According to the American Cancer Society, esophageal cancer will be diagnosed in about 13,200 U.S. patients this year, and 12,500 will die of the disease. Its incidence, while shifting toward adenocarcinomas and away from squamous cell carcinomas, is rising. Yet it remains too rare in industrialized countries to propel the kind of widespread screening credited with lowering, for example, colon cancer mortality in the U.S.
As a result, most esophageal cancers are discovered too late to save the patient. Survival, which depends greatly on the stage and penetration of the disease, is only 12% at five years for patients who are able to undergo resection. For all patients, the mean survival time is five months from diagnosis. Chemo- and radiotherapy followed by resection currently offers the best hope for long-term survival. Earlier diagnosis, more accurate staging, and better management will all be needed to improve the prognosis.
In recent years, FDG-PET has gained acceptance as a useful tool for evaluating esophageal cancer, both in initial staging and post-therapy reassessment. PET isn't perfect: it misses some primary tumors, especially in the distal esophagus, and works best in concert with other imaging modalities such as CT and ultrasound. But PET's strengths, including its ability to spot unsuspected metastases and help evaluate treatment response, have generated research aimed at finding its optimal role in esophageal cancer.
At the 2001 Society of Nuclear Medicine meeting in Toronto, several groups reported encouraging results. Among them, Dr. Ching-Yee Wong and colleagues from the PET center at William Beaumont Hospital in Royal Oak, MI, evaluated 19 esophageal cancer patients (15 adenocarcinomas and 4 others) with both PET and intraluminal ultrasound.
"There is a great need for accurate staging, and staging by CT is poor," Wong said, citing tumor (T) sensitivity of 75%, and nodal (N) sensitivity of 50%, respectively, for CT in esophageal cancer. Ultrasound does somewhat better, at about 90% (T) and 70% (N), he said. Thoracoscopy and laparoscopy reach about 90% sensitivity for both T and N, but are invasive and expensive.
"MRI is worse than CT," he said, citing lower sensitivity for all modalities except PET in the detection of metastatic (M) disease.
The 19 patients in the study (14 male, 5 female, mean age 64) underwent both intraluminal ultrasound (GF UM130 radial echoendoscope, Olympus, Lake Success, NY) and FDG-PET imaging (ECAT 951-31R, Siemens, Hoffman Estates, IL; or Advance, GE Medical Systems, Waukesha, WI).In all, PET visualized 17 of 19 primary lesions. The two lesions without FDG uptake included 1 superficial cancer and 1 mucinous adenocarcinoma. PET also detected 2 positive and 3 negative paraesophageal/perigastric nodes correctly, Wong said, but missed these nodes in 14 patients (74%) compared to ultrasound and surgery (p<0.0001).
In 3 patients, PET also showed unexpected metastases outside the chest region, which were later confirmed by CT, thus avoiding unnecessary surgery in these patients, Wong said.
PET plays an effective complementary role in the staging of esophageal cancer, he said. While ultrasound is superior for T and N staging, PET is best for M staging.
"FDG-PET identified most of the primary lesions and distant metastases," Wong said. "Serious limitations exist in detecting paraesophageal or perigastric nodes, which require ultrasound. In about one-tenth of patients there is no uptake in the primary lesion, hence staging is limited when there is a lack of uptake in the primary tumor."
In discussions following the talk, Wong said there were no established criteria for what constitutes a positive finding in ultrasound.
Gauging treatment response
Next, Dr. Mitsuaki Tatsumi and colleagues from the Osaka University Graduate School of Medicine in Japan sought to gauge the usefulness of PET in evaluating esophageal cancer activity after chemotherapy or chemoradiotherapy, comparing the results to those obtained with CT or intraluminal ultrasound.
Results in all 20 patients were confirmed either by histopathology (Group I, 10 patients who underwent surgery after PET), or by follow-up with conventional imaging (Group II, 10 patients treated only with chemotherapy or chemoradiotherapy).
Following surgery in the Group I patients, PET was 70% accurate in detecting residual disease at the primary sites, and 60% accurate at the lymph-node sites. PET underestimated 2 primary sites in 2 patients, and 5 lymph-node sites in 4 patients, according to Tatsumi. Standardized uptake values (SUV) correlated well with the histopathological tumor grade at 3 primary sites and 5 lymph-node sites, he said.
In comparison, conventional imaging was 70% and 50% accurate in primary and lymph-node sites, respectively. It overestimated 3 primary sites and 3 abdominal lymph-node sites. Compared with conventional imaging, PET was more accurate in 1 patient, the same in 3, and less accurate in 1.
In the nonsurgical patients (Group II), PET was 90% accurate at both primary and lymph-node sites, compared with conventional imaging, which was 70% and 90% accurate at the primary and lymph-node sites, respectively. PET and conventional imaging results were concordant in 8 primary-tumor patients and at all 10 lymph-node sites.
Confirming PET's predictive value in esophageal cancer, all 4 patients with SUV greater than 2.5 (3 primary, 1 lymph node) were found to have progressive disease at those sites within 3 months. For predicting treatment response, PET was more accurate than conventional imaging in two patients, and produced the same results in 5.
"PET is useful in the evaluation of residual disease activity after treatment in patients with esophageal cancer," Tatsumi concluded. "Despite relatively poor sensitivity to tiny lesions, it is able to provide additional or actual information about residual tissue that is equivocal on conventional imaging." It may also allow for the prediction of prognosis following treatment for esophageal cancer, he said.
A group of French researchers reached similar conclusions with 3-D PET, while adding a caveat. Dr. Denis Agostini from the Service de Medicine Nucléaire in Caen presented a study of 18 male patients with proven esophageal cancer, both squamous cell carcinomas and adenocarcinomas, who underwent whole-body 3-D FDG-PET before and four weeks after a combined chemo- and radiotherapy regimen lasting five weeks.
The researchers used a Siemens ECAT Exact HR+ with images acquired 50 minutes after injection of 111 MBq of FDG. Iterative reconstruction was used with attenuation correction (software version 7.2).
The technique detected all known primary tumors, and showed 9 distant metastases that could not be seen on CT. Following therapy, the mean SUV was reduced significantly, Agostini said, from 6 ± 2 to 3 ± 2, and the number of distant nodes and metastases decreased from 1.8 ± to 0.2 ± 0.4.
The technique showed an effective reduction of cancer progression following therapy. However, in patients who underwent subsequent surgery, "an important message is that the decrease in FDG uptake doesn't correlate with the decrease of histopathologic response," Agostini said. "Further studies are needed to assess the histopathologic correlation with quantitative measurements of tumor FDG uptake."
The next study, presented by Dr. Nuri Aslan from the Washington University School of Medicine in St. Louis, sought to answer a similar question. The group analyzed tumor volume and SUV uptake to determine whether FDG PET could differentiate between inflammation and persistent tumor following chemo- and radiotherapy treatment.
"Response to cancer therapy is usually [measured] by the change in tumor size. However, none of the conventional imaging techniques can reliably differentiate viable tumor from post-therapy changes," Arslan said. "Reduction in tumor FDG uptake after successful therapy usually precedes morphological changes."
The team evaluated treatment response in 24 patients with esophageal carcinoma. Images were acquired before and four weeks after therapy using the same scanner model as the French group.
Within a month of acquiring the post-treatment images, 20 patients underwent surgical resection, Arslan said.
A 3-D analysis of tumor volume, peak SUV uptake and average SUV uptake within the tumor volume was performed using an edge-detector threshold of 40% of peak tumor uptake. The team obtained pre- and post-treatment values for all three indices and computed the difference.
Arslan concluded that while quantitative measurement of peak and average SUV predicts the presence of distant disease following neoadjuvant therapy, quantitative evaluation at the primary tumor site cannot differentiate between persistent tumor and post-therapy inflammation.
By Eric BarnesAuntMinnie.com staff writer
August 2, 2001
Related Reading
Chemoradiotherapy before surgery improves survival in esophageal carcinoma, July 31, 2001
FDG-PET improves prognosis of esophageal cancer patients, May 15, 2001
FDG-PET effective for diagnosing recurrent esophageal cancer, January 4, 2001
Copyright © 2001 AuntMinnie.com
