Lessons learned from the maturation of a cancer drug: oxaliplatin in colorectal cancer

The addition of oxaliplatin to a fluoropyrimidine is reference treatment after colon cancer surgery in patients with stage III and in stage II if at least one high-risk criterium for recurrence is present [1–3]. Even if guidelines open for using a fluoropyrimidine alone, there is a tendency to add oxaliplatin whenever adjuvant chemotherapy is indicated, at least in stage III. Three randomized trials comparing adjuvant therapy for six months using a fluoropyrimidine with or without oxaliplatin showed that the combination improved disease-free survival (DFS) by approximately 20% (hazard ratios, HRs 0.80–0.82) [4–6]. Overall survival was also slightly improved (HRs 0.84–0.88). An average risk of recurrence of 30% after fluoropyrimidine treatment alone, typical for many stage III patients, means that six additional patients per 100 treated will not experience any recurrence if oxaliplatin was added. If the recurrence risk is higher, more treated patients are gained, but for many stage III patients, including most patients with stage II and high-risk criteria, the recurrence risk is less, and the gain from adding oxaliplatin is only a few percent. All treated patients are at risk of toxicity, including the oxaliplatin-induced chronic and often disabling neurotoxicity.

The oxaliplatin-induced peripheral neuropathy (OIPN) was considered along with the gains in recurrence risk and survival when different scientific and clinical organizations such as ASCO and NCCN in the US and ESMO in Europe and national groups recommended an oxaliplatin combination for all stage III colon cancers and for stage II cancers with high-risk factors. The extent of the adverse effects, i.e. the downside of adding oxaliplatin was known from the clinical trials where the treating physicians graded the neurotoxicity according to one of a few recommended scales. Typically, acute peripheral neurotoxicity of grade 2 or higher was recorded in 30-50% of the patients and although chronic OIPN was seen, it was not considered extensive [4–6]. In the MOSAIC trial, grade 2 and 3 chronic neuropathy were seen in 4.8 and 1.3%, respectively, after 12 months and in 3.4 and 0.7%, respectively, after 4 years [4].

How much chronic OIPN does up to 6 months of oxaliplatin-containing treatment result in?

The full scale of essentially subjective morbidity cannot be properly evaluated unless the patients report their experiences in validated instruments and how the symptoms affect daily living. The results of retrospective, sometimes cross-sectional studies have been reported through the years, describing clearly more problems than reported from the pivotal clinical trials [7,8]. In this issue of Acta Oncologica, Soveri et al. [9] present the results from a prospective study including 144 Finnish patients with CRC (85% stage III) receiving adjuvant CAPOX (n = 72) or a modified FOLFOX6 regimen (n = 72 matched controls). Ninety-two long-term survivors (median follow-up 4.2 years) responded to the general quality of life (QoL) instrument EORTC-QLQ-C30 and to one of a few frequently used instruments evaluating chemotherapy-induced peripheral neuropathy (CIPN), EORTC-CIPN20. The study is welcome since it reports the morbidity recorded by patients treated in real-life and how it influences their well-being. As expected, acute neurotoxicity was seen in almost all patients (any grade 94%, grade 2+ in 71%) with no difference between the two regimens. The problems were the same whether treatment was given during the two winter months of January or February or not.

Late OIPN, i.e. toxicity beyond 2 months after treatment stop was seen in 69% of the patients, being grade 2 in 24% and grade 3–4 in 9%. Late grade 2+ neuropathy could be seen in some patients who did not have any or minor (grade 0–1) acute toxicity, but the presence of acute toxicity significantly predicted long term neuropathy (grade 2+ late toxicity seen in 39% if grade 2+ acute toxicity vs 19% if grade 0–1). ECOG performance status 1 vs 0 significantly also increased the risk of late neurotoxicity (53% vs 27%). Of the 92 patients responding to CIPN20, 69% reported any grade of persistent neuropathy with numbness in toes and feet as most prevalent (present in 60%, severe in 21%). Patients with OIPN had slightly but significantly poorer physical and role functioning, but not lower global QoL. The results of this study confirm those of previous retrospective studies, namely that chronic OIPN appears as a much greater problem when reported by patients than when evaluated by physicians. “The deeper you dig, the more you will see”, has been repeatedly stated.

How much less is seen if 3 months of oxaliplatin is given?

Even if the magnitude of the chronic problems described in the pivotal adjuvant trials was limited, many clinicians still experienced this to be an outstanding issue, in line with the report by Soveri et al. [9], and at least five different randomized trials comparing a shorter oxaliplatin treatment (3 months) with the standard 6 months of treatment were initiated to show that 3 months did not give inferior results but less OIPN. The rapid inclusion of together well over 10,000 patients [10] clearly tells that both doctors and the patients they informed about the trials considered this research question well worth exploring and participating in. Collectively, the studies also showed that DFS was not significantly inferior after 3 months of treatment for many groups of patients and the OIPN was as expected significantly less.

In one of the trials comparing 3 vs 6 months of adjuvant FOLFOX or CAPOX, the SCOT trial [11], another patient-reported outcome measure than CIPN20, used by Soveri et al. [9], the FACT/GOG NTx-4 scale was used. The scores in the 6-month arm (approximately 60% of patients actually received 6 months oxaliplatin) were about twice as high as in the 3-month arm (approximately 85% received 3 months), showing that longer treatment gives more acute and late neurotoxicity. Global QoL using EORTC QLQ-C30 was worse in the 6-month arm when treatment was ongoing, i.e. between months 3 and 6, but the difference had disappeared at 9 and 12 months. The same development was seen using the EQ-5D self-rated VAS scale at 9 months and beyond up to 6 years. Patients answering the two higher scores (quite a bit or very much numbness, tingling or discomfort in hands or feet) on FACT/GOG NTx-4 had worse global QoL at 12 months than patients scoring less (QLQ-C30 was not administrated beyond 12 months). Using the EQ-5D self-rated VAS scale or the EQ-5D 3L Health index scale, both reflecting aspects of patients’ well-being, patients scoring worse on FACT/GOG NTx-4, i.e. having more problems, scored significantly worse at 1, 3, and 5 years. Thus, severe OIPN affects important aspects of the QoL of the patients, adding further support to a previous systematic overview of associations between CIPN and QoL [12].

Will the clinical problem with chronic neuropathy in CRC be less in the future?

Oxaliplatin is used also in other malignancies, but its largest use is in CRC; CRC is the third most common cancer worldwide. In CRC, it is used both in primary disease, pre- or postoperatively and in metastatic disease (mCRC). In mCRC, long survival depends upon exposure to all available cytotoxic drugs, a fluoropyrimidine, irinotecan, and oxaliplatin and to one or all available biologic agents depending upon RAS and BRAF mutation status [13]. Whether oxaliplatin is used in first- or second line has some bearing on how long patients are at risk suffering from the chronic neuropathy. Oxaliplatin is, however, the primary drug of choice in all instances of neo-adjuvant therapy and in most instances when conversion prior to metastases surgery is required. Many of these patients may receive up to six months of oxaliplatin exposure and a substantial fraction will become long-term survivors. In mCRC, its use will likely continue at the same level as today, although the few patients with mismatch repair deficient tumors may perhaps get less oxaliplatin if and when antibodies targeting PD-1/PDL-1 are approved within EU for this indication.

In the adjuvant setting, two aspects will decrease the use of oxaliplatin. One, and the reason most clinicians likely first think of, is, as described above, the recently reported trials revealing that 3 months of oxaliplatin-containing chemotherapy is for many patients not inferior to 6 months of therapy and that the shorter treatment results in less OIPN [10,11,14]. Since the release of trial data in the spring 2018 and the full publications later in 2018, the duration of adjuvant oxaliplatin-containing treatment has likely been shortened world-wide for many, if not all patients. Controversies still exist about whether the two schedules, CAPOX or FOLFOX should be handled differently and for what subgroups, if any, the reference treatment still is 6 months. According to a sub-study of the SCOT trial [11], the largest of the 5 trials in the IDEA consortium [10], 6 months of therapy is never cost-effective [15]. It is beyond the purpose of this editorial to discuss pros and cons of 3 or 6 months of adjuvant therapy, but the clinical problem of long-lasting OIPN will diminish as a result of the practice changing 3 vs 6 months adjuvant trials.

The second reason that will reasonably diminish the use of adjuvant oxaliplatin is less need for adjuvant therapy thanks to improvements in staging and surgery for colon cancer. The recurrence risk has decreased and, thus, so has the number of patients who will benefit from the addition of oxaliplatin sufficiently to motivate its use [16]. A few studies have recently substantiated that these risks are less than in the past [17–19]. Whether this knowledge will change routines is not known [20], but the Swedish guidelines valid from spring 2019 do not recommend adding oxaliplatin routinely to patients below 70 years of age as often as was the case in the guidelines from 2016 [21]. In patients over 70 years, oxaliplatin is not recommended in the adjuvant setting.

The third reason for less OIPN might emerge if effective neuroprotectors that, at the same time, do not prevent the anti-tumor effects of oxaliplatin, can be developed. Through the years, many attempts have been done, such as calmangafodipir [22], but none has so far been approved for use [23].

Lessons learned from the story of oxaliplatin

The trajectory of oxaliplatin illustrates some general lessons to be learned and kept in mind when new cancer drugs are approved and made available in routine healthcare. As for oxaliplatin, and repeatedly seen for other more recently approved cancer drugs, the initial pivotal trials forming the basis for medical agencies assessment focus on efficacy in highly selected patient populations in trials with high internal but often questionable external validity. While benefit is emphasized, risk is not in focus and mostly reported ‘objectively’ by physician assessments guided by toxicity scales and laboratory blood sampling. The conclusion regarding risk at this stage is almost always that the new treatment is “well tolerated” although the patients’ opinion on this issue is mostly unknown.

When the drug is then made available for broader use outside of the highly experimental settings of the pharma-sponsored randomized trials, clinical experience on efficacy, as well as adverse effects, slowly accumulate and is scientifically reported, often with a broader clinical utility perspective, including the reporting of patients’ views on adverse effects and QoL. This “post-approval” research is important and can more truly reflect the benefit–risk balance for a new treatment, often showing less benefit and more risk, as illustrated from the oxaliplatin trajectory.

Thus, the data from the initial pivotal trials on cancer drugs should be viewed with caution and regarded as preliminary and not the final truth regarding their clinical utility. By time and more research, importantly also involving assessment of patients’ views on adverse effects and QoL, the place of a drug in our armamentarium matures and allows for its judicious use to the benefit of our patients.

Disclosure statement

BG was the principal investigator of the neuroprotective PLIANT trial (22) supported by PledPharma AB, Sweden. No potential conflict of interest was reported by PN.