Pneumocystis jirovecii pneumonia in a patient treated with trastuzumab-deruxtecan

Abstract

Trastuzumab-deruxtecan (T-DXd) is a novel antibody drug conjugate that has improved treatment outcomes in patients with ERBB2-positive cancer, including locally advanced or metastatic gastric and gastro-oesophageal junction adenocarcinoma. One of the reported side effects of this medication is drug-induced pneumonitis. We present in this case report, a diagnostic dilemma of a patient presenting with clinical and radiographical features of drug-induced pneumonitis but was found to have pneumocystis jirovecii pneumonia (PJP). Our case is the first of PJP in a patient treated with T-DXd, highlighting the increasing incidence of this opportunistic infection in patients with solid malignancy. It also highlights the clinical and radiographical similarities between the PJP and drug-induced pneumonitis.

Background

Antibody-drug conjugates (ADCs) are a class of targeted oncology treatments that are a combination of a monoclonal antibody linked to a chemotherapy drug (known as the payload). These are engineered to deliver the cytotoxic drug directly to the target cells; thereby selectively destroying cancer cells with less systemic toxicity compared with conventional chemotherapy medications.¹

Trastuzumab-deruxtecan (known as fam-trastuzumab deruxtecan-nxki in the USA; T-DXd; ENHERTU) is an ERBB2 (formerly HER2) targeted ADC that has shown improved treatment outcomes in the management of patients with ERBB2-positive cancer; including metastatic breast, gastric or gastro-oesophageal cancer.²

One of the reported side effects of T-DXd is drug-induced interstitial lung disease (ILD)/pneumonitis,³ with fatal events reported in 2.6% of the patients who received T-DXd.⁴

Pneumocystis jirovecii pneumonia (PJP) is a life-threatening opportunistic infection that occurs in patients with profound T cell immunodeficiency. It is much less recognised in patients with solid tumours unless they have other major predisposing factors.⁵

We present the first known case report of PJP in a patient with metastatic gastro-oesophageal cancer undergoing treatment with T-DXd. This patient presented with a concern for drug-induced pneumonitis after T-DXd treatment but was eventually found to have PJP. This case report highlights the diagnostic dilemma of drug-induced pneumonitis and its overlapping clinical and radiological features with infectious aetiologies as well as the rise of PJP infections in patients with solid cancers.

Case presentation

A man in his mid-40s with a history of metastatic gastro-oesophageal junction (GEJ) adenocarcinoma (undergoing treatment with T-DXd) and cryptogenic cirrhosis, presented to the oncology clinic with progressive shortness of breath for a few days. He also had intermittent fevers, dry cough and malaise. A CT scan of the chest was obtained which showed ground glass opacities (GGO) and bilateral reverse halo sign (figure 1). He was treated with a 10-day course of levofloxacin empirically, with mild improvement of his symptoms. However, he again had clinical worsening requiring presentation to the emergency department (ED). On arrival to ED, he was febrile with a temperature of 38.6°C, heart rate of 124 beats/min, blood pressure of 95/68 mm Hg, respiratory rate of 24 breaths/min and oxygen saturation of 95% on room air. Workup showed white cell count of 5.4×10⁹ cells/L, absolute neutrophil count 4.65×10⁹ cells/L (reached a nadir of 0.63×10⁹ cells/L on the fourth day of admission, recovered spontaneously to 1.74×10⁹ cells/L by the sixth day of admission), AST (aspartate aminotransferase) 59 IU/L (35 IU/L on the third day), ALT (alanine aminotransferase) 67 IU/L (49 IU/L on the third day), total bilirubin 1.2 mg/dL (0.5 mg/dL on the third day), INR (international normalized ratio) 1.5, albumin 31 g/L, total protein 53 g/L, procalcitonin of 0.25 (units) and negative COVID-19 PCR. CT chest showed worsening bilateral ground-glass infiltrates (figure 2). Echocardiogram was normal with ejection fraction of 60%–65%.

Figure 1.

Chest CT chest showing GGO and bilateral inverse halo sign. GGO, ground glass opacities.

Figure 2.

Chest CT showing worsened bilateral ground-glass infiltrates.

He was started on broad spectrum antibiotics with cefepime and vancomycin, latter being discontinued 2 days later after nasal MRSA (methicillin-resistant Staphylococcus aureus) culture returned negative. However, he became more hypoxemic requiring up to 6 L/min of oxygen supplementation via nasal cannula to maintain oxygen saturations ≥88%. Given that the patient had been on T-DXd therapy for 8 months prior to presentation with the last dose being 2 weeks prior to hospitalisation and the findings on his chest images, drug-induced pneumonitis was considered as the most likely diagnosis and intravenous methylprednisolone was initiated at 125 mg once followed by 60 mg every 6 hours.

Subsequently, bronchoscopy with bronchoalveolar lavage (BAL) of the right upper lobe was performed. Fluid analysis revealed 46% lymphocytes, 30% polymorphic neutrophils, 21% monocytes and 3% eosinophils. Gram stain showed 1+ gram negative rods and moderate white cell count; however, final cultures revealed 1+ normal flora. BAL cytology was negative for malignancy and organisms. Blood cultures, fungal cultures and acid-fast bacilli remained negative. HIV test was negative. Serum Fungitell was positive, and PJP PCR test was added on to the BAL specimen. The patient was discharged on 4 L/min of oxygen and prednisone 80 mg daily.

BAL PJP PCR, which is a send out test at our institution, subsequently returned positive.

The patient had been diagnosed with GEJ adenocarcinoma three-and-a-half years prior to his admission. Baseline imaging with CT chest, abdomen and pelvis with intravenous contrast demonstrated metastatic disease in the liver and abdominal lymph nodes. The tumour was noted to be positive for HER2 amplification by next generation sequencing (NGS) as well as by immunohistochemistry with 3+ staining; NGS was also significant for stable microsatellite expression and PD-L1 staining (combined positive score) of 10. Since his cancer diagnosis, he received various chemotherapy regimens, including FLOT (fluorouracil, leucovorin, oxaliplatin and docetaxel), FOLFIRI (leucovorin, 5-fluorouracil and irinotecan), pembrolizumab, ramucirumab/paclitaxel, and, for the last 8 months prior to admission, T-DXd. Follow-up CT scans at 3 and 6 months after beginning treatment showed partial response and stable disease, respectively. The patient reported no side effects from T-DXd and thus the treatment was continued with plans for repeat CT in another 3 months.

He was diagnosed with cryptogenic cirrhosis about a year prior to his presentation with initial MELD (model for end-staqge liver disease) score of 8 and Child-Pugh Class of B on diagnosis. His MELD score on admission (during treatment with T-DXd) was 11 with the same category of Child-Pugh class.

Differential diagnosis

The top differential diagnosis of this case based on clinical and radiographical presentation was drug-induced pneumonitis.

Other differential diagnoses included other infections, including bacterial, mycobacterial, viral and fungal infections. However, all the other extensive workup remained negative except for PJP. Left ventricular dysfunction causing volume overload was also ruled out as patient had no clinical signs of volume overload and echocardiogram that was normal.

Treatment

The patient was started on trimethoprim-sulfamethoxazole 15 mg/kg for 21 days for treatment of PJP and switched to prednisone taper recommended for PJP management.

Outcome and follow-up

Patient’s symptoms of shortness of breath and fever and need for supplemental oxygen subsequently resolved. Following completion of treatment for PJP and steroid taper, T-DXd was resumed with a dose reduction to 5.4 mg/kg every 3 weeks as compared with the starting dose of 6.4 mg/kg every 3 weeks. The patient has tolerated an additional 3 months of treatment without his symptoms returning and his most recent CT continues to show stable disease.

Discussion

T-DXd is a novel ADC that is composed of a monoclonal antibody directed against ERBB2 (trastuzumab), conjugated to a topoisomerase inhibitor (deruxtecan). T-DXd was approved by the Food and Drug Administration for the management of ERBB2-positive locally advanced or metastatic gastric or gastro-oesophageal junction adenocarcinoma in January 2021.³

T-DXd has a black box warning for ILD and pneumonitis. Other side effects include anaemia, lymphopenia, thrombocytopenia, severe neutropenia and left ventricular dysfunction.⁴ In DESTINY-Gastric01 trial, 10% of patients (n=12/125) in the T-DXd arm developed ILD/pneumonitis of any grade of which one patient developed a grade 4 event.⁴ Based on a systematic review of 14 clinical trials, the overall incidence of all-grade ILD/pneumonitis cases was 11.40% (n=136/1193).³ The incidence and severity of drug-induced ILD/pneumonitis is documented and evaluated based on the National Cancer Institute Common Terminology Criteria for Adverse Events. Grades range from asymptomatic (grade 1), symptoms limiting instrumental activities of daily living (grade 2), severe symptoms limiting self-care (grade 3), life-threatening respiratory compromise (grade 4) and death (grade 5).⁶

Making a diagnosis of drug-induced ILD/pneumonitis can be challenging; signs and symptoms are non-specific and there are no pathognomonic clinical or radiological findings.³ It should be suspected with any patient on ADC who presents with new-onset or worsening respiratory manifestations, and/or suggestive radiological findings. It is a diagnosis of exclusion and hence, a thorough investigation of other possible causes should be performed. Bronchoscopy with BAL can be used as an adjunct to aid the diagnosis and is recommended to be considered, if feasible. BAL cell count analysis, however, is variable and may include lymphocytosis, eosinophilia and/or neutrophilia.⁷ Bronchoscopy, however, is primarily performed to rule out other causes such as infection, eosinophilic pneumonia or diffuse alveolar haemorrhage. In patients with suspected drug-induced pneumonitis, withdrawal of the offending agent is of paramount importance with consideration of corticosteroids depending on the severity of the symptoms. Determination of reusing the drug should be individualised after weighing the risks versus benefits and depending on the availability of alternative agents.

The imaging findings of drug-induced pneumonitis can mimic PJP which poses a diagnostic challenge. The most common imaging feature of PJP in non-HIV patients is symmetric GGO that are more perihilar and apical with peripheral sparing. Other common imaging finding is mosaic pattern of lung attenuation. Less reported imaging findings include atypical distribution of the GGO, consolidations, nodules or septal thickening. Moreover, PJP is difficult to culture, and diagnosis usually depends on different methods including staining (eg, Gomori’s methentamine silver stain), which carries a challenge in non-HIV patients as most of these patients have low organism burden. Therefore, PCR is recently considered the test of choice for its higher sensitivity.⁸

PJP is mainly recognised in patients with profound immunosuppression for example, HIV infection, high doses of glucocorticoid therapy, solid organ or bone marrow transplantation or any condition that results in defects in cell-mediated immunity. Even though patients with haematological malignancies are also considered at risk, it rarely occurs in the solid tumour patient population. However, recent studies suggest that it may be increasing within this subgroup with possible explanation including the new chemotherapies, as well as higher clinical suspicion, improved diagnostic techniques and increased use of immunosuppressant agents. There is currently a lack of consensus on PJP prophylaxis in patients who are immunocompromised due to chemotherapy.⁵

On literature review, we found only one case of PJP in a patient with metastatic breast cancer that was managed with trastuzumab (in addition to weekly paclitaxel), who also received dexamethasone as part of premedication for paclitaxel.⁹ Our patient did receive steroids as per the protocol to administer T-DXd with 8 mg of dexamethasone as a premedication prior to infusion and 4 mg taken orally at home on days 2, 3 and 4 every cycle (21 days) but this is a much lower dose than the dose that is thought to be a significant risk factor for PJP.

Multiple case reports of trastuzumab-induced pneumonitis have been reported. In most of these cases, patients did undergo a bronchoscopy to rule out infection. Some of the case reports did not comment on whether PJP PCR was checked in the BAL aspirate.^10–12

However, in other multiple cases, no bronchoscopy was performed, and infection was ruled out based on blood and sputum cultures.^{13 14} There are only two reported case reports of T-DXd-induced pneumonitis/ILD so far, one of the patients (grade 4) did not undergo bronchoscopy and the pneumonitis proved to be fatal. The other patient had grade 3 ILD and did undergo BAL to rule out infection.¹⁵

It is worth mentioning that severe neutropenia and febrile neutropenia are also reported side effects of T-DXd.⁴ The patient’s neutrophils were normal on admission (absolute neutrophil count 4.65×10⁹ cells/L). It subsequently dropped and reached a nadir of 0.63 on fourth day during admission and recovered spontaneously.

Lastly, our patient also had cryptogenic cirrhosis. Cirrhosis is not a recognised risk factor for PJP. However, there have been several cases of PJP reported in patients with cirrhosis including a recent case series. However, all the eight patients in that series had advanced cirrhosis with MELD score above 16. Our patient’s MELD score ranged between 8 and 11 and we believe it is less likely that it played a major role in him developing PJP.¹⁶ There are no recommended dose adjustments to T-DXd for patients with cirrhosis.

Patient’s perspective.

This was quite an experience. Regardless, my takeaway remains clear. I’m extremely grateful for those who took care of me throughout my hospitalisation. Additionally, the bronchoscopy with BAL and the appropriate laboratory studies from it saved my life. I thought I most likely had pneumonitis from my chemotherapy. I’m grateful that those who cared for me didn’t settle for the low-hanging fruit. My clinical picture was not improving significantly on steroids and broad-spectrum antibiotics. I was on oxygen for the first and only time thus far in my life. Once my PJP diagnostic from the BAL came back positive I was promptly notified by the pulmonologist. He started me on specific PJP treatment (TMP-SMX four times daily for 3 weeks). Though I don’t remember the exact timeframe, it seemed that my oxygen requirement rapidly decreased to zero over several days. Since then I’ve not required oxygen. I’m fortunate to say that I’m currently still on traztuzumab-deruxtecan (T-DXd) for my cancer. I continue to have radiographic evidence consistent with pneumonitis. I have had another bronchoscopy with BAL that came back negative for infectious etiologies, including PJP. I do have dyspnoea on exertion, but breathe comfortably otherwise. Because there is not currently a better chemotherapeutic option for my case I am continuing with T-DXd in addition to prednisone 20 mg daily to help control the pneumonitis for the foreseeable future.

Learning points.

• This case demonstrates the clinical challenge in diagnosing drug-induced pneumonitis/interstitial lung disease (ILD) and the importance to rule out infection using bronchoscopy with bronchoscopy with bronchoalveolar lavage whenever feasible.

• In clinical studies, 2.6% of patients treated with trastuzumab-deruxtecan (T-DXd) developed grade 5 ILD/pneumonitis and our case report should raise the question of how many of these patients may have had undiagnosed pneumocystis jirovecii pneumonia (PJP) as our case is the first reported case of PJP in a patient being treated with T-DXd.

• Our goal with this case is to raise awareness and help educate oncologists and pulmonologists on the possibility of PJP infections in this population and to strongly consider early bronchoscopy if deemed clinically necessary.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s)