Methods
The German sentinel surveillance system for STIs collects data from 60 local health offices, 13 hospital-based STI clinics and 159 private practitioners. Private practitioners such as specialists in dermatovenerology, gynaecology, urology and HIV have been chosen at randomin all federal states. Sentinel physicians are asked to report LGV cases to the Robert Koch-Institut using sentinel reporting forms.
Updates on LGV were published in the German national epidemiological bulletin [5, 6] which is read by local health authorities and private practitioners. All physicians were asked to send their samples to laboratories that perform chlamydia genotyping. Reports on cases were received from sentinel physicians, other physicians, hospitals and laboratories. We also provided information about LGV to magazines aimed at a gay readership.
A possible case was defined as a person with symptoms of proctitis and/or inguinal lymph node swelling and a positive chlamydia serology. A probable case also had a positive chlamydia rectal or urinary polymerase chain reaction (PCR) test. A case was confirmed if the genotype L1-L3 was identified, even if the patient’s symptoms were unknown.
Chlamydia trachomatis infections sent to the Arndt and Partner diagnostic laboratory were diagnosed by DNA amplification of lesional swabs or lymph node aspirates using PCR (Cobas TaqMan CT, Roche, Mannheim) or strand displacement amplification (ProbeTec ET, Becton-Dickinson, MD). C. trachomatis genotypes were identified by sequence analysis of variable omp1 gene regions (VS1, VS2, and VS4), amplified by PCR using primer pairs MF21/MB22 and MF44/MB4 [6].
Additional patient information was obtained by asking physicians. We described confirmed, probable and possible LGV cases by demographic characteristics and symptoms.

Results
Between May 2004 and November 2005, 61 confirmed and 17 probable or possible cases were reported to the Robert Koch-Institut. All confirmed cases were genotype L2. Reports were received from two local health offices, seven hospital-based STI clinics and 17 private practitioners.
The epidemic curve is shown in figure 1. The median number of monthly reported cases has increased from two in 2004 to four in 2005. All cases were in men, and the mean age was 39 years. Main characteristics of LGV patients are shown in the table. All 58 patients from whom an information on sexual orientation was obtained were MSM. Twenty seven of all 78 cases (35%) had proctitis and 31 had unspecified rectal symptoms. Ten patients showed an inguinal lymph node swelling. Of those, one patient showed confluent lymph nodes with signs of extensive inflammation. Whether this patient was MSM is unknown.

Of the 52 patients with a known HIV status, 50 (96%) were known to be HIV-positive. Two patients were HIV-negative while the HIV status of 26 patients remains unknown. Patients originated from eight different federal states. Forty five patients (58%) were diagnosed in Hamburg, 17 (22%) in Berlin and 6 (8%) in Munich. The geographical distribution is shown in figure 2.

Of 24 patients for whom information on sexual contacts in other countries was available, three reported sexual contacts in the Netherlands and in Belgium. The other 21 patients did not report any contacts outside of Germany.

Discussion
In Germany, the first cases of LGV in MSM were observed at approximately the same time as the first cases in the Netherlands, France, Belgium and the United Kingdom [7-10]. It is possible that the outbreak began among MSM in the Netherlands and subsequently spread to Germany, but this remains unproven. Although information on sexual contacts in other countries was only available for a few German patients, the majority of these patients appear to have become infected in Germany. LGV may therefore have become endemic in the MSM community within two years of detection of the first cases in Germany. Furthermore, the number of reported patients with LGV has increased over the past few months and there are no signs that the epidemic is over. Over 80% of the reported cases have been diagnosed in large cities with a substantial MSM community such as Berlin, Hamburg and Munich. LGV may be more prevalent in these cities, but the difference could also reflect a diagnostic bias. Since the main diagnostic laboratory is situated in Hamburg, awareness among physicians may be higher in Hamburg than elsewhere.
About two thirds of LGV patients knew they were HIV positive at the time of their LGV diagnosis, a feature which has been observed in other countries as well [7, 9]. Reported cases of HIV and syphilis among MSM have also increased over the past three years in Germany [11, 12]. Since LGV facilitates HIV transmission, the emergence of LGV in the MSM community in the context of increasing numbers of newly acquired HIV infections should not be ignored. Clinicians and MSM may not be sufficiently aware of the disease, and existing efforts to promote awareness and prevention of sexually transmitted infections and HIV need to be strengthened.