Fox sign in a case of terminal stage pancreatic cancer and suggestions for diagnosis

Dermatologic findings, which are frequently observed during autopsies, can provide valuable differential diagnostic clues regarding their underlying etiology. Such findings sometimes even indicate cause of death [1, 2]. These findings include various skin signs in acute pancreatitis that are defined based on their particular expansion and location of emergence. A commonly accepted view on pancreatitis is that the premature activation of digestive enzymes within the pancreas leads to autodigestion of the organ [3‐5]. The enzymes include proteases such as trypsin, elastase, and pancreatic lipase [6]. Skin signs associated with pancreatitis are presumably caused by the expansion of intra- and retroperitoneal hemorrhages along anatomic structures, which manifest in the subcutaneous tissue at specific locations, depending on the path of expansion [7]. However, these signs do not solely occur in acute pancreatitis, and many other causes have been identified, including necrotizing pancreatitis, intraperitoneal hemorrhage, portal hypertension, liver cirrhosis anticoagulation with warfarin, or deep vein thrombosis [8‐13]. Furthermore, they indicate a poor prognosis [14]; therefore, it seems reasonable to raise awareness regarding the potential clinical impact they entail. Grey Turner’s sign [15] and Cullen’s sign [16] represent the best-known signs and are frequently mentioned in textbooks and journals. A lesser-known skin sign is the Fox sign, which emerges from the extraperitoneal expansion of pancreatic fluid along the fascia of the psoas and iliacus until its subcutaneous manifestation below the inguinal ligament [17]. By contrast, the Fox sign is rather rare in clinical practice, and some physicians may have never encountered such a case themselves.

In 2018, we encountered an autopsy that was ordered by the district attorney’s office owing to alleged medical malpractice and poor hospital care in which the Fox sign was assumed to be present and later used to refute the raised allegations. Furthermore, we identified another case with skin signs associated with suspected pancreatitis and examined how to identify skin signs in suspected pancreatitis using immunohistochemistry.

Two forensic autopsy cases, in which the deceased exhibited skin discolorations at typical locations for skin signs in pancreatitis at the external examination, were identified between 2018 and 2019. The presence of those skin signs and an additional pancreatic disease was the inclusion criterion. Skin samples of the discolorations were obtained, and histologic examinations and immunohistochemical staining were performed. Six autopsy cases served as controls. Five controls had no skin discolorations or relevant organ pathologies; one case showed a reddish-livid discoloration at the right flank but no pancreatic pathologies. The cause of death was hanging in three cases and gunshot to the head, severe head injury, and myocardial infarction in one case each.

The immunohistochemical staining used the following antibodies: rabbit monoclonal anti-pancreatic lipase antibody (EPR6276) ab124915 and rabbit monoclonal anti-trypsin antibody (EPR19498-15) ab211491 (Abcam, Cambridge, United Kingdom). In accordance with the manufacturer’s instructions, the immune complexes were visualized using the Dako Envision + Dual Link System (Dako Denmark, Glostrup, Denmark). After staining and preparation, the slides were microscopically examined using the Olympus BX 53 photomicroscope with the Olympus SC 180 camera (Olympus Corporation, Tokyo, Japan) with a sensor size of 4.8 × 2.7 mm at an eyepiece magnification of 10 × and an objective magnification between 4 × and 20 × .

In forensic analyses, determining the origin and etiology of skin discoloration to exclude or confirm bodily harm can be crucial. This, among other factors, requires knowledge of the described skin signs in pancreatitis and other conditions that may cause these specific discolorations. The literature describes certain conditions that potentially cause the Fox sign, such as necrotizing pancreatitis, deep vein thrombosis, and rupture of an abdominal aortic aneurysm [8, 13, 17].

In case 1, a pulmonary embolism was diagnosed in the hospital before the man’s death. Because pancreatic carcinoma has the highest risk for the development of vein thrombosis among other carcinomas [18] and approximately 93% of thromboses originate in the inferior vena cava and pelvic and leg veins [19], a deep vein thrombosis could also have caused the Fox sign. There were no signs of inflammation indicating a diaper dermatitis in the skin. Immunohistochemical analysis showed a low level of staining for pancreatic lipase in the subcutaneous skin and no staining for trypsin. However, the pancreatic lipase staining was slightly more distinct than in the control cases, and after considering all factors, the discoloration was assumed to be a Fox sign. To the best of our knowledge, identifying a presumed diaper dermatitis as a Fox sign and thus refuting allegations of medical malpractice and poor hospital care appear to represent a novelty. Furthermore, this case emphasizes the importance and explanatory power of forensic autopsies and complementary histological examinations regarding alleged medical malpractice.

The male in case 2 had a fatty liver with the beginning of cirrhosis, a necrotic and inflamed pancreatic head with surrounding hemorrhage, and hemorrhage in the right-sided mesocolon and right renal pelvis. The cause of death was determined as necrotizing pancreatitis with internal bleeding. There were no signs of reddish staining for trypsin; however, there was extensive staining for pancreatic lipase in the epidermis and dermis that was much more intense than in any of the included cases. Considering the immunohistochemical results and because the person had two conditions that are known to cause such discoloration, Grey Turner’s sign was assumed to be present in this case.

Overall, the differential diagnostics of skin discolorations should include the distinction from hematoma or ecchymosis due to blunt force or needle puncture sites because this could feign the origin of skin signs associated with pancreatitis [20]. In both cases, there was no clear evidence of blunt force or needle punctures although small injections are not always visible and therefore cannot be ruled out with absolute certainty.

Regarding the control group, one case showed circumscribed staining of trypsin in the epidermis and very slight disseminated staining of lipase in the dermis. This might be owing to cross-reactions with the trypsin antibody and other serine proteases present in human skin [21]. Talas et al. [22] examined the expression of human elastase 1 in the skin that showed results similar to those found in both our study and control cases. Lipase can be detected in the sebaceous glands and in the external root sheath of the hair follicle [23]. Therefore, lipase enzymes were believed to be dissolved during staining, which resulted in their detection in normal skin. Skin discoloration at the right flank observed in the 89-year-old female may be explained by the use of anticoagulants. The use of an anticoagulating agent with warfarin is described in the literature to form skin discoloration at this location [12]. Unfortunately, no information on anticoagulating agents or other medication was available. Thus, the presence of Grey Turner´s sign was considered unlikely and the discoloration presumably had a different cause.

Based on our findings, minor disseminated staining of lipase and trypsin should be considered regular and alone should not lead to the conclusion of a skin sign associated with pancreatitis. Considering all cases, trypsin does not seem as suitable as lipase for this suggested immunohistochemical method. Furthermore, in the three cases with skin signs associated with suspected pancreatitis, it remained unclear when the discoloration first occurred; therefore, both the antemortem and postmortem degradation of the enzymes must be considered, which might have led to the indistinct staining pattern in two cases.

Conclusively, additional immunohistochemical investigations regarding the presence of pancreatic lipase may be useful in cases with unclear skin discoloration. However, considering the relative rarity of this phenomenon, further and ideally multicenter studies with a sufficient number of cases are warranted to establish and verify reliable staining patterns for the examined enzymes.