Impact of Insulin-Derived Amyloidosis on Glycemic Control and Insulin Dosing

A 57-year-old woman with longstanding type 2 diabetes on basal-bolus insulin therapy for 15–20 years and with a distant history of gastric bypass surgery presented to her primary care physician’s (PCP’s) office with concern about a subcutaneous right-sided abdominal mass. She had first noticed the painless mass about 2 months prior and did not notice any growth afterward. She had no known injury in the area and had not had any surgical procedure in the area.

Physical examination demonstrated a firm mass the size of a golf ball with no overlying skin changes such as erythema, ulceration, or drainage. The PCP ordered a limited abdominal ultrasound to help further characterize the mass. The ultrasound imaging showed a 3.3 × 1.5 × 2.0–cm heterogeneous, hypoechoic structure with similar echogenicity to the adjacent subcutaneous fat (Figure 1). The radiologist noted that it was suggestive of a hernia, although the differential diagnosis could also include a soft tissue mass.

The patient was referred to general surgery for further evaluation and management. A computed tomography scan was ordered that did not show any abdominal wall defect consistent with a hernia but did demonstrate a 3.5 × 3.5 × 2.5–cm hyperattenuating lesion along the right-sided periumbilical soft tissues (Figure 2). Differential considerations from the radiologist included nodular fasciitis or granulomatous tissue.

Because of the uncertainty regarding the diagnosis, the patient underwent surgical excision of the mass by general surgery, and the specimen was sent for further evaluation and characterization by pathology. Tissue sections demonstrated benign adipose and eosinophilic amorphous material. Because of the abundance of eosinophilic material, a Congo-red special stain was ordered and showed apple-green birefringence under polarized light, which is characteristic of amyloidoma (Figure 3). Mass spectrometry confirmed the diagnosis of insulin-type amyloid deposits.

After further discussion with the patient, it was discovered she had been giving her four daily insulin injections—once-daily insulin glargine and thrice-daily insulin aspart before meals—in the same region of her abdomen for years. She was counseled about safe rotation of injection site locations and advised to avoid this region of her abdomen for further injections.

Six months after excision, the patient had outpatient follow-up for her diabetes and confirmed that she was now rotating insulin injections to other regions of her abdomen. She was using 20% less insulin overall, with a total of 33 units/day compared with 41 units/day before the surgery. Despite using less insulin, she had lower blood glucose levels, with an average of 134 mg/dL compared with 146 mg/dL previously and was also having more hypoglycemia, with 24% of readings (11 of 46) <70 mg/dL compared with 8% (2 of 24) previously. Her insulin doses were decreased further to help limit her hypoglycemia risk. During a hospitalization for an unrelated issue 1 month later, her insulin doses were reduced even further, with her insulin requirement ultimately decreasing to 20 units/day on a basal-bolus regimen—half of the total daily insulin dose originally prescribed before surgery.

1. How and why do subcutaneous amyloid deposits develop from insulin injections?

2. How does insulin-derived amyloidosis differ from lipohypertrophy, which can also occur from insulin injections?

Insulin-derived amyloidosis is a rare complication of insulin therapy and is still not fully understood. These subcutaneous amyloid deposits have been seen in areas such as the abdomen, thigh, and arms, which are common sites for insulin injections. However, there has also been a case within breast tissue from insulin injections in this region (1).

The pathophysiology of how insulin is converted into amyloid fibrils is not well understood, although it is suspected that misfolding and polymerization of precursor proteins leads to development of amyloid fibrils (2). Insulin-derived amyloidomas are known to have resulted from repeated injections of insulin analogs, including lispro, aspart, glargine, and detemir insulins (3). It is thought that the process can take between 20 and 23 years to develop for people with type 1 diabetes but may be more rapid in type 2 diabetes, with cases reported 4–18 years after initiation of insulin therapy (4). Our case falls within the later end of this range, with the patient having used insulin for treatment of type 2 diabetes for 15–20 years before this complication was recognized.

Masses typically present as firm, subcutaneous nodules on examination, similar to the clinical presentation for our case. Although diagnosis is based on pathology, there are some clues that may help to distinguish insulin-derived amyloidosis from lipohypertrophy, which can also present similarly. Insulin-derived amyloidosis typically will not regress and may even continue to grow after insulin injections in a region are ceased, whereas lipohypertrophy will typically regress over time (5). Insulin-derived amyloidosis often presents as a firm mass, compared with lipohypertrophy, which is typically softer to palpation (3). However, it should be noted that there are rare cases of insulin-induced amyloid deposits being more diffuse through the tissue, which clinically can appear more like lipohypertrophy (6). In addition, insulin-derived amyloidosis may affect insulin absorption even more than lipohypertrophy, with only about one-third of the insulin absorbed in a prior case of insulin-derived amyloidosis (3,7).

There are several hypothesized mechanisms describing how the amyloidomas may disrupt insulin absorption, including the amyloidoma acting as a physical barrier preventing insulin absorption, injected insulin adhering to amyloid fibrils, or the amyloidoma producing an insulin-degrading enzyme (2).

Imaging is reasonable to consider in these cases to rule out other pathology, although it is typically inconclusive. Ultrasound findings are often similar to those in our case, with an irregularly shaped, heterogeneous lesion with ill-defined margins. Computed tomography will often demonstrate an irregularly shaped lesion with ill-defined margins and may show punctate calcifications, which were not present in our case. Magnetic resonance imaging was not evaluated on our patient, although it may show low signal intensity of the mass compared with adipose tissue on both T1 and T2 weighted images; this can be one of the most specific imaging findings for subcutaneous amyloid deposition (4).

Ultimately, pathologic examination of the tissue is needed to diagnose this condition. Although growth can be limited by rotating injection sites, definitive treatment requires excision of the mass, which otherwise may continue to grow even after halting injections in the affected area (5).

It is important for blood glucose values to be followed carefully by the PCP, diabetologist, or endocrinologist because there may be markedly improved absorption of insulin after removal of the mass or movement of injection locations. Because of this improved absorption or action, hypoglycemia is common after a patient changes injection sites or after the mass is removed (8). It would be reasonable to consider reducing doses in high-risk patients immediately to avoid severe hypoglycemia (8). In our case, despite using lower doses of insulin after the surgery, the patient continued to have more hypoglycemia on her home glucose meter readings and eventually needed her overall total daily dose of insulin halved to maintain euglycemia.

Although there are no specific guidelines for evaluation and management of this condition, it is important for providers to be aware of this complication, which can influence treatment decisions. Patients who develop tissue growth in the region of insulin injections should be counseled to rotate injection sites more aggressively and to avoid injecting into the affected region until it is improved. If it does not improve, it may be reasonable to consider a workup, which may include imaging or excisional biopsy with histopathological confirmation. Patients should be cautioned about the risk of hypoglycemia, and their insulin doses may need to be reduced proactively after counseling them about rotating injection sites because of the increased risk of hypoglycemia with improved insulin absorption or action in unaffected areas.

• Insulin-derived amyloidosis is a rare complication from long-term insulin injections but can have a clinically meaningful impact on insulin absorption, action, and dosing.

• Subcutaneous amyloid deposits are typically firm and can persist after injections in the region are stopped, as opposed to lipohypertrophy, which tends to be softer and more diffuse and will typically regress with rotating injection sites.

• Providers should consider insulin-derived amyloidosis as a differential for patients with local subcutaneous masses in regions of repeated insulin injections.

• Providers should counsel patients on rotating injection sites frequently, especially for patients requiring multiple daily insulin injections, to help avoid insulin-derived amyloidosis and lipohypertrophy.

• Insulin absorption and action can be markedly improved with treating this condition or rotating injections away from the affected region. Patients should be cautioned about hypoglycemia risk and may require significant insulin reductions to limit their hypoglycemia risk.

No potential conflicts of interest relevant to this article were reported.

A.E. and J.U. researched data and wrote the manuscript. C.B. and R.J. reviewed and edited the manuscript. J.U. is the guarantor of this work and, as such, takes responsibility for the integrity of the information presented.