Unexpected neoplasia in autopsies: potential implications for tissue and organ safety.
Abstract: * Context.--Medical examiner cases are increasingly used as tissue donor referral sources to meet the ever-growing need for transplant tissues. The assumption is often made that traumatic and sudden deaths have minimal risk of unsuspected neoplasia. An autopsy of a registered tissue donor with strong preautopsy clinical assessment of a saddle pulmonary embolus revealed unsuspected acute lymphoblastic leukemia, prompting a review of the incidence of unsuspected neoplasia from a regional forensic autopsy practice.

Objective.--To determine the incidence of (1) unsuspected neoplasia, (2) clinical concordance of known neoplasia, and (3) potential donor referral in a regional forensic autopsy service.

Design.--A retrospective, 5-year review of 412 autopsies from a regional, primarily forensic, autopsy service to determine the incidence of unsuspected neoplasia, clinical

concordance of known neoplasia, and the preautopsy assessment of potential donor referral suitability.

Results.--Unsuspected neoplasia rate at autopsy was 7% (29 of 412 patients); cancer was the cause of death in 41% (12 of 29 patients) of these individuals. In patients with a history of cancer, the discordance of cancer diagnosis was 44% (4 of 9 patients [11 patients with known cancer, 2 who refused medical evaluation were excluded from the study]). Nearly 60% (17 of 29 patients) of the unsuspected cancer cases had no apparent reason for deferral of tissue procurement before the autopsy examination.

Conclusions.--The 7% incidence of unsuspected cancer in a forensic autopsy practice raises concern for the potential introduction of neoplastic tissue in the donor pool. To ensure the safety of this vital resource, mandatory complete autopsies on deceased donors are advocated as well as a tissue-recipient registry to track donor-related neoplasia.
Article Type: Case study
Subject: Tumors (Diagnosis)
Tumors (Case studies)
Autopsy (Usage)
Acute lymphocytic leukemia (Diagnosis)
Acute lymphocytic leukemia (Case studies)
Authors: Sens, Mary Ann
Zhou, XuDong
Weiland, Timothy
Cooley, A. Marvin
Pub Date: 12/01/2009
Publication: Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 College of American Pathologists ISSN: 1543-2165
Issue: Date: Dec, 2009 Source Volume: 133 Source Issue: 12
Topic: Canadian Subject Form: Tumours; Tumours
Organization: Organization: College of American Pathologists
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 230246904
Full Text: The medical examiner community is increasingly a source of donor referral for organ and tissue transplantation from cases under their jurisdiction. (1) Although medical examiners are on record as supportive of organ and tissue procurement, (2,3) the role of the autopsy in safeguarding the organ and tissue pool has not been adequately addressed. An index case in our office, where all historical information as well as aspects of limited autopsy examination supported alternate causes of death, led us to examine the incidence of occult malignant disease in our autopsy population.

Many articles document the discordance between clinical diagnosis and autopsy diagnosis in nonforensic populations; a Q-probe study of 2479 autopsies at 248 institutions by the College of American Pathologists revealed nearly 40% of cases with a major unexpected finding contributing to death. (4) Similar rates are found in several other studies at different institutions throughout the world. (5,6) This rate has not changed significantly over time, even with the advent of modern imaging and diagnostic techniques. (7-9) Fewer studies specifically assess the incidence and discordance of malignant neoplasms in an autopsy population, but 2 recent US studies in a primarily nonforensic population documented a 20% to 25% incidence rate of malignant neoplasms at autopsy and a 44% discordance rate between clinical diagnosis of malignant neoplasms and findings at autopsy. (10,11) Similar discordance rates for malignant neoplasms are reported in studies in many countries (12-14) and in different medical eras, (15,16) dating to the 1920s. (17)


A 50-year-old, male triathlete experienced severe leg cramps and generalized malaise during a trans-Pacific air flight. The next day, he attempted a workout but experienced shortness of breath, nausea, and light-headedness, which he attributed to jet lag. He was found unresponsive a short time later, and resuscitation efforts were in vain. Past medical history was remarkable only for untreated dyslipidemia, with a total cholesterol of 205 mg/dL, 3 years earlier. Permission for tissue procurement was granted per patient designation and family permission, although heart and bone donation was not accomplished when review of chart and blood consistency suggested resuscitation hemodilution.

At autopsy, loose pale clots, without lines of Zahn, were present in the heart and pulmonary vessels. Cerebral edema and herniation were present with a hemorrhage in the parietal-temporal lobes. There was moderate coronary atherosclerosis with focal areas of 75% occlusive plaque in the left main and left anterior descending arteries, which had an intramyocardial course for 1.5 cm. The clots of the pulmonary vessels and heart consisted of similar small, blue, primitive cells with scant cytoplasm (Figure 1). A pronounced leukostasis in vascular channels and interstitial infiltrates of small, blue cells were present in nearly all tissues examined, with pronounced infiltrate in the endocardial surface (Figure 2) of the heart. A postmortem white blood cell count was 729000 cells/ [micro]L, and the peripheral smear confirmed a massive, leukemic blast population (Figure 3). Immunohistochemical staining of the peripheral blood clots in the heart and lungs was positive for CD34, terminal deoxynucleotidyl transferase (stem cell markers), and CD79a (an early B-cell marker) and negative for CD117 (c-Kit; myeloid linage) and CD5 (an early T-cell marker), establishing a diagnosis of an acute lymphoblastic leukemia, early B-cell phenotype.

The diagnosis of the index case was apparent only after the complete autopsy and extensive microscopic and special examinations. Several compelling, but erroneous, causes of death were supported by history and partial autopsy findings, all of which would have allowed tissue procurement to proceed. A pulmonary embolus was strongly suspected from the history and partially supported by the gross clots in the pulmonary vessels. The central nervous system findings initially suggested a catastrophic central nervous system bleed. If only cardiac procurement (for valve donation) with gross examination of the heart was done, a cause of death from coronary artery disease and bridging left anterior descending artery is reasonable; if microscopic examination of the heart had occurred, a diagnosis of lymphocytic myocarditis might be entertained (Figure 2). White blood cell infiltrates were present in every microscopic section; however, because blast morphology is not well demonstrated in tissue sections, limited microscopic sections may have been interpreted in isolation as resulting from infection or autoimmune disease. Without a relatively detailed autopsy, the best assessment of the cause of death may have been incorrectly ascribed to many nonneoplastic conditions, potentially allowing the introduction of blast leukemia into the tissue pool.


A 5-year review of the prosector (M.A.S., X.Z.) autopsy files was initiated to ascertain the frequency of cancer in a predominantly forensic autopsy practice. Cases with cancer at autopsy were examined for (1) preautopsy assessment of donor potential, and (2) correlation to any premortem, clinical neoplastic diagnosis.

Autopsy Practice Characteristics

The regional autopsy practice included autopsies ordered by coroners and medical examiners from 2 states, all hospital-requested cases from the regional health system, and private autopsy referrals. All autopsy cases were included, with the exception of intrauterine fetal demise, skeletal remains, and externally referred neuropathology cases. The autopsy practice was predominantly forensic-based with 83% of cases ordered by local coroners and medical examiners (including M.A.S.; Table 1). Histology was performed at the University of North Dakota School of Medicine (Grand Forks, North Dakota) laboratory, except for 120 forensic cases (referred to M.A.S. during the absence of the state medical examiner); histology for these 120 cases was contracted by the state to a private laboratory (Medcenter One, Bismarck, North Dakota). Gross prosection (by M.A.S.) was similar in all cases, although the forensic cases included both higher neck dissections with tongue removal and routine central nervous system examination. None of the cases were prosected by students or residents. Autopsy permit restrictions limited dissection scope in many medical autopsies. The intestines were not routinely opened in either hospital or forensic cases, unless external examination was abnormal or there were specific requests for examination of the opened bowel. Internal genitalia (prostate, testes, uterus, and ovaries) were examined grossly and sectioned; however, microscopic examination was generally done only on a single section of prostate, cervix, or corpus uterus, unless gross examination or clinical history was indicative of organ-based disease.

Evaluation of Neoplasia

All neoplasms recorded on the autopsy final diagnosis were reviewed, with confirmation of the malignant diagnoses by all authors. Renal cortical neoplasms less than 1 cm were not tabulated as malignant. For this retrospective study, if the preliminary autopsy diagnosis did not list a mass or neoplasm, the cancer was tabulated as found by microscopic examination.

Cases Selected for Cancer Review

In all cases where cancer was found at autopsy, the medical records were reviewed for any clinical impression of neoplasia before finalizing the autopsy. If none were found, the case was classified as Unsuspected cancer. When clinical history was positive for a neoplastic process, the case was classified as Clinical cancer. Clinical cancer included accidental, suicidal, or homicidal deaths if the diagnosis of cancer was present in the medical record or was indicated in any way by the forensic investigation, including comments to friends and family or mention of ill health in the suicide note. Cases with remote neoplasia were not included unless cancer was present at autopsy. Unsuspected cancer cases were tabulated by suspected cause of death before autopsy, the cause of death after autopsy, whether the cancer caused the death, whether the cancer was diagnosed only by microscopic evaluation, and whether screening deferral of the decedent for tissue procurement would have been made before autopsy (Table 2). If widely metastatic disease was found at autopsy and there were no superseding fatal trauma or toxicology, the case was classified as resulting from cancer. When death was because of thromboemboli in patients with cancer, the case was classified as likely resulting from cancer if no antecedent coagulopathy was present in the available medical records. In cases with a clinical diagnosis of cancer, the autopsy findings and clinical information was classified as Concordant, when the autopsy and clinical information agreed, or Discordant, when discrepancies existed between the clinical and autopsy findings (Table 3). As shown in Table 2, the cause and manner of death were not considered, only whether a diagnosis of malignancy was known before death. Two patients with suspected terminal malignancies refused further medical evaluation and were omitted from concordance evaluation.

Benign or Indeterminate Neoplasms Contributing to Death

There were 4 patients whose benign neoplasms or neoplasms of indeterminate behavior may have contributed or caused the death of the patient; 1 of these (case 7) had a coexistent malignant neoplasm, and the others are grouped separately in Table 4. These benign neoplasms or neoplasms of indeterminate behavior were not included in malignancy or donor status calculations.




Screening Review for Donation Referral

Decedents without an active diagnosis of cancer at the time of death were classified as Likely excluded from donation referral for any of the following reasons: (1) age 80 or older, (2) any history of malignancy (no time limit), (3) drug abuse or other high-risk behavior apparent from the scene investigation, or (4) known or suspected acute infection. These basic exclusion criteria are used by the tissue and organ procurement organization (OPO) in the study region and are shared by many OPOs; however, exclusion criteria of an individual OPO may differ. More detailed donation assessment, primarily done by the OPOs before donation, such as time since death, detailed family interviews, serology testing, and donor preference or family consent, was not considered.

Data Analysis

Cases were maintained in an Access database (Microsoft Corporation, Redmond, Washington). Basic statistical tabulations were done with Excel software (Microsoft). Systat software (Systat Software, Inc, Richmond, California) was used for 2 X 2 contingency tables and x2 tests of association, with Fisher exact test. Graphic data were created with Prism for Windows (GraphPad Software, Inc, La Jolla, California). Paint Shop Pro software (Corel Corporation, Ottawa, Canada) was used for image montage creation from digital gross (Coolpix, Nikon, Tokyo, Japan) and microscopic (SPOT Insight 2 megapixel camera and Eclipse 80i microscope; Nikon) images.


The immunostaining was performed on a Dako autostainer universal staining system (Dako North America, Inc, Carpinteria, California). Both the ready-to-use primary antibodies, for terminal deoxynucleotidyl transferase, CD34, CD79a, CD117, and CD5, and the detection system, EnVision FLEX visualization system, were obtained from Dako.


Patient Demographics in Autopsy Cases

Sex of the decedents in the autopsy cases studied was 62.5% male and 37.5% female, likely reflective of the forensic-based practice. Race of decedents was reflective of the demographics of the region: 87% white, 6% American Indian/Native American, 1% African American, less than 1% Asian, 3% mixed racial identification, and 4% without racial information; 2% of the decedents were Hispanic, again reflective of the local demographics. The average age of patients with unsuspected benign neoplasms was 63 years; the age of patients with unsuspected malignant neoplasms was 62 years. Patients with a clinical history of cancer were slightly older (average age, 67.9 years). The average age of men (64 years) in the study was slightly greater than that of the women (62 years).

Neoplasia Assessment

Of the 412 autopsy cases, 40 were identified with malignancies, for a total of 45 separate neoplasms, which corresponds to an overall cancer incidence of 9.7% and 1.1 malignancies per patient with a neoplasm. Twenty-nine patients without a clinical history of cancer (Table 2) were found to have 33 malignant neoplasms (1.1 malignancies/ patient), yielding an unsuspected cancer incidence of 7%. This rate varied annually from 6% to 8% during the 5-year period of the study. When unsuspected cancer was present at autopsy, it was the cause of death in 41% of the cases. Of the 33 unsuspected cancers, 7 cancers (21%) were detected by microscopic rather than gross examination; this rate was similar in medical and forensic cases (Table 1). The microscopically detected cancers were from the prostate (n = 3), pancreas (n = 2), lung (n = 1), and the index leukemia case. There was statistically significant variation in the rate of undetected cancer between medical and forensic autopsies (Table 1), with a rate of undetected cancer in the medical autopsies of 16% compared with 5% in forensic cases. The forensic cases with more microscopic slides from the authors' (M.A.S. and X.Z.) laboratory had an undetected cancer rate of 6%; in the forensic cases with contracted histology and fewer microscopic sections, the undetected cancer rate was 4%, however, results were not statistically significant (Table 1).

There were 11 patients with a clinical history of active cancer (3% of all 412 cases; 7.4% of medical cases; 1.7% of forensic cases). In the 11 patients with a history of cancer, 2 patients refused workup and were omitted from concordance evaluation. Clinical diagnosis was discordant with autopsy findings in 4 of the remaining 9 patients (44%; Table 3). Two patients with discordant findings were in the immediate postoperative period for surgical resection of primary cancers; both had unsuspected metastatic disease at autopsy. One of the patients had an unknown primary; 2 malignancies were documented at autopsy (urothelial carcinoma of the renal pelvis and multiple myeloma). One patient was clinically assessed with organ-confined cholangiocarcinoma, but metastatic pancreatic cancer was found at autopsy.

The most common cancers detected at autopsy in this study (Figure 4) were malignant pancreatic and lung neoplasms, with 10 cases of each; 8 of which were unsuspected. All 7 cases of renal cancer were unsuspected; 5 cases of prostate cancer were found, only 1 of which was clinically suspected. Other cancers found in this study were uterine (cervix and endometrial), larynx/tongue, breast, hematopoietic (acute lymphoblastic leukemia, multiple myeloma), and thyroid, with 2 cases each. A single case each of adenocarcinoma of the stomach, adenocarcinoma of the gallbladder, and an urothelial cell carcinoma of the renal pelvis was identified. One patient with occult cancer also had an islet cell tumor of the pancreas thought to be benign; 3 additional patients had clinically undetected neoplasms of benign or indeterminate malignant potential (pancreas, central nervous system) which may have contributed to death (Table 4).

Donor Potential Status

As depicted in Figure 5, of the 29 patients with unsuspected malignant neoplasms, 3 patients (10%) were excluded from donation based on age; 2 patients (7%) were excluded based on known drug use or high-risk behavior. Four patients (14%) would likely have been excluded for known or suspected infectious processes. There were 3 patients (10%) with a remote history of treated malignancy, which may have precluded tissue donation in some, but not all, tissue procurement agencies. At autopsy, these patients had new malignancies unrelated to the original cancer. In 17 patients (59% of occult cancers), there was no history or gross findings that would likely have halted tissue donation.


The overall incidence of 7% unsuspected neoplasia may have significant implications for tissue donation, particularly from medical examiner populations where the index of suspicion for cancer is low. When unsuspected malignant neoplasms were found, they were the cause of death in more than 40% of the patients or 3% of the total autopsy population. The index case, as well as the cases presented in Table 2, demonstrates that many decedents with unsuspected neoplasia were eligible for tissue procurement before the autopsy findings. This raises concern for introduction of neoplasia into the donor pool from cases that are not autopsied or from cases in which the autopsy is not focused on a detailed assessment of neoplastic and infectious disease risk. The risks of neoplasia transmission in organ and tissue transplants is believed to be low, although they are well established in the literature for organ donation. (18) Studies that address the risk of neoplasia transmission in tissue recipients are not readily available. Organ procurement organizations commonly exclude tissue donors with active cancer, although many eye banks do not have similar exclusions for corneal donations. Some OPOs will allow donation from cancer survivors after a period of disease-free time determined by the OPO medical directors. Currently, although OPOs will use autopsy findings in case evaluation and the Uniform Anatomic Gift Act mandates release of any autopsy information to the OPOs, few OPOs require autopsies on potential donors. In light of the proven discordance between autopsy and clinical diagnosis, not obtaining autopsies in organ and tissue donors is suboptimal for quality assurance of this vital resource, particularly when donors are from the forensic population, where contemporary clinical workups are seldom available.

The overall incidence rate of malignancies, 9.71% in this study, is somewhat lower than those of previous studies, where incidence rates of 20% to 25% were found. (10,11) How ever, the medical autopsy cases compared favorably with previous studies, (10-12,14-17) with a 24% overall cancer rate and a 16% undetected cancer rate. The undetected cancer rate in forensic autopsies in this study was 5%. Cancer diagnosis rates are rarely reported within purely medical examiner populations; the authors are unaware of studies of a medical examiner population about undetected cancer rates. Furthermore, the focus of many forensic practices is directed to nonnatural deaths with medicolegal implications, thus underestimating the number of undetected neoplasia in the population. This can occur in several ways, primarily related to office case selection, extent of gross dissection, and scope of histologic examination. Some forensic practices have advocated the elimination of microscopic slides; many others limit sections to a few major organs for cost effectiveness, (19) although other studies suggest histology affects final autopsy diagnosis. (20,21)

The types of neoplasms in this study generally agree with those seen in other postmortem studies but are disparate with overall population cancer incidence and cancer death tabulations. (22) Although both inaccuracy of death certification and low autopsy rates may play a role, autopsy limitations in gross or microscopic examination (because of permit restrictions, difficulty of adequate gross examination, office practices of focusing on medicolegal findings, or limits on histologic examinations) also likely contribute to lowering the incidence and type of malignant disease detected at autopsy. On an organ-for-organ basis, autopsies have fewer slides than the corresponding organ submitted for surgical pathology examination. For example, 1 section of prostate was standard in this study; a transurethral resection of the prostate or radical prostatectomy in surgical pathology would have produced many more slides and, presumably, a higher rate of microscopic cancer detection. These factors, and likely others, contribute to the discrepancy seen in the incidence of malignancies at autopsy and rates of clinically detected cancer and cancer deaths.



Potential donor status was also assessed as part of this review. Donor exclusion criteria vary among individual OPOs, and eye banks generally have more liberal criteria than OPOs. Nearly all OPOs will exclude donors with advanced age, suspected infectious disease, or high risk/ drug behavior. Using these widely accepted criteria, only 9 patients would have been excluded, leaving 69% of patients with unsuspected cancer eligible for tissue donor referral. If patients with a remote history of cancer (all of whom had second, unrelated malignancies) were excluded, 3 additional cases would be deferred; however, 59% of cases with malignant disease, discovered only on a detailed autopsy, would be considered for donation (Figure 5). In this study, tissue procurement in 2 additional patients (cases 5 and 25) was halted or the tissue was retracted when renal cell carcinoma was found at autopsy.

Autopsies of tissue and organ donors can improve the safety of the organ and tissue pool and thus increase the benefits to the numerous recipients of this vital resource. The public health mission of pathology has a clear role in safeguarding the vital resource of tissue and organ donation through autopsy performance. The highly desirable goal of increasing autopsies on organ and tissue donors is complex and problematic in implementation. Clearly, increasing autopsies in numbers and scope has financial consequences. In our opinion, the financial responsibility for donor autopsies rests firmly with the OPO enterprise as a normal business expense for quality management and control. The time and expense of these autopsies are considerably more than that of many routine hospital and forensic-focused autopsies where the cause and manner of death or assessment of disease complications are the objectives. Many forensic practices, from which many potential donors originate, may not have personnel or resources to routinely perform detailed autopsies with extensive histologic sampling in potential donors, whose medicolegal investigations may be met with more focused examinations. The mandate of county and state governments may not encompass expending resources for the private OPO enterprise, especially with the strained economic resources of local governmental bodies. Although direct payment lines from the OPOs to pathologists are a critical step in ensuring autopsies in the donor population, some governmental-supported forensic practices prohibit or limit outside reimbursement. Ethical conflicts are occasionally cited with OPO payment to entities that refer or permit tissue and organ donation. Although payment for baseline referrals from forensic and hospital practices to OPO could be construed as creating ethical conflicts, market value payment for the increased work and assessment of donor postmortem evaluation for neoplasia, infections, and other aspects of donor suitability should not raise ethical concerns because payment for other professionals, such as procurement surgeons and teams, laboratory testing of donors, and similar expenses are routinely accepted as business-related expenses for donor tissue. Further complicating issues in increasing donor autopsies include the availability of facilities and autopsy pathologists. Fewer autopsies have lead some health care entities to eliminate or minimize mortuary facilities and staffing, potentially creating difficulties for autopsy performance, even in the private hospital setting, where no forensic referrals complicate the donor status. The 16% rate of undetected neoplasia for medical autopsies and the 44% discordant clinical versus autopsy diagnosis compare favorably with other studies (10-15) and emphasize the need for autopsy evaluation of the nonforensic donor. Patient and family consent issues are also critical in autopsies of potential donors. For optimal examination, some organs may need fixation and retention for examination. Separate informed consent for these detailed autopsies and organ retention should be part of donation protocols and should not be "piggybacked" on the medicolegal death-investigation systems or routine hospital autopsies, where a more focused examination may be standard, or in the case of hospital autopsies, where restrictions may be imposed by the family. Finally, the scope and timing of these donor autopsies deserve consideration. There is no consensus on extent of histologic or gross examination within autopsy practices. Recommended gross examination techniques and numbers of microscopic sections for each organ system need to be developed for donor autopsies. These would likely be age specific, reflecting optimal estimates for occult neoplasia detection and infections. The examination would likely include more microscopic sections for some organs, such as the prostate, where increased microscopic detection of neoplasia is expected. Autopsy examinations may be considered where, with the consent of family, multiple sections of selected organs and tissues are taken at a forensic autopsy, and the OPO submits this tissue for contracted surgical pathology examination, if autopsy workload or governmental agencies prohibit detailed donor autopsies within a forensic practice. Regardless of the processes developed to increase autopsy evaluation of tissue and organ donors, there will be a mandate for timeliness of results. Autopsies for donor neoplasia or infection need to be completed in the timeframe of surgical pathology, not autopsy pathology, reports.

Finally, in our opinion, a national or international registry of tissue recipients' long-term outcome should be established by a regulatory or governmental agency, such as the Centers for Disease Control and Prevention or the US Food and Drug Administration in the United States, with sufficient regulatory backing and funding to accurately track neoplastic, infectious, or other outcomes of recipients, as well as to coordinate the efforts with analogous agencies in other countries. The first transplant organ tumor registry was initiated by Dr Israel Penn, who collected data on more than 15 000 transplant-related malignancies for decades. This is now an international, open-access, voluntary-reporting registry (the Israel Penn International Transplant Tumor Registry, Cincinnati, Ohio) for malignancies associated with organ transplants. (23) Since 1999, the United Network for Organ Sharing (Richmond, Virginia) has collected data on all solid organ transplant recipients who develop malignancies, expanding the knowledge of donor-related malignancies. (24) No reporting registry exists for larger tissue usage. Infectious disease transmission to recipients, although not examined in this study, also has potentially devastating consequences in organ and tissue recipients. (25) Use of existing registries and formation of similar registries for tissue recipients will assist in identifying the risk of these procedures and may lead to better screening of potential donors by OPOs and focusing of future donor autopsy examinations to reduce risks to recipients.

Pathology's critical role in ensuring patient safety challenges us to devise multiple interventions to increase both the autopsy rate of organ and tissue donors and to maximize the autopsy assessment of donors for infectious and neoplastic disease that may affect tissue and organ transplant in living recipients.


(1.) Shafer T, Schkade LL, Warner HE, et al. Impact of medical examiner/coroner practices on organ recovery in the United States. JAMA. 1994;272(20):16071613.

(2.) Pinckard JK, Wetli CV, Graham MA. National Association of Medical Examiners position paper on the medical examiner release of organs and tissues for transplantation. Am J Forensic Med Pathol. 2007;28(3):202-207.

(3.) Jaynes CL, Springer JW. Decreasing the organ donor shortage by increasing communication between coroners, medical examiners and organ procurement organizations. Am J Forensic Med Pathol. 1994;1 5(2):1 56-1 59.

(4.) Zarbo RJ, Baker PB, Howanitz PJ. The autopsy as a performance measurement tool--diagnostic discrepancies and unresolved clinical questions: a College of American Pathologists Q-Probes study of 2479 autopsies from 248 institutions. Arch Pathol Lab Med. 1999;123(3):191-1 98.

(5.) Grundmann E. Autopsy as a clinical quality control: a study of 15,143 cases. In Vivo. 1994;8(5):945-952.

(6.) Ornelas-Aguirre JM, Vazquez-Camacho G, Gonzalez-Lopez L, et al. Concordance between premortem and postmortem diagnosis in the autopsy: results of a 10-year study in a tertiary care center. Ann Diagn Pathol. 2003;7(4):223 230.

(7.) Kirch W, Schafil C. Misdiagnosis at a university hospital in 4 medical eras. Medicine. 1996;75(1):29-40.

(8.) Shojania KG, Burton EC, McDonald KM, Goldman L. Changes in rates of autopsy-detected diagnostic errors over time: a systematic review. JAMA. 2003; 289(21):2849-2856.

(9.) Sonderegger-Iseli K, Burger S, Muntwyler J, Salomon F. Diagnostic errors in three medical eras: a necropsy study. Lancet. 2000;355(9220):2027-2031.

(10.) Burton EC, Troxclair DA, Newman WP III. Autopsy diagnoses of malignant neoplasms: how often are clinical diagnoses incorrect? JAMA. 1998;280(14): 1245-1248.

(11.) Avgerinos DV, Bjornsson J. Malignant neoplasms: discordance between clinical diagnoses and autopsy findings in 3,118 cases. APMIS. 2001;109(11): 774-780.

(12.) Stenback F, Paivarinta H. Relation between clinical and autopsy diagnoses, especially as regards cancer. Scand J Soc Med. 1980;8(2):67-72.

(13.) Szende B, Kendrey G, Lapis K, Lee PN, Roe FJ. Accuracy of admission and pre-autopsy clinical diagnoses in the light of autopsy findings: a study conducted in Budapest. Hum Exp Toxicol. 1994;13(10):671-680.

(14.) de Pangher Manzini V, Revignas MG, Brollo A. Diagnosis of malignant tumor: comparison between clinical and autopsy diagnoses. Hum Pathol. 1995; 26(3):280-283.

(15.) Bauer FW, Robbins SL. An autopsy study of cancer patients, I: accuracy of the clinical diagnoses (1955 to 1965) Boston City Hospital. JAMA. 1972; 221(13):1471-1474.

(16.) Grundmann E, Menke GG. Autopsy diagnosis versus clinical diagnosis, particularly in malignant disease. Comparison of two periods: 1961-70 and 1 978-87. IARCSci Publ. 1991;(112):81-90.

(17.) Wells HG. Relation of clinical to necropsy diagnosis in cancer and value of existing cancer statistics. JAMA. 1923;80:737-740.

(18.) Buell JF, Gross T, Alloway RR, Trofe J, Woodle ES. Central nervous system tumors in donors: misdiagnosis carries a high morbidity and mortality. Transplant Proc. 2005;37(2):583-584.

(19.) Molina DK, Wood LE, Frost RE. Is routine histopathologic examination beneficial in all medicolegal autopsies? Am J Forensic Med Pathol. 2007;28(1): 1-3.

(20.) Bernardi FD, Saldiva PH, Mauad T. Histological examination has a major impact on macroscopic necropsy diagnoses. J Clin Pathol. 2005;58(12):1261 1264.

(21.) Roulson J, Benbow EW, Hasleton PS. Discrepancies between clinical and autopsy diagnosis and the value of post mortem histology: a meta-analysis and review. Histopathology. 2005;47(6):551-559.

(22.) Centers for Disease Control and Prevention/National Program of Cancer Registries (NPCR). United States cancer statistics: 1999-2004 cancer incidence and mortality data Web-based report. http://apps.nccd.cdc.gov/uscs/. Updated February 9, 2009. Accessed June 14, 2008.

(23.) Witherow BA, Roth GS, Carrozza MA, et al. The Israel Penn International Transplant Tumor Registry. In: AMIA Annual Symposium on Biomedical and Health Informatics--From Foundations to Applications to Policy; November 8-12, 2008; Washington, DC. Bethesda, MD: American Medical Informatics Association; 2003. Abstract 1053.

(24.) Kauffman HM, Cherikh WS, McBride MA, Cheng Y, Hanto DW. Deceased donors with a past history of malignancy: an organ procurement and transplantation network/united network for organ sharing update. Transplantation. 2007; 84(2):272-274.

(25.) Srinivasan A, Burton EC, Kuehnert MJ, et al. Transmission of rabies virus from an organ donor to four transplant recipients. N Engl J Med. 2005;352(22): 1103-1111.

Mary Ann Sens, MD, PhD; XuDong Zhou, MD; Timothy Weiland, MD; A. Marvin Cooley, MD

Accepted for publication February 5, 2009.

From the Department of Pathology, University of North Dakota School of Medicine and Health Sciences, Grand Forks (Drs Sens and Zhou); the Medical Examiner's Offices for Marshall, Kittson, and Red Lake counties, Warren, Hallock, and Red Lake Falls, Minnesota (Dr Sens); the Coroner's Office, Grand Forks, North Dakota (DrSens); and the Department of Pathology, Altru Health System, Grand Forks, North Dakota (Drs Weiland and Cooley).

The authors have no relevant financial interest in the products or companies described in this article.

Presented in part as a College of American Pathologists' APEX case in the Autopsy Pathology Committee. Summary data for the neoplasia incidence was presented in part at the annual meeting of the National Association of Medical Examiners, Savannah, Georgia, October 17, 2007.

Reprints: Mary Ann Sens, MD, PhD, Department of Pathology, University of North Dakota School of Medicine and Health Sciences, 501 N Columbia Rd, Grand Forks, ND 58202-9037 (e-mail: msens@ medicine.nodak.edu).
Table 1. Histologic Assessment of Cases

                                              Decedents       Suspected
                               Autopsy           With         Cancer, %
Case Type                    Cases, No.       Cancer, %          (a)

Medical, histology 1 (b)          68            23.53         7.35 (c)
All forensic cases               344             6.98         1.74 (c)
  Forensic histology 1 (b)       224             7.59         1.79 (NS)
  Forensic histology 2 (d)       120             5.83         1.67 (NS)
All cases                        412             9.71           2.67
                                               (n = 40)       (n = 11)

                              Decedents      Unsuspected
                                With       Cancers Detected
                             Unsuspected   Microscopically,
Case Type                     Cancer, %           %

Medical, histology 1 (b)      16.18 (c)       23.08 (NS)
All forensic cases             5.23 (c)       21.05 (NS)
  Forensic histology 1 (b)     5.80 (NS)      21.43 (NS)
  Forensic histology 2 (d)     4.17 (NS)      20.00 (NS)
All cases                       7.04            21.21
                              (n = 29)         (n = 33)

                             Average No.       Range of
Case Type                    Slides/Case        Slides

Medical, histology 1 (b)         26              3-68
All forensic cases
  Forensic histology 1 (b)       21              3-62
  Forensic histology 2 (d)       10              3-25
All cases

(a) Includes any clinical diagnosis in the medical record, as well as
any indication such as comments to family, friends, or suicide notes
that decedent suspected or knew of cancer diagnosis.

(b) Histology performed in authors' (M.A.S. and X.Z.) laboratory.

(c) Significant difference between decedents with suspected and
unsuspected cancer at P = .005 using Fisher exact test; NS, no
significant differences in forensic cases between the 2 histology
groups in both Suspected cancer and Unsuspected cancer using 2 X 2
contingency tables using x2 association at P = .05. There were no
differences between any groups (forensic, medical, histology source)
in number of microscopically detected cancers.

(d) Contract histology on fee/slide basis for contracted state
forensic cases.

Table 2. Unsuspected Cancer at Autopsy

No.          Malignancy         Suspected COD        Actual COD

 1     Pancreas:                Drowning             Drowning, EtOH
         adenocarcinoma                                intoxication
 2     (1) Lung (NSCC)          Cardiac vs motor     Acute coronary
       (2) Renal (clear cell)     vehicle accident     thrombosis
 3     Pancreas:                Drug overdose        Drug overdose
 4     (1) Renal (clear cell)   Gunshot wound        Gunshot wound
       (2) Lung (NSCC)
 5     Renal (clear cell)       Sudden cardiac       Acute myocardial
                                  death                infarction
 6     Lung (NSCC)              Pneumonia            Metastatic lung
 7     (1) Renal (clear cell)   Cold exposure        Cold exposure,
       (2) Islet cell tumor                            EtOH
           (likely benign)                             intoxication
 8     Lung (NSCC)              COPD                 Widely metastatic
 9     Cervical (SCC)           Acute febrile        Peritonitis,
                                  illness              widespread SCC
                                                       with GI erosion
10     Pancreas:                Carbon monoxide      Carbon monoxide,
         adenocarcinomad                               polypharmacy
         (second cancer of                             over-dose
11     Endometrial carcinoma    Complications        Pulmonary
         (no evidence of          of diabetes          thromboemboli
         breast carcinoma)
12     Lung, bronchoalveolar    Sudden cardiac       Acute myocardial
                                  death                infarction
13     Renal                    ARDS, bowel          ARDS, Candida
         (clear cell, 3 cm)       infarction           sepsis
14     Lung (NSCC)              Sudden cardiac       Widespread NSCC
15     Prostate, 3+4 Gleason    Complications of     Complications of
                                  CABG                 CABG
16     Pancreas,                Gunshot wound        Gunshot wound
17     Thyroid, follicular      COPD, CHF            Pulmonary
18     Lung (NSCC)              Sudden cardiac       Acute myocardial
                                  death                infarction
19     Renal (clear cell,       CHF                  Budd-Chiari
         5 cm) (no evidence                            syndrome,
         residual/recurrent                            pulmonary
         GCT, testes)                                  thromboemboli
20     Acute lymphocytic        Pulmonary            ALL, early B-cell
         leukemia (ALL)           thromboembolus       phenotype
21     Lung (NSCC)              Sudden cardiac       Thromboemboli,
                                  death                metastatic lung
22     Prostate, 3+3 Gleason    Carbon monoxide      Carbon monoxide
23     Pancreas,                CHF                  Pulmonary
         adenocarcinoma                                thromboemboli
24     Prostate, 3+3 Gleason    Klebsiella sepsis    Klebsiella sepsis
25     Renal (clear cell)       Sudden cardiac       Saddle pulmonary
                                  death                embolus
26     Pancreas,                Positional           Positional
         adenocarcinoma           asphyxia             asphyxia
27     (1) Thyroid, papillary   Empyema,             Empyema, Bezold
       (2) Breast, invasive       EtOH abuse           abscess
28     Pancreas,                Dementia             Pulmonary
         adenocarcinoma                                thromboemboli
29     Pancreas,                Sudden cardiac       Acute myocardial
         adenocarcinoma           death                infarction

No.          Malignancy         Cancer COD           Detection

 1     Pancreas:                No                   Gross
 2     (1) Lung (NSCC)          No                   (1) Gross
       (2) Renal (clear cell)                        (2) Gross
 3     Pancreas:                No                   Micro
 4     (1) Renal (clear cell)   No                   (1) Gross
       (2) Lung (NSCC)                               (2) Gross
 5     Renal (clear cell)       No                   Gross
 6     Lung (NSCC)              Yes                  Gross
 7     (1) Renal (clear cell)   No                   (1) Gross
       (2) Islet cell tumor                          (2) Gross
           (likely benign)
 8     Lung (NSCC)              Yes                  Gross
 9     Cervical (SCC)           Yes                  Gross
10     Pancreas:                No                   (1) Micro
         adenocarcinomad                             (2) Micro (d)
         (second cancer of
11     Endometrial carcinoma    Likely               Gross
         (no evidence of
         breast carcinoma)
12     Lung, bronchoalveolar    No                   Micro
13     Renal                    No                   Gross
         (clear cell, 3 cm)
14     Lung (NSCC)              Yes                  Gross
15     Prostate, 3+4 Gleason    No                   Micro
16     Pancreas,                No                   Gross
17     Thyroid, follicular      Likely               Gross
18     Lung (NSCC)              No                   Gross
19     Renal (clear cell,       Likely               Gross
         5 cm) (no evidence
         GCT, testes)
20     Acute lymphocytic        Yes                  Micro
         leukemia (ALL)
21     Lung (NSCC)              Yes                  Gross
22     Prostate, 3+3 Gleason    No                   Micro
23     Pancreas,                Likely               Gross
24     Prostate, 3+3 Gleason    No                   Micro
25     Renal (clear cell)       Likely               Gross
26     Pancreas,                No                   Gross
27     (1) Thyroid, papillary   No                   (1) Gross
       (2) Breast, invasive                          (2) Gross
28     Pancreas,                Likely               Gross
29     Pancreas,                No                   Gross

No.          Malignancy         Age, y               Likely Deferral

 1     Pancreas:                56                   No
 2     (1) Lung (NSCC)          67                   No
       (2) Renal (clear cell)
 3     Pancreas:                48                   Yes
 4     (1) Renal (clear cell)   67                   No
       (2) Lung (NSCC)
 5     Renal (clear cell)       53                   No
 6     Lung (NSCC)              78                   Yes
 7     (1) Renal (clear cell)   51                   No
       (2) Islet cell tumor
           (likely benign)
 8     Lung (NSCC)              54                   No (b)
 9     Cervical (SCC)           51                   Possible (c)
10     Pancreas:                32                   Yes
         (second cancer of
11     Endometrial carcinoma    59                   Possible
         (no evidence of
         breast carcinoma)
12     Lung, bronchoalveolar    86                   Yes
13     Renal                    53                   Yes
         (clear cell, 3 cm)
14     Lung (NSCC)              77                   No
15     Prostate, 3+4 Gleason    87                   Yes
16     Pancreas,                55                   No
17     Thyroid, follicular      79                   No
18     Lung (NSCC)              60                   No
19     Renal (clear cell,       54                   Possible
         5 cm) (no evidence
         GCT, testes)
20     Acute lymphocytic        50                   No
         leukemia (ALL)
21     Lung (NSCC)              68                   No
22     Prostate, 3+3 Gleason    66                   No
23     Pancreas,                62                   No
24     Prostate, 3+3 Gleason    82                   Yes
25     Renal (clear cell)       47                   Noe
26     Pancreas,                57                   No
27     (1) Thyroid, papillary   48                   Yes
       (2) Breast, invasive
28     Pancreas,                68                   No
29     Pancreas,                78                   Possible

No.          Malignancy         Reason Deferral

 1     Pancreas:
 2     (1) Lung (NSCC)
       (2) Renal (clear cell)
 3     Pancreas:                Behavior
 4     (1) Renal (clear cell)
       (2) Lung (NSCC)
 5     Renal (clear cell)       (a)
 6     Lung (NSCC)              Infection
 7     (1) Renal (clear cell)
       (2) Islet cell tumor
           (likely benign)
 8     Lung (NSCC)              Gross
 9     Cervical (SCC)           Likely infection
10     Pancreas:                Behavior
         (second cancer of
11     Endometrial carcinoma    Cancer history,
         (no evidence of          12 y NED
         breast carcinoma)
12     Lung, bronchoalveolar    Age
13     Renal                    Infection
         (clear cell, 3 cm)
14     Lung (NSCC)
15     Prostate, 3+4 Gleason
16     Pancreas,
17     Thyroid, follicular
18     Lung (NSCC)
19     Renal (clear cell,       Testicular
         5 cm) (no evidence       GCT, 25 y
         residual/recurrent       NED
         GCT, testes)
20     Acute lymphocytic
         leukemia (ALL)
21     Lung (NSCC)
22     Prostate, 3+3 Gleason
23     Pancreas,
24     Prostate, 3+3 Gleason
25     Renal (clear cell)
26     Pancreas,
27     (1) Thyroid, papillary   Infection
       (2) Breast, invasive
28     Pancreas,
29     Pancreas,                NHL, 12 y NED

Abbreviations: ARDS, adult respiratory distress syndrome; CABG,
coronary artery bypass graft; CHF, congestive heart failure; COD,
cause of death; COPD, chronic obstructive pulmonary disease; EtOH,
ethanol; GCT, germ cell tumor; micro, microscopic; NED, no evidence
of disease; NHL, non-Hodgkin lymphoma; NSCC, nonsmall cell carcinoma;
SCC, squamous cell carcinoma.

(a) Patient with heart valve and tissue donation before autopsy; renal
cell carcinoma found at autopsy.

(b) Patient eligible for cardiac valve donation; metastatic disease
grossly visible on opening body cavity, donation likely would have

(c) Patient likely deferred for acute febrile episode before death,
which resulted from intestinal perforation and peritonitis from
metastatic squamous cell carcinoma of cervix.

(d) Patient with confirmed, widespread, high-grade, prostatic
intraepithelial neoplasia, likely invasive prostatic adenocarcinoma;
autolysis precluded confirmation. Not included as a confirmed

(e) Patient approved for donation; process halted when node-positive
renal cell carcinoma was found.

Table 3. Clinical Concurrence of Active Cancer

Case                           Malignancy at
No.        Actual COD             Autopsy         Clinical Assessment

30     Sudden cardiac       Lung (NSCC) in        Lobectomy for NSCC
         death, cold          contralateral         <24 h prior;
         exposure, EtOH       lobes, nodes          no residual
         tobacco use
31     Hanging              Stomach,              Stomach,
                              adenocarcinoma        adenocarcinoma
32     TCC, metastatic      Renal pelvis, TCC     Metastatic disease,
         multiple myeloma     multiple myeloma      unknown primary
33     Metastatic           Pancreas,             Cholangiocarcinoma,
         pancreatic           adenocarcinoma        confined
34     GSW                  Pancreas,             Metastatic disease,
                              adenocarcinoma        unknown primary
35     SGW                  Laryngeal carcinoma   Laryngeal carcinoma
36     Epithelial           Pleural epithelial    Likely mesothelioma
         mesothelioma         mesothelioma
37     Metastatic disease   Metastatic breast     Metastatic breast
                              carcinoma             carcinoma
38     Sepsis metastatic    Gallbladder,          Resection,
         disease              adenocarcinoma,       localized
                              metastatic to         adenocarcinoma,
                              liver, pleura         gallbladder,
                                                    3 d postop
39     GSW                  Tongue, SCC           Tongue, SCC
40     GSW                  Prostate, Gleason     Gleason 5 + 5
Case                                               Concurrence
No.        Actual COD       Cancer COD   Age, y      Clinical

30     Sudden cardiac          Yes         55     Discordant
         death, cold
         exposure, EtOH
         tobacco use
31     Hanging                  No         51     Concordant
32     TCC, metastatic         Yes         78     Discordant
         multiple myeloma
33     Metastatic              Yes         78     Discordant
34     GSW                      No         55     Refused workup
35     SGW                      No         66     Concordant
36     Epithelial              Yes         78     Refused workup
37     Metastatic disease      Yes         81     Concordant
38     Sepsis metastatic       Yes         65     Discordant
39     GSW                      No         57     Concordant
40     GSW                      No         80     Concordant

Abbreviations: COD, cause of death; EtOH, ethanol; GSW, gunshot wound;
NSCC, nonsmall cell carcinoma; postop, after surgery; SCC, squamous
cell carcinoma; SCD, sudden cardiac death; SGW, shotgun wound; TCC,
transitional (urothelial) cell carcinoma.

Table 4. Benign/Indeterminate Neoplasms Possibly Contributory to Death

Neoplasm                   Actual COD           Neoplasm COD     Age/y

Pancreas:            Known hypoglycemia       Unlikely            73
  intraductal          episode;
  papillary            complications of
  mucinous tumor,      CAD, hypertension,
  multifocal, 3 cm     DM
Insulin islet cell   Possible                 Possible            77
  tumor                hypoglycemia             hypoglycemia
                     Complications of CAD,
                       DM, hypertension
Pineal tumor         Motor vehicle            Possible, visual    41
                       accident; possible       compromise
                       CNS effect

Abbreviations: CAD, coronary artery disease; CNS, central nervous
system; COD, cause of death; DM, diabetes mellitus.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.