- Email: [email protected]
Empirical treatment of febrile, neutropenic patients with tobramycin and latamoxef
Journal
of Hospital
Empirical
Infection
(1987)
9, 278-284
treatment of febrile, neutropenic with tobramycin and latamoxef
E. G. Rhodes*,
patients
R. I. Harris?, R. S. Welch*, D. J. Perry?, Brown* and B. J. Boughton*?
R. M.
Departments of Haematology and Microbiology, *Queen Elizabeth Hospital, Birmingham, and jCoventry and Warwickshire Hospital, Coventry Accepted for publication
14 July
1986
Summary:
One hundred and two febrile episodes in neutropenic patients were treated with intravenous tobramycin and latamoxef. After 48 h latamoxef at 6 g day-‘, patients were randomized to continue this regimen or latamoxef at 3 g day-‘. Infections responded to these regimens in 67% and 71% of patients, respectively. Two-thirds of the infections which failed to respond were due to coagulase-negative staphylococci in Hickman catheters, a trend which may necessitate the inclusion of additional antibiotics in future empirical regimens. Prolonged prothrombin times due to antibiotic therapy were seen in nine patients but there was only one episode of bleeding and this responded quickly to treatment with vitamin K and fresh frozen plasma. In 35 patients, coagulopathy was present before antibiotics were started, and these cases also responded to vitamin K. The study shows that the response to tobramycin and latamoxef is comparable to other broad-spectrum antibiotic regimens and that a reduction in the dose of latamoxef after 48 h treatment may safely permit cost savings.
Introduction Any infection in neutropenic patients is potentially serious, and intravenous broad-spectrum antibiotics should be given without waiting for the result of microbiological cultures. Latamoxef is a new broad-spectrum, a-lactamase stable cephalosporin which we have assessed in combination with tobramycin in neutropenic patients. In addition, we have examined its adverse effects, particularly those relating to coagulation (Bang & Kammer, 1983). Patients and methods All patients were adults receiving intravenous chemotherapy for acute myeloblastic or lymphoblastic leukaemia. The criteria for trial entry were (a) temperature 3 38°C sustained for at least 2 h and not due to drug $ Correspondence 0195-67Ol/87/030278+07
to Dr B. J. Boughton. 0 1987 The Hospital
SO3.00/0
278
Infection
Society
Latamoxef
in neutropenic
patients
279
reactions or blood products, and (b) neutrophil count < 1.0 x 10” 1-l. Tobramycin 1.6 mg kg-’ g-hourly was given by bolus intravenous injection and subsequently adjusted according to serum levels. Latamoxef 2g g-hourly was given intravenously for 48 h, after which patients were randomized to receive either 1 g or 2 g g-hourly (continued low dose, CLD; continued high dose, CHD). Bacteriological cultures from relevant sites were performed before treatment and repeated twice weekly. Blood cultures were obtained via Hickman catheters when these were in situ. Bacterial sensitivities to tobramycin and latamoxef were measured using the modified Stokes disc diffusion method (Stokes & Ridgway, 1980). Prothrombin times were measured by the Quick one-stage method. The clinicians responsible for the patients had no knowledge of the randomized dose of latamoxef. Infective episodes were classified as (1) microbiologically documented, (2) clinically documented or (3) pyrexias of unknown origin (Gaya, 1984). In patients in whom fever settled and clinical signs of infection improved, antibiotics were usually continued for a further 3 days (Rodriguez et al., 1973). In treatment failures, additional antibiotics or granulocyte transfusions were utilized according to the clinical and bacteriological picture. An independent sequential analysis was conducted, with the intention to close the trial if significant differences emerged between the treatment groups. The study received local ethical committee approval. The significance of differences between values was assessed using Student’s t-test and Fisher’s exact probability test. Results
One hundred and two episodes of fever were seen in 70 patients. One further case was withdrawn before randomization because of Clostridium welchii septicaemia treated with additional penicillin G, and a second patient died before randomization. Forty-six cases were documented bacteriologically, 14 were documented by clinical criteria such as pneumonic shadowing on chest X-ray and 42 were pyrexias of unknown origin (PUO). Seventy episodes responded to treatment, 36/54 (67%) on CHD latamoxef, and 34/48 (71%) on the CLD (Table I). In the CLD patients who responded to treatment, 13/34 (38%) responded within 48 h whilst receiving the pre-randomization high dose treatment, and 21/34 (62%) responded after further treatment with CLD latamoxef. Response rates to CHD and CLD latamoxef in episodes subclassified by clinical and microbiological criteria were similar k2 = 0.08; P> 0.05 (Table I). There were 32 treatment failures of whom 18 received CHD and 14 received CLD latamoxef. A wide variety of micro-organisms were isolated, though the high incidence of coagulase-negative staphylococci and the low incidence of Pseudomonas spp. is unusual. Organisms resistant to tobramycin or latamoxef were seen in 15 and 17 episodes, respectively11 coagulase-negative staphylococci, four Streptococcus faecalis, one
E. G. Rhodes
280
et al.
Table I. Comparative details of 102 febrile neutropenic patients receiving tobramycin and continued high dose or continued Cow dose latamoxef Latamoxef continued high dose (6 g day-9
continued low dose (3 g day-7
Treatment successes (70) Microbiologically documented Clinically documented Pyrexia of unknown origin
14 4 18
16 5 13
Total
36
34
9 3 6
5 5
18
14
Treatment failures (32) Microbiologically documented Clinically documented Pyrexia of unknown origin Total
Staphylococcus aureus and one BaciZlus cereus (Table II). Sixteen treatment failures responded to additional vancomycin, four to ampicillin, four to amphotericin and three to granulocyte transfusions. Five patients died despite additional measures, three treated with CHD and two with CLD latamoxef. Three of these had PUOs, one had cytomegalovirus pneumonia and one patient treated with CHD latamoxef died of septicaemia due to a tobramycin/latamoxef-sensitive Escherichia coli. In the CHD and CLD latamoxef groups there were no significant differences in the degree of neutropenia and thrombocytopenia, nor in the duration of fever or the duration of treatment (P> 0.1, Table III). One patient developed a skin rash, which resolved on withdrawing the drugs, and diarrhoea not associated with Cl. di&ile toxin occurred in a further six cases. There was no antibiotic-induced hepatic or renal toxicity. The prothrombin time (PT) was measured prior to antibiotic treatment in 92 patients and in 35 it was prolonged 3-15 s. These received vitamin K and none showed further prolongation of the PT during antibiotic therapy. In 57 patients with normal coagulation on trial entry, no prophylactic vitamin K was given during the antibiotic chemotherapy and prolongation of the PT was seen in nine. Six of these received CHD and three received CLD latamoxef. Only one patient experienced clinically significant bleeding, where gastrointestinal haemorrhage after 8 days’ treatment with CLD latamoxef responded to vitamin K and transfusions of fresh frozen plasma and platelet concentrates. None of the five deaths in the trial were associated with bleeding despite the fact that both groups were significantly thrombocytopenic (Table III).
II.
22 4 3 1 49
Total
~~~
isolates
Total
Bacteria isolated in 36 episodes additional coagulase-negative
Klebsiella spp. Escherichia coli Enterobacter spp. Aerobic Gram-negative bacilli Proteus spp. n-haemolytic streptococci Pseudomonas spp. .___-. .~~~~~ .~~__ Coagulase-negative staphylococci Streptococcus faecalis Bacillus cereus Staphylococcus aureus
Micro-organisms
Table
~~
30
01
(a) Blood --_ 4 5 1 0 0 2 1 .___ 15 1 ~
19
30
0 ___~~ 7 3
: :,
(b) Other ~~~~ __-~ 0 2
Site
sites
~~
_~~
17
1
:: 0 0 0 ____~ 11 4
8
Latamoxef
Antibiotic
treated with tobramycin and latamoxef. staphylococci were cultured. Organisms resistant
of fever
15
01
Tobramycin ~~ 0 0 0 0 0 0 0 ...._~~ 10 4
resistance
30
20 (1)
13 0
0
: :
Successful ~-.. ~___~ 4 5
Therapeutic
response
in which
19
:
9 4
Failed ~~~ -___~ 0 2 l(1) 0 0 0 l(1)
Figures in brackets show three episodes to antibiotics appear below the line
282
E. G. Rhodes
et al.
Table III. Comparative details of 102 patients receiving continued high dose or continued low dose latamoxef. Mean values IL S.E.M. are shown on trial entry Latamoxef continued high dose (6 g day-‘) Treatment successes Neutrophil count ( x 10’ 1-l) Platelet count ( x lo9 IV’) Febrile period (days) Treatment duration (days)
0.23 56 4.0 7.8
Treatment failures Neutrophil count ( x lo9 1-l) Platelet count ( x 10’ 1-l) Febrile period (days) Treatment duration (days)
0.17f 45 f 9.3 f 10.4 f
f 0.05 f 10.2 f 0.42 f 0.46
0.06 7.0 1.38 1.57
continued low dose (3 g day-7 0.15f 38 f 3.7 zt 7.8 f
0.04 4.9 0.59 0.60
0.13zt 0.06 57 fll.5 6.7 zk 0.73 7.5 f 1.15
Discussion
The 68% response of febrile neutropenic patients to tobramycin and latamoxef reported in this study compares favourably with the 63% response to other broad-spectrum antibiotic combinations (Gaya, 1984). In leukaemic patients central venous catheters have become more widely used in recent years. In our own patients the catheters are inserted aseptically under vancomycin cover, flushed daily with heparinized saline and the skin exit sites sprayed daily with povidone iodine. Despite these measures, catheter infections with coagulase-negative staphylococci remain a major clinical problem. They are often resistant to all broad-spectrum antibiotics and it may, therefore, be necessary for future empirical antibiotic regimens to include additional agents which are active against these organisms. Apart from this, our results confirm the efficacy of latamoxef in a wide range of clinical infections (Kammer, 1982), including those in neutropenic patients (Stambaugh & McAdams, 1982). Th e newer /?-lactam antibiotics can induce multiple resistance and whilst some authors have advised caution over their uncontrolled widespread use (Sanders & Sanders, 1983), we ourselves have not encountered this problem. It is alleged that circulating antibiotic levels in excess of the required MICs ensure good tissue penetration and a rapid bactericidal effect. Since lower dosages may reduce this hypothetical effect, we performed a sequential analysis to carefully monitor the lower dose regimen. We found no significant difference in the response rate if latamoxef was reduced after 48 h to 3 gday-‘, and the lower dose did not result in longer periods of treatment. The numbers in this study give 80% power to detect a 25% difference between the two treatment regimens, at the 5% level (Pocock,
Latamoxef
in neutropenic
patients
283
1979). Within these limits therefore, the lower dose regimen represents a safe reduction. No serious side effects to latamoxef were encountered and coagulation defects were minor. One previous study showed no evidence of abnormal coagulation due to latamoxef (Bang et al., 1982), whilst others have reported hypoprothrombinaemia and defective platelet aggregation (Bechtold et al., 1984; Haupt & Barriere, 1984). Haemorrhage has been reported in some cases (Joehl et al., 1983), and this has led to a recommendation for prophylactic vitamin K (Beeley & Beadle, 1983, Haupt & Barriere, 1984). It has been suggested that like all broad-spectrum antibiotics, latamoxef may interfere with intestinal vitamin K synthesis (Wold, Buening & Hanasone, 1983). Other studies suggest that clotting factor carboxylation in the liver is inhibited by the N-methylthiotetrazole side chain of latamoxef (Lipsky, 1983). In our own study, an initially normal PT became prolonged in 9/57 patients receiving latamoxef. Whilst this was more common in patients receiving higher doses, only one instance of bleeding was recorded in 102 episodes. In a group of thrombocytopenic patients with septicaemia-related liver dysfunction (Jeffries, 1979), this is a low incidence of bleeding. The long PTS we observed in 35 other patients before antibiotics were given shows that it may be misleading to ascribe the coagulation abnormalities to antibiotics. In these patients vitamin K prevented any alteration in the PT during the subsequent antibiotic therapy. In conclusion, therefore, significant clotting abnormalities were seen in a minority of patients and in most cases this was not due to antibiotic therapy. Prothrombin times should, however, be monitored, and vitamin K administered as necessary.
Our thanks are due to the Pharmacy Departments at the Queen Elizabeth Hospital, Birmingham, and at Walsgrave Hospital, Coventry, for their valuable assistance. \Ve would also like to thank Dr P. P. Brown for microbiological advice, Professor C;. A. J. Ayliffe for advice on the manuscript and Bliss Catherine Garyey for her secretarial assistance in preparing the manuscript.
References Bang,
N. C. & Kammer, R. B. (1983). Hematologic complications associated with p-lactam antibiotics. Reviews of Infectious Diseases 5, 538G5393. Bang, N. U., Tessler, S. S., Heidenreich, R. O., Marks, C. A. & Mattler, L. E. (1982). Effects of Moxalactam on blood coagulation and Dlatelet function. Reaiews of Infectious Diseases 4, 546-554. Bechtold, H., Andrassy, K., Jahnchen, E., Koderisch, H., Weilemann, L. S., Sonntag, H. G. & Ritz, E. (1984). Evidence for impaired hepatic vitamin K, metabolism in patients treated with N-methvl-thiotetrazole cephalosporins. Thrombosis and Haemostasis (Stuttgart) 51, 358-361. Beeley, L. & Beadle, F. (1983). Bleeding with latamoxef. British Medical Journal 287, 1028. Gaya, H. (1984). Rational basis for the choice of regimens for empirical therapy of sepsis in granulocytopenic patients. Infections in Haematology. In Clinical Haemutology (Grant Prentice, H., Ed.), pp. 573-586. W. B. Saunders, Philadelphia. Haupt, B. A. & Barriere, S. L. (1984). Effects of Moxalactam on blood coagulation and platelet function. Drug Intelligence and Clinical Phurmucology 18, 590-591.
284
E. G. Rhodes
et al.
Jeffries, G. H. (1979). Diseases of the Liver. In Textbook of Medicine (Beeson, PI B., McDermott, W. & Wyngaarden, J. B., Eds), pp. 1656. W. B. Saunders, Philadelphia. Joehl, R. J. Rasbach, D.A., Ballard, J. O., Weitekemp, M. R. & Sattler, F. R. (1983). Moxalactam: Evaluation of clinical bleeding in patients with abdominal infection. Archives of Surgery 118, 1259-1261. Kammer, R. B. (1982). Moxalactam: clinical summary of efficacy and safety. Reviews of Infectious Diseases 4, 712-719. Lipsky, J. J. (1983). Latamoxef-associated hypoprothrombinaemia. Lancet ii, 624, Pocock, S. K. (1979). Statistical methods in the design and analysis of long term clinical trials. Drug treatment andprevention in cerebral vascular disorders (Tognoni, G. & Garattini, S., Eds). Elsevier L, North Holland Biomedical Press, Amsterdam. Rodriguez, V., Burgess, M. & Bodey, P. (1973). Management of fever of unknown origin in patients with neoplasms and neutropenia. Cancer 32, 1007-1012. Sanders, C. C. & Sanders, W. E. (1983). Emergence of resistance during therapy with the newer fi-lactam antibodies: role of inducible /3-lactamases and implications for the future. Reviews of Infectious Diseases 5, 639-648. Stambaugh, J. E. & McAdams, J. (1982). The efficacy and safety of Moxalactam in the treatment of acute bacterial infections in immunosuppressed patients with cancer. Current Therapeutic Research 31, 864-871. Stokes, E. J. & Ridgway, G. L. (1980). Clinical Bacteriology, 2nd edn, pp. 205. Edward Arnold, London. Wold, J. S. Buening, M. K. & Hanasone, G. K. (1983). Latamoxef-associated hypoprothrombinaemia. Lancet ii, 408.
of Hospital
Empirical
Infection
(1987)
9, 278-284
treatment of febrile, neutropenic with tobramycin and latamoxef
E. G. Rhodes*,
patients
R. I. Harris?, R. S. Welch*, D. J. Perry?, Brown* and B. J. Boughton*?
R. M.
Departments of Haematology and Microbiology, *Queen Elizabeth Hospital, Birmingham, and jCoventry and Warwickshire Hospital, Coventry Accepted for publication
14 July
1986
Summary:
One hundred and two febrile episodes in neutropenic patients were treated with intravenous tobramycin and latamoxef. After 48 h latamoxef at 6 g day-‘, patients were randomized to continue this regimen or latamoxef at 3 g day-‘. Infections responded to these regimens in 67% and 71% of patients, respectively. Two-thirds of the infections which failed to respond were due to coagulase-negative staphylococci in Hickman catheters, a trend which may necessitate the inclusion of additional antibiotics in future empirical regimens. Prolonged prothrombin times due to antibiotic therapy were seen in nine patients but there was only one episode of bleeding and this responded quickly to treatment with vitamin K and fresh frozen plasma. In 35 patients, coagulopathy was present before antibiotics were started, and these cases also responded to vitamin K. The study shows that the response to tobramycin and latamoxef is comparable to other broad-spectrum antibiotic regimens and that a reduction in the dose of latamoxef after 48 h treatment may safely permit cost savings.
Introduction Any infection in neutropenic patients is potentially serious, and intravenous broad-spectrum antibiotics should be given without waiting for the result of microbiological cultures. Latamoxef is a new broad-spectrum, a-lactamase stable cephalosporin which we have assessed in combination with tobramycin in neutropenic patients. In addition, we have examined its adverse effects, particularly those relating to coagulation (Bang & Kammer, 1983). Patients and methods All patients were adults receiving intravenous chemotherapy for acute myeloblastic or lymphoblastic leukaemia. The criteria for trial entry were (a) temperature 3 38°C sustained for at least 2 h and not due to drug $ Correspondence 0195-67Ol/87/030278+07
to Dr B. J. Boughton. 0 1987 The Hospital
SO3.00/0
278
Infection
Society
Latamoxef
in neutropenic
patients
279
reactions or blood products, and (b) neutrophil count < 1.0 x 10” 1-l. Tobramycin 1.6 mg kg-’ g-hourly was given by bolus intravenous injection and subsequently adjusted according to serum levels. Latamoxef 2g g-hourly was given intravenously for 48 h, after which patients were randomized to receive either 1 g or 2 g g-hourly (continued low dose, CLD; continued high dose, CHD). Bacteriological cultures from relevant sites were performed before treatment and repeated twice weekly. Blood cultures were obtained via Hickman catheters when these were in situ. Bacterial sensitivities to tobramycin and latamoxef were measured using the modified Stokes disc diffusion method (Stokes & Ridgway, 1980). Prothrombin times were measured by the Quick one-stage method. The clinicians responsible for the patients had no knowledge of the randomized dose of latamoxef. Infective episodes were classified as (1) microbiologically documented, (2) clinically documented or (3) pyrexias of unknown origin (Gaya, 1984). In patients in whom fever settled and clinical signs of infection improved, antibiotics were usually continued for a further 3 days (Rodriguez et al., 1973). In treatment failures, additional antibiotics or granulocyte transfusions were utilized according to the clinical and bacteriological picture. An independent sequential analysis was conducted, with the intention to close the trial if significant differences emerged between the treatment groups. The study received local ethical committee approval. The significance of differences between values was assessed using Student’s t-test and Fisher’s exact probability test. Results
One hundred and two episodes of fever were seen in 70 patients. One further case was withdrawn before randomization because of Clostridium welchii septicaemia treated with additional penicillin G, and a second patient died before randomization. Forty-six cases were documented bacteriologically, 14 were documented by clinical criteria such as pneumonic shadowing on chest X-ray and 42 were pyrexias of unknown origin (PUO). Seventy episodes responded to treatment, 36/54 (67%) on CHD latamoxef, and 34/48 (71%) on the CLD (Table I). In the CLD patients who responded to treatment, 13/34 (38%) responded within 48 h whilst receiving the pre-randomization high dose treatment, and 21/34 (62%) responded after further treatment with CLD latamoxef. Response rates to CHD and CLD latamoxef in episodes subclassified by clinical and microbiological criteria were similar k2 = 0.08; P> 0.05 (Table I). There were 32 treatment failures of whom 18 received CHD and 14 received CLD latamoxef. A wide variety of micro-organisms were isolated, though the high incidence of coagulase-negative staphylococci and the low incidence of Pseudomonas spp. is unusual. Organisms resistant to tobramycin or latamoxef were seen in 15 and 17 episodes, respectively11 coagulase-negative staphylococci, four Streptococcus faecalis, one
E. G. Rhodes
280
et al.
Table I. Comparative details of 102 febrile neutropenic patients receiving tobramycin and continued high dose or continued Cow dose latamoxef Latamoxef continued high dose (6 g day-9
continued low dose (3 g day-7
Treatment successes (70) Microbiologically documented Clinically documented Pyrexia of unknown origin
14 4 18
16 5 13
Total
36
34
9 3 6
5 5
18
14
Treatment failures (32) Microbiologically documented Clinically documented Pyrexia of unknown origin Total
Staphylococcus aureus and one BaciZlus cereus (Table II). Sixteen treatment failures responded to additional vancomycin, four to ampicillin, four to amphotericin and three to granulocyte transfusions. Five patients died despite additional measures, three treated with CHD and two with CLD latamoxef. Three of these had PUOs, one had cytomegalovirus pneumonia and one patient treated with CHD latamoxef died of septicaemia due to a tobramycin/latamoxef-sensitive Escherichia coli. In the CHD and CLD latamoxef groups there were no significant differences in the degree of neutropenia and thrombocytopenia, nor in the duration of fever or the duration of treatment (P> 0.1, Table III). One patient developed a skin rash, which resolved on withdrawing the drugs, and diarrhoea not associated with Cl. di&ile toxin occurred in a further six cases. There was no antibiotic-induced hepatic or renal toxicity. The prothrombin time (PT) was measured prior to antibiotic treatment in 92 patients and in 35 it was prolonged 3-15 s. These received vitamin K and none showed further prolongation of the PT during antibiotic therapy. In 57 patients with normal coagulation on trial entry, no prophylactic vitamin K was given during the antibiotic chemotherapy and prolongation of the PT was seen in nine. Six of these received CHD and three received CLD latamoxef. Only one patient experienced clinically significant bleeding, where gastrointestinal haemorrhage after 8 days’ treatment with CLD latamoxef responded to vitamin K and transfusions of fresh frozen plasma and platelet concentrates. None of the five deaths in the trial were associated with bleeding despite the fact that both groups were significantly thrombocytopenic (Table III).
II.
22 4 3 1 49
Total
~~~
isolates
Total
Bacteria isolated in 36 episodes additional coagulase-negative
Klebsiella spp. Escherichia coli Enterobacter spp. Aerobic Gram-negative bacilli Proteus spp. n-haemolytic streptococci Pseudomonas spp. .___-. .~~~~~ .~~__ Coagulase-negative staphylococci Streptococcus faecalis Bacillus cereus Staphylococcus aureus
Micro-organisms
Table
~~
30
01
(a) Blood --_ 4 5 1 0 0 2 1 .___ 15 1 ~
19
30
0 ___~~ 7 3
: :,
(b) Other ~~~~ __-~ 0 2
Site
sites
~~
_~~
17
1
:: 0 0 0 ____~ 11 4
8
Latamoxef
Antibiotic
treated with tobramycin and latamoxef. staphylococci were cultured. Organisms resistant
of fever
15
01
Tobramycin ~~ 0 0 0 0 0 0 0 ...._~~ 10 4
resistance
30
20 (1)
13 0
0
: :
Successful ~-.. ~___~ 4 5
Therapeutic
response
in which
19
:
9 4
Failed ~~~ -___~ 0 2 l(1) 0 0 0 l(1)
Figures in brackets show three episodes to antibiotics appear below the line
282
E. G. Rhodes
et al.
Table III. Comparative details of 102 patients receiving continued high dose or continued low dose latamoxef. Mean values IL S.E.M. are shown on trial entry Latamoxef continued high dose (6 g day-‘) Treatment successes Neutrophil count ( x 10’ 1-l) Platelet count ( x lo9 IV’) Febrile period (days) Treatment duration (days)
0.23 56 4.0 7.8
Treatment failures Neutrophil count ( x lo9 1-l) Platelet count ( x 10’ 1-l) Febrile period (days) Treatment duration (days)
0.17f 45 f 9.3 f 10.4 f
f 0.05 f 10.2 f 0.42 f 0.46
0.06 7.0 1.38 1.57
continued low dose (3 g day-7 0.15f 38 f 3.7 zt 7.8 f
0.04 4.9 0.59 0.60
0.13zt 0.06 57 fll.5 6.7 zk 0.73 7.5 f 1.15
Discussion
The 68% response of febrile neutropenic patients to tobramycin and latamoxef reported in this study compares favourably with the 63% response to other broad-spectrum antibiotic combinations (Gaya, 1984). In leukaemic patients central venous catheters have become more widely used in recent years. In our own patients the catheters are inserted aseptically under vancomycin cover, flushed daily with heparinized saline and the skin exit sites sprayed daily with povidone iodine. Despite these measures, catheter infections with coagulase-negative staphylococci remain a major clinical problem. They are often resistant to all broad-spectrum antibiotics and it may, therefore, be necessary for future empirical antibiotic regimens to include additional agents which are active against these organisms. Apart from this, our results confirm the efficacy of latamoxef in a wide range of clinical infections (Kammer, 1982), including those in neutropenic patients (Stambaugh & McAdams, 1982). Th e newer /?-lactam antibiotics can induce multiple resistance and whilst some authors have advised caution over their uncontrolled widespread use (Sanders & Sanders, 1983), we ourselves have not encountered this problem. It is alleged that circulating antibiotic levels in excess of the required MICs ensure good tissue penetration and a rapid bactericidal effect. Since lower dosages may reduce this hypothetical effect, we performed a sequential analysis to carefully monitor the lower dose regimen. We found no significant difference in the response rate if latamoxef was reduced after 48 h to 3 gday-‘, and the lower dose did not result in longer periods of treatment. The numbers in this study give 80% power to detect a 25% difference between the two treatment regimens, at the 5% level (Pocock,
Latamoxef
in neutropenic
patients
283
1979). Within these limits therefore, the lower dose regimen represents a safe reduction. No serious side effects to latamoxef were encountered and coagulation defects were minor. One previous study showed no evidence of abnormal coagulation due to latamoxef (Bang et al., 1982), whilst others have reported hypoprothrombinaemia and defective platelet aggregation (Bechtold et al., 1984; Haupt & Barriere, 1984). Haemorrhage has been reported in some cases (Joehl et al., 1983), and this has led to a recommendation for prophylactic vitamin K (Beeley & Beadle, 1983, Haupt & Barriere, 1984). It has been suggested that like all broad-spectrum antibiotics, latamoxef may interfere with intestinal vitamin K synthesis (Wold, Buening & Hanasone, 1983). Other studies suggest that clotting factor carboxylation in the liver is inhibited by the N-methylthiotetrazole side chain of latamoxef (Lipsky, 1983). In our own study, an initially normal PT became prolonged in 9/57 patients receiving latamoxef. Whilst this was more common in patients receiving higher doses, only one instance of bleeding was recorded in 102 episodes. In a group of thrombocytopenic patients with septicaemia-related liver dysfunction (Jeffries, 1979), this is a low incidence of bleeding. The long PTS we observed in 35 other patients before antibiotics were given shows that it may be misleading to ascribe the coagulation abnormalities to antibiotics. In these patients vitamin K prevented any alteration in the PT during the subsequent antibiotic therapy. In conclusion, therefore, significant clotting abnormalities were seen in a minority of patients and in most cases this was not due to antibiotic therapy. Prothrombin times should, however, be monitored, and vitamin K administered as necessary.
Our thanks are due to the Pharmacy Departments at the Queen Elizabeth Hospital, Birmingham, and at Walsgrave Hospital, Coventry, for their valuable assistance. \Ve would also like to thank Dr P. P. Brown for microbiological advice, Professor C;. A. J. Ayliffe for advice on the manuscript and Bliss Catherine Garyey for her secretarial assistance in preparing the manuscript.
References Bang,
N. C. & Kammer, R. B. (1983). Hematologic complications associated with p-lactam antibiotics. Reviews of Infectious Diseases 5, 538G5393. Bang, N. U., Tessler, S. S., Heidenreich, R. O., Marks, C. A. & Mattler, L. E. (1982). Effects of Moxalactam on blood coagulation and Dlatelet function. Reaiews of Infectious Diseases 4, 546-554. Bechtold, H., Andrassy, K., Jahnchen, E., Koderisch, H., Weilemann, L. S., Sonntag, H. G. & Ritz, E. (1984). Evidence for impaired hepatic vitamin K, metabolism in patients treated with N-methvl-thiotetrazole cephalosporins. Thrombosis and Haemostasis (Stuttgart) 51, 358-361. Beeley, L. & Beadle, F. (1983). Bleeding with latamoxef. British Medical Journal 287, 1028. Gaya, H. (1984). Rational basis for the choice of regimens for empirical therapy of sepsis in granulocytopenic patients. Infections in Haematology. In Clinical Haemutology (Grant Prentice, H., Ed.), pp. 573-586. W. B. Saunders, Philadelphia. Haupt, B. A. & Barriere, S. L. (1984). Effects of Moxalactam on blood coagulation and platelet function. Drug Intelligence and Clinical Phurmucology 18, 590-591.
284
E. G. Rhodes
et al.
Jeffries, G. H. (1979). Diseases of the Liver. In Textbook of Medicine (Beeson, PI B., McDermott, W. & Wyngaarden, J. B., Eds), pp. 1656. W. B. Saunders, Philadelphia. Joehl, R. J. Rasbach, D.A., Ballard, J. O., Weitekemp, M. R. & Sattler, F. R. (1983). Moxalactam: Evaluation of clinical bleeding in patients with abdominal infection. Archives of Surgery 118, 1259-1261. Kammer, R. B. (1982). Moxalactam: clinical summary of efficacy and safety. Reviews of Infectious Diseases 4, 712-719. Lipsky, J. J. (1983). Latamoxef-associated hypoprothrombinaemia. Lancet ii, 624, Pocock, S. K. (1979). Statistical methods in the design and analysis of long term clinical trials. Drug treatment andprevention in cerebral vascular disorders (Tognoni, G. & Garattini, S., Eds). Elsevier L, North Holland Biomedical Press, Amsterdam. Rodriguez, V., Burgess, M. & Bodey, P. (1973). Management of fever of unknown origin in patients with neoplasms and neutropenia. Cancer 32, 1007-1012. Sanders, C. C. & Sanders, W. E. (1983). Emergence of resistance during therapy with the newer fi-lactam antibodies: role of inducible /3-lactamases and implications for the future. Reviews of Infectious Diseases 5, 639-648. Stambaugh, J. E. & McAdams, J. (1982). The efficacy and safety of Moxalactam in the treatment of acute bacterial infections in immunosuppressed patients with cancer. Current Therapeutic Research 31, 864-871. Stokes, E. J. & Ridgway, G. L. (1980). Clinical Bacteriology, 2nd edn, pp. 205. Edward Arnold, London. Wold, J. S. Buening, M. K. & Hanasone, G. K. (1983). Latamoxef-associated hypoprothrombinaemia. Lancet ii, 408.