P. multocida infection can produce a spectrum of human diseases, mostly corresponding to soft tissue infections, but it might also be seriously invasive [30]. Indeed, respiratory, bone and joint infections (septic arthritis) are reported. Generally, septic arthritis of a native joint can occur by hematogenous spread as well as direct inoculation of the joint following trauma, surgery or intra-articular injection, or continuous extension of a bone infection [31]. Most cases of septic arthritis by P. multocida occur after a cat or dog bite or scratch distal to the involved joint, without direct penetrating injury. However, approximately one-third of cases are not preceded by an animal injury. Usually, skin and soft tissue infections have a good outcome, especially in patients without significant risk factors, but mortality rates from P. multocida can range from 9 to 31%, without a prompt and correct diagnosis and treatment [32, 33].

In the case here described the challenge for forensic pathologists consisted in the absence of a clinical diagnosis for septic arthritis that, as shown by our systematic reviews, usually precedes the fatal outcome. Indeed, septic arthritis is not typically included among the primary substrates responsible for an unexplained death. In the case here reported, given the reported shoulder pain and the absence of any other pathological cause of death, the right shoulder was sectioned, leading to the detection of abundant purulent material and giving an input to more detailed examinations, particularly to the microbiological culture.

The isolation of P. multocida from the swab taken on the articular purulent material was initially considered with caution because of the time elapsed between the patient death and the sample collection (8 days), which might have led to a bacterial putrefactive growth or to postmortem bacterial transmigration. Bacterial growth may correspond to a true positive result in post-mortem samples (pure growth of a specific pathogen colonizing an otherwise sterile organ or fluid), but translocation (bacterial migration from the mucosal surface into the blood and internal organs after death) or contamination (incidental introduction of bacteria into the samples when they are obtained using non-sterile tools or operating in non-sterile environments) may also occur [34,35,36,37]. As P. multocida is neither a commensal microbe of human beings nor typical of post-mortem proliferation [38], and relying on the monomicrobial growth of a typical pathogenic microorganism on an isolated body area, as well as on the body storage at 4 °C, the bacterial growth was considered reliable. On the other hand, a blood culture was not performed due to the multiple possible biases arising from it.

Once detected the growth of the potential pathogen, P. multocida, a history of animal contacts was searched but, at first, not confirmed. Only after extensive questioning of the family, it was learned belatedly that the woman had been bitten and scratched by a stray cat.

The first symptoms of soft tissues infections caused by P. multocida usually arise within 24 h from the injury, and other manifestations (edema, swelling cellulitis) follow within 24–48 h [30]. However, the time required for a septic arthritis to develop is not fully defined in the current literature. Weber et al. [30] described 12 cases of patients with septic arthritis caused by P. multocida, in whom the condition was “long lasting” and difficult to diagnose “before a few weeks” after the first injection of the pathogen in the subcutaneous tissues [30].

In our case, the timeframe between the reconstructed exposition to P. multocida and the development of the arthritis was estimated as two to four weeks, consistent both with the prolonged bacterial incubation and the histopathological findings of sub-acute/chronic arthritis. Moreover, at autopsy, no evidence of residual cat bites and scratches were observed, but they might have been healed during the incubation period.

A further complication was represented by the paucity of ante-mortem clinical data. Indeed, when the patient was still alive, the right shoulder pain was attributed to degenerative polyarthritis, as confirmed by X-rays. The woman did not report of recent fever, which, as demonstrated in our review [28], might be absent in septic arthritis; vital signs abnormalities were not detected, and the condition was not further investigated by laboratory testing or ultrasound.

The re-evaluation of the radiological images contributed, however, to the comprehension of the case. Previous case reports of septic arthritis [39, 40] have shown aspecific signs at radiography, such as depression of the humeral head or widening of the sub-acromial space. These signs, additionally noted at the re-examination in our case, were interpreted as indicative of effusion and, therefore, inflammation, further confirming the suspect for septic arthritis.

The autopsy finding of septic arthritis, which typically involves a hematogenous spread of the pathogen, alerted for further analysis in order to establish whether a systemic infection and sepsis had occurred.

The diagnosis of sepsis is mainly clinical and requires multiple information and positive scorings performed in a hospital setting, which were unavailable in our patient [41].

As suggested by Stassi C. et al. [42], for the post-mortal diagnosis of sepsis it is advisable to collect all available data, including microbiological, biochemical analysis, and histopathological examinations, to make the most probable and accurate diagnosis.

Beside the above-mentioned microbiological analysis, the post-mortem biochemistry, conducted on femoral blood, oriented towards the diagnosis of sepsis, despite only two of the three measured markers were found above the reference values: CRP of 147.7 mg/L and TnIH of 16,567.70 ng/L, respectively. CRP is an acute-phase protein produced by the liver in response to inflammation and the elevated levels detected, compared with the literature, were consistent with sepsis prior to death. TnIH might reflect the myocardial damage associated with sepsis, but could be falsely elevated in heart blood in a post-mortem setting. In the present case, the biochemical analyses were conducted on femoral blood, so the values were considered indicative of the pre-mortem condition. PCT levels were within normal limits with 0.09 ng/mL (normal range < 0.5–2 ng/mL). Although PCT in septic arthritis can be of great value, as reported by a recent meta-analysis it cannot and should not be considered the only relevant marker for the diagnosis [43]. Moreover, PCT negativity does not necessarily contradict the hypothesis of septic arthritis, as post-mortem degradation of serum proteins and the low sensitivity (50–60%) must be taken into account [43].

Even if PCR [44,45,46] and troponin I values are not specifically indicative of a spreading bacterial infection, they were deemed highly suggestive of an ongoing inflammatory status at the time of death.

This systemic condition was also sustained by the observation of non-occlusive initial thrombi composed of fibrin and granulocytes within the small pulmonary vessels. These ancillary findings were consistent with the hypothesis of an infectious/septic state [47].

Granulocytes, immunohistochemically marked by CD15, were also found in the renal glomeruli, indicating an ongoing chemotactic process [48, 49].

On this basis, the cause of death was deemed as a P. multocida septic arthritis with likely evolution into sepsis and septic shock. The deceased was previous healthy, except for a moderate osteoarthritis.

Despite the absence of an ante-mortem diagnosis and although several tests were not performed at the ED (e.g. blood analysis and ultrasonography of the shoulder), the suspicion for medical liability was waived as the correct diagnosis in life would have been challenging (clinical signs, symptoms and radiological images were non-specific and the contact with a stray cat was not medically reported) and the mortality of septic arthritis is reportedly high. On the other hand, the local infiltration of methylprednisolone might have facilitated the proliferation of P. multocida in the joint cavity as well as its spreading, by inhibiting the migration and phagocytic function of neutrophils and macrophages.

Our literature review further highlighted that septic arthritis as a cause of death is rarely encountered in forensic pathology, because the diagnosis is usually established ante-mortem. Moreover, other pathogens than P. multocida are usually involved. Indeed, only one other case of fatal septic arthritis due to P. multocida was reported and the pathogen was identified during ante-mortem cultures. However, as emerged in the cases here revised, symptoms of septic arthritis might be non-specific (with no fever, general malaise, signs of arthritis) leading to delayed or missed diagnosis [15, 17, 19, 21, 23, 25, 29]. In similar cases, a judicial autopsy might be requested to clarify the mechanism of death or for a suspicion for medical malpractice.

According to our results, fatal outcomes in septic arthritis are more frequently reported in middle-aged male patients, with associated conditions such as systemic or immune involvement, that should raise a high level of suspicion for infective causes of death in the forensic pathologist.

Individuals with defective immune system are particularly at risk for P. multocida infections, which can be additionally more severe in children or in older people [50, 51], underlyining the need for a greater vigilance in these subpopulations. Nevertheless, even healthy subjects may develop P. multocida widespread septicemia, sepsis, and septic shock, as shown for other pathogens such as Corynebacterium striatum [2, 21, 29, 52, 53].

Beside conditions compromising the liver or immune function, septic arthritis shows a predilection for affecting joints already damaged by trauma [24], surgery [23] or degenerative diseases, as degenerative osteoarthritis in the case here described [54,55,56,57,58]. Similar conditions, with particular reference to degenerative diseases, might be increasingly encountered in the forensic practice with the aging population, and it should be kept in mind that a concurrent infection might occur.

A clear history, e.g. of animal contacts, might not be reported, underlying the need for a greater awareness among doctors regarding the potential complications of animal-related injuries and the importance of a targeted history. Particularly, when preexisting degenerative diseases are considered responsible for the pain in the clinical setting, the autopsy might represent the only mean to achieve a diagnosis. In this setting, risk factors for septic arthritis should be documented during the post-mortem examination or sought for during the forensic investigation, including contacts with animals.

The first indication at autopsy, beside the absence of other pathological causes, might be represented by purulent material at the joint or abscesses and these findings should initiate the careful collection of microbiological samples. Additionally, signs of sepsis in other tissues such as thrombi in the kidneys or in the lungs should be sought, although most postmortem findings reported in the literature are non-specific.

In cases of suspected sepsis with limited pre-mortem information and an unclear focus of infection, the examination of joints during autopsies is recommended and may represent the only opportunity for a forensic diagnosis of the cause of death. This practice can uncover critical information that may have been overlooked during clinical evaluations and can also clarify infection pathways, address clinical discrepancies, and ultimately contribute to enhancing future patient care in septic conditions. The insights gained from such examinations can significantly impact clinical approaches to managing sepsis and related complications.

In conclusion, the autopsy performed in a multidisciplinary manner, corroborated by ancillary tests and thorough circumstantial data, still remains the gold standard examination in explaining the cause of death, despite the complexity of the case. Moreover, injuries provoked by domestic or wild animals should not be underestimated, even if they leave only mild or no sign, and a history of animal scratches and bites, if reported, should raise the suspicion for serious infections and even a septic death.