Race and ethnicity correlate with survival in patients with gastric adenocarcinoma.

BACKGROUND: Asian centers have consistently reported superior gastric cancer outcomes. Our study examines gastric cancer survival among different races and ethnicities in a large, heterogeneous USA population. PATIENTS AND METHODS: Patients with gastric adenocarcinoma treated in Los Angeles County from 1988 to 2006 were identified from the Los Angeles County Cancer Surveillance Program. Patients were categorized by race and ethnicity as White, Asian, Hispanic and Black. RESULTS: Of 13 084 patients, 39% were White, 22% Asian, 28% Hispanic, 11% Black and 2% other. Asian patients demonstrated higher survival than Whites, Hispanics and Blacks [median survival (MS) 16.3 versus 8.4, 8.7 and 7.9 months, respectively; log-rank P values < 0.001]. Multivariate Cox regression analysis showed that Asians had improved probability of survival [hazard ratio (HR) 0.76, 95% confidence interval (CI) 0.72-0.82; P < 0.001]. In patients who underwent curative-intent surgery, Asian patients demonstrated higher survival than Whites, Hispanics and Blacks (MS 32.7 versus 18.8, 19.9 and 18.9 months, respectively; log-rank P values < 0.001). Multivariate Cox regression analysis showed that Asians had improved probability of survival after surgery (HR 0.79, 95% CI 0.71-0.88; P < 0.001). CONCLUSIONS: Asians with gastric adenocarcinoma have superior outcomes in Los Angeles County. These outcomes verify disparities in gastric cancer survival among different races and ethnicities independent of established clinical and pathologic factors.

Robotic-assisted laparoscopic low anterior resection with total mesorectal excision for rectal cancer.

BACKGROUND: With advanced stereoscopic vision, lack of tremor, and the ability to rotate the instruments surgeons find that robotic systems are ideal laparoscopic tools. Because of its high operating cost, however, robotic surgery should be reserved to procedures in which the technology can be of maximum benefit, usually when precise dissections in confined spaces are required. Because conventional laparoscopic total mesorectal excision is a challenging procedure, we have sought to assess the utility of the DaVinci robotic system in laparoscopic low anterior resections for cancer of the rectum. METHODS: Between November 2004 and May 2005 robotic-assisted low anterior resection with total mesorectal excision was performed on six consecutive patients with rectal cancer. These cases were compared with six consecutive low anterior resections performed with conventional laparoscopic techniques by the same surgeon. RESULTS: There were no conversions in either group. Operative and pathological data, complications, and hospital stay were similar in the two groups. Robotic operations appeared to cause less strain for the surgeon. CONCLUSIONS: Robotic-assisted laparoscopic low anterior resection for rectal cancer is feasible in experienced hands. This technique may facilitate minimally invasive radical rectal surgery.

Asymptomatic renal neoplasms in the rectal cancer patient.

Although cancers of the rectum and kidney are common malignancies the incidence of coexistent rectal and renal primary tumors is unclear. Our objective was to determine the true incidence of synchronous neoplasms of the rectum and kidney. The computed tumor registry database at the City of Hope National Medical Center was queried for patients with synchronous rectal cancer and renal neoplasms presenting between August 1990 and August 2000. During the 10-year period there were 182 patients presenting for treatment of rectal carcinoma. Of these seven (3.8%) were found to have an asymptomatic renal neoplasm. Four patients underwent synchronous resection. Three patients underwent staged renal and rectal resections. The pathology of the renal lesions included renal cell carcinoma in six and an oncocytoma in one patient. Rectal lesions were all adenocarcinomas and all were within 10 cm of the dentate line. Three patients required abdominoperineal resections and four were treated with low anterior resections. Two patients presented with hepatic metastasis at the time of diagnosis. Five patients remain free of disease. Two patients died of persistent and recurrent disease 6 months and 40 months after operation. With the exception of one patient who required prolonged intubation because of severe Parkinson's disease there were no major complications after simultaneous resection of both renal and rectal disease. Simultaneous asymptomatic renal neoplasms may be found in up to 3.8 per cent of patients with rectal cancer. Synchronous lesions may be treated simultaneously without significant morbidity.

Transanal excision for low rectal cancers is curative in early-stage disease with favorable histology.

Controversy still exists as to the optimal treatment of early-stage low rectal cancers. Standard resections such as abdominal perineal resections or low anterior resection with coloanal anastomosis can be associated with significant morbidity. Local excision has been considered as a potentially far less morbid option for these patients. We identified 20 patients from our prospective tumor registry database that underwent transanal resection of early rectal cancers between 1988 and 1998. Five patients had carcinoma in situ, and 15 patients had T1 lesions. All T1 lesions were well or moderately well differentiated, and none had lymphovascular invasion. The mean tumor size was 2.65 cm, and the average distance from the anal verge was 4.55 cm. One patient required temporary colostomy because of postoperative complications. Four patients had postoperative adjuvant therapy (radiation alone, two; radiation and chemotherapy, two) for close margins after they refused an abdominal perineal resection. With a median follow-up of 60 months there were no recurrences and no cancer-related deaths. This series suggests that transanal excision is a safe and effective treatment for selected early low rectal cancers with favorable histology.

Esthetic reconstruction after mastectomy for inflammatory breast cancer: is it worthwhile?

BACKGROUND: Because inflammatory breast cancer (IBC) has been viewed as a malignancy with a poor likelihood of longterm survival, few women have been offered esthetic reconstruction after mastectomy for IBC. Recent advances in multimodality therapy have improved the outcomes for women with this disease. The purpose of this review was to assess the results of esthetic breast reconstruction in the population with IBC. STUDY DESIGN: Review of medical records at the City of Hope National Medical Center for the 10-year period ending in May 1997, revealed 23 women who underwent elective esthetic breast reconstruction after mastectomy for IBC. The records of these patients were reviewed retrospectively. Patients requiring reconstruction for large surgical chest wall defects were not included in the review. RESULTS: Treatment for IBC included mastectomy in all patients, chemotherapy in 22, and chest wall radiation therapy in 14. Immediate reconstruction was performed at the time of mastectomy (n = 14) or was delayed (n = 9). The types of reconstruction included transverse rectus abdominis musculocutaneous flap (n = 18), latissimus dorsi flap (n = 2), or prosthetic mammary implant reconstruction (n = 3). Seven women chose to undergo additional reconstruction procedures (ie, nipple reconstruction) after their initial reconstruction. With a median followup of 44 months for survivors, 16 patients developed recurrence after reconstruction. Of these, 6 were local recurrences and 10 were distant failures. Seven patients are currently alive with no evidence of disease, 4 are currently alive with disease, and 12 have died as a result of breast cancer. The median disease-free survival after reconstruction was 19 months. The median overall survival after reconstruction for all patients was 22 months. The only negative predictor of survival was a positive surgical margin at mastectomy. CONCLUSIONS: The significant emotional and esthetic benefits of breast reconstruction should be available to women with IBC. In light of the improving prognosis of IBC with current aggressive multimodality treatment, reconstructive procedures should be offered as part of comprehensive therapy.

Targeting of human p53-overexpressing tumor cells by an HLA A*0201-restricted murine T-cell receptor expressed in Jurkat T cells.

A potent anti-human (hu) p53 CD8+ CTL response develops in HLA A*0201 transgenic (Tg) mice after immunization with peptides corresponding to HLA A*0201 motifs from hu p53. Mice immunized with the hu P53(149-157) peptide develop a CTL response that is of moderately high affinity and is capable of recognizing hu tumor cells expressing mutated p53. In this report, the mRNAs encoding the predominantly expressed T-cell receptor (TCR) sequences were molecularly cloned from a murine (mu) CTL clone derived from immunized Tg mice, which recognized endogenously processed hu p53 restricted by HLA A*0201. The separate A and B chain TCR cDNAs were transfected in the corresponding TCR A- and B- Jurkat-CD3- mutant T-cell lines, and each rescued CD3 surface expression. Both TCR chains were simultaneously introduced into Jurkat-CD3+ cells, and the transfected Jurkat cells recognized hu T2 cells sensitized with the p53(149-157) CTL epitope but not T2 cells sensitized with a nonspecific CTL epitope. Breast, pancreatic, and sarcoma tumor cell lines, which overexpress endogenous mutated p53, were recognized in the presence of anti-CD28 costimulation, only if they also expressed HLA A*0201. Normal hu fibroblasts established from skin cultures were not recognized. These results represent the first time that a p53-specific TCR capable of recognizing hu cancer cells was heterologously expressed in a naive recipient cell, converting that cell to one recognizing hu tumor cells with mutated p53. This TCR represents a candidate molecule for a genetic strategy in combating hu cancer by an adoptive immunotherapy approach, which uses the strong xenorecognition of hu p53 in mice.

Contralateral prophylactic mastectomy improves the outcome of selected patients undergoing mastectomy for breast cancer.

BACKGROUND: Risk factors for contralateral breast cancer (CBC) may indicate a benefit for contralateral prophylactic mastectomy (CPM) at the time of unilateral mastectomy for breast cancer. The purpose of this study is to evaluate the efficacy of CPM in preventing CBC. METHODS: sixty-four patients undergoing CPM and a control group of 182 patients not undergoing CPM and matched for age, stage, surgery, chemotherapy, and hormonal therapy were retrospectively compared for CBC rate, disease-free survival, and overall survival. RESULTS: Thirty-six CBCs occurred in the control group. In the CPM group, 3 CBCs were found at the time of prophylactic mastectomy, but none occurred subsequently (P = 0.005). Disease-free survival at 15 years in the CPM group was 55% (95% confidence interval [CI] 38% to 69%) versus 28% (95% CI 19% to 36%) in the control group (P = 0.01). Overall survival at 15 years was 64% (95% CI 45% to 78%) CPM versus 48% (95% CI 39% to 58%) in controls (P = 0.26). CONCLUSION: CPM prevented CBC and significantly prolonged disease-free survival. Future studies will need to address risk assessment and contralateral breast cancer prevention in patients treated for early breast cancer.

Immunotherapy of bladder cancer targeting P53.

PURPOSE: Superficial bladder cancer is often responsive to immunotherapy with bacillus Calmette-Guerin (BCG). However, some tumors progress despite BCG treatment, and most of these have mutations in the p53 tumor suppressor gene resulting in its over-expression. Overexpressed p53 is therefore a potential target for immunotherapy. The objective of this study was to demonstrate whether human bladder cancer xenografts in SCID mice could be eliminated by cytotoxic T cells (CTL) which recognize over-expressed p53. MATERIALS AND METHODS: Murine CTL which are specific for human p53 were previously generated in our laboratory by peptide immunization of HLA A2.1 transgenic mice. These CTL recognize and lyse human tumor cell lines which over-express p53 in the context of HLA A2.1. The p53 over-expressing HLA A2.1+ human bladder cancer cell line J82 was used to establish subcutaneous or intravesicular tumors in SCID mice. The mice were then administered tail vein injections of 5 x 10(7) p53-specific CTL, control CTL, or phosphate buffered saline (PBS). RESULTS: The subcutaneous tumor mean volume at 5 weeks for the p53-specific CTL treatment group was significantly lower than for both the control CTL or the PBS group (32 mm.3 versus 185 mm.3, p = 0.04 and 32 mm.3 versus 418 mm.3, p = 0.0001). In the mice with intravesicular tumors, a reduction to nonpalpable tumor size in vivo was seen with specific CTL therapy (14% palpable) versus control CTL treatment (86% palpable), the final tumor volume at necropsy was 127 mm.3 versus 246 mm.3 (N.S.). CONCLUSION: The overall response of the human bladder tumors in the SCID mouse model suggests the possibility of targeting p53 in patients with bladder cancer.

A mortality-free decade of pancreatoduodenectomy: is quality independent of quantity?

Pancreatoduodenectomy (PD) for periampullary cancer is a procedure of high morbidity and poor long-term survival. Superior clinical outcome has been described in high-volume institutions or for surgeons with a high case load. All patients undergoing pancreatectomy at the City of Hope National Medical Center (Duarte, CA) between 1987 and 1998 were analyzed retrospectively for postoperative outcome, and correlating or predictive clinicopathological factors were identified. Fifty-four patients underwent pancreatectomy [PD, n = 43; pylorus-preserving PD, n = 8; total pancreatectomy, n = 3]. There were 26 males and 28 females, with a median age of 63 years (range, 19-86). Fifty patients had a malignant diagnosis, and four patients had a benign diagnosis. Nine surgical oncologists performed an average of six pancreatectomies (range, 2-8). There was no perioperative death. Postoperative complications occurred in 30 patients, and infections predominated (n = 17). The median hospital stay was 16.5 days. The median postoperative actuarial survival by cancer site was 56 months (ampullary/ bile duct), 32.5 months (duodenal), 22.5 months (pancreatic), and 23.2 months (others). In this 11-year single institutional experience, PD and total pancreatectomy have been performed without lethal complication. In the setting of an exclusive oncology practice, operative mortality rates and survival outcome can be generated that compare favorably to large center experiences. Quality of outcome after pancreatectomy can be independent of quantity.

An orthotopic in vivo model of human pancreatic cancer.

BACKGROUND: Pancreatic cancer is a highly lethal disease that frequently presents in advanced stages. For most patients, treatment with great clinical efficacy does not exist. Relevant in vivo models to test novel therapies are highly desirable. METHODS: The human pancreatic ductal adenocarcinoma cell line Panc-1 was injected intraperitoneally into SCID mice. The pattern of the resulting peripancreatic as well as metastatic disease was examined. Survival experiments after chemotherapy with gemcitabine or doxorubicin, and after immunotherapy with p53-specific cytotoxic T lymphocytes were performed. RESULTS: All animals developed isolated pancreatic tumor implants within 48 hours after injection. After the formation of invasive pancreatic tumor nodules, peripancreatic and portal adenopathy developed, causing biliary obstruction. All tumor-bearing animals died of disease within 5 to 12 weeks. Survival after gemcitabine treatment and after p53-CTL injection was significantly prolonged, with some animals remaining tumor-free. Doxorubicin treatment did not yield extended survival, but led to significant toxicity. CONCLUSION: Intraperitoneal injection of Panc-1 cells into SCID mice produces a quasi-orthotopic tumor development model that shares many characteristics with human pancreatic cancer. The ease of cell injection, avoidance of cumbersome surgical intervention with its resulting mortality, and the reliable development of obstructive jaundice as a dependent comorbid factor render this a useful model for in vivo testing of novel therapeutic approaches to pancreatic cancer. Our initial therapeutic studies demonstrate that in vitro antitumor efficacy against Panc-1 cancer cells does not necessarily predict the in vivo response, highlighting the preclinical experimental value of this model.

Beyond palliative mastectomy in inflammatory breast cancer--a reassessment of margin status.

BACKGROUND: Inflammatory breast cancer is a locally advanced tumor with an aggressive local and systemic course. Treatment of this disease has been evolving over the last several decades. The aim of this study was to assess whether current therapies, both surgical and chemotherapeutic, are providing better local control (LC) and overall survival (OS). We also attempted to identify clinical and pathologic factors that may be associated with improved OS, disease-free survival (DFS), and LC. METHODS: A 25-year retrospective review performed at the City of Hope National Medical Center identified 90 patients with the diagnosis of inflammatory breast cancer. RESULTS: Of the 90 patients identified with inflammatory breast cancer, 33 received neoadjuvant therapy (NEO) consisting of chemotherapy followed by surgery with radiation (n = 26) and without radiation (n = 7). Fifty-seven patients received other therapies (nonNEO). Treatments received by the nonNEO group consisted of chemotherapy, radiation, mastectomy, adrenalectomy, and oophorectomy, alone or in combination. The median follow-up was 28.9 months for the NEO group and 17.6 months for the nonNEO group. Borderline significant differences in the OS distributions between the two groups were found (P = .10), with 3- and 5-year OS for the NEO group of 40.0% and 29.9% and for the nonNEO group of 24.7% and 16.5%, respectively. DFS and LC were comparable in the two groups. Lower stage was associated with an improved OS (P < .05). The 5-year OS for stage IIIB was 30.9%, compared to 7.8% for stage IV. In those patients with stage III disease who were treated with mastectomy and rendered free of disease, margin status was identified by univariate analysis to be a prognostic indicator for OS (P < .05). The 3-year OS, DFS, and LC for patients with negative margins were 47.4%, 37.5%, and 60.3%, respectively, compared to 0%, 16.7%, and 31.3% in patients with positive margins. CONCLUSIONS: This study suggests that in patients with inflammatory breast cancer and nonmetastatic disease, an aggressive surgical approach may be justified with the goal of a negative surgical margin. Achievement of this local control is associated with a better overall outcome for this subset of patients. The ability to obtain negative margins may further identify a group of patients with a less aggressive tumor biology that may be more responsive to other modalities of therapy.

Targeting p53 for adoptive T-cell immunotherapy.

p53 gene mutations occur in most human cancers and result in an altered protein product that accumulates within the cell. Although the observed endogenous human CTL response to p53 is weak, high-affinity, human p53-specific CTLs have been generated from HLA A2.1 transgenic mice immunized with human CTL epitope peptides. In this study, we examine the ability of HLA A2.1-restricted and human p53-specific CTLs from HLA A2.1 transgenic mice to suppress the growth of p53-overexpressing human tumors in severe combined immunodeficient (SCID) mice. In vitro, murine p53(149-157)-specific CTLs selectively lysed the p53-overexpressing pancreatic carcinoma cell line Panc-1 but did not recognize HLA A2.1- tumor cells or HLA A2.1+ normal human fibroblasts. Furthermore, in vivo, the growth of established human tumor xenografts in SCID mice was significantly reduced and survival was prolonged after the administration of p53-specific CTLs but not after the administration of control CTLs or PBS alone. Following treatment with p53(149-157)-specific CTLs, regressing Panc-1 tumors were infiltrated by the CD8+ CTLs, as demonstrated by immunohistochemistry. These findings suggest that p53(149-157)-specific and HLA A2.1-restricted murine CTLs suppress the growth of established Panc-1 tumors following adoptive transfer into SCID hosts and prolong their survival.

An HLA-restricted, p53 specific immune response from HLA transgenic p53 knockout mice.

BACKGROUND: p53 is over-expressed in most human malignancies and is therefore an attractive target for immunotherapy. Unfortunately, a human cytotoxic T cell response to p53 is difficult to generate. p53 knockout transgenic mice may provide a model to circumvent immunologic tolerance to p53 and develop high-affinity p53-specific cytotoxic T lymphocytes (CTL). METHODS: p53 knockout, HLA A2.1 transgenic mice were generated and immunized with the immunodominant wild-type p53 nonamer peptide epitope p53149-157. Two weeks later splenocytes were harvested and stimulated in vitro with acid-treated, p53 peptide-pulsed syngeneic blast cells. Cultures were restimulated weekly with acid-treated, p53 peptide-pulsed Jurkat cells transfected with the HLA A2.1 gene. Peptide-specific cytotoxic activity was measured by chromium release assay, and the resulting CD8+ effectors were cloned via limiting dilution. RESULTS: P53 peptide-specific CTL were generated against p53149-157. Clones generated from the p53149-157 cell line demonstrated high affinity and specificity for p53149-157 when presented by HLA A2.1+ antigen-presenting cells. The p53149-157 CTL killed only cells overexpressing p53 cells that were HLA A2.1+ and did not kill cells with normal levels of p53 expression or those that were HLA A2.1-. CONCLUSION: HLA transgenic mice not previously exposed to the p53 protein provide a useful model for generating high-affinity p53-specific CTL.

Surgical management of thyroid cancer invading the airway.

BACKGROUND: Locally advanced thyroid cancer invading the tracheal cartilage represents a difficult treatment dilemma during thyroidectomy. METHODS: A retrospective chart review was performed to determine the results of laryngotracheal resection or tracheal cartilage shave with adjuvant radiotherapy in patients with locally advanced thyroid cancer invading the upper airway. RESULTS: Of 597 patients undergoing thyroidectomy for thyroid cancer, 40 were found to have laryngotracheal invasion. Thirty-five patients with superficial invasion underwent cartilage shave procedures with adjuvant radiotherapy; five with full-thickness invasion underwent radical resection, including tracheal sleeve resection (n = 3) or total laryngectomy (n = 2). Histologic subtypes included papillary (n = 32), follicular (n = 2), Hurthle cell (n = 1), medullary (n = 3), and anaplastic (n = 2). Of the cartilage shave group, 25 are currently alive with no evidence of disease at a mean follow-up of 81 months (range 1-290). Six developed isolated local/regional recurrence and were managed with total laryngectomy (n = 1), tracheal resection (n = 1), cervical lymphadenectomy (n = 1), or repeat radiotherapy (n = 3). All six patients remain free of disease at a mean follow-up of 5 years. Of those who underwent initial laryngotracheal resection, four remain free of disease at a mean follow-up of 5 years. The rates of 10-year disease-free survival and overall survival for all patients were 47.9% (95% confidence interval [CI] 24.8, 71.0) and 83.9% (95% CI 70.3, 97.5), respectively. CONCLUSIONS: These data suggest that adequate management of thyroid cancer with laryngotracheal invasion can be achieved with a more conservative surgical approach and adjuvant radiotherapy, reserving more radical resections for extensive primary lesions or locally recurrent disease.

The use of transgenic mice to generate high affinity p53 specific cytolytic T cells.

P53 is an attractive target immunotherapy because it is overexpressed in up to one half of all malignancies, and its overexpression often correlates with a worsened prognosis. We wanted to determine the feasibility of targeting wild-type epitopes p53 on human tumor cells. HLA A2.1 transgenic mice were immunized with the immunodominant wild-type p53 peptide epitopes, p53(149-157) and p53(264-272), along with a pan-DR helper epitope peptide in incomplete Freund's adjuvant (IFA). Twelve days later, splenocytes were harvested and stimulated with syngeneic blast cells that had been acid-treated to remove endogenous peptide and p53 peptide-pulsed. The responding cells were subsequently restimulated weekly with acid washed, peptide-pulsed Jurkat cells transfected with HLA A2.1. Peptide specific activity was tested in a chromium release assay. The resulting cytotoxic T cells (CTL) were cloned by limiting dilution. Peptide specific CTL were generated against both p53(149-157) and p53(264-272. Only p53(149-157) specific CTL were able to recognize and lyse cells that overexpressed endogenous p53. CTL clones derived from the p53(149-157) cell line demonstrated high affinity and specificity for p53(149-157) when presented by HLA A2.1+ cells. The p53(149-157) specific CTL were tested for specificity against a variety of cultured human cell lines. The CTL clones only lysed cells that overexpressed p53 in the context of HLA A2.1 and did not lyse cells with normal p53 expression or cells that lacked HLA A2.1 expression. This study demonstrates the possibility of targeting tumors, which overexpress p53, and raises the possibility transferring the high affinity, p53 specific T cell receptors from the murine CTL to human T cells.

Local recurrence in breast cancer: implications for systemic disease.

BACKGROUND: Recurrence in breast carcinoma follows a pattern of growth marked by local, regional, or widespread dissemination. Local recurrence may be the harbinger of systemic disease or failure of local control. Delineation of these processes may have implications in treatment. METHODS: A retrospective review found 1,171 patients with stages I and II breast cancer from 1978 to 1990 treated at the City of Hope Medical Center. RESULTS: Twenty-seven percent (n = 313) of patients developed recurrences. These were classified as local, including chest wall and regional nodes (n = 40), local and distant (n = 63), and distant (n = 210). Mean follow-up was 60 months. Multivariate analysis demonstrates tumor size was not different between the three groups, but the presence of positive lymph nodes was: local = 51%, local and distant = 78%, and distant = 64%. The disease-free interval was longest in the local group (42 months) versus the local and distant group (23 months) and distant group (39 months). Median survival was calculated from the time of recurrence: local = 90 months, local and distant = 26 months, and distant = 16 months. CONCLUSIONS: A group of patients with local recurrence have improved survival and do not develop distant disease. This group may benefit from aggressive surgical treatment to control local disease. These data suggest that a subset of breast tumors can act locally aggressive without metastatic potential.

Treatment of perianal infection following bone marrow transplantation.

PURPOSE: Bone marrow transplantation (BMT) is often associated with profound neutropenia. Allogeneic transplant recipients also have defects in both humoral and cellular immunity and thus are subject to increased risk of serious, often life-threatening, infection even beyond the period of granulocyte recovery. The current study was undertaken to evaluate patients who required operative intervention for perianal sepsis following BMT. METHODS: The bone marrow transplant database at a single institution was used to identify all patients diagnosed with perianal infections after autologous or allogeneic BMT. Charts were reviewed in a retrospective manner. RESULTS: Over a ten-year period ending in November 1993, 963 BMT were performed at the City of Hope National Medical Center. Twenty-four patients were diagnosed with perianal infections following their transplants. Fifteen patients did not have purulent collections requiring drainage and were treated with antibiotics and supportive measures alone. Nine patients (37.5 percent) required surgical intervention between 10 and 380 days following transplantation. At the time of surgical intervention, seven patients had purulent collections and two patients had acute and chronic inflammation, tissue necrosis, and fibrosis. Of the two patients with an absolute neutrophil count less than 1,000, a purulent collection was found in one of the patients. Cultures taken from perianal abscesses were almost all polymicrobial, and the most common organisms were Escherichia coli, Bacteroides, Enterococcus, and Klebsiella. For those patients undergoing surgical intervention, mean time to complete wound closure by secondary intention was 37.6 days; five patients healed in less than 15 days, two patients healed at 93 and 114 days, and two patients had persistent, open wounds at time of death, which was unrelated to their perianal disease. Five patients were receiving systemic steroids at time of surgical intervention; this did not appear to affect time to wound healing. CONCLUSIONS: Perianal infections are a rare complication of BMT. Majority of these infections are polymicrobial, and organisms isolated are similar to those seen in the perianal infections of nonimmunosuppressed patients. Despite steroid use, granulocytopenia does not exclude the possible presence of purulent collections, and clinical examination should guide the decision for surgical drainage. In general, perianal wound healing is not prolonged in BMT patients.

Survey of physicians' attitudes relative to the field of surgical oncology.

BACKGROUND: Surgical oncology as a distinct field of expertise is fairly young. The current study was designed to gain a better understanding of the attitude of practicing physicians toward the field of surgical oncology. METHODS: Three hundred twenty-seven physicians in the San Gabriel Valley (a suburban area adjacent to Los Angeles) responded to an anonymous survey of opinions regarding surgical oncology. Responses were placed into a computerized database. RESULTS: Of those responding, 179 were primary care physicians, 52 were general surgeons, 78 were gynecologists, and 18 were medical oncologists. Overall, 89% of physicians were familiar with the field of surgical oncology, but only 47% had ever heard of The Society of Surgical Oncology (SSO). Ninety-four percent of the respondents felt that a surgical oncologist should care for patients with complex cancer, and 63% of respondents felt that surgical oncologists should care only for patients with complex cancer. Familiarity with the field of surgical oncology and with the SSO correlated with the percentage of the physicians practice that was cancer related. Only 22% of physicians felt that the field of surgical oncology is redundant to the general surgical specialties. CONCLUSIONS: Results of the survey indicate that there is considerable recognition of the unique expertise of the surgical oncologist by the medical community. Unfortunately, many physicians are not familiar with the SSO. Educating physicians in the community about the SSO may help to further expand the role of the surgical oncologist in the care of the patient with cancer, standardize the expectations of the skills and training of a surgical oncologist, and set a benchmark for the surgical subspecialty.

Salvage abdominoperineal resection following combined chemotherapy and radiotherapy for epidermoid carcinoma of the anus.

BACKGROUND: Up to one-third of patients with anal epidermoid cancer will fail initial chemoradiotherapy (CT-RT) or have local recurrence after treatment. This study evaluates the Memorial Sloan-Kettering Cancer Center (MSKCC) experience with salvage abdominoperineal resection (APR) in these patients. METHODS: Thirty-eight patients who underwent salvage APR following 5-fluorouracil (5-FU), mitomycin C, and radiotherapy over the past 12 years were analyzed by retrospective review. Survival was calculated by the Kaplan-Meier method and comparisons by log-rank analysis. RESULTS: The indications for APR were recurrent disease after CT-RT in 14 patients and persistent disease in 24 patients. Median follow-up time and survival were 47 and 41 months, respectively. The actuarial 5-year survival was 44%. Twenty-three patients had recurrent disease after APR. Inguinal lymphadenopathy at initial presentation (p < 0.05), fixation of tumor to the pelvic sidewall (p < 0.01), and pathologic involvement of the perirectal fat (p < 0.01) adversely affected survival. Age, gender, initial response to CT-RT, initial stage of the primary tumor, histologic levator muscle involvement, status of perirectal lymph nodes, and extent of lymphadenectomy did not affect survival. CONCLUSIONS: Salvage APR can be expected to yield a moderate number of long-term survivors, but the high rate of disseminated failure suggests the need for additional postoperative treatment.