Relative to Role as Provider of Care in Oncologic Emergencies/Nursing Management II
Superior Vena Cava Syndrome
One of two great veins, known as vena cavae, in the general or systemic circulatory system of the body, the other being the inferior vena cava.
The superior vena cava starts at the top of the chest, behind the lower edge of the right first rib and close to the sternum (breastbone).
It travels some
The superior vena cava is formed by the junction of the right and left brachiocephalic (innominate) veins, which themselves are formed from the union of the subclavian veins (draining venous blood from the arms), the jugular veins (draining blood from the head and neck), and several minor veins.
The superior vena cava also receives blood from the azygos vein, which drains much of the chest.
The superior vena cava is a vein in the body that carries deoxygenated blood to the right atrium of the heart, and it is also commonly referred to as the precava.
This vein is short in length; however, it is very wide in diameter.
The right and left brachiocephalic veins form this vein. All of these veins receive blood from the upper body, such as the arms, neck, and the head. Before going into the right atrium of the heart, the precava will join with the azygous vein, and the azygous vein receives its blood supply from the rib cage. These veins are bringing the blood into the heart so that it can pulsate and pump.
The superior vena cava does not have any valves separating it from the right atrium of the heart, and this allows the blood to flow into the right atrium and ventricle of the heart to cause it to contract. This forces blood to flow up through the internal jugular vein and will result in jugular venous pressure.
Superior Vena Cava Syndrome
Superior vena cava syndrome (SVCS) is a common complication of malignancy, especially of lung cancer and lymphoma.
The frequency of SVCS varies depending on the specific malignancy. Approximately 2% to 4% of all patients with lung cancer develop SVCS at some time during their disease course.
The incidence is higher in small cell lung cancer (SCLC), given its predilection for mediastinal involvement and rapid growth; the incidence approaches 10%.5,7,8 SVCS develops in approximately 2% to 4% of non-Hodgkin’s lymphoma (NHL) but is relatively rare in Hodgkin’s lymphoma despite the presence of mediastinal lymphadenopathy.
For primary mediastinal large B-cell lymphomas with sclerosis, the incidence has been reported as high as 57% in one series of 30 patients.10 Together, lung cancer and lymphoma are responsible for over 90% of malignant causes of SVCS.
In the modern era, 60% to 90% of cases of SVCS are caused by malignant tumors, with the remaining cases accounted for largely by fibrosing mediastinitis and thrombosis of indwelling central venous devices and/or pacemaker leads.
Superior vena cava syndrome (SVCS) is the clinical expression of obstruction of blood flow through the superior vena cava (SVC).
Characteristic symptoms and signs may develop quickly or gradually when this thin-walled vessel is compressed, invaded, or thrombosed by processes in the superior mediastinum. The first pathologic description of SVC obstruction, in a patient with syphilitic aortic aneurysm, appeared in 1757.
In 1954, Schechter reviewed 274 well-documented cases of SVCS reported in the literature; 40% of them were due to syphilitic aneurysms or tuberculous mediastinitis. These entities have since virtually disappeared, and lung cancer is now the underlying process in approximately 70% of the patients with SVCS.
ANATOMY AND PATHOPHYSIOLOGY
The SVC is the major drainage vessel for venous blood from the head, neck, upper extremities, and upper thorax. It is located in the middle mediastinum and is surrounded by relatively rigid structures such as the sternum, trachea, right bronchus, aorta, pulmonary artery, and the perihilar and paratracheal lymph nodes. The SVC extends from the junction of the right and left innominate veins to the right atrium, a distance of 6 to
When the SVC is fully or partially obstructed, extensive venous collateral circulation may develop. The azygos venous system is the most important alternative pathway. Carlson found that dogs could not survive sudden ligation of the SVC below the level of the azygos vein, but they tolerated well ligation of the SVC above it.
He could, however, successfully obstruct the SVC and the azygos vein in operations performed in two stages, presumably because time allowed collaterals to form. Other collateral systems are the internal mammary veins, lateral thoracic veins, paraspinous veins, and the esophageal venous network. The subcutaneous veins are important pathways, and their engorgement in the neck and thorax is a typical physical finding in SVCS. Despite these collateral pathways, venous pressure is almost always elevated in the upper compartment if there is obstruction of the SVC. Venous pressures as high as 200 to
ETIOLOGY AND NATURAL HISTORY
SVCS usually has an insidious onset and progresses to typical symptoms and signs. Review of the data from three series shows dyspnea to be the most common symptom.
Dyspnea occurred in 63% of the patients with SVCS. A sensation of fullness in the head and facial swelling were reported by 50% of the patients.
Other complaints were cough (24%), arm swelling (18%), chest pain (15%), and dysphagia (9%). The characteristic physical findings were venous distention of the neck (66%) and chest wall (54%),
Patient who presented with progressively enlarging veins over the anterior chest wall. A diagnosis of a right-sided superior sulcus (Pancoast) tumor compressing the SVC was made.
facial edema (46%), plethora (19%), and cyanosis (19%).
Fig. Superior vena cava syndrome in a person with bronchogenic carcinoma. Note the swelling of his face first thing in the morning (left) and its resolution after being upright all day (right).
These symptoms and signs may be aggravated by bending forward, stooping, or lying down.
Malignant disease is the most common cause of SVCS. The percentage of patients in different series with a confirmed diagnosis of malignancy varies from 78% to 86%. Lung cancer was diagnosed in 65% of 415 patients analyzed in these series. Armstrong and colleagues did a retrospective review of 4100 persons treated for bronchogenic carcinoma between 1965 and 1984, and identified 99 patients (2.4%) with SVCS. Salsali and Cliffton observed SVCS in 4.2% of 4960 patients with lung cancer; 80% of the tumors inducing SVCS were of the right lung. Small cell lung cancer is the most common histologic subtype, and it was found in 38% of the patients who had lung cancer and SVCS. In five large series of small cell cancer, 9% to 19% of patients demonstrated SVCS.
The second most common histologic subtype, squamous cell carcinoma, is found in 26% of lung cancer patients with SVCS.
Lymphoma involving the mediastinum was the cause of SVCS in 8% of the patients reported in the series. Armstrong and Perez found SVCS in 1.9% of 952 lymphoma patients. Perez-Soler and coworkers identified 36 cases (4%) of SVCS among 915 patients with non-Hodgkin’s lymphoma treated at the M.D.
Twenty-three patients (64%) had diffuse large cell lymphoma, 12 (33%) had lymphoblastic lymphoma, and one patient had follicular large cell lymphoma. Of their patients with diffuse large cell lymphoma and lymphoblastic lymphoma, 7% and 21% had SVCS, respectively. In a series of patients with primary mediastinal B-cell lymphoma with sclerosis, SVCS was present in 57% of patients. Hodgkin’s lymphoma commonly involves the mediastinum, but it rarely causes SVCS. Other primary mediastinal malignancies that cause SVCS are thymoma and germ cell tumors. Breast cancer is the most common metastatic disease that causes SVCS. In one report, breast cancer was the cause of SVCS in 11% of cases.
Nonmalignant conditions causing SVCS are not as rare as was previously reported. When the data were collected from general hospitals, as many as 22% of the patients had SVCS of noncancerous causes.
Parish and coworkers reported on 19 patients with benign causes of SVCS, and Schraufnagel included 16 such patients in his series.
Fifty percent of the patients in both reports had a diagnosis of mediastinal fibrosis, which was probably due to histoplasmosis. Parish and associates reported on six patients with thrombosis of SVC, and in five the thrombosis developed in the presence of central vein catheters or pacemakers. [ref: 9] Sculier and Feld reviewed 24 cases of central venous catheter-induced SVC. Of these, 18 were caused by pacemaker catheters. LeVeen shunts, Swan-Ganz catheters, and hyperalimentation catheters were also involved. The increasing use of these devices for the delivery of chemotherapy agents or for hyperalimentation contributes to the development of SVCS in cancer patients. Obstruction of SVC in the pediatric age group is rare and has a different etiologic spectrum. The causative factors are mainly iatrogenic, secondary to cardiovascular surgery for congenital heart disease, ventriculoatrial shunt for hydrocephalus, and SVC catheterization for parenteral nutrition.
In a report of 175 children with SVCS, 70% of cases were iatrogenic. Of the remaining 53 cases, 37 (70%) were caused by mediastinal tumors, 8 (15%) by benign granuloma, and 4 (7.5%) by congenital anomalies of the cardiovascular system. Two thirds of the tumors causing SVCS in childhood are lymphomas. Of 16 children with SVCS at presentation reported on from St. Jude Children’s Research Hospital, 8 were diagnosed with non-Hodgkin’s lymphoma, 4 had acute lymphoblastic leukemia (ALL), 2 had Hodgkin’s disease, 1 had neuroblastoma, and 1 had a yolk sac tumor.
Most children who developed SVCS late in the course of their malignancy had recurrent solid tumors. Issa and colleagues reported that mediastinal fibrosis secondary to histoplasmosis caused SVCS in 7 (5%) of the 150 patients reviewed
DIAGNOSTIC PROCEDURES
Superior vena cava syndrome has long been considered a potentially life-threatening medical emergency. It was common practice to immediately apply radiation therapy with initial high dose fractions, sometimes even before the histologic diagnosis of the primary lesion was established.
Diagnostic procedures such as bronchoscopy, mediastinoscopy, thoracotomy, or supraclavicular lymph node biopsy were often avoided because they were considered hazardous in the presence of SVCS.
The traditional therapeutic philosophy was recently challenged. The reported clinical experience was reassessed, and the safety and importance of diagnostic procedures were reevaluated.
Multidrug chemotherapy, sometimes combined with radiation therapy, is potentially curative for small cell carcinoma of the lung and non-Hodgkin’s lymphoma, even when presented as SVCS. The current practice of using different modalities for different primary causes of SVCS makes accurate histologic diagnosis of SVCS invaluable. Mediastinal irradiation before biopsy precludes proper interpretation of the specimen in almost half of patients.
Clinical identification of SVCS is simple because the symptoms and signs are typical and unmistakable. The chest film shows a mass in most affected patients. Only 16% of the patients studied by Parish and colleagues had normal chest films. The most common radiographic abnormalities are superior mediastinal widening and pleural effusion .
Computed tomography (CT) provides more detailed information about the SVC, its tributaries, and other critical structures such as the bronchi and the cord. The additional information is necessary because the involvement of these structures requires prompt action for relief of pressure. Moncada and associates outlined the advantages of combining CT with CT digital phlebography in SVCS: detailed resolution of the intrathoracic structures and musculoskeletal anatomy, accurate identification of the site and extent of obstructing thrombus in the SVC and of any external compression or invasion by mediastinal mass, contrast opacification of the venous trunks and collateral circulation sufficient to make confident surgical decisions and to determine late graft patency, accurate guidance for percutaneous biopsy techniques, guidance for radiation therapy to ensure that radiation ports fully encompass the disease, monitoring the effect of therapy.
The role of magnetic resonance imaging (MRI) has not been sufficiently investigated but appears promising, especially as this modality is totally noninvasive.
Contrast venography is controversial. It provides important information for determining if the vena cava is completely obstructed or remains patent and is extrinsically compressed. Dyet and Moghissi demonstrated by venography that 41% of patients with SVCS have patent SVCs that are displaced or involved — but not obstructed — by a tumor.
Another 19% have SVC obstruction below the azygos vein, for which collateral venous compression should be adequate. Venography is valuable if surgical bypass is considered for the obstructed vena cava. Lokich and Goodman stated that venograms are relatively contraindicated because the interruption of he integrity of the vessel wall, in the presence of increased intraluminal pressure, may result in excessive bleeding from the puncture site. However, there is no evidence of this complication. Although venography can confirm the clinical diagnosis and outline the anatomy, priority should be given to procedures that help to establish the histologic diagnosis. Radionuclide technetium-99m venography is an alternative, minimally invasive method of imaging the venous system. Although images obtained by this method are not as well defined as those achieved with contrast venography, they demonstrate patency and flow patterns. Collateral circulation can be evaluated in a general manner and quantified to some degree by radionuclide venography. Gallium single-photon emission CT may be of value in selected cases. In 58% of 107 patients reported by Schraufnagel the SVCS developed before the primary diagnosis was established.
Cytologic diagnosis is as accurate as tissue diagnosis for small cell carcinoma. Bronchoscopy supplies the malignant cells for cytologic evaluation in most cases of small cell disease.
In the presence of pleural effusion, thoracocentesis established the diagnosis of malignancy in 71% of patients. Biopsy of a supraclavicular node, especially if there was a suspicious palpatory finding, was rewarding in two thirds of the reported attempts. Small cell carcinoma of the lung and non-Hodgkin’s lymphoma often involve the bone marrow. A biopsy of the bone marrow may provide the diagnosis and stage for these lesions. Mediastinoscopy has a success rate for providing a diagnosis, but Painter and Karpf reported complications in five of nine procedures attempted. In 2 patients, the procedure had to be terminated before completion. In 3 patients, complications occurred after the mediastinoscopy, but these complications were managed successfully and the procedure was diagnostic in each case. Lewis and associates reported their experience with cervical mediastinoscopy in 15 patients with SVCS. All mediastinoscopies were diagnostic, and no complications were observed. A recent report by Jahangiri and coworkers of 34 patients with SVCS whose histologic diagnosis could not be established before mediastinoscopy confirmed the safety and high diagnostic yield of mediastinoscopy. A histologic diagnosis was obtained for all their patients, and only one manageable complication occurred during mediastinoscopy. Percutaneous transthoracic (CT-guided fine-needle biopsy) is emerging as an effective and safe alternative to open biopsy or mediastinoscopy. Successful diagnostic transluminal atherectomy has also been reported If all other procedures failed, thoracotomy is diagnostic.
Ahmann examined the traditional opinion that diagnostic procedures carry with them significant hazard, principally excessive bleeding.
He reviewed 843 invasive and semiinvasive diagnostic procedures and found only 10 reported complications, none of them fatal. Ahmann and others found minimal evidence to suggest that diagnostic procedures such as venography, thoracotomy, bronchoscopy, mediastinoscopy, and lymph node biopsy carry excessive risk for patients with SVCS.
In 163 patients treated for anterior mediastinal mass at Memorial Sloan-Kettering Cancer Center, 44 underwent general anesthesia. There were no deaths, and only 4 patients had prolonged intubation, demonstrating the low risk of modern anesthesia for patients with thoracic disease.
MANAGEMENT
The goals of SVCS management are to relieve symptoms and to attempt cure of the primary malignant process. Small cell carcinoma of the lung, non-Hodgkin’s lymphoma, and germ cell tumors constitute almost half of the malignant causes of SVCS. These disorders are potentially curable, even in the presence of SVCS.
The treatment of SVCS should be selected according to the histologic disorder and stage of the primary process. The prognosis of patients with SVCS strongly correlates with the prognosis for the underlying disease.
SMALL CELL LUNG CANCER
Combination chemotherapy, alone or in conjunction with radiation therapy, is considered standard treatment for small cell lung cancer.
Dombernowsky and Hansen reported the results of the treatment of 26 patients with small cell carcinoma of the lung who presented with SVCS. Of these 26 patients, 22 were initially treated with combination chemotherapy alone and, in all these patients, the resolution of the SVCS was prompt (median 7 days). Maddox and associates reported on 56 patients with small cell lung cancer who presented with SVCS.
Correction of SVCS was obtained in 64% (9 of 16) of patients treated with radiation alone, in 100% (23 of 23) of those given chemotherapy, and in 83% (5 of 6) of those who received combined therapy. The type of treatment did not substantially influence survival. Among 643 patients with small cell lung cancer, Sculier and colleagues identified 55 patients (8.5%) with SVCS.
Half of the patients developed the manifestations of SVCS before the histologic diagnosis was established. In the rest of the patients, the syndrome developed after the pathologic diagnosis of small cell lung cancer was made but before a specific treatment was started. Symptomatic relief of SVCS was achieved in 35 (73%) of 48 patients initially treated with chemotherapy and in 3 (43%) of 7 patients initially treated with radiation. Relief of SVCS occurred within 7 to 10 days after initiation of therapy. Fourteen patients had recurrent SVCS after initial treatment. Recurrent SVCS improved in 8 of 12 patients treated with radiation, one of two patients treated with chemotherapy, and three of four patients treated with combined modality. Spiro and associates analyzed 37 patients with SVCS who, after initial chemotherapy for small cell lung cancer, were randomized to receive chemotherapy alone or radiation therapy followed by more chemotherapy.
The addition of a radiation dose of 40 Gy to the mediastinum did not increase the protection from local recurrence or improve the survival.
In some series of small cell lung cancer, SVCS was a favorable prognostic sign, whereas its presence did not affect survival in other reports.
A recent study of 408 patients with small cell lung cancer by Wurschmidt and colleagues showed that the presence of SVCS independently predicted for better survival. Other independent predictors for better survival were stage and performance status. The reason for the possible association of SVCS with better prognosis remains obscure. Some researchers found a higher incidence of brain metastases at the time of diagnosis in small cell lung cancer patients with SVCS than in patients without SVCS.
Three randomized trials have shown an advantage for combining radiation therapy with chemotherapy over chemotherapy alone for the treatment of limited-disease small cell cancer of the lung.
The optimal sequence of the two modalities and the dose and fractionation of radiotherapy have not been established; however, the use of combination chemotherapy as the initial modality, and subsequent rapid shrinkage of the tumor, may eliminate the necessity for irradiating a large volume of lung tissue. When chemotherapy is administered, the arm veins should be avoided. Veins of the lower extremities provide an alternative for simple venous access.
NON-HODGKIN’S LYMPHOMA
The most extensive experience in SVCS management secondary to non-Hodgkin’s lymphoma is reported from the M.D.
Twenty-two patients with diffuse large cell lymphoma and 8 with lymphoblastic lymphoma were evaluated for results of treatment. The patients were treated with chemotherapy alone, chemotherapy combined with irradiation, or radiotherapy alone. All patients achieved complete relief of SVCS symptoms within 2 weeks of institution of any type of treatment. No treatment modality appeared to be superior in achieving clinical improvement. The presence of dysphagia, hoarseness, or stridor was a major adverse prognostic factor for patients with lymphoma who presented with SVCS. Eighteen (81%) of 22 patients with large cell lymphoma achieved a complete response. Relapse occurred in all 6 patients treated with irradiation alone, in 4 of 7 patients treated with chemotherapy alone, and in 5 of 9 patients treated with chemotherapy and radiotherapy. Median survival time was 21 months. All 8 patients with lymphoblastic lymphoma achieved complete response. Six relapses occurred in this group, all in sites not initially involved. Median survival was 19 months.
From these results, the researchers concluded that SVCS secondary to lymphoma is rarely an emergency that requires treatment before a histologic diagnosis is made. They recommended that the choice of treatment be based on the histologic diagnosis and that, if possible, patients undergo a complete staging work-up before therapy. Lymphangiography should be avoided, to prevent embolization of contrast material that could result in respiratory failure. They advocated chemotherapy as the treatment of choice because it provides both local and systemic therapeutic activity. They suggested that local consolidation with radiation therapy may be beneficial in patients with large cell lymphoma and mediastinal masses larger than
A similarly favorable experience in children with T-cell lymphoma or leukemia (nine patients) and Hodgkin’s disease (two patients) who presented with SVCS was reported from Israel.
Tissue diagnosis was obtained before specific therapy in all children, SVCS responded to chemotherapy within 2 to 10 days and overall 3-year disease-free survival rate was 78%.
NONMALIGNANT CAUSES
Patients with nonmalignant causes of SVCS differ significantly from those who have malignant disease. If the cause is not malignancy the patients often have symptoms long before they seek medical advice, it takes longer to establish the diagnosis, and they survive markedly longer.
Schraufnagel and coworkers reported that average survival was 9 years if the primary process was benign, as compared with average survival of 5 months for patients with lung cancer.
Mahajan and associates reviewed the literature of benign SVCS and reported 16 new cases. Twelve (75%) of these 16 patients had a mediastinal granuloma that was attributed to histoplasmosis. Most had insidious onset of SVCS and were relatively young. Ten patients who were available for follow-up of 1 to 11 years were all doing well at the time of the report. It was suggested that the good prognosis of patients with benign SVCS caused by fibrosing mediastinitis does not provide a role of SVC bypass surgery.
However, Nieto and Doty advocated surgery for SVCS caused by benign disorders, if the syndrome develops suddenly, progresses, or persists after 6 to 12months’ observation for possible development of collaterals. In patients with histoplasmosis complement fixation titers suggestive of active disease, ketoconazole treatment may prevent recurrent SVCS.
CATHETER-INDUCED OBSTRUCTION
In catheter-induced SVCS, the mechanism of obstruction is usually thrombosis. Streptokinase, urokinase, or recombinant tissue-type plasminogen activator may cause lysis of the thrombus early in its formation. Heparin and oral anticoagulants may reduce the extent of the thrombus and prevent progression. Removal of the catheter, if possible, is another option, and it should be combined with anticoagulation to avoid embolization. When patients’ pacemaker electrodes must be changed, the broken wire should be removed to prevent risk of developing SVCS.
Percutaneous transluminal angioplasty, with or without thrombolytic therapy, has been successfully used to open catheter-induced SVC obstructions.
RADIATION THERAPY
In patients with SVCS secondary to non-small cell carcinoma of the lung, radiotherapy is the primary treatment. The likelihood of relieving the symptoms and signs of SVCS is high, but the overall prognosis for these patients is poor.
In Armstrong’s group’s series, the 1-year survival rate for these patients was 17% and the survival rate at 2 years declined to 2%. Radiotherapy has been advocated as standard treatment for most patients with SVCS.
It is used as the initial treatment if a histologic diagnosis cannot be established and the clinical status of the patient is deteriorating; however, recent reviews suggest that SVC obstruction alone rarely represents an absolute emergency that requires treatment without a specific diagnosis. The syndrome may be the earliest manifestation of invasive involvement of additional critical structures in the thorax, such as the bronchi. Under such circumstances, prompt treatment with irradiation may be required without any delay.
The fractionation schedule of radiation usually includes two to four large initial fractions of 300 to 400 cGy, followed by conventional fractionation to a total dose of 3000 to 5000 cGy. Patients treated initially with large dose fractions showed slightly faster symptomatic improvement than patients who received conventional-dose radiation. Improvement within 2 weeks or less was observed in 70% of those treated with large initial dose fractions and in 56% of patients receiving conventional-dose therapy. This difference was not statistically significant.
A recent study evaluated the efficacy of treating patients with SVCS with a short course of hypofractionated irradiation. The study compared a regimen of 8 Gy per fraction once a week (to a total dose of 24 Gy) with a program of delivering only two fractions of 8 Gy (total of 16 Gy) within a week. Transient dysphagia was the main side effect in almost half of the patients in both programs. The 24-Gy regimen resulted in complete resolution of symptoms in 56% of patients and partial response in another 40%. The 16-Gy regimen yielded a complete response in only 28% of patients. The mean time for SVCS recurrence and median overall survival were longer with the higher-dose regimen (6 and 9 months, respectively) as compared with the low-dose regimen (3 months for both). Serial venography and autopsy findings suggest that the symptomatic improvement achieved by radiotherapy is not always due to improvement of flow through the SVC, but it is probably also a result of the development of collaterals, after the pressure in the mediastinum is eased.
The field of radiation for SVCS induced by lung cancer should encompass the gross tumors with appropriate margins plus mediastinal, hilar, and supraclavicular lymph nodes. In Armstrong’s series, supraclavicular failures occurred in 8 (9%) of 91 patients receiving radiation therapy to the supraclavicular fossae, and 2 (33%) of 6 patients who did not receive therapy to these lymph nodes exhibited failure at this site.
TRANSLUMINAL ANGIOPLASTY AND ENDOPROSTHESIS INSERTION
Percutaneous transluminal angioplasty using balloon technique or insertion of expandable wire stents has been used successfully to open and maintain the patency of the SVC, even after maximal-tolerance radiation therapy.
Hennequin and associates obtained relief of symptoms and persistent patency of the SVC in 93% of patients with a malignancy-induced SVCS by inserting a Wallstent endoprosthesis.
Gaines and colleagues successfully treated 18 of 20 patients with SVCS due to malignancy using the Gianturco Z stent. After inserting Z-type-metal stents, Oudkerk and colleagues attained complete or nearly complete resolution of symptoms in 86% of patients with SVCS. In two cases, percutaneous atherectomy was used to establish the histologic diagnosis of SVCS and treatment of the obstruction. In catheter-induced SVCS administering a thrombolytic agent through the angioplasty device may be therapeutically beneficial.
SURGERY
Experience with successful direct bypass graft for SVC obstruction is limited. It was recommended that autologous grafts of almost the same size as the SVC should be used.
Doty and coworkers used a composite spiral graft constructed from the patient’s saphenous vein. They reported 15 years’ experience with this procedure in nine patients with benign obstruction of SVC: vessels of seven were maintained patent, and all patients were relieved of symptoms of SVCS. Avashti reported successful bypass of obstructed SVCs using Dacron prostheses.
Magnan and associates used an expanded polytetrafluoroethylene prosthesis to reconstruct the SVC in nine patients with malignancy-induced SVCS and in one with chronic mediastinitis.
In all patients, symptoms disappeared promptly after the operation, the grafts remained open, and survival rates at 1, 2, and 5 years were 70%, 25%, and 12.5%, respectively.
The preferred bypass route is between an innominate or jugular vein on the left side and the right atrial appendage, using an end-to-end anastomosis. [ref: 42] Piccione and colleagues used the autologous pericardium to reconstruct the SVC after resection for malignant obstruction.
In a patient with malignancy-induced SVCS, surgical intervention should be considered only after other therapeutic maneuvers with irradiation and chemotherapy have been exhausted. Most patients with SVCS of benign origin survive long without surgical intervention however, when the process progresses rapidly or if there is arterosternal goiter or aortic aneurysm, surgical intervention may relieve the obstruction.
THROMBOLYTIC THERAPY
Successful experience with thrombolytic agents is limited to the treatment of catheter-induced SVCS.
A review of the response of SVCS to thrombolytic therapy from the Cleveland Clinic showed that in 8 (73%) of 11 patients with a central venous catheter, the thrombus was effectively lysed, as compared with only 1 of 5 patients who responded to thrombolytic therapy in the absence of a central catheter.
The higher yield for thrombolytic therapy in patients with catheters is probably related to the mechanism of obstruction, the ability to deliver the agent directly to the thrombus, and earlier recognition of SVCS in patients with malfunctioning catheters. In the Cleveland Clinic experience urokinase was more effective than streptokinase, and a delay in administering therapy beyond 5 days of symptom onset was associated with treatment failure.
Favorable experience with recombinant tissue-type plasminogen activator as a thrombolytic agent for catheter-induced SVCS has been reported.
GENERAL MEASURES
Medical measures other than specific chemotherapy may temporarily relieve the symptoms of SVCS. Bed rest with the head elevated plus oxygen administration can reduce cardiac output and venous pressure. Diuretic therapy and reduced-salt diet to reduce edema may have an immediate palliative effect, but the risk of thrombosis enhanced by dehydration should not be ignored. Steroids are commonly used, but their effectiveness has never been properly evaluated. They mayimprove obstruction by decreasing a possible inflammatory reaction associated with tumor or with irradiation; however, Green and associates demonstrated the lack of inflammatory reaction and edema after radiotherapy for experimental SVCS, although rigorous documentation is lacking.
Thrombolytic therapy with urokinase, streptokinase, and recombinant tissue-type plasminogen activator was effective in catheter-induced SVCS.
MANAGEMENT RECOMMENDATION
When no cause for SVCS can confidently be established, an efficient diagnostic effort should be attempted before any specific treatment is given. Three deep-cough sputum specimens should be obtained for cytologic analysis. A positive cytologic evaluation provides reliable pathologic information, particularly in the diagnosis of small cell lung carcinoma.
If there is pleural effusion, thoracocentesis should be performed and the centrifuge-prepared specimen examined for malignant cells. If a suspicious lymph node is palpable, particularly in the supraclavicular area, needle or open biopsy should be the next diagnostic step. In the absence of positive sputum results, pleural effusion, or accessible suspicious lymph node analysis, bronchoscopy should be performed and brushing, washing, and biopsy samples should be obtained for cytologic and histologic analysis. If these efforts do not provide histologic diagnosis of the primary process, percutaneous transthoracic fine-needle biopsy under CT or fluoroscopic guidance is safe and highly effective. In the rare case when less invasive procedures fail to establish the diagnosis, the location of the suspicious lesion in the chest and the experience of the surgical team should determine whether mediastinoscopy or thoracotomy is performed.
During the diagnostic process the patient can benefit from bed rest with the head elevated and with oxygen administration. Some clinicians advocate the use of diuretics and steroids, 6 to 10 mg of dexamethasone given orally or intravenously every 6 hours, as a temporary palliative measure if the patient is uncomfortably symptomatic. Anticoagulation has no proven benefit and may interfere with diagnostic procedures. After the cause of SVCS has been established, treatment of the primary process should promptly follow. Combination chemotherapy with an appropriate regimen is the treatment of choice for small cell lung cancer and NHL. Radiation therapy of the lesion and adjacent nodal areas may enhance control after initial response to chemotherapy. Non-small cell lung cancer causing SVCS is best treated with radiation therapy. The incorporation of CT information into a carefully designed treatment plan may enable the administration of a total radiation dose greater than 5000 cGy, which may provide long-term local control for some patients. Most patients with nonmalignant causes for SVCS have an indolent course and a good prognosis. Percutaneous transluminal angioplasty or stent insertion should be considered an effective alternative to surgery. Surgery is indicated only when the process is progressing rapidly or is caused by retrosternal goiter or aortic aneurysm. If SVCS is induced by a catheter, the catheter should be removed, if possible. Heparin should be administered during the removal of the catheter to prevent embolization. In catheter-induced SVCS, urokinase, streptokinase, or recombinant tissue-type plasminogen activator is of value if used early in the thrombotic process.
The clinical course of SVCS rarely represents an absolute emergency. In these situations, the bronchus is likely to be obstructed by the same basic process, and irradiation may have to be started immediately, even before the histologic diagnosis is established.
Surgical Emergencies
A variety of clinical scenarios may prompt surgical consultation and intervention in cancer patients. Conditions arising from the cancer itself, complications of the therapy, or totally unrelated situations may require more or less urgent intervention. The surgeon becomes part of a multidisciplinary team who must constantly assess the immediate and long-term goals of any therapy in the broader context of the patient’s underlying disease. The first question usually asked of a surgeon is whether or not the problem needs surgical intervention; second, what type of intervention is needed; and third, what is the proper timing? These same questions arise in other serious illnesses but are more complex in the setting of cancer. This chapter focuses primarily on abdominal issues that are unique to the cancer patient. The reader is referred to general surgical textbooks for detailed coverage of disease processes.
Additional resources can be found in reviews of surgical problems in immunocompromised patients, and more specifically in cancer patients.
ASSESSMENT OF THE PROBLEM
To assess the acute symptoms carefully, the physician must have a thorough knowledge of the patient’s cancer history, including past medical and surgical interventions. The current status of the cancer — whether cured or known or suspected to be relapsing, affects the assessment of the acute condition. Knowledge of recent chemotherapy and the expected time course of associated neutropenia and thrombocytopenia are critical for establishing a diagnosis and a treatment plan. The current medications used should be carefully delineated. Patients receiving corticosteroids may have a blunted symptom complex and physical examination. Previous chemotherapy may have resulted in specific organ dysfunction that is transient or permanent.
Thrombocytopenia may require correction before surgical intervention. After the physical examination, laboratory and radiographic studies are obtained. Febrile patients should also have blood and urine taken for culture, and stool for culture as indicated. Plain films of the abdomen and chest are obtained for most patients with abdominal pain or obstructive symptoms. Though the yield from such studies may be low, [ref: 10] reviewing them with a radiologist may add valuable information. Additional radiographic studies may be performed as indicated. Diagnostic laparoscopy is valuable and accurate; its current role is being defined.
Clearly, the diagnostic process is best accomplished with multidisciplinary input and may require repeated and frequent assessments. The patient’s immune function, nutrition status, and cancer prognosis are important factors in formulating a treatment plan. Ethical, social, and spiritual concerns of the patient, family, and physician add a complex dimension to the management plan.
ESTABLISHING A DIAGNOSIS
ABDOMINAL PAIN
The differential diagnosis of abdominal pain in cancer patients is quite extensive. Particularly challenging are patients receiving corticosteroids, whose symptoms are blunted and diagnosis often delayed. Remine and McIlrath reviewed their experience with gastrointestinal perforation in 79 patients receiving corticosteroids. The most consistent clinical finding, abdominal pain, was present in all but one patient. The pain was less severe in patients treated with large doses of steroids, and this led to a significant delay in diagnosis. In general, abdominal pain is the most important indicator of acute abdominal disease in immunocompromised persons. A frequent cause of generalized abdominal pain is constipation. Chemotherapeutic agents are commonly associated with alterations in gut function. Vincristine is one of the drugs most frequently associated with paralytic or adynamic ileus, which may occur in as many as half of patients receiving this vinca alkaloid.
Patients experience colicky abdominal pain (which may be severe), constipation, and abdominal distention within several days of receiving vincristine. Therapy is generally conservative and may require nasogastric tube decompression. Chronic gastrointestinal toxicity secondary to vincristine chemotherapy is rare. Constipation may also be a sequela of the narcotics that frequently are used to treat cancer pain. Immobilization, dehydration, and electrolyte imbalances are additional contributors to constipation. All of these factors should be addressed when treating constipation. Interventions that help maintaiormal bowel function also have great impact on preventing hemorrhoids and anorectal complications, which are discussed later.
Gastrointestinal dysmotility may be neuropathic in origin. Intestinal pseudoobstruction due to paraneoplastic neuropathy of the gastrointestinal tract has been documented in a patient with lung cancer. Pseudoobstruction of the colon was initially described by Ogilvie in
Ogilvie theorized that colonic distention in these patients was secondary to malignant invasion of the celiac axis resulting in sympathetic denervation of the colon. Subsequently, this entity, known as Ogilvie’s syndrome, has been recognized in a multitude of noncancerous conditions. It must be differentiated from gastric distention and colonic volvulus. Colonic decompression to prevent perforation is recommended when cecal distention reaches 12 to
A “nonsurgical” cause of right upper quadrant pain is venoocclusive disease. This entity occurs in 4% to 22% of patients following intensive chemotherapy and bone marrow transplantation. Clinical manifestations include severe right upper quadrant pain, hepatomegaly, ascites, jaundice, and in some cases encephalopathy. This entity is a result of hepatic centrilobular hemorrhagic necrosis leading to obliteration of sinusoids and terminal venules. It usually occurs within 3 weeks of marrow infusion and at a time when patients are severely leukopenic and thrombocytopenic.
The diagnosis is usually made clinically after ruling out other entities such as cholecystitis, pancreatitis, and ulcer disease. Treatment is supportive. Making the correct diagnosis and avoiding unnecessary laparotomy are important in these generally ill patients. An entity with a clinical presentation similar to that of venoocclusive disease is Budd-Chiari syndrome. This syndrome is the result of hepatic venous outflow obstruction secondary to lesions such as hepatoma, adrenal carcinoma, hypernephroma, or thrombosis. Surgical treatment with shunting is advocated because of the high mortality associated with medical management. Left upper quadrant pain or tenderness may be a manifestation of splenomegaly in patients with lymphoma.
Splenic infarct in these patients may lead to exacerbation of pain and referral of pain to the left shoulder. Focal left or right abdominal wall tenderness should raise suspicion of a rectus sheath hematoma occurring in thrombocytopenic patients or those taking coumadin.
Similarly, retroperitoneal bleeding should be considered in patients who have abdominal pain and altered coagulation parameters.
Focal abdominal wall pain that mimics an acute abdomen may be a manifestation of herpes zoster. The lack of peritoneal signs and the subsequent appearance of skin lesions differentiate this condition from one needing surgical intervention. Bulky retroperitoneal adenopathy may cause abdominal pain. Infrequently this pain may be intense and the clinical symptoms may resemble those of a ruptured aortic aneurysm.
Tumor invasion of the pancreas may occasionally lead to pancreatitis manifesting classic signs and symptoms of this inflammatory process. Pancreatitis can also result from therapy with various cytotoxic agents such as L-asparaginase, 6-mercaptopurine, and corticosteroids. Alternatively, pancreatitis after high-dose cyclophosphamide has been attributed to tumor lysis in two patients with lymphoma. Malignant ascites can result in considerable abdominal distention and discomfort. Its management is the subject of a subsequent chapter in this text. Spontaneous bacterial peritonitis, which occurs in up to 10% of patients with cirrhotic ascites, is rare in patients withmalignant ascites.
Urologic causes of abdominal pain, such as obstructive uropathy (due to cancer, scarring, or stones from postchemotherapy hyperuricemia), cystitis, pyelonephritis, and testicular torsioeed to be considered in cancer patients with acute abdominal symptoms. Urologic emergencies are discussed in the next chapter and reviewed elsewhere.
In women the diagnostician must consider gynecologic causes of abdominal pain such as tuboovarian abscess, ectopic pregnancy, hemorrhage from ovarian cysts, and adnexal torsion. Having discussed a variety of conditions that cause abdominal pain in cancer patients, we now direct our attention to those that most often require surgical intervention.
EMERGENT SURGICAL PROBLEMS
Abdominal surgical procedures were performed emergently in 3% to 5% of patients with leukemia or lymphoma in five large series totalling 2227 patients. Selective early surgical intervention was advocated in these series, despite the high postoperative mortality of 17% to 55%. In the presence of a true surgical emergency, the surgeon is left with little choice but to intervene, since conservative (nonoperative) management carries a grim prognosis. Following myelosuppressive chemotherapy, patients are at risk of developing abdominal complications and not infrequently experience abdominal pain. Surgical intervention in neutropenic patients experiencing acute abdominal pain was necessary in 14% to 34% of patients with predominantly hematologic malignancies.
OBSTRUCTION
Obstruction, one of the more frequent reasons for emergency abdominal surgery in cancer patients, accounts for nearly one third of procedures. The cause of obstruction in cancer patients is recurrent disease in 59% to 100%, whereas benign conditions account for as many as 39% of presentations. This is in contrast to the 7% to 9% incidence of malignant obstruction ionselected series of bowel obstruction.
After submitting to careful history taking and physical examination and giving blood for studies, patients admitted with obstructive symptoms such as nausea, vomiting, constipation, crampy pain, and abdominal distention should undergo radiographic evaluation by plain films of the abdomen (supine and upright or left lateral decubitus) and chest. With a tentative diagnosis of ileus or partial bowel obstruction and in the absence of fever, leukocytosis, or peritonitis, patients deserve a trial of conservative therapy. They are monitored with serial physical examinations done by the same physician and daily abdominal radiographs. In nonselected cases of partial small bowel obstruction such treatment may result in resolution of symptoms in as many as 88% of patients.
In the cancer setting, use of nasoenteric decompression, intravenous hydration, and electrolyte replenishment results in spontaneous resolution of obstructive symptoms in 12% to 29% of patients in a mean interval of 3 to 9 days. Bowel strangulation and gangrene are rare in these circumstances (no more than 5%), with the exception of one series reporting a 24% incidence. Preoperative recognition of bowel strangulation cannot be done reliably in patients with complete mechanical obstruction.
The use of nasogastric tube or long intestinal tube (i.e., Miller-Abbott, Cantor, Baker) for decompression is the physician’s. Both nasogastric and long tubes have been reported to have similar success, though long tubes were associated with longer hospitalization. A recent randomized study of adhesive small bowel obstruction failed to identify one tube as superior to the other. The most common practice is to use the nasogastric tube because it is easy to place and effective in decompressing the upper gastrointestinal tract. Long tubes are more difficult to get into position (i.e., to pass beyond the duodenum) but may be useful in delineating a site of obstruction by stopping at the obstructed point and serving as a port for instillation of radiographic contrast agents. Long tubes may also prove useful intraoperatively when the surgeon faces carcinomatosis and dense adhesions.
Locating a suitable segment of small bowel for bypass may be facilitated by this tube. Contrast radiographic studies are best used selectively. Patients with nondiagnostic plain abdominal films, atypical history or examination findings, or protracted courses without resolution of obstructive symptoms are candidates for contrast studies. Coupled with fluoroscopy, contrast agents help answer questions about gut motility and transit time. For patients with suspected colonic obstruction or both small and large bowel obstruction with no perforation, barium enema can be quite useful. One third of patients suspected of having mechanical large bowel obstruction do not have obstruction on contrast enema images. The barium enema should be terminated promptly if the fluoroscopist detects an obstructing lesion. The colon can absorb water from the barium, which then becomes inspissated above a partially obstructing lesion, making subsequent colonic cleansing very difficult.
If no obstructing lesion is visualized in the colon, it is safe to perform an upper gastrointestinal series with small bowel follow-through. In cases of suspected perforation, a water-soluble agent such as gastrografin should be used. Otherwise, barium is the agent of choice because water-soluble agents become progressively diluted with bowel contents and may not provide the necessary detail. Barium that becomes trapped above a small bowel obstruction does not result in inspissation as it would in the colon because the small bowel has less water absorptive capacity.
Successful nonoperative management can be expected in 12% to 29% of cancer patients with obstructive symptoms. 32% to 45% of these patients, however, develop recurrent obstruction, many of whom require surgical intervention.
When an initial trial of conservative therapy for 3 or more days fails, patients are candidates for surgical exploration. Radiographic contrast studies are not necessary but are often useful. The results are particularly helpful intraoperatively when the extent of disease makes definition of the anatomy difficult. Knowledge of fistulas or sites of obstruction may allow a more expeditious and safe operative procedure, whether it be palliative or definitive. As many as one third of cancer patients undergoing surgery may have adhesions or a benign cause of obstruction and the operative success in resolving their symptoms is usually high. Despite the majority of obstructions being tumor related, 55% to 96% of patients can be relieved of their symptoms surgically (see Table 49.4-2). Recurrent obstruction after successful surgical relief of symptoms occurs in 9% to 33% of patients.
Medical management of surgically uncorrectable lesions using pharmacologic agents without intravenous fluids or nasogastric tubes has been successful. Octreotide administered subcutaneously may reduce gastrointestinal secretions and control vomiting.
Home are with intravenous hydration and pharmacologic support has been advocated in some series.
In selected patients with proximal obstruction and failure of pharmacologic support, palliation may be obtained by percutaneous endoscopic gastrostomy placement.
For patients with obstructing colorectal lesions, palliation may be obtained with endoscopic dilatation, diathermy, and laser recanalization or passage of metal stents.
The management of bowel obstruction in advanced cancer has been nicely reviewed by Ripamonti.
Operative mortality (median 16%) and morbidity (median 31%) have remained high and reflect the generally ill nature of this patient population
Mortality is highest (38% to 54%) among elderly patients (older than 70 years) who undergo emergency surgery.
Elective cancer surgery in the elderly is less risky (mortality 3% to 7%) and is largely associated with preexisting conditions.
All patients, particularly the elderly, need very careful preoperative preparation and management of associated medical problems, which are significant contributors to postoperative morbidity and mortality. Age alone should not preclude extensive surgery, which can be justified in carefully prepared elderly patients.
Colorectal, ovarian (gynecologic), and gastric malignancies are the most common (nearly three quarters) primary tumors in patients who subsequently develop intestinal obstruction. Less frequent are a variety of other intraabdominal and extraabdominal tumors, such as pancreatic, breast, lung, mesothelioma, and melanoma. The anatomic sites of obstruction in four combined series were small bowel (59%), colorectal (29%), both small bowel and colorectal (5%), and stomach and duodenum (7%). The most common operative treatments are bypass, stoma creation, resection, and lysis of adhesions.
Gastroduodenal obstructions not caused by a benign condition such as ulcer disease may be secondary to a primary cancer or to direct extension from kidney, pancreas, or colon carcinoma. Surgical resection, if possible, or alternatively bypass, should be undertaken.
The goal is to reestablish gastrointestinal continuity with curative intent or at least to provide palliation. The scope of surgical procedures is beyond this discussion.
The surgical management of malignant small bowel obstruction is resection with primary anastomosis, or bypass if resection is not warranted or possible. Primary small bowel tumor (adenocarcinoma, sarcoma, carcinoid, lymphoma) should be suspected when patients have abdominal pain, weight loss, and occult gastrointestinal bleeding. Metastatic small bowel lesions should be considered in patients with these symptoms and a history of cancer of the ovary, colon, stomach, pancreas, or breast or melanoma, as previously discussed. Small bowel lesions may cause obstruction by their intraluminal or extraluminal location or by intussusception.
Primary colon cancer is the most frequent reason for acute colonic obstruction (78% of patients); extra-colonic malignancy (12%) and benign causes (10%) account for the rest.
Fifteen to sixteen percent of colon cancers present with intestinal obstruction. Intussusception is an infrequent mechanism of obstruction in these patients. A very rare cause of colonic obstruction is metastatic prostate cancer resulting in a separate annular stricture of the rectosigmoid colon.
These lesions are commonly confused with primary colon cancer. They are characterized by normal overlying mucosa and require deep biopsy for successful diagnosis. Initial therapy should be hormonal; surgery is reserved for relief of obstruction.
Surgical management of colonic obstruction has generated controversy. The best approach is one individualized for the patient and comfortable to the surgeon. Right hemicolectomy for obstructing right colon lesions or extended right hemicolectomy for transverse colon lesions followed by primary anastomosis is generally considered safe in emergent conditions without preoperative bowel preparation. Obstructing carcinoma of the splenic flexure can be managed by extended right hemicolectomy and immediate anastomosis, though some surgeons would prefer to include these lesions in left colon resection.
Surgical management of obstructing left colon and sigmoid lesions is evolving from the traditional three-stage procedure to a two-stage one, and, most recently, a single-stage operation.
The presence of peritoneal soiling or inflammation, the condition of the bowel, the stability of the patient, and ultimately the judgment and expertise of the surgeon dictate what approach is used. All three methods have different roles, depending on the clinical setting.
The three-stage procedure for lesions distal to the splenic flexure involves (1) creation of a proximal (transverse colon) decompressing colostomy, (2) elective resection of the lesion in the mechanically cleansed bowel and primary anastomosis, and (3) colostomy closure.
Presently a two-stage procedure is the more common approach; it consists of primary resection with the creation of a proximal end colostomy and Hartmann pouch or mucous fistula.
The second stage takes down the colostomy and reestablishes bowel continuity. Retrospective data as well as those from one randomized study support the use of a two-stage (as opposed to a three-stage) approach.
The most recent trend in treating obstructing left colon lesions has been single-step segmental resection and primary anastomosis. Carefully selected patients can safely undergo a one-stage resection without mechanical bowel preparation.
The most aggressive approach — subtotal or total colectomy and ileocolonic-ileorectal anastomosis — obviates mechanical bowel cleansing. To make a one-stage procedure more widely applicable and retain a low incidence of anastomotic leak, a variety of approaches that allow mechanical bowel cleansing have been advocated. Preoperative techniques include endoscopic laser relief of obstruction, decompression by intralesional passage of a tube or metal stent advanced over a guidewire, colonic irrigation via long intestinal tube, and percutaneous colonic irrigation.
Intraoperative performance of an intracolonic bypass with a Silastic and latex tube has been described. The technique used most frequently and gaining popularity is intraoperative colonic irrigation, as advocated by a variety of groups. The technique consists of proximal (cecal) infusion of saline and distal (above the obstruction) collection of luminal contents in a closed system. Following resection and mechanical cleansing, primary anastomosis is performed. Alternatively, a long intestinal tube inserted preoperatively has been used for intraoperative colonic irrigation. Some patients have undergone proximal diversion with a loop colostomy or ileostomy to protect the anastomosis. At present, one-stage procedures for obstructing left colon lesions are not advocated in the presence of fecal peritonitis, a large contiguous pelvic abscess, or systemic sepsis. For high-risk critically ill patients, proximal diverting colostomy is still a viable alternative.
PERFORATION
Gastrointestinal perforation is one of the more frequent reasons for emergency abdominal surgery in cancer patients.
Perforation at a site involved by tumor is the cause in 33% to 58% of cases.
Nonmalignant causes of perforation include gastroduodenal ulcer, postsurgical bowel anastomosis disruption, colonoscopy, colostomy irrigation, radiation enteritis, diverticulitis, neutropenic colitis, infection (cytomegalovirus, Clostridium difficile), and perforation proximal to an obstructing lesion.
A cause of perforation described recently is paclitaxel therapy.
One of the malignant causes of intestinal perforation reported more frequently is lymphoma, which accounts for almost half of tumor sources.
Primary gastrointestinal lymphoma is rare, accounting for only 2% of non-Hodgkin’s lymphomas in one series of 4234 patients. It is reported that fewer than 20% of patients with primary gastrointestinal lymphoma (of stomach, small bowel, or colon) develop perforation on presentation or during chemotherapy. Perforation appears to be less frequent in patients with leukemia than in patients with lymphoma
Free perforation from gastric cancer is rare (prevalence, 1% to 4% of gastric cancers). Perforation due to colon cancer is also infrequent, occurring in 6% of 533 patients undergoing curative colon surgery ; however, perforation accounts for approximately one fourth of emergency laparotomies for colorectal cancer. The site of perforation is usually at the tumor, though perforation proximal to obstructing colon lesions has been described.
“Extra gut” malignancies that can metastasize to the intestinal wall and are associated with perforation include those of lung, breast, and ovary and melanoma.
Abdominal viscus perforation classically manifests symptoms of generalized peritonitis: severe constant abdominal pain, diffuse tenderness and guarding, ileus and abdominal distention, fever, hemodynamic alterations requiring fluid resuscitation, and leukocytosis. Patients receiving chemotherapy or steroids may not be able to manifest the usual clinical findings associated with perforation, and this poses a diagnostic challenge. Symptoms of perforation were blunted in a series of 79 patients receiving steroids; the result was significant delays in diagnosis. The most consistent finding was abdominal pain, which occurred in all but one of these patients. Free intraabdominal air was visualized radiographically in only 46% of them. In general, half of patients with gastrointestinal perforation do not have free intraperitoneal air visualized.
The presence of pneumoperitoneum, however, is diagnostically quite valuable and usually mandates abdominal exploration. A degree of caution is warranted when acting on this radiographic finding: for 14% of patients with free intraperitoneal air the source is something other than a perforated abdominal viscus. Gastrografin contrast radiographic studies may be useful in delineating the site of perforation and in planning therapy.
The surgical management of intestinal perforation in cancer patients depends on the benign or malignant cause, location of the lesion in the gastrointestinal tract, the degree of associated peritonitis, and the patient’s condition and prognosis. The surgical goal is to perform the simplest lifesaving procedure that controls sepsis, establishes gastrointestinal continuity, and allows for prompt therapy of the underlying malignancy. Definitive or curative surgical therapy may be indicated in some circumstances.
In general, gastroduodenal perforations are caused by acute ulceration and are managed with a simple patch closure procedure. Gastric ulcers that appear chronic should be biopsied and if frozen section shows malignancy, resected.
In precarious clinical situations simple closure of a perforated lymphoma has been performed successfully, though this is not generally advocated. Perforations of the small bowel are best managed by resection and primary anastomosis, reserving exteriorized stomas for cases of severe peritonitis.
Perforation of the large bowel are generally managed by resection, colostomy, and Hartmann’s procedure or mucous fistula. Primary anastomosis has been advocated after resection of right colon perforation in selected cases.
Perforations of the gastrointestinal tract into adjacent structures are infrequent but deserve mention. Malignant duodenocolic fistula is a rare complication of carcinoma of the hepatic flexure, presenting with bilious diarrhea and feculent vomiting. Resection with curative intent is preferable, though palliative diversion has been described for unresectable lesions.
Enterovaginal or enterovesical fistulas may be most amenable to proximal bowel diversion to control sepsis and provide palliation. Duodenal wall or malignant retroperitoneal tumors may invade and perforate the duodenum, resulting in persistent infection. For unresectable tumors a technique of duodenojejunostomy and distal duodenal stapled occlusion has resulted in successful resolution of sepsis.
BLEEDING
Bleeding is one of the most frequent abdominal surgical emergencies of cancer patients. In a series of cancer patients undergoing emergency surgery for intraabdominal bleeding, postoperative hemorrhage was the most frequent source, occurring in two thirds of episodes. In a broader context, serious bleeding has occurred in 15% of 301 patients receiving myelosuppressive chemotherapy for solid tumors.
The cause of bleeding in cancer patients is infrequently the tumor itself: tumor is the source of bleeding in 12% to 27% of patients undergoing endoscopy; gastritis and ulcer disease account for 44% to 76%; and esophagitis and esophageal varices occur in 11% to 17%. An infrequent cause of bleeding is Mallory-Weiss tear at the gastroesophageal junction from chemotherapy-induced retching and vomiting. Infections such as cytomegalovirus enteritis have also been reported to cause intestinal bleeding.
A useful algorithm for the evaluation and treatment of hemorrhage in cancer patients was outlined by Baker. When a patient presents with signs and symptoms of intraabdominal bleeding, the physician must initiate a search for the cause and identify any contributing factors, such as thrombocytopenia or altered coagulation profiles secondary to diminished hepatic synthetic function, malnutrition, or disseminated intravascular coagulation.
Fluid and blood product resuscitation should be initiated as a diagnosis is being established. The clinical findings and the urgency of the situation dictate the subsequent diagnostic approach and therapy, whether it be supportive, with interventional radiology, or surgical.
Deciding whether bleeding is (1) intraluminal or extraluminal and (2) proximal or distal to the ligament of Treitz is useful for selecting diagnostic modalities.
Passage of a nasogastric tube may quickly determine if bleeding is originating in the upper gastrointestinal tract. Identification of a site of intraluminal bleeding should be attempted with esophagogastroduodenoscopy, or proctosigmoidoscopy, or colonoscopy, depending on the site of bleeding. Upper gastrointestinal bleeding can be diagnosed accurately with endoscopy in 74% to 96% of patients.
Fortunately, the majority of bleeding from nonmalignant causes stops spontaneously without therapeutic intervention.
Endoscopic therapy with thermal contact devices, laser, and injection treatments are effective in decreasing bleeding and the need for surgery.
Bleeding peptic duodenal ulcers that need surgical intervention are best treated with truncal vagotomy, pyloroplasty, and oversewing with plication of the ulcer bed.
If endoscopy does not reveal the source of bleeding, imaging studies are warranted. Technetium 99m-labeled red blood cell scans can localize the site of bleeding in 83% of patients and allow repeated scanning during a 24-hour period, which is useful since most bleeding is intermittent.
The authors found this technique superior to angiography, which visualizes only active bleeding at a rate greater than 1 mL per minute. Though angiography is diagnostically successful in only 50% to 65% of patients, it has the therapeutic advantage of vasopressin infusion or embolization to control hemorrhage.
Vasopressin infusion was popular in the early 1970s and in a review of published reports was 55% successful in stopping bleeding from a variety of benign causes.
Gelfoam embolization has been used selectively to control hemorrhage in cancer patients, though iatrogenic bowel infarction has been a complication.
Malignant causes of gastrointestinal hemorrhage are generally managed surgically. Gastric cancer is an infrequent cause of upper gastrointestinal bleeding (approximately 2%), and bleeding is the initial presentation for fewer than 10% of gastric cancers.
Early surgical resection of these lesions is advocated, though emergency procedures are associated with high mortality. Small intestinal hemorrhage is the presumed source of bleeding when upper and lower gastrointestinal tract endoscopy are negative, and particularly when bleeding through the ileocecal valve is noted. Enteroclysis or small bowel barium studies may provide the diagnosis and supplement information obtained from a radionuclide study. Primary small bowel tumors (adenocarcinoma, sarcoma, carcinoid, lymphoma) or metastatic lesions (melanoma, lung, testicular) may present with bleeding.
Major bleeding from colon cancer is uncommon. In one series of 1033 patients with colorectal cancer, fewer than 3% required emergency admission for bleeding. Only 7% of cases of massive lower gastrointestinal hemorrhage from a variety of causes were due to adenocarcinoma of the colon.
Infrequently, colonic lymphoma may present with intestinal bleeding.
When emergent surgery for colonic bleeding is undertaken, resection and primary anastomosis is recommended.
For unresectable low colonic lesions, palliative endoscopic laser therapy may be effective in controlling bleeding.
Unusual causes of gastrointestinal bleeding in cancer patients include hemobilia due to hepatobiliary tumors or instrumentation.
Therapy may include embolization or hepatic resection. Iliac arterial-enteric fistulas secondary to pelvic irradiation are also infrequent sources of bleeding and are best treated with bowel resection, arterial ligation, and extraanatomic bypass.
Aortoduodenal fistula from perforation of a metastatic tumor can result in severe hemorrhage and can require urgent surgical correction. For suspected extraluminal intraabdominal bleeding CT can be a valuable diagnostic tool.
Sources of intraabdominal bleeding include leiomyoblastoma of the stomach, pheochromocytoma, carcinoma and angiomyolipoma of the kidney, pelvic neoplasms (urinary bladder, uterus, ovary, prostate), retroperitoneal leiomyosarcoma, spontaneous splenic rupture (hematologic malignancies, hemangiosarcoma), and spontaneous rupture of hepatocellular carcinoma or hepatic metastasis.
Iatrogenic causes of intraabdominal bleeding include liver and spleen injury following thoracentesis or after elective percutaneous needle biopsy. In one series of 9212 liver biopsies, 0.3% of all patients developed major bleeding.
Malignancy was a risk factor for subsequent hemorrhage: estimated risks for fatal and nonfatal bleeding were 0.4% and 0.6%, respectively. Treatment of intraabdominal bleeding depends on the cause. In a good operative candidate, surgical resection of a bleeding tumor is the preferred approach. Operative control of bleeding that is not amenable to resection mandates an alternative such as packing, direct suture ligation, ligation of arterial inflow, application of topical hemostatic agents such as thrombin and Gelfoam, or fibrin glue delivered as an aerosolized spray.
Alternatively, angiographic embolization may be used to stabilize the patient in anticipation of surgery or as definitive therapy.
Patients with widely metastatic disease whose condition is poor are probably best served with nonoperative techniques.
INFECTION
Infection is reviewed separately in this textbook, and the focus of this section is conditions that prompt surgical consultation. Cancer patients, by virtue of underlying disease or treatment with cytotoxic drugs, are immunocompromised and at risk of developing infections.
The recent rise iosocomial infections, which include organisms such as coagulase-negative staphylococci, Staphyloccus aureus, enterococci, and Candida species, is a concern. The majority of infections requiring abdominal surgery at
The changing frequency of organisms — and in particular the rise in Candida species — emphasizes the need for input from infectious disease consultants and judicious use of antibacterial and antifungal agents. Cancer patients develop infectious processes such as cholecystitis, appendicitis, and diverticulitis that may be coincidental to their cancer diagnosis. Neutropenic enterocolitis, hepatobiliary infections, abscesses, and soft tissue and perianal infections more directly related to the underlying malignancy or its therapy are discussed in detail next. Neutropenic Enterocolitis Abdominal pain is not an infrequent complaint of cancer patients, and a multitude of causes have previously been discussed. Among neutropenic patients who require surgical consultation for abdominal pain, 40% to 44% have had a clinical diagnosis of neutropenic enterocolitis.
The true prevalence of this disease entity is not known, owing to its broad inclusion criteria and lack of pathologic confirmation, but it is generally reported to be 3% to 11%, or, for patients with leukemia, as high as 32%.
The clinical syndrome is characterized by abdominal distention, pain, tenderness (often localized to the right lower quadrant), watery (sometimes bloody) diarrhea, and fever in the setting of neutropenia and thrombocytopenia.
It most often occurs in patients with hematologic malignancies (principally leukemia) and develops approximately 7 to 9 days after onset of neutropenia. The pathogenesis of this septic, inflammatory process is not clear. It generally localizes in the terminal ileum and right colon. When confined to the cecum it is termed typhlitis.
The diagnosis of enterocolitis is made clinically, and the differential diagnosis includes pseudomembranous colitis, appendicitis, and diverticulitis.
Examination of stool for Clostridium difficile toxin helps rule out pseudomembranous colitis.
Differentiating appendicitis from enterocolitis may be impossible preoperatively; both entities occur with similar frequency in leukemic children with right lower quadrant pain. Appendicitis occurred in 3% of 2237 children and adults (collected series) with malignancies.
Plain radiographs of the abdomen may demonstrate an ileus pattern and a distended cecum. Sonography and CT have been helpful in demonstrating colon wall thickening and pericolic fluid.
Sigmoidoscopy and radiopaque contrast studies, though generally avoided because of the risks of perforation, bleeding, and sepsis, have been used selectively. The initial recommended management of neutropenic colitis is nonoperative and consists of bowel rest, decompression, hydration, correction of coagulopathy, nutritional support, and broad-spectrum antibiotics.
Patients who show evidence of perforation, peritonitis, uncontrollable bleeding, increasing fluid, or vasopressor requirements, or who fail to improve after 2 to 3 days’ medical management should undergo surgery. Right colectomy and ileostomy is advocated, though some authors perform primary anastomosis in patients without perforation or gross peritonitis.
Because patients for whom medical management of neutropenic enterocolitis is successful are at risk for recurrence, one author has recommended that subsequent chemotherapy be given in the settings of bowel rest, total parenteral nutrition, and avoidance of antidiarrheal agents, with consideration given to performing a prophylactic diverting ileostomy.
Hepatobiliary Complications In cancer patients hepatobiliary disease may pose a diagnostic challenge. Factors such as the presence of malignancy, gastrointestinal dysfunction secondary to chemotherapy, neutropenia, use of hepatotoxic drugs, steroids, and atypical infections may alter the usual presentation of many disorders. Jaundice in a cancer patient requires an accurate diagnosis so that proper medical, interventional, or surgical treatment can be instituted. Liver dysfunction caused by cytotoxic drugs [ref: 9] or hepatic and systemic sepsis should be differentiated from biliary obstruction of mechanical causes such as primary or metastatic tumors — either intrahepatic or extrahepatic. Obstruction caused by “benign” conditions such as gallstone disease or pancreatitis must be excluded. Metastatic tumors from most solid malignancies and from lymphoma and melanoma have been identified as causes of extrahepatic biliary obstruction.
The initial diagnostic modalities include ultrasound, CT, and in selected patients radionuclide scan. Percutaneous or laparoscopically guided hepatic biopsy may be needed to identify infectious organisms. Endoscopic retrograde cholangiopancreatography or percutaneous transhepatic cholangiography may be useful for defining the level and degree of biliary obstruction. Most patients can be successfully decompressed with endoscopic or combined endoscopic and percutaneous stents.
The preferred approach for biliary decompression in poor-risk patients or those with advanced metastatic disease is with an endoscopically placed endoprosthesis.
If endoscopic stenting is not successful, percutaneous decompression should be performed.
Because nearly half of patients with stents develop stent occlusion (with or without cholangitis), operative biliary decompression has been advocated in selected patients.
Patients expected to survive longer than 6 months or who have chemotherapy-sensitive tumors may benefit from operative approaches. Decompression of the obstructed biliary tree can usually be performed on a semielective basis. Infrequently, suppurative cholangitis complicates malignant biliary obstruction and requires urgent drainage. Acute calculous cholecystitis in a cancer patient generally indicates the need for cholecystectomy. The timing of surgery depends on various factors, including the condition of the patient, response to antibiotics, and recent history of myelosuppressive drugs. For critically ill patients, percutaneous cholecystostomy can be diagnostic and therapeutic or can serve as a temporizing procedure until the patient is able to undergo an operation.
Abscess Postoperative abscess was the most frequent infectious reason for emergency abdominal surgery in patients at
Abscesses were identified as subphrenic, paracolic, pericecal, pancreatic bed, intraloop, and hepatic. The high success rate and low morbidity of percutaneous drainage of intraabdominal abscess makes this the preferred approach. Liver abscesses, especially if limited iumber, may also be treated with percutaneous drainage.
Whether percutaneous drainage should be performed for splenic, pancreatic, or necrotic tumor abscesses remains controversial.
Infected carcinomas, phlegmons, or organized hematomas are infrequently treated adequately by percutaneous techniques. The criteria for percutaneous drainage of intra-abdominal abscesses were recently expanded to include complex and multiloculated abscesses, and it is also used as a temporizing procedure in critically ill patients.
This technique may also be useful in patients with perforated appendiceal, diverticular, or colon cancer abscesses in preparation for a definitive surgical procedure.
The temporizing value of percutaneous catheters is particularly applicable to cancer patients whose condition may be less than optimal for surgery. In all circumstances, however, lack of clinical improvement in 24 to 48 hours should prompt reassessment of the situation and, probably, surgical intervention. Soft Tissue Infection Soft tissue infections such as cellulitis, paronychia, carbuncle, and felon may occur in cancer patients and are managed in standard fashion with antibiotics, warm compresses, and drainage if appropriate. Neutropenic patients deserve close observation because subtle soft tissue infection may rapidly progress to a life-threatening condition that requires extensive surgical debridement. Perianal infections requiring surgical intervention have occurred in 1% to 8% of leukemia patients.
Anorectal complications in patients receiving chemotherapy include cellulitis, abscesses, ulcers, fissures, and thrombosed or bleeding hemorrhoids.
Diarrhea and constipation, common alterations in bowel function, contribute to perianal disease and should be corrected. Preexisting anorectal conditions should be documented before chemotherapy, and digital rectal examination, instrumentation, and enemas should be avoided during treatment.
When symptoms develop, careful perianal inspection and gentle external palpation should be performed and may be facilitated by Valsalva’s maneuver. Fissures or thrombosed hemorrhoids may be evident on examination but rarely if ever require acute surgical intervention unless complicated by uncontrollable hemorrhage. Perirectal infection in febrile and neutropenic patients manifests with intense pain, swelling, and induration. If no fluctuant areas suggestive of abscess can be palpated, therapy is initiated with broad-spectrum antibiotics. Warm compresses, Sitz baths, topical or systemic analgesics, and stool softeners or antidiarrheal agents are beneficial for most perianal conditions. If the initial findings are ambiguous or the patient fails to improve after 24 to 48 hours’ conservative therapy, pelvic CT is helpful in ruling out perianal abscess or an inflammatory process that requires surgical drainage. Persistence of symptoms may necessitate examination under anesthesia, which is much enhanced by the information provided by a CT scan. Infrequently, proximal fecal diversion has been advocated for severe local infection or failure to heal. Suspicious lesions or those that fail to heal should be sampled by biopsy to rule out malignancy.
Radiation therapy was advocated for treating perianal complications in the past but is no longer recommended unless lymphoma can be documented on biopsy. Necrotizing soft tissue infections are rare in cancer patients. Gangrenous cellulitis of the extremity and Fournier’s gangrene of the scrotum and perineum have been described in case reports of patients with hematologic malignancies.
Distant nontraumatic clostridial myonecrosis in patients with colonic or hematologic malignancies is also quite rare.
These entities present initially with pain, swelling, and fever that may not be particularly alarming to the clinician, and then rapidly progress to fulminant tissue necrosis. Early recognition, aggressive surgical debridement, and broad-spectrum antibiotics (including anaerobic coverage) are necessary to reduce the high mortality rate of these infections. The role of hyperbaric oxygen in treating necrotizing infection remains unclear, and randomized studies are needed to determine its usefulness.
RADIATION ENTEROPATHY
The incidence of major gastrointestinal complications following abdominal radiation appears to be decreasing and currently is expected to be less than 5%. Despite this trend, major complications continue to be reported. In a recent study of radiotherapy for rectal carcinoma, 7% (183 patients) required surgery for small bowel obstruction and 31% developed chronic diarrhea.
The most frequent gastrointestinal complications requiring surgery include obstruction (stricture), fistula, perforation, and bleeding.
Acute radiation enteropathy, manifesting as nausea, vomiting, crampy abdominal pain, and watery or bloody diarrhea, is usually self-limited and rarely requires surgical intervention.
Chronic radiation enteritis can present within months to decades but usually manifests 12 to 24 months after completing radiation therapy. Patients experience intermittent abdominal pain, gradual weight loss, and eventually bloody diarrhea and tenesmus. They are best managed symptomatically unless surgical intervention becomes necessary for obstruction, fistula, perforation, or bleeding. Preoperative assessment may include abdominal CT and endoscopy; however, the most helpful — and often mandatory — examinations are radiographic contrast studies from below, above, and through a fistula if present. Surgical interventions to correct complications of radiation have resulted in a high incidence of postoperative anastomotic leaks and fistulas, prompting many surgeons to limit the extent of the operative procedure. For bowel perforation, the operative strategy is to do the least possible surgery.
A proximal diverting ostomy and planned reexploration for subsequent restoration of gastrointestinal continuity may be the safest approach. If patients are stable and the bowel can easily be mobilized, resection may be considered. Since the full extent of bowel radiation injury cannot reliably be determined by visual inspection, anastomosis should ideally be performed to a segment of bowel that is out of the irradiated field. In the setting of acute hemorrhage, a lifesaving resection and diversion is recommended; intestinal continuity can be established later.
Whether to manage obstructed segments of small bowel with resection or bypass has been debated. Extensive adhesiolysis should be avoided in all circumstances because of poor healing and the risk of postoperative perforation. Most authors agree that if resection is performed in the region of the terminal ileum, the procedure should include resection of the right colon and performance of an ileocolic anastomosis. If the small bowel cannot be freed from the pelvis a bypass procedure is preferable. If stenosis or obstruction occurs in the large bowel, diverting colostomy has been advocated.
Bowel cleansing, resection, and anastomosis can be performed later. Fistula management is complex. Detailed radiographic studies should be performed preoperatively and recurrence of cancer ruled out. Initial operations may involve proximal bowel diversion or bypass, with therapeutic intent or in anticipation of preparing the patient for definitive surgery. The goal is resection of the diseased bowel, primary anastomosis, and bringing a new blood supply to the fistula site. At the conclusion of the operation, the anastomosis should not lie at the previous fistula site. Mild cases of radiation proctitis have been managed with steroid enemas and by various conservative means. More severe inflammation, hemorrhage, and tenesmus may be palliated by descending end-colostomy. Laser therapy may control hemorrhage but usually requires multiple treatments.
The gastrointestinal complications of radiation therapy are quite complex. Patients with radiation enteritis are probably best cared for by physicians familiar with this progressive (and at times devastating) complication of cancer treatment. Both diagnosis and therapy of surgical emergencies in cancer patients can be challenging. A careful systematic approach including accurate history taking, thorough physical examination, and judicious use of diagnostic studies can help the physician manage those complex cases. By applying sound surgical principles, and after careful discussion with the entire medical team, the surgeon should be able to determine if and when emergent intervention is needed.
Urologic Emergencies
Urologic emergencies, common in cancer patients, relate mainly to complications of bladder hemorrhage, upper or lower urinary tract obstruction, urinary tract infection, and priapism. The cause, clinical presentation, and management of these emergencies are the focus of this chapter.
BLADDER HEMORRHAGE
Urinary tract hemorrhage can occur in a variety of clinical settings in cancer patients and can rapidly evolve into a life-threatening emergency. Gross hematuria is often the presenting sign of a urinary tract malignancy (kidney, urothelial, prostatic) or may be secondary to direct invasion of colonic and female reproductive system cancers or pelvic sarcomas into the urinary tract. Occasionally, a benign process such as a renal angiomyolipoma or arteriovenous malformation bleeds massively into the urinary tract.
Hemorrhagic cystitis following chemotherapy or radiation therapy or secondary to viral infection in an immunocompromised host causes life-threatening hemorrhage. Disorders of hemostasis due to the systemic manifestations of cancer (i.e., thrombocytopenia or disseminated intravascular coagulation) can cause occult tumors or damaged urothelium to bleed. More than one factor may be responsible for the bleeding.
HEMORRHAGIC CYSTITIS
Fig. Changes associated with irradiation cystitis, which developed after 7200cGy external beam radiation for localized prostate cancer.
Hemorrhagic cystitis, defined as an acute or insidious diffuse bladder inflammation with hemorrhage, can be caused by many toxic agents. Metabolites of chemotherapeutic agents, bladder injury secondary to radiation therapy, and viral infection account for the vast majority of cases of hemorrhagic cystitis in cancer patients. In the early experience using cyclophosphamide as a chemotherapeutic agent or in preparation for bone marrow transplantation, the incidence of hemorrhagic cystitis was as high as 40% to 68%.]
Among bone marrow transplant recipients, massive bladder hemorrhage was associated with a mortality rate as high as 75%.
There are no clinical predictors to indicate which patients will experience this complication. Acute hemorrhage usually occurs during or shortly after treatment, whereas delayed hemorrhage is most common in patients who take oral cyclophosphamide long term.
Acutely, the patient may complain of dysuria and urinary frequency. Microscopic hematuria due to a hyperemic, edematous, and ulcerated bladder mucosa may antedate serious bladder hemorrhage. Of 100 patients treated with cyclophosphamide at the Mayo Clinic who developed hemorrhagic cystitis, bleeding sufficient to require transfusion occurred in 20%. Hemorrhagic cystitis developed after a mean cumulative oral dose of
Liver metabolites of cyclophosphamide and its analogue ifosfamide are phosphoramide mustard and acrolein. Acrolein is the urinary metabolite implicated as the urotoxic substance, but the exact mechanism by which it damages the urothelium is unknown. The greater incidence of hemorrhagic cystitis observed with ifosfamide, as compared with cyclophosphamide, may be due to the preferential excretion of chloroacetaldehyde following ifosfamide metabolism.
Chloroacetaldehyde may also be the metabolite responsible for acute and chronic nephrotoxicity due to direct glomerular, proximal tubular, and distal tubular dysfunction in both children and adults.
Sodium 2-mercaptoethane-sulfonate (mesna) is an effective uroprotectant that does not interfere with the therapeutic efficacy of cyclophosphamide. [ref: 9,10]
Within minutes of intravenous administration, mesna is oxidized in the serum to a stable, inactive disulfide, which is activated in the kidney and binds to urinary acrolein to form an inert thioether. Mesna slows the degradation of the 4-hydroxymetabolites of ifosfamide and cyclophosphamide, further inhibiting their breakdown and the release of acrolein. The only side effects associated with its administration are mild nausea and occasional vomiting.
The serum half-life of mesna is 90 minutes and that of cyclophosphamide, 6 hours. To be effective, mesna must be present in the bladder at the time the acrolein comes into contact with the urothelium. Mesna must be administered before the first dose of cyclophosphamide and continued after the last dose using one of several schedules that employ either continuous intravenous infusion or bolus methods.
Mesna is also capable of detoxifying the nephrotoxic metabolite of ifosfamide, chloroacetaldehyde, but its concentration in the glomerulus and tubules may be insufficient to prevent nephrotoxicity completely. The use of mesna in continuous intravenous infusion may enhance its nephroprotective effects while maintaining its uroprotective ones.
Mesna has dramatically reduced the incidence of hematuria and hemorrhagic cystitis after cyclophosphamide-based chemotherapy (to less than 5%), has reduced cyclophosphamide-induced complications of bone marrow transplantation, and has obviated the chemotherapy dose reductions.
Overhydration to dilute urinary acrolein also is successful in reducing the incidence of major bladder hemorrhage associated with ifosfamide and cyclophosphamide.
In a recently published randomized trial comparing intravenous hydration with mesna to intravenous hydration with a three-way Foley catheter and continuous bladder irrigation with saline, Vose and colleagues reported no difference in episodes of severe hemorrhagic cystitis between the groups, but patients treated with continuous bladder irrigation had a significantly greater incidence of urinary tract infections (27% versus 14%) and bladder spasms or pain (84% versus 2%). These data suggest that the combination of intravenous hydration and continuous infusion mesna is the most effective means of preventing chemotherapy-induced hemorrhagic cystitis.
RADIATION-INDUCED CYSTITIS
Approximately 20% of patients receiving definitive radiation therapy for gynecologic, genitourinary, and rectal cancers experience bladder complications.
Although symptoms of urinary urgency, frequency, and even urinary retention may occur after pelvic irradiation, serious bladder hemorrhage is an unusual acute event. The small vessel injury caused by radiation later leads to interstitial bladder wall fibrosis, reduced bladder capacity, and the formation of friable, telangiectatic blood vessels that course through the bladder mucosa and can spontaneously rupture, leading to massive hemorrhage. Although there are no specific measures to prevent radiation-induced cystitis, acute symptoms of urgency, dysuria, and frequency may be relieved by topical analgesics (phenazopyridine hydrochloride) and antispasmodics (oxybutynin chloride). A period of bladder catheterization may be necessary to relieve symptoms of bladder irritability or treat acute urine retention. Currently under clinical investigation for the treatment of radiation-induced hemorrhagic cystitis are sodium pentosulfanpolysulfate, [ref: 23] hyperbaric oxygen, [ref: 24] and conjugated estrogen, [ref: 25] each of which attempts to stabilize the damaged urothelium and promote healing.
Cystoscopic view of a bladder showing the neovascularity and telangiectasia of radiation cystitis
VIRAL HEMORRHAGIC CYSTITIS
Efforts to understand the late onset of hemorrhagic cystitis in bone marrow recipients focused on a possible viral agent. Although hemorrhagic cystitis secondary to adenovirus type 11 was reported in immunocompetent children, it was detected infrequently in the urine of bone marrow transplant patients. Rice and colleagues identified the BK type of human polyomavirus in the urine of five of six bone marrow recipients, two of whom had hemorrhagic cystitis. In healthy people the BK virus is ubiquitous, persists in the kidney following primary infection, and only occasionally is associated with a mild respiratory illness. The BK virus is activated during periods of immunosuppression (e.g., in organ and marrow transplant patients) and is recoverable from the urine.
A prospective study of 53 bone marrow recipients demonstrated BK virus in the urine of 47%.
Hemorrhagic cystitis of more than 7 days’ duration occurred four times more frequently in patients who excreted BK virus than in those who did not. Urinary BK virus was identified in 55% of patients during episodes of hemorrhagic cystitis but in only 11% of patients who were cystitis free. In a recent report, Bedi and coworkers further defined the association of BK virus with hemorrhagic cystitis in 95 consecutive bone marrow transplant recipients, each of whom had been treated with cyclophosphamide in preparation for transplantation. In 50 of 95 (53%) bone marrow transplant recipients BK virus was detected in the urine, and 38 patients (40%) exhibited persistent BK viruria. Despite prophylaxis with forced hydration or mesna, hemorrhagic cystitis occurred in 19 of 38 patients (50%) with persistent viruria, 1 of 12 (8%) who had a single episode of BK viruria, and ione of the 45 patients who did not shed BK virus. These data strongly implicate the BK virus as the principal etiologic factor in the development of hemorrhagic cystitis in the bone marrow transplant setting, despite adequate uroprotective maneuvers. Persistent BK viruria alone may not be sufficient to induce hemorrhagic cystitis since 50% of patients did not experience this complication. An explanation for this, the investigators postulated, was that hematologic recovery and return to immunocompetence, which often occurs several weeks after transplant and corresponds to the late clinical appearance of hemorrhagic cystitis in these patients, may require competent T lymphocytes to create urothelial damage. This theory may explain the uncommon clinical observation of hemorrhagic cystitis in acquired immunodeficiency syndrome (AIDS) patients, despite the presence of BK viruria.
Strategies are now being formulated to add antiviral agents such as inhibitors of prokaryotic DNA gyrase (e.g., vidarabine) to the uroprotective measures for bone marrow recipients.
TREATMENT.
If bladder hemorrhage is massive and intractable, the patient can develop clot urine retention and complain of severe suprapubic and flank pain. The urologist must reestablish urine outflow quickly by inserting a large-diameter, multiple-hole urethral catheter into the bladder and performing saline lavage and clot evacuation. Often the hemorrhage slows or ceases completely in response to simply removing the clots and decompressing the bladder. When the return lavage fluid is clear or pink tinged, continuous bladder irrigation with a three-way Foley catheter can be effective in removing residual small clots and maintaining free bladder drainage. Bedside lavage is not always effective in evacuating all clots, especially if they have been present for many hours. In this setting continued irrigation may increase intravesical pressure and cause bladder rupture. Instead, the patient should be taken to the operating room for endoscopic clot evacuation under anesthesia. Under direct visualization, the urologist can mechanically disrupt the clots, inspect the bladder for a controllable source of bleeding, and fulgurate any bleeding vessels or tumor. Diffuse bleeding from any cause, despite clot evacuation and fulguration, is an indication for intravesical instillation of a hemostatic agent such as formalin, which acts as chemical cautery to control bleeding from submucosal and mucosal vessels.
In 1969, the use of a 10% formalin instillation was reported effective in controlling intractable bladder hemorrhage in 24 patients treated with radiation therapy for bladder cancer.
Complications of renal papillary necrosis, renal failure, ureteral stenosis, bladder contracture, and bladder rupture with fatal intraperitoneal extravasation tempered the initial enthusiasm for the use of 10% formalin.
In 1974, Fair reported the use of a 1% to 2% formalin solution to treat hemorrhagic cystitis and found it to be effective without the toxicity of 10% formalin.
Complications of formalin appear infrequently when concentrations of 4% or less are used, and for the vast majority of patients with hemorrhagic cystitis (more than 75%) bleeding is thus controlled.
The following technique of formalin instillation currently is used at our institution:
1. Spinal or general anesthesia is required because formalin is caustic to sensory nerves in the bladder. Cystoendoscopy is performed to evacuate clots and fulgurate bleeding vessels.
3. Formalin and formaldehyde are not equivalent compounds. Formalin is a 37% solution of formaldehyde. The stock solution is diluted with sterile water to the desired concentration as follows: 10% formalin is 3.7% formaldehyde and 1% formalin is a 0.37% formaldehyde solution. The order to the pharmacy for the preparation of the desired formalin solution must be written explicitly and the solutions checked by the operating team to avoid serious error.
4. An
5.After the formalin instillation, the bladder is irrigated thoroughly with at least
Other methods of controlling bleeding of hemorrhagic cystitis, including intravesical alum, prostaglandins, phenol, intravesical silver nitrate, hydrostatic distention, iced saline lavage, and parenteral or oral aminocaproic acid, have been utilized with varying degrees of success.
Of these agents, intravesical 1% alum and prostaglandin E(2) and F(2), each of which can be instilled without anesthesia, are the most effective. A 1% alum solution (i.e., the ammonium or potassium salt dissolved in water) delivered by continuous bladder irrigation causes protein precipitation, vasoconstriction, and decreased capillary permeability.
During instillation, bladder spasms may occur, but they are usually controlled well by antispasmodic medication. The principal advantage of alum in the treatment of hemorrhagic cystitis is that it can be delivered without anesthesia — at the bedside using a three-way indwelling catheter. Some urologists recommend initial use of alum for massive bladder hemorrhage and formalin as second-line treatment. A significant disadvantage of alum irrigation is that it may take as long as 7 days to control the bleeding effectively. Serum aluminum levels should be monitored, as rare cases of alum-induced encephalopathy have been reported.
Prostaglandin E(2) and F(2) can also be instilled intravesically without anesthesia and have been reported to be effective in controlling intractable hemorrhagic cystitis, possibly by protecting the microvasculature and epithelium and inhibiting development of tissue edema.
In a recent report Carboprost tromethamine, a commercially available prostaglandin F(2alpha) that can be administered at the bedside, controlled severe bladder hemorrhage in 9 of 18 bone marrow transplant patients.
Severe prostaglandin-induced bladder spasm may limit their overall utility. Despite all conservative efforts, persistent exsanguinating hemorrhage from the bladder may necessitate open surgical attempts to control the bleeding — by open cystotomy with bladder packing, cutaneous ureterostomy, hypogastric artery embolization, or cystectomy with urinary diversion.
Surgical intervention in these already critically ill patients should be performed only when conservative measures have failed. Surgical control of hemorrhage may still not be possible.
OBSTRUCTIVE UROPATHY UPPER TRACT OBSTRUCTION.
Obstruction of one or both ureters in cancer patients may be secondary to direct tumor invasion, compression of the ureter by a retroperitoneal tumor, encasement of the ureter by retroperitoneal or pelvic lymph nodes involved with metastatic disease, or rarely, by direct metastases to the ureter. Seventy percent of the tumors that cause ureteral obstruction are genitourinary (cervical, bladder, prostate) in origin; breast and gastrointestinal lesions and lymphoma account for most of the remainder. Ureteral obstruction may also be secondary to retroperitoneal fibrosis following combinations of surgery, chemotherapy, and pelvic radiation therapy.
In a recent review of 1784 patients with stage IB carcinoma of the cervix treated with external beam irradiation at M.D. Anderson Cancer Center, McIntyre and colleagues described the late complication of severe ureteral stricture which occurred in 29 patients, 4 of whom died of associated complications. The authors calculated the continuous actuarial risk of developing a severe stricture at 0.15% per year; thus, 2.5% of patients who survive 20 years experience this complication. Despite the patient’s achieving a disease-free status, they recommend long-term surveillance for late ureteral obstruction.
Acute ureteral obstruction usually causes flank pain and colic typical of the symptoms of urolithiasis. In contrast, chronic unilateral obstruction is usually a silent event that often is detected incidentally as hydronephrosis with renal cortical atrophy on upper abdominal imaging studies.
Bilateral ureteral obstruction, acute or chronic, is associated with decreased urine output and symptoms of uremia. In one study of 50 patients with acute renal failure secondary to bilateral ureteral obstruction, for 76% malignant disease was the cause of the obstruction and half of these had uremia as the presenting sign of their cancer. Increased renal pelvic pressure may cause rupture of the renal calyceal fornix, leading to extravasation of urine into the renal sinus with the formation of a perinephric urinoma.
Rarely, acute uremia can occur in the absence of imaging studies consistent with ureteral obstruction. If clinical suspicions warrant, ureteral decompression should still be undertaken. This often is followed by clinical diuresis and normalization of renal function.
If infected urine exists in the obstructed system, fever, chills, and eventual urosepsis may ensue, requiring emergency decompression or percutaneous drainage. The diagnosis of ureteral obstruction may be made by intravenous urography, computed tomography (CT), renal ultrasonography, retrograde pyelography, or radionuclide renography.
Of these, abdominal CT has the advantage of better defining any extrarenal lesion that might account for the obstruction. A difficult ethical question arises with the diagnosis of ureteral obstruction in a patient with an incurable malignancy. Is decompression going to facilitate treatment with chemotherapy and palliate symptoms or will it merely address a short-term problem and prolong the patient’s suffering? Palliative urinary diversion or decompression with ureteral stents is justified if improvement in renal function will facilitate the use of chemotherapy or if symptoms of ureteral obstruction (i.e., pain or urosepsis) can be alleviated. Previously, malignant ureteral obstruction was managed by open surgical placement of a nephrostomy tube, a procedure associated with a major complication rate of up to 45%.
Approximately 25% of patients undergoing opeephrostomy died within 30 days of operation, and the average survival time after the procedure was approximately 6 months. Opeephrostomy has beeearly completely replaced by CT or ultrasound-guided percutaneous nephrostomy, with and without internal stent placement.
In the last 15 years, techniques have evolved utilizing both percutaneous and cystoscopic access to the kidney to decompress the obstructed urinary tract. Using these “endourologic” techniques, decompression of obstructed kidneys, removal of obstructing ureteral stones, and dilation of benign and malignant ureteral strictures can be performed without an open operation. The percutaneous tubes and stents are composed of flexible synthetic materials (i.e., polyurethane), which minimize migration in the urinary tract owing to the double-J configuration of the stents .
Recently reported preliminary experience with metallic, self-expanding stents, used alone or in conjunction with double-J stents for malignant ureteral obstruction, confirms their value in maintaining ureteral patency and avoiding percutaneous nephrostomy.
Reports of endourologic decompression of ureteral obstruction in prostate cancer, female reproductive system tumors, colorectal cancer, and breast cancer confirm the value of these techniques.
Flexible guidewires caegotiate narrowed ureteral lumens and sharp bends in the obstructed ureter and aid in the placement of the stents during retrograde cystoscopic or antegrade percutaneous procedures. Endourologists can also incise or balloon dilate and stent benign ureteral strictures or ureteral enteric stenoses after urinary diversion.
In patients with bilateral ureteral obstruction, significant palliation and return of nearly normal renal function are possible following decompression of the obstructed kidney with more substantial remaining cortex as determined by CT or ultrasound examination. Once a ureteral stent or percutaneous nephrostomy is placed, subsequent management requires careful follow-up, including periodic monitoring of renal function, upper tract imaging (CT, renal ultrasound), frequent urine cultures, and stent or tube replacement every 4 to 6 months. If chemotherapy or radiation therapy directed at the primary tumor successfully eliminates the lesion responsible for the ureteral obstruction (i.e., retroperitoneal adenopathy) from breast cancer or leukemic infiltrates of the ureter, then the percutaneous nephrostomy or ureteral stents can be removed, provided imaging studies document restoration of normal urine drainage from the involved kidney. Locally advanced prostate cancer with ureteral obstruction in hormone-naive patients can be treated with emergent total androgen blockade or bilateral orchiectomy, thus obviating long-term ureteral decompression.
Recent reports of short-term high-dose corticosteroid for malignant ureteral obstruction suggest that, either by reducing edema around the tumor or by antitumor effect, formal ureteral decompression procedures may be obviated. For patients with advanced cancer treated with steroids the risk of gastrointestinal bleeding or perforation may make this approach hazardous. Despite the rapid evolution of endourologic techniques, information confirming morbidity associated with the ureteral decompression devices, number of related hospitalizations, patient survival from the time of decompression, and quality of life concerns had not been properly addressed in the literature. Donat and Russo from Memorial Sloan-Kettering Cancer Center examined the outcome of 78 patients referred for ureteral decompression who had locally advanced or metastatic “non urologic” tumors: 68% underwent successful cystoscopic decompression, and 32% required percutaneous nephrostomy. Half of the patients had at least one complication related to the ureteral decompression device, and 67% of all complication events required hospitalization for treatment. The most common complications included stent obstruction, infection, and migration. The mediaumber of hospitalizations per patient for management of the decompression device was eight. Median survival for all patients following the first decompression procedure was 6.8 months; actuarial survival at 1 year was 55% and at 3 years, 30%. Patients with gastric or pancreatic cancer survived a median of 1.4 months (actuarial 1-year survival, 12.5% and 3 year survival zero); these figures were significantly worse than those for patients with colorectal, cervical, sarcoma, lymphoma, or ovarian primary lesions. Given the poor overall survival for metastatic gastric and pancreatic cancer patients, ureteral decompression should not be attempted. In this series, only two patients had effective systemic or regional therapy that allowed their devices to be removed. This experience with malignant ureteral obstruction more clearly defines physician and patient expectations.
Complications of endourologic procedures include gross hematuria with perinephric hematoma, splenic or bowel injury, and hemo- or pneumothorax. Indwelling ureteral stents can also migrate, obstruct with proteinaceous encrustations, become infected, fragment, cause uncomfortable vesicoureteral reflux, and erode through the urinary tract.
Patients may experience bladder spasm from irritation of the trigone, which generally subsides within several weeks of stent placement. Today, opeephrostomy, cutaneous ureterostomy, or operative urinary diversion should be reserved for the unusual case in which endourologic techniques are not successful and the patient is in satisfactory medical condition and has a life expectancy that does not preclude a major surgical procedure. Major operative reconstruction (revision of ileal-ureteral anastomosis, ureteral reimplantation, ileal-ureter interposition) may also be considered in this restricted clinical setting.
BLADDER OUTLET OBSTRUCTION AND URINE RETENTION.
Acute urine retention and bladder outlet obstruction (hesitancy, dribbling, incomplete bladder emptying, overflow incontinence, decrease in the force of the voided stream) can be caused by mechanical or neurophysiologic factors. These factors may be primary manifestations of the malignancy, secondary to treatment, or due to preexisting benign conditions involving the lower urinary tract. A patient with urine retention complains of severe suprapubic pain and has palpable suprapubic fullness secondary to a distended bladder. Pelvic and rectal examination may reveal a genitourinary or rectal tumor that is obstructing the bladder outlet. Complete bladder outlet obstruction can lead to bilateral hydroureteronephrosis with renal insufficiency and should be treated emergently. If a malignant cause of bladder outlet obstruction is excluded, a preexisting benign mechanical condition such as benign prostatic hyperplasia (BPH) or urethral stricture, may exist. Other causes of retention include constipation and neurologic disorders.
It is estimated that 5% to 10% of men at age 40 years and 80% of men at age 80 have BPH, although symptoms vary much and do not correlate directly with prostate size by digital rectal exam.
Many patients have significant obstructive symptoms and may compensate by limiting their fluid intake. During chemotherapy, the combination of antiemetic medication and hydration may precipitate urine retention in patients with preexisting prostatism. Improvements in our understanding of the neurophysiology of micturition have allowed us to diagnose more precisely and treat acute nonmechanical causes of bladder outlet obstruction. Urodynamic evaluation, including uroflow, cystometrography, and perineal sphincter electromyography, are useful in assessing the micturition reflex, which includes sensory nerves in the bladder wall, voluntary cerebral control over the pontine micturition center, sacral parasympathetic nerves (S2-3), cholinergic nerves to the external sphincter, and sympathetic nerves to the bladder neck. Anatomic or neuropharmacologic interference with the micturition reflex occurs frequently in cancer patients and can lead to acute urine retention.
For example, postoperative urine retention, often attributed simply to prostatism, can be caused by
(1) the use during anesthesia of atropine and other anticholinergic agents, leading to depressed detrusor contractility,
(2) pain and stress-induced sympathetic activity leading to increased tone in the bladder neck and proximal urethral muscles,
and (3) perioperative pain medications (i.e., opiates), which can directly inhibit the pontine micturitional center and suppress the urge to void.
Tumors arising in or metastasizing to the brain and or spinal cord can directly interfere with central voluntary control of micturition, coordination of bladder emptying, and sphincter muscle relaxation. This condition, termed detrusor external sphincter dyssynergia, can lead to functional bladder outlet obstruction characterized by bladder wall hypertonicity and hypertrophy with reduced capacity, vesicoureteral reflux, and incomplete bladder emptying. Hypotonic bladder dysfunction may also occur in as many as 50% of patients undergoing abdominoperineal resection for rectal carcinoma or radical hysterectomy: extensive pelvic dissection can disrupt the pelvic parasympathetic nerves necessary for normal detrusor muscle contraction. This effect is permanent in only 10% of patients. In an immunocompromised host viral radiculomyelitis secondary to herpes simplex can also cause a flaccid neurogenic bladder, usually a self-limited process, with the return of normal bladder function expected in approximately 2 weeks.
The first step in the management of urine retention is the passage of a small (
A pathologic response can occur in patients with more extensive renal tubule damage, leading to nephrogenic diabetes insipidus. In these patients, excessive water loss continues even after the retained water and solutes have been eliminated. The urine remains dilute with low osmolality, and administration of antidiuretic hormone is ineffective. Normally, a conscious patient retains a normal thirst mechanism and is able to avoid dehydration until renal tubule function normalizes. Rarely, salt-losing nephropathy occurs, which leads to hypotension and hyponatremia. Close monitoring and fluid resuscitation with normal or hypertonic saline must be undertaken until tubule functioormalizes.
The treatment approach toward the cancer patient in urinary retention should be conservative. Patients are often debilitated by the malignancy or its treatment and are frustrated anew by the inability to urinate. Many patients, when given another chance to void after a period of catheter decompression, are able to urinate normally and may be spared invasive procedures.
Once bladder decompression is accomplished, timing of another voiding trial should be based on the patient’s clinical status. In the postoperative patient with urinary retention, it is prudent to wait for recovery from the operation, a decrease in the requirement for pain medication, and the ability to ambulate before another voiding trial is initiated. For a patient experiencing urinary retention during chemotherapy, the voiding trial should be delayed until the period of hydration is passed and the need for antiemetic medication is lessened. There is no place for bladder training (i.e., clamping and unclamping of the catheter to give the bladder more tone), which was once advocated by some for retention. During periods of bladder catheterization, urine cultures should be obtained every 3 days. If colonization of the urine occurs, an appropriate antibiotic to which the organism is sensitive can be administered at the time of the voiding trial. Continuous use of antibiotics in a patient with an indwelling Foley catheter should be avoided, since this leads to the emergence of resistant organisms in the urine. If the colonizing organism produces urease (i.e., Proteus mirabilis), the urine can become alkaline and magnesium ammonium phosphate (struvite) calculi can form. To prevent this, a short course of an appropriate antibiotic should be administered.
The male patient with a significant antecedent history of prostatism may remain in urinary retention despite voiding trials. Imaging of the upper urinary tract with intravenous urography, ultrasonography, or CT can detect hydronephrosis or renal cortical atrophy.
Ultrasonography has emerged as the most cost-effective means of evaluating the urinary tract and has the added appeal of avoiding the use of iodinated contrast materials, which occasionally cause nephrotoxicity. Most patients’ bladder outlet obstruction can be relieved by transurethral resection of the prostate (TURP) with acceptable morbidity, although complications of bleeding, clot retention, infection, and persistent failure to void make careful case selection important in the cancer patient population. Alternatives to prostatectomy include luteinizing hormone-releasing hormone analogues, inhibitors of 5-alpha-reductase and androgen receptors, alpha-adrenergic receptor blockers, urethral stents, local microwave hyperthermia, and balloon dilation. Although certain patients may benefit from these techniques, alone or in combination, in the short term their overall efficacy appears to be less than that of standard prostatectomy and long-term follow-up data are not available. In patients with locally advanced prostatic adenocarcinoma who present with severe bladder outlet obstructive symptoms or urinary retention, initiation of total androgen blockade (e.g., luteinizing hormone-releasing hormone analogues, alone or in combination with an antiandrogen) can reduce prostatic volume enough so that transurethral resection may be avoided.
While the androgen blockade is working, intermittent catheterization every 4 to 6 hours, Foley catheterization, or an internal prostatic stent (recently described) may be used to achieve bladder drainage. The use of intraprostatic stents has also been used instead of TURP in cases of hormone-refractory prostate cancer, in an attempt to limit morbidity.
Although the intraprostatic stents are effective in alleviating bladder outlet obstruction, patients with stents may be at increased risk of developing urinary tract infection and require frequent monitoring with urine culture. When cause of urinary retention is neurologic, the mainstay of treatment is clean intermittent self-catheterization every 4 to 6 hours, a technique that was introduced nearly 20 years ago by Lapides and colleagues.
This method of alleviating urine retention can preserve kidney function, reduce infection rates, reduce stone formation, and improve the quality of patients’ lives as compared with chronic indwelling Foley catheters. In a recent report of 75 patients with neurogenic bladders followed for a mean interval of 7 years, all patients’ upper urinary tract function was preserved. The complications (20% of cases) included urethral injury or stenosis (13 cases), epididymitis (6 cases), bladder calculi (3 cases), and pyelonephritis (1 case). Prophylactic antibiotics are not given to patients on intermittent self-catheterization since they have not been shown to lower the incidence of clinical urinary tract infection.
As a last resort for terminally ill patients with marginal performance status, a permanent indwelling Foley catheter is justified. With strict attention to catheter care, morbidity can be limited and patient comfort enhanced.
