Off-label drug use in oncology: integrating molecular and genetic analysis in an academic center’s real-world approach

Abstract

Background

Recent advancements in cancer treatment have shifted the paradigm from chemotherapy to targeted therapies and immunotherapies, proving effective even in early-stage cancers. Regulatory agencies like the Food and Drug Administration, European Medicines Agency, and Italian Agency for Drugs have approved several targeted therapies for specific genetic mutations, significantly improving outcomes for several diseases. These advancements highlight the importance of identifying molecular targets, even in historically difficult-to-treat cancers.

Materials and methods

A retrospective observational study was conducted at the University of Campania “Luigi Vanvitelli” from January 2015 to January 2022, involving 83 adult patients with solid tumors. Patients received off-label drugs at no cost. The study analyzed treatment and pathology data, focusing on progression-free survival (PFS) and overall survival (OS), using Kaplan–Meier estimates and the log-rank test for statistical analysis.

Results

Patients averaged 63 years old, with a slight male majority. Off-label drugs were primarily requested for upper gastrointestinal cancers (42%) and colorectal carcinoma (21%). The most common molecular testing method was next-generation sequencing (NGS). Immunotherapy was the predominant treatment (67%), followed by targeted therapy and chemotherapy. The median PFS was 3.4 months [95% confidence interval (CI) 2.0-4.5 months], and the median OS was 10 months (95% CI 7.5-12.2 months). Early-line off-label treatment showed a trend toward longer PFS compared with later lines.

Conclusion

Early access to off-label drugs, particularly for patients with specific genetic alterations, can improve outcomes. The study underscores the importance of molecular tumor boards and multidisciplinary collaboration in selecting off-label treatments. Despite regulatory and ethical challenges, off-label drug use offers significant therapeutic opportunities, necessitating well-designed clinical trials and registries to better understand their efficacy and safety.

Highlights

• •Targeted therapy and immunotherapy are replacing chemotherapy in treating many tumors.
• •In Italy laws enable access to off-label drugs not yet approved for specific uses.
• •Off-label drugs improve PFS, especially in early stages.
• •NGS enables personalized therapies in oncology.

Introduction

In recent years, cancer treatment has undergone a significant paradigm shift. The substantial advent of new targeted therapies and immunotherapy has rapidly supplanted chemotherapy as the standard treatment for many tumors, demonstrating efficacy even in early-stage settings. Several targeted therapies have now received target-specific indications from regulatory agencies such as the Food and Drug Administration (FDA), the European Medicines Agency (EMA), and the Italian Agency for Drugs (AIFA).¹ For instance, in diseases such as non-small-cell lung cancer, the therapeutic algorithm has been transformed by the introduction of next-generation sequencing (NGS) able to identify targetable specific genetic alterations.² Similarly, in melanoma, the approval of therapies targeting BRAFV600E mutation, present in ∼50% of melanoma patients, has significantly impacted outcome of metastatic melanoma patients.³

The identification of molecular targets has become imperative even in diseases that have historically lacked therapeutic successes, such as gastric cancer, which has a poor prognosis. Today, the incorporation of targeted treatments such as anti-human epidermal growth factor receptor 2 (HER2) drugs or immunotherapy in cases with programmed death-ligand 1 (PD-L1) expression exceeding the registration cut-off, has resulted in improvement of survival curves compared with standard chemotherapy.^4,5

Advances in sequencing and analysis of cancer patients’ tissue or blood using different methods have allowed tumor characterization and subsequently the development of personalized therapies. However, despite the existence of approved targeted therapies for many cancer types, there are numerous other diseases with the same alterations that do not receive the same treatment indications. In fact, in this case it is not possible to administer drugs in clinical practice either because definitive data from clinical studies are not yet available or because the alterations present are very rare and do not allow studying the effectiveness of these drugs. The only strategy to obtain the use of these drugs in cancer patients is an off-label request.

In Italy, there are cases where access to a treatment is allowed before AIFA authorizes its commercialization. This can also apply to drugs that have already been authorized but are used for indications that are different from those originally approved in Italy (off-label use). The possibilities for early access to a drug include: Law 648/1996, compassionate use, AIFA National Fund (Law 326/2003 –‘5% Fund’), and non-repetitive use of advanced therapies. Under Law 648/1996 and the AIFA Fund, the cost of the drug is reimbursed by the National Health Service and AIFA, respectively. Compassionate use involves the direct and free supply of drug by its manufacturer. Non-repetitive use of advanced therapies involves the drug being prepared directly by a cell factory, with the requesting clinical center covering the related expenses. Finally, it is possible to access treatment with a drug that is already on the market but for an indication that is different than the one it was authorized for (Law 94/98, Article 3, Paragraph 2—formerly known as the Di Bella Law), even if there are other authorized therapeutic alternatives. In this case, therapy cost is charged to the patient or the health care provider. The choice among different strategies depends on the specific indication.⁶ In the latter case the single approval for the use of individual off-label drugs is evaluated through several steps: physician’s selection of the drug based on phase II or III studies, subsequent approval by the pharmacy, and then by the ethics committee, ensuring transparency in prescriptions.

In summary, the request for off-label drug use can be initiated, in the ways listed above, when there is no available therapeutic alternative; when the drug is innovative (not yet available in Italy but recently approved for marketing in other countries); when the medicine is not yet authorized in Italy but is currently undergoing clinical trials; and when the medicine is administered for a therapeutic indication not listed in the leaflet.⁷

At our institute, from 2015 to 2022, 114 off-label drug requests were made for the treatment of different tumor types. The aim of our analysis is to highlight the types of cancer for which off-label drugs were most frequently requested, the setting and line of therapy in which they were used and the specific gene alterations involved, and to evaluate the overall survival (OS) and progression-free survival (PFS) of the different treatments.

Materials and methods

This is a single-center, retrospective observational study that analyzed a population of 83 adult patients (>18 years of age) with solid tumors treated at Oncology Division of University of Campania “Luigi Vanvitelli” in Naples from January 2015 to January 2022. All patients received the drug completely free of charge due to the purchase made by our facility. The retrospective study protocol was approved by the institutional review board at the main study site (University of Campania “Luigi Vanvitelli”). The study was carried out in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. All patients signed a written informed consent and agreed with the research use of their anonymized data. A descriptive analysis of the study was carried out, based on demographic, treatment, and pathology characteristics extracted from the patient’s therapy and medical records. Of the 102 potential patients, only 82 were included in the analysis, as the remaining were excluded either because they never received the requested off-label treatment or were lost to follow-up.

Most of the drugs administered were off-label due to pathology (n = 72; 88%), another small part due to the type of combination (n = 10; 12%), while no patient received drugs that were off-label due to the setting.

The choice of treatment was made on the basis of the molecular characteristics highlighted on blood or tissue samples of the patients and after a careful review of the clinical studies available in the literature on populations with the same alterations. The techniques used to identify these alterations were NGS, immunohistochemistry (IHC), FISH, and germline analysis.

Median PFS and OS and 95% confidence intervals (CIs) were estimated with the Kaplan–Meier (KM) method, and survival curves were also stratified by number of lines of therapy and were compared with the log-rank test. A P value of <0.05 was considered statistically significant. All statistical analyses were carried out with Graphpad v. 9.5 (GraphPad Software, Inc, San Diego, CA).

Results

The average age of the study population was 63 years (range 20-88 years), with 43 (53%) male and 39 (47%) female patients (Table 1). No patient received more than one off-label drug. Patients with upper gastrointestinal (GI) cancer (esophageal or gastric) were the largest group to receive an off-label drug (n = 34; 43%), followed by colorectal carcinoma (n = 17; 20%). The fewest requests were made for patients with thymoma and prostate cancer (n = 1; 1%, respectively) (Figure 1). The majority of patients (90%) had already undergone at least one line of therapy, with only seven patients (8%) being treatment-naive for metastatic setting. The most commonly used method to study molecular alterations underlying off-label requests was NGS (28 patients, 33%), followed by IHC (27 patients, 32%) carried out on tissue samples. Approximately 17 patients (20%) were candidates for off-label treatment based on clinical trials for advanced lines in an unselected population, without evidence of a specific molecular target. Twenty-six patients (31%) could benefit from off-label immunotherapy treatments based on the identification of microsatellite instability for alterations in mismatch repair genes. The most frequent molecular alteration leading to off-label drug requests was the detection of PD-L1 overexpression in 27 patients (33%) (Table 1). The main reason for treatment discontinuation was disease progression (56 patients; 67%), followed by death (12 patients; 15%). Only four patients discontinued therapy due to toxicity (5%). Patients treated with off-label drugs achieved disease progression as the best response in 51% of cases (43 patients), stable disease in 33% of patients (27 patients), partial response in 14% (11 patients), and complete response in only 2% (2 patients) (Table 1). Median duration of response in patients who only had partial response (PR) or complete response (CR) was 38.6 months (95% CI 7.3-38.6 months). The median duration of response for patients who achieved PR or CR was 38.6 months. Among these 13 patients, 12 had gastric or colorectal cancers and were treated with immune checkpoint inhibitors. Nine received these inhibitors in early lines of therapy, while four received them in advanced lines (Supplementary Table).

Table 1.

Baseline characteristics and clinical outcomes across the entire cohort of patients

Median age, 63 years (20-88 years)
Sex, 47% female; 53% male	8 (10%)
Prior lines of therapy for metastatic disease	n (%)
0	7 (8)
1	28 (34)
2	22 (27)
≥3	26 (30)
Major molecular findings that led to therapy (by involved gene)	n (%)
MSI-H	26 (31)
PD-L1	27 (33)
Unselected	17 (20)
HER2 overexpression	10 (12)
BRCA	2 (2)
RET fusion	1 (1)
Molecular rationale for off-label therapy	n (%)
NGS	28 (33)
IHC	27 (32)
Unknown	17 (21)
FISH	10 (12)
Germline	2 (2)
Type of therapy	n (%)
Chemotherapy alone	12 (14)
Targeted therapy alone	5 (6)
Targeted Therapy with another agent	8 (9)
Immunotherapy alone	55 (67)
Immunotherapy with another agent	3 (3)
Best response to therapy	n (%)
Progressive disease	43 (51)
Stable disease	27 (33)
Partial response	11 (14)
Complete response	2 (2)
Reason for discontinuation	n (%)
Progressive disease	56 (67)
Death	12 (15)
Still on therapy	10 (12)
Toxicity	5 (6)
Median PFS in months (95% CI)
	3.4 (2.0-4.5)
Median OS in months (95% CI)
	10.0 (7.5-12.2)
Alive or deceased as of 30 June 2024	n (%)
Alive	19 (23)
Deceased	63 (76)

BRCA, breast cancer gene; CI, confidence interval; HER2, human epidermal growth factor receptor 2; IHC, immunohistochemistry; MSI-H, microsatellite instability-high; NGS, next-generation sequencing; OS, overall survival; PD-L1, programmed death-ligand 1; PFS, progression-free survival; RET, rearranged during transfection.

Figure 1.

Histopathologic diagnosis (subtypes of tumors).

The majority of patients (n = 55; 67%) received treatment with immunotherapy alone. Two patients (n = 2; 2%) underwent a combination treatment of immunotherapy and tyrosine kinase inhibitor (TKI). Of the remaining patients, 12 (15%) had undergone chemotherapy alone and only 3 (3%) had received a combination treatment based on chemotherapy. The other patients were equally distributed between those who had received targeted therapy [anti-HER2 or poly (ADP-ribose) polymerase inhibitors (PARPi)] alone (n = 5; 6%) or in combination with another agent (n = 8; 9%) (Table 1).

The median PFS of the overall population was 3.4 months (95% CI 2.0-4.5 months), whereas the median OS was 10 months (95% CI 7.5-12.2 months) (Figure 2A and B).

Figure 2.

Medianprogression-free survival (PFS). (A) Median progression-free survival (PFS) and (B) Median overall survival (OS) in the overall population.

CI, confidence interval.

A subgroup analysis comparing PFS in patients receiving off-label drugs in early lines (1-3) versus later lines (>4) demonstrated a trend toward benefit when the drug was administered in earlier lines. The median PFS was 4.16 months for early lines compared with 2.5 months for advanced lines (P value 0.08; hazard ratio 0.70, 95% CI 0.3-1.1), although this difference was not statistically significant (Figure 3).

Figure 3.

Progression-free survival (PFS) comparison between off-label therapy in early (1-3)-line and late (>4)-line cohorts. CI, confidence interval; HR, hazard ratio.

Discussion

In Italy, the number of requests for off-label drugs has been growing rapidly over the years for different reasons, including the increased number of ongoing clinical trials exploring new molecular targets for diseases.⁸ Most of these requests occur in advanced therapy settings, particularly when standard therapeutic options have been exhausted and enrollment in clinical trials is not possible due to the failure to meet inclusion/exclusion criteria or because no active trials are available for that specific setting or disease. Alternatively, these requests often come from attempts to treat patients with rare tumors or those for which there are currently no effective standard therapies, using experimental treatments showing promising results in early trial phases. Our center’s experience mirrors findings from other real-world analyses and systematic reviews of the literature.^9,10

Our study showed that treating patients in early lines with off-label drugs, based on personalized analyses, increases PFS compared with the same treatment used in later lines (Figure 3). This can be attributed to several factors. Firstly, patients start treatment in the initial lines with better performance status. Secondly, they accumulate less toxicity from previous therapies and have a lower likelihood of developing additional concurrent and/or resistant driver mutations.

Choosing targeted therapy in the first lines, aimed at genes driving tumor growth, not only provides a benefit in terms of PFS but may also allow the patient to undergo subsequent lines of treatment, with a potential benefit in PFS 2 as well. In our cohort, 29% (n = 24) of patients had access to a subsequent standard treatment line, with a potential benefit in PFS 2 and OS.

Most off-label drug requests are for immunotherapies, understandable given the recent development of these drug classes and the growing interest in various diseases.¹¹ Gastrointestinal tumors (stomach and colon) account for the most off-label drug requests. In this patient subgroup (total n = 67), 67% (n = 45) were treated with immunotherapy (n = 43; 64% alone, n = 1; 1% with TKI, n = 1; 1% with dual immunotherapy), and 12% (n = 8) were treated with targeted therapy (n = 4; 50% with anti-HER2 + TKI therapy, n = 2; 25% with dual anti-HER2 blockade, n = 1; 12% with PARPi, n = 1; 12% with anti-HER2 + chemotherapy). Only three patients were treated with chemotherapy (one with paclitaxel and two with trifluridine/tipiracil hydrochloride).

Subgroup analyses focusing on the duration of therapy for patients with colorectal and gastric cancer in early (1-3) and later (>4) lines demonstrated longer PFS in the gastric cancer subgroup, particularly in the early therapy lines (Figure 4).

Figure 4.

Swimmer plot for patients with colorectal and gastric cancer treated with immunotherapy in early (1-3) lines and late (>4) lines based onoff-labelrequest for microsatellite instability (MSI).

Most off-label requests were based on phase III studies, underscoring a significant issue: the lengthy approval times by regulatory agencies from the publication of positive clinical trial data to the drug’s commercialization and patient access to treatment.¹²

In recent years, numerous treatments requested for off-label use for our patients have received approval as standard therapies. For instance, pembrolizumab has been approved as a first-line treatment for metastatic microsatellite instability-high (MSI-H)–mismatch-repair-deficient (dMMR) colorectal carcinomas based on the Keynote 177 study.¹³ Similarly, it has been approved for second-line treatment in gastric carcinomas following data from the Keynote 158 study in MSI-H/dMMR patients. The results of this study also highlight a significant advancement in tumor-agnostic oncology, supporting the rationale for a biomarker-based approach in MSI-H/dMMR patients. The objective response rate of 34.3%, along with the sustained duration of responses observed, underscores the efficacy of pembrolizumab across a wide range of tumors with this molecular profile. The FDA’s tumor-agnostic approval of pembrolizumab for MSI-H/dMMR tumors represents a milestone, providing a valuable, personalized therapeutic option that transcends traditional site-based tumor classifications; unfortunately, AIFA has restricted the therapeutic indications of this drug to MSI-H or dMMR tumors in adults with metastatic or unresectable colorectal carcinoma, advanced or recurrent endometrial carcinoma, and unresectable or metastatic gastric, small intestine, or biliary tract carcinoma.¹⁴

Additionally, the combination of atezolizumab and nab-paclitaxel for metastatic triple-negative breast cancer has become a standard therapy, validated by the Impassion131 study.¹⁵ These examples underscore the significant benefits patients can receive from access to off-label drugs before their formal approval.

However, the efficacy of off-label treatments varies, with some patients experiencing substantial improvements while others do not benefit. In our study, ∼50% of patients exhibited disease progression as their best response.

Among the 13 patients who achieved PR or CR, as highlighted in the results section, 9 received these inhibitors in early lines of therapy, while four received them in advanced lines, demonstrating how early access to off-label drugs, which were later approved for these specific cancer types, improved patient prognosis. The remaining patient had oligodendroglioma and was treated with a combination of irinotecan and avastin, the latter of which is currently not approved.

Five patients (6%) discontinued treatment due to toxicity. Notably, three of these patients were on immune checkpoint inhibitors, one on TKI (pralseltinib), and one on anti-HER2 drugs, emphasizing the necessity for careful patient monitoring. Despite these challenges, the safety and tolerability of off-label drugs are generally consistent with those observed in approved uses.¹⁶

Off-label use can thus offer important therapeutic opportunities but requires a careful risk–benefit assessment. This practice poses regulatory and ethical challenges, especially in terms of informed consent and justification of therapeutic choices. For this reason, molecular tumor boards (MTBs) are increasingly active in high-volume centers like ours, involving multidisciplinary collaboration among molecular biologists, pathologists, geneticists, and oncologists to identify the best therapeutic option for each patient.^17,18

Several studies on samples collected at our institution have evaluated the clinical application of gene profiling tests like FoundationOne CDx, a hybrid capture-based NGS, analyzing tissue¹⁹ and liquid.^20,21 tumor samples. These studies confirm that comprehensive genomic profiling (CGP) can be useful in identifying patients with targetable alterations who might benefit from unexpected targeted treatments. These results reinforce the emerging idea that genomic alterations are predictive factors for treatment response in various cancer types. However, well-designed additional clinical studies are needed to provide more robust data on off-label drug use in oncology in terms of outcomes and toxicity. Registries and databases are necessary to monitor and evaluate long-term outcomes, especially compared with the current standard of care.²²

Based on this, several clinical trials have been designed, including the ROME Trial, a prospective phase II multi-basket study, evaluating the efficacy of tailored therapies versus the standard in patients with metastatic solid tumors, following the identification of molecular alterations via NGS (FoundationOne CDx and FoundationOne Liquid) and discussion in an MTB. Therefore, we await promising results in terms of feasibility and treatment outcomes.²³

Our study has several limitations. Firstly, it is an observational and retrospective study, with inherent limitations, reflecting the experience of a single institutional academic center, resulting in a small sample size, but representing our real-world practice of medication use, including patients who might be underrepresented in clinical trials. Additionally, the ability to subject patients to advanced molecular research techniques like NGS and based on these, to propose therapies not yet approved by regulatory agencies but fully supported economically by the belonging institution, does not represent clinical practice in oncology but was possible at our institution as a high-volume university center. Furthermore, the study included a wide range of cancers and drugs, some with limited sample sizes, which may hinder our ability to draw solid conclusions, but allowed us to highlight a beneficial trend and promote further research, and for many drugs, subsequent approval in clinical practice confirmed positively the risk–benefit assessment in off-label drug selection.

In conclusion our study highlights the therapeutic benefits of early access to off-label drugs in our academic center, particularly for patients with specific genetic alterations who lack standard treatment options. The experience at our institution reflects the broader trends in oncology, where MTBs and multidisciplinary collaboration are essential in identifying the best therapeutic options for patients. Our findings also emphasize the importance of careful patient selection, monitoring, and a robust risk–benefit assessment when considering off-label treatments. This practice, while posing regulatory and ethical challenges, offers significant therapeutic opportunities and underscores the need for CGP to tailor treatments effectively. Future directions should focus on well-designed clinical trials and the establishment of registries to provide more robust data on the outcomes and safety of off-label drug use. These efforts will contribute to a more refined and effective approach in oncology, ultimately improving patient outcomes.

Supplementary Data

Supplementary Table