Modern Oncologic Therapy-Induced Pneumonitis – Clinical Features, Management and Outcome of a Series of 12 Cases

1. Introduction

In recent years, the treatment of certain types of cancer has been revolutionized by the approval of targeted therapies and immunotherapy. Although pulmonary toxicity is a rare adverse effect, the increasing use of these therapeutic agents has facilitated the recognition of such reactions.

For active immunotherapy involving immune checkpoint inhibitors, the incidence of therapy-induced pneumonitis ranges between 2.7% and 10%[1,2,3,4,5,6], with conflicting evidence that a higher incidence is observed in cases involving the combination of immunotherapeutic agents[7]. Concomitant chemotherapy seems to reduce the risk of immune-related pneumonitis[7]. In the case of molecular-targeted therapies, incidence rates vary depending on the specific agent. For instance, anti-CD20 agents such as Rituximab have been associated with a risk of inducing non-infectious pneumonitis ranging from 3.9% to 8.4%[8,9]. Conversely, the literature describes a single reported case of pneumonitis caused by the anti-CD20 agent Obinutuzumab[10]. Additionally, CDK 4/6 inhibitors rarely cause pneumonitis; however, the FDA has issued warnings regarding the potential severity of these adverse events[11].

The prompt recognition of pulmonary toxicity is crucial to preventing potentially irreversible or fatal consequences. Pneumonitis is one of the leading causes of immunotherapy discontinuation[3]. Suspicion usually starts with common symptoms (progressive exertional and even at-rest dyspnea, cough, fever) in a patient with cancer undergoing targeted oncologic therapy for several months. The median time to onset is 3 months, but there is a high variability (9 days to 19 months)[3], and pneumonitis can occur even several months after cessation of therapy[12]. Patients can present with mild hypoxemia (PaO2 more than 60 mmHg and SpO2 more than 88-90% but less than normal ~95%) or various degrees of respiratory failure (PaO2 less than 60 mmHg and SpO2 less than 88-90%). Clinical exam of the thorax can be normal, and lung crackles may be audible in rare cases. Postero-anterior chest X-rays can reveal a patchy interstitial pattern with mild opacity consistent mostly with ground-glass attenuation on CT scan, or more intense opacity consistent with lung consolidation. Most papers that present radiologic features of oncologic targeted therapy-induced pneumonitis discuss less simple chest X-ray images and focus mainly on CT scan patterns. Inflammatory markers can have increased values but are not currently used for diagnosis. These patterns can be localized or diffuse. These clinical (symptoms, signs), contextual (cancer following targeted therapy), and paraclinical elements (chest X-ray, inflammatory markers) can suffice for a high-probability diagnosis, but most clinicians follow a more prudent approach and perform other more invasive and costly medical investigations to ”confirm” the diagnosis.

HRCT scan is frequently used to characterize and localize more precisely the lung lesions as it has a higher sensitivity and specificity for interstitial lung diseases[13]; the most typical and frequent patterns are patchy ground-glass attenuation and consolidation zones distributed in both lungs which are quite similar with the aspects encountered in SARS-Cov2 pneumonia (diffuse alveolar damage/acute interstitial pneumonia/acute respiratory distress syndrome – DAD/AIP/ARDS) or organizing pneumonia (OP) [14,15]. Other patterns were also described: nonspecific interstitial pneumonia (NSIP), hypersensitivity pneumonia (HP), bronchiolitis (BL), radiation recall pneumonitis (RRP)[15], diffuse alveolar hemorrhage (DAH) or sarcoid-like reactions (SLR)[17]. Also, localized lung lesions can be more rarely encountered[10].

Bronchoscopy and lung biopsy can be performed to further characterize the existing lung injury by cell distribution patterns but often cannot distinguish other diagnoses such as various types of non-bacterial pneumonia, alveolar hemorrhage, lung congestion, non-cardiogenic pulmonary edema, sarcoidosis, lymphangitic carcinomatosis etc. The bronchoalveolar lavage is important to exclude an infectious cause (viral, bacterial or fungal). The bronchoalveolar lavage reveals lymphocyte-predominant cellularity most of the time [14]. Still, the existence of other patterns cannot exclude the oncologic targeted therapy lung injury and the presence of malignant cells can be a feature of the subsidiary oncologic disease. For these considerations, these furthermore complex medical tests cannot confirm the etiologic diagnosis. Lung biopsies can also be performed, but the high variability of histopathologic lung injury patterns generated by different drugs and even within a specific class of therapeutic agents[3,14] correlated with the risks, makes this approach less desirable in terms of risk/benefit ratio.

The management of these patients is guided by symptomatology, levels of severity or presence of respiratory failure and the extent of pulmonary involvement[15,16] and is guided by severity grades by CTCAE criteria. For asymptomatic patients, imaging surveillance alone is sufficient, without interrupting immunotherapy or targeted therapy. Depending on the severity of symptoms, discontinuation of immunotherapy or targeted therapy and the initiation of corticosteroid treatment should be considered in moderately to severe patients. If considering corticotherapy, currently there is no standard recommendation of treatment for this type of pneumonitis or of discontinuation of the incriminated therapy. Most recommendations are based on clinical experience, consensus opinion or observational studies or case series[16] and include corticotherapy regimens of various doses (0.5-1 mg/kg) for various lengths (minimum of 6 weeks and no maximum duration recommendation) and with no standard dose tapering rhythm (at least 4 weeks, based on the initial severity, clinical response and tendency to relapse with dose decrease)[18,19].

2. Materials and Methods

2.1. Selection and Description of Participants

This case series includes twelve cases of pneumonitis induced by immunotherapy or targeted therapy diagnosed and managed by the authors in the Marius Nasta Institute of Pulmonology in Bucharest, Romania, between January 2022 and June 2025. The medical data of these cases were collected in July 2025 based on the personal data collection consent form signed by the patients on admission to the hospital, which also covers research purposes. Besides this, verbal consent for the purpose of publishing anonymized medical data was obtained from the subjects who were alive at the moment of the decision to start this study. A formal request was sent to the Ethics Committee of Marius Nasta Institute of Pulmonology on 11.07.2025. The study was conceptualized by the main research team members by the end of July 2025. This research was approved by the local ethics committee of Marius Nasta Institute of Pulmonology (reference number 16362/19.08.2025) and was conducted in accordance with the Helsinki Declaration. The manuscript was reviewed by all team members.

The medical data used in this study are available and can be provided by the corresponding author upon request.

The study aims to add more information to the emerging evidence of diagnostic features and management outcomes of these cases.

2.2. Data Collection and Measurements

A retrospective search was performed to identify the clinical cases diagnosed and managed by the authors in the hospital’s electronic database using diagnostic criteria of pneumonitis induced by oncologic immune/targeted therapy. No specific exclusion criteria were used. Clinical, imagistic, biological, treatment, and outcome data were collected and anonymized. A formal approval for data collection was obtained from the Marius Nasta Institute in accordance with the internal and Ethics Committee regulations.

The main characteristics of the patients included in the case series are detailed in Table 1.

Each clinical case was studied in terms of history, clinical picture, basic tests (chest X-rays, blood tests), further tests (CT scans, bronchoscopy with BAL), treatment and outcome.

Imaging studies (chest X-rays and CT scans) were downloaded from the hospital’s database, anonymized, and further studied with the open-source cross-platform imaging software “Weasis – uptodate version.” A descriptive assessment of the initial CT scan was performed using the suggested features of CT patterns previously identified in the literature [15]. Based on the symptomatology, SaO2, and extent of lung involvement, the CTCAE (Common Terminology Criteria for Adverse Events) criteria were used to grade pneumonitis severity at the initial diagnosis (Table 2). A subjective, semi-quantitative comparison of lung involvement was performed for the imaging exams obtained for each patient at the time of diagnosis/start of treatment and after a time interval (during or at the end of treatment) to assess the outcome in imaging terms. Comparative assessments were made preferably between same-type images (chest X-rays or CT scans) based on availability. These features were added to the follow-up clinical assessment (improvement or worsening of symptoms) to establish the outcome (improved, stationary, worsened, death).

The essential collected data was centralized in tables of results and general observations and considerations were issued by the research team. No statistical analysis could be performed on this type of data.

3. Results

3.1. Onset, Symptoms and Severity

The average and median onset duration was 19.4 and 18.5 months, respectively, ranging from 6 to 48 months. One case had an exceptionally long time to onset of 48 months.

The most frequent symptom observed in 10 out of 12 patients was progressive dyspnea, while only two patients reported dry cough as the sole complaint. In addition to dyspnea, 4 patients also presented with cough. Fever was present in a single patient (case 8), and 2 patients reported loss of appetite (case 3) or chest pain (case 4). 3 patients presented with respiratory failure (SpO2≤88%) (Table 3).

Oxygen saturation at diagnosis was available for all 12 patients (median 97%, range 70–99%), and 3 patients (cases 8, 11 and 12) met the criterion for respiratory failure (SpO2 ≤ 88%). Lung function testing was feasible only in part of the series. DLCO was obtained in 8 of 12 patients (median 44.6%, range 17.4–68% of predicted) and was below 80% of predicted in all of them, with a mild reduction (60–79%) in 2 cases, a moderate reduction (40–59%) in 2 cases and a severe reduction (<40%) in 4 cases. FVC was obtained in 8 of 12 patients (median 66.3%, range 46.2–84.7% of predicted) and was below 80% of predicted in 7 of them, consistent with a predominantly restrictive ventilatory pattern.

The degree of functional impairment did not parallel the CTCAE severity grade. The lowest DLCO of the series (case 2, 17.4% of predicted) was recorded in a patient graded 2 whose oxygen saturation was normal (96%), and 3 of the 4 patients with a severe reduction in DLCO (<40%) were graded 2. Lung function tests could not be obtained in 3 patients (cases 6, 11 and 12), including the most severe case (case 11, grade 4) and one of the grade 3 patients (case 12); the reported functional values therefore reflect predominantly the lower severity grades (Table 3).

3.2. Radiologic and Bronchoscopy Features

In 4 of 12 cases, CT scan showed patchy opacities with mild intensity (ground-glass – GGO) predominantly at the periphery of both lung fields, and these were more consistent with the nonspecific interstitial pneumonia (NSIP) pattern. In 7 of 12 cases, the opacities were more intense, expressing features of lung consolidation, and in 2 of them the consolidations were peribronchovascular. These cases were more consistent with an organizing pneumonia (OP) radiologic pattern. Only one case showed diffuse bilateral centrilobular nodules more consistent with a hypersensitivity pneumonitis (HP) radiologic pattern. The CT patterns are centralized in Table 4.

In 3 cases CT scan was not performed at the control visit given the clinician's consideration that the information provided by the clinical picture of the patient and the chest X-ray is sufficient for assessing the improvement.

Only one case (case 11) was not evaluated with a chest CT scan at the moment of pneumonitis diagnosis, as the patient refused to be admitted to the hospital and perform a CT scan, although the severity was of grade 4 and the diagnosis relied only on the clinical data, the nivolumab treatment, and the chest X-ray that showed a patchy interstitial pattern with mild intensity opacities. The patient received corticotherapy and oxygen therapy at home and improved clinically and radiologically at 1 month and the CT scan performed at 1 month showed an improved NSIP pattern. A comparison was made between chest X-rays performed at the diagnosis and at 3-month intervals and showed complete remission of pneumonitis.

The three types of radiologic patterns (OP, NSIP and HP) are illustrated in Figure 1.

Bronchoalveolar lavage (BAL) was performed in 8 of the 12 included patients, revealing lymphocyte-predominant cellularity in all cases, with no detection of neoplastic cells. In all cases with late onset (cases 2,3,4 and 9), after months of oncologic treatment discontinuation, BAL was performed to exclude alternative diagnoses, but only in 4 of 8 cases of pneumonitis that occurred during oncologic treatment, BAL was performed. In the other 3 cases, BAL was not considered necessary based on the high probability of oncologic treatment-associated pneumonitis, and in case 11 the patient declined the procedure, which would in any case have carried an unacceptable risk of further respiratory deterioration given the severe respiratory failure.

3.3. Treatment and Outcome

Pneumonitis treatment consisted of oral methylprednisolone (Table 5) starting at moderate doses (ranging from 0.4 to 0.82 mg/kg with an average of 0.55 mg/kg) in 11 out of 12 patients, with gradual dose tapering throughout 5 to 16 weeks, depending on each patient's clinical and imaging evolution. The treatment regimens were highly variable and are illustrated in Figure 2. The decision to start with a dose close to the inferior or superior limit of the general recommended interval (equivalent to a prednisone dosage of 0.5 to 1 mg/kg) was made by each clinician based on severity and comorbidities and the following regimen was also adjusted by each clinician based on treatment response and clinical experience with different types of pneumonitis treatment with corticotherapy.

Immunotherapy or targeted oncological therapy was discontinued in 6 patients, while in 4 cases, discontinuation was not necessary as pneumonitis developed after completion of treatment. In 2 patients, the decision was made to continue immunotherapy (in case 1 without and in case 5 in parallel with corticotherapy).

It is noteworthy that the mild symptomatic patient in whom targeted therapy was not discontinued was monitored solely through imaging, without corticosteroid administration, and exhibited spontaneous resolution of the interstitial lung lesions (Case 1).

Symptomatic improvement was achieved in all 12 patients, with resolution of parenchymal lesions in 11 out of 12 cases. The imaging evolution of each individual patient is illustrated in the Appendix. In Case 7, the oncologic treatment was discontinued, and the patient had experienced improvement after 2 months of corticotherapy. Immunotherapy was reintroduced, in parallel with methylprednisolone administration, and this led to a worsening of pulmonary lesions, despite an improvement in clinical symptoms. Ultimately, the patient died of cancer complications.

4. Discussion

Reviewing these 12 cases, a series of discussions are raised.

The first aspect is related to diagnosis, which in our opinion should rely mostly on clinical, onco-therapeutic, and radiologic features and only in limited cases should necessitate more invasive techniques such as BAL or biopsy. Common Terminology Criteria for Adverse Events (CTCAE) are suited for initial severity assessment. For high-grade severity cases, the diagnosis criteria should be more permissive, as bronchoscopy with bronchoalveolar lavage could further increase the patient’s risk. Thus, clinical, radiological, and circumstantial data (history of oncologic immune or targeted therapy) could be sufficient for diagnosis.

The second aspect is related to the so-called CT radiologic patterns. Although they can be helpful in imagistic descriptions, some of them are purely descriptive without pathophysiological correspondence. This is the case, for example, of NSIP, which can lead to confusion with idiopathic or connective tissue diseases NSIP which has a different mechanism. Therefore, this pattern should be labeled differently, for example as acute interstitial pneumonia (AIP) as most of the cases have features similar with viral pneumonia and have closer mechanisms that center around T-cell activity.

The third discussion is related to follow-up. In this case series, 3 cases were managed successfully without the need for a CT scan at follow-up. Clinical evaluation and chest X-ray were sufficient to assess the outcome. Although a native CT scan is non-invasive, using only chest X-ray and not a mandatory CT scan could reduce the costs and discomfort of the patients in case of improvement. Cases with non-improvement or worsening necessitate an extensive follow-up including CT scan. Follow-up assessment time interval should be based on severity and based on this assessment (improved, stationary, worsened) the treatment regimen can be adjusted. Criteria for these follow-up temporary outcomes are yet to be established but it is also a matter of triviality. Mandatory Chest X-ray and SpO2 should be performed at follow-up alongside clinical assessment and in case of improvement no further tests are necessary.

The fourth discussion is related to treatment. Currently, there is no standard recommendation on dosages and duration of treatment. In this case series, using highly variable regimens of treatment, the patients had good outcomes, pointing out that following the current recommendations that are flexible based on severity and response to treatment is the proper conduct so far. Data suggest a severity-dose approach with a starting dose close to 1 mg/kg prednisone for high-grade severity cases and close to 0.5 mg/kg for low-moderate grade severity cases. Studies could be performed in the future using lower-dose and shorter duration regimens or more standardized starting doses and tapering schemes to optimize corticotherapy for the lowest necessary cumulative dose to decrease exposure to known side effects (Cushing syndrome, diabetes, osteoporosis).

The fifth discussion concerns whether to maintain or discontinue oncologic treatment. In this subject, there is some controversy, and the decision is a matter of consensus between the pneumologist and oncologist and in our opinion the decision should be based mostly on severity. In mild-to-moderate severity cases, it could be maintained, and for grades 3 and 4, stopping the treatment should be strongly considered.

5. Conclusions

Immunotherapy and targeted therapy have revolutionized cancer treatment, yielding more favorable outcomes for many patients compared to classic chemotherapy. Although pulmonary toxicity is not a frequent adverse effect, it represents a significant cause of morbidity and mortality in this patient population. Current literature reports a low incidence of immunotherapy- or targeted therapy-induced pneumonitis; however, the trend appears to be increasing due to the growing use of these agents—often in combination—as well as improved recognition of clinical symptoms and interstitial lung abnormalities.

Further complex investigations (bronchoscopy, CT scan) may not be needed for the purpose of obtaining a clearer diagnosis and may only increase the financial costs and medical stress for these patients. Also repeated CT scans for follow-up may not be needed and at least for clinically improving patient’s chest X-ray can suffice for confirmatory reason.

Timely identification of symptoms, radiological changes, assessment of severity and prompt start of treatment play a crucial role in the management and prognosis of affected patients. Current guidelines, largely based on expert consensus, recommend discontinuing oncologic treatment in all cases of therapy-induced pneumonitis. However, considering the risk-benefit ratio, our center’s experience suggests that oral corticosteroid therapy in low to moderate doses administered concurrently with immunotherapy or targeted therapy can yield favorable outcomes. Close and continuous monitoring of patients—through regular clinical, functional, and imaging assessments at intervals of one to three months—is essential.

Further studies are needed to develop clear therapeutic protocols, particularly those addressing the continuation, discontinuation, or reintroduction of immunotherapy following the resolution of the acute episode, as well as to identify specific risk factors that may predispose patients to pulmonary toxicity.

Appendix A.1

Figure A1. Case 1: patient with breast cancer treated with Abemaciclib presented with an organizing pneumonia (OP) pattern on chest CT (A). The lesions resolved after 3 months without corticosteroid therapy or interruption of the oncologic treatment (B).

Figure A1. Case 1: patient with breast cancer treated with Abemaciclib presented with an organizing pneumonia (OP) pattern on chest CT (A). The lesions resolved after 3 months without corticosteroid therapy or interruption of the oncologic treatment (B).

Appendix A.2.

Figure A2. Case 2: A patient with non-Hodgkin lymphoma treated with Obinutuzumab, presenting with an organizing pneumonia (OP) pattern on CT and a left pleural effusion (A), with improvement of the pulmonary consolidations and resolution of the pleural effusion after 2 weeks of oral corticosteroid therapy (B).

Figure A2. Case 2: A patient with non-Hodgkin lymphoma treated with Obinutuzumab, presenting with an organizing pneumonia (OP) pattern on CT and a left pleural effusion (A), with improvement of the pulmonary consolidations and resolution of the pleural effusion after 2 weeks of oral corticosteroid therapy (B).

Appendix A.3.

Figure A3. Case 3: patient with non-Hodgkin lymphoma treated with Obinutuzumab presented with an organizing pneumonia (OP) pattern on chest CT (A). The lesions resolved after 2 months with corticosteroid therapy (B).

Figure A3. Case 3: patient with non-Hodgkin lymphoma treated with Obinutuzumab presented with an organizing pneumonia (OP) pattern on chest CT (A). The lesions resolved after 2 months with corticosteroid therapy (B).

Appendix A.4.

Figure A4. Case 4: patient with non-Hodgkin lymphoma treated with Obinutuzumab presented with NSIP pattern on chest CT (A). Complete resolution of ground glass opacities after 3 months of oral corticosteroid therapy (B).

Figure A4. Case 4: patient with non-Hodgkin lymphoma treated with Obinutuzumab presented with NSIP pattern on chest CT (A). Complete resolution of ground glass opacities after 3 months of oral corticosteroid therapy (B).

Appendix A.5.

Figure A5. Case 5: patient with lung cancer treated with Pembrolizumab presented with an organizing pneumonia (OP) pattern on chest CT (A). Improvement of the pulmonary perihilar lesions after 1 month with corticosteroid therapy and no interruption of the oncologic treatment (B).

Figure A5. Case 5: patient with lung cancer treated with Pembrolizumab presented with an organizing pneumonia (OP) pattern on chest CT (A). Improvement of the pulmonary perihilar lesions after 1 month with corticosteroid therapy and no interruption of the oncologic treatment (B).

Appendix A.6.

Figure A6. Case 6: patient with lung cancer treated with Pembrolizumab presented with an organizing pneumonia (OP) pattern on chest CT (A), with improvement of the lesions on the X-ray after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Figure A6. Case 6: patient with lung cancer treated with Pembrolizumab presented with an organizing pneumonia (OP) pattern on chest CT (A), with improvement of the lesions on the X-ray after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Appendix A.7.

Figure A7. Case 7: patient with lung cancer treated with Pembrolizumab presented with an organizing pneumonia (OP) pattern on chest CT (A) with progression of the lesions after reintroduction of the immunotherapy in parallel with corticotherapy (B).

Figure A7. Case 7: patient with lung cancer treated with Pembrolizumab presented with an organizing pneumonia (OP) pattern on chest CT (A) with progression of the lesions after reintroduction of the immunotherapy in parallel with corticotherapy (B).

Appendix A.8.

Figure A8. Case 8: patient with lung cancer treated with Pembrolizumab presented with bilateral reticular and alveolar opacities on chest X-Ray (A). Improvement of the lesions on the X-ray after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Figure A8. Case 8: patient with lung cancer treated with Pembrolizumab presented with bilateral reticular and alveolar opacities on chest X-Ray (A). Improvement of the lesions on the X-ray after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Appendix A.9.

Figure A9. Case 9: patient with renal cancer treated with Ipilimumab/Nivolumab presented with an HP pattern on chest CT (A). The lesions resolved after one month of corticosteroid therapy (B).

Figure A9. Case 9: patient with renal cancer treated with Ipilimumab/Nivolumab presented with an HP pattern on chest CT (A). The lesions resolved after one month of corticosteroid therapy (B).

Appendix A.10.

Figure A10. Case 10: patient with lung cancer treated with Nivolumab presented with an NSIP pattern on chest CT (A). Improvement of the lesions on CT after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Figure A10. Case 10: patient with lung cancer treated with Nivolumab presented with an NSIP pattern on chest CT (A). Improvement of the lesions on CT after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Appendix A.11.

Figure A11. Case 11: patient with lung cancer treated with Nivolumab presented with bilateral reticular opacities on chest X-Ray (A). Improvement of the lesions on the X-ray after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Figure A11. Case 11: patient with lung cancer treated with Nivolumab presented with bilateral reticular opacities on chest X-Ray (A). Improvement of the lesions on the X-ray after one month of corticosteroid therapy and interruption of the oncologic treatment (B).

Appendix A.12.

Figure A12. Case 12: patient with lung cancer treated with Nivolumab presented with NSIP pattern on chest CT (A). The X-Ray showed no new lesions, only small pleural effusion on the right side after one month of corticosteroid therapy and interruption of the oncological treatment (B).

Figure A12. Case 12: patient with lung cancer treated with Nivolumab presented with NSIP pattern on chest CT (A). The X-Ray showed no new lesions, only small pleural effusion on the right side after one month of corticosteroid therapy and interruption of the oncological treatment (B).