Streptococcus equi Infections in Horses: Guidelines for Treatment, Control, and Prevention of Strangles-Revised Consensus Statement.

This consensus statement update reflects our current published knowledge and opinion about clinical signs, pathogenesis, epidemiology, treatment, complications, and control of strangles. This updated statement emphasizes varying presentations in the context of existing underlying immunity and carrier states of strangles in the transmission of disease. The statement redefines the "gold standard" for detection of possible infection and reviews the new technologies available in polymerase chain reaction diagnosis and serology and their use in outbreak control and prevention. We reiterate the importance of judicious use of antibiotics in horses with strangles. This updated consensus statement reviews current vaccine technology and the importance of linking vaccination with currently advocated disease control and prevention programs to facilitate the eradication of endemic infections while safely maintaining herd immunity. Differentiation between immune responses to primary and repeated exposure of subclinically infected animals and responses induced by vaccination is also addressed.

Biotypes and ScM types of isolates of Streptococcus canis from diseased and healthy cats.

Lancefield group G Streptococcus canis is a component of the normal urogenital and pharyngeal flora of the cat. It is also frequently implicated in epizootics of severe disease in closed cat colonies and animal shelters. Given the importance of S canis as a feline pathogen and relative lack of published information on characteristics potentially associated with virulence, the authors have compared isolates from healthy and diseased cats in New York and California using fermentation profiles (biotype) and ScM sequences. With few exceptions, isolates associated with disease were biotype 1. Four alleles of scm were identified of which type 1 dominated in diseased cats. Type 4 allelic variants were found only in healthy cats and all but one were biotype 2. Type 2 and 3 alleles showed extensive N-terminal variation suggesting a plasminogen-binding site as found on the type 1 allele was absent. Cat antisera to ScM were opsonobactericidal, and these potentially protective antibodies increased during convalescence.

Preliminary evaluation of a dual antigen ELISA to distinguish vaccinated from Leptospira infected horses.

Immunogenic proteins of Leptospira interrogans serovar Pomona type kennewicki (Lk) including Sph1, LigA, Hsp15 and LipL45 (Qlp42) are up-regulated in infected horses but are undetectable or expressed in trace amounts on cultured organisms. In contrast, LipL32 is abundant on cultured Lk and elicits infection antibody responses. The aim of this study was to develop an ELISA based on LipL32 or Lk sonicate and host-induced proteins to differentiate vaccine from infection serum antibody. IgG specific for recombinant Sph1, LigA, Lk90 (LigA; 379-1225 a.a), Hsp15, LipL45 and LipL32 of Lk were assayed in sera of horses infected naturally with Lk and before and after immunisation with serovar Pomona bacterin. Infection but not vaccine sera reacted strongly with Sph1, LigA and Lk90. LipL45 and Hsp15 reacted moderately with infection sera and weakly with vaccine sera. Lk sonicate and LipL32 reacted strongly with both infection and vaccine sera. As expected, culture-based vaccine failed to stimulate antibody to host-induced proteins. Therefore a dual antigen ELISA based on Lk sonicate or LipL32 combined with host-induced Sph1 and Lk90 will be valuable in differentiating infection from vaccine responses.

Multiple specificities of immunoglobulin M in equine fetuses infected with Leptospira interrogans indicate a competent immune response.

REASONS FOR PERFORMING STUDY: Foals of mares infected with Leptospira interrogans serovar Pomona type kennewicki (Lk) may be aborted/stillborn or delivered as healthy foals. Is fetal survival explained in part by the immune response of the fetus to Leptospira antigens? OBJECTIVES: To describe an outbreak of Leptospira abortion in which infected mares delivered dead/sick or normal foals and determine specificities of antibody in a collection of 54 fetuses from similar outbreaks. STUDY DESIGN: Outbreak investigation in combination with a case-control study of a larger set of samples from aborted fetuses. METHODS: Serology and polymerase chain reaction (PCR) on urine and amniotic fluids were used to diagnose infection during an outbreak of Leptospira abortion. Specificities of immunoglobulin (Ig)M, IgGa and IgGb for recombinant proteins of Lk in archived fluids of fetuses from similar outbreaks were compared by ELISA with those of fluids of fetuses not infected with Leptospira spp. RESULTS: Five fetuses of 11 infected mares in an outbreak survived in utero in the presence of persistent placental infection and were healthy at foaling. Fetuses of 6 mares in the outbreak were aborted or died soon after birth. Significantly greater (P<0.05) IgM reactivity with all recombinant proteins and with Lk sonicate was observed in 54 archived fluids from Leptospira infected fetuses than in fluids of 30 of non-Leptospira infected fetuses. Low levels of IgGa and IgGb specific for LipL32 and Lk sonicate and traces of LigA and Hsp15 specific IgGa were detected in a minority of archived fluids from Leptospira infected fetuses. CONCLUSION: Although mainly mediated by IgM, a high level of immune competence in aborted fetuses was evidenced by the multiplicity of Leptospira proteins targeted. This is likely to contribute to survival of foals in mares with evidence of placental infection at foaling as detailed in a typical outbreak.

Capsular hyaluronic acid of equine isolates of Streptococcus zooepidemicus is upregulated at temperatures below 35°C.

REASONS FOR PERFORMING STUDY: Streptococcus zooepidemicus causes opportunist respiratory and other infections in the horse. Capsule expression is highly variable and known to affect resistance to phagocytosis. Most clinical isolates producing small, dry colonies at 37°C produce mucoid colonies at temperatures below 35°C. OBJECTIVES: The aim was to understand the molecular basis of increased capsule expression by equine isolates of S. zooepidemicus at temperatures lower than 35°C. STUDY DESIGN: Cross-sectional observational study. METHODS: Capsule production by groups of equine S. zooepidemicus strains was determined at 23, 30, 35 and 37°C. Hyaluronidase (HylC) at 23 and 37°C was measured by quantitative enzyme-linked immunosorbent assay. Expressions of hasA and hylC at 23 and 37°C were measured by quantitative reverse transcriptase-PCR. The covRS genes in representative S. zooepidemicus were sequenced and checked for mutations. RESULTS: Colonies of randomly selected S. zooepidemicus strains became mucoid or showed marked increase in colony mucoidy following the change in temperature to 23°C. Expression of hasA at 23°C was 45- to 700-fold greater than at 37°C. Transcription of hylC at 23°C was 2.5- to 200-fold greater than at 37°C, yet enzyme concentrations in cultures were significantly higher at 37°C (P<0.05), suggesting that production of HylC is regulated post transcriptionally. The covRS gene in S. zooepidemicus was not mutated as seen in isolates of Streptococcus pyogenes with increased capsule production at 25°C. CONCLUSIONS: Sensitivity of capsule expression to temperature above 35°C but not HylC by the general population of equine S. zooepidemicus indicates that capsule is not required for extended colonisation nor for opportunist invasion. Instead, capsule production at lower than body temperature may reflect adaptation to life on skin and mucosal surfaces, where hyaluronic acid contributes to adhesion and resistance to desiccation. Pathogenicity of S. zooepidemicus following opportunist invasion is possibly dependent on factors other than capsule that may be co-regulated with HylC. The Summary is available in Chinese - see Supporting information.

Immunisation of mares with binding domains of toxins A and B of Clostridium difficile elicits serum and colostral antibodies that block toxin binding.

REASONS FOR PERFORMING STUDY: Enterocolitis caused by Clostridium difficile (C. difficile) is a serious, sometimes fatal, disease of neonatal foals and older horses. Toxins A and B (TcdA and B) produced by C. difficile are important virulence factors. Immunisation of mares with receptor binding domains of toxins may prevent or reduce the severity of C. difficile colitis in foals. OBJECTIVES: To determine whether antibodies generated in the pregnant mare to the binding regions of TcdA and B will neutralise TcdA and B toxicity. METHODS: Sequences encoding the binding domains of each toxin were isolated by PCR amplification from C. difficile JF09, a foal isolate, and cloned and expressed into pET15b. Thirteen mares were immunised twice 2 weeks apart with 200 µg of each recombinant protein with Quil A 2 months prior to foaling. Antibodies were assayed in the sera and colostrum by ELISA and for ability to block the cytopathic activity of each of toxin for equine endothelial cells. RESULTS: All mares produced strong serum antibody responses to the binding domain of each toxin. A high level of toxin-specific antibodies was also detected in colostrum and in most foal sera 2 days after suckling. Diluted sera and colostrum premixed with either TcdA or B had no effect on the morphology of equine endothelial cells. Application of the same concentration of toxins alone or premixed with nonimmune mare/foal serum or colostrum led to an unambiguous cytopathic effect that ranged from complete degradation to varying degrees of cell rounding. CONCLUSIONS: Immunisation of pregnant mares with recombinant binding domains of TcdA and B of C. difficile resulted in the production of specific antibodies in serum and colostrum that blocked the cytopathic activity of toxins. POTENTIAL RELEVANCE: Results of studies support the feasibility of a prepartum vaccine against C. difficile enterocolitis in foals.

Surface-associated Hsp60 chaperonin of Leptospira interrogans serovar Autumnalis N2 strain as an immunoreactive protein.

The present study has been formulated in order to detect an immunoreactive protein whose identification can play a major role in the early diagnosis of disease. The identified protein will be produced by recombinant methods and used for the recombinant protein based ELISA. A comparison was made between the developed method and the gold standard MAT test to evaluate the serodiagnosis potential of the protein. The protein profile, immunoblot and MALDI-TOF analysis was carried out to identify the immunoreactive protein. The immunoreactive protein identified was used to develop ELISA for the diagnosis of leptospirosis using patients' sera with various clinical manifestations. The immunoreactive protein was identified as Leptospira GroEL chaperonin of molecular weight 60 kDa. The theoretical/experimental molecular weights, pI were found to be 58.5/60 kDa and 5.41/6, respectively. The overall results of the recombinant GroEL-IgM ELISAs showed cumulative sensitivity, specificity, positive predictive value, and negative predictive values of 90.6%, 94.9%, 94.6%, and 91.0%, respectively. The performance of such ELISA appeared better than that of any other serological tests previously evaluated for the diagnosis of leptospirosis in India. Thus, a highly conserved and immunogenic outer exposed GroEL protein during infection clearly merits further use in the serodiagnosis of leptospirosis.

A clonal outbreak of acute fatal hemorrhagic pneumonia in intensively housed (shelter) dogs caused by Streptococcus equi subsp. zooepidemicus.

An outbreak of acute, fatal, hemorrhagic pneumonia was observed in more than 1,000 mixed breed dogs in a single animal shelter. The Department of Anatomic Pathology at the University of California at Davis School of Veterinary Medicine performed necropsies on dogs that were found moribund in acute respiratory distress or found dead with evidence of nasal bleeding. All dogs had hemothorax and an acute, fibrinosuppurative pneumonia. Large numbers of gram-positive cocci were observed within the lungs of all dogs and within septic thromboemboli of remote organs in about 50% of cases. Bacterial cultures from the dogs and their environment revealed widespread beta-hemolytic Streptococus equi subspecies zooepidemicus (Lancefield Group C). Extensive diagnostic testing failed to reveal the consistent presence of copathogens in individual cases. The clinical, epidemiologic, molecular biologic, and pathologic data indicate that a single clone of S. zooepidemicus was the cause of an acutely fatal respiratory infection in these dogs.

Early pathogenesis of equine Streptococcus equi infection (strangles).

REASONS FOR PERFORMING STUDY: Little is known about entry and subsequent multiplication of Streptococcus equi following exposure of a susceptible horse. This information would have value in design of intranasal vaccines and understanding of shedding and protective immune responses. OBJECTIVES: To determine entry points and sites of subsequent replication and dispersion of S. equi at different times after intranasal infection or commingling exposure. METHODS: Previously unexposed horses and ponies were subjected to euthanasia 1, 3, 20 or 48 h following intranasal inoculation with biotin labelled or unlabelled S. equi CF32. Some ponies were inoculated with suspensions of equal numbers of CF32 and its mutants lacking capsule, S. equi M-like protein or streptolysin S. Others were infected by commingling exposure and subjected to euthanasia after onset of fever. Tonsils and lymph nodes were cultured for S. equi and tissues sectioned for histopathological examination and fluorescent microscopy. RESULTS: Tonsillar tissues of both the oro- and nasopharynx served as portals of entry. Entry was unexpectedly rapid but involved few bacteria. Small numbers of organisms were detected in tonsillar crypts, in adjacent subepithelial follicular tissue and draining lymph nodes 3 h after inoculation. By 48 h, clumps of S. equi were visible in the lamina propria. At onset of fever, tonsillar tissues and one or more mandibular and retropharyngeal lymph nodes were heavily infiltrated by neutrophils and long chains of extracellular S. equi. Mutant S. equi lacking virulence factors were not seen in draining lymph nodes. CONCLUSIONS: Although very small numbers of S. equi entered the lingual and nasopharyngeal tonsils, carriage to regional lymph nodes occurred within hours of inoculation. This observation, together with visual evidence of intracellular and extracellular multiplication of S. equi in tonsillar lymphoid tissue and lymph nodes over the following days, indicates involvement of potent antiphagocytic activity and failure of innate immune defences. RELEVANCE: Future research should logically address the tonsillar immune mechanisms involved including identification of effector cell(s) and antigens.

Detection of equine herpesvirus-specific effector and memory cytotoxic immunity in the equine upper respiratory tract.

Immunological protection of horses from equine herpesvirus-1 (EHV-1) infection and disease depends on the cooperation of virus-specific humoral and cellular immune responses. EHV-specific mucosal immunity may be an important component of such immune responses. This study demonstrates the induction of anti-EHV cytotoxic cellular immune responses in various mucosal and systemic lymphoid tissues associated with the upper respiratory tract (URT) of the horse. Four young horses (1-2 years of age) were inoculated intranasally with the Army 183 strain of EHV-1 and euthanized 1 week later. One untreated foal served as a non-infected control. Mucosa-associated tonsillar tissues, draining lymph nodes and PBMC were harvested. Virus-specific memory and effector cytolytic activity were individually assessed using 4 h chromium release assays, with and without in vitro restimulation with EHV-1, respectively. EHV-specific cytotoxic activity was detected ex vivo in several URT-associated mucosal lymphoid tissues of horses, particularly within the lining of the nasopharynx, a principal site of EHV-1 replication. This activity was also detected in the circulation of some horses 1 week post-challenge. Virus-specific memory cytotoxic activity was elevated in the circulation, and detectable in the draining lymph nodes of all horses following challenge infection.

Histology, immunohistochemistry and ultrastructure of the tonsil of the soft palate of the horse.

The tonsil of the soft palate was an oval, flat structure located centro-rostrally on the oral surface of the soft palate. Its stratified squamous non-keratinized epithelium was perforated by holes or small crypts the deeper parts of which were loosely spongiform inter-digitated with lymphoid tissue. These unusual features have not previously been reported in tonsils of any species. Crypts and reticulated epithelium as found in the lingual and palatine tonsils were not observed. Lectins showed varying affinities for specific layers of the epithelium. M cells were not observed. A few Langerhans cells were distributed among surface epithelial cells. Lymphoid tissue was arranged loosely and in isolated lymphoid follicles in the subepithelial lamina propria mucosae. Although IgA+ cells and macrophages were proportionately more numerous the amount of lymphoid tissue was much less than in the lingual and palatine tonsils. Most of the follicular germinal centres lacked a darkly stained corona. CD4 positive were more numerous than CD8+ lymphocytes and were distributed in the parafollicular and inter-follicular areas. Large clusters of mucus acini positive for glycogen, acidic and neutral mucopolysaccharides separated lymphoid tissue from deeply placed striated muscle. Only a few high endothelial venules were observed in the parafollicular and inter-follicular areas. These had relatively few vesiculo vacuolar or other organelles in their high endothelial cells and few lymphocytes attaching to their walls.

Histology, immunohistochemistry and ultrastructure of the equine tubal tonsil.

The tubal tonsil of the horse surrounds the pharyngeal opening of the eustachian tube and is lined by pseudostratified columnar ciliated epithelium interspersed with areas of follicle-associated epithelium (FAE) heavily infiltrated by lymphocytes but devoid of goblet and ciliated cells. Scanning and transmission electron microscopy revealed microvillous cells and cells with features characteristic of M cells such as reduced microvilli or depressed bare surface, more numerous mitochondria, small vesicles and lysosomes, as well as vimentin filaments and epitopes specific for GS 1-B4 as previously seen in the nasopharyngeal tonsil. M cells were also identified in areas of respiratory epithelium not associated with lymphoid follicles and appeared to be the nasal mucosal counterparts of recently described intestinal villous M cells in the mouse. The underlying lymphoid tissue of the FAE was generally organized as solitary lymphoid follicles without germinal centres in contrast to the diffuse and large amount of organized lymphoid follicles with germinal centres that characterize the nasopharyngeal tonsil. CD8+ T and B-lymphocytes were much fewer than in the nasopharyngeal tonsil. High endothelial venules were mainly oriented towards the parafollicular area and contained much fewer endothelial pores and vesiculo-vacuolar organelles. Finally, scattered small clusters of mucus acini and striated muscles were other features that differentiated the tubal and nasopharyngeal tonsils.

Histology, immunohistochemistry and ultrastructure of the equine palatine tonsil.

The palatine tonsils of five young horses formed 10-12 cm elongated follicular structures extending from the root of the tongue on either side to the base of the epiglottis and lateral to the glossoepiglottic fold. The stratified squamous non-keratinized epithelium of the outer surface was modified into crypts as reticular epithelium by heavy infiltration of lymphoid cells from underlying lymphoid follicles. In places, lymphoid tissue reaching almost to the surface and with only one to two cell layers intact was identified as the lymphoepithelium. Langerhans cells with Birbeck granules were interspersed between epithelial cells. Lymphoid tissue organized in lymphoid follicles constituted the parenchyma of the palatine tonsil. CD4-positive cells were more numerous and CD8-positive lymphocytes less numerous compared with their distribution in the lingual tonsil. B cells and macrophages were also more numerous than in the lingual tonsil and lectins showed a different pattern of attachment. M cells were not observed. High endothelial venules with well-developed vesiculo-vacuolar organelle had structural evidence of transendothelial and interendothelial migration of lymphocytes. Striated muscles as seen in the deeper lamina propria mucosae of the lingual tonsil were absent. The immunohistological and ultrastructural characteristics of the equine palatine tonsil are similar to those of humans but differ from those of the lingual tonsil and are consistent with a role as an effector and inductor immunological organ.

Histology and ultrastructure of the equine lingual tonsil. II. Lymphoid tissue and associated high endothelial venules.

The stratified squamous epithelium of the lingual tonsil of five young horses was infiltrated with CD4 and CD8 positive cells, which were very numerous in the crypt reticular epithelium along with macrophages and IgGb and IgA positive cells. Lymphoid follicles of the lamina propria mucosae consisted of a parafollicular area, corona and germinal centre. The parafollicular area was populated by large numbers of CD4 and CD8 positive lymphocytes as well as macrophages, inter-digitating cells, and a few B-lymphocytes. The germinal centre contained mainly IgGb and IgG(T) positive cells, plasma cells and small numbers of follicular dendritic cells, macrophages, CD4, CD8 and IgA positive cells. Some venules in the parafollicular and inter-follicular areas had features characteristic of high endothelial venules with irregular nuclei and cytoplasmic processes on the luminal surface. In addition to the normal cytoplasmic organelles, a novel vesiculo-vacuolar organelle was observed in small clusters toward the lateral, luminal and abluminal surfaces of these high endothelial venules. These vesiculo organelles along with their stomata and diaphragms, communicated with each other and with inter-endothelial clefts forming a structural basis for enhanced permeability and extravasation of macromolecules. The presence of lymphocytes in the high endothelial venules is consistent with transmural and inter-endothelial passage of these cells.

Histology and ultrastructure of the equine lingual tonsil. I. Crypt epithelium and associated structures.

The microstructural and ultrastructural features of the equine lingual tonsil were studied in five young horses. Located at the root of the tongue it presented an irregular surface with rounded elevations, numerous folds and crypts. Stratified squamous non-keratinized epithelium lining its outer surface was modified by heavy infiltration of lymphoid cells to form reticular epithelium within the crypt. The latter implies a role in initiating and maintaining immune responses to incoming infectious agents and antigens. Lamellated structures resembling Hassall's corpuscle were observed towards the outer surface epithelium. Microplicae were visible by scanning electron microscopy on the surface of both the outer and reticular epithelia. No microvillus cells resembling M cells were observed. The stratum superficiale of the reticular epithelium showed strong affinity for Soybean (SBA), Phosphocarpus tetragonolobus 1 (WBA 1), Ulex europaeus (UEA) and Griffonia simplicifolia 1 isolectin-B4 (GS1-B4). The characteristic lectin binding patterns may be useful for embryological and microbiological investigations. Vimentin filaments were not detected consistent with absence of M cells. Mucus glandular acini in the deeper lamina propria mucosae contained glycogen, acidic, neutral and weakly sulphated mucopolysaccharides. Transmission electron microscopy revealed that the layers of the outer surface and reticular epithelia shared characteristic features except the stratum superficiale, which had nuclei of varying shapes and an abundance of cell organelles. A few mast cells with electron lucent granules and myelinated nerve fibres were localized in the deeper portion.

Host-inducible immunogenic sphingomyelinase-like protein, Lk73.5, of Leptospira interrogans.

Leptospira interrogans causes a variety of clinical syndromes in animals and humans. Although much information has accumulated on the importance of leptospiral lipopolysaccharide in protective antibody responses, relatively little is known about proteins that participate in immune responses. Identification of those proteins induced only in the host is particularly difficult. Using a novel double-antibody screen designed to identify clones in a gene library of L. interrogans serovar Pomona expressing host-inducible proteins, we have characterized a gene (lk75.3) encoding a sphingomyelinase-like preprotein of 648 amino acids with cytotoxic activity for equine pulmonary endothelial cells and weak hemolytic activity for equine and rabbit erythrocytes. lk73.5 was found as a single gene copy in all serovars of L. interrogans but not in other Leptospira spp. except L. inadai. The open reading frame (ORF) for Lk73.5 is followed by another partially homologous sequence containing an ORF (sph-like 2) for a 28.7-kDa peptide. Lk73.5 and Sph-like 2 share 95.1 and 97.7% amino acid identity with putative sphingomyelinases Sph2 and Sph1 (N terminus) from L. interrogans serovar Lai (S.-X. Ren, G. Fu, X.-G. Jiangk, R. Zeng, Y.-G. Miao, H. Xu, Y.-X. Zhang, H. Xiong, G. Lu, L.-F. Lu, H.-Q. Jiang, J. Jia, Y.-F. Tu, J.-X. Jiang, W.-Y. Gu, Y.-Q. Zhang, Z. Cai, H.-H. Sheng, H.-F. Yin, Y. Zhang, G.-F. Zhu, M. Wank, H.-L. Huangk, Z. Qian, S.-Y. Wang, Wei Ma, Z.-J. Yao, Y. Shen, B.-Q. Qiang, Q.-C. Xia, X.-K. Guo, A. Danchinq, I. S. Girons, R. L. Somerville, Y.-M. Wen, M.-H. Shik, Z. Chen, J.-G. Xuk, and G.-P. Zhao, Nature 422:88-893, 2003). Substantial homologies to sphingomyelinases from other leptospiras and other bacteria are also present. Lk73.5 was not detected in leptospiras cultured at 30 or 37 degrees C. The recombinant protein reacted strongly with sera from recently infected mares but not with sera from horses vaccinated with commercial pentavalent bacterin. The host-inducible immunogenic Lk73.5 should have value in distinguishing vaccine from infection immune response.

Safe and cost effective use of alteplase for the clearance of occluded central venous access devices.

PURPOSE: To determine whether cryopreserved solutions of the thrombolytic agent alteplase could be used as a safe, effective, and economically reasonable alternative to urokinase in patients presenting with occluded central venous access devices (CVADs). MATERIALS AND METHODS: Alteplase has been reported as an efficacious alternative to urokinase for treatment of occluded CVADs. However, the practicality of using alteplase as the thrombolytic of choice for this indication remained conjectural. To make this approach economically feasible, alteplase was diluted to 1 mg/mL and 2.5-mL aliquots were stored at -20 degrees C until use. A need to confirm that the cryopreserving and thawing of the reconstituted solution did not compromise the safety and efficacy reported from prior trials was recognized. A quality assessment initiative was undertaken to concurrently monitor the safety and efficacy of this approach. Patients presenting with occluded CVADs received a sufficient volume of the thawed alteplase solution to fill the occluded catheter(s). Data, including efficacy, adverse reactions, dwell time, and catheter type, were collected over a 5-month period. RESULTS: One hundred twenty-one patients accounting for 168 attempted clearances were assessable for safety and efficacy. One hundred thirty-six (81%) of the 168 catheter clearance attempts resulted in successful catheter clearance (95% confidence interval, 74% to 86%). No adverse events were reported. CONCLUSION: Cryopreserved 1-mg/mL aliquots of alteplase are safe and effective in the clearance of occluded CVADs when stored at -20 degrees C for 30 days. The ability to cryopreserve alteplase aliquots makes it an economically reasonable alternative to urokinase in the setting of CVAD occlusion.