Diagnostic utility of cerebrospinal fluid studies in patients with clinically suspected tuberculous meningitis.

OBJECTIVE: To compare yields of cerebrospinal fluid (CSF) studies in the diagnosis of tuberculosis meningitis (TBM). DESIGN: Prospective laboratory study, Kenyatta National Hospital, Kenya. STUDY POPULATION: Consecutive patients with 1) headache, neck stiffness and altered consciousness for more than 14 days, 2) above features plus evidence of tuberculosis elsewhere in the body, and 3) on standard antimeningitic drugs for one week without response, were included. Those with contraindications to lumbar puncture, confirmed causes of meningitis (except TB) and on anti-tuberculosis treatment were excluded. METHODS: CSF cell counts, glucose and protein were assayed. CSF was stained on ZN, cultured on LJ and BACTEC and subjected to PCR and LCR for Mycobacterium tuberculosis DNA sequences. Positive tests for M. tuberculosis were classified as definite and the rest as probable TBM. RESULTS: Fifty-eight patients with a mean age of 33.0 years were recruited. Mean CSF cell count was 71/microl and CSF lymphocyte count up 67%. Mean CFS protein and glucose were 2.10 g/l and 2.05 mmol/l, respectively. BACTEC was positive in 20 cases, LJ 12, LCR eight, and PCR and ZN one each. Twenty-six patients had definite and 32 probable TBM. Patients with definite TBM had significantly higher CSF protein, lower CSF glucose, higher CSF cell count and lower CSF lymphocytes. CONCLUSION: TBM can be confirmed in half of clinically suspected cases. More sensitive tests for confirmation of TBM are required.     

Flow cytometry and cytomorphology evaluation of hematologic malignancy in cerebrospinal fluids: comparison with retrospective clinical outcome

An independent clinical assessment was compared with flow cytometry (FCM) and cytomorphology results obtained on 227 cerebrospinal fluids investigated for hematologic malignancy, in a retrospective longitudinal study with a median observation time of 11 months. A combined method assessment (CMA), defining "positive" a sample if at least one method gave "positive" results, was also tested. Eleven out of 55 screening samples and 53 out of 166 follow-up samples resulted positive at clinical evaluation. FCM and CM were concordant with positive clinical assessment in 68.5% and 51.5% of cases, respectively. According to CMA, 10.5% of samples (resulting false negative by either FCM or cytomorphology) were rescued as true positive. FCM retained significantly higher accuracy than cytomorphology (p=0.0065) and 100% sensitivity when at least 220 leukocytes were acquired. CMA accuracy was higher than FCM accuracy and significantly higher than cytomorphology accuracy in the analysis of all samples (p<0.0001), samples from mature B/T cell neoplasms (p=0.0021), and samples drawn after intrathecal treatment (p=0.0001). When acquiring ≤220 leukocytes, FCM accuracy was poor, and combining cytomorphology added statistically significant diagnostic advantage (p=0.0043). Although FCM is the best diagnostic tool for evaluating CSF, morphology seems helpful especially when clinically positive follow-up samples are nearly acellular.     

New flow cytometry in hematologic malignancies

The complex and heterogeneous biology of acute lymphoblastic leukemia is unfolding as more experimental approaches such as expression profiling, DNA sequencing, and high resolution genomic arrays continue to unveil new recurrent alterations. In this perspective article, Drs. Cools and Vandenberghe examine new diagnostic applications of flow cytometry. See related paper on page 1767.Lymphoblastic leukemia/lymphoblastic lymphoma represents a heterogeneous group of neoplasms arising from lymphoblasts committed to either the B- or T-cell lineage. The term acute lymphoblastic leukemia (ALL) is reserved for those cases with extensive involvement of bone marrow or peripheral blood, with 25% bone marrow blasts as a commonly used threshold for defining ALL. The laboratory diagnosis of ALL is complex and is based on cytomorphological, immunological, cytogenetic and molecular analysis. Detailed immunophenotyping of leukemic blasts distinguishes several differentiation stages of B- or T-lymphoblasts: early precursor B-ALL or pro-B-ALL, common B-ALL, and pre-B-ALL for B-lineage lymphoblasts; and pro-TALL, pre-T-ALL, cortical T-ALL and medullary T-ALL.^1,2 Over the years, the association between the presence of specific genomic abnormalities with distinct clinical or phenotypic features, or therapeutic responses and survival has led to the recognition of several distinct clinicobiological entities. Among the B-ALL with recurrent genetic abnormalities, the new 2008 WHO classification now recognizes seven genetic entities: B-ALL with t(9;22)(q34;q11); B-ALL with t(v;11q23), B-ALL with t(12;21)(p13;q22), B-ALL with hyperdiploidy, B-ALL with hypodiploidy, B-ALL with t(5;14)(q31;q32) and B-ALL with t(1;19)((q23;p13). In addition to these specific entities, many other recurrent genetic abnormalities have been reported in B-ALL as well as in T-ALL. The prognostic significance of specific genomic lesions has been incorporated into risk stratification and risk-adapted therapy.^3,4 In some cases, they also identify specific subgroups that are candidates for targeted therapies such as imatinib or dasatinib, e.g. B-ALL with t(9;22) or T-ALL with NUP214-ABL1 episomal amplification, or gamma-secretase inhibitors for T-ALL with NOTCH1 mutations.⁵ Finally, genetic analysis at diagnosis may identify specific markers that can be used to follow up the kinetics of the response to treatment and minimal residual disease. Detection of these markers requires higly sensitive technologies.A standardized protocol for the cytogenetic analysis of B-ALL has been proposed.⁶ Standard cytogenetic analysis will readily identify cases with hyperdiploidy (more than 50 but usually less than 66 chromosomes), which have a good prognosis; cases with hypodiploidy (less than 44–46 chromosomes), which have a poor prognosis. Cytogenetic analysis will also easily identify t(9;22), t(1;19) and t(4;11). Additional FISH and PCR analyses are mandatory to identify cytogenetically cryptic aberrations such as t(12;21), or cytogenetically challenging aberrations, especially MLL rearrangements, e.g. t(9;11)(p22;q23) or t(11;19)(q23;p13). These techniques can also be very helpful to resolve the karyotype from poor quality metaphases as often seen in ALL. Thanks to their greater sensitivity, quantitative PCR methods are the method of choice for the follow-up of minimal residual disease.The complex and heterogeneous biology of B-ALL continues to unfold as more experimental approaches, such as expression profiling, DNA sequencing, and high resolution genomic arrays continue to reveal new recurrent alterations. It is now established that each of the chromosomal rearrangements that occur in ALL are by themselves not sufficient to cause the development of leukemia, and that additional mutations are present in those leukemias. Genome-wide copy number variation screening approaches have identified a number of additional cryptic lesions, such as the deletion and mutation of IKZF1, EBF1, PAX5 in B-ALL,⁷ and the duplication of MYB in T-ALL.^8,9 The clinical meaning of these aberrations remains to be fully determined, but it is already clear that IKZF1 lesions have an independent prognostic meaning.¹⁰ In addition, these genome-wide screening approaches have also demonstrated that relapse often arises from pre-leukemic clones, and, hence, that not all mutations present at diagnosis will be found again at the time of relapse.^11,12 The latter observation poses yet another analytical challenge for the diagnostic laboratory in that the use of a single method to follow up the status of a patient during remission may not be sufficient. Rather, a combination of methods is desired so that the development of novel clones at relapse is not missed.In this issue, Raponi and co-workers have studied the efficacy, accuracy and sensitivity of the use of a bead-based flow cytometric immunoassay, yet another method that can be used for the detection of a specific fusion protein in a leukemia sample.^13,14 This method uses a commercially available assay in which the BCR-ABL protein is detected in the lysed blood or bone marrow sample. Here, in contrast to the usual routine hematologic flow cytometry, flow cytometry is used to detect an intracellular protein, not a surface marker. In addition, the assay is performed not on intact cells but on a cell lysate obtained from bone marrow or peripheral blood. BCR-ABL proteins are immunocaptured on beads coupled with an anti-BCR capture antibody, and are subsequently detected using a secondary anti-ABL1 detection antibody. This method is in principle very similar to a classical ELISA test, but uses beads instead of multi-well plates in combination with a readout that is based on flow cytometry. This method was found to be very accurate in samples where the percentage of BCR-ABL positive cells was sufficiently high (over 10%), and highly concordant with molecular assays. Yet, it failed to detect the BCR-ABL protein in 2 steroid treated samples with a very low percentage of BCR-ABL positive cells. In a diagnostic setting, the rapid turnaround time, ease and specificity of this assay are extremely attractive. It may also provide a molecular diagnosis in geographical areas where diagnostic PCR is not available. In principle, the assay could be adapted to detect other types of oncogenic fusion proteins, depending on the availability and/or development of dedicated antibodies. The data reported here by Raponi et al. nicely and independently confirm the data published by the EuroFlow Consortium.¹⁵ As indicated by the two cases with discordant flow cytometric and PCR results, the assay is less sensitive than RT-PCR, which needs to be taken into account when performing these assays, especially during follow-up.The papers by Raponi¹⁴ and Weerkamp¹⁵ potentially extend the application field of flow cytometry to the detection of the BCR-ABL fusion protein in t(9;22) positive B-ALL and CML, and to the less common cases of t(9;22) positive T-ALL and AML. Flow cytometry has long been an established cornerstone in the diagnosis of hematologic malignancies, mainly to identify the malignant cell type by detection of cell surface proteins that provide information on its differentiation and/or maturation stage.¹⁶ In ALL, detailed antibody panels are used to determine the presence of B- or T-cell markers, and to further delineate the stage of B- or T-cell differentiation. This is important for diagnosis and allows diagnostic molecular and/or FISH algorithms to be rationally triggered to fulfil clinical needs according to modern standards.The constitutive phosphorylation of the BCR-ABL fusion protein is key to its transforming capacity. In more recent years, powerful assays have been developed to analyze the activation state (phosphorylation state) of signaling proteins within intact cells using fluorescently labeled phospho-specific antibodies. This, so-called, ‘phospho-flow’ analysis was pioneered by Gary Nolan’s group (http://proteomics.stanford.edu/nolan/phospho-flow), and generates novel insights into signaling pathways in growth factor stimulated normal cells and cancer cells.¹⁷ This method takes advantage of the power of flow cytometry to interrogate single individual cells, and thus detects cell populations with different signaling properties within a heterogeneous sample.In the era of tyrosine kinase inhibitor therapy, rapid analysis of the phosphorylation state of oncogenic proteins such as BCR-ABL, or its downstream effectors such as STAT5 or CRKL, is becoming increasingly important. For example, treatment of a patient with BCR-ABL positive B-ALL with an ABL inhibitor will lead to the dephosphorylation of BCR-ABL and its downstream effectors, except in those cases with BCR-ABL mediated resistance. Thus, the analysis of a blood sample before administration of the first drug dose could provide information on the response to be expected, while subsequent samples under treatment could confirm the response to optimize the drug dose or to suggest switching to another inhibitor if no response can be documented. One example of such a test is the analysis of phospho-CRKL, a surrogate marker for the activation of BCR-ABL. Analysis of the phosphorylation of CRKL can be made by Western blot analysis,^18,19 but could also be performed using ‘phospho-flow’ analysis.¹⁷ While Western blot analysis is a time consuming method, it is less sensitive to noise, and possible non-specific binding of the antibody used. In contrast, flow cytometric analysis of phospho-CRKL and/or other signaling proteins requires less hands-on time, and can provide results in just one day. Unfortunately, these techniques are more subject to noise and the quality and specificity of the antibodies used are of utmost importance to obtain a sufficient signal to noise ratio. In addition, the assay requires fresh samples with viable cells. At the moment these are still major hurdles for the analysis of phospho-proteins using flow cytometry. Despite these difficulties, a first demonstration of the application of ‘phospho-flow’ methods for the follow-up of response to ABL inhibitors has been reported,^20,21 and we can expect that further improvements may lead to a more general application of this technology in clinical practice.Provided cells or lysates can be stored under conditions that preserve the phosphorylation status, the flow cytometric assay of Raponi and co-workers could potentially be adapted to measure phosphorylated BCR-ABL instead of or in addition to the presence of BCR-ABL. Although at the expense of single cell analysis, a bead assay for BCR-ABL phosphorylaton could obviate the cumbersome need for viable cells in ‘phospho-flow’ analysis, and might provide a more convenient approach to functional assessment of cellular responses to tyrosine kinase inhibitors, that might be applicable on a wider scale.     

Venous thromboembolism in patients with hematologic malignancy and thrombocytopenia

The optimal management of hematologic malignancy‐associated venous thromboembolism (VTE) in patients with moderate‐to‐severe thrombocytopenia is unclear. This is a retrospective study of 128 adult patients with hematologic malignancies who were diagnosed with VTE. The outcome of patients with significant thrombocytopenia (≤50,000/µL) was compared with those without. Forty‐seven patients (36.7%) had a platelet count ≤50,000/µL during a period of time of perceived need for new or continued anticoagulation. The median nadir platelet count in those with significant thrombocytopenia was 10,000/µL (range 2,000–45,000/µL) versus 165,000/µL (50,000–429,000/µL) in those without (P < 0.001). The median duration of significant thrombocytopenia in the first group was 10 days (1–35 days). Therapy during the period of significant thrombocytopenia included prophylactic‐dose low‐molecular‐weight heparin ( LMWH) (47%), therapeutic‐dose LMWH or heparin (30%), warfarin (2%), inferior vena cava filter (2%), and observation (17%). Patients without thrombocytopenia were managed with the standard of care therapy. At a median follow‐up of more than 2 years, the risk of clinically significant bleeding (11% vs 6%, P = 0.22) including major bleeding (6% vs 2%) and clot progression or recurrence (21% vs 22%, P = 1.00) were similar in patients with or without significant thrombocytopenia. In a multivariate analysis, the risk of recurrence/progression (hazard ratio, HR 0.59, 95% CI 0.21–1.66, P = 0.31) and hemorrhage rate (HR 0.29, 95% CI 0.05–1.56, P = 0.15) did not differ based on the presence of significant thrombocytopenia. Within the limits of this retrospective study, cautious use of prophylactic‐dose LMWH may be safe in thrombocytopenic patients with hematologic malignancy‐associated VTE. Am. J. Hematol. 91:E468–E472, 2016. © 2016 Wiley Periodicals, Inc.     

Diagnostic utility of flow cytometry in low-grade myelodysplastic syndromes: a prospective validation study

Background

The diagnosis of myelodysplastic syndromes is not always straightforward when patients lack specific diagnostic markers, such as blast excess, karyotype abnormality, and ringed sideroblasts.

Design and Methods

We designed a flow cytometry protocol applicable in many laboratories and verified its diagnostic utility in patients without those diagnostic markers. The cardinal parameters, analyzable from one cell aliquot, were myeloblasts (%), B-cell progenitors (%), myeloblast CD45 expression, and channel number of side scatter where the maximum number of granulocytes occurs. The adjunctive parameters were CD11b, CD15, and CD56 expression (%) on myeloblasts. Marrow samples from 106 control patients with cytopenia and 134 low-grade myelodysplastic syndromes patients, including 81 lacking both ringed sideroblasts and cytogenetic aberrations, were prospectively analyzed in Japan and Italy.

Results

Data outside the predetermined reference range in 2 or more parameters (multiple abnormalities) were common in myelodysplastic syndromes patients. In those lacking ringed sideroblasts and cytogenetic aberrations, multiple abnormalities were observed in 8/26 Japanese (30.8%) and 37/55 Italians (67.3%) when the cardinal parameters alone were considered, and in 17/26 Japanese (65.4%) and 42/47 Italians (89.4%) when all parameters were taken into account. Multiple abnormalities were rare in controls. When data from all parameters were used, the diagnostic sensitivities were 65% and 89%, specificities were 98% and 90%, and likelihood ratios were 28.1 and 8.5 for the Japanese and Italian cohorts, respectively.

Conclusions

This protocol can be used in the diagnostic work-up of low-grade myelodysplastic syndromes patients who lack specific diagnostic markers, although further improvement in diagnostic power is desirable.     

Detection of malignant haematopoietic cells in the cerebrospinal fluid by conventional cytology and flow cytometry

Morphological evaluation of cerebrospinal fluid (CSF) samples by light microscopy is the method of choice for the detection of central nervous system involvement in patients with haematological malignancies. In this retrospective study, we assessed the value of three-colour immunophenotyping by flow cytometry in addition to conventional cytology in 28 patients with and 17 patients without pre-existent haematological malignancy in whom CSF analysis was performed because of neurological abnormalities (n = 37) or as part of routine staging procedures (n = 8). Four samples could not be analysed by flow cytometry because of insufficient cell numbers. CSF involvement was detected in a total of 18 patients, in 12 by both cytology and flow cytometry, and in three each by either cytology or flow cytometry alone. Discordant results were particularly frequent in patients with low CSF cell numbers and B cell neoplasms. When combined with conventional cytology, flow cytometric analysis improves the detection rate of CSF involvement in patients with haematological malignancies. It is particularly useful in B cell neoplasms with low CSF cell numbers.     

Thoracentesis in patients with hematologic malignancy: yield and safety

BACKGROUND: Pleural effusions occur in patients with hematologic malignancies, particularly during periods of hospitalization. Thoracentesis is often performed to diagnose infection and to exclude the presence of complicated parapneumonic effusions. The efficacy and safety of thoracentesis in this setting has not been well-studied. DESIGN: Retrospective chart review of hospitalized patients with hematologic malignancies undergoing thoracentesis. The aim of this study was to assess the role of thoracentesis in establishing a diagnosis of infection in this population and to determine the risk of complications. RESULTS: A total of 100 thoracentesis findings were analyzed in patients with lymphoma (52 patients) and leukemia (27 patients), and in patients who had undergone bone marrow or stem cell transplantation (21 patients). The indication for performing thoracentesis was to exclude infection in 69% of cases. Fever was present in 59% of the patients, and a concomitant lung parenchymal abnormality was present in 69% of cases. Effusions were moderate to large in size (87% of cases), and were both bilateral (62%) and unilateral (38%). Exudates were documented in 83%of the cases. A specific diagnosis was found in 21 patients and was more frequently established in those with lymphoma (31%) compared to the other groups of patients. Diagnoses found included malignancy in 14 cases, chylous effusions in 6 cases, and infection in 1 case. The one patient in whom empyema was found required drainage. The criteria for a parapneumonic effusion were not found in any other patients. The complication rate of 9% (pneumothorax, seven patients; hemothorax, two patients) was comparable to that in other populations of patients. CONCLUSIONS: Despite a high propensity for developing pulmonary infections, hospitalized patients with hematologic malignancies rarely developed complex parapneumonic effusions. The etiology of many of the effusions that occurred in this setting was unclear.     

Invasive Aspergillus infections in hematologic malignancy patients

The incidence of invasive Aspergillus (IA) infections in patients with hematologic malignancies continues to increase. The most common species include Aspergillus fumigatus (approximately 90% of cases), A. flavus, A. niger, A. terreus, and A. nidulans. Most infections involve the pulmonary parenchyma, though systemic dissemination of the fungus from a primary pulmonary focus or the paranasal sinuses after hyphal invasion into blood vessels is frequent. Early diagnosis and initiation of appropriate antifungal therapy has been shown to improve the prognosis of patients afflicted with this condition. The definitive diagnosis of IA is based on showing the hyphal invasion in tissue specimens together with a positive culture for Aspergillus species from the same specimen. The detection of circulating fungal antigens and DNA seems to be a promising, rapid, and sensitive diagnostic tool for early diagnosis of aspergillosis. The current antifungals available for the treatment of IA include amphotericin B deoxycholate and lipid formulations, itraconazole and caspofungin acetate. New investigational antifungal drugs include the triazoles voriconazole, posaconazole and ravuconazole, liposomal nystatin, and 2 echinocandin derivatives (anidulafungin [VER-002] and micafungin [FK463]). Preventive measures include reduction of environmental exposure of patients from sources of infection and anti-fungal prophylaxis. Specialized air-handling systems capable of excluding Aspergillus spores, such as high-efficiency particulate air (HEPA) filtration with or without laminar air flow ventilation has proven to be very efficacious. Targeted antifungal prophylaxis for hematologic patients who are at high risk for developing invasive fungal infections is not currently standardized.     

Diagnostic utility of pleural fluid eosinophilia

It has been stated that pleural fluid eosinophilia (defined as greater than 10 percent eosinophils in the pleural white cell differential count) is not helpful in the diagnosis of exudative effusions. By review of the recent literature, it was found that pleural fluid eosinophilia was associated most often with idiopathic effusions or with air previously introduced into the pleural space. Also, a high proportion of "idiopathic" and benign asbestos effusions were characterized by pleural fluid eosinophilia, a previously unrecognized phenomenon. The diagnostic utility of finding eosinophils in the pleural space was assessed from its impact on prior probabilities of disease. Estimates of pretest likelihoods of malignant versus nonmalignant pleural effusions and the prevalence of eosinophilia in effusions of known cause were obtained from extensive literature review. These were modified by using Bayes' rule to estimate the revised probability of disease in the presence of an eosinophilic effusion. The presence of pleural fluid eosinophilia considerably reduced the probability of malignancy or tuberculosis and increased the likelihood of an underlying benign disorder. Pleural fluid eosinophilia is a useful finding that can aid in the diagnosis of an exudative pleural effusion.     

The value of multiparameter flow cytometry of cerebrospinal fluid involved by leukemia/lymphoma cells

The usefulness of multiparameter flow cytometric (FC) analysis of cerebrospinal fluid (CSF) was evaluated in leukemia/lymphoma patients having central nervous system (CNS) involvement of the disease. In 12 specimens of 8 patients with different types of leukemia/lymphoma (one case of T-ALL, 3 cases of early B-cell ALL, one case of AML, and 3 proven or suspicious NHL cases) the presence of pathological clone in CSF has been confirmed or excluded. The phenotypic patterns of CSF cells were defined according to those of bone marrow (BM)/peripheral blood (PB) at diagnosis or during follow-up of the same patients. Furthermore, in one case of suspicious CNS infiltration of NHL, the pathological clone was characterized as a highly suspicious of solid tumor and was proved to be a lung cancer metastasis. The definition was made on the basis of CD45 (common leukocyte antigen) and other studied CD markers negativity. The exact comparison of immunophenotypic profiles of specimens from different sites (CSF, BM, PB) of the same patient has been performed and no phenotypic changes were found. In some CSF specimens, where no cells of suspicious pathological clone were detected, in 4-color analysis only normal lymphocyte population was found even in small cell samples (even if the cellularity was < than 0.3x10-6). In these populations the high values of T-cells (CD3+) predominated and the high prevalence of CD4+ over CD8+ cells, and an almost total lack of B-lymphocytes was found. Our results suggest that positive CSF immunology is a useful indicator of malignancy and reflects leptomeningeal involvement. Simultaneously we demonstrated that FC analysis of CSF in the aim to detect possible CSF seeding of leukemia/lymphoma is a reliable and quick technique.     

Pulmonary alveolar proteinosis: a complication in patients with hematologic malignancy

We present the case of a patient with acute myeloid leukemia and secondary Pulmonary Alveolar Proteinosis (PAP), which is an underestimated cause of a persistent pulmonary infiltrate in patients with hematologic malignancies often accompanied by neutropenia due to therapy. Diagnosis is established by performing Periodic Acid-Schiff reaction (PAS) stains and transmission electron microscopy (EM) on bronchoalveolar lavage (BAL) fluid. We wish to stress the importance of the early recognition of PAP, which is potentially reversible, and routinely performing PAS staining on BAL fluid in patients with a hematologic disease especially myeloid disorders and a persistent lung infiltrate.     

Successful renal transplantation in patients with prior history of malignancy

Eleven selected patients with a history of cancer have received renal transplants. In three patients the malignancies were diagnosed and treated before renal failure developed, two patients underwent bilateral nephrectomies for the treatment of bilateral renal malignancy. In six patients the malignancy developed while they were uremic, and they received transplants after treatment of the tumor. Recurrent tumor developed in only One of the 11 (9 per cent) patients after receiving the transplant. Thus, patients with a history of malignancy can receive a transplant without accelerating growth of the tumor. We recommend a minimum period of one year between tumor treatment and transplantation to observe for tumor recurrences or metastases. A longer period of observation may be necessary in patients who have tumors with a poorer prognosis.     

High-dose etoposide and cyclophosphamide without bone marrow transplantation for resistant hematologic malignancy

Seventy-five patients with resistant acute leukemia or lymphoma received high-dose cyclophosphamide and etoposide to explore the activity of this combination in resistant hematologic malignancies, and to determine the maximum doses of these drugs that can be combined without bone marrow transplantation. Etoposide was administered over 29 to 69 hours by continuous infusion corresponding to total doses of 1.8 g/m2 to 4.8 g/m2. Cyclophosphamide, 50 mg/kg/d, was administered on 3 or 4 consecutive days total 150 to 200 mg/kg ideal body weight). At all dose levels myelosuppression was severe but reversible. Mucosal toxicity was dose-limiting with the maximum tolerated dose level combining etoposide 4.2 g/m2 with cyclophosphamide 200 mg/kg. Continuous etoposide infusion produced stable plasma levels that were lower than would be achieved after administration by short intravenous infusion, and this could explain our ability to escalate etoposide above the previously reported maximum tolerated dose. There were 28 complete (35%) and 12 partial (16%) responses. Median duration of complete response (CR) was 3.5 months (range 1.1 to 20+). Seventeen of 40 patients (42%) with acute myelogenous leukemia (AML) achieved CR, including 6 of 20 (30%) with high-dose cytosine arabinoside resistance. We conclude that bone marrow transplantation is not required after maximum tolerated doses of etoposide and cyclophosphamide. This regimen is active in resistant hematologic neoplasms, and the occurrence of CR in patients with high-dose cytosine arabinoside-resistant AML indicates a lack of complete cross-resistance between these regimens.     

Diagnostic utility of eosinophils in the pleural fluid.

This study was conducted to assess the prevalence of eosinophilia in 358 consecutive samples of pleural fluid (all cases corresponded to first thoracentesis), to review the cause of eosinophilic pleural effusions, and to determine whether the presence of eosinophils increases the likelihood of nonmalignant underlying disorders. Eosinophilic pleural effusions were identified in 45 patients (12.6%): malignant underlying conditions were diagnosed in 11 patients (24.4% with eosinophilic effusions) and benign aetiologies were found in 27 patients. Benign aetiologies included uncomplicated paraneumonic effusion in 10 patients, tuberculosis in seven, complicated paraneumonic in five, liver cirrhosis in three, hydronephrosis in one and pulmonary thromboembolism in one. Seven pleural effusions were idiopathic. There was no difference in the prevalence between eosinophilic and noneosinophilic effusions according to the different diagnoses. With parameters of sensitivity, specificity, pretest and post-test probability and positive and negative predictive values for any prevalence figure using the Bayes' theorem and for any value of eosinophils (both in percentage or absolute numbers) in the pleural fluid (receiver operating characteristic curve) an adequate predictor of benign disease was not found. It is concluded that pleural eosinophilia at the initial thoracentesis cannot be considered as a predictor of an underlying benign disorder.