Evidence Of Ureteral Obstruction Research Articles

Feline ureteral obstruction: a case-control study of risk factors (2016-2019).

Ureteral obstruction (UO) in cats causes acute kidney injury and typically requires surgical intervention. Information is required about potentially modifiable risk factors to inform prevention strategies. A case-control study was performed to assess risk factors associated with feline UO. Cases were defined as cats with either of the following: (1) ureteral obstruction (ureteroliths: 13/18; unknown: 5/18) confirmed with pyelography; or (2) a creatinine concentration >140 µmol/l with both UO (ureteroliths: 6/10; blood clots: 3/10; pyonephrosis: 1/10) and pyelectasia ⩾5 mm on abdominal ultra sonography. Controls were defined as cats without evidence of UO on history, physical examination and abdominal ultrasound. Age, sex, breed (domestic shorthair/longhair), diet (predominantly dry, mixed or predominantly moist food), housing (indoors or mixed) and plasma total calcium were evaluated for their association with UO using multivariable logistic regression. A receiver operator characteristic (ROC) curve was created to evaluate the predictive ability of the final model. In total, 168 cats (28 cases, 140 controls) were included. Age, sex, breed, housing and total calcium were not significantly associated with UO; however, diet was. Compared with cats eating a predominantly moist food diet, cats fed a predominantly dry food diet were 15.9 times more likely to develop a UO (95% confidence interval 2.9-295; P = 0.009). There was no difference in the association between diet and UO in cats fed a mixed diet vs cats fed a predominantly moist food diet (P = 0.25). The area under the ROC curve was 72%. Changes in diet formulation could provide a simple and economical method to reduce the risk of UO.

Sonography First for Acute Flank Pain?

s part of the Sound Judgment Series, this article describes the case of a patient presenting with acute flank pain and a classic clinical picture of a first episode of renal colic. The epidemiologic characteristics, pathophysiologic characteristics, risk factors, and initial assessment of patients with renal colic are briefly discussed. Imaging options are discussed, with emphasis on sonography and computed tomography (CT). Although CT is typically a first-line test in the United States and is very accurate, there is increasing awareness of the radiation risk associated with CT scanning. Sonography may directly visualize kidney stones and/or evidence of ureteral obstruction and may obviate the need for CT scanning. Sonography is typically the first-line test in Europe, even in a first episode of kidney stones. We submit that sonography as an initial imaging modality in suspected kidney stones should be considered more often, particularly in younger and female patients with classic symptoms on first presentation and in patients with symptoms consistent with their prior episodes of renal colic, reserving CT for patients in whom symptoms do not resolve or there is a suspicion of alternative diagnoses. Decisions about imaging may offer an opportunity for shared decision making about what initial imaging modality to use. Clinical Case

TERMINAL LOOP CUTANEOUS URETEROSTOMY IN RENAL TRANSPLANTATION: AN UNDER UTILIZED URINARY DIVERSION TECHNIQUE

We evaluated the effectiveness of terminal loop cutaneous ureterostomy as a means of urinary drainage in kidney transplant recipients during a 20-year period. Five cadaveric and 2 living related patients underwent kidney transplantation with terminal loop cutaneous ureterostomy between 1984 and 2004. These patients had no usable bladder or they were not suitable candidates for intermittent catheterization. Followup was 20 months to 17 years. One patient underwent stomal revision 5 months after renal transplantation. Current serum creatinine 4 years later was 166 mumol/l. The remaining 6 patients had no evidence of ureteral obstruction and rarely had bacteriuria or urinary tract infections. Four patients had a functioning allograft with normal serum creatinine. One patient died with a normally functioning allograft and the remaining patient lost his graft due to chronic rejection. No patient in this series lost the graft due to a urological cause. Overall outcomes included excellent allograft function with minimal infection or stomal stenotic complications. Terminal loop cutaneous ureterostomy is a simple, safe and alternative means of urinary diversion in patients with renal transplant and a defunctionalized lower urinary tract.

Analysis of retroperitoneal tumors operated on in a gynecology clinic

To evaluate the patients with retroperitoneal tumors that presented as pelvic mass and were operated on in our gynecology clinic. Records of 16 patients were analyzed retrospectively. All patients had a palpable mass on physical examination and abdominal pain was the most frequent symptom. The median diameter of the tumors was 15 cm. The postoperative histopathology of the patients was leiomyosarcoma in nine, retroperitoneal fibrosis in three, and paraganglioma, hamartoma, and dermoid cyst in three patients. All the tumors were excised almost completely. The incorrect preoperative diagnosis of the patients in this study is due to the presentation of the patients with non-specific symptoms and not considering the less common differential diagnosis of a pelvic mass. For clinical signs of vascular or rectal displacement, or where there is evidence of ureteric obstruction, retroperitoneal pathology should be kept in mind.

Phase III randomized pilot study comparing interferon α- 2b in combination with radiation therapy versus radiation therapy alone in patients with stage III-B carcinoma of the cervix

This randomized pilot study was designed to determine whether the addition of interferon α-2b to standard radiation therapy offered an advantage in loco-regional control and survival over radiation therapy alone in a homogeneous group of patients with stage IIIB carcinoma of the cervix. Thirty-six patients were treated with a combination of interferon α-2b plus radiation therapy, and 38 patients were treated with radiation therapy alone. Patients with evidence of ureteral obstruction were excluded from the study. Evaluation of loco-regional response was determined by pelvic examination, cervical cytology, biopsies and CT scans when indicated. Survival time was measured from initiation of treatment to date of death or last follow-up. Patient characteristics were comparable between both study arms. The objective complete response rate was 67% in the combined therapy group and 55% in the radiation alone group (P = 0.454). With a median follow-up of 17 months for all patients and 31 months for live patients, 50% of the combined group survived vs. 39.5% of the radiation alone group (P = 0.424). We conclude that the addition of interferon α-2b to standard radiation therapy did not significantly improve loco-regional response or survival, although such a trend was noted. We encourage the design of a larger randomized study with sufficient power to detect meaningful differences to prove whether the tendency observed in the present investigation holds any promise to improve the outcome of these patients.

The role of renal scintigraphy and unenhanced helical computerized tomography in patients with ureterolithiasis.

Unenhanced helical computerized tomography (UHCT) has recently evolved as an accurate imaging modality for determination of the presence or absence of ureterolithiasis in patients with acute flank pain. Functional renal scintigraphy is considered the gold standard for urinary tract obstruction. The objective of this study was to correlate the secondary signs of urinary obstruction on UHCT with findings of functional renal scintigraphy. UHCT was performed in 30 patients admitted to the emergency room with acute flank pain. All patients had a calcified urinary stone identified on UHCT. The location of each urinary stone was classified as ureteral or in the ureterovesical junction. The presence of secondary CT signs of ureteral obstruction was determined for each patient. After oral or intravenous hydration, a technetium-99m diethylene triamine penta-acetic acid renal scan was performed in all patients within 12 h of the CT scan. Follow-up delayed scintigraphic images were obtained at 2 h and 24 h in patients with evidence of ureteral obstruction. The sensitivity, specificity and predictive values of each possible combination of CT findings were determined by comparison with the scintigraphic results. The distal ureter was the most common location for a calculus on UHCT, followed in frequency by the ureterovesical junction, proximal ureter and mid-ureter. The renograms showed high-grade, unilateral obstruction in 12 patients, indeterminate scans in five patients and normal renograms in 13 patients. The sensitivities and specificities of individual CT findings ranged from 50% to 75% and from 8% to 69%, respectively. Perinephric stranding gave the highest positive predictive value (PPV) for obstruction (69% including indeterminate renograms). None of the individual CT findings showed a statistically significant correlation with scintigraphic findings. A combination of one or two positive CT findings had a PPV of only 25% for obstruction. A combination of three or four positive CT findings gave a PPV of 70% for obstruction. Our preliminary study shows that secondary CT signs of ureterolithiasis correlate poorly with the scintigraphic findings and that they do not permit evaluation of the functional status of obstructed kidneys. Even a combination of the most frequent CT findings has a low predictive value, i.e. does not allow a decision to be made as to the most suitable treatment. Therefore, renal scintigraphy should be performed in conjunction with UHCT in all patients with ureteral calculi.

Strategies for managing upper tract calculi in young children.

Pediatric urolithiasis is relatively uncommon and there is little information on the application of modern surgical procedures in young children. We present a single center experience with the surgical management of upper tract calculi in this age group. We reviewed presentation, co-morbidity, treatment, outcome and complications in all prepubertal patients who required surgical treatment for ureteral or renal calculi during a 4-year period. The series consists of 24 girls and 17 boys 17 months to 14 years old (mean age 7.5 years). A total of 26 children were anatomically normal, and 4 had myelomeningocele, 4 had ureteropelvic junction obstruction (in a pelvic kidney in 1), 2 had cloacal anomalies, 2 had vesicoureteral reflux, and 1 each had nonrefluxing megaureter, orthotopic ureterocele and a functioning renal transplant. Extracorporeal shock wave lithotripsy was performed in 24 patients. Stents or nephrostomy tubes were only used in the 4 patients who presented with pyonephrosis. Of the 41 cases 17 were rendered stone-free, 3 had a decreased stone burden and 4 were failures. Ureteroscopic extraction of distal ureteral calculi was successful in 11 of 12 children, of whom the youngest was 2.5 years old. No child had postoperative infection or evidence of ureteral obstruction. Stent placement facilitated stone passage or dissolution in 2 patients, a renal calculus was percutaneously extracted in 2 and 7 required open surgery, mostly for correcting simultaneous anatomical abnormalities or after minimally invasive surgery failed. Some metabolic abnormality was detected in 80% of the children tested. The surgical management of upper urinary tract calculi in young children parallels that in adults. Minimally invasive surgical methods may be safely used even in young infants. Most children do not need elective stenting before lithotripsy. Open procedures are still required in 17% of cases. The majority of children have definable metabolic abnormalities.

STRATEGIES FOR MANAGING UPPER TRACT CALCULI IN YOUNG CHILDREN

Purpose Pediatric urolithiasis is relatively uncommon and there is little information on the application of modern surgical procedures in young children. We present a single center experience with the surgical management of upper tract calculi in this age group. Materials and Methods We reviewed presentation, co-morbidity, treatment, outcome and complications in all prepubertal patients who required surgical treatment for ureteral or renal calculi during a 4-year period. The series consists of 24 girls and 17 boys 17 months to 14 years old (mean age 7.5 years). A total of 26 children were anatomically normal, and 4 had myelomeningocele, 4 had ureteropelvic junction obstruction (in a pelvic kidney in 1), 2 had cloacal anomalies, 2 had vesicoureteral reflux, and 1 each had nonrefluxing megaureter, orthotopic ureterocele and a functioning renal transplant. Results Extracorporeal shock wave lithotripsy was performed in 24 patients. Stents or nephrostomy tubes were only used in the 4 patients who presented with pyonephrosis. Of the 41 cases 17 were rendered stone-free, 3 had a decreased stone burden and 4 were failures. Ureteroscopic extraction of distal ureteral calculi was successful in 11 of 12 children, of whom the youngest was 2.5 years old. No child had postoperative infection or evidence of ureteral obstruction. Stent placement facilitated stone passage or dissolution in 2 patients, a renal calculus was percutaneously extracted in 2 and 7 required open surgery, mostly for correcting simultaneous anatomical abnormalities or after minimally invasive surgery failed. Some metabolic abnormality was detected in 80% of the children tested. Conclusions The surgical management of upper urinary tract calculi in young children parallels that in adults. Minimally invasive surgical methods may be safely used even in young infants. Most children do not need elective stenting before lithotripsy. Open procedures are still required in 17% of cases. The majority of children have definable metabolic abnormalities.

Unilateral Tubular Obstruction: An Unusual Complication of Urographic Contrast Media

OBJECTIVE. Our purpose was to report our experience with four patients with renal colic who had a rare unilateral complication from urographic contrast media, with a dense and striated nephrogram affecting only the healthy kidney. All cases were reversible. CONCLUSION. An unusual complication of urography with a characteristic radiologic appearance is presented. This complication was transient and reversible in our patients. Unilateral tubular blockage induced by a contrast agent must be considered when the contralateral kidney shows evidence of ureteral obstruction. Editorial Comment: The authors describe 4 patients who had a dense, unilateral, striated nephrogram affecting a normal healthy kidney with contralateral ureteral obstruction. This complication was reversible and transient. They attributed this appearance to protein or uric acid precipitation in the renal tubules. Thus, among the patterns of unilateral striated nephrogram that has been reported, this new entity needs to be considered. Other entities that present such a pattern and that may be unilateral or bilateral include obstructed uropathy, acute pyelonephritis, acute thrombosis of the renal vein and so forth. In 1 patient proteinuria was observed, while the other 3 had hyperuricemia with urates in the urinary sediment after excretory urography. I have not seen this phenomenon but it is important to be aware of and include in the differential diagnosis.

Emphysematous pyelonephritis in diabetic patients

To determine the clinico-pathological profile of patients with emphysematous pyelonephritis (EPN). The records of 22 diabetic patients who presented with EPN were reviewed. EPN occurred predominantly in female diabetic patients without evidence of ureteric obstruction. The ages of the patients and the duration of the diabetes were variable. EPN occurred in insulin-dependent as well as non-insulin-dependent patients. The patients presented following a prodromal illness of urinary sepsis, with an acute severe illness with symptoms and signs on the affected side. Dehydration and ketoacidosis were common. The diagnosis was made by recognizing gas in the kidney on an X-ray or ultrasound. Eighteen patients were subjected to emergency nephrectomy. EPN is an uncommon, life-threatening condition characterized by the production of intraparenchymal gas. We believe that vigorous resuscitation and emergency nephrectomy is the treatment of choice.

The significance of ureteral obstruction in invasive transitional cell carcinoma of the urinary bladder.

Ureteral obstruction is a frequent complication of transitional cell carcinoma (TCC), known as a poor prognostic sign and indicative of advanced disease. We investigated retrospectively the medical records of 122 consecutive patients who suffered from invasive TCC of the urinary bladder during a 6-year period. Unilateral or bilateral ureteral obstruction was found in 66 patients (54.1%). High stage (T3-T4) and grade (III-IV) tumors were correlated with ureteral obstruction in 89.4% and 83.3% respectively compared to 67.9% and 66.1%, respectively, among patients with normal upper tracts (P < 0.001); 10.6% of the patients with ureteral obstruction had low stage disease, and all of them proved to have involvement of the ureteral orifices on the affected side. The 5-year survival rate of patients with and without ureteral obstruction was 32.2% and 65.9%, respectively (P < 0.001). The presence of ureteral obstruction, particularly in the absence of intravesical involvement of the ureteral orifices, signified a high stage, muscle invasive, and often metastatic tumor in more than 90% of the patients. Ureteral obstruction was an accurate criterion for poor prognostic and was associated with significantly lower, overall and stage-specific, survival rates, despite of radical surgery. We conclude that evidence of ureteral obstruction is an important staging standard and significant prognostic indicator in transitional cell carcinoma of the urinary bladder.

Primary amyloidosis of ureter

Primary nodular deposits of amyloid have been described as isolated findings in a number of organs including bladder, lung, skin, larynx, salivary glands, breast and ureter. Within the urinary tract the bladder is most frequently involved. A 56 year old woman presented with clinical and radiological evidence of ureteric obstruction due to a mass in the mid-region of the left ureter. She was treated by left nephroureterectomy which on histological examination showed an intramural mass of amyloid material which stained positively with congo red and showed apple green birefringence. This unusual lesion has been identified in 26 previous cases. Amyloid tumour of the ureter occurs predominantly in women (21F:6M). The reported cases have occurred between 12 and 81 years of age with 52% of cases occurring in the 6th and 7th decade. Solitary amyloid tumour occurs more frequently in the left ureter than right (18L:9R, 67%). In the majority of cases the patient has presented in similar fashion and been treated with nephroureterectomy. In three cases intraoperative frozen section of the lesion prevented unnecessary radical resection. It is strongly suggested that frozen section be recommended to urological surgeons when dealing with primary lesions of the ureter.

Endoscopic Correction of Primary Vesicoureteric Reflux: Results in 94 Ureters

Sixty one patients with primary vesicoureteric reflux were treated by endoscopie injection of Mentor Polytef paste. Seventy six of the 94 treated ureters showed complete disappearance of vesicoureteric reflux after one injection. Eight re fluxing ureters required a second injection, one required a third, and one a fourth. Six ureters showed improvement in the grade of reflux after the first injection. Two ureters did not show any change in the grade of reflux after endoscopie injection. The amount of paste injected to correct reflux in the 94 ureters varied from 0-1 to 10 ml (mean 0-33 ml). Patients were followed up for three to 25 months (mean 13*4 months). Follow up micturating cystograms in the 94 ureters showed absence of reflux in 83. Eleven ureters showed recur? rence of reflux. Recurrence of reflux was attributed to early technical difficulties and to insufficient amounts of polytef paste used in some cases. Follow up urograms did not show evidence of ureteric obstruction in the treated ureters. Endoscopie injection of polytef paste is a safe, simple, and effective procedure for primary vesicoureteric reflux and averts the need for open operation.

Predictive value of excretory urography, ultrasonography, computerized tomography, and liver and bone scan in the staging of bilharzial bladder cancer in Saudi Arabia.

The role of ultrasonography, computed tomography (CT), and radioisotopic scanning in the staging of bilharzial bladder cancer has not been reported previously. Forty patients with invasive bladder cancer seen at the King Faisal Specialist Hospital and Research Centre between January 1978 and June 1981 underwent complete preoperative workup for staging of their tumors prior to radical cystectomy. The preoperative radiologic investigations included excretory urography (IVP), ultrasonography (US), CT of the pelvis, and liver and bone scans. The results of these investigations were compared with the operative and pathologic staging. Ninety-three percent of the patients with bilharzial cancer had evidence of ureteric obstruction on IVP compared with 22% of the nonbilharzial cancer patients. The presence of ureteric obstruction in these patients did not correlate with the stage of the disease with 83% of the patients with superficial tumors (T1 and T2) having hydroureteronephrosis. Ultrasonography and CT had an 83% accuracy in the staging of superficial tumors. Stage T3 tumors were understaged in 14% of the cases. Ultrasonography did not differentiate Stages T3 and T4 tumors while CT scan differentiated these two stages in 57% of the cases. Bone scan failed to reveal evidence of metastatic disease in any of the bilharzial cancer patients. Liver scan was suspicious for liver metastases in two patients with bilharzial cancers in whom open liver biopsy revealed only hepatic bilharziasis. Of all the radiographic studies, US or preferably CT scan seem to be of some value in the staging of bilharzial tumors localized to the bladder. Bone and liver scans are probably of no cost effective benefit.

Growth of rats' kidneys after unilateral uretero‐caval anastomosis.

1. The relative significance of reduced excretion of urinary constituents and reduced renal mass, as stimuli to growth of one kidney after the other has been removed, has been investigated. 2. To abolish the excretory function of one kidney without removing it, the right ureter was drained into the vena cava through a compound cannula for 6 weeks. Uretero-caval anastomoses were performed in twenty-four male rats at 10 weeks of age: six survived without evidence of ureteric obstruction (and a further five with minimal obstruction). 3. The rats with anastomoses grew less than six other rats from which the right kidney had been removed or six which had been submitted to a sham operation, and their plasma urea and creatinine concentrations were higher. 4. Relative to body weight, the dry weight of each kidney after uretero-caval anastomosis without obstruction was 18% greater than after sham operation; taking both kidneys together, the total increase was almost as much as in the left kidney alone after right nephrectomy (46%). 5. Histologically and in terms of DNA concentration, the growth of both kidneys after uretero-caval anastomosis was of the same kind as in the left kidney after right nephrectomy. 6. The return of urine from one kidney into the blood provided a powerful stimulus to renal growth in spite of the restraining effect of increased renal mass.

Conservative management of renal tuberculosis.

The conservative management of 60 patients suffering from renal tuberculosis is reported. Following chemotherapy, bacteriological conversion to negative was obtained in all patients. 2 patients (3-3%) relapsed due to failure to take adequate treatment. Evidence of ureteric obstruction was obtained in 30 patients (50%)--23 at the time of diagnosis, 6 during treatment and 1 at the time of relapse. Following corticosteroid treatment relief of obstruction was obtained in 72%. Nephrectomy was carried out in 4 patients (6-7%) but in only 1 of these was the operation carried out for reasons which we now consider valid.

Emphysematous Pyelonephritis: Report of 3 Cases Treated by Nephrectomy

Three cases of emphysematous pyelonephritis are presented. All 3 patients were diabetics, they all had Escherichia coli and there was evidence of ureteral obstruction in 1 case only. All 3 patients underwent nephrectomy and survived. Initial evaluation should be performed to rule out ureteral obstruction. Treatment should consist of a brief trial of high dose antibiotic therapy with serious consideration of nephrectomy in the face of persistent symptoms or gas. Incision and drainage are reserved for poor surgical risk patients. All patients should remain on antibacterial therapy and have frequent followup examinations.

Obstructive uropathy secondary to a false aneurysm in 12 years after abdominal aneurysmectomy.

Long-tern follow-up of patients having abdominal aneurysmectomy and graft replacement will show that a considerable number develop false aneurysms at the anastomotic sites. This is the particularly true in situations in which silk suture was used for the vascular anastomoses. Successful management of this problem necessitates an awareness that unsuspected, asymptomatic ureteral obstruction can be a complicating factor. Preoperative work-up of all abdominal aortic problems should include an intravenous pyelogram. When evidence of ureteral obstruction is encountered, retrograde ureteral catheterization is important to precisely identify the point of obstruction and to aid in the ureterolysis at the time of operation. Inadvertent injury and contamination from an obstructed ureter is a situation which must be avoided when prosthetic grafting is anticipated.

Unusual case of juvenile G.P.I.

ureter was apparently normal ; the pelvis and calices alone were dilated ; the uretero-pelvic junction was well marked ; and yet no cause of ureteric obstruction was found. Jewett (1940), after careful study, showed that normally there is no absolute uretero-pelvic junction, the pelvis blend ing into the ureter in a funnel-shaped prolongation. He believed that when such a junction exists, as it did in the cases described here, stenosis should be considered, particu larly if there is pyelectasis. The fact that he considered this stenosis often congenital, since there were no acquired factors to explain it, would make such a mechanism an exceedingly attractive explanation for the present series of familial cases. However, no evidence of stenosis was found. Faute de mieux* we must therefore fall back on such a con ception as inadequacy of the ureter as an efficient urinary duct which has been posited as a causal factor in hydro nephrosis ; or we must postulate a functional defect in the neuromuscular mechanism of the renal pelvis, as the familial and inherited primary aetiological factor, though there is no positive evidence for either supposition. Consideration of the familial nature of these cases does, however, serve to show that, contrary to general opinion, a pelvic and caliceal hydronephrosis of this type can occur as a congenital condi tion without evidence of ureteric obstruction.

The lymphatic spread of carcinoma of the cervix and of the body of the uterus: A study of 420 necropsies

1.1. The necropsy study of 420 cases of carcinoma of the cervix (356 cases) and the corpus uteri (64 cases) is presented with the incidence of lymph node involvement, distant metastases, and the assigned cause of death.2.2. Careful dissection and multiple sections of the lymph nodes in twenty-six nontreated and fifteen treated cases of cervical carcinoma emphasize the frequency of early metastasis, the probable impedimentary effect of irradiation, and traces the routes of lymphatic spread. The parametrium was involved in 77 per cent of the nontreated cases and 33 per cent of the treated cases. Distant metastasis occurred in 27 per cent of the nontreated and 53 per cent of the treated cases.3.3. In the series of 356 nontreated and treated cases of cervical carcinoma, the probable arresting effect of irradiation is emphasized by the difference in the node group involvement. The incidence of distant metastases varies little in the two groups.4.4. The lymphatic spread, from cervical carcinoma, follows a constant course, the parametrium, the nodes of the primary group, and finally the secondary group nodes, before extension beyond the pelvis.5.5. Uremia, the end result of ureteral compression, was the assigned cause of death in 58.5 per cent of nontreated cases and 49.3 per cent of the treated cases. Evidence of ureteral obstruction and hydronephrosis was found in 82.8 per cent of the nontreated and 78.6 per cent of the treated cases.6.6. The routes of lymphatic spread of endometrial carcinoma are less constant. The three main channels permit involvement of the same node groups as found in cervical carcinoma, or a by-passing of the primary groups with metastases directly to the inguinal nodes or the aortic nodes.7.7. Uremia and/or pyelonephritis was the most frequent assigned cause of death in cases with endometrial carcinoma.8.8. The disintegrating effects of irradiation does not occur in all of the treated cases, and similar changes, though usually of a lesser degree, are present in the nontreated cases. Histopathological changes are more frequent when there is a history of infection.9.9. Histopathological evidence is lacking to support the claim that metastatic nodes may be sterilized by irradiation. The necropsy findings suggest the possibility of this effect.10.10. The numerous intercommunicating lymphatic vessels account for the unpredictable sites of metastases.