Infusions Of Infliximab for Pyoderma Gangrenosum Leading to Multi-System Thrombosis and Death

Aim: To present a case of a rare side effect from the tumor necrosis alpha inhibitor infliximab causing probable multi-organ system thrombosis which eventually led to death.

Methods: Infliximab infusion at 5mg/kg was given twice at 2-week intervals for the treatment of pyoderma gangrenosum.

Results: Both the first and second infusion caused immediate excruciating pain in both legs with the presumptive diagnosis of microvascular thrombosis of small vessels in the legs. The second infusion led to acute renal failure, possibly from renal artery thrombosis. She also had a thrombotic stroke. Various treatments for this complication which led to a cascade of other adverse events leading to death of a 51-year-old woman with a history of rheumatoid arthritis.

Conclusion: This case emphasizes the importance of heeding the label warning not to give a second course of infliximab if the first one caused a serious adverse event.

Infliximab has been used to treat pyoderma gangrenosum [1]. According to the regulatory prescribing information (package insert/FDA label) there are case reports of infliximab leading to thrombotic events, including venous thrombosis, arterial thrombosis, renal artery thrombosis/infarction, pulmonary embolism, stroke, and microvascular thrombosis.

Vasculitis is another uncommon side effect of infliximab. Both thrombosis and vasculitis can lead to temporary or permanent kidney failure. 

Presented here is a case of a 51-year-old woman who was treated with intravenous infusion of infliximab, resulting in multisystem thrombosis which precipitated a series of events that led to a critical morbid state with eventual death within two months of her first infliximab injection.

A 51-year-old woman suffered from joint pain from rheumatoid arthritis since her mid-twenties. She did not respond very well to a high dosage of corticosteroids. She was also treated by etanercept and adalimumab, but each one led to a severe local reaction with marked swelling, erythema, and pain at the injection site. They were eventually stopped not only because of the local reaction, but they provided no relief of her arthritis. Thus, she continued for years on high dosage corticosteroid for minimal but somewhat decreased arthritic pain. Related to various side effects from chronic prednisone therapy, she had consulted our group, and she was placed on dextroamphetamine sulfate as previously described [2]. 

From the fatigue and joint pain standpoint, she had a moderate improvement with the combination of low dosage prednisone (5-10mg) and 30 mg amphetamine salts immediate release tablet. Eventually, over the later years she increased the amphetamine salts to 60 mg per day. For the last 20 years on this combination her fatigue and joint pain was moderately improved.

Because of the previous usage of high dosage prednisone and the continued usage of lower dosage glucocorticoids, she did have some complications e.g., osteonecrosis of the jaw requiring partial jaw replacement. She also suffered from very thin skin which bruised and cut easily and did not heal well.

At the age of 51 she noticed a lesion on each shin. At first, she thought she must have injured her shins, and they were not healing well. It presented as nodules that turned into ulcers. They were painful but bearable, and she did not require any type of pain medication.

She consulted a dermatologist who biopsied the lesions and concluded that she had pyoderma gangrenosum. As an outpatient, he gave her an intravenous infusion of infliximab at the dosage of 5 mg/kg. 

The next day she complained of excruciating pain in both legs. The dermatologist had her admitted to the hospital where he had staff privileges, which is a top-ranked university hospital. The presumptive diagnosis by the dermatologist was that she was suffering from a secondary bacterial infection from her pyoderma gangrenosum lesions which he believed was supported by her very high white blood cell count of over 30,000 per microliter with predominantly neutrophils. However, the infectious disease consultants found no evidence of infection and believed that her severe bilateral leg pain (which was so severe she needed a wheelchair because she could not stand up) was an adverse side effect related to the infliximab. 

Nevertheless, the infectious disease physicians, in deference to the opinion of the dermatologist, treated her with a wide variety of intravenous antibiotics. However, the white blood cell count remained elevated, and her pain was only mildly improved despite the addition of opiates. She was finally discharged from the hospital with a prescription for oral antibiotics and oxycodone. However, the dermatologist still was of the opinion that the key to treating the “infection” from the pyoderma gangrenosum lesions was to heal the pyoderma so he gave her another course of infliximab. Thus, two weeks from the first injection as she was given a second infusion of infliximab prior to her discharge from the hospital. 

We do not have medical privileges at the hospital in which she was admitted. However, she called to get our opinion, we advised her our opinion was that the severe pain sounds suspicious of an adverse reaction to the infliximab. Thus, we advised her not to take a second course of infliximab. She was advised even in the package insert; it states not to take another infusion of infliximab if there was a reaction to the first dosage.

Unfortunately, the dermatologist, steadfast in his opinion that she did not have an adverse event from the infliximab, suggested that without the second course of infliximab her pyoderma gangrenosum will not heal and infection will progress. Thus, he convinced her to take a second course of infliximab. Her pain at the end of the two weeks before the second course of infliximab had been considered to be only mildly relieved but bearable with the addition of opiates. The day following the second infliximab infusion, the pain was so intense in her legs that it was 2 to 3 times worse than the first one, even with opiates, which she did not have the first time. She stated that the only method to reduce her pain was to scream as loud as she could.

She did not wish to return to the hospital, so she consulted us by telehealth, and we added the dopamine agonist carbidopa levodopa 10/100 mg to the 60 mg of amphetamine salt (which she resumed). Her very beneficial response to this treatment has been published in another case report, so all the details will be omitted [3]. However, the pain was so much better, and the lesions were healing so that she no longer thought she needed opiates. She knew she could not stop them suddenly because of the risk of withdrawal side effects, so she began weaning off the opiates. More details have been provided in the aforementioned case report [3]. 

Despite what appeared to be clinical improvement, she was found comatose. The rescue team gave her naloxone, and she immediately woke up. She was still taken to the closest hospital, which is also a university-based hospital. The cause of her apparent “overdose” of oxycodone, despite reducing her dosage, was evident when it was found that she was in acute renal failure. This was followed by a thrombotic stroke.

Following anticoagulant therapy with unfractionated heparin, her kidney function returned to normal, and she did not seem to have any obvious permanent damage from the stroke. The hospitalists physicians discontinued the dopaminergic drugs and her pyoderma lesions returned. However, they were not associated with painful legs suggesting that the previous pain following infliximab infusion was an adverse event from the drug and not infectious because her white blood cell count remained over 30,000 cells per microliter.

Her hemoglobin dropped from 10 g/DL to 4.5 g/DL without evidence of bleeding. Though there was no rise in serum bilirubin or the presence of schistocytes, the hospitalist first thought that the drop in hemoglobin was related to an autoimmune process and treated her with a high dosage of corticosteroids which persisted throughout her course of this hospital stay despite a drop in serum calcium to a very dangerous life-threatening level.

While she was at the first hospital a small infiltrate of unknown origin was noted in her upper left lung. Possibly related to the combination of immunosuppressives i.e., infliximab, coupled with high dosage corticosteroids, the lung lesions progressed despite a multitude of anti-bacterial and anti-viral agents and amphotericin B for possible fungal infection. Meanwhile, she developed severe thrombocytopenia. She required both transfusions of red cells and platelets.

The pneumonia continued to progress, requiring her to be placed on a ventilator. She died from the pneumonia. A limited autopsy failed to find her bone marrow infiltrated with white blood cells, i.e., no evidence of leukemia. Fungal cultures of her sputum were taken prior to death, but the test results were not available before she died. 

The perceived complication from taking infliximab is not known for certain. In our opinion, the infliximab caused microvascular thrombosis starting in her legs causing severe pain from ischemia. Though vasculitis is another possibility, the sudden pain favors thrombosis as does the development of a stroke.

Though the kidney failure could have been from thrombosis or vasculitis, the quick return to normal kidney function after anticoagulation favors thrombosis in the authors’ opinion. Related to the quick recovery of renal function, no radiographic studies of the kidney or biopsies were performed. Though microvascular thrombosis could account for the sudden kidney failure leading to lack of proper excretion of the opiates resulting in an overdose, other possibilities include renal artery thrombosis, or even venous thrombosis. Though eventual use of high dosage corticosteroids could have resolved vasculitis, kidney function returned after anticoagulant therapy and before the use of corticosteroids.

One cannot completely exclude an infectious cause of acute kidney injury, but she had already been taking antibiotics for a month, and the antibiotics did not stop the progression of pneumonia. There was no kidney biopsy, so it is not clear whether there was damage to the glomerular capillaries or small renal vessels or both.

The immunosuppression from the infliximab, coupled with the use of high dosage corticosteroids used not only to treat the pneumonia, but to hypothetically restore red cell and platelet function that may have been the result of an autoimmune complication from heparin or the multitude of different antibiotics she received initially to treat a possible skin infection from pyoderma gangrenosum (which probably was not necessary because the pain was more likely from small vessel thrombosis and subsequent ischemia). All of these factors contributed to her demise by not allowing her immune system to fight the pathogen(s), i.e., infliximab and high dosage corticosteroids. 

It is sad that the dermatologic expert did not know about the possibility of using dopaminergic drugs for various skin disorders because possibly he would have not treated her with the infliximab knowing that previously she could not tolerate TNF alpha inhibitors or at least not giving her the second infusion “because there was no other treatment option.”. 

There are only a few published case reports of infliximab causing thrombosis. One reported acute renal artery occlusion following infliximab infusion for ulcerative colitis [4]. Another case reported cerebral venous sinus thrombosis following infliximab for Crohn’s disease [5]. There are also some rare case reports of vasculitis, causing renal impairment, following infliximab therapy for Crohn’s disease [6]. A patient treated with infliximab for Behcet’s disease developed bilateral renal artery thrombosis [7]. There are other publications linking thrombosis to infliximab treatment without necessarily causing kidney failure [8]. Thrombosis has also occurred with other TNF inhibitors [9,10]. 

The patient had leukocytosis with WBC of over 20,000 for 3 years before it was recently found to be over 30,000. A full autopsy was not performed. Possibly there could have been some underlying occult malignancy secreting ectopic granulocyte colony stimulating factor responsible for the increase in neutrophils thus possibly, though increased, these white blood cells were compromised in their function.