Predictors of severe and febrile neutropenia during primary chemotherapy for ovarian cancer

Objective

To identify factors that increase the risk of neutropenic events in women with advanced ovarian carcinoma receiving initial chemotherapy.

Methods

Multi-center retrospective study of women with FIGO stage III-IV epithelial ovarian cancer treated postoperatively with multi-agent intravenous chemotherapy from 1995 to 2008. Outcomes were severe (SN; absolute neutrophil count [ANC] < 500/mm³) and febrile neutropenia (FN; ANC < 1000/mm³ and temperature > 38.1 °C). Cumulative risk of neutropenic events was estimated by Kaplan Meier method. Multivariate analysis was by Cox proportional hazard regression.

Results

Three hundred twenty-six patients met inclusion criteria. There were 251 SN events among 140 (43%) patients and 24 FN events among 22 (7%) patients. Univariate predictors of SN were body surface area < 2.0 m² (p = 0.03), body mass index (BMI) < 30 kg/m² (p < 0.01), Caucasian race (p < 0.01), treatment on research protocols (p < 0.01), non-carboplatin-containing regimens (p < 0.01), and planned relative dose intensity (RDI) > 85% of standard (p = 0.02). Women over age 60 were more likely to develop FN (p = 0.05). Multivariate predictors of SN were treatment on research protocols (hazard ratio [HR] 1.93; p < 0.01), Caucasian race (HR 2.13; p = 0.01), and planned RDI > 85% (HR 1.69; p = 0.05); predictors of FN were age > 60 (HR 2.84; p = 0.05) and non-carboplatin containing regimens (HR 4.06; p < 0.01).

Conclusion

While SN is fairly common, FN occurs infrequently in women with EOC undergoing taxane and platin-based chemotherapy and primary prophylactic growth factor support is not indicated. However, women older than 60 years of age receiving non-carboplatin containing regimens are at higher risk for FN and warrant closer surveillance.     

Effects of pacing tachycardia and balloon valvuloplasty on pulmonary artery impedance and hydraulic power in mitral stenosis.

BACKGROUND. Mitral stenosis is characterized by progressive pulmonary hypertension and eventual right ventricular failure. However, the correlation between right ventricular failure and the level of pulmonary hypertension is poor, suggesting that factors other than those recognized from nonpulsatile hemodynamic parameters may contribute to impaired right ventricular performance in this condition. METHODS AND RESULTS. We studied 16 patients with severe mitral stenosis (mean valve area, 1.0 +/- 0.2 cm2) at supine rest and during pacing tachycardia using high-fidelity catheter recordings of pulmonary artery (PA) pressure and flow velocity. Pulmonary impedance spectra, wave reflection properties, and hydraulic power data were derived from Fourier analysis of signal-averaged data. Pacing tachycardia (baseline heart rate, 81 +/- 11 beats per minute; pacing, 132 +/- 11 beats per minute) significantly raised pulmonary wedge and mean PA pressures. There was no change in pulmonary vascular resistance (209 +/- 144 to 232 +/- 164 dyne-sec/cm5) or PA characteristic impedance (62 +/- 25 to 55 +/- 28 dyne-sec/cm5). However, first harmonic impedance (Z1) significantly decreased (134 +/- 71 to 100 +/- 68 dyne-sec/cm5; p < 0.001). Accordingly, oscillatory and total dissipated hydraulic power per unit forward flow (WT/CO) fell during tachycardia (2.6 +/- 1.6 to 2.3 +/- 1.4 mW/ml.sec-1; p = 0.06) despite acute pulmonary hypertension. Reflected pressure waves returned earlier to the proximal PA, suggesting increased vessel stiffness. Immediately after percutaneous balloon mitral valvuloplasty (PBV) in eight of the patients, baseline and pacing data were again recorded. Compared with the pre-PBV baseline state, post-PBV resting data demonstrated no change in resistance or characteristic impedance, but there was a significant fall in Z1 (166 +/- 75 to 103 +/- 45 dyne-sec/cm5; p < 0.05) and in the magnitude of pulmonary wave reflections. WT/CO tended to decrease after PBV, and pacing after PBV produced a further decrease in WT/CO, again in association with lower Z1. CONCLUSIONS. These data demonstrate that 1) increased pulmonary characteristic impedance, although a feature of mitral stenosis, is not exacerbated by the acute effects of increased distending pressure; 2) pacing tachycardia in mitral stenosis causes little change in the pulmonary impedance spectrum except at low frequencies, where decreased impedance lowers power requirements per unit flow; and 3) relief of mitral stenosis produces immediate improvement in low-frequency impedance and in hydraulic power requirements. These findings suggest that although characteristic impedance may be a measure of the long-term effects of pulmonary hypertension on the pulmonary circulation, acute increases and decreases in PA pressure produce effects on right ventricular load that are best described in terms of the low-frequency properties of the PA system. Improvement in low-frequency impedance diminishes hydraulic power requirements and thus reflects improved ventricular-vascular coupling, irrespective of distending PA pressure. Efforts to treat or prevent right heart failure in the presence of pulmonary hypertension should take account of the potential benefit of changes in low-frequency impedance characteristics of the pulmonary vascular bed.     

The development of a protocol to assess reproductive effects of toxicants in the rat

The determination that a chemical poses a reproductive risk to man typically relies upon fertility studies using rodents. However, fertility in rodents is often difficult to disrupt and more sensitive indicators of reproductive function should be included in the risk assessment process. The present discussion compares the sensitivity of fertility to other endpoints following exposure to known reproductive toxicants. In our studies rats were dosed from weaning through puberty, gestation, and lactation. The reproductive function of the male, the female, and the offspring was assessed. The effects of methoxychlor, carbendazim (MBC), dibutyl phthalate (DBP), and lindane are discussed and compared to fertility. For each compound a ratio (SR = sensitivity ratio) of the lowest effect level (LEL) for infertility or reduced fecundity to the LEL for the most sensitive physiologic endpoint was calculated. The SR should be large when a compound produces effects over a wide range of doses, but should equal unity when the dose-response curve is very steep. For methoxychlor, which blocked implantation, pubertal landmarks and estrous cyclicity provided rapid and sensitive indicators of the subsequent reproductive failure. The SR = 8 (100/12) for methoxychlor using data from females. In contrast, DBP and MBC directed altered testicular function, and for these compounds, sperm and testicular measures provided sensitive indicators of toxicity. The SR for MBC was 2 (100/50), while DBP had a SR of 1 (500/500). In the lindane study, fertility was not reduced but most of the pups (F1) died shortly after birth. The SR for lindane is equal to 0.5 (10/20). At 20 mg/kg the treated females were larger and their estrous cycles were erratic. The data from these studies indicate that no single endpoint will consistently be the most sensitive indicator of reproductive toxicity. Studies must include a number of well validated endpoints that provide a comprehensive assessment of the entire reproductive system of the male and female.     

Effect of cadmium and other metal cations on in vitro Leydig cell testosterone production.

In vivo assessment of toxicant action on Leydig cell function is subject to homeostatic mechanisms which make it difficult to determine whether any changes seen in serum testosterone (T) concentration are due to extragonadal endocrine alterations or to a direct effect on the Leydig cell. For example, metal cations administered in vivo have been shown to depress serum T concentration and alter serum concentrations of pituitary hormones in laboratory animals. The studies reported here use a testicular cell culture technique to evaluate Leydig cell testosterone biosynthesis in the presence of several metal cations. To determine the site of toxic action, the Leydig cells were stimulated to produce testosterone by using human chorionic gonadotrophin (hCG), dibutyl cyclic adenosine monophosphate (db-cAMP), or several substrates required for the biosynthesis of testosterone. hCG was chosen because resultant T production requires an intact membrane receptor and db-cAMP was used to test for post LH receptor defects caused by the metals. The other substrates were chosen to isolate the effect of metals on enzymatic pathways. Collagenase dispersed testicular cells (15% Leydig cells) were incubated with metal cations (1 to 5000 microM) for 3 hr in the absence and presence of maximally stimulating concentrations of hCG, db-cAMP, 20 alpha-hydroxycholesterol (HCHOL), or pregnenolone (PREG), and T concentration was determined by radioimmunoassay. In one separate experiment we also tested the effect of the substrates progesterone, 17 alpha-hydroxy-progesterone, and androstenedione on Cd2(+)-treated Leydig cells. The results show no change in Leydig cell viability with any metal cation treatment during the 3-hr incubation. Ca2+, Cr3+, Fe3+, Mg2+, Na+, or Pb2+ had no effect on stimulated testosterone. Dose-response depression in both hCG- and db-cAMP-stimulated T production were seen with Cd2+, Co2+, Cu2+, Hg2+, Ni2+, and Zn2+ treatment. Surprisingly, Cd2+, Co2+, Ni2+, and Zn2+, which caused a depression in hCG- and db-cAMP-stimulated T production, caused significant increases in HCHOL- and PREG-stimulated T production over untreated and similarly stimulated cultures. This indicates that these cations may act at multiple sites within the Leydig cell.