Atezolizumab-induced scleroderma: a rare complication

Abstract

Few cases of programmed death-ligand 1 inhibitor-induced scleroderma have been reported and their clinical features remain unpublished. Optimal management is, therefore, unknown and an autoantibody association has yet to be identified. We present the case of a female in her 60s who developed skin thickening after starting atezolizumab for metastatic non-small cell lung cancer. Skin biopsy 7 months after symptom onset showed histological changes consistent with scleroderma. Anti-PM/SCL-75 antibody was positive. Atezolizumab was discontinued and treatment was started with mycophenolate mofetil. After 5 months, she experienced mild improvement in skin thickening. Earlier identification of this complication may limit morbidity in this disease process, which otherwise has limited treatment options. In suspected cases, obtaining scleroderma-associated autoantibodies may help with earlier diagnosis.

Background

Sclerodermoid reactions are a rare adverse effect of immune checkpoint inhibitor (ICI) therapy, usually in association with programmed cell death protein 1 (PD-1) inhibitors. Two cases have been observed with the programmed death ligand 1 (PD-L1) inhibitors atezolizumab and durvalumab, but their clinical features were not documented. In published cases of ICI-induced sclerodermoid reactions, skin changes include morphea as well as the confluent thickening more typical of scleroderma, that is, limited cutaneous systemic sclerosis and diffuse cutaneous systemic sclerosis. Per review of previously published cases, there does not appear to be a pattern of antibody profiles correlating with extent of skin disease. Steroid tapers have been used with varying levels of success and addition of mycophenolate mofetil (MMF) and cyclophosphamide was found to be useful in poor responders.^{1 2} We present the case of a female in her 60s who developed scleroderma induced by treatment with the PD-L1 inhibitor atezolizumab.

Case presentation

We encountered a woman in her 60s who noted skin thickening of both feet 15 months after initiation of atezolizumab therapy for treatment of metastatic large cell neuroendocrine lung cancer (LCNEC). This complication progressed 3 weeks later to the right upper extremity from the mid-bicep distally, then to the left upper extremity. Abdominal wall swelling was also noted, and the patient was prescribed furosemide 20 milligrams twice daily for presumed anasarca. It was thought that this symptom was a side effect of the atezolizumab. There was transient improvement in her symptoms, though not in upper extremity thickening. Due to progression of skin thickening, atezolizumab was stopped after 21 months of therapy, that is, 6 months after symptom onset.

The patient was initially diagnosed with non-small-cell lung cancer of the left upper lobe, pT2a pN0 M0, stage IB in June 2017. In the following months, she underwent left upper lobectomy and adjuvant chemotherapy with cisplatin and etoposide. She had a CT scan of the chest, abdomen, and pelvis, as well as a positron emission tomography scan done in December 2018, which was suggestive of recurrence. Biopsy of a large left adrenal lesion at this time showed metastatic large-cell neuroendocrine carcinoma. One month later, she was started on carboplatin, etoposide, and atezolizumab for disease recurrence. This was continued for cycles 1–4, and starting with cycle 5, she was switched to maintenance atezolizumab. Repeat imaging in April 2019 showed significant reduction in the large left adrenal mass, and resolution of a previously noted left periaortic lymph node. Repeat CT scans of the chest, abdomen, and pelvis done at 3–4 months intervals during 2020 showed stable disease with no new masses or nodules. During oncology follow-up in October 2020, the patient was noted to have progressives skin changes, prompting a referral to rheumatology.

The patient’s first rheumatology evaluation occurred 3 weeks after discontinuation of atezolizumab. At that visit she reported dry skin, reddish discolouration of the fingers and toes, skin tightness of the face and neck with difficulty opening her mouth, a change in voice, decreased appetite with a 60-pound weight loss and dysphagia to solids. She denied Raynaud’s or arthralgias.

Medical history was significant for chronic obstructive pulmonary disease, anxiety and depression. She did not report a family history of autoimmune disease. Surgical history was significant for left lung upper lobectomy and left adrenalectomy. Of note, the patient received cisplatin and etoposide for LCNEC 2 years before atezolizumab initiation, and received carboplatin and etoposide during the induction phase with atezolizumab.

Physical examination revealed restricted opening of the oral aperture, indurated skin of the lower extremities from the distal leg to foot (seen in figures 1 and 2), indurated skin of the upper extremities from the upper arm to fingers, positive prayer sign, extensive xerotic ichthyosis form changes of the trunk and thighs, and erythema of the digits extending distally from the metacarpophalangeal and metatarsophalangeal joints. There were also stage II ulcerations on the inferior lateral buttocks and thighs. There was no telangiectasias, calcinosis, synovitis or muscle weakness.

Figure 1.

Indurated skin of bilateral feet.

Figure 2.

Indurated skin of the lower extremity.

Investigations

Lab work revealed antinuclear antibody titre of 1:40 with homogeneous pattern, C reactive protein 1.2 mg/dL (reference range <0.5), erythrocyte sedimentation rate 33 mm/hour (reference range 0–25), elevated aldolase of 12.5 U/L (reference range 1.0–7.5), low creatine kinase of 20 U/L (reference range 96–140) and moderately positive anti-PM/Scl-75 of 40 Units. SCL-70, RNA polymerase III, ribonucleoprotein (RNP) antibodies, anti-Th/To, anti-U1 RNP, anti-U3 RNP and anti-Ku were negative. Thyroid-stimulating hormone was within normal limits.

Punch biopsies of the right hip, forearm and thighs were performed and showed loss of pilosebaceous units accompanied by dermal thickening and a mild superficial and deep perivascular and perieccrine lymphoplasmacytic infiltrate, histological findings consistent with morphea or scleroderma (findings shown in figure 3). In addition, the right hip and right forearm biopsies showed lichen sclerosis et atrophicus-like changes. There was no evidence of eosinophilic fasciitis or a process mimicking eosinophilic fasciitis.

Figure 3.

Pathology slide showing the loss of normal epidermal rete architecture and dermal thickening accompanied by enlarged/swollen collagen bundles.

CT chest with contrast showed stable postsurgical scarring and volume loss, no evidence of residual or recurrent tumour post left upper lobectomy and no findings suggestive of interstitial lung disease. Transthoracic echocardiogram showed a normal right ventricular systolic pressure of 27 mm Hg. Gastrointestinal involvement could not be confirmed as the patient declined the offer for barium swallow to reduce exposure to the healthcare system during the coronavirus pandemic.

A diagnosis of checkpoint inhibitor-induced scleroderma was made based on the temporal relation of her symptoms to the initiation of atezolizumab.

Treatment

At follow-up visit 2 months after cessation of atezolizumab therapy, the patient noted very mild improvement of skin thickening. Further management was difficult to determine as optimal treatment has not yet been elucidated in these circumstances. The risks and benefits of immunomodulating therapy versus steroid therapy were discussed with the patient. Due to the risk for scleroderma renal crisis associated with the use of steroid therapy in classic scleroderma, the patient declined treatment with prednisone. She was agreeable to starting MMF at low dose, with taper up to maintenance dose. MMF was started at 500 mg two times per day and titrated over 10 days to a maintenance dose of 1500 mg two times per day. The importance of regular lab monitoring was discussed with the patient, given the risk of developing cytopenias with MMF, particularly in the setting of malignancy posing an increased infection risk.

Outcome and follow-up

Four weeks after initiation of MMF, the patient reported improvement in voice changes and dysphagia. At 5-month follow-up, she reported resolution of dysphagia and improvement of skin thickening. Objective assessment was limited as follow-up occurred via video visit in keeping with the patient’s preferences.

Discussion

Since 2011, multiple ICIs have been Food and Drug Administration (FDA) approved for the treatment of various tumour types. These include drugs that target cytotoxic T-lymphocyte-associated protein 4, PD-L1 and PD-1.³ Activation of the CTLA-4 pathway, which involves binding of CTLA-4 with the ligands CD80 and CD86, results in an inhibition of T cells that are involved in the immune response against cancer cells. Activation of the PD-1 pathway, which involves interaction of PD-1 with the ligand PD-L1, also inhibits these T cells involved in the antitumour response.⁴ By inhibiting these pathways, ICIs have had a large role in the treatment of multiple types of malignancies.⁵

By increasing antitumour activity of T-cells, these agents produce expected inflammatory side effects, known as immune-related adverse events (irAEs). Rheumatic disorders, including inflammatory arthritis, polymyalgia rheumatica (PMR), myositis, Sjögren’s syndrome, vasculitis, sarcoidosis and scleroderma, have all been described among patients receiving treatment with these agents. Inflammatory arthritis, PMR and myositis are most frequently described, with musculoskeletal manifestations occurring in 5%–7.7% of these patients. In most cases of inflammatory arthritis and PMR, ICI therapy can be continued, and patients typically improve with less than 20 mg prednisolone daily. In cases of myositis, however, ICIs are usually discontinued, and patients may require treatment with higher dose steroids.⁶ ICI-induced scleroderma has only rarely been described in the literature, and there is no definitive management.²

In their 2020 review of the WHO pharmacovigilance database, Terrier et al identified 35 cases of ICI-induced scleroderma. Thirty of these cases were observed after treatment with the PD-1 inhibitors nivolumab and pembrolizumab, two were observed after treatment with the PD-L1 inhibitors atezolizumab and durvalumab, and three were observed after treatment with the CTLA-4 inhibitor ipilimumab. Terrier et al reported two additional cases of scleroderma post-treatment with pembrolizumab, bringing the total number of cases described in the literature to six, each after treatment with a PD-1 inhibitor. To our knowledge, ours is the first published case of scleroderma observed after treatment specifically with a PD-L1 inhibitor. The six previously described cases are summarised in table 1.

Table 1.

1 7 11 12

Patient information	Type of malignancy	ICI used	Autoantibody profile	Treatment	Outcome
Male in 70s	Proto-oncogene B-Raf (BRAF) wild-type stage IV metastatic melanoma	Pembrolizumab 2 mg/kg every 3 weeks	Negative: ANA, anticentromere, anti-RNP, anti-topoisomerase 1	Held pembrolizumab; started prednisone 1 mg/kg daily	Attempts to taper prednisone were unsuccessful; required multiple escalations to 60 mg daily.
Male in 60s	BRAF positive stage IV metastatic melanoma	Pembrolizumab 2 mg/kg every 3 weeks	Negative: ANA, anticentromere, anti-RNP, anti-topoisomerase 1	Discontinued pembrolizumab; started intravenous immunoglobin 0.4 mg/kg daily for 5 days monthly, mycophenolate mofetil 1000 mg two times per day	Melanoma was in complete remission and skin thickening subjectively improved, however, the patient later developed diffuse muscle weakness; further workup was deferred as the patient chose hospice care.
Male in 60s	Oligometastatic renal cell carcinoma	Nivolumab (dose not specified)	Negative: ANA, anticentromere, anti-RNP, anti-topoisomerase 1	Discontinued nivolumab; started prednisone 1 mg/kg daily with slow taper	Symptoms improved until prednisone tapered to 10 mg daily, after which skin tightness and oedema worsened. Mycophenolate then added with slow prednisone taper.
Male in 70s	Malignant melanoma	Nivolumab 2 mg/kg every 3 weeks	Negative: ANA, anti-dsDNA, anticentromere, anti-topoisomerase 1, anti-RNA polymerase III, anti-U1 RNP, anti-Ro, anti-La	Nivolumab was continued; started 20 mg prednisolone daily	Symptoms improved within 2 weeks; prednisolone was tapered down and the patient had no recurrence of symptoms.
Female in 50s diagnosis of lcSSc prior to initiation of ICI	Metastatic non-small cell lung cancer	Pembrolizumab 2 mg/kg every 3 weeks	Positive ANA, RNA polymerase III	Pembrolizumab was discontinued; intravenous CYC initiated	Progression to proximal skin thickening and SRC before initiation of immune suppression. Improved skin thickening after six cycles of CYC.
Female in 40s	Metastatic non-small cell lung cancer	Pembrolizumab 2 mg/kg every 3 weeks	Positive ANA 1:640 Negative Scl70, centromeres, fibrillarin, RNA polymerase III	Pembrolizumab was discontinued; started prednisone 10 mg daily	Progression of skin thickening 6 weeks later despite steroid therapy. Intravenous CYC added to prednisone.

ANA, antinuclear antibody; CYC, cyclophosphamide; ICI, immune checkpoint inhibitor; lcSSc, limited cutaneous systemic sclerosis; RNP, ribonucleoprotein; SRC, scleroderma renal crisis.

The mechanism by which ICIs induce scleroderma remains unclear. It has been previously demonstrated that increased serum levels of soluble PD-1, which antagonises PD-1, has been associated with increased severity of skin sclerosis in systemic sclerosis. Therefore, it would be expected that inhibition of the PD-1 pathway could result in the observed outcomes described in this case report.² In regard to this pathway, it has also been suggested that inflammation caused by PD-1 inhibition could lead to transforming growth factor beta activation within the skin, resulting in a profibrotic cascade.⁷

One retrospective study⁸ determined there may be a relationship between irAEs and the HLA genotypes of patients. Specifically, inhibitor-induced pneumonitis was correlated with expression of HLA-B*35 and HLA-DRB1*11. In patients with autoimmune disease, the product of the B*35 gene has been associated with expansion of inflammation-associated cell lineages. Meanwhile, expression of DRB1*11 has been associated with systemic sclerosis, as well as impairment of the T cell cytotoxic response to specific viruses. Another retrospective study⁹ performed a retrospective analysis of patients with metastatic non-small cell lung cancer who received PD-1 inhibition with nivolumab. It was demonstrated that the progression-free survival) of patients was strongly correlated to expression of HLA-A*01 and/or A*02 alleles, while patients with less frequent expression of these alleles had a poor prognosis.

Furthermore, a case report from June 2021¹⁰ discussed the case of a patient with metastatic bladder cancer who developed myositis signs and myasthenia-like syndrome during treatment with PD-1 inhibitor pembrolizumab. Of note, an analysis of this patient’s HLA genotype revealed an autoimmune haplotype associated with an increased risk of myositis and myasthenia gravis (HLA A*02, B*08, C*07, DRB1*03).

Considering the information from both retrospective studies and the case report, it has been suggested that further research of HLA haplotypes prior to initiation ICI therapy could provide further insight on the risk of developing irAEs. Further studies will need to be done in this area to develop a better understanding of the mechanism behind these findings, as well as to analyse the incidence of sclerodermoid reactions among patients being treated with ICIs based on which protein is targeted.

Patient’s perspective.

I’ve never experienced anything as difficult as this time in my life. I’ve had three surgeries over the last two years and that was nothing compared to this. This really impacted every aspect of my life and made day to day function challenging. I haven’t had the same life I had before all of this happened. After stopping the cancer drug and getting the treatment for this, I’m doing a little better, and getting closer to normal every day. I’m happy to share my story for others who may be experiencing the same thing.

Learning points.

• New oedema without identifiable cause in patients on immune checkpoint inhibitors (ICIs) should prompt suspicion for ICI-induced scleroderma.

• Consider early cessation of ICI in patients with ICI induced scleroderma as these patients are usually resistant to steroid monotherapy even after ICI is discontinued.

• Consider treatment other than ICI in patients with a history of scleroderma as therapy may lead to worsening of existing scleroderma.

• In patients with ICI-induced scleroderma without a history of autoimmune disease, there is no definite evidence yet of systemic involvement beyond skin thickening.

• Continued antibody profiling in patients with ICI-induced scleroderma may help determine those at risk for resistant skin disease and additional organ involvement.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s)