Electromagnetic stabilization of weakly conducting fluids.

Classical hydromagnetic theory predicts that the flow of dilute aqueous electrolyte in a slit can be stabilized by application of a strong, transverse magnetic field. However, recent experiments indicate that stabilization can be achieved with the use of a much weaker field in the presence of a small lateral current. A revised theory describes how the magnetic and electric fields interact to eliminate natural convection.

Epidemiological features of Wilms' tumor: results of the National Wilms' Tumor Study.

Nearly 2,000 children with Wilm's tumor registered in a national clinical trial during 1969-81 showed high rates of aniridia, hemihypertrophy, cryptorchidism, hypospadias, and other genitourinary anomalies. Patients with bilateral disease, who constituted 5% of the total, had younger ages at diagnosis and an increased incidence of congenital anomalies and renal blastemal rests. Those with multicentric unilateral lesions had more blastemal rests but were otherwise indistinguishable from the unicentric cases. The 20 familial cases had none of the features usually associated with genetic tumors: neither younger ages nor an increase in bilaterality nor associated congenital anomalies. These observations suggest that the fraction of Wilm's tumors that is due to an inherited mutation may be substantially smaller than previously supposed and support the concept that the disease arises from a variety of pathogenetic pathways.

Clonal expansion and attenuated apoptosis in Wilms' tumors are associated with p53 gene mutations.

The p53 gene product is required for activation of an apoptotic pathway triggered by oncogenes and cytotoxic agents. Wilms' tumor, a pediatric renal malignancy, provides a paradigm for evaluating genetic events involved in tumor progression. This malignancy is generally not associated with p53 mutations, and even in advanced disease states is quite responsive to current treatment regimens. The anaplastic histological variant of Wilms' tumor, however, is frequently associated with p53 gene mutations and shows poor prognosis. We analyzed seven Wilms' tumors for which we had paired samples from nonanaplastic and anaplastic regions. p53 mutations were detected in six of these tumors, five of which demonstrated mutations restricted to anaplastic regions. Nonanaplastic cells of the sixth sample were heterozygous for a p53 mutation, whereas the anaplastic area of this tumor showed reduction to homozygosity. These results indicate that progression to anaplasia is associated with clonal expansion of cells which have acquired a p53 mutation. We demonstrated that tumor cells with p53 mutations show attenuated apoptosis, suggesting that such lesions may provide a selective advantage in vivo by decreasing cell death.

Age distribution of Wilms' tumor: report from the National Wilms' Tumor Study.

Median ages at diagnosis were 36.5 and 42.5 mo for 1523 males and 1678 females with unilateral Wilms' tumor registered between October 1969 and December 1985; they were 23.5 mo and 30.5 mo for 100 males and 141 females with bilateral disease. The median age for multicentric, unilateral cases was intermediate between the bilateral and unicentric medians. Patients with hemihypertrophy in addition to their Wilms' tumor had a typical age distribution, whereas those with aniridia or characteristic genitourinary anomalies were substantially younger. Patients with perilobar nephroblastomatosis had a median age of 35.5 mo and those with intralobar nephroblastomatosis, a median age of 18.5 mo. Most of the bilateral disease occurred in the presence of one or both of these precursor lesions. These findings suggest heterogeneity in the pathogenesis of Wilms' tumor, having implications for genetic counselling, and call into question certain aspects of Knudson and Strong's two-stage mutational model for the origin of Wilms' tumor.

Increased birth weights of National Wilms' Tumor Study patients suggest a growth factor excess.

An analysis of over 1800 patients with Wilms' tumor revealed significantly higher birth weights than newborns in the general United States population. The highest birth weights were found not only in patients diagnosed with the Beckwith-Wiedemann syndrome (mean, 3.78 kg), as had been expected, but also in those with hemihypertrophy (3.80 kg) or perilobar nephrogenic rests (3.56 kg) in addition to their Wilms' tumor. The birth weights of Wilms' tumor patients with intralobar nephrogenic rests (3.43 on average kg) and of those without associated anomalies (3.45 kg) were slightly but still significantly higher on average than national birthweights (overall mean, 3.35 kg) adjusted for gender, race, and year of birth in each subgroup. Birth weights of children with aniridia and Wilms' tumor (2.99 kg) were lower than the national mean. Among more than 3000 patients with Wilms' tumor, heights and weights at diagnosis were significantly higher for the subgroups of patients with Beckwith-Wiedemann syndrome or hemihypertrophy, and height was lower for those with aniridia or characteristic genitourinary anomalies, when compared to other patients with Wilms' tumor. These data suggest prenatal effects of growth factors on the development of Wilms' tumors, or vice versa, and provide further epidemiological support for heterogeneity in the pathogenesis of Wilms' tumors associated with perilobar nephrogenic rests versus intralobar nephrogenic rests.

ETV6-NTRK3 gene fusions and trisomy 11 establish a histogenetic link between mesoblastic nephroma and congenital fibrosarcoma.

Congenital mesoblastic nephroma (CMN) is an infantile spindle cell tumor of the kidney that is subdivided into "classical" and "cellular" forms based on the degree of cellularity and mitotic activity. The histogenesis of CMN remains obscure, but relationships to other pediatric renal neoplasms have been proposed. However, cellular CMN is virtually identical histologically to congenital fibrosarcoma (CFS), a malignant tumor of fibroblasts in children of the same age group. Moreover, cytogenetic studies have reported common trisomies in CFS and cellular CMN, particularly of chromosome 11. We show here that t(12;15)(p13;q25)-associated ETV6-NTRK3 gene fusions described in CFS are also present in cellular CMN. ETV6-NTRK3 chimeric transcripts were detected in 8 of 9 cellular CMNs and 2 of 2 mixed CMNs. In contrast, all of the four classical CMNs tested were negative, as were cases of Wilms' tumor and clear cell sarcoma of the kidney. Moreover, we found trisomy 11 only in cellular or mixed CMNs with the ETV6-NTRK3 gene fusion. Our studies indicate that classical and cellular CMN have different genetic features and support the concept that cellular CMN is histogenetically related to CFS. They also provide insight into potential mechanisms involved in the transformation of the classical into the cellular form of CMN.

Treatment outcomes in patients less than 2 years of age with small, stage I, favorable-histology Wilms' tumors: a report from the National Wilms' Tumor Study.

PURPOSE: Retrospective analyses were performed to determine the effect of tumor weight and therapy modifications on outcome in patients less than 2 years of age with stage I favorable-histology Wilms' tumors. PATIENTS AND METHODS: The 4-year relapse-free and overall survival percentages for patients randomized to different treatment regimens in National Wilms' Tumor Studies (NWTS)-1, -2, and -3 were calculated and compared. RESULTS: The 4-year relapse-free survival percentages of patients whose specimen weight was less than 550 g were found to be 89.1% on NWTS-1, 96.0% on NWTS-2, and 93.2% on NWTS-3. There was no evidence that the relapse-free survival of these patients had improved over time (P value for trend = .99). The 4-year relapse-free survival percentage for similar age and stage patients whose specimen weight was 550 g or greater was significantly poorer than that of patients with smaller tumors (P = .02). CONCLUSION: Changes in the NWTS treatment regimens over a period of more than 20 years have not improved on the excellent prognosis of patients who are less than 2 years of age at diagnosis and who have a stage I, favorable-histology Wilms' tumor with specimen weight less than 550 g. These data could be used as the basis for a future trial in which a subgroup of such patients is treated with nephrectomy only.

Anaplastic Wilms' tumor: clinical and pathologic studies.

A review of almost 1,200 children participating in the first and second National Wilms' Tumor Study (NWTS-1 and -2) has demonstrated a number of significant differences in the clinical presentation and response to therapy of anaplastic and nonanaplastic Wilms' tumor. Compared to their counterparts, children with anaplastic Wilms' tumor were generally one to two years older at diagnosis, more were non-white, and more had lymph node metastases at diagnosis. Consistent with previous studies, children with anaplastic Wilms' tumor survived for a significantly shorter time than those with non-anaplastic Wilms' tumor. A hopeful outlook, however, was suggested by the NWTS-2 experience since the more aggressive chemotherapies used in this study appear to have substantially improved the survival of patients with diffusely anaplastic tumors. Also, the survival of NWTS-2 patients with anaplastic Wilms' tumor was determined in part by clinicopathologic stage. It may be possible therefore to refine therapy according to stage so as to provide children with localized disease a chance for cure with fewer untoward treatment-related sequelae.

Treatment of children with stages II to IV anaplastic Wilms' tumor: a report from the National Wilms' Tumor Study Group.

PURPOSE: To evaluate the effect of the combination of vincristine, dactinomycin, and doxorubicin with (regimen J) or without (regimen DD-RT) cyclophosphamide on the relapse-free survival of children with stages II to IV Wilms' tumor and focal or diffuse anaplasia. PATIENTS AND METHODS: We reviewed the clinical courses of all randomized patients from National Wilms' Tumor Study (NWTS)-3 and NWTS-4 with stages II to IV anaplastic Wilms' tumor, and determined the 4-year relapse-free survival rate separately for those with focal or diffuse anaplasia. Anaplasia was evaluated using newly developed topographic definitions for focal and diffuse anaplasia. RESULTS: The 4-year relapse-free survival rate for five children with focal anaplasia who received regimen DD-RT was 80.0%, compared with 100.0% for eight children who received regimen J (P = .68). The 4-year relapse-free survival rate for 29 children with diffuse anaplasia treated with regimen DD-RT was 27.2%, compared with 54.8% for 30 children treated with regimen J (P = .02). CONCLUSION: We conclude that children with focal anaplasia have an excellent prognosis when treated with vincristine, doxorubicin, and dactinomycin. The addition of cyclophosphamide to the three-drug treatment regimen improved the 4-year relapse-free survival rate of children with stage II to IV diffuse anaplasia. This result suggests that further intensification of the treatment regimen for children with diffuse anaplasia may result in an additional improvement in prognosis.

Treatment of children with clear-cell sarcoma of the kidney: a report from the National Wilms' Tumor Study Group.

PURPOSE: To evaluate the effect of the sequential addition of doxorubicin (DOX) and cyclophosphamide (CTX) to the combination of vincristine (VCR) and dactinomycin (AMD) on the relapse-free survival of children with clear-cell sarcoma of the kidney (CCSK). PATIENTS AND METHODS: We determined the 6-year relapse-free survival rate for patients with CCSK treated on National Wilms' Tumor Study (NWTS)-1, NWTS-2, or NWTS-3 with the combination of VCR and AMD, with or without DOX, and for patients treated on NWTS-3 with the combination of VCR, AMD, and DOX with (regimen J) or without (regimen DD-RT) CTX. RESULTS: The 6-year relapse-free survival rate for the eight children with CCSK treated with VCR, AMD, and radiation therapy was 25.0%, compared with 63.5% for the 58 children treated with VCR, AMD, DOX, and radiation therapy (P = .09). The 6-year relapse-free survival rate for children with CCSK treated on regimen DD-RT was 64.6%, compared with 58.2% for those treated on regimen J (P = .79). CONCLUSION: We conclude that the addition of DOX to the combination of VCR plus AMD appeared to improve the 6-year relapse-free survival rate of children with CCSK. The addition of CTX in the dose and schedule used in NWTS-3 did not improve the 6-year relapse-free survival rate of children with CCSK. Because 30% of relapses occurred more than 2 years after diagnosis, prolonged follow-up evaluation of patients with CCSK is necessary.

Prognosis for Wilms' tumor patients with nonmetastatic disease at diagnosis--results of the second National Wilms' Tumor Study.

Multivariate statistical methods were used to study prognosis for 632 patients entered on the second National Wilms' Tumor Study who had nonmetastatic, unilateral disease at diagnosis. Separate analyses were conducted for each of four endpoints: abdominal recurrence, distant metastasis, relapse without regard to site, and death. The two most important predictors for metastasis and general relapse were an unfavorable (anaplastic or sarcomatous) histology and the presence of microscopically confirmed disease in the regional lymph nodes. Operative spillage of tumor increased the rates of abdominal recurrence and death, even after accounting for histology and lymph node effects. The presence of a tumor thrombus in the renal vein or IVC increased the risk of metastasis, and intrarenal vascular invasion was associated with general relapse after accounting for histology, lymph nodes, and spillage. However, these latter associations were weaker, and some uncertainty remains regarding the true prognostic import of such findings due to a high degree of collinearity among variables. By contrast to the results of a similar data analysis for the first National Wilms' Tumor Study, specimen weight had no bearing on outcome, and the effects of age at diagnosis were entirely explained by the association of age with other more critical factors.

Nuclear morphometry and prognosis in favorable histology Wilms' tumor: A prospective reevaluation.

PURPOSE: This study was designed to evaluate the ability of a previously published nuclear morphometry discriminant function to predict disease-free survival in patients with Wilms' tumor. PATIENTS AND METHODS: We identified 218 patients with stage I-IV Wilms' tumor of favorable histology who were entered onto the National Wilms' Tumor Study (NWTS) between January 1, 1990 and April 15, 1994. The nuclear morphometry score was calculated for each patient as follows: MV(f) = (0.02 x AGE) + (1.17 x SNRF) + (90.6 x LEFD) - 94, with AGE denoting age at diagnosis in months, SNRF the skewness of the nuclear roundness factor, and LEFD the lowest value of nuclear ellipticity as measured by the feret diameter method. Relative risks of relapse were estimated for the total score and for each of its components. Sensitivity and specificity were determined for the criterion of "MV(f) is greater than -0.35" as a predictor of relapse. RESULTS: By contrast with previously published results, neither the SNRF nor the LEFD made any contribution to the prediction of disease-free survival. Sensitivity and specificity of the criterion of "MV(f) is greater than -0.35" were 71% and 56%, respectively. CONCLUSION: Re-evaluation of a published nuclear morphometry score showed that it did not predict disease-free survival in patients with Wilms' tumor. The earlier study very likely overestimated the predictive power of nuclear morphometry by using the same data set both to develop the score and to evaluate its properties. Because of the huge number of combinations of nuclear morphometry measurements that may enter into the multivariate discriminant function, use of appropriate statistical methods is essential to estimate accurately the sensitivity and specificity.

Effect of duration of treatment on treatment outcome and cost of treatment for Wilms' tumor: a report from the National Wilms' Tumor Study Group.

PURPOSE: National Wilms' Tumor Study (NWTS)-4 was designed to evaluate the efficacy, toxicity, and cost of the administration of different regimens for the treatment of Wilms' tumor (WT). PATIENTS AND METHODS: Between August 6, 1986 and September 1, 1994, 905 previously untreated children aged younger than 16 years with stage II favorable histology (FH) WT (low-risk [LR]), stages III to IV FH WT, or stages I to IV clear-cell sarcoma of the kidney (high-risk[HR]) were randomized after the completion of 6 months of chemotherapy to discontinue (short) or continue for 9 additional months (long) treatment with chemotherapy regimens that included vincristine and either divided-dose (standard [STD]) courses (5 days) or single-dose (pulse-intensive [PI]) treatment with dactinomycin. HR patients also received either divided-dose (STD) courses (3 days) or single-dose (PI) treatment with doxorubicin. RESULTS: The 4-year relapse-free survival (RFS) rates after the second randomization for LR patients were 83.7% for the 190 patients treated with short and 88.2% for the 187 patients treated with long chemotherapy (P = .11). The 4-year RFS rates after the second randomization for HR FH patients were 89.7% for the 256 patients treated with short and 88.8% for the 246 patients treated with long chemotherapy (P = .87). The charge for treatment with the short PI treatment regimens for all children with stages I through IV FH WT was approximately one half of that with the long STD treatment regimens. CONCLUSION: The short administration schedule for the treatment of children with WT is no less effective than the long administration schedule and can be administered at a substantially lower total treatment cost.

Comparison between single-dose and divided-dose administration of dactinomycin and doxorubicin for patients with Wilms' tumor: a report from the National Wilms' Tumor Study Group.

PURPOSE: The National Wilms' Tumor Study (NWTS)-4 was designed to evaluate the efficacy, toxicity, and cost of administration of different regimens for the treatment of Wilms' tumor (WT). PATIENTS AND METHODS: Between August 6, 1986 and September 1, 1994, 1,687 previously untreated children less than 16 years of age with stages I to II/favorable histology (FH) or stage I/anaplastic histology WT (low-risk [LR] group) or stages III to IV/FH WT or stages I to IV/clear cell sarcoma of the kidney (high-risk [HR] group) were randomized to treatment that included vincristine and either divided-dose (standard [STD]) courses (5 days) or single-dose (pulse-intensive [PI]) treatment with dactinomycin. HR patients also received either STD courses (3 days) or PI treatment with doxorubicin. RESULTS: The 2-year relapse-free survival (RFS) rates for LR patients were 91.3% for 544 randomized to treatment with PI and 91.4% for 556 randomized to treatment with STD chemotherapy (P = .988). The 2-year RFS rates for HR patients were 87.3% for 299 randomized to treatment with PI and 90.0% for 288 randomized to treatment with STD chemotherapy (P = .865). CONCLUSION: We conclude that patients treated with PI combination chemotherapy for LR or HR WT or clear cell sarcoma of the kidney have equivalent 2-year RFS to those treated with STD regimens. PI drug administration is recommended as the new standard based on demonstrated efficacy, greater administered dose-intensity, less severe hematologic toxicity, and the requirement for fewer physician and hospital encounters.

Treatment with nephrectomy only for small, stage I/favorable histology Wilms' tumor: a report from the National Wilms' Tumor Study Group.

PURPOSE: Children younger than 24 months with small (< 550 g), favorable histology (FH) Wilms tumors (WTs) were shown in a pilot study to have an excellent prognosis when treated with nephrectomy only. PATIENTS AND METHODS: A study of nephrectomy only for the treatment of selected children with FH WT was undertaken. Stringent stopping rules were designed to insure closure of the study if the true 2-year relapse-free survival rate was 90% or lower. RESULTS: Seventy-five previously untreated children younger than 24 months with stage I/FH WTs for which the surgical specimen weighed less than 550 g were treated with nephrectomy only. Three patients developed metachronous, contralateral WT 1.1, 1.4, and 2.3 years after nephrectomy, and eight patients relapsed 0.3 to 1.05 years after diagnosis (median, 0.4 years; mean, 0.51 years). The sites of relapse were lung (n = 5) and operative bed (n = 3). The 2-year disease-free (relapse and metachronous contralateral WT) survival rate was 86.5%. The 2-year survival rate is 100% with a median follow-up of 2.84 years. The 2-year disease-free survival rate (excluding metachronous contralateral WT) was 89.2%, and the 2-year cumulative risk of metachronous contralateral WT was 3.1%. CONCLUSION: Children younger than 24 months treated with nephrectomy only for a stage I/FH WT that weighed less than 550 g had a risk of relapse, including the development of metachronous contralateral WT, of 13.5% 2 years after diagnosis. All patients who experienced relapse on this trial are alive at this time. This approach will be re-evaluated in a clinical trial using a less conservative stopping rule.

Wilm's Tumor: An Update.

Major advances have been made against Wilms' tumor as a result of treatment methods developed by single institutions that then have been confirmed and extended by national cooperating groups. Better survival rates have been achieved, and therapy has been refined so that treatment can be reduced in early stage disease without jeopardizing tumor control. This results in fewer short- and long-term complications, an especially important consideration in children. Their organs and tissues are vulnerable to anti-mitotic treatments such as chemo- and radiotherapy, that can produce disabling if not lethal dysfunctions. This progress has been the result of the cooperative efforts by multiple specialists, and provides evidence of the value of such integrated studies. They have changed the outlook from a 90% death rate in the early years of this century to the 90% survival rate now possible with modern management.

Metanephric stromal tumor: report of 31 cases of a distinctive pediatric renal neoplasm.

We report 31 cases of a novel pediatric renal neoplasm, metanephric stromal tumor (MST). Mean patient age was 2 years, and the most common presentation was that of an abdominal mass. Gross examination typically revealed a fibrous lesion centered in the renal medulla containing smooth-walled cysts (mean tumor size, 5.5 cm). MST is histologically identical to the stromal component of metanephric adenofibroma (MAF, previously termed nephrogenic adenofibroma) and is an unencapsulated spindle cell lesion that entraps native kidney. Characteristic histologic features of MST include alternating cellularity that imparts a nodular low-power appearance, onion-skin cuffing around entrapped renal tubules, heterologous differentiation (glia or cartilage), and vascular alterations (angiodysplasia of entrapped arterioles, juxtaglomerular cell hyperplasia in entrapped glomeruli). Three tumors in which the vascular alterations were particularly florid were associated with extrarenal vasculopathy and attendant morbidity. A majority of cases stained for CD34, although the degree of staining was variable. Most patients were treated with surgical excision alone, and none experienced recurrence or metastasis. Recognition of this entity can spare a child potentially toxic adjuvant chemotherapy that might be used for lesions in its differential diagnosis, specifically clear cell sarcoma of the kidney.

Nephrogenic adenofibroma. A novel kidney tumor of young people.

Nephrogenic adenofibroma is a novel kidney tumor of young people (mean age of presentation, 13 years), who present with polycythemia, hypertension, or hematuria, which resolve following nephrectomy. The typical nephrectomy specimen contains a solitary, nonencapsulated, vaguely circumscribed, irregularly shaped or spherical, firm mass with either tan, gray-white, or pale yellow coloration. Cysts are sometimes present within the tumor. The histologic appearance is distinctive and characterized by a marked proliferation of spindled mesenchymal cells resembling the classical type of congenital mesoblastic nephroma, encasing discrete nodules of embryonal epithelium similar to the hyperplastic nephrogenic rests (nephroblastomatosis) usually associated with Wilms' tumor. The mesenchymal component consists of a fascicular proliferation of tightly interlaced, uniform, benign-appearing spindled cells that immunostatin for vimentin and fibronectin, but not desmin or actin. The epithelial component consists of discrete islands of blastemal cells that are partially or fully differentiated toward tubular, tubulopapillary, or papillary structures. Psammoma bodies are plentiful. Embryonal epithelium immunostains for cytokeratin but not epithelial membrane antigen. The overall histologic appearance of the mesenchymal and epithelial components is benign, and preliminary clinical data suggest that the tumor has a benevolent course. Two cases, however, contained small, well-circumscribed papillary lesions near the renal pelvis that resembled low-grade collecting duct carcinoma. The clinical implications of the latter finding are unclear.

The clinicopathologic spectrum of putative extrarenal rhabdoid tumors. An analysis of 42 cases studied with immunohistochemistry or electron microscopy.

The existence of extrarenal rhabdoid tumor (ERRT) as a discrete pathologic entity has been controversial despite frequent reports of its occurrence. We performed immunohistochemistry, electron microscopy, or both on 42 cases with this diagnosis sent in consultation to us. Only 12 of the 42 neoplasms had the histological findings of "classic" malignant rhabdoid tumor of the kidney; the remainder displayed a variety of neural, epithelial, myoid, mesenchymal, or ependymal patterns. Electron microscopy also showed that most possessed neural, epithelial, or ependymal features. Immunohistochemistry generally revealed marked polyphenotypia, with immunoreactivity to a wide array of antibodies against neural, epithelial, glial, and myogenic markers. A specific tissue-based diagnostic category could not be assigned in only 11 of the 42 cases, seven of which lacked material for a comprehensive ultrastructural or immunohistochemical study. We conclude that tumors currently diagnosed as ERRT represent a heterogeneous group of neoplasms that may form unique subsets of known entities within the specific site where they arise or that may defy classification into a specific alternative category. Our findings lead us to believe that the term ERRT is not valid as representing a specific diagnostic entity and to prefer the term "poorly differentiated neoplasm with rhabdoid features" for undifferentiated tumors.

Rhabdoid tumor of kidney. A report of 111 cases from the National Wilms' Tumor Study Pathology Center.

We review 111 cases of rhabdoid tumor of kidney (RTK), including 79 entered on the National Wilms' Tumor Study (NWTS). Median age at diagnosis was 11 months, with a range from 0 to 106 months. The male:female ratio was 1.5:1. Gross features included a characteristic involvement of perihilar renal parenchyma. A wide histological spectrum was encountered, including nine major morphological patterns (classical, epithelioid, sclerosing, lymphomatoid, histiocytoid, etc.). These appearances invite confusion with other renal neoplasms. Ultrastructural studies were performed in 20 cases; immunocytochemical studies were performed in 11. Vimentin was demonstrated in all tumors; epithelial membrane antigen was seen in 7. Nonspecific decoration of cytoplasmic inclusions by a variety of immunostains was found in several cases. Several findings suggested that RTK might arise from primitive cells involved in formation of the renal medulla. There was no evidence of a histogenetic relationship to Wilms' tumor, although RTK may overlap with mesoblastic nephroma and clear cell sarcoma. Of the 70 NWTS patients with adequate follow-up, 56 (80%) have died. Every patient presenting with distant metastases died, whereas 10 of 20 with negative nodes survived. Survival rates were higher for girls (56.3% versus 11.1%). None of the histological variables had independent prognostic significance.