Control visits after 4 and 12 weeks revealed again “suspicion of UTI with many leucocytes.” Treatment was continued with antibiotics and with medication for bladder voiding and reduction of BPH. After 4 months, the PSA level showed only a small decrease to 17.8 ng/ml. Three weeks later, the urologist recorded his telephone call to the patient as follows: “Prostate biopsy necessary! The patient is unreasonable and does not agree.”

Two and a half months later, the patient visited the urologist again. He did not want a physical examination but only a new prescription of his previous medication. In the record, it was not mentioned that on this occasion the patient was made aware by several means of the suspected prostate cancer diagnosis, nor was the patient’s family physician informed accordingly.

Four years later, the patient’s physician referred him again to the urologist because of a high PSA level, now measuring 300 ng/ml! The DRE revealed a hard prostate, very suspicious of cancer also considering the high PSA level. The subsequent biopsy discovered a carcinoma of the prostate (Gleason 9). Radical surgery showed an advanced prostate tumor (pT3b) with tumorous tissue in both lobes of the prostate, infiltration of the seminal vesicles, and regional lymph node metastases.

In summary, the patient was urologically examined because of his voiding problems due to BPH. A clearly elevated PSA level, leukocyturia, and bacteriuria (without reported symptoms) were found. The DRE of the prostate was not suspicious of cancer. Wrongly, the urologist associated the PSA elevation with UTI and initiated antibiotic treatment (piperacillin/tazobactam) and medical therapy of the voiding disturbances. There was, however, no significant reduction of the PSA in the following several months and still no biopsies were performed for histological clarification. Thereafter, the patient withdrew from urological follow-up. Four years later, he returned with a PSA level of 300 ng//ml and a biopsy revealed carcinoma of the prostate. The question arises: Is antibiotic therapy justified in cases of elevated PSA (>10 ng/ml) when UTI or prostatitis is suspected but not confirmed with adequate diagnostic measures?

A 59-year-old man with a 3-year history of elevated PSA levels of around 30 ng/ml was found to have a prostate of firm consistency on DRE. Exprimate urine contained many leucocytes. Because of highly suspected cancer, three sets of prostate biopsies were performed over the years, each time without any evidence of a tumor. Therefore, the urologist assumed that the elevated PSA was due to “prostatitis.” But no relevant clinical symptoms of either acute or chronic prostatitis had been recorded and no further specific diagnostic procedures in this direction were carried out. The patient received repeated courses of antibiotic treatment.

When hematuria occurred, a cystoscopy and a transrectal ultrasonography were performed. By sonography, an enlarged prostate with an adjoining tumor area was found. Endoscopic examination revealed a stilted exophytic tumor of the colliculus seminalis growing toward the bladder. The tumor was resected transurethrally. Histologic examination showed a prostate cancer originating from the prostatic urethra or colliculus seminalis growing toward the bladder, a so-called ductal carcinoma of the prostate gland.

In summary, during 3 years, highly significant PSA elevations were observed. Because of suspicious DRE, a prostate cancer was suspected, and in the course of time, repeated prostate biopsies were carried out, none of which indicated tumors. Because of pyuria in the exprimate urine, prostatitis was assumed and the patient was treated with antibiotics [2].

Regarding the connection between UTI/prostatitis/epididymitis and PSA elevation, the question of pathogenesis must be raised. PSA is a specific marker (protease) of the prostate tissue. The antigen is produced by the epithelial cells of the prostate gland and drained into the ductuli. Here, the protease plays a role in the liquidification of seminal coagula [31]. A stimulating androgen effect on PSA formation exists (gene activation). The usual normal serum level of PSA is <4 ng/ml. However, the upper limit is also age specific: <50 years 2.5 ng/ml; 50–59 years 3.5 ng/ml; 60–69 years 4.5 ng/ml; 70–79 years 6.5 ng/ml [32].

Different kinds of damage of the prostate cells, e.g., infection, inflammation, malignant degeneration, or other injuries, lead to the release of PSA into the stroma of the organ. From there, it enters the vascular system via lymphatic vessels. Elevated serum PSA levels appear in prostatitis, BPH (increased volume), and especially in prostate cancer. Febrile UTI, even without characteristic prostatic pain, can raise PSA levels, although in such cases, an inflamed prostate is expected [40]. How high PSA elevation that can be anticipated in asymptomatic bacteriuria, for example, in patients with indwelling catheter or in lower UTI, has not yet been documented in convincing studies.

An acute or chronic inflammatory process of the prostate gland can cause a pathologic elevation of the PSA level without clinical signs and symptoms of prostatitis [13, 16, 27]. This is explained by the release of PSA from the prostatic cells damaged by inflammation. If a PSA level is above normal age range and an inflammatory prostate process is assumed, relevant examinations are necessary to establish the diagnosis of an acute or chronic bacterial prostatitis before any antibiotic therapy is initiated.

Patients with acute bacterial prostatitis complain of typical clinical symptoms of acute UTI with irritative and/or obstructive voiding disturbances. Frequently, they also have symptoms of a systemic infection with fatigue, nausea, vomiting, fever, chills, and even urosepsis. In addition, there can be perineal and suprapubic pain which can radiate to the outer genitals. On digital-rectal examination, patients complain of strong tenderness of the prostate. Massage of the prostate is contraindicated since this can cause sepsis [1, 10]. The bacteriological examination is done with mid-stream urine. Infravesical obstructions or an intraprostatic abscess have to be ruled out or treated accordingly. Following diagnosis, an immediate broad spectrum antibiotic therapy including all measures required is necessary [12, 28].

The diagnostics of chronic-bacterial prostatitis should include medical history, registration of symptoms according to the NIH symptom score, a standardized microscopic and bacteriological analysis using the four glass test (first-voided urine, midstream urine, prostate secretion after prostate massage and prostate exprimate urine), or a two-glass test (midstream urine and exprimate urine) [20, 23, 25, 30]. Suspicious findings (e.g., erythrocyturia) require urethrocystoscopy and/or transrectal ultrasonography [2].

Elevated PSA levels due to UTI raise the question whether this antigen, a protease, originates also from other organs than the prostate, especially from renal and urinary tract structures. Evidence of PSA in extraprostatic tissue was provided by immune reaction in the urethra, in the periurethral glands (male and female), and in adenoid cell carcinomas of the urethra and urinary bladder [3]. Schmidt et al. [35] detected PSA (mean 0.29 ng/ml) in the urine of 11 % of 217 women without sexual intercourse during the last 48 h. The detection of PSA in the urethra and urinary bladder, however, is fairly weak and not constant. This phenomenon is due to different specificity and sensitivity of the antibodies used [3].

Concerning the pathogenesis of elevated PSA levels due to inflammatory changes of the urinary tract, the studies of Ulleryd et al. [38] are important. In 83 % of 70 male patients with febrile UTI, the authors found elevated PSA levels with a median of 14 ng/ml (range 0.54–140 ng/ml). The patients had at least one of the typical symptoms of an acute UTI, such as dysuria, flank pain, fever of at least 38 °C and a significant bacteriuria, mainly due to E. coli [17]. Only six (9 %) of the patients complained of pain during DRE of the prostate. These investigations show that the prostate in men is involved during febrile UTI episodes which are demonstrated by elevated PSA levels [40]. In contrast, serum PSA levels were not above 1 ng/ml in 16 control patients with febrile infections of other origin.

Pilatz et al. [33] investigated 237 patients with acute epididymitis concerning etiology and outcome. PSA levels dropped within 3 months from a mean of 2.1 (range 0.8–5.6) ng/ml in antibiotic-naive patients to about half of the initial values in those cases with detected bacterial pathogens but remained more or less stable in those cases without evidence for bacterial pathogens. In pretreated patients, PSA levels declined to a similar extent, as witnessed in antibiotic-naive patients with pathogens, irrespective of whether a pathogen was detected or not. In patients where pathogen detection was based on 16S rDNA analysis, PSA values declined to a comparable degree. Elevated PSA levels in epididymitis may be explained by the fact that epididymitis is usually considered an ascending infection from the prostate.

In febrile UTI (i.e., body temperature of 38 °C or more, flank pain, with or without painful palpation of the prostate, and significant bacteriuria with cfu ≥10⁴/ml due to E. coli), the prostate (prostatitis) is often involved as documented by distinct elevated PSA levels. In asymptomatic bacteriuria with or without leukocyturia or in a simple lower UTI, prostatic involvement is not expected. Convincing studies, however, are missing.

PSA levels may increase due to acute or chronic bacterial prostatitis/epididymitis and/or febrile UTI and decrease again during and after antibiotic therapy, which, however, takes some time. One month after therapy, 43 % of 70 men treated for febrile UTI still showed increased PSA levels, which finally dropped to a median of 1.5. ng/ml only after 12 months. The reductions in PSA and prostate volume were significantly correlated. The slow decline of PSA levels in some patients indicates a protracted healing process, according to the authors. Interestingly, in one patient with disseminated prostate cancer who had raised baseline levels of PSA responded to the infection with a further increase in serum PSA, which was interpreted in a way that different prostatic affections may have an additive effect on PSA levels [38, 39].

On this background, it was assumed that the part of PSA, which is increased due to infection, can be reduced to normal after the infection is cured by a suitable antibiotic therapy, whereas the part of PSA increased due to a concomitant prostatic cancer will remain the same. Later on, however, antibiotic therapy was not only performed, in case of bacterial prostatitis (acute or chronic) or febrile UTI, but also for any kind of elevated PSA as sort of differential therapy to reduce that part of PSA, which was probably increased due to infectious microfoci. If the PSA levels returned to normal after therapy, no prostatic biopsy was needed. This approach seemed to be even more justified, because increased PSA levels have also been found in asymptomatic inflammatory changes of the prostate [13, 19]. Serretta et al. [36] treated 99 patients due to elevated PSA for 3 weeks with ciprofloxacin. In 65 % of cases, histology detected small foci of prostatitis. No cancer was detected if PSA decreased below 4 ng/ml or more than 70 %. Nickel et al. [29], however, could not find any correlation between kind and grade of inflammation and the total PSA or PSA density in patients with BPH and LUTS. The PSA increase in BPH patients was mainly explained by hyperplasia and not so much by the asymptomatic prostatitis [16, 24, 26].