A bi-level multi-follower optimization model for R&D project portfolio: an application to a pharmaceutical holding company.

The need for a study of project portfolio optimization in pharmaceutical R&D has become all the more urgent with the outbreak of COVID-19. This study examines a new model for optimizing R&D project portfolios under a decentralized decision-making structure in a pharmaceutical holding company. Specifically, two levels of decision makers hierarchically decide on budget allocation and project portfolio selection-scheduling to maximize their profit, and we formulate the problem as a bi-level multi-follower mixed-integer optimization model. At the upper level, the investment company has complete knowledge of the subsidiaries' response, acts first, and decides on the best budget allocation. At the lower level, each subsidiary responds to the allocated budget and decides on its portfolio scheduling. Since the lower level represents several mixed-integer programming problems, solving the resulting bi-level model is challenging. Therefore, we propose an efficient hybrid solution approach based on parametric optimization and convert the bi-level model into a single-level mixed-integer model. To validate it, we solve a case and discuss the optimal strategy of each actor. The experimental results show that the planned project portfolio for each subsidiary of the holding company is drastically affected by the allocated budget and its decisions. Supplementary Information: The online version contains supplementary material available at 10.1007/s10479-022-05052-0.

Estudio observacional multicéntrico retrospectivo sobre el manejo clínico y terapéutico de los diferentes tipos de estatus epiléptico en la práctica clínica / Retrospective multicentre observational study on clinical management and treatment of different types of status epilepticus in clinical practice

Introducción: El estatus epiléptico es una urgencia neurológica asociada a una mortalidad y morbilidad significativa. Analizamos las características en nuestra población. Métodos: Se recogieron los datos de manera retrospectiva de la historia clínica electrónica de adultos con diagnóstico de estatus epiléptico en 5 centros hospitalarios durante 4 años. Resultados: Se obtuvieron datos de un total de 84 episodios en 77 pacientes, con edad media de 60,3 años. El 52,4% tenían historia previa de epilepsia. Clasificación según el tipo de estatus: 47,6% tónico-clónico; 21,4% parcial complejo; 17,9% parcial motor; 6% parcial simple; 3,6% mioclónico y 3,6% sutil. Si analizamos el momento que finalizó el estatus según las fases definidas para este estudio obtenemos: 13,1% precoz (hasta 30 min); 20,2% establecido (entre 30-120 min); 41,7% refractario (más de 120 min) y 13,1% superrefractario (continúan o recurren después de más de 24 h de anestesia). Diez casos (11,9%) fallecieron sin haberse controlado el estatus. El porcentaje acumulativo de éxito alcanzado con el primer tratamiento fue de 8,3%; segundo 27,3%; tercero 48,7%; cuarto 58,2%; quinto 70,1%; sexto 80,8%; séptimo 83,2% y octavo 84,4%. Conclusiones: En nuestro estudio encontramos que el estatus no se controló en las primeras 2 h en casi la mitad de los casos, y un 11,9% fallecieron sin controlarse, sin haber diferencias significativas entre el tipo de estatus. En casi la mitad se logró el control del estatus con el tercer tratamiento, pero en algún caso se precisó hasta 8. Son necesarios registros amplios que permitan analizar el manejo en los distintos tipos y fases (AU)

Introduction: Status epilepticus (SE) is a neurological emergency associated with significant mortality and morbidity. We analyse characteristics of this entity in our population. Methods: Data from electronic medical records of adults diagnosed with SE were collected retrospectively from 5 hospitals over 4 years. Results: Data reflected 84 episodes of SE in 77 patients with a mean age of 60.3 years. Of this sample, 52.4% had a previous history of epilepsy. Status classification: 47.6% tonic-clonic, 21.4% complex partial, 17.9% partial motor, 6% partial simple, 3.6% myoclonic, and 3.6% subtle SE. Based on the duration of the episode, SE was defined in this study as early stage (up to 30 min) in 13.1%, established (30-120 min) in 20.2%, refractory (more than 120 min) in 41.7%, and super-refractory (episodes continuing or recurring after more than 24h of anaesthesia) in 13.1%. Ten patients (11.9%) died when treatment failed to control SE. The cumulative percentage of success achieved was 8.3% with the first treatment, 27.3% for the second, 48.7% for the third, 58.2% for the fourth, 70.1% for the fifth, 80.8% for the sixth, 83.2% for the seventh, and 84.4% for the eighth. Conclusions: In our study, we found that SE did not respond to treatment within 2h in approximately half the cases and 11.9% of the patients died without achieving seizure control, regardless of the type of status. Half the patients responded by the third treatment but some patients needed as many as 8 treatments to resolve seizures. Using large registers permitting analysis of the different types and stages of SE is warranted (AU)

Retrospective multicentre observational study on clinical management and treatment of different types of status epilepticus in clinical practice. / Estudio observacional multicéntrico retrospectivo sobre el manejo clínico y terapéutico de los diferentes tipos de estatus epiléptico en la práctica clínica.

INTRODUCTION: Status epilepticus (SE) is a neurological emergency associated with significant mortality and morbidity. We analyse characteristics of this entity in our population. METHODS: Data from electronic medical records of adults diagnosed with SE were collected retrospectively from 5 hospitals over 4 years. RESULTS: Data reflected 84 episodes of SE in 77 patients with a mean age of 60.3 years. Of this sample, 52.4% had a previous history of epilepsy. Status classification: 47.6% tonic-clonic, 21.4% complex partial, 17.9% partial motor, 6% partial simple, 3.6% myoclonic, and 3.6% subtle SE. Based on the duration of the episode, SE was defined in this study as early stage (up to 30min) in 13.1%, established (30-120min) in 20.2%, refractory (more than 120min) in 41.7%, and super-refractory (episodes continuing or recurring after more than 24h of anaesthesia) in 13.1%. Ten patients (11.9%) died when treatment failed to control SE. The cumulative percentage of success achieved was 8.3% with the first treatment, 27.3% for the second, 48.7% for the third, 58.2% for the fourth, 70.1% for the fifth, 80.8% for the sixth, 83.2% for the seventh, and 84.4% for the eighth. CONCLUSIONS: In our study, we found that SE did not respond to treatment within 2h in approximately half the cases and 11.9% of the patients died without achieving seizure control, regardless of the type of status. Half the patients responded by the third treatment but some patients needed as many as 8 treatments to resolve seizures. Using large registers permitting analysis of the different types and stages of SE is warranted.

Bilateral cochlear implantation in a patient with petrous bone cholesteatoma in the only hearing ear: case report.

OBJECTIVE: We report a case of successful cochlear implantation in a patient with petrous bone cholesteatoma in the only hearing ear. CASE REPORT: A 63-year-old man presented with a four-year history of right-sided, progressive hearing loss in his only hearing ear. Computed tomography and magnetic resonance imaging revealed a right supralabyrinthine petrous bone cholesteatoma, with erosion of the superior semicircular canal and the roof of the internal auditory canal. Due to the high risk of post-operative right-sided deafness, we decided first to perform left cochlear implantation. Five months later, the patient had a 40 per cent score for open-set two-syllable word recognition and an 85 per cent score for sentence recognition. Given these good performances, we decided to eradicate the cholesteatoma via a translabyrinthine approach, with insertion of a second cochlear implant, as a single-stage procedure. A successful outcome was achieved. CONCLUSION: Cochlear implantation can be an effective method of hearing rehabilitation in patients with petrous bone cholesteatoma, following total eradication of disease, if the cochlea remains intact. To our best knowledge, this is the first English language report of cochlear implantation in a patient with petrous bone cholesteatoma.

Immunologically mediated aplastic anemia in mice: evidence of hematopoietic stromal injury and injury to hematopoietic stem cells.

Immunologically mediated aplastic anemia (AA) results when lymph node cells (LNC) from C3H/He mice are injected intravenously (i.v.) into H-2 identical CBA/J mice previously given 600 cGy sublethal total-body gamma irradiation (TBI). Previously, we showed that T lymphocytes injure pluripotent hematopoietic stem cells and cause severe pancytopenia and death in 80 to 100% of mice within 3 to 4 weeks, with changes in the bone marrow suggesting stromal injury. The following models were used to study the stroma: (1) Transplantation of femurs from AA mice into normal syngeneic CBA/J mice. After 6 weeks, colony-forming unit-spleen (CFU-S) levels in the femur implants were measured in both AA and control mice (600 cGy TBI only). (2) Development of Dexter long-term bone marrow cultures from AA and control mice, which were used to support hematopoietic bone marrow cells (colony-forming units-granulocyte/macrophage [CFU-GM]) from normal mice. (3) Cellulose ester membranes (CEM) were coated with hematopoietic stroma from AA and control mice and then implanted intraperitoneally (i.p.) into syngeneic CBA/J mice. Six months later, the CEM were removed and analyzed for the presence of trilineal hematopoiesis and bone. Injury to the hematopoietic stroma was documented by the following: (1) Femurs from AA mice had a decreased number of CFU-S compared to controls; (2) Dexter cultures from AA mice formed abnormal stromal layers with a decreased capacity to support CFU-GM from normal donor mice; and (3) CEM coated with stromal cells from AA mice had a decreased capacity to support trilineal hematopoiesis and bone compared to CEM coated with marrow stroma from control mice.

Increased cytosolic free sodium concentrations in platelets from type 2 (non-insulin-dependent) diabetic patients is associated with hypertension.

Cytosolic free sodium concentrations ([Na+]i) in intact platelets from 32 type 2 (non-insulin-dependent) diabetic patients and from 27 age- and sex-matched non-diabetic control subjects were measured with the novel sodium-sensitive fluorescent dye sodium-binding-benzofuran-isophthalate. [Na+]i was significantly higher in platelets from type 2 diabetic patients compared with control subjects (40.6 +/- 2.4 vs 32.0 +/- 2.0 mmol/l, means +/- S.E.M., P < 0.03). Both systolic and diastolic blood pressure were significantly elevated in diabetic patients compared with control subjects. Analysis of diabetic patients showed a significant association between [Na+]i and diastolic blood pressure (P = 0.026). Stimulation of Na/H exchange by thrombin increased [Na+]i in both groups. After inhibition of Na/K/ATPase by ouabain (1 mmol/l), [Na+]i was significantly increased both in diabetic patients and non-diabetic subjects in a similar way (by 40.2 +/- 7.3 and 31.7 +/- 5.3 mmol/l respectively). It is concluded that increased [Na+]i in cells from type 2 diabetic patients may be related to hypertension.

Hematopoiesis on cellulose ester membranes. XIII. A combination of cloned stromal cells is needed to establish a hematopoietic microenvironment supportive of trilineal hematopoiesis.

A mixture of stromal cells from murine bone marrow placed upon cellulose ester membranes (CEM) and then implanted intraperitoneally (i.p.) in mice results in a regenerated hematopoietic microenvironment which supports trilineal hematopoiesis. We used this model to study the capacity of 5 cloned murine stromal cell lines of marrow origin to support hematopoiesis in vivo: MBA-1 (fibroblast); MBA-2 (endothelial); MBA-13 (fibroendothelial); 14F1.1 (endothelial-adipose); and 14M1.4 (macrophage).10(7) stromal cells of a single cell line were applied to 1.5 cm2 CEM, which were folded into tubes and implanted i.p. into mice. Similarly, combinations of 4, 3 and 2 stromal cell lines were applied to CEM and implanted i.p. Single lines were implanted into syngeneic hosts of the same murine strain from which the clone was derived and into nude mice. Combinations of stromal cells were implanted only in nude mice to avoid allogeneic incompatibility. CEM implants were removed after intervals of 5 to 36 weeks and examined histologically. 1) Stromal cells of a single phenotype did not develop hematopoiesis. 2) A combination of 4 stromal phenotypes (MBA-1, MBA-2, MBA-13 and 14F1.1) formed a hematopoietic microenvironment supportive of trilineal hematopoiesis and bone. 3) The combination of 14F1.1 (endothelial adipose) + a second stromal phenotype--MBA-1 (fibroblast) or MBA-2 (endothelial) or MBA-13 (fibroendothelial) also supported trilineal hematopoiesis and bone. 4) CEM coated with MBA-13 or MBA-1 developed bone but no hematopoiesis. The endothelial-adipose phenotype appears to be essential to support hematopoiesis but requires other types of stromal cells--fibroblast, fibroendothelial or endothelial phenotype.

Effect of spironolactone on cytosolic free sodium concentration in platelets from hypertensive patients with primary aldosteronism.

Cytosolic free sodium concentration ([Na+]i) was investigated in intact platelets from 5 hypertensive patients with primary aldosteronism (unilateral adenoma in 3 patients, and adrenal hyperplasia in 2 patients) and 21 normotensive control subjects. [Na+]i was measured using a novel sodium-sensitive fluorescent dye technique. [Na+]i was significantly decreased in platelets from patients with primary aldosteronism compared to control subjects (21.9 +/- 4.1 mM vs 35.8 +/- 2.2 mM, mean +/- SEM, P < 0.05). After administration of the mineralocorticoid antagonist spironolactone in 4 patients [Na+]i tended to be higher in platelets although the differences did not reach statistical significance (26.3 +/- 7.2 mM vs 18.2 +/- 2.4 mM, P = 0.125). From the present results it may be concluded that intracellular sodium is decreased by aldosterone-induced activation of Na-K-ATPase. That activation may be partly blocked by spironolactone.

[The spatial redistribution of Christians in Lebanon]. / La redistribution territoriale des chretiens au Liban.

PIP: The redistribution of the Christian population in Lebanon since 1975, primarily as a result of war, is described. The author notes that there has been a general movement of Christians from isolated communities toward the Christian redoubt in north central Lebanon.^ieng

Effect of inhibition of Na, K-ATPase on cytosolic free sodium and calcium in platelets of spontaneously hypertensive rats.

Cytosolic free sodium concentrations ([Na+]i) in intact platelets of 18 spontaneously hypertensive rats (SHR) and of 18 age-matched normotensive Wistar-Kyoto rats (WKY) were measured using the sodium-sensitive fluorescent dye sodium-binding-benzofuran-isophthalate. In resting platelets [Na+]i tended to be higher in SHR compared to WKY (20.5 +/- 3.5 mmol/L v 15.1 +/- 1.9 mmol/L, mean +/- SEM), but the differences were not statistically significant. Stimulation of the Na-H-exchange by 1.0 U/mL thrombin increased [Na+]i in SHR by 22.9 +/- 4.3 mmol/L and in WKY by 35.0 +/- 5.6 mmol/L in a similar way. After inhibition of Na, K-ATPase by 1 mmol/L ouabain there was a significant rise of [Na+]i both in platelets of SHR to 38.0 +/- 5.1 mmol/L (P < .01 compared to resting platelets) and in platelets of WKY to 26.5 +/- 4.3 mmol/L (P < .01). However, no significant difference could be observed between these two groups. Using the calcium-sensitive dye fura-2, resting cytosolic free calcium concentrations ([Ca2+]i) were found to be significantly higher in platelets of SHR compared to WKY (171.9 +/- 21.5 nmol/L v 93.14 +/- 19.7 nmol/L, P < .05). After the addition of ouabain [Ca2+]i was significantly higher in SHR compared to WKY (245.5 +/- 32.6 nmol/L v 159.6 +/- 22.5 nmol/L, P < .05). The results do not support the hypothesis that altered sodium-calcium exchange causes elevated cytosolic free calcium in SHR.

Reduced cytosolic free Na+ concentration in intact platelets of essential hypertensives.

OBJECTIVE: The role of intracellular Na+ concentration in the pathogenesis of essential hypertension is a point of considerable discussion. DESIGN: Since the novel fluorescent dye technique offers the possibility of measuring cytosolic free Na+ concentration in intact living cells, the role of Na+ was reinvestigated in resting and stimulated human platelets. METHODS: Cytosolic free Na+ concentration was measured in intact blood platelets of 20 essential hypertensive patients and 21 age- and sex-matched normotensive control subjects using the fluorescent dye Na(+)-binding benzofuran isophthalate. RESULTS: Cytosolic free Na+ concentration was significantly reduced in hypertensives compared with normotensives. Inhibition of Na+,K(+)-adenosine triphosphatase by ouabain elevated cytosolic free Na+ concentration in hypertensives and normotensives in a similar way. Addition of thrombin increased cytosolic free Na+ concentration both in hypertensives and normotensives. CONCLUSIONS: Previous concepts concerning the role of Na+ in the pathogenesis of essential hypertension based upon measurements in destructed cells need to be reinvestigarted using new techniques in living cells.

Cultured cell lines from human breast cancer biopsies and xenografts.

Eighty-five breast cancer specimens were processed as part of a program in tumor acquisition, propagation, and preservation for biotherapy. Nine long-term culture cell lines were developed. Four cell lines were from solid tumor metastases, two lines were from pleural fluid specimens, and three were from xenograft tumors grown in nude mice. Two of the xenograft-derived cell lines were from biopsies which produced tumor cell lines as well. Success in establishing cultures did not correlate with the viability of the biopsy received. Poor tumor cell attachment to culture plastic was the most common problem. For certain specimens, attachment and growth were enhanced on collagen and extracellular matrix substrates. Collagen was beneficial in the development of one cell line. The cell lines were characterized and each of the lines contained more nuclear DNA than found in normal cells. Four of five lines tested were tumorigenic in nude mice. Five of nine were clonogenic in soft agar. Each of the cell lines tested reacted with at least two anti-tumor monoclonal antibodies. Xenograft and biopsy-derived cell lines from the same tumor were similar in their characteristics. While breast cancers are indeed difficult to establish and propagate in culture, the use of xenografts and special substrates appears to be beneficial in the development of cell lines from some tumors.

Cultured breast cystosarcoma phylloides cells and applications to patient therapy.

Malignant cystosarcoma phylloides (CP) is a relatively rare cancer of the breast. A CP tumor was processed as part of a tumor acquisition, propagation, and preservation program in patient biotherapy. Two tissue culture cell lines were developed from this tumor, one directly from the biopsy, another from a xenograft tumor grown in athymic mice. The two cell lines were similar in character. There was strong immunochemical reactivity with antibodies to vimentin, type I collagen, and type III collagen. There was no reactivity with antibodies to cytokeratin and epithelial membrane antigen. Both cell lines were aneuploid, clonogenic in soft agar, and tumorigenic in nude mice. 5 alpha-dihydrotestosterone and thyroxine added to the culture medium stimulated growth, while testosterone, 17 beta-estradiol, and 4-hydroxytamoxifen were without effect. Dexamethasone and cortisol were inhibitory at high doses (10(-6) M). Dibutyryl cyclic AMP, theophylline, and vitamin C were all inhibitory. The biopsy contained tumor-infiltrating lymphocytes which proliferated in cultures containing interleukin 2. The expanded lymphocytes were activated T cells which had the capacity to lyse tumor cells. These results suggest possibilities in the therapy of cystosarcoma phylloides involving vitamin C, certain hormones, and tumor-infiltrating lymphocytes.

Hematopoiesis on cellulose ester membranes (CEM). XII. Effect of membrane enrichment by purified matrix proteins.

Cellulose ester membranes (CEM) were enriched with the following purified matrix proteins: collagen I, II, IV, proteoglycan and laminin. Fifteen milligrams of each were placed on CEM which were then folded into open-ended tubes implanted i.p. and s.c. CEM were removed after 3, 6 and 12 months and examined histologically. There was no evidence of hematopoiesis or new bone formation on the implanted, enriched CEM at any of the intervals examined. Collagen I and proteoglycan-enriched CEM showed evidence of increased sinusoid-like vascular structures.

Tumor acquisition, propagation, and preservation. The culture of human colorectal cancer.

Fourteen new colorectal cancer cell lines were developed as part of a tumor acquisition, propagation, and preservation program for biotherapy. Fifty-six specimens were received. Nine cell lines were generated from biopsies; seven of these cell lines were from metastatic lesions. Five additional cell lines were developed from xenografts grown in nude mice. Biopsies that produced three of these xenografts gave rise to parallel culture cell lines. Biopsy-derived and xenograft-derived cell lines from the same tumor behaved similarly in culture and exhibited similar markers when assessed immunohistochemically. Collagen substrate was beneficial in the primary culture of 50% of the specimens tested. Collagen was required for the successful propagation of two cell lines.

Hematopoiesis on cellulose ester membranes. XI. Induction of new bone and a hematopoietic microenvironment by matrix factors secreted by marrow stromal cells.

Cellulose ester membranes (CEM) were coated with stromal cells from bone marrow (BM) or bone and implanted intraperitoneally (IP) in CAF1 mice for intervals of 1 to 6 months. Previous studies indicated that matrix factors [glycoproteins (GPs), proteoglycans (PGs), and glycosaminoglycans (GAGs)] were secreted by the regenerating stromal cells and adsorbed by the CEM. After 1 to 6 months, the CEMs were removed, scraped free of adherent cells, and irradiated in vitro with 40 Gy. The scraped and irradiated CEMs were then reimplanted IP or subcutaneously (SC) for periods of 1 to 6 months in secondary syngeneic murine hosts. They were then removed for histologic study. CEMs reimplanted in SC sites developed bone and hematopoiesis as early as 1 month after implantation. Maximum hematopoiesis and bone formation was observed after 3 months. CEMs coated during the initial implantation with bone-derived stromal cells contained more bone and hematopoietic cells than did CEMs coated with marrow-derived stromal cells after SC implementation. Neither the CEMs coated with bone stromal cells nor those coated with marrow stromal cells developed new bone or trilineal hematopoiesis after being implanted IP. A few CEMs contained small foci of granulopoiesis only. We conclude that noncellular matrix substances deposited on CEMs by bone, and to a lesser degree by marrow cells, can induce prestromal cells in the SC tissues to produce a microenvironment suitable for trilineal hematopoiesis.

Immunohistochemical phenotyping of human solid tumors with monoclonal antibodies in devising biotherapeutic strategies.

A panel of 14 monoclonal antibodies (MoAbs) (4 raised against breast cancer, 6 against colon cancer and 4 against melanoma) were used to phenotype frozen sections of tumor biopsies obtained from 110 patients, by avidin-biotin-peroxidase complex techniques. We observed heterogeneity of antigen expression among the multiple metastatic lesions of single patients, as well as among tumor lesions from different patients with similar tumor histotypes. A wide range of cross-reactivity of anti-(breast-carcinoma) and anti-(colon-carcinoma) MoAbs with other carcinoma histotypes and limited reactivity with melanoma and sarcoma was detected. Some of our anti-melanoma MoAbs were also found to cross-react with selected carcinomas. Nine of the 14 MoAbs most reactive with carcinomas of diverse histotypes have been identified. A mixture or 'cocktail' of different MoAbs could be selected for each individual patient in order to achieve binding of MoAbs with most, if not 100% of tumor cells. This study illustrates the approach that we have taken to individualize the cocktail of MoAbs for the development of patient-specific therapeutic immunoconjugates.

The ultrastructure of hematopoietic stroma on cellulose ester membranes implanted intraperitoneally into Sl/Sld and Sl+/Sl+ mice.

The structural features of hematopoietic stromal elements forming on cellulose ester membranes (CEM) implanted intraperitoneally into hematopoietically impaired, anemic Sl/Sld mice and their normal Sl+/Sl+ littermates were compared by combined light and electron microscopy. The generally thicker, multilayered stroma lining the Sl+/Sl+ CEM implants developed from a bed--a syncytium--of large, highly pleomorphic macrophage-type lining cells whose filopodial extensions exhibited extensive interactions (i.e., nurse cell interactions) with both stromal and hematopoietic elements. In contrast, the thinner stromal layers lining the CEM of Sl/Sld mice formed from a base of dysplastic lining elements. These CEM-lining macrophage-type cells had much reduced cytoplasmic volumes, less extensive interactive surface projections, and an absence of select types of cytoplasmic organelles (e.g., membrane-bound crystalline inclusions). These observations suggest that the reduction of cell layering and, in turn, hematopoietic support activity, is due to an impaired interactive capacity of these elemental lining cells, i.e., pleomorphic macrophagic cell types, in the hematopoietically impaired strain of Sl/Sld mice.

Hematopoiesis on cellulose ester membranes (CEM). IX. Enrichment of membranes with adherent layers from long-term marrow culture from normal or drug-treated mice (hematopoietic microenvironment study II).

Cellulose ester membranes (Millipore) or polytetrafluoroethylene (Mitex) membranes were coated with adherent layers taken from Dexter-type long-term cultures, 4-5, 8 or 12 weeks after initiation of culture. The cultures were established with marrow taken from untreated mice or, in some cases, from mice treated with a single lethal dose (LD10) of carmustine (BCNU) or cyclophosphamide. In the studies using untreated mice, the cultures went for 8 or 12 weeks and in the drug studies, for 4-5 weeks. The 8 and 12 week cultures were reseeded at 4 weeks. The membranes were implanted into the peritoneal cavities of mice for 3-12 months after which they were removed, fixed, sectioned and stained for histologic study. After 6 months of implantation, about 40% of the membranes coated with cells from non-drug-treated mice and 60% of the membranes coated with cells from drug-treated mice contained hematopoietic elements; often there were foci of trilineal hematopoiesis. Hematopoiesis never occurred without bone formation, but the reverse was not true. Membranes coated with adherent layers established from marrow of mice treated with cyclophosphamide or BCNU showed two main characteristics: 1) they supported hematopoiesis normally, and 2) the regeneration of stroma and hematopoiesis occurred earlier than in membranes coated with stroma derived from normal mice, perhaps because the cells from the drug-treated mice spent a shorter time in culture. In vitro culture may damage cells required to condition the membrane for hematopoiesis.