Dermatological adverse events are often reported after treatment with immune checkpoint inhibitors. This brief communication reports on two patients who developed a cutaneous rash with fever during treatment with neoadjuvant chemoradiotherapy and atezolizumab for resectable esophageal adenocarcinoma.
Abstract
Dermatological adverse events have frequently been reported after immune checkpoint inhibition. When an adverse event occurs during combination of immune checkpoint inhibition with chemotherapy, the question arises which agent is responsible. Unnecessary withdrawal of either chemotherapy or immunotherapy could lead to suboptimal treatment outcomes. Here we report on two patients who developed a cutaneous drug reaction with fever during treatment with paclitaxel, carboplatin, radiotherapy, and PD‐L1 inhibition (atezolizumab) for resectable esophageal adenocarcinoma. In the first case atezolizumab was suspected, and in the second paclitaxel. We discuss the clinical manifestation, treatment, and pathophysiology underlying both cases.
Introduction
The CROSS trial demonstrated the superiority of neoadjuvant chemoradiotherapy (nCRT) compared with surgery alone for resectable esophageal and gastroesophageal junctional (GEJ) cancer [1]. However, after nCRT, more than half of all patients experience progression [1]. To further improve survival for patients with resectable esophageal adenocarcinoma (rEAC), we are currently enrolling patients in the phase II PERFECT trial combining atezolizumab (PD‐L1 inhibitor) with nCRT (NCT03087864). Treatment consists of 5 weeks of standard dose CROSS regimen (five cycles: carboplatin, area under the curve = 2 mg/mL per min; paclitaxel, 50 mg/m2 with premedication and 23 fractions of 1.8 Gy of radiotherapy) combined with atezolizumab (1,200 mg) in weeks 1 and 4 (Fig. 1) [1]. Atezolizumab monotherapy is continued after nCRT in weeks 7, 10, and 13. Surgery is scheduled in week 14–16. We present two cases enrolled in the PERFECT trial who developed a cutaneous rash with fever during treatment.
Figure 1.
Treatment scheme in the PERFECT trial.
Abbreviations: A, atezolizumab; C, carboplatin; P, paclitaxel; RT, radiotherapy; S, surgery.
Case One
A 60‐year‐old man with a cT3N1M0 adenocarcinoma located in the lower third of the esophagus developed an erythematous rash, 1 week after the fifth cycle of chemotherapy and 2 weeks after the fourth cycle of chemotherapy with atezolizumab. Topical steroids, triamcinolone, and betamethasone were started with marginal effect. Seven days later, the patient developed a fever of 38.6°C and was hospitalized. Patient was neutropenic (neutrophil count 0.60 × 109 cells per L), and antibiotics were started until cultures were negative. No infectious focus was found. The rash increased in severity with erythematous plaques on the thorax and back (Fig. 2A). Target lesions were identified on the dorsum and the palms of his hands (Fig. 2B). Bullae were identified on his scalp. Prednisone 90 mg once per day was started. A biopsy of the abdominal skin was taken, which showed lichenoid dermatitis with extensive apoptosis of keratinocytes, reminiscent of erythema multiforme (EM; Fig. 3A). Patient was discharged after 2 days while continuing treatment with prednisone. After 2 weeks the skin was recovering with severe desquamation, and prednisone dose could be tapered. Because of high suspicion of atezolizumab being the causative agent of the cutaneous drug reaction, further treatment with this agent was cancelled.
Figure 2.
Erythema multiforme cutaneous reaction in the first case. (A): Erythematous plaques on the thorax. (B): Target lesions on the dorsum of the hands.
Figure 3.
Histopathologic evaluation of the two cases. (A): Extensive epidermal apoptosis consistent with an erythema multiforme‐like reaction in the first case. (B): Dermal infiltration with eosinophils (arrows) consistent with urticarial reaction in the second case.
Case Two
A 48‐year‐old man with a cT3N1M0 adenocarcinoma of the GEJ presented to the emergency department with a fever of 39.5°C, malaise, and nausea. For 2 days he suffered from an erythematous rash with urticaria sparing only his face and the palms of his hands. On the day of presentation, the patient had received his fourth cycle of paclitaxel, carboplatin combined with atezolizumab. On physical examination no focus for infection was found, and laboratory examination showed low infection parameters (C‐reactive protein, 2.1 mg/L; neutrophils, 5.8 × 109 cells per L) and negative blood culture results. The chest X‐ray showed no abnormalities. A biopsy of the abdominal skin, taken 1 day after presentation, revealed a mild superficial perivascular dermatitis with eosinophils and focal epidermal spongiosis with a few apoptotic keratinocytes, suggesting focal interface reaction (Fig. 3B). Patient was hospitalized for observation. After 2 days the fever and skin reaction subsided spontaneously. Patient was discharged in good condition from the hospital. After his fifth cycle of carboplatin and paclitaxel, without atezolizumab, the patient developed again a fever of 38.8°C together with the cutaneous rash. After 8 mg of dexamethasone and 2 mg clemastine taken orally, that same evening and the morning after, the fever and skin reaction subsided. Patient received atezolizumab monotherapy in weeks 7, 10, and 13 without any fever or cutaneous reaction.
Discussion
We report on two patients who developed a cutaneous rash with fever during treatment with nCRT and atezolizumab for rEAC. The first case most likely suffered from an EM‐like reaction. The second case developed a delayed hypersensitivity reaction. First, we will discuss the various cutaneous manifestations associated with cytotoxic agents used during nCRT and checkpoint inhibition. Second, we will discuss the potential relation between the administered treatment and the observed adverse events (AEs). Third, we discuss the clinical management of cutaneous reactions with fever during nCRT combined with checkpoint inhibition.
During conventional nCRT according to the CROSS scheme, the most common nonhematological grade 3 or higher AEs were leukopenia (6%), anorexia (5%), and fatigue (3%) [1]. In the publication of the CROSS trial, hypersensitivity or cutaneous AEs were not mentioned [1]. The cytotoxic agents used in the CROSS trial (paclitaxel and carboplatin) can, however, trigger the aforementioned AEs.
Taxanes can induce several types of cutaneous reactions, including, among others, hypersensitivity reactions, maculopapular eruptions, phototoxicity, drug‐induced lupus erythematosus, radiation recall dermatitis, and, in rare instances, EM, toxic epidermal necrolysis (TEN), or Stevens‐Johnson syndrome (SJS) [2]. The incidence of taxane‐induced cutaneous AEs varies significantly in the literature from 6% up to 81%, which is probably related to differences in premedication, dosing intensity and frequency, combination therapy, and cancer type [2]. We will discuss the most common cutaneous reaction, the hypersensitivity reaction, in more detail below, as this AE is most relevant for our case report.
Taxane‐induced hypersensitivity reactions are often seen after the first or second cycle, and onset is very rapid during or soon after infusion [3], [4]. Delayed hypersensitivity reactions are uncommon but can occur after several hours and even 10 days after infusion [3]. Adequate premedication includes corticosteroids with an antihistamine which can reduce the incidence of infusion reactions to less than 10% [2]. Clinical symptoms are usually skin related (urticaria, maculopapular rash, pruritus, angioedema, flushing) in combination with systemic compromise (hypotension, back pain, dyspnea, bronchospasms) [3]. Delayed reactions usually manifest as a macropapular rash, urticaria, or angioedema [3]. For immediate reactions clinical recovery is rapid after termination of infusion [3]. Usually it is possible to restart the infusion on the same day. If patients develop the same reaction after rechallenge, substitution of paclitaxel with docetaxel or vice versa (watch out for cross‐reactivity), switching to nano‐albumin bound (nab)‐paclitaxel, or a drug desensitization protocol may be considered [3], [4]. Regarding severe reactions, including anaphylaxis, subsequent treatment with taxanes should be carefully discussed in a multidisciplinary oncology meeting. Anaphylatoxins, direct mast cell activation, and IgE‐mediated type I hypersensitivity have been implicated as potential pathophysiological mechanisms [5], [6], [7]. The solvents Cremophor EL (polyethoxylated castor oil, BASF, Germany) for paclitaxel and polysorbate 80 for docetaxel have been linked to complement and histamine activation [6], [8]. However, paclitaxel without the solvent was also capable of inducing histamine release in one patient [9]. It thus seems that both the solvent and the agent itself might be responsible for inducing a hypersensitivity reaction.
Carboplatin can also elicit a hypersensitivity reaction. Most commonly, this occurs after six to eight cycles [4], [10]. Incidence is reported to be 1% in patients who receive fewer than five cycles of carboplatin [11]. Symptoms mimic an anaphylactic reaction, including flushing, urticaria, swelling, dyspnea, and hypotension [11]. After interruption of infusion, symptoms might need several hours to resolve. Management of a hypersensitivity reaction may include the administration of antihistamines and corticosteroids [10]. For severe anaphylaxis it may be necessary to administer epinephrine, bronchodilators, and oxygen [10]. Skin testing with intradermal injection of carboplatin after six cycles can help identify patients at risk for hypersensitivity reactions [12]. In patients who experienced a severe AE to carboplatin, desensitization protocols, switching from carboplatin to cisplatin, and cessation of treatment may be considered [4], [10], [13]. Most likely, carboplatin‐related reactions are IgE related (type I hypersensitivity) [10].
It may also prove challenging to identify the agent responsible for a hypersensitivity reaction in combination treatment with carboplatin and paclitaxel [4]. However, there are clues that may help distinguish between compounds: severity (carboplatin usually more severe than paclitaxel), timing of first symptoms (paclitaxel, minutes after infusion; carboplatin varies between minutes to days), and disappearance of symptoms after interruption of infusion (paclitaxel, rapid clinical improvement; carboplatin, improvement within hours).
Immunotherapy, specifically checkpoint inhibition, is particularly well known for the occurrence of cutaneous adverse events [14]. Incidence varies between 14% and 50% and differs between compounds (more common in CTLA‐4 inhibitors compared with PD‐1/PD‐L1 inhibitors) [14], [15]. Severe cases (grade 3–4) are relatively uncommon (below 10%) [14]. Dermatological AEs usually occur early in the first few weeks of treatment [15]. The most common cutaneous reaction is a maculopapular rash, which starts on the trunk and spreads to the extremities [15]. Usually, it is accompanied by pruritis. Histopathological analysis may show features of spongiotic dermatitis with a superficial T‐cell infiltrate and some eosinophils [15]. Pathophysiology is yet unknown, but there is translational evidence that T cells, antibodies, and cytokines may be involved in inducing autoimmune AEs [16]. Treatment may consist of topical or systemic corticosteroids (depending on severity) and antihistamine agents [15]. Resolution of symptoms is usually slow and may take several weeks to months. Interruption of treatment is usually not necessary. In more severe cases, retrospective data suggest that switching between checkpoint inhibitors (CTLA‐4 to anti‐PD‐1) may also provide an opportunity to continue treatment with another agent without the recurrence of the observed AE [16]. Other cutaneous reactions during checkpoint inhibition may include bullous pemphigoid, psoriasis, lichenoid dermatitis, Grover's disease, TEN, or SJS [15].
Combining chemotherapy with a checkpoint inhibitor may also increase the number of serious AEs. In the IMPOWER‐131 trial for advanced‐stage lung cancer, serious treatment‐related AEs were seen in 22.3% in patients who received carboplatin, paclitaxel, and atezolizumab; in 20.4% who received carboplatin, nab‐paclitaxel with atezolizumab; and in 10.5% of patients who received carboplatin with nab‐paclitaxel (data from American Society of Clinical Oncology abstract) [17]. Rash was one of the most common AEs (number of events for each group not available in abstract). Two trials investigating the addition of pembrolizumab to chemotherapy, with either cisplatin/carboplatin and pemetrexed (KEYNOTE‐189) or carboplatin and paclitaxel/nab‐paclitaxel (KEYNOTE‐407) versus chemotherapy alone for advanced‐stage lung cancer, found that 2.2%–2.5% of patients randomized to pembrolizumab with chemotherapy experienced a grade 3–4 severe rash or infusion reaction [18], [19]. In the chemotherapy alone group this was 2% (KEYNOTE‐189) and 0.7% (KEYNOTE‐407) [18], [19]. In the PACIFIC trial, which administered durvalumab after chemoradiotherapy for stage III lung cancer, only one patient experienced a grade 3–4 rash AE [20]. It thus seems that simultaneous administration of chemotherapy with a checkpoint inhibitor might increase the number of grade 3–4 infusion and rash AEs compared with chemotherapy alone.
In our first case, in which the patient suffered from an EM‐like cutaneous reaction, the most likely causative agent was atezolizumab. The reaction started after finishing nCRT combined with two cycles of atezolizumab. EM‐like reactions, mainly maculopapular eruptions, have been reported after checkpoint inhibitor monotherapy [15]. After taxane or carboplatin treatment, EM‐like reactions have rarely been reported, and for paclitaxel, only on sun‐exposed body parts [21]. Our patient had lesions on body parts that did not receive sun exposure. The most severely affected areas were on the thorax. The patient received radiation therapy for rEAC around the chest. However, because of the extensive nature of the cutaneous reaction, including on nonirradiated locations such as the hands, feet, and scalp, radiation cannot solely be responsible. In fact, target lesions on the hands are pathognomonic for EM [22]. The main hypothesis regarding the pathophysiology of drug‐induced EM is the generation of autoreactive T cells [22]. The normal function of the PD‐1/PD‐L1 axis is to inhibit the proliferation of T cells and their effector function [23]. The blockage of PD‐L1 by atezolizumab might have induced a proinflammatory state with autoreactive T cells targeting the keratinocytes in the epidermis causing an EM‐like cutaneous reaction (Fig. 3A). Nevertheless, we cannot rule out the combination of modalities: chemotherapy, radiotherapy, and atezolizumab contributed to the inflammatory response. An alternative hypothesis might involve the generation of drug‐specific T cells in response to paclitaxel or carboplatin administration. The disinhibition of the immune system through blockage of PD‐L1 might have increased the susceptibility of developing a cutaneous reaction in response to nCRT.
In our second case, the delayed occurrence of urticaria and fever might be related to paclitaxel administration and could potentially have been aggravated by atezolizumab. The reaction occurred several hours after combined infusion of chemotherapy and atezolizumab. Moreover, it recurred after administration of chemotherapy alone and not after atezolizumab monotherapy. Delayed reactions to paclitaxel or carboplatin have been reported [10]. However, these reactions are relatively uncommon, especially after administration of nCRT according to the CROSS scheme without atezolizumab [1]. It seems unlikely that carboplatin is solely responsible, as hypersensitivity reactions are usually more severe and occur after multiple infusions of carboplatin [11]. Paclitaxel‐related hypersensitivity is usually accompanied by urticaria, as observed in our patient [2]. Mast cells, which express PD‐L1, can induce urticaria [24], [25]. Hypothetically, the inhibition of PD‐L1 might have aggravated the reaction to paclitaxel or its solvent by loss of T‐cell inhibition or increased release of histamine.
These two cases demonstrate that combining nCRT with atezolizumab might lead to more severe but manageable cutaneous reactions. In line with a recent review, we advocate performing a thorough evaluation in cases with an atypical erythematous rash, including a biopsy [15]. Clinical and histopathological analysis can help identify the most likely causative agent. Target lesions on the skin are pathognomonic for EM‐like reactions [22]. Urticaria should be reminiscent of a hypersensitivity reaction [2]. Extensive apoptosis of keratinocytes on histopathological examination is suggestive of an EM‐like reaction; dermal infiltration with eosinophils of a hypersensitivity reaction [2], [22]. Severe grade 3–4 EM‐like reactions with fever should be treated with systemic corticosteroids, and an infectious focus should be excluded. Treatment with atezolizumab should be interrupted and carefully reconsidered after resolution of symptoms and tapering of corticosteroids. Recovery of the skin is expected to take several weeks. Delayed hypersensitivity reactions with urticaria and fever are most likely related to paclitaxel and could be triggered or aggravated by concurrent treatment with atezolizumab. An infectious focus should be excluded, and in severe cases corticosteroids and antihistamines should be prescribed. Resolution of symptoms is within days. For subsequent infusions of chemotherapy, an extra dosage of premedication (dexamethasone and antihistamines) should be administered before and after infusion. Based only on the experience from our case, checkpoint inhibitor monotherapy might be continued, although close follow‐up for any signs of cutaneous toxicity relapse is recommended as well as giving indications to patients for alarm signs that may allow for early detection of cutaneous and noncutaneous immune‐related AEs.
Conclusion
Our two cases demonstrate that combining checkpoint inhibitors with chemoradiotherapy may lead to additional drug‐induced cutaneous AEs. Clinicians should avoid early judgement about the relation between a certain therapeutic agent and toxicity. This could lead to unnecessary interruption of treatment, potentially reducing treatment efficacy. Clinical manifestation and histopathological findings can aid in finding the agent most likely responsible for the observed AE.
Acknowledgments
The PERFECT study (NCT03087864) is financially supported by Roche.
Disclosures
Stef P. Menting: Abbvie (H), Novartis (C/A); Hanneke W.M. van Laarhoven: Bristol‐Myers Squibb, Eli Lilly & Co., Nordic Pharma (C/A), Bayer, Bristol‐Myers Squibb, Celegene, Janssen, Eli Lilly & Co., Nordic Pharma, Philips, Roche (RF). The other authors indicated no financial relationships.
(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board
References
- 1.van Hagen P, Hulshof MC, van Lanschot JJ et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 2012;366:2074–2084. [DOI] [PubMed] [Google Scholar]
- 2.Sibaud V, Leboeuf NR, Roche H et al. Dermatological adverse events with taxane chemotherapy. Eur J Dermatol 2016;26:427–443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Picard M. Management of hypersensitivity reactions to taxanes. Immunol Allergy Clin North Am 2017;37:679–693. [DOI] [PubMed] [Google Scholar]
- 4.Lee C, Gianos M, Klaustermeyer WB. Diagnosis and management of hypersensitivity reactions related to common cancer chemotherapy agents. Ann Allergy Asthma Immunol 2009;102:179–187; quiz 187–179, 222. [DOI] [PubMed] [Google Scholar]
- 5.Weiss RB, Donehower RC, Wiernik PH et al. Hypersensitivity reactions from taxol. J Clin Oncol 1990;8:1263–1268. [DOI] [PubMed] [Google Scholar]
- 6.Szebeni J, Muggia FM, Alving CR. Complement activation by cremophor EL as a possible contributor to hypersensitivity to paclitaxel: An in vitro study. J Natl Cancer Inst 1998;90:300–306. [DOI] [PubMed] [Google Scholar]
- 7.Prieto Garcia A, Pineda de la Losa F. Immunoglobulin E‐mediated severe anaphylaxis to paclitaxel. J Investig Allergol Clin Immunol 2010;20:170–171. [PubMed] [Google Scholar]
- 8.Eschalier A, Lavarenne J, Burtin C et al. Study of histamine release induced by acute administration of antitumor agents in dogs. Cancer Chemother Pharmacol 1988;21:246–250. [DOI] [PubMed] [Google Scholar]
- 9.Essayan DM, Kagey‐Sobotka A, Colarusso PJ et al. Successful parenteral desensitization to paclitaxel. J Allergy Clin Immunol 1996;97:42–46. [DOI] [PubMed] [Google Scholar]
- 10.Boulanger J, Boursiquot JN, Cournoyer G et al. Management of hypersensitivity to platinum‐ and taxane‐based chemotherapy: CEPO review and clinical recommendations. Curr Oncol 2014;21:e630–e641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Markman M, Kennedy A, Webster K et al. Clinical features of hypersensitivity reactions to carboplatin. J Clin Oncol 1999;17:1141. [DOI] [PubMed] [Google Scholar]
- 12.Zanotti KM, Rybicki LA, Kennedy AW et al. Carboplatin skin testing: A skin‐testing protocol for predicting hypersensitivity to carboplatin chemotherapy. J Clin Oncol 2001;19:3126–3129. [DOI] [PubMed] [Google Scholar]
- 13.Lee CW, Matulonis UA, Castells MC. Carboplatin hypersensitivity: A 6‐h 12‐step protocol effective in 35 desensitizations in patients with gynecological malignancies and mast cell/IGE‐mediated reactions. Gynecol Oncol 2004;95:370–376. [DOI] [PubMed] [Google Scholar]
- 14.Friedman CF, Proverbs‐Singh TA, Postow MA. Treatment of the immune‐related adverse effects of immune checkpoint inhibitors: A review. JAMA Oncol 2016;2:1346–1353. [DOI] [PubMed] [Google Scholar]
- 15.Sibaud V. Dermatologic reactions to immune checkpoint inhibitors: Skin toxicities and immunotherapy. Am J Clin Dermatol 2018;19:345–361. [DOI] [PubMed] [Google Scholar]
- 16.Postow MA, Sidlow R, Hellmann MD. Immune‐related adverse events associated with immune checkpoint blockade. N Engl J Med 2018;378:158–168. [DOI] [PubMed] [Google Scholar]
- 17.Jotte RM, Cappuzzo F, Vynnychenko I et al. IMPOWER131: Primary PFS and safety analysis of a randomized phase III study of atezolizumab + carboplatin + paclitaxel or nab‐paclitaxel vs carboplatin + nab‐paclitaxel as 1L therapy in advanced squamous NSCLC. J Clin Oncol 2018;36(suppl 18):LBA9000. [Google Scholar]
- 18.Gandhi L, Rodriguez‐Abreu D, Gadgeel S et al. Pembrolizumab plus chemotherapy in metastatic non‐small‐cell lung cancer. N Engl J Med 2018;378:2078–2092. [DOI] [PubMed] [Google Scholar]
- 19.Paz‐Ares L, Luft A, Vicente D et al. Pembrolizumab plus chemotherapy for squamous non‐small‐cell lung cancer. N Engl J Med 2018;379:2040–2051. [DOI] [PubMed] [Google Scholar]
- 20.Antonia SJ, Villegas A, Daniel D et al. Durvalumab after chemoradiotherapy in stage III non‐small‐cell lung cancer. N Engl J Med 2017;377:1919–1929. [DOI] [PubMed] [Google Scholar]
- 21.Cohen PR. Photodistributed erythema multiforme: Paclitaxel‐related, photosensitive conditions in patients with cancer. J Drugs Dermatol 2009;8:61–64. [PubMed] [Google Scholar]
- 22.Samim F, Auluck A, Zed C et al. Erythema multiforme: A review of epidemiology, pathogenesis, clinical features, and treatment. Dent Clin North Am 2013;57:583–596. [DOI] [PubMed] [Google Scholar]
- 23.Sun C, Mezzadra R, Schumacher TN. Regulation and function of the PD‐L1 checkpoint. Immunity 2018;48:434–452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Church MK, Kolkhir P, Metz M et al. The role and relevance of mast cells in urticaria. Immunol Rev 2018;282:232–247. [DOI] [PubMed] [Google Scholar]
- 25.Nakae S, Suto H, Iikura M,et al. Mast cells enhance T cell activation: Importance of mast cell costimulatory molecules and secreted TNF. J Immunol 2006;176:2238–2248. [DOI] [PubMed] [Google Scholar]



