Measures and models for causal inference in cross-sectional studies: arguments for the appropriateness of the prevalence odds ratio and related logistic regression

Background  

Several papers have discussed which effect measures are appropriate to capture the contrast between exposure groups in cross-sectional studies, and which related multivariate models are suitable. Although some have favored the Prevalence Ratio over the Prevalence Odds Ratio -- thus suggesting the use of log-binomial or robust Poisson instead of the logistic regression models -- this debate is still far from settled and requires close scrutiny.     

Contribution of interleukin-1 to activation of coagulation and fibrinolysis, neutrophil degranulation, and the release of secretory- type phospholipase A2 in sepsis: studies in nonhuman primates after interleukin-1 alpha administration and during lethal bacteremia

Although studies with interleukin-1 receptor antagonist (IL-1ra) in animal models have shown that IL-1 contributes to mortality in sepsis, the mechanisms whereby IL-1 mediates lethal effects are not well established. A possible mechanism is that IL-1 enhances the activation and release of other inflammatory mediator systems such as coagulation, fibrinolysis, neutrophils, and secretory-type phospholipase A2 (sPLA2). We investigated this possibility by assessing the effect of intravenously injected recombinant human IL-1 alpha (rhIL-1 alpha) on these plasma parameters in baboons. In addition, we examined the course of these inflammatory parameters in baboons after a challenge with a lethal dose of Escherichia coli and while receiving a 24-hour constant infusion of IL-1ra or placebo. Intravenous administration of IL-1 alpha (10 micrograms/kg) induced the formation of thrombin, as evidenced by the appearance of thrombin-antithrombin III (TAT) complexes into the circulation (peak levels, 188 +/- 92 ng/mL at 2 hours), as well as the activation of fibrinolysis, assessed by circulating plasmin-alpha 2- antiplasmin complexes (PAP complexes; peak levels, 0.4% +/- 0.03% of fully activated plasma at 1 hour), the release of tissue-type plasminogen activator (t-PA; peak levels, 6 +/- 2 ng/mL at 2 hours), and its inhibitor, plasminogen activator inhibitor (PAI; peak levels, 724 +/- 246 ng/mL at 4 hours). Il-1 alpha administration also induced the release of sPLA2 (maximal levels, 336 +/- 185 ng/mL at 8 hours), but not degranulation of neutrophils. In the septic baboons, a significant reduction of the formation of thrombin (peak TAT levels decreased from 582 +/- 78 ng/mL to 219 +/- 106 ng/mL; P < .005), the release of t-PA (peak levels decreased from 37 +/- 11 ng/mL to 17 +/- 2 ng/mL; P < .001), and its inhibitor, PAI (peak levels decreased from 2,639 +/- 974 ng/mL to 1,110 +/- 153 ng/mL; P <.001), was observed in the group receiving IL-1ra compared to that receiving placebo. The release of neutrophilic elastase was also significantly attenuated in IL-1a-treated animals (peak levels, 1,024 +/- 393 and 655 +/- 104 ng/mL in control and treatment groups, respectively; P < .05). The difference between sPLA2 levels in both groups, although higher in the controls (maximal levels, 3,140 +/- 1,435 ng/mL in control v 2,217 +/- 1,375 ng/mL in IL-1ra-treated group), was not significant. Thus, IL-1 contributes to activation of various other mediator systems in severe sepsis in nonhuman primates. We propose that these effects may explain the lethal actions of IL-1 in this sepsis model and suggest a similar role for IL-1 in severe human sepsis.     

Transduction of primitive human hematopoietic cells with recombinant adenovirus vectors

We have examined the ability of recombinant adenoviral vectors to transduce human hematopoietic cells. Our findings indicate that adenovirus readily infects a large proportion of CD34+ cells. Using adenovirus vectors that transduce either a lacZ or an alkaline phosphatase reporter gene, we observed up to 45% of total CD34+ cells infected. Upon more detailed analysis, we observed comparable levels of transduction for CD34+/CD38- cells and for CD34+ cells in G(zero) phase of the cell cycle. Importantly, exposure to adenovirus resulted in negligible levels of toxicity as assayed by propidium iodide staining and colony-forming ability. Using adenovirus vectors, we also describe a model system for regulated gene expression in early hematopoietic tissues. CD34+ cells were simultaneously infected with two viruses, one carrying a TetR/VP16 transactivator (tTA) and the second carrying a tTA- dependent lacZ reporter gene. Using this approach, beta-gal expression was only observed upon coinfection with the transactivator vector. In addition, as shown previously (Gossen and Bujard, Proc Natl Acad Sci USA 89:5547, 1992), tetracycline was able to inhibit tTA mediated induction, thereby providing an effective means to regulate expression of the reporter gene. We conclude that recombinant adenovirus is an effective vehicle for transiently expressing genes in primitive human hematopoietic cells.     

The myeloid blood cell differentiation-inducing protein MGI-2A is interleukin-6

The mouse myeloid blood cell differentiation-inducing protein, macrophage and granulocyte inducer, type 2A (MGI-2A), was purified, and the amino acid sequence of a CNBr cleavage peptide (22 residues) was determined. This amino acid sequence is identical to the sequence found in positions 73 to 94 of mouse interleukin-6 (IL-6). Recombinant mouse IL-6 protein induces differentiation of mouse myeloid leukemic cells that are induced to differentiation by MGI-2, and monoclonal antimouse- MGI-2 antibody, which neutralizes MGI-2, also completely neutralizes this IL-6-induced differentiation. These results show that the major type of mouse myeloid differentiation-inducing protein (MGI-2A) and IL- 6 are very similar and most likely identical proteins. Recombinant human IL-6 (also called interferon-beta 2 or B-cell differentiation factor), which shows only a 41% similarity to mouse IL-6, has 11 identical amino acid residues out of the 22 in the mouse MGI-2A peptide and also induces differentiation of the same myeloid leukemic cells.     

Purification of common acute lymphoblastic leukemia antigen positive cells from normal human bone marrow

Mononuclear cells expressing the common acute lymphoblastic leukemia antigen (CALLA) were purified from normal adult human bone marrow, where they constitute a small fraction of the total population. This was accomplished by a two-step purification from Ficoll-Hypaque- isolated mononuclear cells. Isolated mononuclear cells were first labeled with a mixture of monoclonal antibodies (MoAb) specific for myeloid and erythroid precursor cells, and immune rosettes were then formed with sheep erythrocytes coated with rabbit anti-mouse antibodies (R/M-SRBC). Sedimentation through Ficoll-Hypaque then eliminated the majority of mature myeloid cells. The second step consisted of labeling the remaining rosette-negative cells with CALLA-specific MoAb and purifying CALLA+ cells by fluorescence activated cell sorting. Alternatively, CALLA+ cells were purified in a second R/M-SRBC rosette sedimentation step. The purified CALLA+ cells, which morphologically were medium to large lymphoid cells, were subsequently studied using dual fluorescence techniques to identify surface markers as well as intracytoplasmic staining to detect terminal deoxynucleotidyl transferase enzyme (TdT) and intracytoplasmic mu. While the CALLA+ cell suspensions contained very few mature myeloid cells or T lymphocytes, the finding that 5% to 11% of them were cyto-mu+ and 13% to 22% expressed the B1 differentiation antigen clearly indicated that at least some of these cells were B cell precursors. Because 48% to 63% of the cells were TdT+ and practically all of them expressed Ia antigen, it appears that these cells are a mixture of very early lymphoid precursor cells as well as more differentiated pre-B cells. The phenotype of these normal cells is very similar to that of common ALL cells. Differences in the surface marker phenotypes between adult and fetal CALLA+ cells that have previously been purified were also identified.     

Induction of proliferation of purified human myeloid progenitor cells: a rapid assay for granulocyte colony-stimulating factors

The proliferation and differentiation of granulocyte and monocyte progenitor cells (CFU-C) in vitro is dependent on the presence of a group of closely related glycoproteins termed colony-stimulating factors (CSF). In order to investigate the interaction of these factors with CFU-C, we purified CFU-C from the peripheral blood of chronic myeloid leukemia patients with an immune rosette technique using specific monoclonal antibodies (mean 74-fold enrichment, 45% cloning efficiency). Colony formation by purified CFU-C demonstrated an absolute dependence on an exogenous source of CSF. Liquid culture of small aliquots of enriched CFU-C with CSF-containing medium resulted in a rapid, time- and concentration-dependent induction of DNA synthesis as measured by 3H-thymidine incorporation. This specific CSF induction of DNA synthesis by enriched CFU-C was used to develop a microassay system for CSF activity. CSF activity could be reproducibly quantitated in 24-48 hr. The proliferating cells in this assay system were shown to be myeloid progenitor cells by examining the morphology of their progeny and by determining the surface antigen phenotype of the responding cells (Ia+, T3-, B1-, Mo1-). This microassay provides a quantitative assessment of CSF activity that may be useful in the purification of human CSF and in the generation of monoclonal antibodies to CFU-C surface structures.     

Analysis of antigenic determinants on human monocytes and macrophages

Mo1, 2, 3, and 4, and Plt-1 are a series of five distinct antigens detected on the surface of human peripheral blood monocytes by mouse monoclonal antibodies. Mo2 and 3 are restricted to the monocyte- macrophage series, while Mo1, as previously reported, is also expressed by human granulocytes and null cells. Mo3, as distinguished from Mo1 and Mo2, is weakly expressed by virgin peripheral blood monocytes but becomes well expressed if monocytes are cultured overnight at 37 degrees C. Mo4 is coexpressed by monocytes and platelets, while Plt-1 appears to be a platelet-specific antigen whose detection on monocytes reflects adherence of platelets to monocyte membranes. That Mo2-4 are true monocyte antigens is demonstrated by their resynthesis following protease treatment of monocytes (Mol expression is resistant to proteolytic digestion). During myeloid-monocyte differentiation, the Mo antigens are infrequently expressed by immature myeloid cells but are found at higher frequency on leukemic monocytic forms. Macrophages from cultured peripheral blood monocytes and HL-60 cells exposed to lymphokines or phorbol diester express Mo1-4, but noncirculating peritoneal macrophages lack Mo3. The Mo antigens are differentiation markers whose expression reflects membrane heterogeneity during myeloid- monocyte-macrophage maturation.