Biomarker driven population enrichment for adaptive oncology trials with time to event endpoints.

The development of molecularly targeted therapies for certain types of cancers has led to the consideration of population enrichment designs that explicitly factor in the possibility that the experimental compound might differentially benefit different biomarker subgroups. In such designs, enrollment would initially be open to a broad patient population with the option to restrict future enrollment, following an interim analysis, to only those biomarker subgroups that appeared to be benefiting from the experimental therapy. While this strategy could greatly improve the chances of success for the trial, it poses several statistical and logistical design challenges. Because late-stage oncology trials are typically event driven, one faces a complex trade-off between power, sample size, number of events, and study duration. This trade-off is further compounded by the importance of maintaining statistical independence of the data before and after the interim analysis and of optimizing the timing of the interim analysis. This paper presents statistical methodology that ensures strong control of type 1 error for such population enrichment designs, based on generalizations of the conditional error rate approach. The special difficulties encountered with time-to-event endpoints are addressed by our methods. The crucial role of simulation for guiding the choice of design parameters is emphasized. Although motivated by oncology, the methods are applicable as well to population enrichment designs in other therapeutic areas.

Intravitreal vascular endothelial growth factor.

PURPOSE: To evaluate whether a specific pre-analytical stabilization regimen is needed for naïve vitreous taps to detect true values of intrinsic VEGF levels. METHODS: Fourteen consecutive patients with different vitreomacular pathologies without blood-retina-barrier breakdown were scheduled for standard 23-gauge three-port pars plana vitrectomy, and naïve vitreous taps were sampled at the beginning of each procedure. The extracted vitreous specimen was split; one half was immediately stored in a -20 °C freezer (unstabilized samples) and the other half was instantly stabilized with albumin (2.5 % final conc.), followed by arginine stabilization (1.25 M final conc.) and consecutively stored in a -20 °C freezer (stabilized samples). RESULTS: Intravitreal VEGF was detected in all 14 analyzed samples (100 %). VEGF levels were shown to be 46.5 pg/ml ± 62.3 pg/ml (MV ± SD; range: 5.99-232.3 pg/ml) in unstabilized, and 120.4 pg/ml ± 94.4 pg/ml (range: 42.9 pg/ml-289.6 pg/ml) in stabilized vitreous samples. Intravitreal VEGF levels in stabilized vitreous samples were on average 2.6-fold, and thus significantly higher than in unstabilized taps of same eyes (p = 0.001, Wilcoxon test). VEGF levels in stabilized vitreous samples can be up to 8.5 times higher than in corresponding unstabilized vitreous taps of same eyes (bootstrap analysis). Intravitreal VEGF levels in unstabilized samples correlate with those in stabilized vitreous taps (r = 0.594; p = 0.025; Pearson). CONCLUSIONS: An adequate pre-analytic stabilization regimen is needed to evaluate the most accurate intravitreal VEGF levels. This in turn will result in a better understanding of the physiological as well as pathological role of VEGF within the eye. Furthermore, knowing the true value of intravitreal VEGF levels will help to calculate the dosage of intravitrealy applied anti-VEGF drugs.

Intravitreal functional plasminogen in eyes with branch retinal vein occlusion.

PURPOSE: To evaluate whether intravitreal functional plasminogen is elevated in eyes with branch retinal vein occlusion (BRVO) and to discover whether intravitreal plasminogen activities are correlated with the extent of blood-retina barrier (BRB) breakdown. METHODS: Our study is a prospective case series of 20 consecutive patients with BRVO and 10 consecutive patients serving as controls. Vitreous taps were extracted from the central vitreous body and plasminogen was functionally determined in an innovative, ultrasensitive p-nitroanilide reaction after activation with streptokinase (100% of normal, %N = functional plasminogen in pooled normal citrated plasma). Intravitreal VEGF levels were assayed to estimate BRB breakdown. RESULTS: Intravitreal functional plasminogen was detected in all analyzed samples (n = 30) and mean (±SD) plasminogen activities were found to be 0.97 ± 1.06%N (range: 0.03-3.9%N). Patients suffering from BRVO exhibited significantly higher intravitreal plasminogen (1.35 ± 1.11%N) in comparison with controls (0.20 ± 0.21%N, p < 0.001). Intravitreal VEGF concentrations in the BRVO group (576 ± 547 pg/ml) were significantly higher than these in controls (111 ± 120 pg/ml, p = 0.003). There was a significant correlation between intravitreal functional plasminogen and intravitreal VEGF levels (r = 0.519, p = 0.003). CONCLUSIONS: Intravitreal functional plasminogen is significantly elevated in eyes suffering from BRVO and correlates with the extent of BRB breakdown. The induction of posterior vitreous detachment by using intravitreally administered recombinant tissue plasminogen activator (enzymatic vitreolysis) should be explored in further investigations.

α4 integrin levels on mobilized peripheral blood stem cells predict rapidity of engraftment in patients receiving autologous stem cell transplantation.

Rapidness of leukocyte engraftment in patients receiving peripheral blood stem cell transplantation is clinically important because the risk of fatal opportunistic infections increases with time to engraftment. Adhesion receptor molecules on hematopoietic stem cells (HSCs) have been shown to modulate homing and engraftment of HSCs. Therefore, we correlated expression levels of α4 (CD49d) and α6 (CD49f) integrins in the CD34(+) HSC compartment with time to engraftment. Leukapheresis products from 103 patients were retrospectively analyzed for CD34, CD38, CD3, CD49f, and CD49d surface molecules by multiparameter flow cytometry. High expression levels of α4 integrin, but not α6 integrin on CD34(+) cells, were associated with regular engraftment of leukocytes (days 8-19), whereas low surface expression correlated with delayed recovery (> 19 days; P < .0005). We show that α4 integrin expression levels on HSCs in leukapheresis products predict the engraftment capacity of mobilized peripheral blood stem cells in peripheral blood stem cell transplantation patients.

Intracameral concentrations of the fibrinolytic system components in patients with age-related macular degeneration.

PURPOSE: To detect the intracameral concentrations and activities of plasminogen and other components of the fibrinolytic system, and to investigate whether those concentrations and activities are higher in patients with age-related macular degeneration (AMD) in comparison to healthy controls. METHODS: Prospective case series of 93 patients scheduled for standard phacoemulsification. RESULTS: Mean plasminogen activity in patients with non-exsudative AMD (n = 24) revealed to be 0.06%N, in patients with exudative AMD (n = 7) 0.03%N and in healthy controls (n = 43) 0.02%N (p = 0.38, ANOVA). Plasminogen activator inhibitor I (PAI-1) was detected in neither group. Alpha2-antiplasmin activity was 1.61 U/ml in the non-exudative AMD group (n = 27), 0 U/ml in the exudative AMD group (n = 7) and 0.54 U/ml in the control group (n = 48) (p = 0.1, ANOVA). Concentrations of plasmin-a2-antiplasmin complex (PAP) were detected at levels of 17.91 ng/ml in the non-exudative AMD group (n = 11), of 16.6 ng/ml in the exudative AMD group (n = 5), and of 17.43 ng/ml in the control group (n = 14) (p = 0.92, ANOVA). CONCLUSIONS: Plasminogen is present with a very low activity in aqueous humor. There are no significant differences in aqueous humor concentrations or activities of plasminogen and other components of the fibrinolytic system between patients with non-exudative AMD, exudative AMD, and healthy controls. Further studies should investigate vitreous samples instead of aqueous humor samples. A careful and accurate workup of obtained intraocular fluids is needed to detect the low concentrations and activities of the parameters analyzed.

Incidence of bisphosphonate-related osteonecrosis of the jaws (BRONJ) in patients taking bisphosphonates for osteoporosis treatment - a grossly underestimated risk?

PURPOSE: The aim of this study was to detect the incidence of bisphosphonate-related osteonecrosis of the jaws (BRONJ) in association with osteoporosis in 2009 in the rural district of Marburg-Biedenkopf, Germany. In addition, the therapeutic regimen of dentists in this area was compared to the treatment guidelines of established international associations. METHODOLOGY: A postal survey including 129 dental offices within the named investigation area was conducted. Additionally, the dentists were asked to contribute to this study during a course of retraining. Ultimately, 107 colleagues participated in this study. RESULTS: A total of 37 BRONJ cases were revealed, 37.4 % of those were linked to malignant diseases, 62.6 % to osteoporosis. Noticeably, 30.1 % of the BRONJ cases linked to osteoporosis were connected with intravenous application. In total, 62.6 % of all cases were associated with intravenous and 37.4 % with oral application. Considering the estimated number of 1.014 patients using bisphosphonates for osteoporosis treatment in Marburg-Biedenkopf in 2009, the specific incidence of BRONJ could be narrowed down to about 2.27 %. In proportion to the increasing risk potential of the three patient groups, participants conducted fewer surgical interventions themselves but tended to refer patients to colleagues. CONCLUSION: This study reveals the incidence of BRONJ in association with osteoporosis as being grossly underrated so far, especially in connection with intravenous bisphosphonate treatment. The therapeutic regimen of the dentists who participated correlated with the established guidelines. CLINICAL RELEVANCE: The interface between dentistry and medicine may profit from our study's results which will help to improve interdisciplinary communication. With regard to this, we wish to contribute to an informative discussion since a main focus was to ensure colleagues in their decision making, especially in case of complications after elective dentoalveolar surgery-like implantations. Besides medical there are also economic and political aspects, i.e. the financial responsibility in case of unexpected or inevitable complications, possibly leading to BRONJ, that may become part of future discussions based on this study.

Sampling aqueous humor: anterior segment anatomy, anesthetic and surgical technique, and rates of yield.

PURPOSE: To evaluate anterior segment anatomy and anesthetic and surgical techniques with respect to the amount of aqueous humor (AH) that can be sampled out of the anterior chamber (AC) at the beginning of standard cataract removal procedures (phacoemulsification). METHODS: In a prospective survey, volumes of sampled AH from 123 eyes (110 patients) were analyzed in regard to AC anatomy (anterior chamber depth, ACD) and different anesthetic techniques (local and general anesthesia). RESULTS: 107 eyes (87%) were included into our analysis, and 16 eyes (13%) had to be excluded due to failure of AH collection. We found a significant positive association between ACD and obtained AH volume (p=0.007). In general anesthesia, a strong trend to acquire more AH in comparison to local anesthesia was apparent, but statistical significance failed (p=0.167). Different anesthetic techniques seem to have no significant influence on ACD (p=0.169). No training curve for the individual surgeon was obtained. No complications were observed. CONCLUSION: When AH sampling is performed in eyes with a deep AC and when the procedure is performed under general anesthesia, more AH can be aspirated.

The impact of posterior vitreous adhesion on ischaemia in eyes with retinal vein occlusion.

PURPOSE: To evaluate whether posterior vitreous adhesion (PVA) contributes to retinal ischaemia in eyes suffering from central (CRVO) or branch retinal vein occlusion (BRVO). METHODS: Retrospective patient chart analysis of eyes with CRVO/BRVO receiving pars-plana vitrectomy (ppV). Prior to surgery fluorescence angiography was conducted to classify RVO as ischaemic or not. RESULTS: Sixty eyes were included, thereof 36 (60%)/24 (40%) with CRVO/BRVO. In the CRVO group, 17 (47%)/19 (53%) eyes were classified as ischaemic/non-ischaemic. Respective results for BRVO-affected eyes were 16 (67%)/8 (33%). PVA/posterior vitreous detachment (PVD) was found in 33 (92%)/3 (8%) eyes with CRVO and in 23 (96%)/1 (4%) patients suffering from BRVO. Value differences of PVA/PVD between ischaemic- and non-ischaemic-typed RVO failed statistical significance for both, CRVO (p = 0.095) and BRVO (p = 1.0). CONCLUSIONS: Posterior vitreous adhesion had no impact on retinal ischaemia in this investigation. As an attached posterior vitreous cortex acts as a scaffold and thus significantly increases neovascularization (NV) development in ischaemic-typed RVO, a prospective study evaluating the effect of enzymatic vitreolysis is indicated.

Changes in corneal endothelium cell characteristics after cataract surgery with and without use of viscoelastic substances during intraocular lens implantation.

PURPOSE: To evaluate whether the use of balanced salt solution (BSS) or an ophthalmic viscoelastic device (OVD) during hydrophilic acrylic intraocular lens (IOL) implantation variously impacts corneal endothelial cell characteristics in eyes undergoing uneventful phacoemulsifications. METHODS: Prospective nonrandomized observational clinical trial. Patients were assigned either to the BSS plus(®) or to the OVD Z-Celcoat™ group depending on the substance used during IOL implantation. Corneal endothelium cell characteristics were obtained before, 1 week, and 6 weeks after surgery. Intraoperative parameters (eg, surgery time, phacoemulsification energy) were recorded. RESULTS: Ninety-seven eyes were assigned to the BSS plus and 86 eyes to the Z-Celcoat group. Preoperative corneal endothelium cell density (ECD) and endothelium cell size were 2,506±310 cells/mm(2)/2,433±261 cells/mm(2) and 406±47 µm(2)/416±50 µm(2) (P=0.107/P=0.09). After 1 and 6 weeks, ECD decreased and endothelium cell size increased significantly in both groups (each P<0.001) without significant differences between both groups (each P>0.05). Irrigation-aspiration suction time (30.3±16.6 versus 36.3±14.5 seconds) and overall surgical time (7.2±1.2 versus 8.0±1.4 minutes) were significantly longer in the OVD Z-Celcoat group (each P<0.001). No complications or serious side effects occurred. CONCLUSION: Implantation of a hydrophilic acrylic IOL under BSS infusion seems to be a useful and faster alternative in experienced hands without generating higher ECD loss rates.

NT-proANP and NT-proBNP as prognostic markers in patients with acute decompensated heart failure of different etiologies.

BACKGROUND AND PURPOSE: Peak NT-proANP and NT-proBNP plasma levels after hospital admission may be of additional prognostic value in patients with acute decompensation of heart failure. The time-course of natriuretic plasma levels after hospital admission, and a possible influence of the underlying etiology on the time-course have not been sufficiently investigated. METHODS AND RESULTS: Natriuretic peptide plasma levels of 85 patients with decompensated heart failure from ischemic and non-ischemic origins were measured at baseline and at 12h after hospital admission. NT-proBNP plasma levels on admission were lower compared to 12-hour-plasma levels, whereas NT-proANP plasma levels on admission were higher compared to 12-hour-plasma levels. Twenty-six patients (31%) died within the first 30 days. In patients who died within the first 30 days after admission NT-proANP and NT-proBNP plasma levels on admission and 12h later were significantly higher compared to survivors. Irrespective of different etiologies NT-proANP on admission and NT-proBNP 12h after admission were highest and demonstrated superior impact with respect to the prediction of 30-day-mortality. CONCLUSIONS: NT-proANP and NT-proBNP are powerful markers of 30-day-mortality in patients with acute heart failure of ischemic and non-ischemic origins. With respect to the prediction of 30-day-mortality, NT-proBNP plasma levels at 12h after admission are comparable with NT-proANP plasma levels on admission. These data underline the fact that with regard to etiology-dependent hemodynamic changes and plasma half-time, the determination of peak plasma levels is of highest importance for the estimation of the impact of natriuretic peptides on the prognosis of patients with decompensated heart failure.

Evaluation of a posterior vitreous detachment: a comparison of biomicroscopy, B-scan ultrasonography and optical coherence tomography to surgical findings with chromodissection.

PURPOSE: To find the most reliable and efficient noninvasive technique to clinically detect a posterior vitreous detachment. METHODS: In a prospective study of 30 eyes in 30 patients with macular pucker or macular hole formation, the posterior vitreous cortex was examined 1 day prior to a scheduled vitrectomy. Three independent investigators classified the posterior vitreous cortex of each eye as 'attached' or 'detached' via slit-lamp biomicroscopy (BM), 10-MHz B-scan ultrasonography (I³ Innovative Imaging Inc.), and optical coherence tomography [OCT III Stratus(®) (Carl Zeiss Meditec Inc.) and RTVue-100 OCT (Optovue Corp.)]. These preoperative findings were then compared during a triamcinolone acetonide-assisted vitrectomy 1 day later. RESULTS: Triamcinolone acetonide-assisted vitrectomy showed in 60% a posterior vitreous detachment and in 40% an attached posterior vitreous cortex. Preoperatively conducted B-scan ultrasonography and BM revealed the highest, correct evaluation of the posterior vitreous status. The prediction of the OCT was confirmed intraoperatively in 12.5%. In all other cases, the evaluation by OCT was not possible or was inadequate. CONCLUSION: The prognostic most reliable but investigator-dependent methods to clinically detect whether the posterior vitreous cortex is detached are B-scan ultrasonography and BM. The objective technique of the high-resolution, two-dimensional time-domain OCT allows only in a few cases a clear differentiation of preretinal structures.

The vitreomacular interface in retinal vein occlusion.

PURPOSE: To evaluate the posterior vitreous adhesion status in patients with a history of central or branch retinal vein occlusion and to compare the results with the natural time-course of posterior vitreous detachment in healthy age-related controls. METHODS: A retrospective chart review in terms of the posterior vitreous adhesion status was performed in 132 patients (133 eyes) with a history of a central (CRVO) or branch (BRVO) retinal vein occlusion. All patients underwent vitrectomy. Based on the operation reports, the vitreous adhesion status was classified as attached, partially detached or completely detached. The results were compared to the natural time-course of posterior vitreous detachment development in healthy age-related controls. RESULTS: Eighty-one eyes met the inclusion and exclusion criteria. Fifty-two eyes (64%) had a history of CRVO and 29 eyes (36%) a history of BRVO, respectively. In the CRVO group, the posterior vitreous was attached in 47 eyes (90%) and completely detached in five eyes (10%). In the BRVO group, the posterior vitreous was attached in 27 eyes (93%), partially detached in 1 eye (3%) and completely detached in another eye (3%). A subdivision into age classes and a comparison with healthy age-related controls [data by Weber-Krause & Eckardt (1997) Ophthalmologe, 94, 619-623] showed in patients between 65 and 69 years of age an attached posterior vitreous cortex in 72% in healthy eyes, in 100% in CRVO (p = 0.109) and in 89% in BRVO (p = 0.440), in patients between 70 and 79 years of age an attached posterior vitreous cortex in 56% in healthy eyes, in 86% in CRVO (p = 0.010) and in 100% in BRVO (p = 0.038) and in patients between 80 and 89 years of age an attached posterior vitreous cortex in 43% in healthy eyes, in 100% in CRVO (p = 0.191) and in 67% in BRVO (p = 0.582) (Fisher's exact t-test). CONCLUSION: In patients with a history of CRVO or BRVO, the posterior vitreous cortex stays attached more frequently in all age groups in comparison with the healthy age-related controls.