JOURNAL OF CLINICAL MICROBIOLOGY, OCt. 1992, p. 2753-2755 0095-1137/92/102753-03$02.00/0
Vol. 30, No. 10
Letters to the Editor Enhanced Detection of Cytomegalovirus in MRC-5 Shell Vials West and coworkers (6) reported differences in the sensitivities of MRC-5 monolayers to cytomegalovirus (CMV) among different suppliers (i.e., Whittaker Bioproducts, Bartels, and ViroMed). The authors further suggested that pretreatment of shell vials with the (glucocorticoid) dexamethasone plus dimethyl sulfoxide (DEX-DMSO) and DEXDMSO plus Ca enhanced CMV recovery and that this effect was most pronounced in shell vials supplied by Whittaker. On the basis of recent studies by this worker and colleagues (4, 5), some important questions in experimental design should be brought to the attention of the readership of the Journal of Clinical Microbiology before the conclusions by the SmithKline investigators are drawn. Firstly, and perhaps most importantly, studies by Leonardi and Lipson (4) and Lipson et al. (5) using laboratoryadapted and wild CMV strains as well as freshly collected peripheral blood demonstrated that MRC-5 monolayer sensitivity may be enhanced following a medium enrichment effect. In our studies specifically, monolayer sensitivity was enhanced and CMV isolation was increased in shell vials pretreated with Eagle minimal essential medium supplemented with 10% heat-inactivated fetal bovine serum (EMEM-10% FBS) compared with EMEM-2% FBS. Regarding the West et al. (6) study, I cannot help but question whether refeeding with an enriched culture medium, that is, EMEM-10% FBS (toxic cellular metabolites would be removed from the system by this step as well), might have increased the sensitivity and overall recovery rate of CMV in the untreated shell vials to the levels obtained by using monolayers pretreated with EMEM-2% FBS plus DEXDMSO or DEX-DMSO-Ca. According to our data, for example, no significant differences were identified in the rates of CMV recovery from blood specimens which were inoculated into shell vials pretreated with either DMSO or EMEM-10% FBS (5). A second point arises with the fact that most diagnostic virology laboratories do not recommend that potentially CMV-positive clinical specimens remain in storage (at a refrigeration temperature [i.e., 4°C]) longer than from 24 to 48 h. Ideally, such specimens should be received and processed by the laboratory within hours of collection (1, 5). The testing of 1- to 2-week-old preselected positive clinical specimens (the authors claim that "the number of [shell] vials available from each supplier was limited"), although extremely convenient for such comparative studies, may not necessarily reflect that which may take place when freshly collected clinical specimens are tested. Although the data of West et al. (6) imply differences in the sensitivities of MRC-5 cultures supplied by different manufacturers, the development of artifacts (possibly due to, e.g., the formation of particulates-precipitates or changes in pH) in the 1- to 2-week-old clinical specimens cannot be ignored as possible confounding factors which might influence interpretation of the data. In summary, a valid protocol to investigate the issues raised by West and colleagues may be appropriately addressed by the performance of comparative testing utilizing different concentrations of FBS (i.e., at the very least, the
routinely used EMEM-2% FBS and an enriched culture medium containing 10% FBS) and DEX, as well as the screening of fieshly collected clinical specimens. We are in agreement with others (2, 3) that additional testing utilizing clinical specimens secured from varied sources is needed to substantiate the recommended use of dexamethasone as an enhancement agent of CMV infectivity. REFERENCES 1. Hodinka, R. L., and H. M. Friedman. 1991. Human cytomegalovirus, p. 829-837. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 2. Leland, D. S. 1992. Concepts of clinical diagnostic virology, p. 3-43. In E. H. Lennette (ed.), Laboratory diagnosis of viral infections. Marcel Dekker, Inc., New York. 3. Leannette, D. A. 1991. Preparation of specimens for virological examination, p. 818-821. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 4. Leonardi, G. P., and S. M. Lipson. 1992. Enhanced detection of cytomegalovirus in shell vial culture following MRC-5 monolayer pretreatment with glucocorticoids. Zentralbl. Bakteriol. 277:9099. 5. Lipson, S. M., M. H. Kaplan, J. K. Simon, Z. Ciamician, and L.-F. Tseng. 1992. Improved detection of cytomegalovirus viremia in AIDS patients using shell vial and indirect immunoperoxidase methodologies. J. Med. Virol. 38:36-43. 6. West, P. G., R. A. Hartwig, and W. W. Baker. 1992. Comparison of sensitivities of three commercial MRC-5 cell lines grown in shell vials to cytomegalovirus and responses to enhancing agents. J. Clin. Microbiol. 30:557-560. Steven M. Lipson Diagnostic Vrology Laboratory Division of Infectious Diseases Department of Medicine North Shore University HospitalCornell University Medical College Manhasset, New York 11030
Author's Reply In reply to Dr. Lipson's letter concerning our work on enhanced cytomegalovirus (CMV) detection with MRC-5 cells, I will try to address his objections in the order that he set them down in his letter. He states that he has some important questions about the experimental design of our last paper (5). He describes his own studies and compares his results with our results. I fail to see what any of this has to do with our experimental design. We have found that our cell treatments consistently enhance CMV detection (2, 4, 5). I do not dispute the fact that Dr. Lipson's method (or anyone else's) may also enhance the recovery of this virus, but our papers have dealt with our own enhancement procedure, and our experiments have been built around this procedure alone. Secondly, Dr. Lipson objects to our use of 1- to 3-weekold positive specimens for our experiments. If he has read out first two papers dealing with dexamethasone (DEX)2753
2754
LETTERS TO THE EDITOR
dimethyl sulfoxide (DMSO) enhancement of CMV and herpes simplex virus, he will be aware that both of these studies were done entirely with fresh specimens (2, 3). We saw absolutely no difference in the appearance of CMV inclusions with these fresh specimens and that of CMV inclusions with our older specimens and no evidence of false positives caused by "artifacts" or "particulates-precipitates" with our monoclonal anti-CMV antibodies. Neither have we found any accounts of such false positives in the literature or heard of them from colleagues. What we have seen, rather, is a loss in viability of the virus with storage, resulting in specimens becoming negative. We took advantage of this fact to push our system to the limit and to create a more rigorous test of sensitivity by including a larger percentage of these borderline positives. Another reason for using previously determined positives was the expense of shell vials. We did not wish to waste large numbers of them on negative specimens which would not figure into our data. Another point worth touching on is the fact that methods of virus recovery may vary considerably with the type of specimen involved. Judging from the titles of Dr. Lipson's papers and abstract (1), two out ot three of these articles deal with CMV in blood specimens. None of our studies have
included blood specimens. Along the same lines, Dr. Lipson seems to question the efficacy of DEX (we use DEX not alone but in conjunction with DMSO and calcium) as an enhancing agent, yet the title of one of his papers is "Enhanced Detection of Cytomegalovirus in Shell Vial Culture following MRC-5 Monolayer Pretreatment with Glucocorticoids." Unfortunately, since his papers are not yet published, it is
J. CLIN. MICROBIOL.
impossible for me to further compare our work with Dr. Lipson's until his data and methodologies become available. REFERENCES 1. Lipson, S. M., et al. 1992. Rapid detection of cytomegalovirus viremia by shell vial and indirect immunoperoxidase methodologies, abstr. C-150, p. 445. Abstr. 92nd Annu. Meet. Am. Soc. Microbiol. 1992. American Society for Microbiology, Washington, D.C. 2. West, P. G., B. Aldrich, R. Hartwig, and G. J. Haller. 1988. Enhanced detection of cytomegalovirus in confluent MRC-5 cells treated with dexamethasone and dimethyl sulfoxide. J. Clin. Microbiol. 26:2510-2514. 3. West, P. G., B. Aldrich, R. Hartwig, and G. J. Haller. 1989. Increased detection of herpes simplex virus in MRC-5 cells treated with dimethyl sulfoxide and dexamethasone. J. Clin. Microbiol. 27:770-772. 4. West, P. G., and W. W. Baker. 1990. Enhancement by calcium of the detection of cytomegalovirus in cells treated with dexamethasone and dimethyl sulfoxide. J. Clin. Microbiol. 28:17081710. 5. West, P. G., R. A. Hartwig, and W. W. Baker. 1992. Comparison of sensitivities of three commercial MRC-5 cell lines grown in shell vials to cytomegalovirus and responses to enhancing agents. J. Clin. Microbiol. 30:557-560. Patricia G. West
SmithKline Beecham Clinical Laboratories 400 Egypt Road Nornstown, Pennsylvania 19403
Evaluation of Manufacturer's Recommended Growth Value Thresholds for BACTEC Media McGowan and Metchock recently reported that individual laboratory evaluation of manufacturer's recommended growth value thresholds (GVTs) for BACTEC media might have an impact on detection of positive blood cultures and assessment of laboratory efficiency in this area (2). We would add that any assessment of optimal GVTs should consider the impact of the timing of test results on patient therapy and clinical outcome. We recently assessed these issues with BACTEC PEDS Plus (PP) medium. PP medium is used with the BACTEC NR 660 system (Becton Dickinson Diagnostic Instruments, Sparks, Md.) to diagnose bloodstream infection in children. PP medium contains less broth and less sodium polyanetholesulfonate than other BACTEC media and contains resins, and comparative clinical studies have generally shown increased microbial recovery with reduced detection time (1, 3, 5). When PP medium is used, the manufacturer recommends a GVT of 20 be used to denote positivity for vials contain ing 1 to 2 ml of blood and a lower GVT be used for vials with less blood. Because the volume of blood submitted from children, particularly neonates, may be less than 1 ml (1, 4), we evaluated the benefits of lowering the GVT of the first laboratory run of blood cultures submitted in PP medium on work load, yield of pathogens, and impact on therapy. From October 1990 through August 1991 all specimens received in PP medium were classified as PR or PN (Peds routine or neonate), the latter if they were from neonates or if
there was visibly little blood. Specimens arriving before 6:00 a.m. were sampled twice on day 1; those arriving from 6:00 to 9:30 a.m. were sampled once. All vials were sampled twice on day 2 and once on days 3, 5, and 7. For vials run twice on day 1, the initial GVT was set at 10 for PN specimens and 15 for PR specimens; all subsequent GVTs were set at 20. During the study, 9,458 bottles were processed; 8,409 (8,037 PR and 372 PN) were sampled twice on day 1. Of these latter bottles, 167 (2%) were positive with an initial GVT of
Vol. 30, No. 10
Letters to the Editor Enhanced Detection of Cytomegalovirus in MRC-5 Shell Vials West and coworkers (6) reported differences in the sensitivities of MRC-5 monolayers to cytomegalovirus (CMV) among different suppliers (i.e., Whittaker Bioproducts, Bartels, and ViroMed). The authors further suggested that pretreatment of shell vials with the (glucocorticoid) dexamethasone plus dimethyl sulfoxide (DEX-DMSO) and DEXDMSO plus Ca enhanced CMV recovery and that this effect was most pronounced in shell vials supplied by Whittaker. On the basis of recent studies by this worker and colleagues (4, 5), some important questions in experimental design should be brought to the attention of the readership of the Journal of Clinical Microbiology before the conclusions by the SmithKline investigators are drawn. Firstly, and perhaps most importantly, studies by Leonardi and Lipson (4) and Lipson et al. (5) using laboratoryadapted and wild CMV strains as well as freshly collected peripheral blood demonstrated that MRC-5 monolayer sensitivity may be enhanced following a medium enrichment effect. In our studies specifically, monolayer sensitivity was enhanced and CMV isolation was increased in shell vials pretreated with Eagle minimal essential medium supplemented with 10% heat-inactivated fetal bovine serum (EMEM-10% FBS) compared with EMEM-2% FBS. Regarding the West et al. (6) study, I cannot help but question whether refeeding with an enriched culture medium, that is, EMEM-10% FBS (toxic cellular metabolites would be removed from the system by this step as well), might have increased the sensitivity and overall recovery rate of CMV in the untreated shell vials to the levels obtained by using monolayers pretreated with EMEM-2% FBS plus DEXDMSO or DEX-DMSO-Ca. According to our data, for example, no significant differences were identified in the rates of CMV recovery from blood specimens which were inoculated into shell vials pretreated with either DMSO or EMEM-10% FBS (5). A second point arises with the fact that most diagnostic virology laboratories do not recommend that potentially CMV-positive clinical specimens remain in storage (at a refrigeration temperature [i.e., 4°C]) longer than from 24 to 48 h. Ideally, such specimens should be received and processed by the laboratory within hours of collection (1, 5). The testing of 1- to 2-week-old preselected positive clinical specimens (the authors claim that "the number of [shell] vials available from each supplier was limited"), although extremely convenient for such comparative studies, may not necessarily reflect that which may take place when freshly collected clinical specimens are tested. Although the data of West et al. (6) imply differences in the sensitivities of MRC-5 cultures supplied by different manufacturers, the development of artifacts (possibly due to, e.g., the formation of particulates-precipitates or changes in pH) in the 1- to 2-week-old clinical specimens cannot be ignored as possible confounding factors which might influence interpretation of the data. In summary, a valid protocol to investigate the issues raised by West and colleagues may be appropriately addressed by the performance of comparative testing utilizing different concentrations of FBS (i.e., at the very least, the
routinely used EMEM-2% FBS and an enriched culture medium containing 10% FBS) and DEX, as well as the screening of fieshly collected clinical specimens. We are in agreement with others (2, 3) that additional testing utilizing clinical specimens secured from varied sources is needed to substantiate the recommended use of dexamethasone as an enhancement agent of CMV infectivity. REFERENCES 1. Hodinka, R. L., and H. M. Friedman. 1991. Human cytomegalovirus, p. 829-837. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 2. Leland, D. S. 1992. Concepts of clinical diagnostic virology, p. 3-43. In E. H. Lennette (ed.), Laboratory diagnosis of viral infections. Marcel Dekker, Inc., New York. 3. Leannette, D. A. 1991. Preparation of specimens for virological examination, p. 818-821. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 4. Leonardi, G. P., and S. M. Lipson. 1992. Enhanced detection of cytomegalovirus in shell vial culture following MRC-5 monolayer pretreatment with glucocorticoids. Zentralbl. Bakteriol. 277:9099. 5. Lipson, S. M., M. H. Kaplan, J. K. Simon, Z. Ciamician, and L.-F. Tseng. 1992. Improved detection of cytomegalovirus viremia in AIDS patients using shell vial and indirect immunoperoxidase methodologies. J. Med. Virol. 38:36-43. 6. West, P. G., R. A. Hartwig, and W. W. Baker. 1992. Comparison of sensitivities of three commercial MRC-5 cell lines grown in shell vials to cytomegalovirus and responses to enhancing agents. J. Clin. Microbiol. 30:557-560. Steven M. Lipson Diagnostic Vrology Laboratory Division of Infectious Diseases Department of Medicine North Shore University HospitalCornell University Medical College Manhasset, New York 11030
Author's Reply In reply to Dr. Lipson's letter concerning our work on enhanced cytomegalovirus (CMV) detection with MRC-5 cells, I will try to address his objections in the order that he set them down in his letter. He states that he has some important questions about the experimental design of our last paper (5). He describes his own studies and compares his results with our results. I fail to see what any of this has to do with our experimental design. We have found that our cell treatments consistently enhance CMV detection (2, 4, 5). I do not dispute the fact that Dr. Lipson's method (or anyone else's) may also enhance the recovery of this virus, but our papers have dealt with our own enhancement procedure, and our experiments have been built around this procedure alone. Secondly, Dr. Lipson objects to our use of 1- to 3-weekold positive specimens for our experiments. If he has read out first two papers dealing with dexamethasone (DEX)2753
2754
LETTERS TO THE EDITOR
dimethyl sulfoxide (DMSO) enhancement of CMV and herpes simplex virus, he will be aware that both of these studies were done entirely with fresh specimens (2, 3). We saw absolutely no difference in the appearance of CMV inclusions with these fresh specimens and that of CMV inclusions with our older specimens and no evidence of false positives caused by "artifacts" or "particulates-precipitates" with our monoclonal anti-CMV antibodies. Neither have we found any accounts of such false positives in the literature or heard of them from colleagues. What we have seen, rather, is a loss in viability of the virus with storage, resulting in specimens becoming negative. We took advantage of this fact to push our system to the limit and to create a more rigorous test of sensitivity by including a larger percentage of these borderline positives. Another reason for using previously determined positives was the expense of shell vials. We did not wish to waste large numbers of them on negative specimens which would not figure into our data. Another point worth touching on is the fact that methods of virus recovery may vary considerably with the type of specimen involved. Judging from the titles of Dr. Lipson's papers and abstract (1), two out ot three of these articles deal with CMV in blood specimens. None of our studies have
included blood specimens. Along the same lines, Dr. Lipson seems to question the efficacy of DEX (we use DEX not alone but in conjunction with DMSO and calcium) as an enhancing agent, yet the title of one of his papers is "Enhanced Detection of Cytomegalovirus in Shell Vial Culture following MRC-5 Monolayer Pretreatment with Glucocorticoids." Unfortunately, since his papers are not yet published, it is
J. CLIN. MICROBIOL.
impossible for me to further compare our work with Dr. Lipson's until his data and methodologies become available. REFERENCES 1. Lipson, S. M., et al. 1992. Rapid detection of cytomegalovirus viremia by shell vial and indirect immunoperoxidase methodologies, abstr. C-150, p. 445. Abstr. 92nd Annu. Meet. Am. Soc. Microbiol. 1992. American Society for Microbiology, Washington, D.C. 2. West, P. G., B. Aldrich, R. Hartwig, and G. J. Haller. 1988. Enhanced detection of cytomegalovirus in confluent MRC-5 cells treated with dexamethasone and dimethyl sulfoxide. J. Clin. Microbiol. 26:2510-2514. 3. West, P. G., B. Aldrich, R. Hartwig, and G. J. Haller. 1989. Increased detection of herpes simplex virus in MRC-5 cells treated with dimethyl sulfoxide and dexamethasone. J. Clin. Microbiol. 27:770-772. 4. West, P. G., and W. W. Baker. 1990. Enhancement by calcium of the detection of cytomegalovirus in cells treated with dexamethasone and dimethyl sulfoxide. J. Clin. Microbiol. 28:17081710. 5. West, P. G., R. A. Hartwig, and W. W. Baker. 1992. Comparison of sensitivities of three commercial MRC-5 cell lines grown in shell vials to cytomegalovirus and responses to enhancing agents. J. Clin. Microbiol. 30:557-560. Patricia G. West
SmithKline Beecham Clinical Laboratories 400 Egypt Road Nornstown, Pennsylvania 19403
Evaluation of Manufacturer's Recommended Growth Value Thresholds for BACTEC Media McGowan and Metchock recently reported that individual laboratory evaluation of manufacturer's recommended growth value thresholds (GVTs) for BACTEC media might have an impact on detection of positive blood cultures and assessment of laboratory efficiency in this area (2). We would add that any assessment of optimal GVTs should consider the impact of the timing of test results on patient therapy and clinical outcome. We recently assessed these issues with BACTEC PEDS Plus (PP) medium. PP medium is used with the BACTEC NR 660 system (Becton Dickinson Diagnostic Instruments, Sparks, Md.) to diagnose bloodstream infection in children. PP medium contains less broth and less sodium polyanetholesulfonate than other BACTEC media and contains resins, and comparative clinical studies have generally shown increased microbial recovery with reduced detection time (1, 3, 5). When PP medium is used, the manufacturer recommends a GVT of 20 be used to denote positivity for vials contain ing 1 to 2 ml of blood and a lower GVT be used for vials with less blood. Because the volume of blood submitted from children, particularly neonates, may be less than 1 ml (1, 4), we evaluated the benefits of lowering the GVT of the first laboratory run of blood cultures submitted in PP medium on work load, yield of pathogens, and impact on therapy. From October 1990 through August 1991 all specimens received in PP medium were classified as PR or PN (Peds routine or neonate), the latter if they were from neonates or if
there was visibly little blood. Specimens arriving before 6:00 a.m. were sampled twice on day 1; those arriving from 6:00 to 9:30 a.m. were sampled once. All vials were sampled twice on day 2 and once on days 3, 5, and 7. For vials run twice on day 1, the initial GVT was set at 10 for PN specimens and 15 for PR specimens; all subsequent GVTs were set at 20. During the study, 9,458 bottles were processed; 8,409 (8,037 PR and 372 PN) were sampled twice on day 1. Of these latter bottles, 167 (2%) were positive with an initial GVT of
