66_04 sequele

Sequelae of Treatment

Irradiation Alone

Acute Sequelae

Descriptions of sequelae vary among institutions because toxicity grading scales are not uniform and the scoring system for complications is not clearly stated in all reports.

Dusenbery et al. (132) reported 21 (6.4%) life-threatening complications in 327/462 patient implants.

Lanciano et al. (348), in 95 tandem and ovoid insertions for cervical cancer in 91 patients and for endometrial cancer in four, observed two uterine perforations and a vaginal laceration in two patients. Twenty-four percent of implants in 16 patients were associated with temperatures higher than 100.5°F. Five implants (5%) were removed because of presumed sepsis, pulmonary disease, arterial hypotension, change in mental status, and myocardial infarction.

Jhingran and Eifel (282), in 4,043 patients with carcinoma of the cervix who had undergone 7,662 intracavitary procedures, observed 11 (0.3%) documented or suspected thromboembolism, resulting in four deaths; the incidence of postimplant thromboembolism did not decrease significantly with the routine use of minidose heparin prophylaxis. Other life-threatening perioperative complications included myocardial infarction (one death in five patients), cerebrovascular accident (two patients), congestive heart failure (three patients), and halothane liver toxicity (two deaths). Intraoperative complications included uterine perforation (2.8%) and vaginal laceration (0.3%), which occurred more frequently in patients 60 years of age or older (p <0.01).

Wollschlaeger et al. (659) reported morbidity during hospitalization in 128 patients with cervical carcinoma undergoing 110 LDR intracavitary brachytherapy insertions. Forty-two implants (24.7%) were associated with acute problems; the most common were fever/infection (14.1%) or gastrointestinal problems (5.9%).

Acute gastrointestinal side effects of pelvic irradiation include diarrhea, abdominal cramping, rectal discomfort, and occasionally rectal bleeding, which may be caused by transient enteroproctitis. Patients with hemorrhoids may experience discomfort earlier than other patients. Diarrhea and abdominal cramping can be controlled with the oral administration of diphenoxylate hydrochloride, with loperamide, atropine sulfate, or opium preparations or emollients such as kaolin and pectin. Proctitis and rectal discomfort can be alleviated by small enemas with hydrocortisone and anti-inflammatory suppositories containing bismuth, benzyl benzoate, zinc oxide, or Peruvian balsam. Some suppositories contain cortisone. Small enemas with cod liver oil are also effective. A low-residue diet with no grease or spices and increased fiber in the stool (psyllium, polycarbophil) usually help to decrease gastrointestinal symptoms.

Genitourinary symptoms, secondary to cystourethritis, are dysuria, frequency, and nocturia. The urine is usually clear, although there may be microscopic or even gross hematuria. Methenamine mandelate and antispasmodics such as phenazopyridine hydrochloride or a smooth muscle antispasmodic such as flavoxate hydrochloride, hyoscyamine sulfate, oxybutynin chloride, or tolterodine tartrate can relieve symptoms. Fluid intake should be at least 2,000 to 2,500 mL daily. Urinary tract infections may occur; diagnosis should be established with appropriate urine culture studies, including sensitivity to sulfonamides and antibiotics. Therapy should be promptly instituted.

Erythema and dry or moist desquamation may develop in the perineum or intergluteal fold. Proper skin hygiene and topical application of petroleum jelly, petrolatum, or lanolin should relieve these symptoms. U.S.P. zinc oxide ointment and intensive skin care may be needed for severe cases.

Management of acute radiation vaginitis includes douching every day or at least three times weekly with a mixture of 1:5 hydrogen peroxide and water. Douching should be continued on a weekly basis until the mucositis has resolved or for 2 or 3 months as necessary. Superficial ulceration of the vagina responds to topical (intravaginal) estrogen creams, which stimulate epithelial regeneration within 3 months after irradiation. Use of vaginal dilators several times daily, started during the course of treatment, prevents vaginal stenosis.

Psychoeducational intervention and motivation improve the compliance in use of dilators (278). More severe necrosis may require debridement on a weekly basis until healing takes place. Judicious use of biopsies is recommended to rule out persistent or recurrent cancer.

Late Sequelae

The incidence of major late sequelae of radiation therapy for stage I and IIA carcinoma of the cervix ranges from 3% to 5%, and for stage IIB and III, between 10% and 15%. The most frequent major sequelae for the various stages are listed in Tables 66.25 and 66.26.

Injury to the gastrointestinal tract usually appears within the first 2 years after radiation therapy, whereas complications of the urinary tract are seen more frequently 3 to 5 years after treatment (327,469).

Pedersen et al. (467), in a review of morbidity of radiation therapy in 442 patients with cervical cancer stages IIB, III, and IVA, recommended that actuarial estimates rather than frequency of sequelae be reported to avoid underestimation of risks of late morbidity after radiation therapy in long-term survivors.

In fact, Eifel et al. (140), in 1,784 patients with stage IB carcinoma of the cervix, noted that the greatest risk of sequelae is in the first 3 years after therapy. The risk of rectal complications declined after the first 2 years of follow-up to 0.6% per year, whereas the risk of major urinary tract complications for survivors continued at 0.3% per year, with a 20-year actuarial risk of major complications of 14.4%.

Montana et al. (417), Perez et al. (475), and Pourquier et al. (500) noted a greater incidence of complications with higher doses of irradiation.

Perez et al. (469) and Pourquier et al. (500)

P.1584

reported that with doses below 75 to 80 Gy delivered to limited volumes, grade 2 and 3 complications in the urinary tract and rectosigmoid were approximately 5%. However, the incidence increased to over 10% with higher doses of irradiation to these organs (Fig. 66.32). Doses higher than 60 Gy were also correlated with a greater incidence of small bowel injury (Fig. 66.33).

The same analysis showed that patients who experienced sequelae of therapy had slightly better survival rates than patients without any complications. This was related to improved tumor control with higher doses of irradiation (469).

Perez et al. (475) quantitated the impact of total doses of irradiation, dose rate, and ratio of doses to bladder or rectum and point A on sequelae in 1,456 patients treated for cervical cancer with external-beam irradiation plus two LDR intracavitary insertions to deliver 70 to 90 Gy to point A. Median follow-up was 11 years. In stage IB, the frequency of grade 2 morbidity was 9%, and grade 3, 5%; in stages IIA, IIB, III, and IVA, the frequency of grade 2 morbidity was 10% to 12%, and of grade 3, 10%. The most frequent grade 2 urinary/rectal sequelae were cystitis and proctitis (0.7% to 3%). The most common grade 3 sequelae were vesicovaginal fistula (0.6% to 2% in patients with stage I to III tumors), rectovaginal fistula (0.8% to 3%), and intestinal obstruction (0.8% to 4%). In the bladder, doses below 80 Gy correlated with a <3% incidence of morbidity and 5% with higher doses (p = 0.31). In the rectosigmoid, the incidence of significant morbidity was <4% with doses below 75 Gy and increased to 9% with higher doses. For the small intestine, the incidence of morbidity was <1% with 50 Gy or less, 2% with 50 to 60 Gy, and 5% with higher doses to the lateral pelvic wall (p = 0.04). Multivariate analysis showed that dose to the rectal point was the only factor influencing rectosigmoid sequelae, and dose to the bladder point affected bladder morbidity (Table 66.25).

In a review of the Patterns of Care Study, Lanciano et al. (349) observed a 5-year actuarial rate of 14% for major late complications in 1,558 patients treated with irradiation for invasive carcinoma of the cervix. Women younger than 40 years of age or with a history of prior surgery or laparotomy for staging had a greater incidence of significant morbidity (15% to 18% vs. 8% to 9%). Also, EBRT dose per fraction >2 Gy, paracentral doses of 85 Gy or greater, and lateral parametrial doses higher than 60 Gy were independently associated with a higher complication rate.

Lee et al. (369), using 3 Gy fractions with EBRT, calculated the rectal point dose in the anterior wall at the level of the cervical os and noted that total higher BED (142.7 Gy) were associated with rectal sequelae compared with BED <131 Gy.

Mitchell et al. (413) evaluated 398 patients with stage I to III cervical carcinoma treated with radiation therapy. Patients were divided into nonelderly (35 to 69 years of age; n = 338) and elderly (≥70 years of age; n = 60) groups. The frequency and severity of acute and chronic sequelae were equivalent in both groups.

When late radiation proctitis occurs, initial treatment is the same as for acute proctitis. If the symptoms and rectal bleeding persist, laser treatment of rectal telangiectasis or ulcers is frequently beneficial.

Roche et al. (515) treated six patients with hemorrhagic radiation-induced proctitis using outpatient intrarectal application of formaldehyde 4%. In four cases the bleeding ceased after the first formaldehyde application; two patients continued to bleed, but another application 3 weeks later definitively controlled the hemorrhage. There were no complications, such as burns or late stenosis of the deep layers of the rectum, and this technique was well tolerated.

Rubinstein et al. (531) and Seow-Choen et al. (550) also reported

P.1585

treatment of radiation proctitis with a similar technique. Patients are sedated, a local anesthetic block is administered, and a sponge moistened with 4% formalin is applied for 4 minutes to each bleeding area of the rectum. Care is taken to protect the perianal skin from any caustic effects of the formalin.

Occasionally, a colostomy is necessary if conservation management fails. The importance of performing colonoscopy in patients with rectal bleeding, to exclude other lesions in the colon, including polyps or cancer, is emphasized.

Anal incontinence is occasionally observed. This sequela must be assessed in light of a report by Nelson et al. (438), who in a survey of 6,959 nonirradiated patients identified 153 (2.2%) who reported anal incontinence, without specific etiology. Thirty percent of incontinent subjects were older than 65 years of age, and 63% were women. Of those with anal incontinence, 36% were incontinent to solid feces, 54% to liquid feces, and 60% to gas.

Kim et al. (317) investigated the effects of radiation on anorectal function using manometry in 24 patients with carcinoma of the uterine cervix who had late radiation proctitis. These data were compared with those from 24 age-matched nonirradiated female volunteers. Regardless of the severity of proctitis symptoms, 75% of irradiated patients exhibited abnormal manometric parameters for sensory or motor functions. Radiation damage to nerves and to the external sphincter muscle were considered to be an important cause of motor dysfunction.

Quilty (504) noted a greater incidence of pelvic complications in patients treated with higher doses to the whole pelvis (40 to 50 Gy). The author commented that the intracavitary radium dose was not correlated with severe complications.

Similar observations were made by Stryker et al. (584) in 132 patients, who recorded a 9% incidence of fistulas and a 14% incidence of grade 2 and 3 complications after delivery of 50 Gy or higher to the whole pelvis (1.8-Gy daily dose) combined with intracavitary insertion. They recommended that the whole pelvis dose should not exceed 40 to 45 Gy when doses of approximately 40 Gy are delivered to point A with LDR intracavitary insertions.

Kuske et al. (344), comparing results of therapy in 99 patients with carcinoma of the cervix on whom minicolpostats were used, noted a 15% higher incidence of grade 2 and 3 complications, which was higher than the 8% incidence noted in a similar group of patients treated with regular colpostats during the same period (p = 0.08).

Perez et al. (469) reported that the incidence and type of complications with interstitial therapy at Washington University were approximately the same as in patients treated with intracavitary technique only.

On the contrary, Kashibhatla et al. (300) noted 6% small bowel obstruction in 36 women with gynecological cancer treated with EBRT and interstitial brachytherapy, which was aggravated by previous abdominopelvic surgery. The 3-year risk of rectovaginal fistula was 18%, and it was significantly higher in patients who received total doses >76 Gy (100% vs. 7%; p = 0.009).

Irradiation of the para-aortic lymph nodes has been reported to cause increased morbidity, particularly if it is done after transperitoneal staging para-aortic lymphadenectomy.

In a randomized study reported by Rotman et al. (527), a somewhat higher incidence of grade 2 and 3 complications was reported in 170 patients (10 complications) given 45 Gy to the para-aortic area in addition to standard pelvic irradiation, compared with five complications in 167 patients treated by pelvic irradiation only. The incidence of fatal (grade 5) complications was four and one, respectively. In a similar randomized

P.1586

P.1587

study by Haie et al. and the EORTC (219), the incidence of grade 3 small bowel injury was 2.3% in the para-aortic irradiation group and 0.9% in the pelvic irradiation–only group. The overall incidences of severe complications were 9% and 4.8%, respectively.

Although it is common knowledge in the radiation oncology community, only a few sketchy reports have been published on the possible intolerance to pelvic irradiation in patients with inflammatory bowel disease (134,252).

Willett et al. (655) reported on 28 patients with inflammatory bowel disease (10 with Crohn's disease, 18 with ulcerative colitis) who underwent external-beam abdominal or pelvic irradiation. Patients were treated either by specialized techniques (16 patients) to minimize small and large bowel irradiation or by conventional approaches (12 patients). The overall incidence of severe toxicity was 46% (13/28 patients), and six (21%) experienced severe acute toxicity necessitating cessation of radiation therapy. Late toxicity requiring hospitalization or surgical intervention was observed in 8/28 patients (29%). For patients treated with conventional approaches, the 5-year actuarial rate of late toxicity was 73% versus 23% for patients treated by specialized techniques (p = 0.02). In patients with inflammatory bowel disease undergoing abdominal or pelvic irradiation, judicious use of this modality must be observed.

On the contrary, Song et al. (567), in a review of 24 patients with a history of inflammatory bowel disease who received RT (median dose of 45 Gy in 1.8 to 3 Gy fractions) to fields encompassing some portion of the gastrointestinal tract, noted that five patients (21%) experienced acute intestinal toxicity of grade 3 or greater and two (8%) had grade 3 or greater late intestinal toxicity. Fifteen patients also received concurrent chemotherapy. The authors believed that the gastrointestinal toxicity in these patients was more modest than generally perceived.

Also, Tiersten and Saltz (612) noted that five patients with inflammatory bowel disease and gastrointestinal malignancy completed planned radiation therapy (30 to 54 Gy), usually with concurrent 5FU, without difficulty.

Salama et al. (542) reported preliminary observations on acute toxicity with extended field IMRT in 13 patients with gynecological cancer. With median follow-up of 11 months, two patients treated with chemoradiation experienced grade 3 or higher morbidity and one (with history of previous surgeries) developed small bowel obstruction.

Levenback et al. (374) identified 116/1,784 patients (6.5%) with stage IB carcinoma of the cervix treated with irradiation in whom hemorrhagic cystitis developed, 23% grade 2 (repeated minor bleeding), and 18% grade 3 (hospitalization required for medical management). The median interval to onset of hematuria was 35.5 months. The risk of severe hematuria requiring surgical intervention was 1.4% at 10 years and 2.3% at 20 years. Minor episodes of hematuria are managed by antibiotic therapy. Cystoscopic, laser, or cautery treatment of bleeding points is indicated. Clot evacuation and continuous bladder irrigation are important elements in the acute management of patients with heavy bleeding. Occasionally, a urinary diversion is required for intractable severe hematuria.

In 13 patients with hemorrhagic cystitis treated with hyperbaric oxygen, all but one experienced durable cessation of hematuria (647).

Lee et al. (364) also noted that, in 16/20 patients (80%) with hemorrhagic radiation cystitis, significant improvement was observed after treatment with hyperbaric oxygen at 2.5 atm (44 sessions).

Ureteral stricture at 20 years was observed in 2.5% of 1,784 patients with stage IB carcinoma of the cervix treated with irradiation (274 followed for up to 20 years or longer) (406). The most common presenting symptoms were flank pain and urinary tract infection. In five patients, ureteral stricture was complicated by a vesicovaginal fistula. Seven of 43 patients who had no evidence of cancer and hydroureter or hydronephrosis died of radiation complications. Treatment of ureteral stenosis may consist of stenting or resection of the fibrotic segment and reimplantation of the ureter either with a ureteroneocystostomy or ureteroileocystostomy. In approximately half of the patients, diversion of urinary stream and ileal conduits are necessary. Occasionally, a nephrectomy is performed for removal of a nonfunctional kidney.

Buglione et al. (60) reported a 10% incidence of late urinary morbidity and 1% ureteral fibrosis grade III or IV in 191 patients. They postulated the role of TGF-β1 factor in the activation of fibroblasts and remodeling of extracellular matrix, which may be important in the induction of these sequelae.

Patients with gynecologic malignancies, including those receiving radiation therapy, are prone to development of urinary tract infections.

Prasad et al. (501) collected 216 urine samples from 36 patients receiving pelvic irradiation, 12 of whom had urinary tract infection. The most common organisms isolated were Escherichia coli, followed by Enterococcus species. Appropriate urine bacterial studies and cultures are indicated in patients suspected of having superimposed urinary tract infection during the course of radiation therapy.

Parkin et al. (462) reported a 26% incidence of severe urinary symptoms (urgency, incontinence, and frequency) in patients treated with irradiation alone for cervical carcinoma. They carried out urodynamic studies in 42 women and compared them with 28 women having urodynamic evaluations before and after treatment. There was no difference in the mean maximum flow rate or mean residual volume in the two groups. However, mean volume of full bladder sensation was significantly lower in the postirradiation group than in the pretreatment group, as was the mean maximum cystometric capacity.

This same dysfunction may be noted in approximately 10% of the general female population, and the incidence increases in older women (478).

Ureteroarterial fistula is a rare occurrence, and it is associated with a high mortality rate. When profuse urinary tract bleeding occurs in patients previously diagnosed with a gynecologic malignancy and treated with radiation therapy and extensive surgery, ureteroarterial fistula should be considered in the differential diagnoses (118).

Although extremely rare, lumbosacral plexopathy has been occasionally reported in patients treated for pelvic tumors with doses of 60 to 67.5 Gy. At Washington University, this syndrome was observed in four of 2,410 patients with cervical or endometrial carcinoma receiving 45 Gy to the para-aortic lymph nodes (without spinal cord shielding) or external pelvic irradiation (60 Gy to the parametria) and brachytherapy, with the lumbosacral plexus receiving total doses of 70 to 79 Gy (176).

Lower extremity paralysis secondary to lumbosacral plexopathy was reported in one patient after standard radiation therapy for cervical cancer (3).

Patients previously reported as having radiation myelopathy to the lumbar spine may have suffered a lumbar and sacral nerve plexopathy instead of or in addition to the spinal cord injury. The differential diagnosis of plexopathy with recurrent tumors is sometimes difficult. In a comparison of 20 patients with lumbosacral plexopathy after irradiation and 30 patients with plexus damage from pelvic malignancy,

Thomas et al. (610) noted that indolent leg weakness occurred early in radiation-induced plexopathy (pain occurred initially in 10% of patients, although ultimately it was present in 50%), whereas pain was most frequently associated with tumor plexopathy. Muscular weakness, numbness, and paresthesia are common in both groups. Electromyography showed abnormal myokymic discharges in 57% of patients, whereas this finding was very unusual in tumor-induced plexopathy. CT is extremely helpful in the detection of pelvic masses or bone destruction caused by tumor. The authors also reported extensive retroperitoneal fibrosis of the lumbosacral plexus in two patients and femoral nerve fibrosis with plexopathy in one patient. Although cystometrograms have demonstrated bladder atonicity in some cases,

P.1588

several authors have failed to observe bladder or rectal sphincter disturbances. Unfortunately, as in radiation myelopathy, the neurologic deficit is irreversible, and no effective therapy other supportive care has been found.

Other types of less severe, but still clinically significant sequelae have been described.

Bruner et al. (57), in 90 patients treated with intracavitary irradiation for either carcinoma of the cervix (42 patients) or endometrial carcinoma (48 patients), 78 of whom also received external pelvic irradiation (44.5-Gy mean dose), noted that vaginal length decreased in most patients (at 24 months, in endometrial carcinoma from 8.8 cm to 7.8 cm, and in cervical carcinoma from 7.6 cm to 6.2 cm). Pretreatment sexual activity was reported by 31% of women in comparison with 43% after treatment. However, 22% of women reported a decrease in sexual frequency, and 37% a decrease in sexual satisfaction. This was correlated with increased dyspareunia, which was noted in 31% of women treated for carcinoma of the cervix and 44% of those treated for endometrial carcinoma.

Grigsby et al. (204) described complex problems with sexual adjustment in women with gynecologic tumors treated with radiation therapy, with decreased frequency of sexual intercourse, desire, orgasm, and enjoyment of intercourse in 16% to 47% of patients.

Regular vaginal dilation is widely recommended to maintain vaginal health and sexual functioning; however, the compliance rate with this recommendation is not consistent.

In a study to test the effectiveness of an “information-motivation-behavior skills” model, 32 women compliance was significantly improved (278,514).

There was no evidence that the experimental intervention improved global sexual health. Jensen et al. (281) described persistent sexual dysfunction throughout 2 years after RT in 118 women; 85% had low or no sexual interest, 35% lack of vaginal lubrication, and 55% mild to severe dyspareunia. However, 63% of the sexually active patients before RT remained active, although with decreased frequency.

Grigsby et al. (203), in 1,313 patients with gynecologic tumors treated with radiation therapy, identified 207 who received pelvic irradiation to the inguinal areas, including the hips. Femoral neck fractures developed in 10 patients (4.8%); four were bilateral. The cumulative actuarial incidence of fracture was 11% at 5 years and 15% at 10 years. Most of the fractures occurred in patients receiving 45 to 63 Gy, and although radiation dose could not be correlated with the occurrence of fracture, no fractures were noted in patients receiving <42 Gy. Cigarette smoking and osteoporosis were significant prognostic factors for increased risk of fracture.

Blomlie et al. (40) described radiation-induced insufficiency fractures of the pelvis on MRI (characterized by edema on T1-weighted images) in 16/18 women (nine premenopausal and nine postmenopausal) with advanced cervical carcinoma. During the study, the fractures associated with edema subsided without treatment in 41/52 (79%) lesions in 15/16 (94%) patients.

Moreno et al. (420) also described eight patients with pelvic cancer who developed insufficiency fractures after pelvic irradiation. The bone and CT scan showed abnormalities in the sacroiliac joint in all cases, and in the pubis in three cases. In five patients, the initial diagnosis was bone metastases, which was incorrect. Treatment, consisting of nonsteroidal anti-inflammatory drugs and rest, led to symptomatic relief in all cases.

These data should be evaluated in light of a report by Cummings et al. (103) of 9,516 women, 192 of whom sustained hip fractures not due to motor vehicle accidents. The incidence of hip fracture ranged from 1.1 per 1,000 woman-years when no risk factors were present to 27 per 1,000 woman-years in women with risk factors and low bone density.

Bye et al. (65) assessed health-related quality of life (HRQOL) 3 to 4 years after pelvic radiation therapy for carcinoma of the endometrium and cervix in 94 survivors, 79 (84%) of whom answered a survey. The treated women scored lower than the general population on role functioning (81.5 vs. 90.6; p <0.01) and higher on diarrhea (23.8 vs. 9.5; p <0.01). Compared with pretreatment conditions, an increase in cases with pain in the lower back, hips, and thighs was seen and was associated with deterioration in HRQOL.

Sequelae of High–Dose-Rate Brachytherapy

Petereit et al. (481) reported 16 acute events (9.5%) in 169 patients treated with HDR brachytherapy (128 with cervical cancer, also receiving external irradiation, and 41 medically inoperable endometrial carcinomas). The overall 30-day morbidity rate for the patients with cervical cancer was 5.5%, and the 30-day mortality rate was 1.6% (two patients, one died of pulmonary edema 12 days after first HDR insertion and the other had enteritis and died in a nursing home).

The complication rates for HDR and LDR techniques are usually equivalent (165,455).

Petereit et al. (482) observed equivalent morbidity with LDR or HDR brachytherapy (Table 66.27).

However, in the series by Cikaric (87), the rectal complication rate was significantly higher in the LDR group. Bladder complication rates reported, in general, are lower than rectal complication rates; again, except for the series by Cikaric (87) showing a higher complication rate with the LDR technique, there were no significant differences with the two techniques.

Ogino et al. (450), in 253 patients with invasive carcinoma of the cervix treated with HDR brachytherapy, noted that grade 4 rectal complications were not observed in patients with a time–dose factor below 130 or biologic equivalent dose lower than 147, assuming an α/β ratio of 3 Gy for late reactions.

Spontaneous intraperitoneal rupture of the urinary bladder, an extremely rare event, was reported by Fujikawa et al. (166) after radiation therapy for cervical cancer in 6/148 patients treated with HDR intracavitary brachytherapy combined with EBRT. All six patients underwent laparotomy and repair of the perforation; however, rerupture of the bladder occurred in three of these patients.

Clark et al. (88) reported on 43 patients treated with pelvic EBRT (46 Gy) and three HDR intracavitary treatments given weekly combined with concomitant chemotherapy (cisplatin, 30 mg/m2 weekly) for advanced carcinoma of the cervix. At 40 months after treatment, 9/13 patients who received a dose to the rectal reference point greater than the prescribed point A

P.1589

dose had a 46% actuarial rate of serious (grade 3 and 4) rectal complications, compared with 14% in the remainder. A strong dose response was observed with a threshold for complications at a brachytherapy dose of 8 Gy per fraction.

Surgery

With improved anesthesia, surgical techniques, and antibiotic therapy, the mortality rate for radical hysterectomy with pelvic lymphadenectomy has decreased to 1% or less. The most frequent sequela after radical hysterectomy is urinary dysfunction, as a result of partial denervation of the detrusor muscle. Patients may have various degrees of loss of bladder sensation, inability to initiate voiding, residual urine retention, and incontinence.

Magrina et al. (392), in 375 patients treated with a modified radical hysterectomy for various gynecologic disorders, observed some form of postoperative (within 42 days of surgery) complications in 89 patients (24%). Patients who had a pelvic lymphadenectomy experienced a greater incidence of lower extremity lymphedema than those who did not undergo this procedure. Preoperative or postoperative pelvic irradiation was a significant predisposing factor for urinary tract infection, lymphedema, and bowel obstruction in these patients compared with those who did not receive pelvic irradiation.

Some loss of defecatory urge associated with chronic rectal dysfunction was observed by Barnes et al. (29) after radical hysterectomy. Manometric studies suggest a disruption of the spinal arcs controlling defecation.

Other complications include ureterovaginal fistula (the incidence of which has decreased to <3%), hemorrhage, infection, bowel obstruction, stricture and fibrosis of the intestine or rectosigmoid colon, and bladder and rectovaginal fistulas. Postsurgical complications are usually more amenable to correction than are late complications after irradiation.

Combined Irradiation and Surgery

When irradiation is combined with surgery, the complication rate tends to be somewhat higher, particularly because of injury to the ureter or the bladder (ureteral stricture or ureterovaginal or vesicovaginal fistula) (140). The dose of irradiation, technique, and the type of surgical procedure performed are important in determining the morbidity of combined therapy.

Jacobs et al. (276), in 102 patients with invasive cervical carcinoma treated with low-dose preoperative irradiation and a radical hysterectomy with lymphadenectomy or high-dose preoperative irradiation and a conservative extrafascial hysterectomy, noted a major complication rate of 5%.

As discussed previously, a significant number of complications are associated with pretherapy staging laparotomy, particularly if irradiation (over 55 Gy) is given to metastatic para-aortic lymph nodes. The incidence of complications is between 5% and 20%, depending on the extent of the para-aortic lymph node dissection, use of transperitoneal or retroperitoneal approach for the operation, and dose of irradiation given (651).

Postoperative Irradiation

When postoperative radiation therapy is given to selected patients, further complications of the additional therapy are expected because of intestinal adhesions to denuded surfaces in the pelvis. Enteric complications, such as obstruction, fistula, or dysfunction, were observed in 24% of patients reported by Fiorica et al. (158). Other investigators, however, have reported no increase in the incidence of severe complications in patients treated with postoperative irradiation (355,563).

Montz et al. (419) evaluated bowel obstruction in 98 patients undergoing radical hysterectomy for a nonadnexal gynecologic malignancy. The incidence of small bowel obstruction was significantly higher (p <0.05) in patients who received concomitant radiation therapy (20%). None of these patients had recurrent disease at the time of small bowel obstruction. Findings at surgery consisted of minimal incisional adhesions but extensive matted small bowel adherent to the pelvic operative sites.

After combined treatment, some degree of lymphedema is frequently observed (30% to 40%).

Hormonal Replacement after Treatment of Cervical Cancer

After pelvic irradiation or bilateral salpingo-oophorectomy, usually carried out with a radical hysterectomy in patients treated for carcinoma of the uterine cervix, symptoms of menopause may occur. They can be treated with replacement hormones, although some gynecologists have expressed reservations. During the past 25 years, hormonal replacement therapy has been shown to reduce the risk of cardiovascular diseases, osteoporotic fractures, and colon carcinoma. On the other hand, there is a significant increase of the risk in breast cancer with prolonged use longer than 5 years.

Burger et al. (61) concluded that squamous-cell cancers of the cervix, vulva, and vagina are unlikely to be influenced by hormonal replacement therapy. In a study of women with ovarian cancer, 50 years of age or younger, estrogen replacement therapy did not have a negative influence on disease-free survival. Long-term hormonal replacement therapy in women treated for a gynecologic cancer must be based on the medical history of and discussion of risk with the individual patient (and her family when warranted). Usually, 0.625 to 1.25 mg of coagulated estrogen daily is sufficient (491,538,619).

Palliative Irradiation

Frequently, the radiation oncologist is faced with the challenge of treating a patient with stage IVB or recurrent carcinoma who requires palliation of pelvic pain or bleeding. If vaginal bleeding is the main concern, a single LDR intracavitary insertion with tandem and colpostats for approximately 6,000 mgh (55 Gy to point A) suffices. If irradiation was delivered previously, lower intracavitary doses should be prescribed (4,000 to 5,000 mgh). Grigsby et al. (202) used two fractions of HDR brachytherapy with a ring applicator (once weekly) with control of bleeding in 14/15 patients.

Several high-dose fractionation schedules with external-beam radiation have been used.

Spanos et al. (576) reported on a phase II study of daily multifractionated split-course irradiation in 142 patients with recurrent or metastatic disease in the pelvis. Irradiation consisted of 3.7 Gy per fraction given twice daily for 2 consecutive days, repeated at 3- to 6-week intervals for a total of three courses, aiming to a total tumor dose of 44.4 Gy. Occasionally, this regimen was combined with an LDR intracavitary insertion (4,500 mgh), blocking the midline for the last 14.4-Gy external dose. Twenty-seven patients survived more than 1 year. There were only two recorded cases of grade 3 toxicity (lower gastrointestinal tract).

This study was expanded to a phase III protocol randomizing 136 patients between a short (2-week) or a longer (4-week) rest period between the split courses of irradiation (577). There was a trend toward increased acute toxicity in patients with shorter rest periods (5/58 vs. 0/68; p = 0.07). Late toxicity was not significantly different in the two groups. Pelvic tumor response was comparable in both groups (34% vs. 26%).

Spanos et al. (575) reported a 6% complication rate in 290 patients treated in RTOG Protocol 85-02. No patient receiving <30 Gy experienced late toxicity. There was no significant difference in the incidence of complications for patients with a 2- or 4-week rest (p = 0.47).

66_03 results of therapy

Results of Therapy

When therapeutic results in invasive carcinoma of the cervix are evaluated, a direct comparison of surgically treated or irradiated patients is fraught with many uncertainties, including patient selection, reporting of surgical cases using staging determined by laparotomy findings, and different treatment techniques (670).

The impact of patient selection in results of surgical series was illustrated by Whitney and Stehman (654), who evaluated the frequency with which intended radical hysterectomy for cervical cancer is abandoned and the outcomes for those selected patients. In 1,127 patients with stage IB carcinoma of the cervix entered on GOG Protocol 49, 98 women (8.7%) were found at surgery to have extrauterine disease, and the proposed radical operation was abandoned. Subgroups of patients with extrapelvic disease (30) and pelvic extension (26), including grossly positive pelvis nodes (12), other pelvic implants (8), and gross serosal extension (2), were identified. Sixty-three (93%) patients subsequently underwent pelvic radiation therapy and brachytherapy. Para-aortic fields were added for eight patients who were found to have positive para-aortic nodes. The disease-free survival was shorter for patients whose radical procedure was abandoned than for those patients who underwent radical hysterectomy.

Stage IA

In 47 patients with microinvasive carcinoma treated at Washington University, 20 with intracavitary therapy only and 27 patients with combined external irradiation and intracavitary brachytherapy, only one patient had a pelvic recurrence and distant metastases 10 years later; the 5-year disease-free survival rate was 96% (201).

Webb et al. (644) analyzed lymph node status and survival rates of women with microinvasive cervical adenocarcinoma (FIGO stages IA1 and IA2) from the SEER database between 1988 and 1997. Among reported cases, 131 had stage IA1 and 170 had IA2 disease. Simple hysterectomy was done in 54 women with IA1 and 64 with IA2 disease and radical hysterectomy in 50 and 83 women, respectively. Only 1/140 women who had lymphadenectomy had a single positive lymph node. There were four tumor-related deaths (one with IA1 and three with IA2 disease). The survival rate was 98.7%.

Stages IB and IIA

The important contribution of external-beam irradiation to improve pelvic tumor control in larger lesions has been documented. Hamberger et al. (226), in 151 patients with stage IA or IB lesions <1 cm in diameter treated with intracavitary therapy

P.1567


alone to high doses (8,640, 9,340, and 13,680 mgh), noted no failures in 41 patients with stage IA disease, and only 4/93 patients (4%) with stage IB, small-volume disease. However, 3/17 patients (18%) with more extensive stage IB lesions, treated with intracavitary therapy only, had regional failures. Only 3/151 patients (0.2%) had grade 3 complications.

Volterrani and Lombardi (639) reported 5-year survival of 82.6% in 23 patients with occult stage IB carcinoma of the cervix treated with intracavitary 226Ra only (7,500 mgh) in contrast to only 64.8% in larger stage IB tumors and 50% in stage II. Unfortunately, the authors did not report the exact location of the failures. It is obvious that intracavitary therapy alone is grossly inadequate to irradiate larger primary tumors, including stage IB1.

With EBRT and BT, the usual 5-year survival rate for stage IB is 86% to 92%, and for stage IIA, approximately 75%. The overall pelvic failure rate in stage IB is approximately 5% to 8%, and in stage IIA, 15% to 20% (in half of the patients combined with distant metastases). Either surgery or adequate irradiation is equally effective in the treatment of stage IB and IIA carcinoma of the cervix; numerous noncontrolled studies support the merits of either modality with no significant difference in survival or pelvic tumor control (Tables 66.16 and Fig. 66.23). Saibishkumar et al. (541) published a retrospective review of 1,069 patients with cervical cancer treated with EBRT and BT to median point A dose of 81 Gy; 5-year overall pelvic tumor control was 63.9%, DFS 49.4%, and overall survival 51.8%. Late toxicity was observed in 1% to 2% of the patients.

Randomized Studies

A few randomized studies of radical operation and irradiation have been published; outcome with the two modalities is comparable. Newton (441) and Roddick and Greenlaw (516) reported, in prospectively randomized studies, equivalent survival and pelvic recurrence rates in patients with stage IB and IIA carcinoma of the uterine cervix treated with a radical hysterectomy or irradiation alone.

Landoni et al. (352) published results of a prospective, randomized trial of radiation therapy versus surgery; 469 women with stage IB and IIA cervical carcinoma were referred for treatment and 343 were randomized (172 to surgery and 171 to radiation therapy). Postoperative irradiation was delivered after surgery for women with surgical stage pT2b or greater, <3 mm of safe cervical stroma, and cut-through margins or positive pelvic nodes. Scheduled treatment was delivered to 169 and 158 women, respectively; 62/114 women with cervical diameters of <4 cm and 46/55 with >4 cm received radiation

P.1568


therapy. After a median follow-up of 87 months (range, 57 to 120 months), 5-year overall and disease-free survival rates were nearly identical in the surgery and radiation therapy groups (83% and 74%, respectively); recurrent disease developed in 86 women: 42 (25%) in the surgery group and 44 (26%) in the radiation therapy group (Fig. 66.24). Forty-eight patients (28%) in the surgery group had severe morbidity compared with 19 (12%) in the radiation therapy group (p = 0.0004; Table 66.17). The combination of surgery and radiation therapy had the worst morbidity, especially urologic complications.

Nonrandomized Studies

Kielbinska et al. (313), in a long-term study of 792 women treated with irradiation and 789 women treated with hysterectomy and irradiation for stage I cervical carcinoma, found no difference in survival, general health, incidence of recurrent carcinoma, or appearance of second primary malignancies.

Piver et al. (489) treated 103 women with stage IB cervical carcinoma with either radical hysterectomy and pelvic lymphadenectomy (if tumor <3 cm in greatest diameter) or

P.1569


irradiation (tumor >3 cm or medically inoperable). The 5-year disease-free survival rate was 92.3% for the surgical group and 91.1% for the radiation therapy group. Equivalent overall 5-year survival rates were noted.

In stage IB and IIA disease after a hysterectomy and lymphadenectomy (even combined with irradiation), patients with metastatic lymph nodes have survival rates that are approximately 50% of those of patients with negative nodes (284).

Einhorn et al. (147), in a nonrandomized study, observed a 100% 5-year survival rate in 49 patients with stage IB disease receiving combined therapy in comparison with 81% in 64 patients treated with irradiation alone. No difference was observed in 25 patients with stage IIA tumor treated with combined therapy and 40 patients treated with irradiation alone (5-year survival rate, 75%).

Perez et al. (471) reported on a prospectively randomized study of 118 patients with stage IB or IIA carcinoma of the uterine cervix in which patients were treated with RT alone or irradiation and surgery (20 Gy to the whole pelvis, one intracavitary insertion for 5,000 to 6,000 mgh, followed by a radical hysterectomy with pelvic lymphadenectomy 2 to 6 weeks later). In stage IB, the 5-year tumor-free survival was 80% and 82% (p = 0.23), respectively, and in stage IIA, 56% and 79%, respectively (p = 0.13). The incidence of grade 2 or 3 complications radiation alone was 13.8% and with preoperative irradiation and surgery 11%.

Subsequently, Perez et al. (472) described results in 415 patients with stage IB or limited stage IIB treated with preoperative or postoperative irradiation and surgery. The 10-year cause-specific survival rate for patients with stage IB nonbulky tumors treated with irradiation alone or irradiation combined with surgery was 84% with either modality. With bulky tumors (>5 cm), the 10-year rates were 61% and 68%, respectively (p = 0.5). For patients with stage IIA nonbulky tumors, the 10-year cause-specific survival rates were 66% and 71%, respectively, and with bulky tumors, 69% and 44%, respectively (p = 0.05). In patients with stage IIB nonbulky tumors treated with irradiation alone or combined with surgery, the 10-year cause-specific survival rates were 72% and 65%, respectively.

Stages IB and IIA (Bulky)

Mendenhall et al. (408) compared 75 patients in each group treated with irradiation alone or combined with surgery for bulky tumors and reported local tumor control of 74% and 76%, and absolute 54% and 52%. The authors currently reserve combining irradiation with an extrafascial hysterectomy for patients who have <25% tumor regression at the time of the first intracavitary application, who are medically operable, and in whom it is thought adequate surgical margins may be obtained.

Thoms et al. (611) reported on 363 patients with bulky endocervical carcinoma treated with curative intent (246 with irradiation alone and 117 with irradiation and surgery); 10-year survival was 45% and 64%, respectively. In a subset of 48 patients with similar tumors treated with irradiation alone and 45 with irradiation and surgery, the 10-year survival rates were comparable, and the pelvic tumor control rates were 90% and 87%, respectively.

P.1570


Eifel et al. (144) evaluated 1,526 patients, of whom 371 had tumors 6 cm or greater. There were biases in treatment selection, but a statistically significantly higher 10-year survival rate was noted in patients treated with irradiation and surgery (64% vs. 45%). Tumor diameter was highly significant as a prognostic factor, and they concluded that only patients with lesions >8 cm in diameter benefited from adjuvant hysterectomy. In the same study, 98 patients with stage IB and IIB bulky endocervical carcinomas (≥6 cm in diameter) were treated with RT alone. Twenty-four patients received <6,000 mgh of intracavitary treatment, and 73 received higher doses. Despite having somewhat more favorably treated tumors, patients who received <6,000 mgh had a higher rate of pelvic recurrence at 5 years (33%) than those who received higher doses (16%; p = 0.03). Actuarial 5-year survival rates were 44% and 60% for low- and high-dose groups, respectively (p = 0.14).

Kim et al. (318) assessed the prognostic factors for pelvic tumor control in 40 patients with FIGO stage IB or IIA carcinoma, and 25 patients with stage IIB carcinoma classified as barrel-shaped (i.e., at least 5 cm in diameter) treated with curative intent. Seventy-two percent were treated with RT alone and 28% with RT and extrafascial hysterectomy. The extent of tumor regression after external-beam radiation therapy correlated with the likelihood of local tumor control (p = 0.02). For patients treated with radiation therapy alone, increased brachytherapy dose was associated with better local tumor control. The 10-year overall and cause-specific survival rates were 53% and 68%, respectively, and did not differ significantly between treatment groups.
Paley et al. (458) reported on 57 patients with barrel-shaped (mean diameter, 5 to 9 cm) cervical carcinoma treated with preoperative EBRT and BT (mean dose to point A, 79.6 Gy) followed by extrafascial hysterectomy 6 to 8 weeks later. Residual disease was present in 35 (61%) of the hysterectomy specimens; tumor sterilization correlated significantly with the mean dose to point A (p = 0.016). Ninety-five percent of the patients with negative specimens remained clinically free of disease at their last follow-up versus 31% of those with residual disease (p < 0.001).
The GOG and RTOG conducted a randomized phase III clinical trial in which 282 patients with carcinoma of the cervix measuring 4 cm or greater (exophytic or barrel-shaped) were treated with either external-beam and intracavitary irradiation or a slightly lower dose of intracavitary irradiation and the same pelvic EBRT followed by an extrafascial hysterectomy (311). The survival rates were 61.4% for irradiation alone and 64.4% for the combined irradiation and surgery group. The incidence of recurrences was 43.3% in the irradiation group compared to 34.5% with combined therapy (p = 0.081). The incidence of local recurrences was 25.8% 14.4%, respectively. The incidence of grade 3 and 4 sequelae of therapy was 10.5% and 9.8%, respectively. Thus, the addition of hysterectomy to standard irradiation did not significantly affect survival, although there was a small reduction in the local recurrence rate.

When combined therapy is used, the dose of irradiation delivered to the lymph nodes, the time of the operation, and the pathologic examination of the specimens are critical in

P.1571


determining the presence of postirradiation residual tumor (Table 66.18).

Perez et al. (470) noted that in patients with primary carcinoma of the uterine cervix who had endometrial stromal invasion or tumor only in the curettings, the addition of a hysterectomy did not improve the survival rate because most of the patients failed at distant sites.

Stages IIB, III, and IVA

Most patients with stage IIB tumors are treated with irradiation alone, and the 5-year survival rate is 60% to 65%. The pelvic failure rate ranges from 18% to 39%. In an analysis of the Patterns of Care Study in 157 patients who had stage IIB disease, Coia et al. (90) reported a better 4-year survival rate (67% and 54%) and in-field tumor control rate (78% and 68%) in patients with unilateral versus bilateral parametrial involvement, respectively.

Similarly, in a review of 1,178 patients with stage IIB disease treated at Washington University, the 5-year survival rates were 70% with medial parametrial and 58% with lateral parametrial involvement (p = 0.004) (533).

Kim et al. (320), in patients with stage IIB, found a correlation of point A dose and incidence of pelvic failures (Fig. 66.25). Before 1965, a pelvic lymphadenectomy was carried out at M.D. Anderson Hospital after a full course of radiation therapy, but this procedure did not improve survival over irradiation alone and the complication rate was somewhat higher (161).

In stage IIIB carcinoma, the 5-year survival rates range from 25% to 48%, and pelvic failure rates range from 38% to 50% (161,417).

Hanks et al. (228), reporting on the Patterns of Care Study, noted a 28% probability of 5-year survival in patients with stage III carcinoma of the cervix treated in a large number of facilities in the United States versus 60% survival in selected large centers (extended survey).

Later, Komaki et al. (334) reported a significant increase in local pelvic tumor control (69%) in patients with stage III carcinoma of the cervix treated in 1983, compared with 37% and 49% in earlier periods (p = 0.03). The 5-year survival rate increased from 25% to 47% (p = 0.02). The improvement in pelvic tumor control may be associated with higher external beam doses, but more likely is related to the substantial increase in the percentage of patients receiving brachytherapy (96%) and more careful dosimetry and dose calculations for intracavitary therapy. They noted a decrease in major complications from 15% in the 1973 and 13% in the 1978 Patterns of Care Surveys to 7% in 1983.

Montana et al. (418) reported that calculation of doses to the bladder and rectum were performed in 80% and 76% of patients, respectively, in the 1983 survey, which may also have resulted in decreased toxicity.

P.1572


Arthur et al. (23), in 89 patients with stage IIIB carcinoma of the cervix treated with external irradiation and brachytherapy, observed a locoregional tumor control rate of 22.5% and a disease-free survival rate of 15% in 16 patients treated with 78 Gy or lower doses to point A, in comparison with 53% and 47%, respectively, in 24 patients receiving higher doses.

Horiot et al. (259) reported the results of a French cooperative study of 1,383 patients with invasive carcinoma of the uterus treated with irradiation alone following the M.D. Anderson Hospital treatment guidelines. Survival and locoregional tumor control were similar in both groups, except in stage III, in which the pelvic and central failure rates were lower in the French patients, probably because of different tumor volumes or socioeconomic factors. Major urinary complications were noted in 2% of the patients and grade 3 bowel complications in 3% of the patients with stage I and IIA disease, and bowel complications in 7% of patients with IIB and III disease.

Barillot et al. (28) updated the results in 642 patients; the analysis was divided into three periods: 1970 to 1978 (use of standard prescriptions), 1979 to 1984 (implementation of individual adjustments), and 1985 to 1994 (systematic individual adjustments). There was a significant reduction of the external radiation dose (above 40 Gy in 47% of patients before 1979 vs. 36% after 1984), use of parametrial boost (55% vs. 39%), of use of vaginal cylinder (28% vs. 11.5%), and of combined intracavitary and external irradiation volume (842 cm3 vs. 503 cm3 on average). The 5-year actuarial toxicity rates were grade 2, 23.5%; grade 3, 10%; and grade 4, 3%. The three main predictive factors for rectal and bladder sequelae were increased external radiation dose, higher dose rate at reference points, and whole-vagina brachytherapy. The disease-free survival rates observed with irradiation alone at Washington University Medical Center are shown in Figure 66.26.

Marcial et al. (399) described results of a randomized trial in 301 patients with stage IIB, III, and IVA carcinoma of the uterine cervix treated with split-course irradiation (10 fractions of 2.5 Gy, five weekly doses up to 25 Gy, followed by a rest period of 2 weeks, and an additional 25 Gy delivered in the same manner) or continuous irradiation (30 fractions of 1.7 Gy daily, five times per week, total dose 51 Gy) combined with LDR brachytherapy for 30 Gy to point A. There was no significant difference in tumor control, acute or late complications, or survival in the two groups.

In patients with stage IVA disease, the 5-year survival rates range from 18% (333) to 34%, and pelvic failures from 60% to 80% after definitive irradiation. Million et al. (412) reported 18/53 patients (34%) with bladder involvement surviving without disease after definitive irradiation, results comparable with those obtained with exenteration.

Upadhyay et al. (625) noted 43% local tumor control and 18% 5-year survival rates in 44 patients with stage IVA carcinoma of the cervix treated with definitive radiation therapy.

Kramer et al. (337) reported on 48 patients with stage IVA carcinoma of the cervix treated with definitive RT. Patients with minimal parametrial involvement had a 5-year survival rate of 46% compared with only 5% for those with extensive parametrial tumor. The major complication rate was 22%, consisting mostly of vesicovaginal fistula in five patients.

Crozier et al. (102) described equivalent 5-year survival rates after salvage pelvic exenteration (37% in 35 patients with adenocarcinoma and 39% in 70 patients with squamous-cell carcinoma). In the adenocarcinoma group, 14/22 patients, and in the squamous cell-carcinoma group, 14/30 patients had distant metastases after pelvic exenteration.

Treatment of Elderly Patients

Oguchi et al. (451) reported on 23 patients 90 years of age or older treated with cervix carcinoma. Definitive radiation therapy was completed in 13 of the patients, and local tumor control at 6 months was attained in nine patients. Palliative RT was completed in 7/11, and palliation was observed in nine patients (81%). Seven patients were alive for 15 to 67 months. Fourteen patients died because of intercurrent disease or senility associated with active cancer, and two because of senility without evidence of cancer. The 2-year overall and relapse-free survival rates were 30% and 21%, respectively.

Multivariate Analysis

Fyles et al. (170), in 965 patients with invasive carcinoma of the cervix, identified FIGO stage as the most significant prognostic factor, followed by dose of irradiation to point A and overall time of radiation therapy. The 10-year survival rate was 62% in 743 patients receiving doses to point A of 85 Gy or higher, in contrast to 53% for 222 patients receiving lower doses.

Patients with higher hemoglobin levels not receiving a transfusion (595) had a 10-year survival rate of 60% in contrast to 42% in 353 patients who were given transfusions for lower hemoglobin levels.

Chatani et al. (75), in 216 patients with stage IIB-III cervical carcinoma treated with a combination of external-beam and HDR brachytherapy, noted that overall treatment time was the most highly significant factor for local tumor control in multivariate analysis (p = 0.0005). Concerning relapse-free survival, stage classification (p = 0.0001), overall treatment time (p = 0.0035), and hemoglobin level (p = 0.0174) were the three

P.1573


most important prognostic factors; there was no relationship between treatment time and late complications.

Kapp et al. (299), in a study of 181 patients with FIGO stages IB to IV carcinoma of the cervix, documented that prognostic factors for patients treated with HDR are similar to those in previous series with LDR brachytherapy. In multivariate analysis, tumor size was the most powerful for pelvic tumor control and incidence of distant metastasis.

Interstitial Implants

Interstitial brachytherapy, discussed in detail in Chapter 20, has been used in the treatment of patients with cervical cancer (591).

Because of the inability to insert an intracavitary tandem interstitial needle, implants were used at Washington University in 30 patients with stage IIB and in 37 with stage III carcinoma to deliver interstitial irradiation in the parametrium to supplement the dose delivered by external-beam and intracavitary brachytherapy. Despite the fact that the patients treated with interstitial implant were in a high-risk group, local tumor control was comparable with that of patients treated with standard techniques (474).

Pierquin et al. (484) described locoregional recurrences in 6% of 53 patients with T1, 11% in 47 with T2, and 42% of 19 patients with T3 primary tumors of the uterine cervix treated with a combination of external-beam irradiation and the Creteil method for interstitial implantation of 192Ir sources in a plastic cervical-vaginal moulage and a uterine tandem.

Prempree (502) reported a 96% local tumor control rate and 61% 5-year disease-free survival rate in 23 patients with stage IIIB carcinoma of the cervix treated with a combination of external irradiation and intracavitary and interstitial implants to the parametrium. Overall, major complications were noted in 8% of the patients.

Martinez et al. (402), using the Martinez Universal Perineal Interstitial (MUPIT) applicator, treated 37 patients with advanced or recurrent carcinoma of the cervix and 26 with vaginal-urethral tumors. Doses of approximately 35 Gy were given, in addition to external irradiation (36 Gy to the whole pelvis and 14 Gy to the pelvic sidewall). They reported six local failures in the patients with cervical lesions and five in the group with vaginal-urethral tumors. The overall complication rate was 5.1%.

Nag et al. (430) reported on 31 patients with carcinoma of the cervix and eight with carcinoma of the vagina treated with external-beam radiation therapy and fluoroscopically guided interstitial brachytherapy. With a median follow-up of 36 months, 16 patients (51%) with cervical and five patients (62.5%) with vaginal carcinomas had local tumor control. The 5-year actuarial survival rates were 34% and 38% for cervical and vaginal cancers, respectively. Only one patient experienced grade 3 complications (2.5%).

Recio et al. (508) used laparoscopy at the time of interstitial brachytherapy in six patients with FIGO stages IIB to IVA cervical carcinoma, after completion of whole pelvis radiation; a total of 98 needles were inserted to deliver a median interstitial brachytherapy dose of 20 Gy. Eleven perforations in the pelvic peritoneum or bladder were identified during surgery in five of the six patients, leading to immediate repositioning of needles. No acute or short-term morbidity related to the procedure was noted.

Californium-252 or Neutrons

Maruyama and Muir (404) reported on 41 patients with stage IB cervix cancer treated with 40 to 50 Gy to the whole pelvis followed by a 5- to 15-Gy boost to the lateral pelvic wall and a single 252Cf-neutron brachytherapy insertion in approximately 8 hours. Nearly total tumor clearance was achieved in over 90% of the patients; tumor regression was more rapid in the 252Cf group than in similar patients treated with 137Cs and the same external-beam irradiation dose.

Maor et al. (398) published results in 156 patients with locally advanced cervical carcinoma treated at five institutions and randomized to receive external photons only to the pelvis (50 Gy in 25 fractions in 5 weeks) or mixed-beam external irradiation (three fractions a week of photons) to a total relative biologic effectiveness adjusted dose of 50 Gy over 5 weeks. All patients were scheduled to receive LDR intracavitary brachy-therapy. Of 146 evaluable patients, 80 were treated with mixed-beam irradiation and 66 with photons. Only 50% of the patients in the mixed-beam group and 75% in the photon group underwent brachytherapy. The local tumor control at 2 years was 45% in the mixed-beam group and 52% in the photon group. Severe complications occurred in 19% of the mixed-beam and 11% of the photon-beam patients (p <0.13). It was thought that the inferior outcome with neutrons may have resulted from the use of horizontal beams of varying energy and penetration.

Heavy Ions

Kato et al. (302) reported on 44 patients with locally advanced cervical cancer treated with carbon ion radiation therapy. Total whole pelvis dose was 52.8 to 72 Gy equivalent (GyE) in 16 fractions of 2.2 to 3 GyE and eight fractions local boost or 68 to 72.8 GyE (44.8 GyE and additional 24 or 28 GyE boost). The 5-year local tumor control was 45% and 79%, respectively. Eight patients developed major intestinal complications, which were surgically salvaged; they were associated with doses >60 GyE.

External-Beam Irradiation Alone

Occasionally, brachytherapy procedures cannot be performed because of medical reasons or unusual anatomic configuration of the pelvis or the tumor (i.e., extensive lesion, inability to identify the cervical canal). These patients may be treated with higher doses of external-beam irradiation alone, although treatment results are less than optimal.

Coia et al. (90), in an analysis of 565 patients with various stages of cervical carcinoma treated in the Patterns of Care Study, reported better survival (67%) and pelvic tumor control (78%) when patients were treated with external irradiation and brachytherapy compared with patients who had no intracavitary brachytherapy applications (36% 4-year survival and 47% in-field failure rates). Patients treated with two intracavitary applications had a higher 4-year survival rate (73%) and in-field tumor control rate (83%) than those receiving only one application (60% 4-year survival rate and 71% in-field tumor control rate).

Hanks et al. (228) and Montana et al. (417) reported a higher incidence of central pelvic recurrences in patients with stage III cervical carcinoma treated with external-beam therapy alone than in patients receiving brachytherapy in addition to external-beam irradiation (Table 66.19). The incidence of major complications was similar in both groups of patients.

Akine et al. (9) treated 104 patients with carcinoma of the uterine cervix with external irradiation alone (anteroposterior–posteroanterior or four-field box techniques) because of inability to perform intracavitary brachytherapy. Average doses delivered were 50 Gy to the whole pelvis, followed by additional doses with reduced portals to deliver a total of 60.8 Gy in 6 weeks, 72.3 Gy in 7.5 weeks, or 80.5 Gy in 8 weeks, with a daily dose of 1.9 or 2 Gy. The local tumor control rate was 27% for stage II, 19% for stage III, and 15% for stage IVA disease. The 5-year survival rates were 36%, 17%, and 5%, respectively. Four patients had major complications (usually proctitis) that required surgical treatment, and one patient died of rectal bleeding. Eight of 23 patients treated with conformal therapy

P.1574


had control of the tumor and survived 5 years without major complications.

Saibishkumar et al (540) treated 146 patients with cervix cancer with EBRT alone (60 to 66 Gy), because of unsuitability for brachytherapy; 5-year pelvic tumor control was 21.9% and DFS 11.6%.

Impact of Tumor Size on Outcome

Perez et al. (474), in an update of a previous report (476), reviewed 1,499 patients (stages IA to IVA) treated with definitive irradiation (combination of external-beam irradiation plus two intracavitary insertions to deliver doses of 70 to 90 Gy to point A). There was a close correlation between tumor size and extent with pelvic tumor control, incidence of distant metastasis, and disease-free survival in all stages. Eifel et al. (140), in a review of 1,526 patients with stage IB squamous-cell carcinoma of the uterine cervix treated with radiation therapy alone, noted pelvic tumor control in 97% of tumors <5 cm and in 84% of tumors 5 to 7 cm.

Impact of Prolongation of Treatment Time on Outcome

Several studies have described lower pelvic tumor control and survival rates in invasive carcinoma of the uterine cervix when the overall time in a course of irradiation is prolonged (115,167,179,350).

Fyles et al. (167) reported approximately 1% loss of tumor control per day of prolongation of treatment time beyond 30 days in 830 patients with cervical carcinoma treated with irradiation alone.

Lanciano et al. (350), in an analysis of 837 patients with squamous-cell carcinoma of the cervix from the Patterns of Care Study who were treated with irradiation and received doses of 66 Gy or greater, also described a 4-year actuarial in-field recurrence increase from 6% to 20% when total treatment time increased from 6 weeks or less to 10 weeks (p = 0.0001); this translated into significantly decreased survival.

Girinsky et al. (179), in 386 patients with stage IIB or III carcinoma of the cervix, also observed that the 10-year local recurrence–free survival rate decreased when overall treatment time exceeded 52 days. A 1.1% loss of pelvic tumor control per day was also observed in their regression analysis.

Perez et al. (473), in 1,330 patients treated with definitive irradiation, noted a major impact of prolongation of treatment time on pelvic tumor control in stages IB, IIA, and IIB (Fig. 66.27A,B). In stage III, although the rate of pelvic failure was higher with prolongation of treatment time, the difference was not statistically significant. There was also a strong correlation between overall treatment time and survival (Fig. 66.28A,B). Regression analysis confirmed previous reports that prolongation of overall treatment time resulted in an increased failure rate of 0.59% per day in stage IB and IIA and 0.86% per day in stage IIB disease. Performance of all intracavitary insertions within 4.5 weeks from initiation of irradiation yielded lower pelvic failure rates (8.8% vs. 18% in stage IIB tumors; p ≤0.01).

In patients treated with radiation therapy, overall treatment time should be as short as possible, and any planned or unplanned interruptions or delays should be avoided. Timely integration of external-beam and intracavitary irradiation in patients with carcinoma of the uterine cervix is an important factor in improving pelvic tumor control.

Metastases to Para-Aortic Lymph Nodes

Para-aortic lymph node metastases are frequently combined with distant dissemination but are clinically apparent in only 10% to 20% of patients who have recurrences.

Nelson et al. (437) reported on 104 patients with stage II and III cervical carcinoma who had exploratory laparotomy and para-aortic lymph node biopsies; 12.5% of patients with stage IIA disease, 14.9% with stage IIB, and 38.4% with stage III disease had para-aortic lymph node metastases. They were treated with 60 Gy to the para-aortic region. Within 4 years, 50% of these patients had distant metastases and only one out of 13 was alive. There was no significant increase in complications in the patients receiving para-aortic irradiation (39% and 32%). They concluded that the main goal of exploratory laparotomy and para-aortic lymph node biopsy is to define the extent of disease.

Lovecchio et al. (388) noted a 50% 5-year survival rate in 36 patients with stage IB and IIA cervical carcinoma who had histologically confirmed para-aortic lymph node metastases treated with RT (including 45 Gy to the para-aortic lymph nodes). Fourteen of 31 evaluable patients had pelvic recurrences (12 combined with distant metastases). Unfortunately, the authors did not specify how many patients had para-aortic recurrences, although they reported four abdominal failures.

Stryker and Mortel (585) determined survival after extended-field treatment of para-aortic lymph node metastasis plus brachytherapy or pelvic boost in 35 patients; 5-year survival was 41.7% for 12 patients with microscopic para-aortic lymph node metastasis and 26.1% for 23 with grossly enlarged lymph nodes. Three patients (8.6%) had grade 4 morbidity.

Grigsby et al. (199) reviewed 43 patients with cervical cancer and biopsy-proven positive para-aortic lymph nodes treated with external irradiation to the pelvis and para-aortic regions (45 to 50 Gy) combined with brachytherapy. The 5-year overall survival rate was 32%, and the cause-specific survival rate was 49%. Tumor recurrence occurred in 20 patients (three in the pelvis, nine in pelvis and distant metastasis, and eight

P.1575


distant metastasis only). Severe grade 3 complications occurred in two patients (one had an enterovaginal fistula and the other radiation myelitis).

Hacker et al. (216), in 437 patients with invasive cervical carcinoma, 222 treated with radical hysterectomy and lymphadenectomy, identified 34 in whom resection of bulky pelvic or para-aortic lymph nodes was carried out without a complete lymphadenectomy. Thirty-three patients received pelvic external irradiation, and 28 combined pelvic and para-aortic extended-field irradiation (50.4 Gy in 1.8-Gy fractions using a four-field technique). Four cycles of cisplatin were administered to 23 patients. The 5-year survival was 80% in patients with pelvic and common iliac nodes and 48% in those with positive para-aortic lymph nodes. Serious long-term morbidity occurred in six patients (18%). Radiation enteritis was observed in five patients, leading to small bowel obstruction necessitating resection.

Grigsby et al. (196) evaluated twice-daily external irradiation to the pelvis and para-aortic nodes (1.2 Gy at 4- to 6-hour intervals, 5 days per week) combined with brachytherapy and concurrent chemotherapy in 29 patients with carcinoma of the cervix and positive para-aortic lymph nodes. EBRT doses were 24 to 48 Gy to the whole pelvis, 12 to 36 Gy parametrial boost, and 48 Gy to the para-aortics with additional boost to a total dose of 54 to 58 Gy to known metastatic para-aortic sites. One or two LDR brachytherapy applications were performed to deliver a total dose of 85 Gy to point A. Cisplatin (75 mg/m2, days 1 and 22) and 5FU (1,000 mg/m2 per 24 hours for 4 days; days 1 and 22) were given for two or three cycles. Hyperfractionated external radiation therapy was completed in 86% (25/29). Radiation therapy toxicity was grade 2 in 34%, grade 3 in 21%, and grade 4 in 28%. An unacceptably high rate (31%, 9/29) of grade 4 nonhematologic toxicity was recorded. With a median follow-up of 18.9 months, at 2 years the overall survival rate was 47%, and the probability of locoregional failure was 49%.

Malfetano et al. (395) treated 67 patients with carcinoma of the cervix (44 with stage IIB and 23 with stage III disease) with cisplatin (1 mg/kg up to 60 mg weekly) and extended-field radiation therapy, including the para-aortic nodes, and brachytherapy; 75% were alive without evidence of disease with a mean follow-up of 47.5 months.

Chou et al. (82) treated 19 patients with isolated para-aortic lymph node metastasis from cervix cancer, 14 of them with chemoradiation, four with chemotherapy, and one with irradiation alone. Seven of the 14 patients receiving chemoradiation survived.

P.1576


Goodman et al. (183) compiled survival statistics on patients with para-aortic lymph node metastasis and found an average 5-year survival rate of approximately 40% (Table 66.20).

Elective Para-Aortic Lymph Node Irradiation

Rotman et al. (527) for the RTOG (527) updated results of a randomized study of 337 patients with stage IIB carcinoma of the uterine cervix with no clinical or radiographic evidence of para-aortic lymphadenopathy who, in addition to standard pelvic irradiation, were randomized to electively receive or not 45 Gy to the para-aortic region (1.6- to 1.8-Gy fractions). The 10-year survival rate was 55% for patients receiving elective para-aortic irradiation and 44% for those treated to the pelvis only (p = 0.02; Fig. 66.29). The locoregional tumor control rate was similar (69% in the para-aortic node–irradiated group and 65% for the pelvis-irradiated group). The 10-year grade 4 or 5 (major) complication rate was 8% in the group receiving para-aortic irradiation compared with 4% in patients treated with pelvic irradiation alone (p = 0.06).

A similar randomized study was reported by Haie et al. (219) and the European Organization for Research and Treatment of Cancer (EORTC) on 441 patients with cervical carcinoma, including stage III, who had no evidence of para-aortic lymph node involvement. In the study group, the para-aortic area either received or did not receive 45 Gy with external-beam irradiation. No statistically significant difference was found between the two treatment arms with regard to local tumor control, distant metastases, or survival. However, the incidence of para-aortic and distant metastases without pelvic failure was significantly higher in patients receiving pelvic irradiation alone. The incidence of small bowel injury was 0.9% in the pelvic irradiation and 2.3% in the pelvic and para-aortic

P.1577


irradiation groups. A severe complication rate of 9% was observed in patients receiving para-aortic irradiation compared with 4.8% in those treated to the pelvis only.

Sood et al. (571) treated 54 patients with cervix cancer using extended fields (45 Gy) and HDR BT; 44 received concurrent cisplatin (20 mg/m2 per day for 5 days during the first and fourth week and once after the second HDR insertion). During a median follow-up of 28 months, six patients had died. The 3-year local tumor control was a 100% and 85%, respectively. Late toxicity was 10% and 6%, respectively.

Adenocarcinoma of the Cervix

Several authors have reported similar survival rates for equivalent stages of adenocarcinoma or squamous-cell carcinoma; clinical stage, volume of disease, and dose of irradiation were the most important prognostic factors (200,315). Grigsby et al. (200) found no difference in 5-year DFS in patients with adenocarcinoma of the cervix compared with squamous-cell treated with RT alone or combined with surgery (Table 66.21).

Similar observations were reported by Eifel et al. (138) in 229 patients with stage IB
adenocarcinoma of the cervix. The 5-year survival rates were 72% versus 81% for squamous-cell carcinoma. The incidence of pelvic recurrence was similar (17% for adenocarcinoma and 13% for squamous-cell carcinoma). However, distant metastases were more frequent in patients with adenocarcinoma (37% vs. 21%; p <0.01). The survival rate in 165 patients who underwent adjuvant hysterectomy (78%) was not significantly different from that of patients who did not have a surgical procedure (71%).

Later Eifel et al. (141) studied 334 patients with adenocarcinoma of the cervix. The 5-year relapse-free survival and locoregional control rates were 88% and 94%, respectively, in 91 patients with a normal-sized cervix, 64% and 82%, respectively, in 102 patients with lesions 3 to 5.9 cm in diameter, but only 45% and 81% in 22 patients with tumors >6 cm in diameter.

Fifty-eight patients with adenocarcinoma of the cervix treated with LDR or HDR brachytherapy and external pelvic irradiation were studied by Nakano et al. (433). The 10-year survival rates for stages I, II, III, and IV were 85.7%, 60%, 27.6%, and 9.1%, respectively. The local tumor control rate with HDR treatment was 45.5%, significantly lower than with LDR (85.7%) or mixed–dose-rate treatments (72.7%).

Kilgore et al. (315), in a study of 162 patients with adenocarcinoma compared with matched patients with squamous-cell carcinoma, found that clinical stage and lesion size were the most important prognostic factors. In patients with stage I tumors, no significant difference in survival was found when they were treated with radical surgery, irradiation alone, or irradiation combined with hysterectomy.

In contrast, Kjorstad et al. (327) reported a worse 5-year survival rate in 102 patients with adenocarcinoma (51%) compared with that of 1,900 patients with squamous-cell or other differentiated carcinomas (68%).

Comparison of Low–Dose-Rate and High–Dose-Rate Brachytherapy Results

Randomized Studies

A few randomized studies have been published comparing HDR and LDR brachytherapy for carcinoma of the cervix (215).

A report by Shigematsu et al. (556), in patients with stage IIB or III disease treated with the HDR technique, showed higher 1-year local control (90%) with HDR versus 77% with LDR. The 5-year survival rate was 55% for both groups.

Another trial including patients with stage IB, IIA, IIB, and III disease by Gupta et al. (215) showed similar local tumor control rates for both HDR and LDR (80% and 85%, respectively). However, the stage distributions in each group and the survival and complication rates were not described.

Teshima et al. (603) reported on a prospective, randomized study of 430 patients with carcinoma of the uterine cervix treated with either LDR (171 patients) or HDR (259 patients) brachytherapy combined with external irradiation. Cause-specific and overall survival rates were comparable for each clinical stage with either modality, except for stage I overall survival (Fig. 66.30). The conversion factor of total intracavitary dose from LDR to HDR was 0.5 to 0.53. With HDR, four fractions usually were delivered, and with LDR two fractions. The incidence of pelvic failures was comparable in both groups. The incidence of grade 2 and 3 morbidity was somewhat higher in the HDR group (approximately 10%) than in the LDR group (4%; p = 0.002).

Patel et al. (463) published a randomized trial of 482 patients with invasive squamous-cell carcinoma of the cervix. The overall local tumor control rate with LDR brachytherapy was 79.7% compared with 75.8% with HDR. The 5-year survival rates were 73% with LDR and 78% with HDR in stage I, 62% and 64%, respectively, in stage II, and 50% and 43% in stage III. The only statistically significant difference was the incidence of overall rectal complications, which was 19.9% for LDR compared with 6.4% for HDR. However, the incidences of more severe grade 3 and 4 complications were not significantly different (2.5% and 0.4%, respectively). Bladder morbidity was similar in both groups.

Hareyama et al. (231) conducted a randomized study in 132 patients with stage II or IIIB cervical carcinoma treated with LDR or HDR BT and identical pelvic EBRT. The conversion factor from LDR to HDR was 0.588. The 5-year DSS with HDR for stage II and IIIB was 69% and 51%, respectively, and with LDR 87% and 60%, respectively. Pelvic tumor control was 89% and 73% and 100% and 70%, respectively, and grade 3 or

P.1578


greater morbidity was 10% and 13%, respectively (differences were not statistically significant).

Lertsanguansinchai et al. (373) randomized 237 patients with cervical cancer to be treated with LDR (109 patients) or HDR (112 patients) brachytherapy and EBRT. Median follow-up was 40 and 37 months, respectively. Three-year pelvic tumor control was 89% and 86.4% and relapse-free survival 69% in both groups. Grade 3 or 4 morbidity was noted in 2.8% of LDR and 7.1% of the HDR patients (p = 0.23).

Nonrandomized Studies

Many nonrandomized studies compared the results of HDR with those of historic or concurrent control patients receiving LDR at the same institution (21,87,165,180,247,289). Most studies used point A as a reference point, although the definition of point A may have differed from center to center.

HDR dose per fraction at point A was 3 to 10 Gy; the number of fractions ranged from two to 13, and the number of fractions per week varied from one to three. Most centers used a schedule of 6 to 7 Gy per fraction per week for four to six fractions.

Table 66.22 illustrates the 5-year survival rates of patients who received HDR or LDR brachytherapy combined with external-beam irradiation for carcinoma of the cervix. Stage for stage, 5-year survival rates of patients treated with the HDR technique are comparable with those of historic or concurrent nonrandomized patients treated with LDR.

In an analysis of 198 patients treated with LDR brachytherapy at the University of Wisconsin, the 3-year survival rate was 66% versus 77% for 40 patients treated with HDR brachytherapy (545). Pelvic tumor control rate were 80% and 77%, respectively. The incidences of complications requiring hospitalization or surgery were 10% (20/198) and 2.5% (1/40), respectively.

Petereit et al. (482) updated the University of Wisconsin experience with 191 patients receiving LDR brachytherapy and 173 receiving HDR brachytherapy with equivalent external-beam radiation therapy techniques. The results are summarized in Tables 66.23 and 66.24. Pelvic tumor control and survival rates were comparable with the two techniques, except in stage III; in this subgroup, outcome was better with LDR brachytherapy, but this may be related to a lower HDR equivalent dose administered.

P.1579


Le Pechoux et al. (359) treated 130 patients with cervical cancer with HDR brachytherapy (for stage I, 30 Gy in six weekly sessions) in combination with EBRT (50-Gy mean dose with midline shielding). Patients with more advanced disease received four sessions of biweekly brachytherapy for a total dose of 18 to 24 Gy and external irradiation (20 to 30 Gy to the whole pelvis, 50 to 66 Gy to parametria with midline shielding). The 5-year survival rates were 82% for patients with stage IIB and 47% with stage IIIB disease. There were four rectovaginal or vesicovaginal fistulas and one case of proctitis requiring colostomy. Survival, local tumor control, and morbidity were equivalent in 76 patients treated with 6 Gy once a week and in 54 patients receiving twice-weekly brachytherapy of 5 Gy per session.

Hsu et al. (266) treated 92 patients with cancer of the cervix with HDR brachytherapy, six fractions of 7 Gy per fraction (42 Gy) at point A (HDR-6); 57 received four fractions of 8 Gy per fraction (32 Gy) at point A (HDR-4). A twice-daily program was used for all patients receiving HDR in two split courses. A historic control group of 259 patients was treated with LDR brachytherapy (40 Gy in two split courses). All patients received whole pelvis external irradiation of 36 to 45 Gy (mostly 40 Gy) before brachytherapy. Five-year local tumor control rates were equivalent in the three groups (82%, 85.5% for HDR, and 89.5% for LDR). Five-year survival rates were also comparable (67.7%, 77.9%, and 74.1%, respectively). However, late complications were lower in the HDR-4 group, which received treatment more biologically equivalent to the LDR regimen, than in patients in the HDR-6 group (11% vs. 25.6%).

Selke et al. (548) published results in 187 patients with primary carcinoma of the cervix treated with whole pelvis irradiation (46 Gy) and HDR brachytherapy with a dose rate to point A of 1.6 Gy per minute, decreasing to approximately 0.8 Gy per minute at the end of the 5-year study. Three HDR fractions (8 to 10 Gy to point A per fraction) were concurrently administered with the last 2 to 3 weeks of external irradiation. The 5-year actuarial survival rates were 72% for stage IB, 65% for

P.1580


IIA, 66% for IIB, 66% for IIIA, and 45% for stage IIIB. With a median follow-up of 54 months, 23 patients had 25 complications; 13 (7.6%) were grade 3 or 4. Rectal complications were significantly higher in patients who received a total rectal dose higher than 54 Gy (p = 0.045).

Choi et al. (81) treated 136 patients with carcinoma of the cervix with external-beam whole pelvis irradiation (46 Gy in 23 fractions) and three weekly applications of HDR brachytherapy of 7 or 8 Gy per fraction to point A. The actuarial 5-year survival was 85% in stage IB, 64% in stage IIA, 70% in stage IIB, and 53% in stage IIIB. Grade 3 or higher complications occurred 3% to 7% of the patients. The most significant determinants of severe rectal complications were the addition of a lower vaginal tandem (p <0.01), uterine tandem length longer than 5 cm, a total biologically effective dose to the rectum of more than 120 Gy, or stage III disease.

Kagei et al. (290) reported on 217 patients with carcinoma of the cervix (71 patients with stage II and 146 with stage III disease) who received whole pelvis EBRT (40 Gy in 20 fractions or 39.6 Gy in 22 fractions) and an additional 10 Gy in five fractions to the parametria followed by HDR brachytherapy. Cause-specific 5-year survival rates were 77% for stage II and 50% for stage III. Pelvic failure rates were 13% and 36% respectively. The rates of severe (grade 4) late complications were 2% for the rectum, 1% for the small intestine or sigmoid colon, and 1% for the bladder.

Takeshi et al. (595) treated 265 patients with stage III cervical carcinoma with external-beam radiation therapy (50.3 Gy) and intracavitary HDR brachytherapy (19.8 Gy). The 5-year overall survival, relapse-free survival, and locoregional event–free rates were 50.7%, 57.1%, and 71.2%, respectively. The 5-year incidence of major complications was 2.6% for bladder and 8.3% for rectum. The radiation dose in the subgroup with rectal complications was significantly greater than that in the subgroup without complications.

Wang et al. (641) reported treatment results in 173 patients with cervical carcinoma treated with HDR brachytherapy, whole pelvis irradiation (40 to 44 Gy in 20 to 22 fractions) followed by pelvic wall boost (6 to 14 Gy in three to seven fractions with central shielding). HDR brachytherapy delivered 7.2 Gy to point A in each of three applications, 1 to 2 weeks apart. Five-year pelvic tumor control rates were 94%, 87%, and 72% for stages IIA, IIB to IIIA, and IIIB to IVA, respectively. Five-year actuarial survival rates were 79%, 59%, and 41%, respectively. Sixty-six (38%) had rectal complications and 19 (11%) bladder complications. The 5-year actuarial rectal complication rates were 15%, 4%, and 3% for grades 2, 3, and 4, respectively.

Potter et al. (499) reported results in 148 patients treated with HDR brachytherapy and EBRT (48.6 to 50 Gy). Small tumors were treated with five to six fractions of 7 Gy at point A (25 Gy in the brachytherapy volume), which is isoeffective to 76 to 86 Gy at point A. Large tumors received three to four fractions of 7 Gy after 50 Gy of EBRT, which is isoeffective to 82 to 92 Gy at point A. Three-dimensional treatment planning for brachytherapy was based on conventional x-rays, and in 181/189 patients on CT scan. The mean brachytherapy dose was 16.2 Gy at the ICRU rectum reference point and 14.4 Gy at the ICRU bladder point. Taking into account the dose for EBRT, the mean isoeffective dose at the ICRU rectum reference point was 69.9 Gy. After a mean follow-up of 34 months, the actuarial late complication rate for grades 3 and 4 was 2.9% for bladder, 4% for bowel, 6.1% for rectum, and 30.6% for the vagina (shortening and obliteration).

Lorvidhaya et al. (386) reported the results in 1,992 patients with carcinoma of the cervix treated by external irradiation and HDR brachytherapy. There were 211 with stage IB, 225 with stage IIA, 902 with stage IIB, 14 with stage IIIA, 675 with stage IIIB, 16 with stage IVA, and 16 (0.8%) with stage IVB. With a median follow-up of 96 months, the actuarial 5-year disease-free survival rates were 70%, 59.4%, 46.1%, 32.3%, 7.8%, and 23,1%, respectively. The late complication rates (RTOG) for bowel and bladder combined were 7% for grade 3 and 1.9% for grade 4 complications.

Leborgne et al. (362) described a 4-year pelvic control rate of 93%, and a disease-free survival rate of 88% for 59 patients with stage IB to IIA disease treated with 18 Gy whole pelvis and 22 Gy to the parametria combined with six HDR fractions (14 Gy per hour to point A) of 7 Gy to point A, two in each treatment day with 6-hour intervals. The corresponding parameters for 29 patients with stage IIB disease were 79%, 75%, and 75%. The actuarial 4-year late grade 2 and 3 complication rate was 4.7%.

Ferrigno et al. (154) carried out a retrospective study of 190 patients treated with LDR and 118 with HDR brachytherapy in combination with pelvic EBRT for cervical cancer. For stage I or II patients, there was no difference in outcome; however, in the stage III group local tumor control was 58% with LDR and 50% with HDR (p = 0.19) and DFS was 49% versus 37% (p = 0.03). At 5 years, rectal sequelae were 16% versus 8% (p = 0.03), bladder 6% and 3% (p = 0.13), and small bowel 4.6% and 8.9% (p = 0.17).

Nakano et al. (434), in 1,148 patients with squamous-cell cervical cancer treated with external RT and HDR brachytherapy with 22 years median follow-up, the 10-year pelvic tumor control was 93% for stage IB, 82% for stage II, and 75% for stage III. Cause-specific survival was 89%, 74%, and 59%, respectively. Major sequelae were 4.4% in the rectosigmoid, 0.9% in the bladder, and 3.3% in the small intestine.

Kapp et al. (298) analyzed 720 192Ir HDR applications in 331 patients with gynecologic tumors to evaluate the dose to normal tissues. The ratio of bladder-base dose to bladder-neck dose was 1.5 for intracervical and 1.46 for intravaginal applications. CT-assisted dosimetry showed that the maximum doses to bladder and rectum were generally higher than those

P.1581


obtained from orthogonal films, with an average ratio of 1.44 for the bladder neck, 2.42 for the bladder base, and 1.37 for the rectum. If conventional methods are used for dosimetry, the authors recommended that doses to the bladder base should be routinely calculated because single-point measurements at the bladder neck seriously underestimate the dose to the bladder. Also, the rectal dose should be determined at several points over the length of the implant because of the wide range of anatomic variations.

Wright et al. (664) developed a questionnaire to elicit patient preference for two brachytherapy methods (one LDR or three HDR fractions and two HDR or five HDR fractions, assuming both methods to be isoeffective). The questionnaire was completed by 90 female staff members at their center, 18 previously treated patients, and 20 newly diagnosed patients. When both methods were assumed to be isoeffective, only 34% of the 38 patients preferred three HDR fractions to one LDR fraction. However, when HDR was assumed to be 2% more curative or 6% less toxic, 50% said they would prefer the HDR therapy. Both preference and strength of preference for LDR were significantly associated with a greater traveling distance for treatments.
Studies on resource utilization, such as published by Bastin et al. (31), and cost-effectiveness comparing LDR and HDR treatment should be carried out. These considerations have more relevance as we continue to debate the cost–benefit issues of HDR brachytherapy and health care costs.

Irradiation after Radical Hysterectomy

Only one randomized study has shown improved survival with postoperative pelvic irradiation (46 to 50.4 Gy in 23 to 28 fractions) after radical surgery in the presence of positive pelvic nodes or node-negative high-risk factors in women with stage IB cervical cancer treated by radical hysterectomy and pelvic lymphadenectomy. There were 277 eligible patients with at least two of the following risk factors: greater than one-third stromal invasion, capillary lymphatic space involvement, and large clinical tumor diameter; 137 patients randomized to pelvic radiation therapy and 140 to no further treatment.

The results were updated by Rotman et al. (528); 24 (17%) patients in the irradiation group and 43 (30.7%) in the no further treatment group had cancer recurrences. In the radiation therapy group 27 patients died of cancer, and in the no further treatment group 40 died from cancer. There was a statistically significant reduction in risk of recurrence in the irradiation group, with recurrence-free rates at 2 years of 88% versus 79% for the irradiation and no further treatment groups, respectively. Overall survival difference did not reach statistical difference (p = 0.074) (Fig. 66.31). Severe or life-threatening (GOG grade 3 or 4) adverse effects occurred in nine patients (6.6%) in the radiation therapy group and three (2.1%) in the observation group.

SundfΦr et al. (588) conducted a randomized study in which 122 patients with stage IIA and 20 with stage IIB cervix cancer were treated with intracavitary radium followed by either radical pelvic surgery including lymphadenectomy (group A, 72 patients) or EBRT (40 Gy) to the pelvis (group B, 70 patients). Postoperative RT (40 to 50 Gy) was given to patients in group A found to have node metastasis at operation. Fourteen patients in group A and 23 in group B died of recurrent cancer. The 10-year survival was 84% and 69%, respectively.

Treatment techniques may have an effect on outcome. Yamazaki et al. (667) compared 34 patients with cervical cancer treated with irregularly shaped four-field whole pelvis radiation therapy using CT simulation and 40 patients receiving whole pelvis EBRT with parallel-opposed fields in a nonrandomized study of postoperative radiation therapy consisting of 50 Gy in 25 fractions in 6 weeks. With a mean follow-up of 60 months, the actuarial 5-year pelvic tumor control was 94% with the two-field technique and 100% for the irregularly shaped four-field technique. The incidence of grade 2 or 3 bowel complications in the irregularly shaped technique group (2.9%, 1/34) was significantly lower than that in the two-field technique group (17.5%, 7/40; p <0.05).

Burnett et al. (62) described a prosthetic silicone plastic device that is filled with saline and Renografin for x-ray visualization (capacity between 750 and 1,500 mL) to conform to the pelvis and exclude the small bowel from the irradiated volume. The device remains in place throughout the radiation therapy course and is removed through a small incision after draining the contents of the prosthesis. Seven devices had been placed to date of the report. In the postoperative period, there was one pulmonary embolism. All seven patients completed planned radiation therapy. The devices have been removed with no adhesions to the prosthesis.

Many nonrandomized reports on postoperative pelvic irradiation have been published (376,510); some are highlighted here.

Snijders-Keilholtz et al. (563) described results in 233 women who underwent radical hysterectomy for stage I or IIA cervical carcinoma; 156 were treated with surgery alone, and 77 received adjuvant radiation therapy for tumor-related high-risk prognostic factors. The most important prognostic factor for survival and disease-free survival was pelvic lymph node positivity; additional factors were depth of invasion and positive

P.1582


surgical margins. Twelve patients recurred after surgery alone, all in the pelvis (100%). Of the 23 recurrences after surgery and adjuvant radiation therapy, 13 were in the pelvis (56%; p = 0.003). Ten patients with poor prognostic factors and negative nodes received adjuvant radiation therapy, and none of these patients recurred. The incidence of severe gastrointestinal radiation-related side effects was 2%. The incidence of lymphedema of the leg was 11%, which was similar to that in the surgery-alone group.

Garipagaoglu et al. (173) investigated prognostic factors in 100 patients with stage IB or IIA cervical carcinoma treated with radical hysterectomy and postoperative irradiation. The 5-year overall survival, disease-free survival, and pelvic tumor control rates were 83.6%, 82.8%, and 91.8%, respectively. Pelvic lymph node metastasis (p = 0.008), interval between surgery and irradiation (p = 0.001), overall radiation therapy time (p = 0.007), and tumor size (p = 0.028) were significant factors for pelvic tumor control as well as for overall survival.

Kahousen et al. (291) reported on a GOG prospectively randomized, multicenter trial in which patients with stage IB or IIB cervical cancer treated with radical hysterectomy who had pelvic lymph node metastases or vascular invasion randomly received adjuvant chemotherapy (400 mg/m2 carboplatin, and 30 mg bleomycin), or external pelvic radiation therapy, or no further treatment. After a median follow-up of 4.1 years (range, 2 to 7 years), there were no statistically significant differences (p = 0.9539) in disease-free survival rates among the three treatment arms, suggesting that adjuvant chemotherapy or radiation does not improve survival or recurrence rates in high-risk patients with cervical cancer after radical hysterectomy.

Gonzales et al. (182) reported that in 89 patients with stage IB or IIA cervical cancer with positive lymph nodes receiving postoperative irradiation, the 5- and 10-year survival rates were 60% and 51%, respectively. By comparison, 43 patients with negative lymph nodes had a survival rate of 85%. In the surviving patients, there were four gastrointestinal and seven genitourinary severe complications requiring surgical correction. In four patients, asymptomatic stenosis of the ureters was detected by IV pyelography performed routinely every year.

Bianchi et al. (37), in 60 patients receiving external irradiation for pelvic node metastasis after radical hysterectomy, observed a 65% 5-year survival rate. In contrast, in 15 patients who refused postoperative irradiation, only three (20%) survived 5 years.

Chatani et al. (76) reported on 128 patients with stage T1B to T2B carcinoma of the cervix who underwent radical hysterectomy with bilateral pelvic lymphadenectomy and postoperative EBRT. The 5-year local and distant failure rates were, respectively, 2% and 12% for negative nodes, 23% and 25% for one positive node, and 32% and 57%, for two or more positive nodes (p = 0.0029 and p = 0.0051, respectively). The 5-year cause-specific survival rates were 90%, 59%, and 42%, respectively (p = 0.0001). The most common complication was lymphedema of the lower extremity, experienced by half of the patients (42% at 5 years and 49% at 10 years).

Uno et al. (624) evaluated results of postradical hysterectomy irradiation in 98 patients with stage IB to IIB cervical cancer; all the patients had at least one pathologic risk factor for pelvic recurrence. The 5-year overall survival was 82%. There were pelvic recurrences in five cases and distant metastases in 15 cases. The 5-year overall survival rates for patients with or without pelvic lymph node metastasis were 76% and 89%, respectively (p = 0.018).

Kinney et al. (323) compared results of therapy in 82 patients with stage IB or IIA carcinoma of the cervix found to have pelvic lymph node metastases at Wertheim hysterectomy and bilateral lymphadenectomy without additional adjuvant therapy with 103 similar patients who received 50 Gy to the pelvis after surgery. The 5-year survival rate was 72% for the surgery-only patients and 64% for the group receiving adjuvant irradiation. The incidences of pelvic recurrences were 67% and 27%, respectively. The lack of impact on overall survival in the irradiated patients is most likely related to a higher incidence of distant metastases, which may be a reflection of higher short-term survival and high-risk patient selection.

Kinney et al. (324), in 117 patients treated with radical hysterectomy and pelvic lymphadenectomy, noted histologically proven nodal metastatic disease in 51 patients (44%; squamous cell in 35 and nonsquamous in 16). Nodal involvement was bilateral in 24 patients (47%). Para-aortic lymph node dissection was performed in 14 patients, and five had tumor involvement. Postoperative pelvic irradiation was administered to 29/51 patients (51.2 Gy, two fractions). Extended fields to the para-aortic area were used in six patients. The 5-year survival rates were 33% for the group receiving irradiation and 50% for the not irradiated group. Only one patient treated with postoperative irradiation had a pelvic failure, in contrast to seven patients not irradiated.

Inoue and Morita (269) described results in 72 patients treated with extended-field irradiation after radical surgery for nodal metastases from cervical cancer stage IB (37 patients), IIA (six patients), and IIB (29 patients). The median dose to para-aortic lymph nodes was 43.5 Gy and to the pelvis 45 Gy. The 5-year disease-free survival rates were 72% in 61 patients with squamous-cell carcinoma and 27% in 11 patients with nonsquamous-cell carcinoma. The 10-year disease-free survival rates were 88% for 22 patients with one positive node, 67% for 15 with two or three positive nodes, 64% for 16 with four to 17 positive nodes, and 20% for 10 patients with unresectable lymph nodes. Nineteen severe complications occurred in 17 patients; five were attributed to surgery, five to irradiation, and nine to both modalities. Four patients (5%) died of severe complications. Another six patients (8%) underwent major abdominal surgery for rectovaginal and ureterovaginal fistulas.

Mitsuhashi et al. (414) described an analysis of 108 patients with carcinoma of the cervix treated with postoperative EBRT to the pelvis followed by intravaginal cone boost with electron beam to the vaginal cuff. The 5-year cause-specific survival rates were 89% for 89 patients undergoing elective radiation therapy and 56% for 19 patients undergoing salvage irradiation (p <0.001). Recurrent tumors at the vaginal cuff were observed in only two patients in the elective irradiation group. Vesicovaginal fistula developed in four patients; only one patient had grade 2 rectal complications.

It appears that a modest gain in survival may be observed in patients with pelvic lymph node metastasis from carcinoma of the uterine cervix who receive irradiation after various types of surgery (182,526). More recently, studies involving a comparison of radiation versus radiation with chemotherapy have addressed this issue and are discussed later.

Postoperative Intracavitary High–Dose-Rate Brachytherapy

HDR brachytherapy after surgery is particularly suited for patients with cervical cancer because it prevents the prolonged immobilization required for LDR brachytherapy. In some patients at higher risk for parametrial tumor or lymph node metastases, HDR brachytherapy is combined with external-beam pelvic irradiation (20 Gy to the whole pelvis and additional 30 Gy to the parametria with midline shielding tailored to the geometry of the brachytherapy applications).

Hart et al. (234) described results in 83 patients who received postoperative RT for early stage cervical cancer with positive surgical margins, positive pelvic or para-aortic lymph nodes, lymphovascular space invasion, or deep stromal invasion, or for disease discovered incidentally at simple hysterectomy. Twenty-eight patients were treated with LDR brachytherapy with or without EBRT and 55 with EBRT to the pelvis and HDR intracavitary. Of these 83 patients, 66 were evaluable (20 LDR and

P.1583


46 HDR patients). Mean follow-up time was 101 months for the LDR group and 42 for the HDR group. The 5-year disease-free survival rate was 89% and 72%, local tumor control was 90% (18/20) and 89% (41/46), respectively. Three of 20 (15%) patients receiving LDR and 4/46 (9%) receiving HDR experienced grade 2 or 3 late treatment-related complications. No patient in either group had grade 4 or 5 complications.

Busch et al. (63) studied the outcome of 68 patients with cervical carcinoma—48 treated with radical hysterectomy and, because of risk factors, with postoperative RT (group 1), and 20 patients (group 2) pretreated with standard hysterectomy—admitted to the hospital for postoperative radiation therapy of the whole pelvis. Postoperative pelvic RT consisted of 39.6 Gy (box technique) and 6-Gy external-beam therapy to the pelvic lymph nodes, sparing the midline plus two HDR applications (7.5 Gy each), and survival, locoregional tumor control, and metastatic disease rates were nearly identical in both groups. Patients with positive lymph nodes had a worse prognosis (75% 3-year survival rate).

Atkovar et al. (25) described results in 126 patients treated with postoperative irradiation (median of 50 Gy in 5 weeks); 37 received vaginal cuff HDR brachytherapy (three fractions of 8 to 10 Gy at 5 mm, weekly). Overall and disease-free survival and locoregional tumor control rates were 71%, 69.9%, and 78.1%, respectively. Grade 2 and 3 complications developed in 5.5% of patients. Survival was the same in 67 patients treated with total abdominal hysterectomy and bilateral salpingo-oophorectomy and in 59 patients treated with radical hysterectomy and pelvic lymphadenectomy.