Clinical response of patients with gonococcal endocervicitis and endometritis-salpingitis-peritonitis to doxycycline

Clinical response of patients with gonococcal to doxycycline GILLES R. G. MONIF, M.D. SCSAN L. WELKOS HERMAN BAER. M.D. Gaine.l<•ilfe. Florida The clinical response to single-drug therapy with doxycycline was evaluated in 25 patients with gonococcal endocervicitis and was correlated with the bacterial isolates present within the cul-de-sac. The 10 patients with gonococcal salpingitis and the three patients with gonococcal peritonitis exhibited excellent clinical responses. When polymicrobial infection coexisted with gonococcal peritonitis or functioned without the concomitant presence of Neisseria gonorrhoeae, a significantly altered therapeutic response was observed in four of the 12 patients. Four patients in this grouping exhibited either secondary temperature elevations or lack of a signifi~ant alteration of the white blood cell count, two features which were not characteristic of the patients with either gonococcal salpingitis or peritonitis. In the cases of polymicrobial peritonitis, there was a poor correlation between ensuing clinical response and in vitro resistance of one or more cul-de-sac isolates to doxycycline. The presence of a resistant organism did not preclude a good or satisfactory clinical response. The absence of a resistant organism correlated well with a good clinical response. (AM. J. 0BSTET. GYNECOL. 129: 614, 1977.)

THE BIOLOGICAL VARIABILITY of individual response to infection, particularly when more than one etiologic agent has been incriminated, t-:J the inability to assess antecedent underlying structural damage to the h!llopian tubes} the limitation of the routine utilization of surgical procedures (laparoscopy) to assess the magnitude of tubal occlusion." and the fact that one is dealing with a self-limiting disease have rendered relatively nebulous the criteria for evaluating the efficacy of therapy in patients with acute pelvic inflammatory disease. The development of relatively sophisticated anaen>bic technology. coupled with the use of cui-

docentesis, permits the evaluation of one of the probable variables governing therapeutic response to infection, namely, the bacteria present within the peritoneal cavity. the cul-de-sac, at the time of acute infection. This paper reports on the correlation betl\·een culde-sac bacterial isolates and therapeutic response to single-drug therapy in 25 patients with endometritissalpingitis-peritonitis in whom Neiswria gononho,·a,· was cultured from the endocenix of the cul-de-sac.

Material and methods Patient population. The study population was derived from women presenting to either the Shands Teaching Hospital in Gainesville or the Cniversity Hospital in jacksonville for medical care between November·. 1974. and May. 1976. The criteria f(H· inclusion were: (I) the presence of acute pelvic inflammatory disease (endocervicitis-endometritis-salpingitis) with physical signs of peritoneal involvement which was of sufficient severity to warrant hospitalization; (2) bacteriological evaluation of the cul-de-sac as a result of culdocentesis; and (3) the recovery of N. gonorrhol'at' from the endocervix and/or endometrial cavity. Therapy consisted of 200 mg. of doxycycline on ad-

From thl' Laboratmy of Infntious Disrases!Anaerobic Laboratur)' and the Departments of Obstrtrits and Cynemlogy and Microbiology at the Univenitv of Florida College ofMedicine. SupportNl in part by a grant from Pfizer Lab., Neu• York, ,Vew York. Reaivl'd for publicatwn Decembtr 27. I 976. Revised May 13, 1977.

A(fr:ptedjune Y. i977. Reprint rrquest.1: Dr. Gille.~ R. G. Monif. Department of ObstPlrirs and Gvnernlofrl. Universitv o( Florida Collef'e of Medirine, Cafnesvill;,~ Florida 3i6/o. o

614

Volume 1~9 !\;umber fi

mission followed by 100 mg. administered every twelve hours. The intravenous route was used exclusively for the first 48 hours. Chlam_vdia trarhomatis, Mycoplasma hmnini.1, and T-mycoplasma have all been recovered from the cul-de-sac of patients with acute pelvic inflammatory disease. 2 The use of a tetracycline reduced the number of nonbacterial infectious agents which may have been present in the cul-de-sac of the patients. Bacteriologic techniques. An endocervical culture was obtained with a sterile cotton applicator and plated directly on Thayer-Martin agar* (Shands Teaching Hospital) ur Lester-Martin agart (University Hospital). Culdocentesis was performed under direct visualization with a No. 18 gauge needle and syringe through a surgical field which had been dried with cotton drumsticks and washed with Betadine:j: solution. One to 4 mi. of the fluid obtained by culdocentesis was injected into two Ana port bottles.§ One of the Ana port bottles 11 as then sent to the hospital microbiological laboratorv, and the other was forwarded to rhe Anaerobe Laboratory of the Departments of Obstetrics and Gynecology and Microbiology. The specimen receiYed bv the hospital laboratory was immediately subcultured for N. gonorrlu)('at, where as that forwarded to the Anaerobe Laboratory was plated 24 to 48 hours after collection. Processing of the cul-de-sac specimens. The cul-desac specimens in the Anapnrt transport bottlf's werf' passed into an anaerobic chamber-glove box.!! The specimen was retrieved with a needle and syringe and plated on various selective rnedia. 6 In addition, thioglywllate broth~ containing hemin (5 f.Lg per millilitf'r) and menadione (0.5 f.Lg per milliliter) was inoculated. The plating media for anaerobic culture included: Columbia agar base# containing five per cent sheep blood, cvsteine hydrochloride (0.5 Gm. per liter), palladium chloride (0.33 Gm. per liter). and vitamin K-hemin solution (ABAP); ABAP supplemented with vancomycin hydrochloride (7.5 rng. per liter) and kanamycin sulfate (75 mg. per liter) and containing laked sheep blood; lecithin-lactose agar 7 ; Veillonella agar~; and Rogosa Selective Lactobacillus agar.~ All media for

*Pfizer Diagnostics. New York, New York. tTransgrow. Scott Laboratories. Inc., Fiskeville, Rhode Island. tPurdue-Frederick Co., Yonkers, New York. §Anaport, Scott Laboratories, Inc. !!Cov 1\!fg. Co., Ann Arbor, Michigan. ~ Difco Labs., Detroit, Michigan. #Baltimore Biological Laboratories. Cockeysville. Maryland.

Clinical response to doxycycline 615

anaerobic culture were placed in the reducing conditions of the glove-box or an anaerobic jar containing a Gas-Pak generator envelope* for at least 24 hours prior to use. All inoculated anaerobic cultun· plates were placed in an anaerobic jar; these \I ere incubated at 37° C. In addition, the cul-de-sac specimens were cultured for aerobic organisms on the following media: tiYe pet· cent sheep blood agar (BAP). MacConkey agar.t Pfizer Selective Enterococcus medium,:j: and chocolate agar or Thayer-Martin media. The inoculated media were incubated in air containing carbon dioxide at a temperature of 37° C. An aliquot of specimen in the transport container was stained by the Gram tnf'thod to directly determine the predominant morphologic types of organisms present. Identification of strains. Aerobic isolates were identified by the standard classification scheme." Isolated colonies from anaerobic plates were subcultured for identification beginning 48 hours after inoculation. The primary plates were re-examined at intervals up to !0 days after inoculation. Colonies were picked and transferred to ABAP for purity. to peptone-yeast extract glucose (PYG) broth, and to BAP for aerotolerance testing. The anaerobic isolates were identified following the methods of the Virginia polytechnic Institute Anaerobe Laboratory, 9 Morphologic, cultural, and biochemical properties were studif'd for iclentitication. Carbohydrate fermentation reactions were read with the use of a pH meter. 10 In addition. gas-liquid chromatography was performed on extracted PVC broth cultures to detect fatty acid production. A gas chromatograph with a thermal conductivity detector~ was used to detect fatty acid end products of metabolism. In vitro susceptibility testing: Doxycycline. Anaerobic and facultative anaerobic bacteria i-;olatnl from the cul-de-sac in patients were tested for their susceptibility to doxycy1inc by one of tV\'o agar dilution methods. The procedure used initially was a modification of the method of Sutter and associates. 11 Doxycycline (doxycycline hydrate+) was prepared as a I 00-f(Jid stock solution in sterile water and stored at -20° C:. The plating medium consisted of Columbia agar* adjusted to pH 7.2 and containing hemin (5 f.Lg per milliliter), menadione (0.5 f.Lg per milliliter), and five per cent laked sheep blood. Dilutions of the antibiotic were in*Baltimore Biological Laboratories. tDifco Labs. :j:Pfizer Diagnostics. §Anabac. Dohnnann Instruiitents Co., Houstoii,

1

t·xas

616

November 15, 1977 Am. J. Obstet. Gynecol.

Monif, Welkos, and Baer

Table I. GonoC
Duration of {evn (hr.)

27

37.7

12-!6

2

23

38.3

II

3

27

37.\l

4-6

Agl' of patient ()'T.)

Case.Vu.

of LMP* and onset of symptomology (days)

:'vfarked resolution of lendtrnrss in 24 to 48 hr.

Seriaii"''BCt Dati'

I

Respon.1e [\lo.!cu. mm.

lo therapy

Good

4

19

6

38.8

H

5

20

4

38.2

16

6

20

20

7

24

39.8

12-16

H

23

38.7

6-H

9/19 9/20

\l

18

39.1

16

·U23

16,600 9,200 10,000 6,700 4,300 12,500 8,400 17,900 7,900 15,600 7,600 13.100 5.700 13,400 6.200 11.:}00 8.300 14.500

4/25

7.~100

10

4

2/09 2!10

38

3 14

32

4/29 5/l 2/0tl

+

611

+

6/2 7114

7/16 8/23 H/25 Fi/18 '1/20

+

39.1

6/:}0

+

7/:{

22

15,800 9,600

8/30

+

Good Good Good (;ood Good Good GoOf!

Good Good

*LMP = Last menstrual period. tWBC = White blood cell count.

Table II. Gonococcal endocervicitis/endometritis/peritonitis (the latter defined as recoverv of N. gononhoeae alone or in conjunction with faculative anaerobes) Age o{

Interim bel'u.•een onset of

Cas1'

patient

No.

(yr.)

!vlarked Duration

resolution of

LMP and omet of

Highest temperature

offever

s~mptomology (days)

(o C.)

(hr.)

abdominal tenderness in 24 to 48 hr.

38.8

30

+

39.3

12-14

+

18 2

26 18

3 8

39

8-10

corporated into the molten agar medium before pouring to yield plates with antibiotic concentrations ranging from 0.19 to 25 Mg per milliliter. The plates were poured to a thickness of 5 to 6 mm. Strains to be tested were grown for 24 to 48 hours in thioglycollate medium 135-C,* to which 5 ~J-g of hemin per milliliter, 0.5 iJ,g of menadione per milliliter, and l mg. of NaHC(\ per milliliter were added after autoclaving. Inocula were adjusted to a No. 1 McFarland nephelometer standard; the concentrations of occasional strains were confirmed by colony counts performed on ABAP. The agar plates were inoculated *Baltimore Biological Laboratories.

+

SerialWBC Date

I

No. leu. mm.

4!11 4114 9/22 9/24

16,200 7,100 11,800 5,900

11/02

9,800 4,700

11/05

Cul-de-sm i.
Response to therapy

.V. gonorrhoea!'

Good

N. gonorrhoear,

Good

alpha-hemolytic streptococci N. gorwrrhoeae

Good

with I 0 4 to 105 viable cells by means of a replication apparatus,* which delivers 0.0025 mi. The inoculated plates were incubated in Gas-Pak jarst at 37° C for 48 hours. at which time the plates were read. The minimal inhibitory concentration (MIC) was defined as the lowest concentration which completely suppressed growth. With the assumption that levels of 2 to 4 Mg of doxycycline per milliliter of blood can be achieved, 12 · ta bacteria with a MIC value of 3.2 Mg per milliliter or less were considered resistant. Control plates were inoculated with each MIC de*Steers. tBaltimore Biological Laboratories.

Volume 129 Number ti

Clinical response to doxycycline 617

DAY 1

DATE HOUR

0

0

0 N 0

0

"' 0

0 0

0 ...;

0

0 -:1"

...;

DAY 2 0

0

0

0

co

N

...;

N

0 0 N

0

0

0

"' 0

..-f

0

0 0

0

0 -:1" ..-f

DAY 3 0

0

0

0 0

0 0

N N

N

..-f

"'

0

co

0

0

0

0 0

..-f

DAY 4

0

0

""'"

"'

0 ..-f

0

0 0

0 0

0

N N

N

..-f

"'

0

0

0

0

0 0

0 0

-:1"

..-f

..-f

0

0

a:>

..-f

0 0

N N

40° 39° 38°

Fig. I. Temperature curve of No. 2 from Table II (gonococcal endocervicitis). Neisseria gonorrhoeae and alpha hemolytic streptococcus were isolated from the cul-de-sac.

DAY 1

DATE HOUR

0

0

0 0

0

0 "'

N

0 0 0

..-f

0 0 -:1" ...;

DAY 2 0

0 0

co

..-f

0

N N

0 0

N

0

0 0

0 "'

0 0 0

..-f

0 0

'..-f""'

DAY 3 0 0

0 0

00 ..-f

N N

0 0

N

0

0

0

"' 0

0 0 0

.....

0 0 -:1" ..-f

DAY 4 0

0

0 0

0 0

...;

N N

N

a:>

0

0 0

"' 0

0 0 0

..-f

0 0

-:1" .....

0 0

0 0

..-f

N N

a:>

40° THERAPY

39°

ANTI ~CAA

38°

_/'

GE OF IOTICS

r\

/

[\

-

-- -- ------------- --

Fig. 2. Temperature curve of No. 2 from Table IV (gonococcal endocervicitis). Pmteus mirab1bs. Group B beta hemolytic streptococci, enterococci, Peptostreptococcus spp. No. l, hemolytic streptococci, Peptostreptococcus spp. No. 2, Peptococcus prevotii, Staphylococcus, coagulase positive, and Bacteroides spp. were isolated from the cul-de-sac.

termination. Anaerobic and aerobic plates without antibiotic were inoculated in order to test for the purity and viability of the inoculum. In addition, an anaerobic reference strain with a known MIC was included in each determination. Anaerobic strains to be tested for their susceptibility

to doxycycline by MIC determinations were stored in blood and frozen at -70° C. until susceptibility tests could be performed. The second agar dilution method used in the latter part of the study was the same as that recently described by Hauser and associates. 14 This procedure in-

618

!\iovembcr !:>. 1'177 Am. J. Obstet. (;ynerol.

Monif, Welkos, and Baer

Table III. Gonococcal endocervicitis/endometritis/peritonitis and polymicrobial peritonitis (the latter defined as the recovery of N. gonorrhoeae in conjunction with facultative and strict anaerobes) Ca.1e

,1ge of patient

JVo.

(yr.)

18

4

15

2

21

18

15

5

*GPNS

Interim between onset of LMP and onset of symptomology (days)

=

7

8

17

Highest temperature (0 C.)

Duration offever (hr.)

38.6 (secondary febrile spikes)

14-16

39.7

22-24

37.8

22

39

38.5 (secondary febrile spikes)

12-16

4

MarkRd resolution of' tendernes.1 in 24 to 48 hr.

+

+

+

Snia/IVBC

Date

I

.Vo.lcu. mm.

4/l'l 4/20 4/23

16,600 21.000

12/9 12/12

15,600 5.100

2/01 2/02

14,700 7,700

4/15 4/17

19,200 10,500

4/18

6,900

10/21 10/22

13300

24,000 10,200

Gram-positive nonsporulating.

corporates the use of quadrant plates and requires testing with fewer antibiotic dilutions. The basal medium was Mueller-Hinton agar enriched with 0.5 per cent yeast extract and 0.5 ml. of laked sheep erythrocytes. The concentrations of magnesium and calcium in the basal medium were determined by atomic absorption spectrophotometry. The resulting values for the lots of Mueller-Hinton agar and yeast extract employed were found to be similar to those recommended by D' Amato and associates. 15 Therefore, calciurn and r11agnesiur11 supplementation was not required. Dilutions of the antibiotic \AJere incorporated into the molten agar medium before it was poured into the quadrant plates. Seven final concentrations of doxycycline were tested (0.01, 0.1, 1, 2, 4, 8, and 16 /-Lg per milliliter). Two quadrant plates were poured with these dilutions, and one quadrant plate contained medium without antibiotic dilution to serve as a control. Inoculation, incubation, and examination of the plates were as described above. Response to therapy. The two parameters deemed indicative of a patient's response to therapy were: (1) the duration of the febrile response and (2) the time

interval required for the amelioration of the signs of peritoneal irritation and deep organ tenderness. The therapeutic response was judged good if the patient's fever had returned to a base-line temperature of 37.4° C., and there was significant change in the physical findings within the time reference (roughly comparable to that of patients with gonococcal peritonitis). A patient's response was considered satisfactory if the above criteria were met, even if there was a subsequent febrile elevation (this is defined by a temperature of 38° C. or greater on two consecutive readings taken four hours apart) after an initial defervescence of the presenting fever and subsequent absence of fever or if the white blood cell count taken within 48 to 72 hours exceeded 10,000 per cubic millimeter. If a patient had a fever of 38° C. for 36 hours or there was not significant resolution of abdominal findings within 36 hours, the antimicrobial therapy was deemed inadequate. Results

On the basis of the bacterioiogic findings derived from the cul-de-sac cultures, the patients with gonococcal endocervicitiS vvere subdivided into four

Volume 129 "'umber 6

Clinical response to doxycycline

Cul-de-sac isolates!MIC of doxvcycline Neissrria gonorrhoeae Bacteroides spp. No. 1 Bacteroides spp. No. 2 Enterococci Peptostreptococcus spp. Nriss!'ria gonorrhoeae Staphylococcus, coagulase negative Bacteroides spp. No. I Bacteroides spp. No. 2 a-hemolytic Streptococcus, not group D

gonorrhotJae Peptostreptococcus spp. Bacttroides pneumosiutes Corynebacterium spp. Gram-negative coccus GPNS* rods Eubacterium lentum N eissaia wmorrhoeru• Hmwphilus vaginalis

Coru:entrations ( p~g!mL)

Peptorocni.r~

pre-ootii

therap)' Inadequate

Died out 6.25 0.39 >25 Good

Died our Died out >16 Good

JVr~isseria

Eubacl1>rium aerofariens Bactcroide1 Jmgiiis, ss. fragilis .'\!r•is.1tria gmwrrh ome Eschairhia coli Bacteroides spp. No. I Bacteroides spp. No.2 Eubacterium spp.

Response to

I I I >16 I

Good 3.12 3.12 1.56 1.56 Satisfactory 4 >16 4

2

groups: Group I, patients with no organisms isolated from the cul-de-sac; Group 2, patients from whom N. gonorrhome alone or N. gonorrhoeae in conjunction with an aerobic organism was isolated; Group 3, patients from whom N. gonorrhoeae and multiple facultative (Classes I and II) and/or potentially obligatory (Classes

II and III) anaerobic bacteria 'vere isolated; Group 4, patients from whom only multiple facultative and/or potentially obligatory anaerobic bacteria were isolated. Group 1. Gonococcal endocervicitis-endometritissalpingitis (Table I). The 10 patients with gonococcal endocervicitis and presumably endometritis and salpingitis whose culdocentesis yielded no fluid on repeated attempts or clear fluid demonstrated lysis of fever within 36 hours and loss of rebound tenderness and significant lessening of pelvic organ tenderness on deep palpation within 24 to 48 hours. The culdocentesis tap of No. 10 was purulent; however, microscopic examination of Gram-stain smears failed to demonstrate the presence of any bacteria, and no isolates were achieved from either of the Ana port cultures. This patient had a more protracted febrile course. The onset of symptomotology was 14 days from the last

619

menstrual period. Nine of the 10 patients in this category had elevations of the white blood cell counts. The white blood cell count returned to a normal level or dropped significantly within 48 to 72 hours in all cases in vvhich the observations \vere made. In the four instances in which the white blood cell count was repeated within 24 hours, significant reduction in the total count was demonstrated in each case. Four of the 10 patients in this category had a prior history or a documented episode of acute pelvic inflammatory disease. Three of the patients utilized oral contraceptives, three used an intrauterine contraceptive device, and the remaining patients utilized no mode of contraception. Group 2. Gonococcal endocervicitis-endometritissalpingitis-peritonitis (Table II). All three patients in this category demonstrated lysis of fever within 36 hours and loss of rebound tenderness and significant lessening of organ tenderness within 24 to 48 hours. Two of the three patients had elevated white blood cell counts. In all three cases the white blood cell count dropped significantly \vi thin 48 to 72 hours. Tv;o of the three patients had a prior history of acute pelvic inflammatory disease. Only one of tht> patients was practicing contraception (oral contraceptives). Group 3. Gonococcal endocervicitis-endometritissalpingitis-peritonitis plus polymicrobial peritonitis (Table III). All five patients in this category exhibited lysis of fever within 36 hours; however, two of the five had secondary febrile spikes, which were not observed in the previous two groups. In Case No. I, significant amelioration of physical findings did not occur within the first 48 hours, nor did the white blood cell count change appreciably over approximately an 82 hour period. Two of the three patients had an antecedent history of acute pelvic inflammatory disease. Only one patient utilized contraception (oral contraceptives). Group 4. Gonococcal endocervicitis and polymicrobial peritonitis (Table IV). T\AlO of the seven patients in this category had a protracted febrile course beyond 36 hours (Nos. 4 and 6). Three of the seven patients did not exhibit significant amelioration of physical findings on abdominal examination within the first 48 hours. In Nos. 2 and 7, secondary febrile spikes occurred. The white blood cell counts of Nos. 1, 3, 6, and 7 did not show a reduction comparable to that observed for patients in Groups I and 2. Three of the seven patients had a prior history or documented episodes of acute pelvic inflammatory disease. None of the patients practiced any mode of contraception. Clinical profiles. The patients with either gonococcal endocervicitis-endometritis-salpingitis and those with concomitant gonococcal peritonitis differed from

620

November I:,, I
Monif, Welkos, and Baer

Am.

J.

Obstet. (;ynecol.

Table IV. Gonococcal endocervicitis and polymicrobial peritonitis Cas1•

No.

2

Age of

Interim between onset

patient (yr.)

ofLMP and on-

set ofsymptomolog_~ (days)

Highest ternperature (0 C.)

19

14

39.H

15

6

38.8

Duration of} ever (hr.)

Afarked rnolulion of tend,'m''" in 2-1 to -18 hr.

+

35 (Secondary spike)

3

18

II

37.7

16

·1

18

12

37.8

44

5

18

26

38.8

8-12

6

31

7

37.H

7

36

4

37.8 (Secondary spike)

24-28

+

+

Serial WBC

Date 7/17

I

.Vo./ru. mm.

7/l H

12.700 11.600

9/7 9/8

16,000 7,200

8/16 H/17 8/18 H/19

18,200 15,100 13,800 9,700

4/23 4/27

17,900 I 0,400

IIIlO llll4

11,300 11,800

12/9

13,300

12/ll

10,500

11/10 11/12

11,100 9,600

Volume 12\l Number 6

Clinical response to doxycycline 621

Cul-de-sac isolates!MIC/of doxyc.vcline

Concentrations (p.g/ml.)

Bactnoid,,s pneumosintes Peptocorrus asaccharolyt-

>25 12

Response to theraP.v

Good

zrus

Peptostreptococcus spp. Peptococcus variabilis

GPNS rods

Proteus mirabilis

Group B beta hemolytic streptococcus Enterococci Peptostreptococcus spp. :-.lo. 1 a-hemolytic streptococci Peptostreptococcus spp. 1\:o. 2 PejJlococnts preJiolii

Staphylococcus, coagulase positive Bacteroides spp. Enterococci Staphylococcus, coagulase negative Peptostreptococcus spp.

Bactaoidl'l fragili.l

GP:\'S rods

Clostridium malerwminatum

0.1

12.5

Died out Swarming >25 4

>25 >25

Inadequate: secondary spikes, Rxdswitched to cephalosporin and kanamycin

1.56

0.39 1..56 12.5 1.56 >25 0.7R

Inadequate: switched to penicillin and kanamycin

0.7R

>25

Comment

1

a-hemolytic streptococcus, not Group D

Inadequate

Haemophi/w vaginali.1

GPNS rods Veillonella Veillonella pa.rvuuJ

Peptococcus morbillorum

GPNS rods

Peptorocn1s anaerobiw

Microaerophilic streptococc·us

>16 I

4 >16 >16 4 >16

Escherichia coli Bifido bactrrium longum Bifido bacterium adolescentis Bacteroides fragilis,

Not done

Bartnoides fragilis,

Died out

ss. vulgatus

ss. unknown Bacteroides spp.

Eubactt>riu m it'IIW' Peptocoaus asaccharolyticus No. I Peptocoau.< asaccharo/ytim' No. 2 St~phylococcus, coag-

ulase negative Group B beta hemolytic streptococcus Microaerophilic corvnebacterium

those with polymicrobial infection in two minor respects. First, nine of the 13 patients with gonococcal infection had an interim of four days or less between the onset of the last menstrual period and the onset of the clinical symptomology. In contrast, three of the 12 patients with polymicrobial peritonitis did, at the onset of the disease, have a comparable proximity to the menstrual period. Second, while eight of the 13 patients with nonpolymicrobial peritonitis used some form of contraception, only one of the 12 patients with polymicrobial peritonitis had utilized a form of contraception. Correlation of in vitro resistance with clinical response (Tables I to IV). With the exceptions of No. 4 of Group 3 and No. 6 of Group 4, whose evaluations were incomplete because .three key cultures died out prior to in vitro testing, in the cases of polymicrobial infection there was at least one bacterial isolate which exhibited significant in vitro resistance to doxycycline. The in vitro resistance of one or more cul-de-sac isolates to doxycycline did not preclude a good or satisfactory clinical response; however, the absence of a resistant organism, as in the case of gonococcal peritonitis, correlated with a good clinical response.

2 I

Died out

Died out Died out 1

16 >16 16

Peptoc:ocrus preJiotii

Peptostreptococcus

16 16

spp. Lactobacillus acidophilus

2

Good

Inadequate: switched to clindamycin and gentamicin

Satisfactory

When monoetiologic disease resulting from N. gonorrhoeae was dealt with, the response to doxycycline was prompt and efficacious. The occurrence of polymicrobial peritonitis in which one or more of the bacteria exhibited in vitro resistance to doxycycline appeared to alter the therapeutic response to doxycycline. The data on 10 patients with gonococcal endocervicitis-endometritis-salpingitis and on the three patients with gonococcal endocervicitis-endometritissalpingitis-peritonitis were interpreted as the anticipated therapeutic base line (Fig. 1). In the studv material, patients who exhibited febrile spikes after initial lysis of the fever or inadequate response to therapy invariably had a polymicrobial infection within the cul-de-sac (Fig. 2). One major reservation which must be maintained in interpreting the significance of a given bacterial isolate from the cul-de-sac is that the data reported arc qualitative and not quantitative. At any given time, one organism may be of more significance than another for the presumed bacterial synergism. 3 Chow and associates, 16 in their study on the bacteriology of acute pelvic inflammatory disease and clinical response to parenteral doxycycline, similarly could not demonstrate a clear-cut correlation between clinical response and in-vitro susceptibility of the cul-de-sac isolates. The criteria for ciinicai efficacy were resolution of symptoms and signs within 48 hours. However,

622

'JoYemhc·t 10. l\l7i

Monif, Welkos, and Baer

in their study, five patients had what was termed a good response "with resolution of symptoms and signs within four days." 16 These patients would be deemed therapeutic failures by our criteria. If one accepts the contention that there is an anticipated base-line response to pelvic infection with N. gonorrhoeae, then one is confronted with possible significance of an altered therapeutic response. The critical site of inflammation in patients with salpingitis-peritonitis is interstitial rather than intralumifial.17 As in patients with pyelonephritis, the

Am. J. Ohstet. Cmecnl.

consequences of basement membrane destruction secondary to the inflammatory process within the interstitium is healing by fibrosis. Following a single episode of acute pelvic inflammatory disease. 12.H per cent of women have demonstrable tubal ocdusion. 1 " After two episodes, the percentage increases to cEi.3 per cent. It is not improbable that the prolonged infectious morbidity rate observed in patients \\·ith polymicrobial peritonitis may translate into augmental structural alteration of normal fallopian tubal anhitecturc.

REFERENCES I. Chow, A. W .. Malkasian, K. L., Marshall, J. R., and Guze,

2.

3.

4. 5. 6.

7. 8. 9.

10.

L. B.: The bacteriology of acute pelvic inflammatory disease. Value of cul-de-sac cultures and relative importance of gonococcal and other aerobic or anaerobic bacteria, AM. J. 0BSTET. GYNECOL. 123: 876, 1975. Eschenbach, D. A., Buchanan, T. M., Pollack, H. M., Forsyth, P. S., Alexander, E. R., Lim, J. S., Wang, S. P., Wentworth, B. B., McCormack. W. M., and Holmes, K. K.: Polymicrobial etiology of acute pelvic inflammatory disease, N. Engl. J. Med. 293: 166, 1975. Monif, G. R. G., Welkos, S. L., Baer, H., and Thompson, R. J.: Cul-de-sac isolates from patients with endometritis-salpingitis-peritonitis and gonococcal endocervicitis, AM. J. 0BSTET. GYNECOL. 124: 838, 1976. Jacobson, L.: Laparoscopy in the diagnosis of acute salpingitis,· Acta Obstet. Gynecol. Scan d. 43: 160, 1964. Rees, E., and Annels, E. H.: Gonococcal salpingitis, J. Am. Vener. Dis. Assoc. 45: 205, 1969. Ellner, P. D., Granato, P. A .. and May, C. B.: Recovery and identification of anaerobes: A system suitable for the routine clinical laboratory, Appl. Microbiol. 26: 904, 1973. Ellner, P. D., and O'Donnell, E. D.: A selective differential medium for histotoxic clostridia, Am . .J. Clin. Pat hoi. 56: 197. 1971. Lennette, E. H., Spaulding, E. H., and Truant, J. P.: Manual of Clinical Microbiology. ed. 2, Washington. D. C., 1974, American Societv for Microbiolo!!:v. Holdeman, L. V., and Moo're, W. E. C., edit~~s: Anaerobe Laboratory Manual, Blacksburg, Virginia, 1975, Virginia Polytechnic Institute and State University Anaerobe Lahoratory. Fay, D. F .. and Barry, A. L.: Methods for detecting car-

11.

12.

13.

14. 15.

16.

17.

18.

bohydrate fermentation hy gram-negative nonsporeforming anaerobes, Appl. Microbiol. 27: 6tH. 1974. 1974. Sutter, V. L., Kowk, Y-Y., and Finegold, S. l\1.: Standardized antimicrobial disc susceptibility testing of anaerobic bacteria. I. Susceptibility of B. fragih, to Tetracycline, Appl. Microbiol. 23: 268, 1972. Webb, C. D.: In vitro susceptibility testing with doxycycline and tetracycline: Doxycycline: Recent irwestigations and clinical experiences. New York, 1975. Pfizer, Inc .. pp. 10-16. Leibowitz, B. J., Hakes, J. L., Calm, M. l\1., et al.: Doxycycline blood levels in normal subjects after intravenous and oral administration, Curr. Ther. Res. 14: R20. 1972. Hauser, K. J., Johnston, J. A .. and Zabransky. R. J.: Economical agar dilution for susceptibility testing of anaerobes, Antimicrob. Agents Chemother. 7: 712, 1975. D'Amato, R. F., Thornsberry. C., Baker, C. :\'.,and Kirven, L.A.: Effect of calcium and magnesium ions on the susceptibility of pseudomonas species to tetracycline, gentamicin, polymixin B. and carbenicillin. Antimicrob. Agents Chemother. 7: 596, 197 5. Chow, A. W .. Malkasian. K. L., Marshall,]. R .. and (;me. L. B.: Acute pelvic inflammatory disease and clinical re· sponse to parenteral doxycycline, Ant;microb. Agents Chemother. 7: 133, 1975. Schwarz, R. H.: Acute pelvic inflammatory disease. in Monif, G. R. G., editor: Infectious Diseases in Obstetrics and Gynecology, Hagerstown, Maryland, 1974, Harper & Row, Publishers, pp. 384-387. Westrom, I..: Effects of acute pelvic inflammator} di;.ease on fertility, AM. j. 0BSTET. GYNECOL. 121: 707, 197!\.