The presence of visual neglect after thrombolytic treatment in patients with right hemisphere stroke.

Visual neglect (VN) is a common consequence of right hemisphere (RH) stroke. The aims of this study were to explore the presence of VN after RH stroke in the patients with (T+) or without (T-) thrombolytic treatment, and to determine whether thrombolysis is a predictor of VN. The study group consisted of 77 RH infarct patients. VN was evaluated with six conventional subtests of the Behavioural Inattention Test (BIT). Stroke severity was assessed using the National Institute of Health Stroke Scale (NIHSS). In the neuropsychological examination, 22% of all RH stroke patients had VN. VN was present in 15% of the patients in the T+ group and in 28% of the patients in the T- group, but the difference was not statistically significant. Despite that, patients in the T- group had a higher risk of VN than patients in the T+ group. Our results suggest that thrombolysis independently predicted absence of VN.

Recovery from visual neglect after right hemisphere stroke: does starting point in cancellation tasks change after 6 months?

In the acute phase of stroke, patients with left visual neglect (VN) automatically orient to the right hemispace. This study examined the presence of rightward bias after right hemisphere stroke within 10 days of stroke onset and after 6 months. Our sample comprised 43 patients and 49 healthy controls. Presence of VN was evaluated with the six conventional subtests of the Behavioral Inattention Test (BITC). Starting points were determined in three BITC cancellation tasks by measuring the distance between the starting point and the median line of the stimulus sheet in centimeters. Activities of daily living (ADL) were assessed with the Barthel Index. At baseline VN patients showed more robust rightward bias than patients without VN. The magnitude of rightward bias decreased clearly in the VN patients at follow-up. A favorable ADL outcome was observed in 90% of the patients with VN and in all of the patients without VN. The magnitude of rightward bias differed clearly between the patient groups and controls. Our result implies that VN was likely to have improved as measured by BITC sum scores, but symptoms of rightward attention bias were still detected. We therefore suggest that, for clinical purposes, it is important that attention bias is measured accurately after right hemisphere stroke.

Estradiol and cognition during androgen deprivation in men with prostate carcinoma.

BACKGROUND: The adverse effects of hormonal manipulation in prostate carcinoma need to be established in view of its increasing use as an adjuvant treatment. This prospective study investigated the association of androgen deprivation-induced estradiol decline with cognition in prostate carcinoma. METHODS: Cognitive testing of prostate carcinoma patients was carried out at baseline and at 6 and 12 months on androgen deprivation (AD). Cognitive performances were evaluated with standardized measures of information processing, including working memory and attention, visual and verbal skills, and memory performances in 31 tests. Testosterone and estradiol changes during AD were measured with the DELFIA (PerkinElmer, Inc., Wellesley, MA) system. Associations between changes in cognitive performances and estradiol decline were studied. RESULTS: Cognitive performances, which were significantly associated with decline in estradiol, included visual memory of figures (r = -0.52; P = 0.022) and recognition speed of numbers, which were impaired, (r = -0.57; P = 0.030) at 6 months, and improvement in verbal fluency (r = -0.52; P = 0.019) at 12 months. Other cognitive domains appeared unaffected by estradiol decline. The character of change (impairment or improvement) depended on the magnitude of estradiol decline. CONCLUSIONS: The cognitive domains of verbal fluency, visual recognition, and visual memory were associated with decline in estradiol during androgen deprivation. The results suggest selective associations among testosterone decline, estradiol, and cognitive performance. Documentation of these associations has implications for informed patient support in hormonally treated prostate carcinoma.

Effect of adjuvant androgen deprivation on thyroid function tests in prostate cancer patients.

Androgen deprivation (AD) used in the treatment of prostate cancer is known to alter concentrations of sex hormones and their binding globulins. Less is known as to its effect on thyroid hormones. In this prospective study the effects of AD on thyroid function were clarified. Levels of serum thyroid stimulating hormone (TSH), free thyroxine (FT4) and thyroid binding globulin concentrations were measured in prostate cancer patients treated with either radical radiotherapy and androgen deprivation for 12 months (AD) or radical radiotherapy alone (RT). Measurements were made at baseline, and at 3, 6 and 12 months. At baseline and at 3 months the results of thyroid function tests did not differ significantly between groups. A significant decline in serum testosterone in the AD group was accompanied by a significant decline in FT4 at 6 and 12 months, while no significant changes in thyroid function were observed in the RT group. The decline in FT4 among AD patients did not evoke a normal TSH response. Prolonged use of AD hampers the interpretation of thyroid test results. This finding has substantial implications for the follow-up of patients in hormonally treated prostate cancer.

The normal population distribution of PRNP codon 129 polymorphism.

OBJECTIVES: The common prion protein gene (PRNP) codon 129 polymorphism modifies the susceptibility to and the phenotype of prion diseases. However, no truly representative normal population-based data, or data stratified according to age or gender are available on the distribution of this polymorphism. MATERIAL AND METHODS: Allelic variation of codon 129 in three Finnish populations representing different age groups, and among Finnish, British and Irish blood donors were examined. RESULTS: The PRNP codon 129 genotype distribution in the total Finnish sample was 49% for methionine-methionine (MM), 42% for methionine-valine (MV) and 9% for valine-valine (VV), for the UK blood donors 42% for MM, 47% for MV and 11% for VV, and for the Irish blood donors 34% for MM, 56% for MV, and 10% for VV. CONCLUSIONS: The genotype frequencies were almost identical in all three Finnish populations of different ages, with no gender differences, and did not differ from corresponding figures for the Finnish blood donors. However, the PRNP codon 129 genotype distribution in Finland differed significantly from that of the British and the Irish blood donors and the previously published blood donor data on other Western Europeans and Americans.

Androgen deprivation and cognition in prostate cancer.

Androgen deprivation (AD) is commonly used in neoadjuvant and adjuvant setting with prostate cancer (PC) radiotherapy. This prospective study assessed whether cognitive functioning is impaired during 12 months of AD therapy. Longitudinal testing of 25 patients treated with AD and curative radiotherapy was undertaken at baseline, and at 6 and 12 months. CogniSpeed software was used for measuring attentional performances. Other cognitive performances were evaluated using verbal, visuomotor and memory tests. The Beck depression inventory was employed to evaluate depressive mood and EORTC QLQ-C30 for quality of life (QoL). During longitudinal testing of the AD group, no impairment in cognitive performances was found. Instead, improvement was observed in object recall (immediate, P=0.035; delayed, P<0.001), and in semantic memory (P=0.037). In QoL, impairment in physical function was observed. Androgen deprivation of 12 months appears to be associated with preserved cognitive functioning, although physical impairment occurs. These results have implications for counseling and psychosocial support of patients in the context of treatment options in PC.

Increased expression of FGF-8 isoforms and FGF receptors in human premalignant prostatic intraepithelial neoplasia lesions and prostate cancer.

SUMMARY: Fibroblast growth factor 8 (FGF-8) is implicated in growth of prostate cancer. Alternative splicing of the human FGF-8 gene potentially allows coding for four protein isoforms (a, b, e, and f). These isoforms differ in their binding to FGF receptors (FGFR) and in their mitogenic and transforming capacity in transfection assays. Here, we used RT-PCR and immunohistochemistry to study the expression of FGF-8 and FGFR isoforms in human prostate cancer (n = 31). Nonmalignant prostate specimens from cystoprostatectomies (n = 24) were examined as controls. Most prostate cancer samples and some control prostates also contained prostatic intraepithelial neoplasia (PIN) lesions. FGF-8a and e were expressed at significantly higher frequencies in prostate cancer (FGF-8a, 55%; FGF-8e, 45%) than in control samples (FGF-8a, 17%, p = 0.0052; FGF-8e, 8%, p = 0.0031). On the contrary, FGF-8b was found at an equal frequency in prostate cancer (55%) and in control prostates (50%). Furthermore, a combination of two or three FGF-8 isoforms (a, b, and/or e) was also expressed at a higher frequency in prostate cancer than in control samples (45% and 8%, respectively, p = 0.0031). Immunohistochemistry with an antibody recognizing all FGF-8 isoforms was more strongly immunoreactive in prostate cancer cells and PIN lesions than in normal-type epithelium. The receptor splicing variants FGFR1IIIc and FGFR2IIIc, which are activated by FGF-8, were found both in prostate cancer and control samples. Interestingly, immunoreactivity for FGFR1 and FGFR2 was much stronger in prostate cancer cells and PIN than in normal epithelium. These results demonstrate, for the first time, that FGF-8 isoforms and their receptors FGFR1IIIc and FGFR2IIIc are expressed at an increased level not only in prostate cancer but also in premalignant PIN lesions. These data suggest that FGF-8 may have an important autocrine role in the development of human prostate cancer. In addition to FGF-8b, the FGF-8 isoforms a and e may be involved in this process.

Interferon-alpha inhibits cyclooxygenase-1 and stimulates cyclooxygenase-2 expression in bladder cancer cells in vitro.

The enzymes cyclooxygenase-1 (Cox-1) and cyclooxygenase-2 (Cox-2) catalyze the initial step in the formation of prostaglandins (PGs). PGs are known to be involved in numerous processes, for example inflammation, immune responses, carcinogenesis, and tumor angiogenesis. The formation of PGs is stimulated in various cancers since the expression of Cox-2 is upregulated. Interferon (IFN)-alpha is used in the treatment of bladder cancer, although not all of the effects of such treatment are thoroughly known. Therefore, we investigated the expression of cyclooxygenases in two bladder cancer cell lines, 5637 and T24, under basal conditions and in the presence of human recombinant IFN-alpha (100, 1,000, and 10,000 U/ml). The mRNA of Cox-1 and Cox-2 was expressed in both cultured bladder carcinoma cell lines. The level of Cox-1 expression was low in 5637 cells and higher in T24 cells. In contrast, Cox-2 expression was prominent in 5637 cells and low in T24 cancer cells. The highest IFN-alpha concentration (10,000 U/ml) decreased the expression of Cox-1 to 47 and 28% of the control levels in 5637 and T24 cells, respectively. In contrast, Cox-2 expression increased in both cell lines. In 5,637 cells, Cox-2 expression increased 1.3-fold with 10,000 U/ml of IFN-alpha. In T24 cells, the maximum effect was achieved by 1,000 U/ml of IFN-alpha, which increased the expression of Cox-2 up to 2.4-fold. These findings may have relevance in the outcome of patients treated with IFN-alpha because upregulated Cox-2 expression may suppress the cell-mediated defense system. On the other hand, the inhibition of Cox-1 could be beneficial because Cox-1 is known to stimulate angiogenesis.

Increased expression of cyclooxygenase-2 and nitric oxide synthase-2 in human prostate cancer.

Cyclooxygenase-1 (COX-1), cyclooxygenase-2 (COX-2) and inducible nitric oxide synthase-2 (NOS-2) each have an important role in angiogenesis. The expression of these genes was investigated in human prostate cancer by immunohistochemistry, the expression of COX-1 and COX-2 being confirmed by mRNA analysis. Prostate cancer specimens from 12 patients were compared to control prostates from 13 patients operated on for bladder carcinoma. The intensity of COX-2 and NOS-2 immunostaining was significantly stronger in prostate cancer cells than in the non-malignant glandular epithelium of the control prostates. COX-2 and NOS-2 were clearly also expressed in the lesions of prostatic intraepithelial neoplasia (PIN) in control prostates. COX-2 was detected in the muscle fibres of the hyperplastic stroma of some control prostates. No significant difference was detected in COX-1 expression between control and cancer prostates. These results indicate that the expression of COX-2 and NOS-2 is elevated in prostatic adenocarcinoma and in PIN.

Management of testicular cancer--16 years' experience from southwest Finland.

OBJECTIVE: This study investigated the outcome of testicular cancer treatment in Finland. MATERIAL AND METHODS: Data on 88 testicular cancer patients treated in Turku University Central Hospital between 1976 and 1992 were studied to analyse outcome and survival. RESULTS: The histological diagnosis was seminoma for 39 patients and non-seminoma for 49 patients. Two seminoma patients relapsed (5%) and one patient died of progressive disease (3%; initially stage II seminoma). Eleven non-seminoma patients relapsed (22%), nine of whom were cured with chemotherapy. Four non-seminoma patients died of progressive disease (8%; initially one stage I non-seminoma and three stage III non-seminomas). The median time to relapse after the completion of treatment was 9 months (range 3-50 months). Non-seminoma patients had significantly more relapses than seminoma patients (p = 0.03). Most relapses (73% of the non-seminoma relapses) were found among the stage I non-seminoma patients who had not received adjuvant chemotherapy, while none of the stage I seminoma patients relapsed (p = 0.007). CONCLUSIONS: Close surveillance is important for all non-seminoma patients to guarantee the early detection and treatment of recurrent disease. Treatment and surveillance should be covered by national guidelines and be conducted in centres with special interest in this rare but mostly curable cancer.

Measurement of circulating forms of prostate-specific antigen in whole blood immediately after venipuncture: implications for point-of-care testing.

BACKGROUND: The purpose of this study was to validate the use of whole-blood samples in the determination of circulating forms of prostate-specific antigen (PSA). METHODS: Blood samples of hospitalized prostate cancer and benign prostatic hyperplasia patients were collected and processed to generate whole-blood and serum samples. Three different rapid two-site immunoassays were developed to measure the concentrations of total PSA (PSA-T), free PSA (PSA-F), and PSA-alpha(1)-antichymotrypsin complex (PSA-ACT) to detect in vitro changes in whole-blood samples immediately after venipuncture. The possible influence of muscle movement on the release of PSA from prostate gland was studied in healthy men by measuring the rapid in vitro whole-blood kinetics of PSA forms before and after 15 min of physical exercise on a stationary bicycle. RESULTS: Rapid PSA-T, PSA-F, and PSA-ACT assays were designed using a 10-min sample incubation. No significant changes were detected in the concentrations of PSA-T, PSA-F, and PSA-ACT from the earliest time point of 12-16 min compared with measurements performed up to 4 h after venipuncture. Physical exercise did not influence the concentrations of the circulating forms of PSA. Hematocrit-corrected whole-blood values of PSA-T and PSA-F forms were comparable to the respective serum values. Calculation of the percentage of PSA-F (PSA F/T ratio x 100) was similar irrespective of the sample format used, i.e., whole blood or serum. CONCLUSIONS: We found that immunodetectable PSA forms are likely at steady state immediately after venipuncture, thus enabling the use of anticoagulated whole-blood samples in near-patient settings for point-of-care testing, whereas determinations of PSA (e.g., PSA-T, PSA-F, or PSA-ACT) performed within the time frame of the office visit would provide results equivalent to conventional analyses performed in serum.

Human complement factor H related protein test for monitoring bladder cancer.

PURPOSE: The BTA stat test is a rapid, noninvasive, qualitative urine test that detects bladder tumor associated antigen (human complement factor H related protein) in urine. We compared BTA stat test to voided urine cytology in patients monitored for bladder cancer in a prospective trial, and determined whether this test is effective in detection of recurrence not seen by regular cystoscopy. MATERIALS AND METHODS: A total of 445 consecutive patients with bladder cancer were studied. A voided urine sample was obtained before cystoscopy and divided for culture, cytology and BTA stat testing. In cases of a positive BTA stat test but negative cystoscopy, excretory urography or renal ultrasound, random biopsies and collected ureteral urine samples for ureteral cytology were obtained. The overall sensitivity and specificity as well as positive and negative predictive values for BTA stat test, cytology and their combination were calculated. RESULTS: Of the 445 patients 118 (26.5%) had bladder cancer recurrence on cystoscopy, which was detected by BTA stat test and cytology in 63 (53.4%) and 21 (17.8%), respectively. Of the remaining 327 patients not having recurrent tumor on cystoscopy 81 (24.8%) had a positive BTA stat test. Excretory urography or renal ultrasound and random biopsies in 48 (59.3%) of these patients revealed 7 recurrences, making the total number of recurrent tumors 125 of 412 (30.3%). The overall sensitivities and specificities for the BTA stat test, cytology and their combination were 56.0%, 19.2%, 60.0% and 85.7%, 98.3% and 85.0%, respectively. CONCLUSIONS: The sensitivity for detection of recurrent tumor on BTA stat test is superior to that of voided urine cytology in all bladder cancer categories, whereas the specificity of voided urine cytology is higher than that for BTA stat test. However, a sixth of the patients with apparent false-positive BTA stat test results chosen for further investigation had recurrent tumors that were not found on routine cystoscopy. Although the sensitivity and specificity were highest when both tests were used, the differences were not significant overall. Therefore, the BTA stat test could potentially replace urine cytology for followup of superficial bladder cancer.

Leiomyoma of the urinary bladder.

We report a case of rare benign bladder leiomyoma. The patient was a 42-year old man complaining impotence. Digital rectal examination revealed a palpable pelvic tumor. Transurethral ultrasonography, computed tomography (CT), and magnetic resonance image (MRI) examinations all showed a cystic tumor measuring 10 x 8 cm. On the basis of these findings leiomyoma was suspected already before operation. The tumor was removed by open resection of bladder, and the diagnosis was confirmed by histology and positive immunohistochemistry.

Leiomyoma of the urinary bladder.

We report a case of rare benign bladder leiomyoma. The patient was a 42-year old man complaining impotence. Digital rectal examination revealed a palpable pelvic tumor. Transurethral ultrasonography, computed tomography (CT), and magnetic resonance image (MRI) examinations all showed a cystic tumor measuring 10 x 8 cm. On the basis of these findings leiomyoma was suspected already before operation. The tumor was removed by open resection of bladder, and the diagnosis was confirmed by histology and positive immunohistochemistry.

Cathepsin H expression distinguishes oncocytomas from renal cell carcinomas.

BACKGROUND: Because the differential diagnosis of oncocytoma and renal cell carcinoma lacks specificity, new methods supporting correct diagnostic decisions are welcome. MATERIAL AND METHODS: Sixteen cases of renal oncocytoma, and 16 sex-, age-, and stage- matched controls of renal cell carcinoma (T1-2N0M0) were studied. The minimum follow up exceeded ten years. There were no deaths due to neoplasm among oncocytomas, but 4 patient died with metastatic disease among cancer patients. Immunohistochemical staining for cathepsin H was quantified by 3 histoscores. The histoscores evaluated: 1) even cytoplasmic staining of neoplastic cells, 2) granular staining, or 3) total staining. RESULTS: 100% distinction was possible with the even cytoplasmic staining score. Total staining distinguished 87.5%, and granular staining 25% of neoplasms. CAM 5.2 cytokeratin, or vimentin distinguished 84.4% or 56.6% of these tumors, respectively. CONCLUSIONS: Cathepsin H histoscore on even cytoplasmic immunostaining is an excellent method for the distinction of oncocytomas and renal cell carcinomas.

The role of Borrelia burgdorferi in interstitial cystitis.

BACKGROUND/AIMS: Borrelia burgdorferi spirochete has been found both in bladder biopsies and the urine of patients with Lyme disease (LD) as well as in experimental animals. The urological symptoms in borreliosis resemble those of interstitial cystitis (IC): frequency, urgency and nocturia. The aim of this studies is to find the role of B. burgdorferi in interstitial cystitis. METHODS: We studied antibodies against B. burgdorferi from serum samples of 50 IC patients with two separate EIA tests. Patients with positive serology in both tests underwent cystoscopy and a bladder biopsy was taken. The presence of borrelia DNA was studied with borrelia-specific polymerase chain reaction (PCR), and with universal bacterial PCR. RESULTS: IgM class antibodies to B. burgdorferi were not found, but IgG antibodies were found in four samples (8%). This was higher than in the control material (2%). One patient's sample was strongly positive, whereas three samples were weakly positive. Bladder biopsies taken from the 4 patients were negative for borrelia DNA in both PCR tests. None of the seropositive patients had any symptoms consistent with LD. CONCLUSION: These results indicate that persistent infection of B. burgdorferi has no role in the etiology of IC. On the other hand a connection with a past borrelia infection and IC is not excluded.

Brain ethanol levels after voluntary ethanol drinking in AA and Wistar rats.

Brain ethanol was monitored in the nucleus accumbens with one minute microdialysis and headspace gas chromatography in male Wistar and alcohol preferring AA (Alko Alcohol) rats after voluntary limited access consumption without food restriction. The rats drank 0.93 +/- 0.14 (Wistar) and 0.73 +/- 0.07 g/kg (AA), with a resulting mean maximal brain ethanol level of 15.9 mM and 14.1 mM, respectively. Maximum brain ethanol levels for individual AA rats were in the range 9.4-33.6 mM, median 15.5 mM and for the individual Wistar rats in the range 2.5-35.2 mM, median 17.8 mM. There was a significant but not perfect correlation between the amount ethanol drunk and the resulting ethanol level in the nucleus accumbens, probably because of the rats not being food deprived before the experiment. The results show that rats drink pharmacologically meaningful doses in a voluntary limited access situation and that blood samples can give us a hint about the level attained in the brain, but to know the early brain concentration after drinking, microdialysis is an excellent tool.

Prospective randomized trial of interferon alfa-2a plus vinblastine versus vinblastine alone in patients with advanced renal cell cancer.

PURPOSE: The combination of interferon alfa-2a (IFNalpha2a) plus vinblastine (VLB) induces objective tumor responses in patients with advanced renal cell cancer. However, no prospective randomized trial has shown that this treatment prolongs overall survival. We compared overall survival after treatment with IFNalpha2a plus VLB versus VLB alone in patients with advanced renal cell cancer. PATIENTS AND METHODS: We prospectively randomized 160 patients with locally advanced or metastatic renal cell cancer to receive either VLB alone or IFNalpha2a plus VLB for 12 months or until progression of disease. In both groups, VLB was administered intravenously at 0.1 mg/kg every 3 weeks, and in the combination group IFNalpha2a was administered subcutaneously at 3 million units three times a week for 1 week, and 18 million units three times a week thereafter for the second and subsequent weeks. For patients unable totolerate IFNalpha2a at 18 million units per injection, the dose was reduced to 9 million units. RESULTS: Median survival was 67.6 weeks for the 79 patients receiving IFNalpha2a plus VLB and 37.8 weeks for the 81 patients treated with VLB (P =.0049). Overall response rates were 16. 5% for patients treated with IFNalpha2a plus VLB and 2.5% for patients treated with VLB alone (P =.0025). Treatment with the combination was associated with constitutional symptoms and abnormalities in laboratory parameters, but no toxic deaths were reported. CONCLUSION: The combination of IFNalpha2a plus VLB is superior to VLB alone in the treatment of patients with locally advanced or metastatic renal cell carcinoma. This is the first study to demonstrate that survival can be prolonged by using IFNalpha2a for these patients.

Urinary bladder transitional cell carcinogenesis is associated with down-regulation of NF1 tumor suppressor gene in vivo and in vitro.

The NF1 gene product (neurofibromin) is known to act as a tumor suppressor protein by inactivating ras. The best documented factors involved in urinary bladder transitional cell carcinoma (TCC) are ras proto-oncogene activation and p53 suppressor gene mutations. This is the first study reporting alterations in NF1 gene expression in TCC. We examined NF1 gene expression in a total of 29 surgical urinary bladder TCC specimens representing grades 1 to 3 and in three cell lines, RT4, 5637, and T24 (representing grades 1 to 3, respectively). Decreased NF1 gene expression was observed in 23 of 29 (83%) TCC specimens as estimated by immunohistochemistry, the decrease being more pronounced in high-grade tumors. NF1 mRNA levels were markedly lower in TCC tissue compared with adjacent non-neoplastic urothelium, as studied by in situ hybridization for grade 3 TCC. Immunohistochemistry and Western blotting demonstrated that TCC cell lines expressed NF1 protein at different levels, expression being almost undetectable in T24 (grade 3) cells. Northern blotting for cell lines demonstrated reduced NF1 mRNA levels in grade 3 TCC cells. Reverse transcription polymerase chain reaction for cell lines and selected grade 2 and grade 3 tissue samples demonstrated NF1 type II mRNA isoform predominance in all samples studied. Our results show that both NF1 mRNA and protein levels are decreased in high-grade TCC, suggesting that alterations of NF1 gene expression may be involved in bladder TCC carcinogenesis.

The cytostatic effect of 9-cis-retinoic acid, tretinoin, and isotretinoin on three different human bladder cancer cell lines in vitro.

Retinoids have been shown to have activity in both preclinical and clinical bladder cancer studies but their exact role in its treatment and prevention remains obscure. In this study cytostatic activity of a novel 9-cis-retinoic acid (9-cis-RA) was compared with two other retinoids: tretinoin and isotretinoin, in three different bladder cancer cell lines: RT4 (well differentiated), 5637 (moderately differentiated) and T24 (poorly differentiated). The three retinoids were incubated at concentrations of 0.3, 3 and 30 microg/ml with bladder cancer cells in microtitre plates for 3 and 6 days. The cytostatic effect was estimated by using luminometric measuring of ATP activity of viable cells in suspension. Compared with the older retinoids, tretinoin and isotretinoin, the highest concentration of 9-cis-RA had a cytostatic efficacy in all three bladder cancer cell lines tested. A clear dose response relationship was observed in isotretinoin-treated cultures after 6 days and in all 9-cis-RA-treated cultures. Tretinoin was either ineffective or had a stimulating effect on poorly differentiated tumour cells. To conclude, isotretinoin and 9-cis-RA had a cytostatic effect on human bladder cancer cells in vitro. However, the possibility of stimulating cancer growth at small doses, at least with tretinoin, and toxicity at high doses must be considered when planning clinical trials.