Rapid Desensitization to Antitumoral Agents. Result from a Retrospective Study, DESARCh

Abstract

Antitumoral drugs (ADs) can induce drug hypersensitivity reactions (DHRs). Rapid drug desensitization (RDD) protocols represent an important option to mitigate recurrent DHRs thus allowing the safe administration of ADs at therapeutic doses. The aim of this retrospective study was to assess the effectiveness of the RDD protocols performed at our institution. The “DESARCh” study was a retrospective, observational study that included consecutive patients who underwent RDD protocols from January 2011 to December 2022 at IRCCS Ospedale Sacro Cuore Don Calabria in Negrar di Valpolicella, Verona, Italy. The RDD protocol consisted of a 5-step protocol with 5 different concentrations of the drugs at 1:1, 1:10, 1:100, 1:1,000 and 1:10,000 dilution given intravenously over a 1-hour infusion each, with concentrations increasing from the most diluted to the most concentrated form, preceded by a 30-min premedication regimen. A total of 66 RDD protocols were administered to 25 female patients with ovarian (64%; n = 16/25), breast (12%; n = 3/25), endometrium (8%; n = 2/25), cervix (8%; n = 2/25), uterine (4%; n = 1/25) and fallopian tubes (4%; n = 1/25) cancers. A known history of atopy/allergy was reported by 36% (n = 9/25) of patients. Patients received RDD protocols because of DHRs to carboplatin (n = 23/66, 34.85%), paclitaxel (n = 18/66, 27.27%), pegylated liposomal doxorubicin (n = 3/66, 4.55%), and trastuzumab (n = 22/66, 33.33%). DHRs were mild-moderate, severe and life-threatening in 60.72%, 28.57% and 10.71% of cases, respectively. The success rate of RDD protocols, defined as the rate of complete administration of full target dose with no breakthrough reactions, was 81.82% (n = 54/66). Success rate was lower for carboplatin compared to other drugs (65.22% vs 90.7%; P = .017678). The RDD protocol used in our institution was found to be safe, with a meaningful success rate. However, further research is needed to better understand the underlying mechanisms of DHRs and to enhance effectiveness, particularly for patients experiencing DHRs to platinum compounds. This study was approved by the ethics committee of Verona and Rovigo (Italy) with approval number 15476 on 10/03/2023 and it was registered with the Register of Observational Studies of the Italian Medicines Agency (AIFA) (available since 31 January 2023), with ID n. 109, on 28/02/2023 (https://www.aifa.gov.it/en/registro-studi-osservazionali).

Introduction

Cancer is a leading cause of death worldwide, with an estimated 20 million new cases of cancer and 9.7 million deaths estimated in 2022. ¹ Antitumor drugs (ADs) represent a milestone in the treatment of cancer, both in early and advanced stages. Unfortunately, ADs can trigger drug hypersensitivity reactions (DHRs), which are unforeseen reaction not attributable to the known toxicity of the drug. Most ADs have been associated with DHRs, with platinum compounds and taxanes being the most frequently implicated, followed by monoclonal antibodies (moAbs). ² DHRs can significantly limit the administration of potentially effective medications to cancer patients. Therefore, managing DHRs is a crucial aspect of oncology.

DHRs can occur through IgE-mediated activation of basophils/mast cells or through non-IgE-mediated mechanisms, including direct activation of basophils/mast cells, cytokine release, IgG-mediated mechanisms leading to complement activation, and mixed forms. ³ Acute DHRs typically manifest within 1 hour, presenting a broad spectrum of symptoms ranging from mild to severe, including rash, angio-oedema, dyspnoea, rhinitis, chest pain, hypo- or hyper-tension, and anaphylactic reactions. The timing, characteristics, and overall incidence vary depending on the specific agent involved. ⁴ For platinum compounds, the overall incidence of DHRs varies between 1% and 44%, depending on the specific agent administered, the number of previous courses and the underlying cancer type. ⁵ DHRs to taxanes manifest in 5% to 10% of patients treated with docetaxel or paclitaxel, respectively, when administered with standard premedication. ³ Regarding moAbs, DHRs typically occur at the first or second infusion, and the incidence varies depending on the type of moAb, with low incidences observed in humanized or fully human moABs, whereas higher incidences are reported for chimeric moABs. ⁶ The majority of HSRs to ADs are typically observed in gynecological cancers (ovarian, uterine, cervical), with a global incidence in females of 29.1%, ⁷ likely attributed to the high cumulative doses of platinum derivatives and taxanes in the specific chemotherapy regimens used to treat these conditions. ⁸

Several rapid drug desensitization (RDD) protocols have been developed for ADs implicated in DHRs, both IgE- and non-IgE mediated, with the aim of inducing temporary tolerance thus allowing ADs administration at therapeutic doses in a relatively short period.^9-11

In general, the following aspects should always be considered when applying RDD protocols: first, due to the intrinsic risk of recurrent DHRs, RDD protocols should be considered only in the absence of equally effective therapies and after careful evaluation of the risk-benefit balance; furthermore, RDD protocols should be administered by experienced healthcare professionals, possibly in an intensive care unit (ICU) to mitigate the risk of severe complications, at least for the first RDD protocol administration. ¹²

Beyond these widely accepted considerations, the various RDD protocols developed to date differ significantly in the number of steps and drug concentrations used. Currently, there are no comparative studies among these different RDD protocols, and substantial knowledge gaps impede the development of personalized desensitization protocols tailored to the specific DHR mechanisms of each drug in individual patients. In this context, we believe it is crucial to report on the efficacy and safety of different RDD protocols to advance understanding in this field. Here, we present the results of a retrospective study aimed at analyzing the safety and efficacy of a 5-step RDD protocol developed at our institution.

Materials and Methods

Study Design and Population

The “DESARCh” study included consecutive patients with any solid tumor, who underwent at least one RDD protocol administration at our institution (IRCCS Ospedale Sacro Cuore Don Calabria, Negrar di Valpolicella, Verona, Italy) from January 2011 to December 2022, due to a DHRs to ADs. For patients who experienced a DHR while receiving combination therapy, the drug responsible for the reaction was identified using temporal association criteria based on the onset of the reaction in relation to the administration of each drug. Patients who received RDD protocol as a primary desensitization due to previous DHRs to drugs other than the one targeted for the desensitization were excluded. No specific diagnostic tool (ie, skin testing, specific or total IgE assessment, tryptase determination, or basophil activation testing) was required for diagnosis of DHRs. Patients could receive one or more RDD protocols, based on clinical decision.

Data were retrospectively collected by electronic medical records. The primary objective of the study was to assess the outcome of the RDD protocol in terms of success rate The outcome of the RDD protocol was classified as positive when the full target dose was administered without breakthrough reactions that necessitated the interruption of AD administration. Conversely, the outcome was considered negative if the patient experienced any DHRs that led to the interruption of AD administration. The success rate was defined as the percentage of RDD protocols with a positive outcome.

This study conformed to the ethical guidelines of the Declaration of Helsinki, it was approved by the ethics committee of Verona and Rovigo (Italy) with approval number 15476 on 10/03/2023 and it was registered with the Register of Observational Studies of the Italian Medicines Agency (AIFA) (available since 31 January 2023), with ID n. 109, on 28/02/2023 (https://www.aifa.gov.it/en/registro-studi-osservazionali). ¹³

Desensitization Protocol

The RDD protocol used (Figure 1) consisted of five steps, each involving a different dilution of the drug. The protocol was as follows: 1:1 (250 ml bag), 1:10 (100 ml bag), 1:100 (100 ml bag), 1:1,000 (100 ml bag) and 1:10 000 (100 ml bag), as described by Gastaminza et al. ¹⁴ However, each preparation was administered intravenously over a period of 1 hour, with concentrations increasing from the most diluted to the most concentrated form. A premedication regimen consisting of 20 mg dexamethasone and 8 mg ondansetron was also scheduled before starting ADs, with the addition of 10 mg of chlorphenamine in case of taxanes or moAbs. RDD protocol was administered in the ICU and lasted about 6 hours. Patients were then closely monitored for 1 hour.

Figure 1.

Example of Rapid Drag Desensitization (RDD) Protocol used (inspired by Gastaminza et al. ¹⁴ ), adapted for a total dose of 300 mg.

Mild adverse drug reactions (ADRs) were managed without treatment interruption, whereas moderate-severe ADRs led to treatment interruption. ADR occurring during the RDD protocol administrations were reported through the National Pharmacovigilance Network (RNF; https://servizionline.aifa.gov.it/schedasegnalazioni/#/). ¹⁵

Statistical Analysis

Descriptive statistics were used for analysis. Results were described and evaluated in terms of range, median (interquartile range [IQR]) and percentage.

The American Joint Committee on Cancer (AJCC) staging system 7th and 8th editions (depending on year of tumor diagnosis in different patients) was used to assess the clinical or pathological stage.^16,17 Adverse event severity was assessed using National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE; version 5.0).^18,19

Cancer site data are presented as frequency (number) and percentage of patients, as well as protocols out of the total included. Conversely, ADs included in the RDD protocols, grade of DHRs and outcome, were expressed as number and percentage of protocols out of the total. Fisher exact test, performed using MedCalc^® (version v14.8.1), was used to compare categorical variables. The level of significance was set at P < .05.

Results

During the 10-year study period, 29 patients receiving 77 RDD protocols were screened. Two patients were excluded from the study because received RDD protocol as primary prevention after DHRs with a different drug, and two patients were considered not evaluable for RDD outcome, one for a rapid clinical deterioration due to disease progression after only one RDD protocol, and the other one because lost to follow-up after two RDD protocols. Therefore, 25 patients treated with 66 RDD protocols were included in the final analysis (Figure 2).

Figure 2.

PRISMA flowchart of the process of patients/protocols selection.

*One patient lost to follow-up after 2 RDD; 1 patient lost to clinical deterioration due to disease progression.

Overall, 25 patients underwent single drug RDDs (accounting for 58 protocols) and 3 patients underwent RDD to more than 1 drug (accounting for 8 protocols).

Median age was 61 years (IQR = 58.75-66 years). Cancer primary site was: ovary, n = 16/25 patients (64%) and 28/66 protocols (42.42%); breast, n = 3/25 patients (12%) and 26/66 protocols (39.39%); endometrium, n = 2/25 patients (8%) and 3/66 protocols (4.55%); cervix, n = 2/25 patients (8%) and 4/66 protocols (6.06%); uterus, n = 1/25 patients (4%) and 4/66 protocols (6.06%); and fallopian tubes, n = 1/25 patients (4%) and 1/66 protocols (1.52%) (Table 1 and Figure 3A and B).

Table 1.

Patients and Protocols Characteristics.

Patient ID	RDD protocol ID §	Sex	Age (year)	Allergy history	Diagnosis (AJCC)	Line of therapy	Chemotherapy regimen	No. of chemotherapy cycle until initial DHRs	DHRs symptoms	Grade of DHRs (based on CTCAE)	Drug in RDD	No. of RDD protocols	RDD outcome
1	1	Female	46	—	Infiltrating ductal breast cancer (IIa)	1	T	4	Rash, itching	1	T	2	Positive × 2
2	2	Female	61	Iodinated contrast media and doxazosin	Serous ovarian cancer (IIIc)	1	CT	1	Allergic reaction	1	C	1	Positive × 1
2	3	Female	61	Iodinated contrast media and doxazosin	Serous ovarian cancer (IIIc)	1	CT	1	Allergic reaction	1	T	1	Positive × 1
3	4	Female	49	Iodinated contrast media	Poorly differentiated serous ovarian cancer (IIIc)	1	CT	2	Rash, itching	1	C	1	Negative × 1
4	5	Female	70	—	Ovarian cancer (IIIc)	3	CT	9	Rash, itching	1	C	2	Positive × 2
5	6	Female	65	Amoxicillin/clavulanic acid	Serous ovarian cancer (IIIc)	1	CDx	1	Facial flushing, rash on limbs, itching	2	C	1	Negative × 1
6	7	Female	64	—	Endometrioid adenocarcinoma (IV)	1	CT	2	Rash, itching	1	T	2	Positive × 2
7	8	Female	74	Ibuprofen, metoclopramide, mites	Serous ovarian cancer (IIIc)	1	CDx	1	Facial flushing, discomfort, back pain, hypotension	3	Dx	1	Negative × 1
8	9	Female	62	—	Poorly differentiated serous ovarian cancer (IIIb)	1	CT	6	Discomfort, epigastric pain, palmar rash	3	C	2	Positive × 2
9	10	Female	68	—	Serous ovarian cancer (IIIc)	1	CT	4	Whole body rash, facial flushing, palmar itching, nausea	3	C	2	Positive × 2
10	11	Female	54	—	Locally advanced squamous cell cervical cancer (IIIb)	1	CT	1	Hot flashes, abdominal-ache	3	T	2	Positive × 2
11	12	Female	66	—	Poorly differentiated serous fallopian tube cancer (IIIc)	1	CT	7	Facial flushing, rash on limbs, itching	2	C	1	Negative × 1
12	13	Female	61	Iodinated contrast media, carbocaine	Uterine carcinosarcoma (Ib)	1	CT	1	Anaphylactic shock	4	C	2	Positive × 2
12	14	Female	61	Iodinated contrast media, carbocaine	Uterine carcinosarcoma (Ib)	1	CT	1	Anaphylactic shock	4	T	2	Positive × 2
13	15	Female	66	—	Poorly differentiated cervical adenocarcinoma (III)	1	CT	1	Whole body rash, discomfort, back pain, dyspnea	3	T	2	Positive × 2
14	16	Female	61	—	Ovarian adenocarcinoma (IIIc)	1	CT	2	Whole body rash, discomfort, epigastric pain, nausea	2	C	1	Positive × 1
14	17	Female	61	—	Ovarian adenocarcinoma (IIIc)	3	Dx	1	Whole body rash, itching	1	Dx	1	Negative × 2
15	18	Female	76	—	Advanced breast cancer (IV)	1	T	1	Whole body rash, bronchospasm, dyspnea	3	T	2	Positive × 2
16	19	Female	50	Pollen	Poorly differentiated ovarian cancer (IIb)	1	CT	1	Facial flushing, whole body rash, chest pain	2	T	2	Positive × 2
17	20	Female	58	—	Poorly differentiated ovarian cancer (IV)	1	CT	8	Rash, itching	1	C	2	Positive × 1, Negative × 1
18	21	Female	67	—	Moderately differentiated, papillary serous ovarian cancer (IIc)	3	CT	2	Whole body rash, itching, discomfort, epigastric pain	2	C	2	Positive × 1, Negative × 1
19	22	Female	59	—	Advanced ovarian cancer (IIIc)	1	CT	2	Abdominal-ache, emesis and fainting	3	C	1	Negative × 1
20	23	Female	61	Nimesulide	High-grade ovarian adenocarcinoma (IIIc)	1	CT	1	Rash, itching	1	T	2	Positive × 2
21	24	Female	65	Tramadol and paclitaxel	Infiltrating ductal breast cancer (IIIa)	1	PrTr	15	Anaphylactic shock	4	Tr	22	Positive × 22
22	25	Female	70	—	Poorly differentiated serous ovarian cancer (IIIc)	3	CT	9	Rash, itching	1	C	2	Positive × 1, Negative × 1
23	26	Female	35	Pregabalin and metoclopramide	Endometrial cancer	3	C	1	Rash, itching	1	T	1	Negative × 1
24	27	Female	41	—	Ovarian high-grade (IIIc) endometrioid cancer (III)	2	CG	2	Rash, itching	1	C	3	Positive × 2, Negative × 1
25	28	Female	59	—	Serous ovarian cancer (IIIc)	1	CDx	1	Rash, itching	3	Dx	1	Negative × 1

No. = number; C = carboplatin; T = paclitaxel; Dx = pegylated liposomal doxorubicin; CT = carboplatin + paclitaxel; CG = carboplatin + gemcitabine; PrTr = pertuzumab + trastuzumab; CDx = carboplatin + pegylated liposomal doxorubicin; AJCC = American joint committee on cancer staging system 7th edition 2010 and 8th edition 2017; DHRs = drug hypersensitivity reactions; CTCAE = common terminology criteria for adverse events; RDD = rapid drug desensitization.

^§RDD Protocol ID: The same patient may have received RDD protocols for different cancer drugs, so we ordered all protocols with a sequential number.

Figure 3.

Patients stratified for primary site of cancer (panel A) and protocols stratified for primary site of cancer (panel B), drug in desensitization (panel C) and grade of hypersensitivity reaction (panel D). Frequency data are expressed as number and percentage of total patient ID (n = 25; panel A) total RDD ID protocols (n = 28; panel D), total RDD protocols (n = 66; panels B and C).

A known history of atopy/allergy was reported by 36% (n = 9/25) of patients (54.55%; n = 36/66 of RDD protocols) and was mainly related to iodinated contrast media. No patient received a specific test for hypersensitivity to ADs.

The ADs implicated in DHRs were: carboplatin, n = 23/66 (34.85%); paclitaxel, n = 18/66 (27.27%); pegylated liposomal doxorubicin, n = 3/66 (4.55%); and trastuzumab, n = 22/66 (33.33%) (Table 1 and Figure 3C). Reported DHRs included body rash, facial flushing, itching, discomfort, anaphylactic shock, and were classified according to CTCAE grade as follows: grade 1 42.86% (n = 12/28), grade 2 17.86% (n = 5/28), grade 3 28.57% (n = 8/28) and grade 4 10.71% (n = 3/28); (Table 1 and Figure 3D). Moreover, 56% (n = 14/25) of patients presented a DHR during the first course of cancer drug.

Median RDD protocol courses were 2 (IQR = 1-2). The success rate of RDD protocols was 81.82% (n = 54/66) overall. In particular, the success rate of RDD protocols with carboplatin, paclitaxel, pegylated liposomal doxorubicin and trastuzumab was 65.22% (n = 15/23), 94.44% (n = 17/18), 0% (n = 0/3) and 100% (n = 22/22), respectively (Table 1 and Figure 4). Therefore, success rate of RDD protocol was 65.22% (n = 15/23) for carboplatin and 90.7% (n = 39/43) for non-carboplatin treatment (p = 0.017678; Figure 4). No life-threatening hypersensitivity reactions or deaths occurred during the procedure.

Figure 4.

Protocols outcomes expressed as overall percentage of success and as percentages of success for the different RDD protocols analysed: carboplatin, paclitaxel, pegylated liposomal doxorubicin and trastuzumab. A statistical analysis of the comparison between the success rate among carboplatin versus non-carboplatin treatments was performed using the Fisher exact test. ***P = .017678 versus carboplatin treatment success rate.

Discussion

RDD may mitigate the risk of recurring DHRs, thereby enabling the administration of ADs at therapeutic doses.

Numerous RDD protocols have been developed for ADs, each involving a different number of steps with increasing drug concentrations. The effectiveness of these RDD protocols varies across different studies, ranging from 56% to 100%. Such variability is likely due to differences among studies in primary cancer types, AD types, RDD methodologies, and the definition of efficacy endpoints. Indeed, while some studies consider the absence of any-grade breakthrough DHRs as a positive outcome of RDD protocols, others include only the absence of severe recurring DHRs or the completion rate of RDD at the full-target dose, independently of breakthrough DHRs. Based on available evidence, around 90% of administrations by RDD protocols occur with no DHRs or grade 1 DHRs, and the majority of the patients are generally able to complete the protocol receiving the therapeutic dose. ²⁰

In the present study, we used 5-step RDD protocols for 66 desensitizations, with an overall success rate of 81.82% (n = 54/66), defined as the absence of any-grade recurring DHRs. Notably, our study revealed a statistically significant lower success rate (P = .017678) among patients treated with carboplatin (65.22%) compared to those treated with other oncology drugs (90.7%).

Consistent with our findings, other studies have reported a higher incidence of recurring DHRs with platinum salts. In a prospective observational study, among 1027 RDD protocols, breakthrough reactions occurred in 28 out of 73 patients reactive to platinum salts (52%), compared to 20 out of 73 patients reactive to taxanes (23%). ²¹ Similarly, in other series, the majority of breakthrough reactions were observed with platinum salts. ²² The reason for the higher incidence of breakthrough reactions with platinum salts compared to other ADs remains unknown, but it may be attributed to distinct mechanisms underlying DHRs to platinum salts versus those to other ADs, such as taxanes or moABs. Indeed, the clinical presentation of DHRs to platinum salts differs from that of other ADs, often occurring after repeated exposure to the drug, whereas for other ADs, it frequently manifests after the first or second infusion. However, we did not conduct any specific diagnostic tests to elucidate the mechanisms of DHRs to carboplatin compared to other drugs or to identify patients at higher risk for breakthrough reactions. In some studies, candidate patients include those who present with type I HSRs IgE dependent and non-IgE mediated HSRs during the chemotherapy infusion or shortly after. ²³ Our study also suggests a particularly poor efficacy of our RDD protocol for liposomal doxorubicin, with a success rate of 0%, although the reason for this is unclear. However, the limited number of protocols performed for this drug (n = 3) prevents definitive conclusions but suggests the hypothesis that a different desensitization strategy may be necessary for this drug.

Some evidence suggests that RDD protocols do not adversely affect the overall survival (OS) of cancer patients. For instance, a retrospective study comparing 67 patients experiencing oxaliplatin-related DHRs who received the drug via RDD and 143 non-allergic patients receiving oxaliplatin through standard infusion protocols demonstrated similar median OS between the 2 groups (23.7 vs 34.5 months, HR 1.42; 95% CI: 0.93-2.17; P = .104). ¹⁰ Likewise, a comparison of life expectancy between 155 carboplatin-desensitized patients with recurrent ovarian cancers and 81 controls using propensity scores did not reveal significant differences in 5-year survival. ²⁴ Notably, a retrospective study involving 91 ovarian cancer patients found that those who underwent carboplatin desensitization had a longer OS compared to patients without carboplatin hypersensitivity who did not undergo desensitization. ²⁵ However, since we did not collect survival data, we cannot draw conclusions regarding how the RDD protocol may have impacted OS in our study.

Another important and controversial issue is whether it is possible to safely resume the standard schedule after a positive outcome with RDD protocols, and if so, how many successful RDD protocols are required before it is safe to resume the standard treatment regimen. However, our study was not designed to address this question. In our study, some patients underwent multiple RDD protocols based on clinical decisions, but the specific reasons for these decisions were not collected, as this was beyond the scope of the study.

In conclusion, the findings of our study add to the existing evidence supporting safety and effectiveness of RDD protocols for patients experiencing DHRs to ADs. However, our study has important limitations: first, the retrospective nature of the study and the limited sample size; second, the absence of specific diagnostic tools for DHRs diagnosis; third, the non-comparative design of the study that prevents comparison with a matched cohort of non-allergic patients treated with standard infusion protocols in terms of OS. Further research is needed to investigate the mechanisms underlying DHRs to specific ADs aiming to refine and tailor RDD protocols.

Conclusion

RDD represents the best option for the patients to continue a potentially effective treatment avoiding an early switch to successive lines that can be less effective. This procedure should be managed through a close multidisciplinary collaboration between clinicians and pharmacists. Efforts should be made to educate oncologists, pharmacists, allergists and nurses in order to universal use of rapid desensitization protocols and to uncover the cellular and molecular mechanisms underlying the temporary tolerization induced by rapid desensitization to antitumoral agents.