68Ga-PSMA PET/CT compared with MRI/CT and diffusion-weighted MRI for primary lymph node staging prior to definitive radiotherapy in prostate cancer: a prospective diagnostic test accuracy study

BACKGROUND

The aim was to compare the diagnostic accuracy of ⁶⁸Ga-PSMA PET/CT with conventional cross-sectional imaging and diffusion-weighted MRI (DW-MRI) for detecting lymph node metastasis (LNM) to stage prostate cancer patients. Twenty consecutive, newly- diagnosed prostate cancer patients were prospectively enrolled and underwent ⁶⁸Ga-PSMA-11 PET/CT, anatomical MRI or contrast-enhanced CT, and DW-MRI prior to laparoscopic, template-based, extended lymph node dissection. Histopathological findings served as the reference test.

RESULTS

Histopathology showed LNM in 13 of 20 patients (19 high-risk, 1 intermediate risk). Five patients had metastasis-suspected lymph nodes on ⁶⁸Ga-PSMA PET/CT. Patient-based analysis showed that the sensitivity and specificity for detecting LNM were 39% and 100% with ⁶⁸Ga-PSMA PET/CT, 8% and 100% with MRI/CT, and 36% and 83% with DW-MRI, respectively. The positive and negative predictive values were 100% and 49% with ⁶⁸Ga-PSMA PET/C, 100% and 37% with MRI/CT, and 80% and 42% with DW-MRI. Of 573 dissected lymph nodes, 33 were LNM from 26 regions. True-positive LNM on ⁶⁸Ga-PSMA PET/CT was 9-11 mm in diameter, whereas false-negative LNM had a median diameter of 4 mm, with only 3 of 30 lymph nodes being larger than 10 mm. LNM were positive for PSMA by immunostaining.

CONCLUSIONS

The sensitivity of ⁶⁸Ga-PSMA PET/CT was notably better than that of MRI/CT and comparable to that of DW-MRI. Some false positive findings with DW-MRI reduced its specificity and positive predictive value compared with those of ⁶⁸Ga-PSMA PET/CT and MRI/CT.

Phase III Trial of PROSTVAC in Asymptomatic or Minimally Symptomatic Metastatic Castration-Resistant Prostate Cancer

PURPOSE

PROSTVAC, a viral vector–based immunotherapy, prolonged median overall survival (OS) by 8.5 months versus placebo in metastatic castration-resistant prostate cancer in a phase II study. This phase III study further investigated those findings.

PATIENTS AND METHODS

Patients were randomly assigned to PROSTVAC (Arm V; n = 432), PROSTVAC plus granulocyte-macrophage colony-stimulating factor (Arm VG; n = 432), or placebo (Arm P; n = 433), stratified by prostate-specific antigen (less than 50 ng/mL v 50 ng/mL or more) and lactate dehydrogenase (less than 200 v 200 U/L or more). Primary end point was OS. Secondary end points were patients alive without events (AWE)—namely, radiographic progression, pain progression, chemotherapy initiation, or death—at 6 months and safety. The study design was a superiority trial of PROSTVAC (Arm V or Arm VG) versus Arm P. Three interim analyses were planned.

RESULTS

At the third interim analysis, criteria for futility were met and the trial was stopped early. Neither active treatment had an effect on median OS (Arm V, 34.4 months; hazard ratio, 1.01; 95% CI, 0.84 to 1.20; P = .47; Arm VG, 33.2 months; hazard ratio, 1.02; 95% CI, 0.86 to 1.22; P = .59; Arm P, 34.3 months). Likewise, AWE at 6 months was similar (Arm V, 29.4%; odds ratio, 0.96; 95% CI, 0.71 to 1.29; Arm VG, 28.0%; odds ratio, 0.89; 95% CI, 0.66 to 1.20; placebo, 30.3%). Adverse events were similar for the treatment and placebo groups, with the most common being injection site reactions (62% to 72%) and fatigue (21% to 24%). Arrhythmias were the most common cardiac-related events (1.4% to 3.5%). There were no reports of either myocarditis or pericarditis. Serious treatment-related events occurred in less than 1% of all patients.

CONCLUSION

Whereas PROSTVAC was safe and well tolerated, it had no effect on OS or AWE in metastatic castration-resistant prostate cancer. Combination therapy is currently being explored in clinical trials.

Prospective comparison of 68Ga-PSMA PET/CT, 18F-sodium fluoride PET/CT and diffusion weighted-MRI at for the detection of bone metastases in biochemically recurrent prostate cancer

PURPOSE

To prospectively compare diagnostic accuracies for detection of bone metastases by ⁶⁸Ga-PSMA PET/CT, ¹⁸F-NaF PET/CT and diffusion-weighted MRI (DW₆₀₀-MRI) in prostate cancer (PCa) patients with biochemical recurrence (BCR).

METHODS

Sixty-eight PCa patients with BCR participated in this prospective study. The patients underwent ⁶⁸Ga-PSMA PET/CT, a ¹⁸F-NaF PET/CT and a DW₆₀₀-MRI (performed in accordance with European Society of Urogenital Radiology guidelines, with b values of 0 and 600 s/mm²). Bone lesions were categorized using a three-point scale (benign, malignant or equivocal for metastases) and a dichotomous scale (benign or metastatic) for each imaging modality by at least two experienced observers. A best valuable comparator was defined for each patient based on study-specific imaging, at least 12 months of clinical follow-up and any imaging prior to the study and during follow-up. Diagnostic performance was assessed using a sensitivity analysis where equivocal lesions were handled as non-metastatic and then as metastatic.

RESULTS

Ten of the 68 patients were diagnosed with bone metastases. On a patient level, sensitivity, specificity and the area under the curve (AUC) by receiver operating characteristic analysis were, respectively, 0.80, 0.98-1.00 and 0.89-0.90 for ⁶⁸Ga-PSMA PET/CT (n = 68 patients); 0.90, 0.90-0.98 and 0.90-0.94 for ¹⁸NaF PET/CT (n = 67 patients); and 0.25-0.38, 0.87-0.92 and 0.59-0.62 for DW₆₀₀-MRI (n = 60 patients). The diagnostic performance of DW₆₀₀-MRI was significantly lower than that of ⁶⁸Ga-PSMA PET/CT and ¹⁸NaF PET/CT for diagnosing bone metastases (p < 0.01), and no significant difference in the AUC was seen between ⁶⁸Ga-PSMA PET/CT and ¹⁸NaF PET/CT (p = 0.65).

CONCLUSION

⁶⁸Ga-PSMA PET/CT and ¹⁸F-NaF PET/CT showed comparable and high diagnostic accuracies for detecting bone metastases in PCa patients with BCR. Both methods performed significantly better than DW₆₀₀-MRI, which was inadequate for diagnosing bone metastases when conducted in accordance with European Society of Urogenital Radiology guidelines.

Diagnostic test accuracy study of 18F-sodium fluoride PET/CT, 99mTc-labelled diphosphonate SPECT/CT, and planar bone scintigraphy for diagnosis of bone metastases in newly diagnosed, high-risk prostate cancer

The aim of this study was to prospectively compare planar, bone scan (BS) versus SPECT/CT and NaF PET/CT in detecting bone metastases in prostate cancer. Thirty-seven consecutive, newly diagnosed, prostate cancer patients with prostate specific antigen (PSA) levels ≥ 50 ng/mL and who were considered eligible for androgen-deprivation therapy (ADT) were included in this study. BS, SPECT/CT, and NaF PET/CT, were performed prior to treatment and were repeated after six months of ADT. Baseline images from each index test were independently read by two experienced readers. The reference standard was based on a consensus decision made by a multidisciplinary team on the basis of baseline and follow-up images of the index tests, the findings of the baseline index tests by the experienced readers, and any available imaging, biochemical, and clinical data, including the response to ADT. Twenty-seven (73%) of the 37 patients had bone metastases according to the reference standard. The sensitivities for BS, SPECT/CT and NaF PET/CT were 78%, 89%, and 89%, respectively, and the specificities were 90%, 100%, and 90%, respectively. The positive predictive values of BS, SPECT/CT and NaF PET/CT were 96%, 100%, and 96%, respectively, and the negative predictive values were 60%, 77% and 75%, respectively. No statistically significant difference among the three imaging modalities was observed. All three imaging modalities showed high sensitivity and specificity. NaF PET/CT and SPECT/CT showed numerically improved, but not statistically superior, sensitivity compared with BS in this limited and selected patient cohort.

(18)F-fluoride positron emission tomography/computed tomography and bone scintigraphy for diagnosis of bone metastases in newly diagnosed, high-risk prostate cancer patients: study protocol for a multicentre, diagnostic test accuracy study

BACKGROUND

For decades, planar bone scintigraphy has been the standard practice for detection of bone metastases in prostate cancer and has been endorsed by recent oncology/urology guidelines. It is a sensitive method with modest specificity. (18)F-fluoride positron emission tomography/computed tomography has shown improved sensitivity and specificity over bone scintigraphy, but because of methodological issues such as retrospective design and verification bias, the existing level of evidence with (18)F-fluoride positron emission tomography/computed tomography is limited. The primary objective is to compare the diagnostic properties of (18)F-fluoride positron emission tomography/computed tomography versus bone scintigraphy on an individual patient basis.

METHODS/DESIGN

One hundred forty consecutive, high-risk prostate cancer patients will be recruited from several hospitals in Denmark. Sample size was calculated using Hayen's method for diagnostic comparative studies. This study will be conducted in accordance with recommendations of standards for reporting diagnostic accuracy studies. Eligibility criteria comprise the following: 1) biopsy-proven prostate cancer, 2) PSA ≥ 50 ng/ml (equals a prevalence of bone metastasis of ≈ 50% in the study population on bone scintigraphy), 3) patients must be eligible for androgen deprivation therapy, 4) no current or prior cancer (within the past 5 years), 5) ability to comply with imaging procedures, and 6) patients must not receive any investigational drugs. Planar bone scintigraphy and (18)F-fluoride positron emission tomography/computed tomography will be performed within a window of 14 days at baseline. All scans will be repeated after 26 weeks of androgen deprivation therapy, and response of individual lesions will be used for diagnostic classification of the lesions on baseline imaging among responding patients. A response is defined as PSA normalisation or ≥ 80% reduction compared with baseline levels, testosterone below castration levels, no skeletal related events, and no clinical signs of progression. Images are read by blinded nuclear medicine physicians. The protocol is currently recruiting.

DISCUSSION

To the best of our knowledge, this is one of the largest prospective studies comparing (18)F-fluoride positron emission tomography/computed tomography and bone scintigraphy. It is conducted in full accordance with recommendations for diagnostic accuracy trials. It is intended to provide valid documentation for the use of (18)F-fluoride positron emission tomography/computed tomography for examination of bone metastasis in the staging of prostate cancer.

Subcutaneous bleeding: first sign of prostate cancer

We present here a case of subcutaneous bleeding being the first recognized sign of prostate cancer.

Short-chain fatty acids stimulate mucosal cell proliferation in the closed human rectum after Hartmann's procedure

Hartmann's procedure in humans results in a closed rectum deprived of its natural short-chain fatty acid source. This induces atrophy of the entire rectal wall. Ki-67 is a monoclonal antibody directed towards proteins in the cell nucleus that are present only during cell proliferation. This study investigated the effects of short-chain fatty acids on mucosal cell proliferation in the human rectum after Hartmann's procedure by means of Ki-67. Eight patients in whom Hartmann's procedure was performed were treated with placebo and then short-chain fatty acids for 14 days. Biopsies specimens were taken from the rectum before and after treatment; these were prepared with Ki-67 and labeling index was determined. The treatment was found significantly to increase proliferative activity in the rectal mucosa (P<0.01); the increase was principally in the middle (P<0.01) and upper crypt (P<0.05) compartments.

Lewis antigen mediated adhesion of freshly removed human bladder tumors to E-selectin

PURPOSE

Twenty fresh surgical specimens of human bladder tumors were tested for their ability to adhere to recombinant P and E-selectin. The adhesion was correlated to immunological detection of carbohydrate structures.

MATERIALS AND METHODS

A static titertray assay with immobilized selectins and appropriate controls was used for bladder tumor cell adhesion. On the same tumors expression of carbohydrate structures was examined by immunohistochemistry and Western blotting.

RESULTS

No tumor bound to P-selectin. Nine tumors showed a high number of cells binding to E-selectin, 5 showed intermediate binding, and 6 showed only rare binding. The specificity of the binding was verified by inhibition with EDTA, by blocking antibodies to E-selectin, and by an acrylamide based sLe(x) (Galbeta1-4 [Fucalpha1-3]GlcNAc-) polymer. The binding was significantly more frequent (p <0.045) in superficial tumors than in invasive tumors. The binding property was correlated to the detection of carbohydrate structures in Western blots and tissue sections of the same tumors, using six different monoclonal antibodies: anti-sLe(a), anti-sLe(x), anti-Le(a), anti-Le(x) (two different clones) and anti-Le(b). Most blot-stainings were smeared indicating a mucin-type carrier molecule, but 115, 55 and 40 kDa bands carrying Le(a) and/or Le(b) epitopes were present in all tumors that bound. The Le(a) structure, as detected by blotting, was the only structure necessary for binding in the center of the wells (p <0.001), and was correlated to number of bound cells (p <0.006). A weaker correlation was found between Le(b) and number of bound cells (p <0.032), whereas it was remarkable that no correlation was found with Le(x) or sLe(x). Immunohistological staining of Le(a) on cell membranes correlated with frequent binding (p <0.003), whereas no correlation was found to secretor and Lewis genotypes.

CONCLUSIONS

These data on clinical specimens indicate that Lewis antigen mediated E-selectin adhesion may play a role in the human bladder cancer disease.

Surgical repair of vesicovaginal fistulae--a ten-year retrospective study

OBJECTIVE

Vesicovaginal fistulae in the western world generally occur as complications to pelvic surgery or radiation therapy of pelvic cancers. We have reviewed our results of vesicovaginal fistula closure procedures over a 10-year period.

PATIENTS AND METHODS

From 1985 to 1996, 55 patients were referred to our department due to vesicovaginal fistulae. Five patients had fistulae due to malignant recurrence and one patient was considered inoperable. Thus, 49 patients were operated on. Thirty patients had fistulae resulting from pelvic surgery. Nineteen of the 25 patients admitted with fistulae secondary to radiation therapy of pelvic cancers were operated on.

RESULTS

Of the 30 patients with postoperative fistulae, 23 had an abdominal repair and 7 a vaginal repair. A success rate of 90% was achieved after a first closure procedure, as 3 patients within a month experienced a recurrence. These three recurrences were all successfully closed in a second operation, augmenting the success rate to 100% in this group of patients. In the group of patients with fistulae caused by irradiation, a urinary diversion was performed in 12 patients, and in 7 patients a primary attempt to close the fistula was made, either by an abdominal approach (2 patients) or by a vaginal approach (5 patients). The fistula recurred in 6 of these 7 patients. Despite several additional attempts to close the recurrent fistulae, only one patient was successfully operated on.

CONCLUSION

It seems that vesicovaginal fistulae resulting from pelvic surgery, in our hands, can be managed successfully either by an abdominal or vaginal approach. For patients with vesicovaginal fistulae resulting from radiation therapy, a urinary diversion appears to be the method of choice.

Aseptic laparoscopic colon resection with intraabdominal anastomosis. An experimental study in pigs

BACKGROUND

We evaluated a new aseptic method for laparoscopic left colon resection in terms of technical feasibility and outcome.

METHODS

Ten pigs were operated on under general anesthesia. Pre- and postoperative body weight, stools, behavior, and need for analgesics were recorded. Fourteen days later, the animals were killed. At autopsy, the degree of intraabdominal adhesions was noted. The anastomoses were sent for histological examination. The entire procedure was performed intracorporeally, and no antibiotics were given. After division of the mesocolon, the segment to be resected was invaginated down through the colon. This was facilitated by a custom-made instrument that was introduced into the bowel via the anus; it consisted of a pull-out device and a modified diathermy wire. The anastomosis was completed at the invagination fold by a row of hernia staples that were covered by an interrupted suture. Then the invaginated bowel was transected by the diathermy wire and delivered through the anus.

RESULTS

One animal was killed before completion of the operation because of a colonic perforation. The remaining nine animals had an uneventful and rapid recovery. They ate from the 1st postoperative day and gained weight rapidly. Stools were normal after 2 days (median), and normal behaviour was noted in all animals from the 1st postoperative day. At the postmortem examination, intraabdominal adhesions were observed in two animals. In one case, the adhesions extended from a hematoma in the mesentery to the abdominal wall. There were no adhesions to the anastomosis or the colon. In the other case, the anastomosis adhered to the right uterine tube and a loop of small intestines.

CONCLUSIONS

The method is technically feasible, but a modification is suggested for cases where the invagination is impossible. Recovery after the operation is rapid.

A two-site lectinoenzymatic assay for determination of tumour marker glycoproteins in rectal secretions

A method is described for a titre-tray based two-site lectinoenzymatic assay of glycoproteins. WGA lectin, reacting with the core-part of glycans, was combined with lectins PNA and DBA, the latter two reacting with terminal parts of glycans. A standard curve was obtained with bovine submaxillary gland asialomucin, and measurements of human rectal secretion were calibrated against this curve. The assay showed an intra-assay reproducibility of 2.4-7.5%, and inter-assay reproducibility of 3.9-20.8%. Recovery tests showed a linearity close to predicted values. The selected standard was ideal as inhibition of lectin binding by monosaccharides showed similar inhibition profiles for human rectal secretion and for asialomucin standard. Neuraminidase treatment dramatically increased the PNA binding to human rectal secretion immobilized on WGA. Western blotting of human rectal secretion demonstrated a large range of lectin-reactive glycoproteins, the main fraction reacting with all lectins being approximately 250 kDa. The assay described is well suited for studies of the glycan part of tumour marker glycoproteins, and changes occurring in these. It has a high sensitivity by ignoring that the glycans may be present on different molecules. Examination of rectal secretions from various cancer patients showed significantly increased PNA binding, as well as an increased PNA/DBA binding ratio, in patients with colorectal cancer (p<3x10(-3)) and, unexpectedly, in patients with other cancers (p<5x10(-3)).

Loss of ABH antigen expression in bladder cancer is not caused by loss of heterozygosity of the ABO locus

The significance of loss of heterozygosity (LOH) of chromosome 9 as an early event in human bladder tumors has been demonstrated by several studies. We have studied LOH of the ABO gene locus at 9q34.2 by means of polymerase chain reaction (PCR) genotyping of tumor preparations and leukocytes. Twenty-two tumors, all of which were immunohistochemically negative for ABH antigens, were examined. Eleven tumors were from blood-group O individuals, and were examined by denaturing gradient gel electrophoresis (DGGE). They showed normal band patterns with no sign of single base mutations or LOH. All II A and B tumors were sorted on a fluorescence-activated cell sorter (FACS) according to DNA content, thereby making it possible to study differences in genotype between subpopulations of cells in the same tumor. In 2 genotypically AO cases, we found 2 aneuploid subpopulations. In both cases, the most abnormal, with the highest DNA content, showed complete loss of the O allele, leaving the A allele intact. As all tumors were ABH antigen-negative, this study demonstrates that LOH of the ABO locus on chromosome 9q34 is not the cause of loss of blood-group ABH expression in human bladder cancer.

Localization and identification of T-(Gal beta 1-3GalNAc alpha 1-O-R) and T-like antigens in experimental rat bladder cancer

A mouse monoclonal antibody and a rabbit polyclonal antibody against the T-antigen (Gal beta 1-3GalNAc alpha 1-O-R) were used to study the distribution of T-antigens in an experimental rat bladder cancer model. Neoplasia was induced in 28 rats by intravesical installation of N-nitroso-N-methylurea (NMU) dissolved in acetate buffer. Fifteen rats were installed with acetate buffer, and served as controls. Urothelial samples were taken from all animals, the atypia was graded and detailed data on the location of the antibody binding structures were obtained by immunohistochemical methods. In addition, Western Blots of glycoproteins and thin-layer-chromatography (TLC) immunostainings of glycolipids extracted from normal and malignant tissue were performed to characterize the molecules presenting T-antigens. Examination of the histologic distribution of T-antigens showed that both the monoclonal and the polyclonal reagents reacted with atypical cells in proportion to the grade of atypia, but showed no reaction in invasive cells. These results confirm previously obtained data on the T-antigen using peanut (arachis hypogaea) agglutinin (PNA), and support the structure identity as being the classical O-linked mucin type T-antigen. Western blots of tumor glycoproteins showed that the monoclonal and the polyclonal antibody reacted with epitopes different from that of PNA, but all the probes correlated with atypia. In addition PNA, as the only anti-T reagent, bound to glycolipid. By using well characterized and highly specific immunological reagents the present study shows that the T-antigen is a highly selective marker of urothelial atypia.

Human urinary bladder carcinoma glycoconjugates expressing T-(Gal beta(1-3)GalNAc alpha 1-O-R) and T-like antigens: a comparative study using peanut agglutinin and poly- and monoclonal antibodies

T- and T-like antigens on glycoproteins and glycolipids were examined in extracts of human urinary bladder tumors and normal tissue by Western blot analysis and reagent binding to thin layer chromatograms. Three different anti-T-reagents were used: peanut (Arachis hypogaea) lectin (PNA) and mono- and polyclonal antibodies specific for T-antigen (Gal beta(1-3)GalNAc alpha 1-O-R). Immunodetection with the T-specific reagents in nitrocellulose replicas of bladder tumor glycoproteins, separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, demonstrated tumor-specific T-antigen-bearing glycoproteins compared to normal urothelial glycoproteins. In addition, a remarkable difference in binding was found between the immunological reagents and PNA lectin. PNA showed major reactivity to a 28-kD glycoprotein extracted from tumors. Monoclonal anti-T-antibody (clone HH8) showed major reactivity with an M(r) 34,000 glycoprotein, and polyclonal anti-T-antibody showed major reactivity with an M(r) 36,000 glycoprotein. PNA agarose column affinity-purified tumor glycoproteins did not bind the antibodies. Glycoproteins, M(r) 28,000 and 34,000, were shown to be O-linked by stepwise deglycosylation. In solid phase monosaccharide inhibition tests, galactose followed by N-acetyl-galactosamine were the most potent monosaccharides inhibiting binding to immobilized bladder tumor glycoproteins. None of the anti-T-reagents reacted with glycolipids extracted from tumor tissue. It is concluded that PNA lectin, in addition to the T-disaccharide, reacts with other protein-anchored carbohydrate structures in carcinomas.

Expression of histo-blood-group-A/B-gene-defined glycosyltransferases in normal and malignant epithelia: correlation with A/B-carbohydrate expression

Malignant transformation of oral and bladder epithelia is often associated with loss of histo-blood-group-A- and -B-carbohydrate antigens, whereas these antigens, which are absent in normal adult distal colon (but present in fetal colon) reappear in malignant distal colon. In order to gain insight into the genetic basis of the biosynthetic regulation for these changes, we have correlated the expression of the A- and B-carbohydrate antigens with that of the A/B-gene-defined glycosyltransferases in colon, bladder and oral carcinomas by immunohistology. A newly developed anti-A/B-transferase monoclonal antibody (MAb) was used to demonstrate the in situ localization of transferase expression at the individual cell level with correlation to carbohydrate antigen expression, and gave the essential information that the transferase is derived from the ABO gene complex. The reappearance of A- and B-carbohydrate antigens in carcinomas of the distal colon was found to be unrelated to the expression of the A/B-transferase proteins, which were expressed throughout normal adult colon in accordance with previous enzymatic studies. In contrast, the loss of A- and B-carbohydrate antigens in malignant bladder and oral epithelia was accompanied by concordant loss of enzymes.

Nuclear volume and expression of T-antigen, sialosyl-Tn-antigen, and Tn-antigen in carcinoma of the human bladder. Relation to tumor recurrence and progression

The T-antigen system and the mean nuclear volume have been proposed as risk variables in bladder tumors. This study includes 34 patients with initially noninvasive (Ta) transitional cell carcinomas who experienced different courses of disease. Tissue specimens of primary tumors were analyzed for the expression of T-antigen, Tn-antigen, and sialosyl-Tn-antigen using monoclonal antibodies (MoAb) and the lectin peanut agglutinin (PNA) in an indirect immunoperoxidase method. In addition, the mean nuclear volume was estimated by morphometry. Tissue from 7 of 13 patients (54%) who had invasive disease during a follow-up period of 5 years expressed T-antigen, as defined by MoAb HH8 in the primary tumor, whereas tissue of only 3 of 21 patients who did not have invasive disease expressed the antigen (P less than 0.02). No association was found between tumor progression to invasion and the expression of Tn-antigen or sialosyl-Tn-antigen. Tn-antigen expression was partially lost in invasive tumors (P less than 0.03) when compared with the expression in primary noninvasive tumors. A high mean nuclear volume in tissue specimens of primary tumors correlated with a progression to invasive disease (P less than 0.01). A significantly (P less than 0.003) higher mean nuclear volume was found in tumor areas that were positive for PNA compared with areas that were negative for PNA in primary tumors. A significantly lower mean nuclear volume was found in Tn-antigen-positive invasive Grade 3 tumor areas than in Tn-antigen-negative areas (P less than 0.005). The combined use of T-antigen expression and mean nuclear volume is of potential clinical interest for determining patients who are at high risk of disease progression.

Nuclear volume and expression of T-antigen, sialosyl-Tn-antigen, and Tn-antigen in carcinoma of the human bladder. Relation to tumor recurrence and progression

The T-antigen system and the mean nuclear volume have been proposed as risk variables in bladder tumors. This study includes 34 patients with initially noninvasive (Ta) transitional cell carcinomas who experienced different courses of disease. Tissue specimens of primary tumors were analyzed for the expression of T-antigen, Tn-antigen, and sialosyl-Tn-antigen using monoclonal antibodies (MoAb) and the lectin peanut agglutinin (PNA) in an indirect immunoperoxidase method. In addition, the mean nuclear volume was estimated by morphometry. Tissue from 7 of 13 patients (54%) who had invasive disease during a follow-up period of 5 years expressed T-antigen, as defined by MoAb HH8 in the primary tumor, whereas tissue of only 3 of 21 patients who did not have invasive disease expressed the antigen (P less than 0.02). No association was found between tumor progression to invasion and the expression of Tn-antigen or sialosyl-Tn-antigen. Tn-antigen expression was partially lost in invasive tumors (P less than 0.03) when compared with the expression in primary noninvasive tumors. A high mean nuclear volume in tissue specimens of primary tumors correlated with a progression to invasive disease (P less than 0.01). A significantly (P less than 0.003) higher mean nuclear volume was found in tumor areas that were positive for PNA compared with areas that were negative for PNA in primary tumors. A significantly lower mean nuclear volume was found in Tn-antigen-positive invasive Grade 3 tumor areas than in Tn-antigen-negative areas (P less than 0.005). The combined use of T-antigen expression and mean nuclear volume is of potential clinical interest for determining patients who are at high risk of disease progression.

Lewis antigen expression in benign and malignant tissues from RBC Le(a-b-) cancer patients

Eight red blood cell (RBC) Le(a-b-) individuals were selected from a series of patients with bladder or colon cancer. Defined by the presence or absence of alpha 1-4-L-fucosyltransferase activity in saliva, four of these patients were characterized as non-genuine Lewis negative [RBC Le(a-b-) with alpha 1-4-L-fucosyltransferase activity in saliva], and four as genuine Lewis negative [RBC Le(a-b-) with no alpha 1-4-L-fucosyltransferase activity in saliva]. Stainings of paraffin embedded formalin fixed tissue sections for Lea and Leb antigens were performed by means of an indirect immunohistochemical method on all malignant and benign tissue previously removed from these eight patients. Leb antigens were always expressed independently of both the Lewis and the secretor status of the individual. Lea antigens, on the other hand, showed a different staining pattern. Although primarily expressed in non-genuine Le(a-b-) individuals, Lea antigens were expressed in genuine Le(a-b-) individuals as well--to a limited extent, but still detectable. Thus, these findings seem to show that the Lewis antigen expression is tissue dependent, and it is not possible to predict tissue Lewis antigen expression by merely examining erythrocytes or saliva.

Frequency and mechanism of Lewis antigen expression in human urinary bladder and colon carcinoma patients

Changes in the expression of Lewis antigens have been associated with cancer diseases, and recent results have pointed at a possible increased risk of cancer development among Lewis negative patients. The frequency of the erythrocyte Lewis phenotypes Le(a- b+), Le(a+ b-) and Le(a- b-) was analysed in patients suffering from urinary bladder cancer (82), colon cancer (21), and benign urological diseases (45). An increased frequency of Lewis negative individuals was found among colon cancer patients (P less than 0.004) and bladder cancer patients (P = 0.05). The Lewis negative phenotype was shown to be associated with unfavourable disease parameters: invasion (P less than 0.02) and high grade of atypia (P less than 0.01) in bladder cancer patients, and high Dukes stage (P less than 0.05) in colon cancer patients. alpha 1-4fucosyltransferase activity (Lewis transferase) was shown to be present in saliva from four out of eight erythrocyte Lewis negative cancer patients, indicating that some patients with advanced cancer disease may have converted from a Lewis positive to a Lewis negative phenotype.

Cellular localization of PNA binding in colorectal adenomas: comparison with differentiation, nuclear:cell height ratio and effect of desialylation

The lectin Arachis Hypogaea (Peanut Agglutinin, PNA) was used to study the cellular localization of the Thomsen-Friedenreich (T) disaccharide Gal-beta (1-3)-GalNAc alpha 1-R in 22 formalin-fixed paraplast-embedded colorectal adenomas of varying cellular dysplasia. An indirect immunoperoxidase method was used prior to and after neuraminidase treatment. Detailed information on the cellular localization of PNA binding was obtained. In addition, morphometric measurements of the nuclear: cell height ratios were performed on staining-filtered micrographs of crypts from all adenomas. We found 1) a statistically significant increase in the nuclear:cell height ratio with increasing grade of dedifferentiation (p less than 0.003), 2) a statistically significant smaller nuclear:cell height ratio in crypts that were PNA-positive in the Golgi region when these were compared to crypts that were PNA-positive on luminal cell membranes, 3) a decreasing number of crypts expressing PNA binding sites in the Golgi region with increasing dedifferentiation, leading to complete absence of PNA binding sites in Grade IV adenomas, 4) neuraminidase pretreatment increased the number of crypts expressing PNA binding sites in cytoplasm and on luminal membranes, whereas no changes were detected in crypts expressing PNA binding sites in the Golgi region. Our results confirm the general concept of accumulation of precursors of carbohydrate antigens in dedifferentiated cells. On the basis of the results presented, we conclude that the nuclear:cell height ratio shows a good correlation with the cellular localization of PNA binding, cellular differentiation and classic pathologic grading.

O-linked mucin-type glycoproteins in normal and malignant colon mucosa: lack of T-antigen expression and accumulation of Tn and sialosyl-Tn antigens in carcinomas

The expression of carbohydrate core-structures on O-linked glycoproteins was examined in fetal (n = 6), infantile (n = 2), normal adult (n = 15), and malignant (n = 22) colorectal tissue by means of monoclonal antibodies (MAbs) specific for Tn (GalNAc alpha 1-O-R), sialosyl-Tn (NeuAc alpha 2-6GalNAc alpha 1-O-R), and T (Gal beta 1-3GalNAc alpha 1-O-R) antigens. Immunolabelling of solubilized malignant tissue, separated by SDS-PAGE, showed expression of Tn and sialosyl-Tn antigens on 3 molecules of similar mw (230, 210, and 170 kDa), whereas no T antigens could be detected. Immunohistochemical techniques showed that fetal colon mucosa expressed Tn antigens but no sialosyl-Tn antigens. Infantile colon mucosa, however, expressed Tn as well as sialosyl-Tn antigens, and normal adult colon mucosa cells expressed no Tn antigens but sialosyl-Tn in 2 out of 6 biopsies from cecum, which indicates occurrence post partum of alpha-6-NeuAc-transferase. Endothelium in normal adult mucosa showed expression of both Tn and sialosyl-Tn antigens; 82% of carcinoma tissue sections expressed Tn antigens, and 73% expressed sialosyl-Tn antigens in mucin or cytoplasm, or on luminal cell membranes. T antigens could be detected neither in normal mucosa cells at any stage of development, nor in carcinomas. The possibility exists that this could be due to masking of T antigen. Mucin-type blood-group A antigens which contain an internal T-disaccharide were demonstrated in 4 out of 4 A1 tumors by means of MAb HH5. However, urea-containing SDS-PAGE analysis demonstrated an HH5 binding to molecules different from those binding anti-Tn. In remote morphologically normal and abnormal crypts in colons from carcinoma patients, both Tn and sialosyl-Tn antigens were expressed in secreted mucin in 40% of the cases. The data indicate an expression of O-linked Tn and sialosyl-Tn core structures in fetal and infantile colon and in colorectal carcinomas.

Binding of wheat and peanut lectins to human transitional cell carcinomas. Correlation with histopathologic grade, invasion, and DNA ploidy

The binding of peanut (PNA) and wheat germ (WGA) lectins to tissue sections was examined in biopsy specimens from normal urothelium (ten patients) and from tumor tissue of noninvasive (17 patients) and invasive bladder (31 patients) carcinomas. The results were correlated to DNA content, histopathologic grade, and the presence or absence of invasion. Significant alterations in lectin binding associated with the development of cancer were found. A gradual loss of both PNA and WGA binding was found to correlate with higher grades of atypia (P less than 0.001). The loss of WGA binding was significantly correlated with both tumor aneuploidy (P less than 0.001) and the presence of invasion (P less than 0.05), whereas no significant correlation was found between loss of PNA binding and these variables. We concluded that the loss of WGA binding structures associated with bladder cancer shows a better correlation with known risk factors (aneuploidy and invasion) than the loss of PNA binding does.

[Treatment and socioeconomic consequences of skiing injuries in the County of Vejle during the 1986/1987 season]

In a prospective study conducted in the County of Vejle in the skiing-season 1986/1987, the hospitals were contacted by 129 persons who had skiing-injuries. Forty-one patients were primarily treated abroad and 35 patients were primarily examined or treated in Denmark after their return. Fifty-three patients who were injured in Denmark were primarily examined or treated in the casualty ward. The injuries the most often seen were knee and thumb-injuries. Twenty-three operations were undertaken, 15 on knees, and five on thumbs. A total of 139 bed-days were registered, 114 for injuries abroad and 25 for Danish injuries. Fifty-nine injured persons lost 1,481 days on account of sickness. The yearly nationwide costs for the health services in connection with skiing-injuries are calculated to be 9.2 million kroner (approximately 760,000 pounds). The expenses for loss from production are calculated to be 8.6 million kroner (approximately 716,000 pounds).

Immunohistochemistry and cytochemistry of experimental rat bladder cancer: binding of the lectins PNA and WGA and of a Le(Y) mouse monoclonal antibody

Two lectins, peanut agglutinin (PNA) and wheat germ agglutinin (WGA), and a mouse monoclonal antibody against blood group Le(Y) were used to study the distribution of carbohydrate antigens in an experimental rat bladder cancer model. Neoplasia was induced in 28 rats by intravesical installation of N-nitroso-N-methyl-urea (NMU). Fifteen rats were installed with the NMU solvent and served as controls. Urothelial samples were taken from all animals, the atypia were graded and detailed data on the location of the lectin-binding structures and the binding of the monoclonal antibody were obtained by immunohistochemical methods. Urine samples were collected at different times during tumor development and data on the cytological location of the lectin-binding structures and the binding of the monoclonal antibody were obtained by immunocytochemical methods. Examination of the histological distribution of lectin binding structures and Le(Y) antigen showed a characteristic change in carbohydrate antigen expression associated with the development of urothelial atypia. In normal urothelium carbohydrate antigens were present in cytoplasm, whereas they became expressed on cell membranes in non-invasive carcinomas. Immunocytochemistry of urine specimens revealed a lack of correlation between the antigen expression found in exfoliated cells and that found in tissue. Morphologically normal cells from NMU-treated animals were found to express a carcinoma associated antigen pattern.

Lectinohistochemistry of human bladder cancer: loss of lectin binding structures in invasive carcinomas

With the purpose of studying changes in the expression of glycoconjugate structures in urothelium, nine different lectins (PNA, WGA, VFA, GSA II, STA, UEA I, LCA, DBA and HPA) with specificity for mono- or oligo-saccharides were used on formalin-fixed, paraffin-embedded tissue sections from 47 patients who had undergone surgical resection for bladder tumors and on normal urothelial biopsies from 10 patients. The tumors were graded and a lectinohistochemical method using biotinylated lectins and avidin-biotin-peroxidase complex was used to demonstrate the lectin binding. Positive staining reactions of cells in cytoplasm and on membranes were evaluated in the basal, the intermediate, and the luminal cell layers, respectively. In both normal and atypical urothelium lectin binding predominated in the luminal cell layer and decreased towards the basal cell layer. In normal urothelium all lectins stained greater than 66% of the cells in the luminal cell layer in cytoplasm and between 5 and 100% of the cells on membranes depending on the lectin used. A gradual loss of lectin-binding structures was seen with increasing grade of atypia. The range of this decrease varied considerably from one lectin to another, but it was consistently found that the percentage of cells stained in cytoplasm and on membranes decreased. A significantly lower percentage of cells stained in cytoplasm was found in invasive tumor cell-islands compared to normal urothelium. In invasive tumor cell-islands staining of cells on membranes was completely absent, except for HPA lectin that stained less than 10% of the cells. In conclusion, we demonstrate a dramatic decrease in lectin-binding carbohydrate structures associated with urothelial malignant progression.