Relatively few antibiotics are effective in the treatment of chlamydial genitourinary infections, and many are associated with adverse reactions. Tetracyclines, including minocycline and doxycycline, have excellent in-vitro antichlamydial activity [2], and all have been used successfully in men and women with urethritis and cervicitis. The tetracycline analogs, doxycycline hyclate and minocycline hydrochloride, have longer serum half-lives than tetracycline and allow less frequent administration [3]. The advantage of less frequent administration has been shown in a study comparing doxycycline with tetracycline for the treatment of gonococcal infections. Patient compliance improved with doxycycline, accounting for lower failure rates in patients treated with this drug [3]. If drug administration could be further reduced to a single daily dose, one could expect compliance to increase further.

No data exist to support the rationale for a course of tetracycline therapy exceeding 7 days, although treatment of at least 5 days is necessary to effectively eradicate these slow-growing organisms [2]. Studies using 14 or 21 days of therapy have reported similar rates of recurrence if follow-up is continued for 6 weeks. Adverse effects are associated with all tetracyclines, but vestibular toxicity is a special concern with minocycline. These s

Doxycycline, twice a day for 7 days, is currently considered the "drug of choice" for nongonococcal urethritis and mucopurulent cervicitis. Our randomized, double-blind study was designed to compare the efficacy and tolerability of minocycline, 100 mg nightly compared with doxycycline, 100 mg twice daily, each for 7 days in men with nongonococcal urethritis and women with mucopurulent cervicitis.

Methods

Top Methods Results Discussion Author & Article Info References

Study Criteria

We conducted our study at the Sexually Transmitted Disease Clinics in Edmonton, Alberta and Vancouver, British Columbia. Men visiting the clinics with complaints of urethral discharge or dysuria for less than 1 month whose urethral smear showed 4 or more polymorphonuclear leukocytes per high-power field (x 1000) in five or more fields were eligible for study entry. The urethral smear also had to be negative for intracellular gram-negative diplococci. In addition, male contacts of women with mucopurulent cervicitis or with cervical C. trachomatis infections were also considered eligible regardless of whether urethritis was documented.

Women who were diagnosed as having mucopurulent cervicitis were eligible. The diagnosis of mucopurulent cervicitis required both the presence of endocervical mucopus and 10 or more polymorphonuclear leukocytes per high-power field in a Gram stain of cervical mucus that contained fewer than 100 vaginal squamous epithelial cells when collected from a nonmenstruating woman. The presence of mucopus was determined by an off-white or yellow endocervical mucopurulent discharge when viewed on a white-tipped swab. In addition, female contacts of men with nongonococcal urethritis or with urethral C. trachomatis infections were also considered eligible regardless of whether mucopurulent cervicitis was documented.

All study participants were older than 16 years, were willing to attend the clinic regularly for 7 weeks, had not received antimicrobial agents with anti-chlamydial activity in the preceding month, had negative cultures for gonorrhea, and were not allergic to tetracyclines. Women were required to use an effective form of birth control and could not enter if they were pregnant or breast feeding. Women with vaginitis, vaginosis, or pelvic inflammatory disease were excluded, as were men with orchitis or epididymitis.

Examination and Laboratory Studies

At the pretreatment evaluation, patients had a standardized interview and genital examination. In men, urethral discharge was classified as difficult to milk through to heavy spontaneous. Urethral exudate was examined by Gram stain and cultured on Chocolate and Thayer-Martin medium to exclude Neisseria gonorrhoeae. A calcium alginate urethrogenital swab was inserted into the urethra 1 to 2 cm beyond the fossa navicularis and left in place for 10 to 15 seconds. It was then placed in chlamydia transport media and immediately inoculated in McCoy cells or frozen at –70°C for subsequent isolation of C. trachomatis in McCoy cells [4]. The same swab was then used to isolate genital mycoplasmas [5]. At the initial visit only, rectal and pharyngeal cultures were also obtained for N. gonorrhoeae. At each subsequent visit, the urethral Gram stain as well as urethral cultures for chlamydia and the genital mycoplasmas were repeated. All men were asked not to void for 4 hours before their clinic visit.

Women had a pelvic and bimanual examination at each visit. Endocervical discharge was described by color and quantity. Ectocervical erythema, edema, and friability (induced mucosal bleeding) were recorded as well as the area of ectopy. An endocervical Gram smear was obtained at each visit, as was a vaginal wet mount and pH. Endocervical, urethral, rectal, and pharyngeal cultures for N. gonorrhoeae were obtained at the initial visit only. Urethral and endocervical swabs were obtained at each visit for C. trachomatis isolation. Finally, urethral and vaginal cultures for the genital mycoplasmas were obtained at each visit.

Follow-up

Patients were asked to return 14 ± 3, 28 ± 5, and 49 ± 7 days after the start of therapy. All study visits for women were scheduled so as to fall at nonmenstrual times of their cycle. At each visit, patients were questioned regarding symptoms, adverse effects of therapy, and interim sexual activity.

Study Drugs

At each site, patients were randomly assigned in a double-blind fashion to one of two treatment groups. Each patient received a blister package of 14 identically appearing capsules on the first visit. There were seven vertical columns and two horizontal rows. Each horizontal row was clearly labeled am or pm. Patients were instructed to take one capsule in the morning and evening. The blister packages were returned at the first follow-up visit. The two treatment groups were minocycline, 100 mg nightly, and doxycycline, 100 mg twice daily, each administered for 7 days.

Evaluation

The number of polymorphonuclear leukocytes in urethral or endocervical Gram-stained smears was used as an objective measure of response. The primary end point was a clinical cure defined as clinical symptoms subsiding or resolving by day 14 with subsequent complete resolution of symptoms that did not reappear during the remainder of the follow-up period. A secondary end point was bacteriological cure for C. trachomatis, defined for a patient with chlamydia as a microbiological cure at day 14 ± 3 and at all subsequent follow-up visits.

A clinical symptoms and signs score was calculated at each visit. Each symptom and sign was assigned a score from 0 to 3 (absent, mild, moderate, or severe). For men, symptoms included urethral discharge, urethral itch, and dysuria with signs consisting of both quantity and quality of urethral discharge. Symptoms in women included vaginal discharge, external and internal dysuria, deep and introital dyspareunia, abnormal bleeding, and lower abdominal pain, whereas signs comprised ectocervical erythema, edema, and friability. Side effects were elicited by asking the patients if they felt different in any way since beginning the medications.

Statistical Analysis

The Cochran-Mantel-Haenszel test was used in the combined gender group to compare rates in the two treatment groups; the Fisher exact test was used when men and women were considered separately. Logistic regression analysis was performed on the incidence of adverse drug reactions to determine significant factors within each treatment group. The paired t-test was used for analysis of the change from baselines in clinical symptoms and signs scores within each treatment arm. An analysis of covariance adjusting for the baseline score was used to test for differences between treatment arms at each return visit.

Results

Top Methods Results Discussion Author & Article Info References

Initial Microbiological Results

A total of 253 patients were enrolled between March 1989 and October 1990 (151 men and 102 women). The mean ages for men and women were 28.2 years (SD, 8.3) and 23.5 years (SD, 5.1), respectively. Eighty-eight percent of the patients were white and 97% were heterosexual. The mean number of sexual partners in the past year was 6 for the men (range, 1 to 50; SD, 8) and 2.7 for women (range, 1 to 11; SD, 1.7). Demographics and baseline microbiologic results did not differ for the two treatment groups or for the two study centers.

The results of initial cultures for C. trachomatis and the genital mycoplasmas are shown in Table 1. An equal proportion of men and women (31% and 40%) had C. trachomatis isolated. More women than men had either U. urealyticum or Mycoplasma hominis isolated. No organisms were isolated from 55 (37%) of the men enrolled, whereas only 10 (10%) women had negative cultures.

Genital mycoplasmas were isolated from 30% (14 of 46) of men with and 47% (48 of 103) of men without C. trachomatis (P = 0.07). For women, the corresponding figures were 93% (37 of 40) and 84% (51 of 61) for those with and without C. trachomatis, respectively (P = 0.2).

Of the 253 patients enrolled, 162 (64%) completed the study. Follow-up at the day 14 ± 3 visit occurred for 87% of study entries. This proportion decreased to 75% on day 28 ± 5 and to 64% on the final day 49 ± 7 visit. The proportion of individuals followed in each treatment group was equal (doxycycline, 61%; minocycline, 67%; P = 0.2).

Clinical Response to Therapy

Table 2 shows that the rates for clinical cure differed little between the two treatment groups (men: doxycycline, 82%; minocycline, 88%; 95% CI, –9% to 22%. Women: doxycycline, 90%; minocycline, 91%; CI, –16% to 19%. Combined: doxycycline, 85%; minocycline, 89%; 95% CI, –7% to 16%).

Men

Of the 151 men entered, 129 (85%) had a urethral smear that showed more than four polymorphonuclear leukocytes per high-power field in five or more fields. The remaining 22 patients were entered on the basis of disease in their sexual partner. Men with a microscopic diagnosis of nongonococcal urethritis had higher symptom and sign scores at baseline (P < 0.001 for both). They were also more likely to have a positive urethral culture for C. trachomatis alone (32 of 127 [25%] versus 0 of 22; P = 0.004) or either or both of C. trachomatis and the genital mycoplasmas (82 of 127 [65%] versus 12 of 22 [55%]; P > 0.2) than the 22 patients with no evidence of nongonococcal urethritis.

Follow-up at day 14 ± 3 occurred in 110 men (Table 2). Twenty-five patients had persistent urethritis; 26% in the doxycycline group versus 19% in the minocycline group (CI, –10% to 25%). This figure remained stable at the subsequent two follow-up visits. The two treatment groups did not differ at any of the three follow-up visits (P > 0.2 for all). At the end of the study, 28% and 31% of men in the doxycycline and minocycline groups, respectively, continued to show a substantial polymorphonuclear leukocyte response.

Women

Of the 102 women entered, 81 (79%) had more than 10 polymorphonuclear leukocytes in an endocervical smear. Although symptom and sign scores did not differ at baseline (P > 0.2 for both), women with a positive smear were more likely to have chlamydia isolated (39 of 81 [48%] versus 1 of 21 [5%]; P < 0.001). Of the 52 women with mucopurulent cervicitis, 49 (94%) had a positive culture versus 42 of 49 (86%) women without mucopurulent cervicitis.

In women, a striking 59 (80%) continued to have a Gram-stain diagnosis of cervicitis at the first follow-up visit. By day 49 ± 7, this percentage had decreased to 62% and 69% for the minocycline and doxycycline groups, respectively. As in the men, the two treatment groups did not differ at any of the follow-up visits (P = 0.08, P = 0.2, and P > 0.7).

In direct comparison to the persistence of cervical polymorphonuclear leukocytes, endocervical mucopus responded more rapidly. At the first follow-up visit, 12% and 4% of women who were administered minocycline and doxycycline, respectively, continued to have endocervical mucopus (P > 0.2, CI, –10% to 27%). By the third follow-up visit, none of 19 women receiving minocycline and only 2 of 20 women receiving doxycycline continued to show this finding (P > 0.2).

Symptoms and Signs Scores

Figure 1 outlines the response of symptoms and signs based on the mean scores. For men, treatment groups did not differ at any time (P > 0.2). Symptoms and signs both responded dramatically to treatment and scores were then stable at all follow-up visits. The stable score of 2 for signs at follow-up is probably not an indication of significant residual inflammation. The grading for signs in men was urethral discharge quantity 1 (difficulty milking) to 3 (heavy spontaneous) and quality 1 (clear) to 3 (green-yellow). Therefore, men who returned for follow-up and were virtually asymptomatic may still be expected to have a minimal clear urethral discharge that was evident after milking. This finding would result in a sign score of 2.

View larger version (14K): [in this window] [in a new window]   Figure 1. Mean scores and signs. Top. Men. Bottom. Women. Scores are shown at baseline and for each follow-up visit. The numbers above the bars (first one indicated by an asterisk) indicate the standard error of the adjusted mean.

The symptoms and signs in women responded just as quickly by day 14 and then gradually tapered to the last follow-up visit. The only statistically significant finding occurred at day 28 when the change in signs score from baseline was greater in the minocycline treatment group ( –89% versus –74%,P = 0.03). However, this significance disappeared by day 49 (P > 0.2).

Microbiological Response to Therapy

Cultures for C. trachomatis from all but two men and one woman were negative throughout the follow-up period (CI, –5% to 15%) (Table 3). Follow-up cultures were negative at day 14 for Ureaplasma urealyticum in 85% and 79% of men who had received minocycline and doxycycline, respectively. This percentage decreased to 63% for both groups on day 28 and deteriorated at the final day 49 ± 7 visit. Because the numbers of men with persistent or recurrent urethritis were small at each follow-up visit, it was not statistically valid to calculate whether this occurred more frequently in men with U. urealyticum.

An equal proportion of women in each treatment group had negative cultures for U. urealyticum at day 14 (P > 0.2), but by day 49 only 44% and 53% of the patients in the minocycline and doxycycline groups continued to have negative cultures (P > 0.2).

The number of men with M. hominis was too small to draw any conclusions. The proportion of women with persistently positive cultures did not change over time, and the two treatment arms did not differ by day 49 (P > 0.2).

Recurrence in Relation to Sexual Exposure after Therapy

A similar proportion of men and women (38% and 48%) had unprotected sexual contact at some point during their participation in this study. Unprotected sexual contact was not shown to affect clinical cure rates. An equal proportion of patients in each treatment arm engaged in unsafe sexual contact with similar rates of cure. In men, 82% of those who either abstained or used condoms were cured compared with a 90% cure rate for those who had unprotected contact (P > 0.2). However, the reverse occurred in women, with a 97% cure rate for the no-sex or safer-sex group versus an 85% cure rate for the group who had unprotected contact (P = 0.2). Similarly, sexual contact had no identifiable effect on the reisolation of either of the genital mycoplasmas.

Side Effects

Of the 253 patients initially entered, 237 (94%) were evaluable for toxicity (111 in the minocycline group and 126 in the doxycycline group). Table 4 outlines the side effects voluntarily reported by the patient or elicited by direct questioning. The most significant adverse effect was gastrointestinal upset, consisting of nausea, vomiting, abdominal pain or cramps, diarrhea, dyspepsia, increased flatulence, and taste alteration. Vomiting alone (P = 0.004) or any gastrointestinal upset including vomiting (P < 0.001) occurred significantly more often in the doxycycline group. The occurrence of dizziness or other adverse effects did not differ between the two groups (P > 0.1). When adverse effects were examined by drug regimen and gender, significance was again shown. Men and women who were administered doxycycline reported side effects more frequently than those in the minocycline group (P = 0.05 and 0.009, respectively).

Results of the logistic regression analysis confirmed that the difference between the two treatment groups in the incidence of adverse drug reactions could be primarily attributed to the study drugs. In women, concomitant medication (mainly oral contraceptives) was a second factor in predicting an adverse reaction.

Patients reported the adverse effects as mild in 77% of cases, moderate in 14%, and severe in the remaining 9%. The occurrence of adverse drug effects was responsible for five patients withdrawing from the study; two received minocycline and three received doxycycline. One man who received minocycline developed severe depression and mood swings associated with headaches and bad dreams. The women who withdrew from the minocycline group reported severe nausea, indigestion, and flatulence. The only woman who discontinued doxycycline therapy reported severe nausea and indigestion. The two men withdrawn after receiving doxycycline complained of moderate penile edema and numbness in the extremities and lips as well as some shortness of breath.

Discussion

Top Methods Results Discussion Author & Article Info References

In this study, results of clinical and laboratory variables with minocycline administered as a single dose of 100 mg nightly were virtually identical to those of doxycycline, 100 mg twice a day, each given for 7 days in the treatment of both chlamydia-positive and -negative nongonococcal urethritis and mucopurulent cervicitis.

Both drugs achieved similar clinical cure rates in men at day 49 ± 7 (28% and 32%). These results are similar to those of Bowie and colleagues [6], whose study showed that 7 days of single-dose minocycline was equally effective as 21 days. Although many published reports have appeared of therapeutic modalities for nongonococcal urethritis and mucopurulent cervicitis, few have had an adequate follow-up to 6 weeks. Suppression of chlamydial infection may occur during antimicrobial therapy without complete eradication. Assessment of efficacy immediately after therapy may falsely indicate cure. Microbiological evaluation at a longer interval after completion of therapy may better detect the presence of continued infection, although exogenous reinfection will be difficult to distinguish from suppressed infection. We were able to follow 64% of patients initially enrolled to the final follow-up visit at day 49 ± 7. Considering that the patient population was recruited from sexually transmitted disease clinics, this figure is considered very good. Because organism isolations did not differ for the two treatment groups and equal numbers of participants returned in both groups, we do not think that our results are compromised. The individuals who failed to return were in all likelihood asymptomatic and if they had returned, cure rates may have been even higher. In our study, sexual activity was not an apparent factor for relapse. The three individuals from whom C. trachomatis was reisolated at one of the follow-up visits had all engaged in unprotected sexual contact, but all their partners had recently received therapy with an antichlamydial antibiotic.

Other studies have shown that minocycline eradicated C. trachomatis from the male urethra with failure rates of 1% to 8% [7, 8]. However, urethritis recurred in up to 24% of men after 3 weeks of follow-up [7]. Cervical C. trachomatis was eradicated in 100% of women reported in the literature, but neither of these studies followed women with cervical Gram stains [2, 7, 9]. The reisolation of ureaplasma after apparently successful therapy suggests that the drug may have initially suppressed the growth of U. urealyticum without eradicating it. The same conclusion can also be drawn for M. hominis.

The greatest resolution of urethritis based on Gram smear was seen at day 14. By day 49 ± 7, the recurrence rate had increased from 19% to 32% in the minocycline group but only from 26% to 28% in the doxycycline group.

The results in women are more difficult to interpret. The original research on mucopurulent cervicitis published by Brunham [10] reported a strong and independent correlation between endocervical mucopus and the presence of 10 or more polymorphonuclear leukocytes in an endocervical smear with cervical C. trachomatis infection. Moscicki and colleagues [11] found that the presence of five or more polymorphonuclear leukocytes was the only significant factor associated with cervical chlamydia. However, our study shows that at day 49 ± 7, 62% to 69% of women continue to have more than 10 polymorphonuclear leukocytes in an endocervical smear despite the absence of chlamydia and, in all but two cases, endocervical mucopus. These findings suggest that 10 polymorphonuclear leukocytes is too low a cutoff for the presence of cervicitis; and we would be cautious in the over-interpretation of an endocervical Gram stain. Perhaps a cutoff of 15 to 20 polymorphonuclear leukocytes would be more appropriate in this group. Unfortunately, the confidence placed in the urethral Gram stain cannot be extrapolated to women. It is also extremely important to exclude any vaginal contamination of cervical smear. This was certainly achieved in our study as there were no smears grossly contaminated by vaginal epithelial cells or vaginal flora. Alternatively, perhaps the number of polymorphonuclear leukocytes is used appropriately to make a decision whether to treat or not initially but should not be used after treatment.

Adverse reactions are reported to vary from 3% to 15% for minocycline [7, 12] and from 8% to 35% for doxycycline [1, 13-15]. Vestibular toxicity is by far the most common significant adverse reaction reported for minocycline. In our study, patients receiving doxycycline experienced significantly more adverse drug reactions than did those who were randomized to receive minocycline. Although 30% of patients receiving minocycline, 100 mg at bedtime, did report adverse drug reactions, almost all were related to gastrointestinal upset and were mild. Only four women reported mild vestibular toxicity. We therefore think that the administration of a single 100-mg capsule of minocycline at bedtime largely overcomes the vestibular toxicity historically associated with this drug. Gastrointestinal upset was the major problem with doxycycline, although again, most reactions were mild.

Infection or colonization with the genital mycoplasmas was strikingly different for men and women. Cultures for ureaplasma initially became negative in 74% to 85% of patients; however, by day 49 they were positive again in 41% to 56% of patients. Mycoplasma was not initially isolated in a large proportion of men, but all with initial positive cultures were culture negative at the first follow-up visit. However, success of organism eradication in women did not vary during the follow-up. No significant difference in eradication of either of these organisms was seen regardless of which antibiotic the patient had received.

Although the tetracyclines are active against C. trachomatis and no clinically significant antibiotic resistance has yet been identified in this bacterial species, new antimicrobial regimens are required to improve patient compliance. The findings of this study support the administration of minocycline 100 mg as a single nightly dose for 7 days in the therapy of both culture-negative and -positive nongonococcal urethritis and mucopurulent cervicitis. Patient convenience with single nightly dosing would also be expected to improve compliance. Further progress in the therapy of these conditions would be facilitated by the discovery of the cause of C. trachomatis-negative, U. urealyticum-negative nongonococcal urethritis and mucopurulent cervicitis so that specific therapy and microbial response can be evaluated.