising new treatments for neuroendocrine tumors. No early predictors for treatment response have been recognized and tumor-shrinkage after radiation therapy appears slowly. In some solid tumors a decline in tumor perfusion was found predictive of final treatment response but the gold standard multiphase computed tomography (CT) has a high radiation burden. Therefore we evaluated the ability of contrast-enhanced ultrasound (CEUS) to evaluate tumor perfusion as a response criteria. MATERIALS AND METHODS: 14 patients with hepatic neuroendocrine tumor (NET) metastases were enrolled in the retrospective study. Eleven patients were treated with beta-emitting (177)Lu/(90)Y-DOTATOC, either intravenous (i.v.) (n = 5) or intra-arterial (i.a.) (n = 6) and three patients received alpha-emitting (213)Bi-DOTATOC (i.a.). CEUS and contrast-enhanced CT (CE-CT) were performed before and 3 months after treatment. RESULTS: CE-CT and CEUS presented comparable results in the baseline study and in the assessment of perfusion changes due to the different treatment regimes. A therapy related decrease in tumor perfusion is an early predictor of longterm morphologic response. CONCLUSION: CEUS is available and radiation free technique which showed comparable results for perfusion and diameter of liver metastases compared to CE-CT. Intensity reduction in an arterial phase CEUS can be seen as a positive sign indicating long term tumor response to treatment. Therefore CEUS may be considered as an imaging modality for monitoring early treatment after focal alpha and beta targeted therapy.

Impact of peptide receptor radionuclide therapy on the 68Ga-DOTATOC-PET/CT uptake in normal tissue.

AIM: Positron-emission tomography/computed tomography (PET/CT) with [68Ga]DOTA0-Phe1-Tyr3-octreotide (68Ga-DOTA-TOC) became a standard for somatostatin receptor imaging. We investigated the potential changes of normal tissue uptake in patients with neuroendocrine tumor undergoing peptide receptor radionuclide therapy (PRRT). METHODS: Sixteen patients underwent [68Ga]-DOTA-TOC-PET/CT prior and after 4-6 cycles of PRRT (mean administered activity: 13.8 GBq 90Y+ 9.6 177Lu). The maximum standardized uptake values (SUVmax) of pituitary, thyroid, spleen, liver parenchyma, pancreas, kidneys and adrenals were determined, respectively. RESULTS: SUVmax values prior and after PRRT were in pituitary (5, 56±2,91/ 4,47±2,53), thyroid (2.05±1.11/ 2.49±2.47), spleen (24.95±14.20/20.06±8.53), liver (7.13±3.96/6.62±2.63), pancreas (6.96±1.99/6.83±2.00), kidneys (13.0±3.85/11.31±3.31) and adrenals (9.65± 4.20/7.10±2.86). A comparison of pre- and post treatment values revealed no significant differences (P>0.05) in any of these organs. CONCLUSION: The uptake of [68Ga] DOTA-TOC in normal tissue is not significantly affected by PRRT. This is relevant with regards to therapeutic monitoring were tumor-to-non-tumor ratio seems to be the most robust biomarker.

Dynamic contrast enhanced-MRI for the detection of pathological complete response to neoadjuvant chemotherapy for locally advanced rectal cancer.

OBJECTIVE: To determine the ability of dynamic contrast enhanced (DCE-MRI) to predict pathological complete response (pCR) after preoperative chemotherapy for rectal cancer. METHODS: In a prospective clinical trial, 23/34 enrolled patients underwent pre- and post-treatment DCE-MRI performed at 1.5T. Gadolinium 0.1 mmol/kg was injected at a rate of 2 mL/s. Using a two-compartmental model of vascular space and extravascular extracellular space, K(trans), k(ep), v(e), AUC90, and AUC180 were calculated. Surgical specimens were the gold standard. Baseline, post-treatment and changes in these quantities were compared with clinico-pathological outcomes. For quantitative variable comparison, Spearman's Rank correlation was used. For categorical variable comparison, the Kruskal-Wallis test was used. P ≤ 0.05 was considered significant. RESULTS: Percentage of histological tumour response ranged from 10 to 100%. Six patients showed pCR. Post chemotherapy K(trans) (mean 0.5 min(-1) vs. 0.2 min(-1), P = 0.04) differed significantly between non-pCR and pCR outcomes, respectively and also correlated with percent tumour response and pathological size. Post-treatment residual abnormal soft tissue noted in some cases of pCR prevented an MR impression of complete response based on morphology alone. CONCLUSION: After neoadjuvant chemotherapy in rectal cancer, MR perfusional characteristics have been identified that can aid in the distinction between incomplete response and pCR. KEY POINTS: Dynamic contrast enhanced (DCE) MRI provides perfusion characteristics of tumours. These objective quantitative measures may be more helpful than subjective imaging alone Some parameters differed markedly between completely responding and incompletely responding rectal cancers. Thus DCE-MRI can potentially offer treatment-altering imaging biomarkers.

Prognostic significance of adrenal gland morphology at CT in patients with three common malignancies.

OBJECTIVES: To determine whether minor alterations in adrenal gland morphology at baseline CT in three common cancers indicate early metastasis. METHODS: 689 patients (237 with lung cancer, 228 with breast cancer, 224 with melanoma) underwent baseline and follow-up CTs that included the adrenals. Two readers independently scored each adrenal at baseline CT as normal, smoothly enlarged, nodular or mass-containing. Adrenals containing a mass >10 mm were excluded. The appearance of each adrenal on the latest available CT was assessed for change since baseline. Cox models were used to assess the association between adrenal morphology at initial CT and subsequent development of adrenal metastasis (defined as new mass >10 mm, corroborated by follow-up imaging). κ statistics were calculated to assess inter-reader agreement. RESULTS: Initial and follow-up CT evaluations were recorded for 1317 adrenals (median follow-up, 18.6 months). At initial CT, Readers 1 and 2 interpreted 1242 and 1230 adrenals as normal, 40 and 57 as smoothly enlarged, 29 and 25 as nodular, and 6 and 5 as containing masses ≤ 10 mm, respectively. κ-values were 0.52 (moderate) at initial CT and 0.70 (substantial) at follow-up. The hazard ratio for developing a metastasis at follow-up CT given an abnormal adrenal assessment at baseline was 0.7 [95% confidence interval (CI) 0.2-2.1; p = 0.47] for Reader 1, and 2.0 (95% CI 0.8-4.7; p = 0.12) for Reader 2. CONCLUSION: Minor morphological abnormalities of adrenals at initial CT did not represent early adrenal metastasis in most patients in this population.

Data sets for the qualification of volumetric CT as a quantitative imaging biomarker in lung cancer.

The drug development industry is faced with increasing costs and decreasing success rates. New ways to understand biology as well as the increasing interest in personalized treatments for smaller patient segments requires new capabilities for the rapid assessment of treatment responses. Deployment of qualified imaging biomarkers lags apparent technology capabilities. The lack of consensus methods and qualification evidence needed for large-scale multi-center trials, as well as the standardization that allows them, are widely acknowledged to be the limiting factors. The current fragmentation in imaging vendor offerings, coupled with the independent activities of individual biopharmaceutical companies and their contract research organizations (CROs), may stand in the way of the greater opportunity were these efforts to be drawn together. A preliminary report, "Volumetric CT: a potential biomarker of response," of the Quantitative Imaging Biomarkers Alliance (QIBA) activity was presented at the Medical Imaging Continuum: Path Forward for Advancing the Uses of Medical Imaging in the Development of New Biopharmaceutical Products meeting of the Extended Pharmaceutical Research and Manufacturers of America (PhRMA) Imaging Group sponsored by the Drug Information Agency (DIA) in October 2008. The clinical context in Lung Cancer and a methodology for approaching the qualification of volumetric CT as a biomarker has since been reported [Acad. Radiol. 17, 100-106, 107-115 (2010)]. This report reviews the effort to collect and utilize publicly available data sets to provide a transparent environment in which to pursue the qualification activities in such a way as to allow independent peer review and verification of results. This article focuses specifically on our role as stewards of image sets for developing new tools.

Change in lung tumor volume as a biomarker of treatment response: a critical review of the evidence.

SPECIFIC AIM: To review the evidence indicating that volumetric image analysis of computed tomography scans meets specifications for qualification as a biomarker in clinical trials or the management of individual patients with lung cancer. METHODS: Claims of value were broken down into questions about technical feasibility, accuracy, the precision of measurement, sensitivity, the correlations with health outcomes, and the risks of producing misleading information. For each claim, the pertinent literature was reviewed. RESULTS: Technical feasibility has now been shown, but only in limited contexts. Accuracy has been demonstrated, but only for tumors with favorable anatomical features. Measurement error still makes the assessment of change in small nodules precarious in diagnostic settings unless rigorous image acquisition and analysis procedures are followed. Precision is sufficient in some larger masses to make volumetrics a sensitive biomarker. In a few trials, correlations with clinical outcomes have been higher for volumetric-based measures than for unidimensional or bidimensional diameters. Value in clinical practice settings and clinical trials has been suggested, but not proven. CONCLUSION: The weight of the evidence indicates there are circumstances in which volumetric image analysis adds value to clinical trial science and the practice of medicine.

Regional chemotherapy for unresectable primary liver cancer: results of a phase II clinical trial and assessment of DCE-MRI as a biomarker of survival.

BACKGROUND: This study reports the results of hepatic arterial infusion (HAI) with floxuridine (FUDR) and dexamethasone (dex) in patients with unresectable intrahepatic cholangiocarcinoma (ICC) or hepatocellular carcinoma (HCC) and investigates dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) assessment of tumor vascularity as a biomarker of outcome. PATIENTS AND METHODS: Thirty-four unresectable patients (26 ICC and eight HCC) were treated with HAI FUDR/dex. Radiologic dynamic and pharmacokinetic parameters related to tumor perfusion were analyzed and correlated with response and survival. RESULTS: Partial responses were seen in 16 patients (47.1%); time to progression and response duration were 7.4 and 11.9 months, respectively. Median follow-up and median survival were 35 and 29.5 months, respectively; 2-year survival was 67%. DCE-MRI data showed that patients with pretreatment integrated area under the concentration curve of gadolinium contrast over 180 s (AUC 180) >34.2 mM.s had a longer median survival than those with AUC 180 <34 mM.s (35.1 versus 19.1 months, P = 0.002). Decreased volume transfer exchange between the vascular space and extracellular extravascular space (-DeltaK(trans)) and the corresponding rate constant (-Deltak(ep)) on the first post-treatment scan both predicted survival. CONCLUSIONS: In patients with unresectable primary liver cancer, HAI therapy can be effective and safe. Pretreatment and early post-treatment changes in tumor perfusion characteristics may predict treatment outcome.

Evaluation of lymph nodes with RECIST 1.1.

Lymph nodes are common sites of metastatic disease in many solid tumours. Unlike most metastases, lymph nodes are normal anatomic structures and as such, normal lymph nodes will have a measurable size. Additionally, the imaging literature recommends that lymph nodes be measured in the short axis, since the short axis measurement is a more reproducible measurement and predictive of malignancy. Therefore, the RECIST committee recommends that lymph nodes be measured in their short axis and proposes measurement values and rules for categorising lymph nodes as normal or pathologic; either target or non-target lesions. Data for the RECIST warehouse are presented to demonstrate the potential change in response assessment following these rules. These standardised lymph node guidelines are designed to be easy to implement, focus target lesion measurements on lesions that are likely to be metastatic and prevent false progressions due to minimal change in size.

New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).

BACKGROUND: Assessment of the change in tumour burden is an important feature of the clinical evaluation of cancer therapeutics: both tumour shrinkage (objective response) and disease progression are useful endpoints in clinical trials. Since RECIST was published in 2000, many investigators, cooperative groups, industry and government authorities have adopted these criteria in the assessment of treatment outcomes. However, a number of questions and issues have arisen which have led to the development of a revised RECIST guideline (version 1.1). Evidence for changes, summarised in separate papers in this special issue, has come from assessment of a large data warehouse (>6500 patients), simulation studies and literature reviews. HIGHLIGHTS OF REVISED RECIST 1.1: Major changes include: Number of lesions to be assessed: based on evidence from numerous trial databases merged into a data warehouse for analysis purposes, the number of lesions required to assess tumour burden for response determination has been reduced from a maximum of 10 to a maximum of five total (and from five to two per organ, maximum). Assessment of pathological lymph nodes is now incorporated: nodes with a short axis of 15 mm are considered measurable and assessable as target lesions. The short axis measurement should be included in the sum of lesions in calculation of tumour response. Nodes that shrink to <10mm short axis are considered normal. Confirmation of response is required for trials with response primary endpoint but is no longer required in randomised studies since the control arm serves as appropriate means of interpretation of data. Disease progression is clarified in several aspects: in addition to the previous definition of progression in target disease of 20% increase in sum, a 5mm absolute increase is now required as well to guard against over calling PD when the total sum is very small. Furthermore, there is guidance offered on what constitutes 'unequivocal progression' of non-measurable/non-target disease, a source of confusion in the original RECIST guideline. Finally, a section on detection of new lesions, including the interpretation of FDG-PET scan assessment is included. Imaging guidance: the revised RECIST includes a new imaging appendix with updated recommendations on the optimal anatomical assessment of lesions. FUTURE WORK: A key question considered by the RECIST Working Group in developing RECIST 1.1 was whether it was appropriate to move from anatomic unidimensional assessment of tumour burden to either volumetric anatomical assessment or to functional assessment with PET or MRI. It was concluded that, at present, there is not sufficient standardisation or evidence to abandon anatomical assessment of tumour burden. The only exception to this is in the use of FDG-PET imaging as an adjunct to determination of progression. As is detailed in the final paper in this special issue, the use of these promising newer approaches requires appropriate clinical validation studies.

Recommendations for the assessment of progression in randomised cancer treatment trials.

Progression-free survival (PFS) is an increasingly important end-point in cancer drug development. However, several concerns exist regarding the use of PFS as a basis to compare treatments. Unlike survival, the exact time of progression is unknown, so progression times might be over-estimated (or under-estimated) and, consequently, bias may be introduced when comparing treatments. In addition, the assessment of progression is subject to measurement variability which may introduce error or bias. Ideally trials with PFS as the primary end-point should be randomised and, when feasible, double-blinded. All patients eligible for study should be evaluable for the primary end-point and thus, in general, have measurable disease at baseline. Appropriate definitions should be provided in the protocol and data collected on the case-report forms, if patients with only non-measurable disease are eligible and/or clinical, or symptomatic progression are to be considered progression events for analysis. Protocol defined assessments of disease burden should be obtained at intervals that are symmetrical between arms. Independent review of imaging may be of value in randomised phase II trials and phase III trials as an auditing tool to detect possible bias.

Imaging response assessment in oncology.

The role of imaging in the clinical setting as well as in the drug development process is expanding rapidly. Imaging technology now exists that is capable of detecting tumor response within hours. In parallel with this advance, a new array of more targeted and specific therapies are being developed. This paradigm shift in turn demands a more sophisticated way of quantifying response. There is a need to update and modify the current response evaluation criteria in solid tumors (RECIST), which rely solely on anatomic size measurement of tumors. In addition, response assessment guidelines will need to be increasingly disease-specific. Response assessment by imaging is now intimately involved with all stages of the drug development process, from exploratory drug discovery through clinical trials, as well as in clinical use. Imaging biomarkers and surrogate endpoints have the potential to speed drug approval significantly. The major funding institutions and the pharmaceutical industry are working more and more with researchers to help maintain progress in this multidisciplinary area involving oncologists, radiologists, molecular imaging specialists, medical physicists, and computer scientists.

Measuring tumor response and shape change on CT: esophageal cancer as a paradigm.

BACKGROUND: Accurate response assessment is essential for evaluating new cancer treatments. We evaluated the impact of Response Evaluation Criteria in Solid Tumors (RECIST), World Health Organization (WHO) criteria and tumor shape on response assessment in patients with metastatic esophageal cancer. PATIENTS AND METHODS: In 19 patients with metastatic esophageal cancer in a phase II trial of bryostatin-1 and paclitaxel, response was retrospectively assessed for 89 lesions with RECIST and WHO criteria on baseline and serial follow-up CT scans. The eccentricity factor (EF) was introduced for measuring the degree to which tumor shape diverges from a perfect sphere [EF = radical1-(LPD/MD)(2), where LPD is the largest perpendicular diameter and MD is the maximal diameter]. RESULTS: The disagreement rate in best overall response categorization between RECIST (unidimensional) and WHO (bidimensional) criteria was 26.3%. Change in eccentricity was significantly greater (P < 0.01) for patients with disagreement (mean 0.31, range 0-0.91). When the short axis was used for unidimensional lymph node measurement, disagreement between WHO and RECIST lessened. CONCLUSIONS: Response assessment by WHO and RECIST differs substantially. Greater change in eccentricity is associated with greater discordance between WHO and RECIST. The discordance between WHO and RECIST may impact on how effective a therapy is judged to be.