Sparfloxacin as alternative treatment to standard therapy for community-acquired bacteremic pneumococcal pneumonia


  • Sparfloxacin Study Group Members: 1. Garau (Hospital Mutua de Tarrasa, Spain), C. Grassi (San Mattco University Policlinic, Pavia, Italy), J. Hosie (1980, Great Western Road, Glasgow, UK), G. Huchon (Hôpital Ambroise Paré, Boulogne, France), N. Legakis (National University of Athens, Greece), 5. Segev (Chain Sheba Medical Centre, Tel Hashomer, Israel), G. Wijnands (Stichting Deventer Hospltal, Holland), C. Regamey (Hôpital Cantonal, Fribourg, Switzerland), R. Verster (Libanon, South Africa), T. May (Maladies Infectieuses, Nancy, France), M. Valtonen (University Central Hospital, Helsinki, Finland), O. Cars (Uppsala University Sweden), M. Wahl (Ostra Hospital, Goteberg, Sweden), P. Saikku (National Public Health Institute, Oulu, Finland), C. Jean and M. P. Richard (Rhône-DPC, Antony, France).

Corresponding author and reprint requests: H. Lode, Pneumologie l, Lungenklinik Heckeshorn, Akademisches Lehrkrankenhaus der FU Berlin, Zum Heckeshorn 33, D-14109 Berlin, Germany. Tel: +49 30 8002 2339 Fax: +49 30 8002 2623


Objective: To assess the effectiveness of a single daily dose of sparfloxacin in comparison with standard antibacterial therapy for the treatment of pneumococcal bacteremic community-acquired pneumonia (CAP).

Methods: The results were analyzed of four comparative trials in CAP, in which 107 adult patients with CAP confirmed by blood cultures positive for Streptococcus pneumoniae were included. Sparfloxacin was given at a loading dose of 400 mg followed by 200 mg daily. Comparator drugs included amoxycillin 3 g/day, amoxycillin/clavulanate 1.5/0.375 g/day and erythromycin 2 g/day. Dosing was for 7–14 days (mean 10 days). Success was determined by a combination of clinical and microbiological assessment and radiologic changes.

Results: Sparfloxacin was as effective as the comparator drugs, with an overall success rate of 80% at the end of treatment (comparators 78%), and a 79% success rate at follow-up (76% for comparators). There were no pneumococcal isolates resistant to sparfloxacin, but eight of 56 were either resistant (four) or had reduced susceptibility to penicillin G, and two strains were resistant to erythromycin. Tolerance to sparfloxacin was good, with fewer patients reporting drug-related adverse events (15.8%) than with the comparator drugs (33.3%).

Conclusions: This analysis suggests that sparfloxacin would be an alternative candidate for empirical therapy in moderately severe CAP.


Community-acquired respiratory tract infections (RTIs) are a major cause of morbidity worldwide, and most require empirical therapy, since they can be acute, of rapid onset, and serious [1]. Community-acquired pneumonia (CAP) is the most serious of the respiratory infections, frequently requiring hospitalization and carrying a 5–15% mortality rate. Empirical therapy is therefore generally regarded as essential [2]. Although traditionally the penicillins have been the mainstay of treatment for many CAPS, recent years have seen a shift towards a greater use of compounds with a broader spectrum of activity. A number of factors have contributed towards this, including: the increasing numbers of β-lactamase-producing strains of Haemo-philus influenzae, Movaxella catarrhalis and Staphylococcus aureus; the growing threat of strains of Streptococcus pneumoniae with reduced susceptibility or resistance to penicillin; and the recognition of the contribution to the pathogenesis of RTI by ‘atypical’ pathogens such as Chlamydia pneumoniae, Legionella pneumophila and Mycoplasma pneumoniae.

The more recent quinolones have a spectrum of activity that includes most of the major pathogens involved in RTI. However, a drawback to the empirical use of some of these compounds (such as ciprofloxacin and ofloxacin) for KTI is their limited activity relative to penicillins against Streptococcus pneumoniae [3], still the most prevalent and most serious bacterial pathogen in CAP [1,4]. The new quinolone sparfloxacin, a derivative of pyridone carboxylic acid, has better activity than other marketed quinolones against Streptococcus pneumoniae [5], with MIC90 values of 0.25–0.5 mg/L [6,7]. It retains activity against strains with resistance to or decreased susceptibility to penicillins and macrolides [8,9], and this good activity in vitro has been confirmed experimentally in vivo, using a mouse model of pneumonia [10].

The activity of sparfloxacin has also been demonstrated in the clinic against various types of RTI, including CAP [11–14], and its safety in these trials has been summarized by Rubinstein [15]. When an oral loading dose of 400 mg is given, followed by a daily dose of 200 mg, steady state with a peak serum concentration of 1.4 mg/L is reached after the second dose. The trough concentration is 0.5 mg/L and the T1/2 is approximately 20 h [16]. The compound penetrates respiratory tract tissues to a greater extent than other quinolones [17], achieving concentrations that exceed the plasma levels considerably; levels can thus be achieved at infective sites that exceed the MIC for many pathogens for prolonged periods. These pharmacokinetic characteristics allow once-daily dosing with sparfloxacin.

Clinical trials of CAP have varied widely in their approach, with respect to both defining and diagnosing pneumonia [18] and assessing the efficacy of antibiotics [12,19]. Many trials assess efficacy as cure, improvement or failure, but the classification of ‘improvement’ is difficult, since it is highly subjective and does not reflect normal clinical practice. In the study reported here, we have used a simple binary expression of cure or failure for overall assessment (this approach is explained in detail by Genevois et al [19]). In this analysis, we present results on the use of sparfloxacin in the treatment of CAP where the infection had been confirmed by blood culture as having been caused by Streptococcus pneumoniae. The patients were selected from four comparative trials in 11 countries (Europe, Israel and South Africa) carried out between 1990 and 1993. Various comparator drugs were used.


Study design

The analysis was performed on four double-blind, randomized, multicenter trials of CAP involving a total of 1654 patients. The trials were conducted mostly in Europe (Finland, France, Germany, Greece, The Netherlands, Norway, Spain, Sweden and Switzerland), with a few patients being recruited from Israel and South Africa, and the results of those individual clinical trials have been published elsewhere [11–14]. Only those patients with Streptococcus pneumoniae pneumonia confirmed by blood culture have been included in the present assessment.


Patients of either sex, aged over 18 years, with acute CAP were included. Pneumonia was defined as the presence of typical respiratory signs and symptoms (including purulent bronchial or pulmonary secretions, cough, chills, dyspnea, sputum production, chest pain), a temperature equal to or exceeding 38°C, and new infiltrates (solid or patchy) evident on chest radiographs.

Patients with severe underlying diseases (such as cancer, tuberculosis or AIDS, although HIV-positive non-AIDS patients were included), severe infections requiring parented or intensive care therapy (such as infections associated with acute respiratory distress syndrome, meningitis, septic shock, mechanical ventilation, or intensive bilateral consolidation on radiographs), and those with large pleural effusions on radiographs were excluded. Other exclusion criteria included treatment with systemic corticosteroids, antacids and iron salts, and pregnancy or lactation. Patients who had been hospitalized immediately before the onset of this episode of CAP were also excluded to avoid the possibility of their pneumonia being acquired in hospital. All patients gave their informed consent before entry to the study.

Study drugs

Patients in these trials were randomized to receive oral treatment with either sparfloxacin or a control antibiotic. A loading dose of 400 mg of sparfloxacin (Rhône DPC, Paris, France) was given on day 1, followed by a daily dose of 200 mg for 7–14 days, with a mean value of 10 days. The comparator antibiotics were also administered for 7–14 days, with a mean value of 10 days, and were: amoxycillin (Smith Kline Beecham Laboratories, France). 1 g three times a day; amoxycillin/clavulanate (Smith Kline Beecham Laboratories, France), 500/125 mg three times a day; amoxycillin 1 g three times a day plus ofloxacin (Roussel Uclaf, France) 200 mg twice a day; roxithromycin (Roussel Laboratories, Denmark) 150 mg twice a day; and erythromycin (August Wolff Laboratories, Germany), 1 g twice a day. Patients were advised to avoid exposure to sunlight or to tanning (UV) lamps for the duration of the trial.

Study conduct

Patients in the trials were treated on an inpatient or an outpatient basis as convenient, but most were inpatients. Patients were seen on admission (visit 1), after 48 h of treatment (i.e. at 3 days, or visit 2), at the completion of (or soon after the completion of) treatment (visit 3) and at a follow-up visit 4–6 weeks after the end of treatment (visit 4).

At visit 1, the patient's medical and surgical history was noted, and blood samples collected for biochemical and hematologic analysis. A complete physical examination, including temperature, respiratory rate, and determination of the clinical signs and symptoms of respiratory infection (as noted above), was performed at each visit. These symptoms were classified as mild, moderate or severe, and sputum as none, mucoid, mucopurulent, purulent or hemoptysis. A chest radiograph was taken on admission and at the follow-up visit. If the investigator thought it necessary, a radiograph was taken at the end of treatment (visit 3), and if this showed resolution or improvement, the radiograph at follow-up was not taken. Blood and urine samples were collected for hematologic and biochemical analysis at the end of treatment. If any abnormalities were evident in these samples, a further sample was taken at the follow-up visit. Additional blood samples for microbiological culture were taken only from those patients who did not respond to treatment.

Patients were withdrawn from the study if protocol violations occurred, or if deemed necessary because of an adverse event, or if they failed to respond to treatment. Any patients discontinuing the study were examined at a follow-up visit.

Analysis of efficacy

Efficacy analyses were performed on both the intention-to-treat population and the evaluable population at two intervals; from study entry (visit 1) to end of treatment (visit 3) and from study entry (visit 1) to follow-up (visit 4). An overall efficacy classification was used that combined both clinical and radiologic changes and was expressed simply as ‘success’ or ‘failure’ [11–14, 19]. This classification was determined by the application of predefined criteria. An external steering committee reviewed the patients when the data were contradictory [19].

Clinical effects were classified as cure, improvement or failure using the predetermined criteria. Cure included the absence of fever, chills, chest pains, cough or dyspnea, and the absence of sputum production at the follow-up visit. However, patients with cardiac failure, with concomitant bronchopulmonary disease, or who smoked, and who showed persistence of mild chest pain, cough, dyspnea, or the presence of non-purulent sputum, could still be classified as cured provided the symptoms were no worse than before the onset of acute illness. Improvement was classified as the absence of chills and fever, but other symptoms could be present at a predetermined level of severity. All other patients were classified as failures. Radiologic changes were classified directly by the investigators as worse, no change, improvement or resolution.

Overall success was defined as those with resolution or improvement on X-ray examination together with a clinical cure, or resolution on X-ray examination together with a clinical improvement. All clinical failures, irrespective of radiologic findings, were classified as overall failures. Those patients with conflicting radiologic or clinical data at visit 3 were defined by the steering committee at visit 4. All other patients were deemed overall failures.

Microbiological assessment

All patients included in this analysis were those in whom Streptococcus pneumoniae had been cultured from the blood sample collected prior to commencing treatment. Wherever possible, blood samples were cultured from any patients not responding to treatment. The minimum inhibitory concentrations (MICs) for a range of antibiotics against the pneumococcal strains isolated were determined using a standard microtiter technique.

Safety evaluation

Blood samples collected immediately prior to treatment (visit 1) and at the end of treatment (visit 3) were analyzed for the usual biochemical and hematologic parameters. The urine samples were tested for protein, blood and glucose using a dipstick method. All adverse experiences, elicited by non-specific questioning or volunteered by the patients, were noted at each visit. These were classified by the investigator as mild, moderate, or severe/life-threatening, and their relationship to the drug administered was classified as unrelated, possibly related, or probably related. The international COSTART terms were used to describe adverse events.

Statistical analysis

All analyses, including clinical and microbiological, were descriptive. Overall success rates are presented for each group at the end of treatment and follow-up, in both the intention-to-treat population and evaluable patients. The bacteriologic response is presented in the clinically and bacteriologically evaluable patients. All patients with a premature withdrawal of the treatment because of an adverse event, whose consent was withdrawn, or who were lost to follow-up, were classified automatically as a failure in the intention-to-treat analysis, but were excluded from other efficacy analyses in the evaluable population.



One hundred and seven patients with Streptococcus pneumoniae pneumonia confirmed by blood culture were included in this assessment. Fifty-eight patients had been randomized to receive sparfloxacin and 49 to receive one of the comparator antibiotics. In the comparator groups, erythromycin had been administered to 17 patients, amoxycillin to 19, amoxycillin/clavulanate to seven, amoxycillin plus ofloxacin to three, and roxithromycin to three.

Of the 107 patients included in the analysis, 94 were deemed evaluable for efficacy, and 13 patients (six in the sparfloxacin group and seven in the comparator group) were not considered to be evaluable for number of reasons (listed in Table 1), including various protocol violations and the occurrence of adverse events. A further two patients (one in each group) were not evaluable at the end of treatment, and seven more patients (four in the sparfloxacin group and three in the comparator group) were not evaluable at follow-up, leaving a total of 85 patients evaluable after the follow-up visit.

Table 1.  Patients evaluable for efficacy and safety
  1. aPatient wrongly included with a mildly febrile pulmonary edema from cardiac insufficiency. bVomiting. cRenal failure.

Patients entering study5849107
Patients evaluable for safety and efficacy   
 Intention-to-treat analysis5849107
Not evaluable for overall efficacy6713
 No pneumonia112
 Treated >3 days224
 Pulmonary edema1a-1
 Other antibiotics-22
 No risk factors-22
 Early withdrawal, adverse event2b-2
Evaluable for overall efficacy524294
Not evaluable at end of treatment (visit 3)112
 Concurrent illness-11
 Missing data1-1
Evaluable at end of treatment (visit 3)514192
 Not evaluable for follow-up (visit 4)437
 Lost to follow-up224
 Other inedication-11
 Concurrent illness2-2
Evaluable at follow-up (visit 4)473885

Baseline data

The demographic data for the two groups (sparfloxacin and comparator antibiotics) are detailed in Table 2. The mean age of the patients was 49 years, they were predominantly male, and approximately 90% were Caucasian. No statistically significant differences were evident between the two groups, although sputum production was more frequent in the comparator group (79.6%) than in the sparfloxacin group (65.5%), as was an association with chronic bronchitis (18.4% versus 10.3%).

Table 2.  Demographic and baseline data of patients on admission
ParameterSparfloxa (n=58)Comparators (n=49)
Age in years; mean (SEM)49(2)49(3)
Males (%)63.867.3
Temperature (°C); mean (SEM)39.4(0.09)39.3(0.12)
Chills (%)74.179.6
Respiratory rate (min); mean (SEM)24.1(0.8)25.6(0.9)
Dyspnea (%)74.177.6
Chest pain (%)82.889.8
Sputum production (%)65.579.6
Associated chronic bronchitis (%)10.318.4
Associated asthma (%)6.98.2
Pleural effusion (%)10.36.1
Unilateral pneumonia (X-ray) (%)91.493.9

The chest radiographs revealed an alveolar type of lobar pneumonia in over 95% of the cases. A single-lobe alveolar image was present in approximately 40% of the patients and more extensive alveolar images (more than one lobe, uni- or bilateral extension) were present in 50–60% of the patients (Table 3).

Table 3.  Rahologic findings of patients on admission
 Sparfloxacin (n=58)Comparators (n=49)
Alveolar—undated25 (43.1%)30(61.2%)
Alveolar—bilateral3 (5.2%)1 (2%)
Alveolar-one lobe28 (48.3%)16 (32.7%)

The MIC distribution for various antibiotics against the strains of Streptococcus pneumnoniae isolated is detailed in Table 4, and shows that sparfloxacin MIC values were 1 mg/L or below for all except one strain, for which the MIC was 2 mg/L. The MIC90 was 0.5 mg/L, and 59 of 67 strains (88%) were inhibited by 0.25 mg/L. Although penicillin G, amoxycillin and amoxycillin/clavulanate had good activity against the majority of the strains tested (MIC90 0.25, 0.5 and 0.25 mg/L), 8 of 56 (penicillin G), 10 of 61 (amoxycillin) and 8 of 59 (amoxycillin/clavulanate) strains showed reduced susceptibility to the penicillins, MIC values ranging between 0.125 and 2 mg/L. The majority of strains were sensitive to erythromycin, but two strains were resistant, the MICs against them being 8 and 32 mg/L. In addition, one other strain showed decreased susceptibility to erythromycin, only being inhibited by 1 mg/L. Only seven strains were tested for sensitivity to roxithromycin, and all were susceptible to 0.25 mg/L or less. Ofloxacin was tested against only two strains; the MICs were 1.0 and 2.0 mg/L.

Table 4.  Distribution of MICs of antibiotics against Streptococcus pneumoniae (g/L)
  1. -, denotes no test performed at that concentration; 0, denotes no strain with that MIC value

Sparfloxacin (67)4114212843100-0.5
Penicillin G (56)--435211220000.25
Amoxycillin (61)---51122320000.5
Amoxycillin/clavulanate (59)---51122120000.25
Erythromycin (61)-1115191730100110.25
Roxithromycin (7)--1132000000-
Ofloxacin (2)00000001100--

Analysis of efficacy

Overall success rates in the evaluable populations were similar in both the sparfloxacin (80%) and the comparator groups (78%) at the end of treatment, dropping only slightly at the follow-up to 79% for sparfloxacin and 76% for the comparators. The values for the intention-to-treat populations were marginally lower, at approximately 74% for both groups at the end of treatment and 69% at follow-up (Table 5).

Table 5.  Overall efficacy at end of treatment and at follow-up; intention-to-treat and evaluable population
No. of patients in study5849
Intention-to-treat: end of treatment5849
 Success43 (74.1%)36 (73.5%)
 Failure12 (20.7%)10 (20%)
  Early failure3 (5.2%)3 (6.1%)
Intention-to-treat: follow-up5849
 Success10 (69%))34 (69.%)
 Failure15 (25.9%)12 (24.5)
  Early failure3 (5.2%)3 (6.1)
Evaluable: end of treatment5141
 success41 (80.4%)32 (78%))
 Failure7 (13.7%)6 (14.6%)
  Early failure3 (5.9%)3 (7.3)
Evaluable; follow-up4738
 Success37 (78.7%)29 (76.3%)
 Failure7 (14.9%)6 (15.8%)
  Early failure3 (6.4%)3 (7.9%)

There were three early failures in both the sparfloxacin group (failure on day 3) and the comparator group (failure at 3–4 days). All three sparfloxacin patients still had an elevated temperature (38–38.4°C), two had mucopurulent sputum, and the radiographs were either no better or worse than at the time of entry. The Streptococcus pneumoniae strains were all sensitive to sparfloxacin (MICs 0.125–0.5 mg/L). In one patient the strain was eradicated, but in the other two the outcome was unknown, as no blood cultures were performed at the end of treatment. An assay of the sparfloxacin serum concentration was performed in two of these patients: in one patient the concentration was 0.4 mg/L just before drug administration (i.e. a trough level), and in the other 0.02 mg/L 1 h after drug administration.

There were seven other failures in the evaluable sparfloxacin group, and these presented a variable picture. One patient with worsening radiographs was suspected of having a fungal infection and was given fluconazole, while another patient had persistently mucopurulent sputum. The other five patients had either a persistent fever or radiographs that did not improve. All isolates from these patients were sensitive to sparfloxacin.

In the comparator group, one early failure was treated with erythromycin and two with amoxycillin. All three had temperatures of 39°C or more and radiographs that had either deteriorated or not improved. The patient treated with erythromycin was infected with a strain of pneumococcus that was resistant to erythromycin (MIC 32 nig/L); a subsequent culture was not obtained.

There were six other failures in the evaluated patients treated with comparators, three with erythromycin and three with amoxycillin/clavulanate. Three patients failed at 5–6 days with fever and non-improving radiographs; in addition, one of these patients had a persistent cough and dyspnea. The other three patients continued to have a slight fever and/or chest pain, although chest radiographs were improved. In one patient treated with erythromycin, the pneumococcal strain was resistant to erythroniycin (MIC 8 mg/L).

Microbiological response

Although the protocol specified that d blood sample should be collected from patients in whom treatment was not a success, for many patients no follow-up sample was obtained; the number of non-evaluable responses was therefore high. If, however, the patient was classified as a success, then it was presumed that the organism was eradicated. Table 6 lists the microbiological responses to the therapy at the end of treatment and at follow-up.

Table 6.  Microbiological assessment; isolation of Streptococcus pneumoniae
 End of treatmentFollow-up (visit 4)
  1. aA negative culture but a clinical failure.

  2. bPersistent positive blood culture at end of treatment.

  3. cClinical relapse with positive blood culture after end of treatment in an HIV-positive patient.

  4. dClinical failure with negative culture at end of treatment. but not evaluable at follow-up.

Presumed eradication38303731
Negative culturea12-1d

There were no patients with a persistent infection in the sparfloxacin group. In one patient treated with erythromycin, the blood culture was still positive after 2 days of treatment; the patient was withdrawn from the study and treated with other antibiotics. Another patient who was HIV positive and treated with erythromycin relapsed after the end of treatment with meningitis and had a blood culture positive for Streptococcus pneumoniae. A patient treated with amoxycillin/ clavulanate had a persistent bacteremia after 5 days of treatment, although responding clinically and radio-logically.

Eight strains of Streptococcus pneumoniae had reduced susceptibility to or were resistant to penicillin G and amoxycillin (MICs 0.125–2.0 mg/L) and two strains were resistant to erythromycin (MICs 8 and 32 mg/L). One of the erythromycin-resistant strains also had reduced susceptibility to penicillin (MIC 0.5 mg/L). Both of the patients infected with the erythromycin-resistant strains were randomized to receive erythromycin, and failed treatment. Of the four patients infected with penicillin-resistant pneumococci (MIC 1.O and 2.0 mg/L), three received sparfloxacin and one erythromycin; all four responded successfully to therapy. Three patients infected with pneumococci with reduced penicillin susceptibility (MICs 0.125- 0.25 mg/L) were treated successfully, two with erythromycin and one with amoxycillin.

Safety evaluation

Overall, the incidence of adverse events reported was similar in both groups, with a total of 32 patients in the sparfloxacin group reporting 49 adverse events, and 31 patients in the comparator group reporting 60 adverse events. Of these adverse events, only 15 in the sparfloxacin group were regarded as being possibly or probably related to the drug compared with 17 events in the comparator group. Fewer patients reported drug-related adverse events in the sparfloxacin group (15.8%) compared with the comparator drugs (33.3%) (Table 7). There were slightly more gastrointestinal disturbances reported by patients receiving comparator drugs, and slightly more skin-related events in patients receiving sparfloxacin, but none of these were reports of phototoxicity.

Table 7.  Summary of adverse events
  1. aVomiting in two patients, clinical signs of failure in three patients.

  2. bclinical signs of failure.

Total number of patients5748
Number stopping therapy5a3b
Total number of adverse events reported4960
Number of patients reporting adverse events3231
% of patients reporting adverse events5664.5
Number of drug-related adverse events1517
Number of patients reporting drug-related adverse events916
% of patients reporting drug-related adverse events15.833.3

Three patients suffered severe or life-threatening adverse events that were regarded by the investigator as probably related to the treatment. One elderly male patient receiving sparfloxacin had severe nausea, vomiting and weight loss; the drug was withdrawn and he recovered uneventfully. The other two patients both had severe or life-threatening pneumonia; one was receiving erythromycin and one sparfloxacin. Both drugs were discontinued. Two patients died of causes unrelated to their therapy. One HIV-positive non-AIDS patient receiving erythromycin, after responding to therapy, relapsed and died of meningitis, and one chronic bronchitic patient with emphysema who received sparfloxacin died of respiratory failure and pulmonary aspergillosis.


The patients in the study reported here all had well-documented CAP with blood cultures positive for Streptococcus pneumoniae, and the severity of the disease in the two treatment groups was similar at baseline. Sparfloxacin at a single oral daily dose of only 200 mg produced a success rate that was at least equal to that seen with the comparator drugs; amoxycillin 3 g/day, amoxycillin plus ofloxacin (3 g plus 400 mg/day), amoxycillin/clavulanate 1.510.375 g/day, erythromycin 2 g/day or roxithromycin 300 mg/day. In the evaluable patients the success rate was 80% for sparfloxacin versus 78% for the comparators, and in the intention-to-treat population it was 74% for sparfloxacin versus 73.5% for the comparators at the end of treatment.

Since CAP can be such a serious disease, progressing rapidly if not treated immediately, and since Streptococcus pneumoniae is still an important and major pathogen [1], it is essential that the antibacterial agent given has good activity against the pneumococcus. Although penicillins have proved the mainstay of therapy against Streptococcus pneumoniae for decades, the increasing occurrence of strains with reduced susceptibility or resistance to penicillin worldwide [20] has compromised their value. The incidence of resistance varies between surveys and between countries, but Geslin and Leophonte, in a survey of 14 786 strains isolated in France between 1984 and 1993, found an increasing incidence of penicillin resistance from 0.5% in 1984 to 12% in 1990 and 25% in 1993 [21]. Resistance has also appeared to erythromycin [20], and in the same French survey, resistance to erythromycin increased from 18.9% in 1984, to 26.2% in 1990 and 34.8% in 1993 (211. Geslin and Leophonte also examined 1182 invasive strains of Streptococcus pneumoniae (isolated from the blood), and again saw an increase in resistance, with 22.9% resistant to erythromycin and 14.8% resistant to penicillin in 1993 [21]. Multiresistant strains of pneumococci are also increasing, and these are more likely to be community acquired [22].

The doses of the comparator drugs were chosen to reflect current clinical practice. Amoxycillin/clavulanate is widely used at 500/125 mg three times a day, and is frequently favored because of its cover of pneumococci and β-lactamase-producing strains of H. influenzae and Moraxella catarrhalis. However, the decreasing susceptibility of the pneumococcus noted above has led to a high dose of amoxycillin (3 g/day) being favored by some. Erythromycin is generally given at 2 g/day for CAP, and is used for its good cover of the atypical pathogens as well as Streptococcus pneumoniae. The dose of sparfloxacin was chosen from preliminary studies showing good tolerance [15] and desirable pharmacokinetics [16].

Both macrolides and quinolones have the ability to penetrate cells and accumulate in tissues. They also have good activity against most of the ‘atypical’ pathogens that can be involved in respiratory tract infections, and erythromycin has long been used as an alternative to penicillins. However, as noted above, the increasing number of macrolide-resistant strains of pneumococci are a cause for concern when using erythromycin. The quinolones currently available do not have sufficiently high activity against Streptococcus pneumoniae to allow them to be used as first-line empirical therapy [3], and indeed their use for this purpose has been heavily criticized [23]. In contrast, sparfloxacin has been shown to have good activity against a range of pneumococci, including penicillin-sensitive, penicillin-resistant, and multiresistant strains [5–9], whilst still retaining the usual broad spectrum of quinolones, which includes most respiratory pathogens.

In this trial all of the pneumococcal strains tested were susceptible to sparfloxacin (67), but two of 68 were resistant to and one of 68 had reduced susceptibility to erythromycin. In addition, eight of 56 strains had decreased susceptibility to penicillin, with four of them being resistant to penicillin (MICs 1 and 2 mg/L). Of the patients infected with these resistant strains, three were treated successfully with sparfloxacin and one with erythromycin. Although the rise in pneumococci with reduced susceptibility and resistance to penicillins gives cause for concern, it is debatable as to whether this is always reflected in a reduced response to penicillins in cases of moderate CAP, especially when the organism is of intermediate resistance. The evidence for a lack of response to penicillins in more severe infections, particularly meningitis, or in the immunocompromised, especially when caused by fully resistant strains, is better documented [22,24]. There are, however, claims that resistance to erythromycin can lead to a poor therapeutic outcome [22,24]; it is thus of interest to note that in this trial the two erythromycin-resistant strains were isolated from patients who were treated unsuccessfully with erythromycin.

Overall, there were too few isolates resistant to sparfloxacin or other agents to draw any firm conclusions regarding the value of sparfloxacin, although these results are encouraging. The individual cases of early failures observed in the patients treated with sparfloxacin are unexplained, except possibly for one patient in whom the serum level 1 h post-dosing was unusually low (approximately 0.02 mg/L). Trough levels of sparfloxacin following a daily dose of 200 mg are generally approximately 0.5 mg/L [16], so it is possible that the low levels of sparfloxacin in this patient may have been inadequate.

Since so few serum samples were available at the end of treatment or at follow-up, it was difficult to determine in many cases whether the pneumococcus had been eradicated. There were, however, no cases of persistent bacteremia in patients treated with sparfloxacin, but two cases occurred in the comparator group, one treated with erythromycin and one with amoxycillin/clavulanate. In addition, one HIV patient in the comparator group relapsed post-therapy and was bacteremic at follow-up.

Sparfloxacin was well tolerated, the number of patients reporting drug-related adverse events being less in the sparfloxacin group (15.8%) than in the comparator group (33.3%)). There were more gastrointestinal disturbances with comparator compounds than with sparfloxacin. Patients were advised to avoid exposure to the sun, and no cases of phototoxicity were reported.

In conclusion, this analysis of a subset of patients from four wider trials has demonstrated that sparfloxacin, given in a single daily dose, for an average of 10 days, is at least equivalent to standard antibacterial treatment for bacteremic pneumococcal pneumonia. Although some pneumococcal strains were resistant to penicillin or to erythromycin, all were sensitive to sparfloxacin. Sparfloxacin was better tolerated than the comparator drugs, with fewer gastrointestinal effects being reported. The drug's pharmacokinetic profile [16], its ability to penetrate cells and accumulate in respiratory tissues in high concentrations [17], and once-daily dosing make it an alternative candidate for the empirical therapy of moderately severe CAP.