Spontaneous subcapsular and perirenal haemorrhage with retroperitoneal haematoma in a patient with ovarian melanoma metastases

Abstract

A 47-year-old woman was admitted to our clinic for intensive pain in the left flank region. The transvaginal ultrasound showed a left adnexal solid mass with ascites. She had undergone surgical removal of skin melanoma in 2008, but in September 2019, intracardiac metastasis resulting from it had been discovered. CT performed in March 2020 had been negative for other metastases. A full abdomen ultrasound was not performed. During the night, the patient began to show signs and symptoms of hypovolaemic shock. The patient was urgently transferred to the operating room for a video laparoscopy. A vast left retroperitoneal haematoma was diagnosed along with voluminous enlargement of the left ovary. We proceeded with a left adnexectomy and blood transfusion. Subsequent contrast-enhanced CT revealed a left subcapsular, perirenal haematoma and a voluminous retroperitoneal haematoma. Kidney metastasis was also seen. The final histological diagnosis was metastatic amelanotic malignant melanoma of the ovary.

Background

Spontaneous perirenal haemorrhage (SPH) is a relatively uncommon but often diagnostically challenging condition. The incidence of this disease is low, and most of the available literature is represented by case reports.¹ Renal neoplasms are the most common causes of SPH, accounting for 60%–65% of all cases. Renal angioleiomyolipoma is the most common benign neoplasm responsible for SPH (accounting for 30%–40%), whereas renal cell carcinoma is the most common malignant neoplasm (accounting for 25%–35%).¹ Kidney metastasis should also be considered a neoplastic cause.² Vascular diseases are the second common cause of SPH, accounting for 20%–30% of cases.³

As the perinephric space lies centrally in the retroperitoneal area, the pathology can traverse fascial planes, which can rapidly accumulate fluid collections. Clinical manifestations of SPH and spontaneous retroperitoneal haemorrhage are highly variable and may be unclear especially in the early stages, thus resulting in delayed diagnosis. Clinical signs and symptoms vary from abdominal or back pain to catastrophic hypovolaemic shock. Some patients may present with Lenk’s triad, which consists of acute flank or abdomen pain, a palpable flank mass and hypovolaemia.^{4 5} However, the presence of all three symptoms is uncommon, occurring in only 20% of cases. Even an ovarian mass can manifest with flank and abdominal pain. The ovary is a common site of metastases from malignant tumours with 5%–20% of ovarian masses being metastases from primary tumours in other organs. The usual primary sites causing ovarian metastases are the stomach, colon and breast.⁶ The ovary is a rare metastatic localisation for cutaneous melanomas as only 3% of melanomas secondarily involve the female genital system. Ultrasound diagnosis of ovarian metastasis from melanomas is difficult because malignant melanoma cells are described as having numerous patterns and varied cell types. In a large study conducted on 16 categories of metastatic tumours in ovaries, Young and Young describe the extremely broad differential diagnoses posed by metastatic malignant melanoma in the ovary. These include steroid cell tumours, small cell tumours, sarcomas, undifferentiated carcinomas and adult granulosa cell tumours.⁷

In this case report, we present a detailed ultrasound description of a late ovarian metastasis of melanoma in a young woman with SPH and a massive retroperitoneal haematoma, according to the International Ovarian Tumor Analysis (IOTA) classification.

Few published articles describe the ultrasound characteristics of ovarian melanoma metastasis. Although approximately 20% of female patients dying of malignant melanoma have ovarian involvement at postmortem examination, the diagnosis of ovarian metastatic melanoma is seldom made before autopsy.⁸

In addition, we aim to highlight any critical issues in the management of the patient.

Case presentation

A 47-year-old woman was hospitalised in the gynaecology department for left flank pain for 2 days, with a solid left ovarian mass and ascites. The day before, she had visited the emergency room and been discharged with a diagnosis of left kidney colic after pain management. In 2008, the patient had had a cutaneous epithelioid melanoma removed from the left arm—Clark stage II, Breslow depth of invasion 0.59 mm and mitotic index 1×10 HPF. In August 2019, the patient had undergone Angio-CT of the pulmonary arteries for shortness of breath and asthenia, and a diagnosis of thromboembolism with dense tissue present in the right ventricle had been made. A cardiac nuclear MRI had further confirmed the presence of a voluminous right endoventricular mass. The mass was biopsied and histological examination confirmed the cardiac location of malignant melanoma metastases. The patient had begun therapy with enoxaparin (1 mg/kg subcutaneously two times per day), dabrafenib 150 mg orally two times per day and trametinib 2 mg a day.

In the hospital’s gynaecology department, the patient had a blood pressure of 120/78 mm Hg, HR of 78 beats/minute and oxygen saturation of 99% on room air. Laboratory investigations revealed a white cell blood count of 12×10⁹/L; haemoglobin level of 100 g/dL; red cell count of 3.31×10¹²/L; haematocrit of 25.9%; creatinine levels of 1.24 mg/dL (range 0.50–1.20); and estimated glomerular filtration rate (eGFR) of 52 mL/min (normal >60 mL/min/1.73 m²). The patient had been complaining of left flank pain for 2 days. The transvaginal and transabdominal ultrasound performed by a gynaecologist demonstrated a voluminous solid mass shifted to the right, above the uterus, 9×6 cm in size, and mobile with respect to the surrounding structures. It was tender to the targeted pressure of the transvaginal probe with regular external profiles and mixed, non-uniform echogenicity containing anechogenic cystic areas; and was moderately vascularised. Ovarian crescent sign was absent (figure 1) and ascites present. The right ovary was of regular size, surface, and functional activity. The risk of malignancy of the mass appeared high and the ultrasound findings could have been compatible with ovarian metastasis. The radiology department was contacted to perform an ultrasound of the abdomen but the radiologist dismissed the request and recommended performing abdominopelvic contrast-enhanced CT instead. Due to the presence of mild kidney failure, the radiologist recommended starting a drug protocol for kidney protection from the contrast medium and the CT was scheduled for the next day. Overnight, the woman’s condition worsened: the flank pain became increasingly intense and the patient began presenting with pallor, sweating, tachycardia and hypotension (systolic blood pressure <90 mm Hg). Laboratory results showed a severe drop in haemoglobin levels to 5.8 g/dL. The patient was urgently transferred to the operating room to undergo a video laparoscopy with Palmer access. A voluminous retroperitoneal haematoma was seen stretching from the left hypochondrium towards the pelvis. The left ovary was noticeably enlarged in volume, positioned above the uterus, and to the right of the midline. The right ovary showed regular volume and surface with signs of functional activity. We performed a left annexectomy through retroperitoneal access. The left ureter had a regular course and calibre. During the surgery, two bags of concentrated erythrocytes were transfused.

Figure 1.

Ultrasound features of the left ovarian mass: solid mass of 9×6 cm with regular external profiles and mixed, non-uniform echogenicity containing anechogenic cystic areas. Ovarian crescent sign absent.

Treatment

At the end of the surgery, it was suspected that the retroperitoneal haematoma could have been caused by a kidney bleed or that it could compress the kidney itself. Therefore, a decision was made together with the urologist to place a left ureteral stent. Finally, a contrast-enhanced CT was performed. The CT scan showed a large retroperitoneal haematoma (figure 2,) and the presence of a perirenal and subcapsular haematoma of the left kidney with rupture of the kidney capsule (figure 3). Inside the haematomas, hyperechogenic spots were displayed, as in active bleeding. Enoxaparin was stopped and the patient was transfused with another two bags of red blood cells. At the end of the blood transfusion, the value of haemoglobin was 8.4 g/dL. After 12 hours, the haemoglobin was 5.9 g/dL. A second contrast-enhanced CT was performed and showed an increase in the size of the retroperitoneal haematoma without active spread, and the presence of left kidney metastasis (figure 3). The patient was transfused with another five bags of red blood cells and started on 5000 units heparin subcutaneously every 12 hours. Over the following 2 days, creatinine level and eGFR normalised and haemoglobin levels remained stable at 9.7 g/dL. After 7 days of surgery, the patient was transferred to the oncology department of a tertiary hospital. The final histological diagnosis was metastatic amelanotic malignant melanoma of the left ovary with large necrotic–haemorrhagic areas.

Figure 2.

Axial contrast-enhanced CT image showing voluminous left retroperitoneal haematoma measuring 11×8 cm.

Figure 3.

Axial contrast-enhanced CT image showing perirenal, subcapsular haemorrhage (white arrow) and left renal metastasis (dark arrow).

Outcome and follow-up

The patient died in the oncology department 1 month after the surgery.

Discussion

SPH is a spontaneous, non-traumatic, subcapsular and perirenal haemorrhage. Symptoms include unilateral flank pain, general malaise with hypovolaemic shock and a palpable lumbar mass (Lenk’s triad). However, the presence of all three symptoms is uncommon, occurring in only 20% of cases. A meta-analysis of SPH showed that 83% of patients had acute onset of flank pain, 19% had haematuria and 11% had symptoms and signs of hypovolaemic shock.¹ The patient described in our case study above had been reporting of pain in her left flank for 2 days with stable vital parameters. The ultrasound evidence of an ovarian mass associated with flank pain misled us to think of an exclusively gynaecological problem. In the medical literature, the most frequently described ovarian metastases are from the breast, stomach, lymphatic system and colon. In a study involving 67 women with metastatic ovary tumours, Testa et al found that metastases from the stomach, breast, lymphatic system and uterus were solid; whereas metastases from the colon, rectum, appendix and biliary tract were multilocular or multilocular–solid cyst.⁹ Our patient had a solid mass with non-uniform echogenicity containing anechogenic cystic areas. These ultrasound characteristics did not correspond to the ultrasound descriptions of most frequent ovarian metastases described. To our knowledge, only one article had reported in detail, the ultrasound features of ovarian melanoma metastases. The authors reported a bilateral multilocular–solid cyst with a prevalent solid and partially hyperechoic component, and multiple small and equal anechoic loculi.¹⁰ Our case presented as a unilateral solid mass with regular external profiles, and cystic areas, and with an absent ovarian crescent sign. Most case reports published in the literature on metastatic ovarian melanoma described a unilateral ovarian mass.^{11 12}

Still, in a multicentric study on 1377 adnexal masses published in 2010, the authors concluded that the presence of ovarian crescent sign decreases the likelihood of invasive malignancy in adnexal masses.¹³ The emergency situation, the presence of a normal right ovary and the young age of the patient led to the execution of a unilateral annexectomy. Unilateral salpingo-oophorectomy has been proposed as an appropriate treatment for metastatic melanoma involving the ovary if there is no evidence of contralateral ovarian involvement or extra ovarian spread.^{14 15} The delay in diagnosis was not only due to non-specific clinical symptoms in the haemodynamically stable patient, but also due to the exclusion of the abdominal ultrasound. Even if contrast-enhanced CT is the first choice of imaging examination in emergency patients to identify possible causes of bleeding, ultrasound is highly sensitive for the identification of perinephric haematomas.¹⁶ Ultrasound diagnosis of a perirenal haematoma would have led clinicians to suspect a kidney disease. This would have further led to better clinical management of the patient and avoidance of high-risk urgent surgery. Nervertheless, when there is distant metastasis of malignant melanoma, the cancer is generally considered incurable with a 5-year survival rate of less than 10%. The median survival time is 6–12 months and the treatment is palliative, focusing on life extension and quality of life.

In conclusion, this case illustrates the sonographic features of ovarian metastases from melanomas in young women with a recent normal CT abdominopelvic scan. Ovarian metastatic melanoma should be suspected in any patient who presents with a history of malignant melanoma and an ovarian mass. In addition, an abdominal ultrasound should always be performed in a patient with abdominal pain resistant to pain management.

Patient’s perspective.

(N.B: All names of locations below have been replaced by anonymised identifiers—that is, capitalised initials—to grant the subjects’ privacy).

Dear Doctor,

I would like to thank you again for everything you have done and, most of all, for the professional and humane attitude you have shown, even after my wife was transferred to Hospital of M. I am speaking on behalf of both of us. My wife was always happy to receive your messages enquiring about her health. She felt reassured.

The whole story started about 1 year ago. I received a phone call from my wife asking me to take her to the emergency room (ER) as her general practitioner had detected something unusual at the cardiac level. He suspected pericarditis. Once in Hospital of S. ER; she underwent an ECG, after which a request for a cardiology consultation was requested and the first heart ultrasound was done, spotting a thrombotic left ventricle. The following days were quite traumatic, shifting from a thrombosis diagnosis—curable with anticoagulant therapy—to something unexpected and dreadful. In fact, a CT scan spotted a solid mass that did not leave much room for hopeful thinking, particularly because of an arm melanoma that had been removed 11 years ago that could now be connected to the new diagnosis. The doctor told me that my wife would be transferred to Hospital of B. for a biopsy to understand better what we were dealing with. I remember asking her: ‘What size is it? Can it be removed?’. The doctor informed me that the mass occupied 70% of the ventricle’s surface and was well rooted in it and removing it would mean a heart transplant. However, in the case of confirmation of melanoma, the only possibility would have been pharmacological. At that moment, my blood froze: I understood everything. After transferring to Hospital of B, all the exams were repeated producing exactly the same results. The intracardiac biopsy confirmed the presence of melanoma and the doctor told me that my wife had no chance. She was then moved to the oncology ward. She was exhausted by then. The ward, which already had bad connotations, looked 50 years old with rooms for six patients, four of which were terminal. Oncologists proposed molecular therapy as the tumour showed a BRAF-V600 mutation; they reassured us that there was a good chance that this therapy would work. My wife clung to this idea and was hopeful. We then enquired about where she could receive this therapy close to home and were told that the Hospital of M was our destination. Unfortunately, I found out, after some research, that the therapy would not cure her; it would have just increase her life expectancy. Of course, I had to keep this to myself; it was my secret. At the first follow-up CT scan, only 30% of the ventricle’s surface was covered by the mass, and no longer 70%. My wife went back to almost normal life: she went back to work and we went skiing for the New Year’s holidays, as we usually do. The situation seemed better, even if I was aware that we had the sword of Damocles hanging over our heads and that worried me. Often I woke up terrified that she was not breathing. Everything continued regularly; monthly checkups and the therapy prescribed. Unfortunately, the COVID-19 pandemic worsened the situation, as I had to remain outside during her visits and had no chance to speak to the doctors. The intracardiac mass detected by the CT scan, which appeared to have diminished in size, after 4 months seemed slightly bigger. After a short time, my wife reported continuous pain in her left side around the kidney. I took her immediately to the E.R., terrified that it was related to her known condition. The doctors suspected renal colic and after a few hours of painkillers, she was sent home. At 2:00 the same night, I took her back to the E.R., and unfortunately, she never made it back home again. Here, again, an ongoing series of brutal events occurred. She had to be hospitalised in the gynaecology department for the presence of an ovarian cyst and her continuous pain. After a few hours, I was informed by phone call that she would shortly undergo surgery. I rushed to the hospital, also worrying about the condition of her heart not being optimal. Luckily, that surgery was successful. Unfortunately, every step forward was followed by bad news. CT scan of the following day revealed an advancement of her condition with a much larger metastasis that had not been present in the previous CT scan done. Doctor, when you gave me that piece of news the world collapsed around me. Certainly, it was not your fault, but we had just started hoping for the best and instead.

In those days, I realised the end was close. Under COVID-19 restrictions, the visits allowed were few and far between and it was only me and fortunately, a close friend of ours, a nurse in the very same hospital. Unfortunately, social distancing rules made my presence at my wife’s side difficult; making me feel even more useless and helpless, furthered by a feeling of a lack of empathy from some nurses. After 2 weeks, she was transferred to oncology department. She was unhappy about that, probably being aware of the situation. In the oncology department, her therapy was changed one more time to oncological immunotherapy, as the molecular one was no longer effective. After speaking to the oncologist, I went back to my wife who enquired about her condition. I told her the truth: her illness had relapsed and she had to fight against it until the new therapy started to kick in. She told me she had sensed it and that she did not intend to leave this world just yet. I kept on working and taking no lunch breaks so I could be at the hospital by 16:00. I finished my working shift every day with the hope of finding her better and being able to bring her back home for at least a few weeks, but instead, her condition was getting worsened day by day. The last week was heart wrenching. I nevertheless have to say that she never lost her positivity: she was an optimist and had a cheerful personality. As promised, I remained by her side until her very last breath, knowing that I could not do anything else.

Farewell my love, you will be forever missed by all of us.

Learning points.

• A solid ovarian mass in a patient with a history of skin melanoma should create suspicion of metastatic localisation.

• Spontaneous perirenal haemorrhage is a rare, acute onset and potentially life-threatening condition with a myriad of causes.

• Even if contrast-enhanced CT is the first choice of imaging examination for emergency patients to identify possible causes of bleeding, the ultrasound is highly sensitive for the identification of perinephric haematoma.